FORM 2
THE PATENTS ACT, 1970
(39 OF 1970)
5 &
THE PATENT RULES, 2003
COMPLETE SPECIFICATION
(See section 10 and rule 13)
10
“METHOD AND SYSTEM FOR AUTOMATICALLY GENERATING ONE OR MORE
SCOREBOARD TEMPLATES”
15
Jio Platforms Ltd., of address Office-101, Saffron, Nr. Centre Point, Panchwati 5 Rasta,
20 Ambawadi, Ahmedabad-380006, Gujarat, India.
25
The following specification particularly describes the invention and the manner in which
it is to be performed:
2
METHOD AND SYSTEM FOR AUTOMATICALLY GENERATING ONE OR MORE
SCOREBOARD TEMPLATES
PRIORITY APPLICATION
This application claims the benefit of Indian Patent Application no. 202321024956
filed on March 31, 2023 titled “METHOD AND SYSTEM FOR DELIVERING
CONTEXTUAL CONTENTS TO VIEWERS OF A LIVE SPORTS EVENT”. The entire
contents of the aforementioned applications are incorporated herein by
reference.
TECHNICAL FIELD
Embodiments of the present disclosure generally relate to processing a live
multimedia stream. More particularly, embodiments of the present disclosure
relate to automatic content recognition from the live multimedia stream to
generate scoreboard templates.
BACKGROUND
The following description of the related art is intended to provide background
information pertaining to the field of the disclosure. This section may include
certain aspects of the art that may be related to various features of the present
disclosure. However, it should be appreciated that this section is used only to
enhance the understanding of the reader with respect to the present disclosure,
and not as admissions of the prior art.
Over the years, there has been a steady rise in the viewership of sports events
across the globe. Thanks to the advent of technology and the widespread
3
availability of internet services, people from all corners of the world can now
access live sports events at their convenience. Additionally, the globalisation of
sports has also contributed to the increase in viewership, as fans now have a wider
range of sports to follow and teams to support. The growing interest in sports has
also led to the emergence of new sports leagues, which have gained massive
followings in a short period of time. With more people tuning in to watch sports
events, the sports industry is expected to continue to grow, with even more
opportunities for athletes and sports enthusiasts alike. Further, sports
personalities and athletes have become some of the most influential figures in the
world today, with their achievements and personalities reaching a global
audience. With the increasing popularity of sports, there has been a significant
increase in brand sponsorship contracts for sports personalities and athletes.
Companies are now investing heavily in sports marketing, with sports figures
serving as the faces of their brands. This trend is not limited to established stars,
as upcoming athletes are also attracting brand endorsements, given their
potential for success and appeal to younger audiences. As a result, sports
personalities and athletes have become important influencers. This trend is
expected to continue as sports continue to grow in popularity and brands
recognise the power of aligning themselves with successful athletes and sports
personalities.
Further, with increase in viewership, advertising during sports telecasts has
become increasingly important for brands to increase their revenue. Sports
telecasts draw a massive audience, providing a platform for brands to showcase
their products and services to a global audience. In addition, the logos of brands
displayed during sports telecasts can have a significant impact on brand
recognition and recall, as they are viewed by millions of people worldwide. With
the advent of technology, brands can now target specific demographics and
regions, increasing their reach and impact. The revenue generated from
4
advertising during sports telecasts is significant, and it has become a critical
revenue stream for deliverers and sports teams alike. With the increasing
popularity of sports, the value of advertising during sports telecasts is expected to
continue to grow, providing brands with an opportunity to increase their revenue
and build their brand equity.
Further, contextual contents such as advertisements (hereinafter also referred to
as ad/ads) has proven to be more effective in terms of viewer engagement and
revenue compared to non-contextual advertisement. By providing viewers of a
media stream of an event with ads that are relevant to their interests and needs,
contextual advertising creates a more personalised experience, increasing the
likelihood of viewer engagement. Studies have shown that contextual ads are
more likely to be clicked, resulting in higher click-through rates and, ultimately,
higher conversion rates. Additionally, contextual advertising is often less intrusive,
making it more likely to be accepted and viewed positively by viewers. This leads
to a better viewer experience, which can increase the chances of repeat visits and
customer loyalty. Also, in terms of revenue, contextual advertising can lead to a
higher return on investment for brands, as it enables them to target their audience
with greater precision, resulting in more effective ad campaigns and increased
revenue. Overall, the relevance of contextual advertising has been proven by its
ability to drive engagement and revenue, making it a valuable tool for brands
looking to reach their target audience effectively. Additionally, one of the most
important purposes of contextual advertising is to allow brands to reach their
target audience with greater precision, leading to more effective ad campaigns
and increased brand awareness. With the rise of digital marketing, contextual
advertising has become an increasingly important tool for brands to connect with
their audience and drive revenue. Further, in the currently known method and
system, contextual advertisements targeted to a specific audience are based only
on the viewer's search or browsing history, location, and other relevant data
5
collected from the viewer's device. Further, the currently known solutions aim to
present ads that are relevant to the viewers interests and needs based on their
demographic regions to make them more likely to engage with the advertisement
and fails to take the current emotion of the viewer into account while delivering
contextual advertisement, which is an important part of any sports event. The
currently known methods and systems for contextual advertising aim to present
ads based on their usefulness and value to the viewer based on their search history
or regional background, with the objective of higher engagement rates and,
ultimately, a better return on investment for advertisers; however, none of the
current solutions appreciate the emotion of the viewer or context of a particular
sports event and its impact on a viewer of a region that is associated with the live
streaming of the particular sports event.
Further, an essential part of a sports telecast is a superimposed scoreboard, which
is a common and essential feature in sports telecasts and provides viewers with
real-time updates on the score, time remaining, and other pertinent game
information. The scoreboard is typically displayed as a graphic overlay on the
screen and updated by a dedicated team that monitors the game and inputs the
data. However, advances in technology have allowed for the automatic detection
of game events through the analysis of the scoreboard using techniques such as
optical character recognition. This technology enables deliverers to quickly and
accurately identify when a goal is scored, a penalty is called, or any other
important event that impacts the game. By using these techniques, sports
telecasts can provide viewers with a more immersive and engaging experience
while also enhancing the accuracy and efficiency of the deliverer. Further, event
detection from scoreboards during sports telecasts is crucial for deliverers as it
allows them to deliver targeted advertisements to viewers based on the current
state of the game. For example, if a team scores a goal, advertisers can display ads
during the celebration. This targeted approach to advertising has a significant
6
impact on the enhancement of brand visibility and ultimately on the revenue
generated from such advertisements. By providing viewers with relevant ads,
deliverers can increase engagement, brand recognition, and ultimately revenue
for their advertisers. Furthermore, event detection from scoreboards enables
deliverers to capitalise on key moments during a game, such as a player's recordbreaking performance, thereby increasing the visibility and value of the associated
brand. In conclusion, the importance of event detection from scoreboards cannot
be overstated, as it has a direct impact on the success of sports telecast and the
profitability of advertisers.
Although current solutions of event detection from scoreboards during sports
telecasts do allow for targeted advertisements, there is a flaw in the system as
many ads displayed are unrelated to the actual game being played. The existing
approaches fail to incorporate the context of the event, and viewers may find the
ads disruptive and irrelevant to the current context of the game. As a result, there
is a growing need for a more contextual approach to sports advertising. For
example, if a player is injured during a game, it may be more appropriate to display
ads for sports medicine or medical services. By incorporating contextual
information into the advertising process, deliverers can provide more relevant and
useful ads to viewers, enhancing the overall viewing experience and increasing the
effectiveness of advertisements. As the sports delivering industry continues to
evolve, it is essential that advertisers and deliverers of the sports event work
together to improve the relevance and context of advertising during sports
telecasts.
In the known solutions for event detection from scoreboards during sports
telecasts has limitations in terms of detecting and incorporating contextual
information for targeted advertising. The current known solutions fail to efficiently
analyse the media feeds of the sports events, detect the faces of athletes and logos
7
of brands etc., and then deliver relevant contextual advertisements based on this
information. For example, if a popular athlete is playing with a logo on their attire
of brand X, in the current known solution, the deliverers fail to detect the athlete
and display ads for the brand they endorse or deliver ads of the athlete even if
they are of a competitor band, i.e., brand Y. Further, the current known solutions
also fail to detect the region for delivering the contextual ad, such as an ad for
brand Y of the athlete, in the region or nation where the athlete is admired. On
the contrary, the current known solution may deliver an ad even if it is not related
to the context of the game, athlete, or brand visible in the media stream.
Having said that, and in view of the aforementioned shortcomings, enigmas, and
drawbacks inherent in the existing techniques, there exist unembellished needs
for an improved and enhanced mechanism to facilitate contextual and eventspecific advertising to the viewer. Further, there is a requirement to propose a
system and method for facilitating contextual content that not only considers the
context of the event and related attributes of the live media stream from the event
but also provides selective content on the basis of the occurrence of the event.
Furthermore, there is a requirement to propose a system for facilitating the
delivery of relevant contextual contents to relevant viewers of the live event based
on the detection of an information from a live media stream of a sport telecast
and/or the detection of an occurrence of an event in the live media stream of the
sport telecast.
OBJECTS OF THE INVENTION
Some of the objects of the present disclosure, which at least one embodiment
disclosed herein satisfies are listed herein below.
8
It is an object of the present disclosure to provide a system and a method that
facilitates for delivering one or more contextual contents to a set of viewers of a
live sports event based on at least one of a player face, a brand logo, and a real
time information of the live sports event.
It is an object of the present disclosure to provide a system and a method that
identifies in real time player faces, brand logos, and a real time information of the
live sports event.
It is an object of the present disclosure to provide a system and a method that
facilitates identifying a real time information of a critical event of the live sports
event and a context associated with the critical event.
It is also an object of the present disclosure to provide a system and a method that
facilitate verification of one or more contextual contents that are delivered to a
set of viewers of a live sports event.
It is also an object of the present disclosure to provide a system and a method that
encompasses use of technologies such as deep learning technologies, computer
vision and/or other artificial intelligence technologies for delivering one or more
contextual contents to a set of viewers.
Also, an object of the present disclosure is to provide contents such as
advertisement/banner-advertisement outside a live media stream region without
modifying the live media stream. For example, at a bottom of a screen displaying
the live media stream.
Another object of the present disclosure is to prioritize event notifications to show
contents such as advertisements in a video stream. For example, a batsman about
9
to score a century may be prioritized over other batsman hitting a boundary to
show an advertisement.
Another object of the present disclosure is to provide in a live stream, team/player
endorsed brand contents based on the teams/players detected from a scoreboard
of the live event, wherein the live event may be a live sports match like cricket,
hockey, football match etc.
Yet another object of the present disclosure is to provide a resolution agnostic
scoreboard detection, such that irrespective of a resolution of a live media stream,
a scoreboard information from the live media stream is efficiently detected.
SUMMARY
This section is provided to introduce certain aspects of the present disclosure in a
simplified form that are further described below in the detailed description. This
summary is not intended to identify the key features or the scope of the claimed
subject matter.
An aspect of the present disclosure may relate a method for delivering one or
more contextual contents to a set of viewers of a live sports event. The method
comprises receiving, at a processing unit, a real time media stream of the live
sports event, wherein the real-time media stream comprises one or more frames.
The method further comprises identifying in real time, by an identification unit, at
least one of one or more player faces, one or more brand logos, and a real time
information of the live sports event. Further, the method encompasses analysing,
by the processing unit, at least one of the one or more player faces, the one or
more brand logos, and the real time information of the live sports event. Further,
the method comprises detecting, by a detection unit, at least one of a content
display priority of the one or more player faces, a content display priority of the
10
one or more brand logos, and a content display priority of the real time
information of the live sports event, based on the analysis. Further, the method
comprises detecting, by the detection unit, one or more content-triggers, based
on at least one of the content display priority of the one or more player faces, the
content display priority of the one or more brand logos, and the content display
priority of the real time information of the live sports event. Further, the method
encompasses retrieving, by the detection unit from a storage unit, the one or more
contextual contents based on the one or more content-triggers, wherein the one
or more contextual contents are based on at least one of the one or more player
faces, the one or more brand logos and the real time information of the live sports
event. Thereafter, the method encompasses delivering, by a content-delivery unit,
the one or more contextual contents to the set of viewers of the live sports event.
Another aspect of the present disclosure may relate to a system for delivering one
or more contextual contents to a set of viewers of a live sports event. The system
comprises a processing unit, configured to receive a real time media stream of the
live sports event, wherein the real-time media stream comprises one or more
frames. Further, the system comprises an identification unit connected at least to
the processing unit, wherein the identification unit is configured to identify in real
time, at least one of one or more player faces, one or more brand logos, and a real
time information of the live sports event. Furthermore, the processing unit as
disclosed is further configured to analyse at least one of the one or more player
faces, the one or more brand logos, and the real time information of the live sports
event. Further, the system comprises a detection unit connected at least to the
processing unit and the identification unit, wherein the detection unit is
configured to detect at least one of a content display priority of the one or more
player faces, a content display priority of the one or more brand logos, and a
content display priority of the real time information of the live sports event, based
on the analysis. Further, the detection unit is configured to detect one or more
11
content-triggers, based on at least one of the content display priority of the one
or more player faces, the content display priority of the one or more brand logos,
and the content display priority of the real time information of the live sports
event. Furthermore, the detection unit as disclosed by the system is further
configured to retrieve from a storage unit, the one or more contextual contents
based on the one or more content-triggers, wherein the one or more contextual
contents are based on at least one of the one or more player faces, the one or
more brand logos and the real time information of the live sports event. Further,
the system encompasses a content-delivery unit connected at least to the
processing unit the identification unit and the detection unit, wherein the contentdelivery unit is configured to deliver the one or more contextual contents to the
set of viewers of the live sports event.
Another aspect of the present disclosure may relate to a user device for delivering
one or more contextual contents to a set of viewers of a live sports event, the user
device comprises a system. The system comprises a processing unit, configured to
receive a real time media stream of the live sports event, wherein the real-time
media stream comprises one or more frames. The system further comprises an
identification unit connected at least to the processing unit, wherein the
identification unit is configured to identify in real time, at least one of one or more
player faces, one or more brand logos, and a real time information of the live
sports event. The processing unit is further configured to analyze at least one of
the one or more player faces, the one or more brand logos, and the real time
information of the live sports event. Further the system comprises a detection unit
connected at least to the processing unit and the identification unit, wherein the
detection unit is configured to: detect at least one of a content display priority of
the one or more player faces, a content display priority of the one or more brand
logos, and a content display priority of the real time information of the live sports
event, based on the analysis; detect one or more content-triggers, based on at
12
least one of the content display priority of the one or more player faces, the
content display priority of the one or more brand logos, and the content display
priority of the real time information of the live sports event; and retrieve from a
storage unit, the one or more contextual contents based on the one or more
content-triggers, wherein the one or more contextual contents are based on at
least one of the one or more player faces, the one or more brand logos and the
real time information of the live sports event. The system further comprises a
content-delivery unit connected at least to the processing unit, the identification
unit and the detection unit, wherein the content-delivery unit is configured to
deliver the one or more contextual contents to the set of viewers of the live sports
event.
Further, an aspect of the present disclosure relates to a non-transitory computer
readable storage medium storing instructions for delivering one or more
contextual contents to a set of viewers of a live sports event, the storage medium
comprising executable code which, when executed by a processor, causes the
processor to: receive a real time media stream of the live sports event, wherein
the real-time media stream comprises one or more frames; enable an
identification unit to identify in real time, at least one of one or more player faces,
one or more brand logos, and a real time information of the live sports event;
analyse at least one of the one or more player faces, the one or more brand logos,
and the real time information of the live sports event; enable a detection unit to:
1) detect at least one of a content display priority of the one or more player faces,
a content display priority of the one or more brand logos, and a content display
priority of the real time information of the live sports event, based on the analysis,
2) detect one or more content-triggers, based on at least one of the content
display priority of the one or more player faces, the content display priority of the
one or more brand logos, and the content display priority of the real time
information of the live sports event, and 3) retrieve from a storage unit, the one
13
or more contextual contents based on the one or more content-triggers, wherein
the one or more contextual contents are based on at least one of the one or more
player faces, the one or more brand logos and the real time information of the live
sports event; and enable a content-delivery unit to deliver the one or more
contextual contents to the set of viewers of the live sports event.
Another aspect of the present disclosure relates to a method for providing one or
more notifications to a content-delivery unit. The method comprises receiving, at
a processing unit, a real time media stream of the live sports event, wherein the
real-time media stream comprises one or more frames. Further, the method
comprises receiving, by the processing unit, scoreboard proportions. Further, the
method encompasses cropping, by the processing unit, a scoreboard area from
the one or more frames based on the scoreboard proportions. Further, the
method comprises detecting, by the processing unit using one or more text
detection techniques, one or more region of interests (ROIs) from the scoreboard
area. The method further comprises extracting, by the processing unit using one
or more text recognition techniques, one or more texts from the one or more ROIs.
The method further encompasses mapping, by the processing unit, the one or
more texts with one or more pre-defined texts. The method further comprises
determining, by the processing unit, a scoreboard-based information based on the
mapping. The method further encompasses detecting, by the processing unit, at
least one of one or more players and one or more scoreboard-based events based
at least on the scoreboard-based information. Thereafter the method
encompasses providing, by the processing unit, the one or more notifications to
the content- delivery unit based at least on the at least one of the detected one or
more players and the detected one or more scoreboard-based events.
Another aspect of the present disclosure may relate to a system for providing one
or more notifications to a content-delivery unit. The system further comprises a
processing unit, configured to receive a real time media stream of the live sports
14
event, wherein the real-time media stream comprises one or more frames. The
processing unit is further configured to: receive scoreboard proportions, and crop
a scoreboard area from the one or more frames based on the scoreboard
proportions. Further, the processing unit of the system is configured to: detect
using one or more text detection techniques, one or more region of interests
(ROIs) from the scoreboard area, and extract using one or more text recognition
techniques, one or more texts from the one or more ROIs. Further, the processing
unit of the system is configured to: map the one or more texts with one or more
pre-defined texts, determine a scoreboard-based information based on the
mapping, and detect at least one of one or more players and one or more
scoreboard-based events based at least on the scoreboard-based information.
Thereafter, the processing unit of the system is configured to provide the one or
more notifications to the content delivery unit based at least on the at least one
of the detected one or more players and the detected one or more scoreboardbased events.
Another aspect of the present disclosure relates to a user device for providing one
or more notifications to a content delivery unit, the user device comprising: a
system, wherein the system comprises: a processing unit, configured to: receive a
real time media stream of the live sports event, wherein the real-time media
stream comprises one or more frames; receive scoreboard proportions; crop a
scoreboard area from the one or more frames based on the scoreboard
proportions; detect using one or more text detection techniques, one or more
region of interests (ROIs) from the scoreboard area; extract using one or more text
recognition techniques, one or more texts from the one or more ROIs; map the
one or more texts with one or more pre-defined texts; determine a scoreboardbased information based on the mapping; detect at least one of one or more
players and one or more scoreboard-based events based at least on the
scoreboard-based information; and provide the one or more notifications to the
15
content delivery unit [208] based at least on the at least one of the detected one
or more players and the detected one or more scoreboard-based events.
Another aspect of the present disclosure relates to a non-transitory computer
readable storage medium storing instructions for providing one or more
notificationsto a content delivery unit, the storage medium comprising executable
code which, when executed by a processor, causes the processor to: receive a real
time media stream of the live sports event, wherein the real-time media stream
comprises one or more frames; receive scoreboard proportions; crop a scoreboard
area from the one or more frames based on the scoreboard proportions; detect
using one or more text detection techniques, one or more region of interests
(ROIs) from the scoreboard area; extract using one or more text recognition
techniques, one or more texts from the one or more ROIs; map the one or more
texts with one or more pre-defined texts; determine a scoreboard-based
information based on the mapping; detect at least one of one or more players and
one or more scoreboard-based events based at least on the scoreboard-based
information; and provide the one or more notifications to the content delivery unit
based at least on the at least one of the detected one or more players and the
detected one or more scoreboard-based events.
Another aspect of the present disclosure may relate to a method for automatically
generating one or more scoreboard templates for a live sports event. The method
further comprises receiving, at a processing unit, a media stream of the live sports
event, wherein the media stream comprises one or more frames. The method
further encompasses cropping, by the processing unit, a pre-defined portion of
the one or more frames. Further the method comprises detecting, by a detection
unit using one or more text detection techniques, one or more region of interests
(ROIs) in the cropped pre-defined portion. The method further comprises
detecting, by the detection unit using one or more text recognition techniques,
one or more texts and a position associated with each of the one or more texts
16
from the one or more ROIs. The method further comprises detecting, by the
detection unit, a set of template coordinates based on the one or more texts, the
position associated with each of the one or more texts, one or more pre-defined
texts and a target template mask value associated with the one or more predefined texts. Thereafter, the method encompasses automatically generating, by
the processing unit, the one or more scoreboard templates based at least on the
set of template coordinates wherein each of the one or more scoreboard
templates comprises at least one of a position of one or more components of a
scoreboard of the live sports event and an order of the one or more components.
Another aspect of the present disclosure may relate to a system for automatically
generating one or more scoreboard templates for a live sports event. The system
comprises a processing unit, configured to receive a media stream of the live
sports event, wherein the media stream comprises one or more frames. The
processing unit is further configured to crop a pre-defined portion of the one or
more frames. Further, the system comprises a detection unit, configured to detect
using one or more text detection techniques, one or more region of interests
(ROIs) in the cropped pre-defined portion. The detection unit is further configured
to detect using one or more text recognition techniques, one or more texts and a
position associated with each of the one or more texts from the one or more ROIs.
Further, the detection unit is configured to detect a set of template coordinates
based on the one or more texts, the position associated with each of the one or
more texts, one or more pre-defined texts and a target template mask value
associated with the one or more pre-defined texts, wherein the processing unit is
further configured to automatically generate, the one or more scoreboard
templates based at least on the set of template coordinates wherein each of the
one or more scoreboard templates comprises at least one of a position of one or
more components of a scoreboard of the live sports event and an order of the one
or more components.
17
Another aspect of the present disclosure may relate to a user device for
automatically generating one or more scoreboard templates for a live sports
event. The user device further comprises a system, wherein the system comprises
a processing unit, configured to receive a media stream of the live sports event,
wherein the media stream comprises one or more frames. The processing unit is
further configured to crop a pre-defined portion of the one or more frames.
Further, the system comprises a detection unit, configured to detect using one or
more text detection techniques, one or more region of interests (ROIs) in the
cropped pre-defined portion. Further, the detection unit is configured to detect
using one or more text recognition techniques, one or more texts and a position
associated with each of the one or more texts from the one or more ROIs. The
detection unit is further configured to detect a set of template coordinates based
on the one or more texts, the position associated with each of the one or more
texts, one or more pre-defined texts and a target template mask value associated
with the one or more pre-defined texts. The processing unit is further configured
to automatically generate, the one or more scoreboard templates based at least
on the set of template coordinates wherein each of the one or more scoreboard
templates comprises at least one of a position of one or more components of a
scoreboard of the live sports event and an order of the one or more components.
Yet another aspect of the present disclosure may relate to a non-transitory
computer readable storage medium storing instructions for automatically
generating one or more scoreboard templates for a live sports event, the storage
medium comprising executable code which, when executed by a processor, causes
the processor to: receive a media stream of the live sports event, wherein the
media stream comprises one or more frames; crop a pre-defined portion of the
one or more frames; enable a detection unit to: 1) detect using one or more text
detection techniques, one or more region of interests (ROIs) in the cropped predefined portion, 2) detect using one or more text recognition techniques, one or
more texts and a position associated with each of the one or more texts from the
18
one or more ROIs, and 3) detect a set of template coordinates based on the one
or more texts, the position associated with each of the one or more texts, one or
more pre-defined texts and a target template mask value associated with the one
or more pre-defined texts; and automatically generate, the one or more
scoreboard templates based at least on the set of template coordinates wherein
each of the one or more scoreboard templates comprises at least one of a position
of one or more components of a scoreboard of the live sports event and an order
of the one or more components.
DESCRIPTION OF THE DRAWINGS
The accompanying drawings, which are incorporated herein, and constitute a part
of this disclosure, illustrate exemplary embodiments of the disclosed methods and
systems in which like reference numerals refer to the same parts throughout the
different drawings. Components in the drawings are not necessarily to scale,
emphasis instead being placed upon clearly illustrating the principles of the
present disclosure. Also, the embodiments shown in the figures are not to be
construed as limiting the disclosure, but the possible variants of the method and
system according to the disclosure are illustrated herein to highlight the
advantages of the disclosure. It will be appreciated by those skilled in the art that
disclosure of such drawings includes disclosure of electrical components or
circuitry commonly used to implement such components.
FIG.1 illustrates an exemplary block diagram depicting an exemplary network
architecture diagram [100], in accordance with an embodiment of the present
disclosure.
19
FIG.2 illustrates an exemplary block diagram of a system [200] for delivering one
or more contextual contents to a set of viewers of a live sports event is shown, in
accordance with an embodiment of the present disclosure.
FIG.3 illustrates an exemplary method flow diagram [300], for delivering one or
more contextual contents to a set of viewers of a live sports event, in accordance
with an embodiment of the present disclosure.
FIG.4 illustrates an exemplary method flow diagram [400], for providing one or
more notifications to a content-broadcasting unit [208], in accordance with an
embodiment of the present disclosure.
FIG.5 illustrates an exemplary method flow diagram [500], for automatically
generating one or more scoreboard templates for a live sports event, in
accordance with an embodiment of the present disclosure.
DETAILED DESCRIPTION
In the following description, for the purposes of explanation, various specific
details are set forth in order to provide a thorough understanding of embodiments
of the present disclosure. It will be apparent, however, that embodiments of the
present disclosure may be practiced without these specific details. Several
features described hereafter may each be used independently of one another or
with any combination of other features. An individual feature may not address any
of the problems discussed above or might address only some of the problems
discussed above.
20
The ensuing description provides exemplary embodiments only, and is not
intended to limit the scope, applicability, or configuration of the disclosure.
Rather, the ensuing description of the exemplary embodiments will provide those
skilled in the art with an enabling description for implementing an exemplary
embodiment. It should be understood that various changes may be made in the
function and arrangement of elements without departing from the spirit and
scope of the disclosure as set forth.
Specific details are given in the following description to provide a thorough
understanding of the embodiments. However, it will be understood by one of
ordinary skill in the art that the embodiments may be practiced without these
specific details. For example, circuits, systems, processes, and other components
may be shown as components in block diagram form in order not to obscure the
embodiments in unnecessary detail.
Also, it is noted that individual embodiments may be described as a process which
is depicted as a flowchart, a flow diagram, a data flow diagram, a structure
diagram, or a block diagram. Although a flowchart may describe the operations as
a sequential process, many of the operations may be performed in parallel or
concurrently. In addition, the order of the operations may be re-arranged. A
process is terminated when its operations are completed but could have
additional steps not included in a figure.
The word “exemplary” and/or “demonstrative” is used herein to mean serving as
an example, instance, or illustration. For the avoidance of doubt, the subject
matter disclosed herein is not limited by such examples. In addition, any aspect or
design described herein as “exemplary” and/or “demonstrative” is not necessarily
to be construed as preferred or advantageous over other aspects or designs, nor
is it meant to preclude equivalent exemplary structures and techniques known to
21
those of ordinary skill in the art. Furthermore, to the extent that the terms
“includes,” “has,” “contains,” and other similar words are used in either the
detailed description or the claims, such terms are intended to be inclusive—in a
manner similar to the term “comprising” as an open transition word—without
precluding any additional or other elements.
As used herein, a “processing unit” or “processor” or “operating processor”
includes one or more processors, wherein processor refers to any logic circuitry
for processing instructions. A processor may be a general-purpose processor, a
special purpose processor, a conventional processor, a digital signal processor, a
plurality of microprocessors, one or more microprocessors in association with a
DSP core, a controller, a microcontroller, Application Specific Integrated Circuits,
Field Programmable Gate Array circuits, any other type of integrated circuits, etc.
The processor may perform signal coding data processing, input/output
processing, and/or any other functionality that enables the working of the system
according to the present disclosure. More specifically, the processor or processing
unit is a hardware processor.
As used herein, “a user equipment”, “a user device”, “a smart-user-device”, “a
smart-device”, “an electronic device”, “a mobile device”, “a handheld device”, “a
wireless communication device”, “a mobile communication device”, “a
communication device” may be any electrical, electronic and/or computing device
or equipment, capable of implementing the features of the present disclosure. The
user equipment/device may include, but is not limited to, a mobile phone, smart
phone, laptop, a general-purpose computer, desktop, personal digital assistant,
tablet computer, wearable device or any other computing device which is capable
of implementing the features of the present disclosure. Also, the user device may
contain at least one input means configured to receive an input from at least one
of a transceiver unit, a processing unit, a storage unit, a detection unit and any
22
other such unit(s) which are required to implement the features of the present
disclosure.
As used herein, “storage unit” or “memory unit” refers to a machine or computerreadable medium including any mechanism for storing information in a form
readable by a computer or similar machine. For example, a computer-readable
medium includes read-only memory (“ROM”), random access memory (“RAM”),
magnetic disk storage media, optical storage media, flash memory devices or
other types of machine-accessible storage media. The storage unit stores at least
the data that may be required by one or more units of the system to perform their
respective functions.
It should be noted that the terms "content" and "contents" as used in the present
disclosure may comprise various types of media, including but not limited to one
or more advertisements, one or more banners, one or more live-feeds, one or
more graphics, one or more videos, one or more avatars, and other forms of digital
media. The person skilled in the art will appreciate that the terms "content" and
"contents" are not intended to be restrictive, and that the present invention may
be applied to a wide range of different types of media and content. The use of the
terms "content" and "contents" is intended to be illustrative only and should not
be construed as limiting the scope of the invention in any way. Accordingly, it is to
be understood that any reference to "content" or "contents" in the present
disclosure should be interpreted broadly, and that the invention is not limited to
any particular type or format of media or content.
The term "advertisement/s" or "ad/s" is used interchangeably throughout this
patent specification. It is important to note that this is purely for convenience and
does not indicate any difference in meaning between the two terms. The use of
one term over the other does not imply any specific limitations or restrictions on
23
the scope of the invention, and both terms are intended to encompass any and all
forms of advertising or promotional messaging that may be delivered during a live
sports event. Therefore, any reference to "advertisement/s" or "ad/s" within this
patent specification should be considered interchangeable and equivalent in
meaning.
The term "advertisement" as used in this patent specification should be
understood to encompass any and all forms of promotional messaging, including
but not limited to video or image content, banners, or sound recordings that are
displayed or played on a part or whole of an output display alongside the live or
recorded event, or that are superimposed over the live or recorded event. It is
important to note that the use of the term "advertisement" in this specification is
intended to be broad and inclusive, and should not be limited to any specific
format or type of promotional messaging. Therefore, any reference to
"advertisement" within this patent specification should be understood to include
any and all forms of promotional messaging that are displayed or played during a
live or recorded event, regardless of their format or placement on the output
display.
The terms "live media stream," "live stream," "live event," or any other words or
phrases having a similar meaning are used interchangeably throughout this patent
specification. It is important to note that these terms are used synonymously and
do not imply any differences in meaning between them. The purpose of using
these terms interchangeably is to provide clarity and facilitate understanding of
the invention. The use of one term over the other does not restrict or limit the
scope of the invention in any way, and all such terms are intended to cover any
type of live streaming media or event. Therefore, any reference to "live media
stream," "live stream," "live event," or similar terms within this patent
specification should be considered interchangeable and equivalent in meaning.
24
The present invention relates to a system and a method for delivering one or more
contextual contents to a set of viewers of a live sports event. The set of viewers
may include one or more digital viewers of the live sports event. The present
invention provides a novel solution and method for delivering contextual content
during the live sports event, which overcomes the limitation of the prior known
solution. The invention enables advertisers to target specific demographics and
regions with contextual contents, thereby increasing the effectiveness of
advertising campaigns. The invention uses a novel solution and novel method that
analyses in real-time input media stream i.e., a media stream of the live sports
event to determine the most relevant contents to display during the live sports
event based on identification of a player face, a brand logo or a critical event from
the live sports event. The system is designed to integrate seamlessly with existing
content delivering technology, allowing for a hassle-free implementation of the
method. The patent specification details the technical aspects of the invention,
including the novel implementation, as well as the potential technical
advancement of this innovative technology. Overall, the present invention
represents a significant advancement in the field of contextual advertising and has
the potential to revolutionize the way advertisers reach their target audience
during live sports events.
Referring to Figure 1, the Figure 1 illustrates an exemplary block diagram depicting
an exemplary network architecture diagram [100], in accordance with exemplary
embodiments of the present disclosure. As shown in Figure 1, the exemplary
network architecture diagram [100] comprises at least one user equipment [102]
connected to at least one server entity [106] via at least one network [104],
wherein in an implementation the server entity [106] further comprises a system
[200] configured to implement the feature of the present invention. Also, in an
implementation the system [200] may reside in the server entity [106] or the user
25
device [102] or both or partially in either the server entity [106] or the user device
[102].
Also, in Figure 1 only the single user equipment (or may be referred to as user
device) [102], the single network [104] and the single server entity [106] are
shown, however, there may be multiple such user equipment [102], server entities
[106], and/or networks [104] or there may be any such numbers of said user
equipment [102], server entities [106] and/or networks [104] obvious to a person
skilled in the art or as required to implement the features of the present
disclosure. Further, in the implementation where the system [200] is present in
the server entity [106], based on the implementation of the features of the
present disclosure, one or more contextual contents may be provided by the
system [100] to a set of viewers of a live sports event, by displaying the one or
more contextual contentsin the live sports event displayed on the user equipment
[102].
Referring to Figure 2, an exemplary block diagram of a system [200] for delivering
one or more contextual contentsto a set of viewers of a live sports event is shown,
in accordance with the exemplary embodiments of the present invention. The
system [200] comprises at least one processing unit [202], at least one
identification unit [204], at least detection unit [206], at least one storage unit
[210] and at least one content-delivery unit [208]. Also, all of the components/
units of the system [200] are assumed to be connected to each other unless
otherwise indicated below. Also, in Fig. 2 only a few units are shown, however, the
system [200] may comprise multiple such units or the system [200] may comprise
any such numbers of said units, as required to implement the features of the
present disclosure. Further, in an implementation, the system [200] may be
present in a server device to implement the features of the present invention. The
26
system [200] may be a part of a service platform/ or may be independent of but
in communication with the service platform.
The system [200] is configured for delivering the one or more contextual contents
to the set of viewers of the live sports event, with the help of the interconnection
between the components/units of the system [200].
The processing unit [202] is configured to receive a real time media stream of the
live sports event, wherein the real time media stream comprises one or more
frames. The term "real time media stream" as used in this patent specification
includes but is not limited to images, videos, and one or more frames captured
continuously or at a time interval. It is important to note that this definition is not
intended to limit the scope of the invention in any way, but rather to provide
clarity and context to the reader. Therefore, any references made to "real time
media stream" in this patent specification should be interpreted as encompassing
all forms of media, whether captured continuously or at a time interval, including
but not limited to images and videos. For ease of understanding, let's consider an
example wherein the processing unit [202] receives a real time media stream of a
cricket match played between two teams named for e.g., Dolphin vs. Whales.
Further, the real time media stream may comprise one or more frames
encompassing at least one of one or more players of either or both teams i.e.,
Dolphin and Whales, a scoreboard of the cricket match and one or more brand
logos.
Further, the system [200] comprises an identification unit [204] connected at least
to the processing unit [202], wherein the identification unit [204] is configured to
identify in real time at least one of one or more player faces, one or more brand
logos, and a real time information of the live sports event. Further, in a preferred
implementation of the present invention, the processing unit [202] as disclosed by
the present disclosure may be configured to analyse at least one of the one or
27
more player faces, wherein to analyse the one or more player faces, the processing
unit [202] is further configured to determine at least one of a player name
associated with the one or more player faces, a detail associated with the one or
more player faces (such as for example a location of the one or more player faces
in the one or more frames), and a face score associated with the one or more
player faces. The face score associated with the one or more player faces may
indicate a correctness of the identified one or more player faces.
Further, in a preferred implementation of the present invention, the processing
unit [202] as disclosed by the present disclosure may be configured to analyse at
least one of the one or more brand logos, and the real time information of the live
sports event, wherein to analyse the one or more brand logos, the processing unit
[202] is further configured to determine at least one of a logo name associated
with the one or more brand logos, a logo location associated with the one or more
brand logos in the one or more frames, and a logo score associated with the one
or more brand logos. The logo score associated with the one or more brand logos
may indicate a correctness of the identified one or more brand logos.
In a preferred implementation of the present invention, the identification unit
[204] is further configured to identify the one or more player faces in real time
using at least one of a face recognition technique and a face detection technique.
In an exemplary implementation of the present invention, the system [200] may
include a face recognition module configured specifically for identifying player
faces in real time from the one or more frames of the real time media stream of
the live sports event. Further, the face recognition module used to identify player
faces in real time from the one or more frames of the real time media stream of
the live sports event may perform a two-step process consisting of face detection
followed by face recognition to accurately identify the players from the one or
more frames of the real time media stream of the live sports event. In an
exemplary implementation of the present invention, the system [200] may use
28
one or more known face detection techniques to detect the player faces from the
one or more frames of the real time media stream of the live sports event. Further,
the face recognition module processes the result to compare the detected faces
with one or more pre-stored faces in a database of known player faces to
accurately identify the player face. Furter, in an implementation of the present
invention, the processing unit [202] as disclosed by the present disclosure may
also be configured to analyse at least one of the one or more player faces.
Furthermore, in an implementation to analyse the one or more player faces, the
processing unit [202] is further configured to determine at least one of a player
name associated with the one or more player faces, a face detail associated with
the one or more player faces by comparing the detected faces with one or more
pre-stored faces in a database of known player faces to accurately identify the
player face. Once the player faces are identified, the face recognition module
captures each of the identified player face in a specific format, such as with a
player name, a face location in the one or more frames, and a face score (or may
be referred to as a player face score) indicating an accuracy level of the identified
face. Therefore, the face recognition module generates an output in a specific
format, which includes the player name, the face location, and the face score for
each identified player. This format makes it easy to analyse and interpret the
output and enables the system [200] to seamlessly integrate with other
applications or systems [200]. For ease of understanding continuing from the
above example, wherein the processing unit [202] receives the real time media
stream of the cricket match played between two teams named Dolphin vs.
Whales. Now, let’s say a frame from the real time media stream of the cricket
match comprises two players, John and David. The identification unit [204]
identifies the face of John and shows the output as "Face_name": "John,"
"Face_location": "657, 244, 728, 314," "Face_score": 0.75", wherein the face
location" ("657, 244, 728, 314") depicts the location of the identified face in the
29
frame and the face score" ("0.75") depicts a match score of the identified face with
the one or more pre-stored faces in the database of known player faces.
Further, in accordance with the above implementation it will be appreciated by
the person skilled in the art that the system [200] is capable of providing accurate
real time identification of players, making it suitable for use in sports applications
where fast and reliable identification is essential. Further, a person skilled in the
art would appreciate that the specific format mentioned in the present disclosure,
includes the player name, location, and accuracy of identification, can be one of a
user-defined, pre-defined, or a format common across all formats. The choice of
format will depend on various factors such as the specific use case, the
requirements of the system [200], and preferences of a user. Therefore, the
specific format mentioned in the previous paragraph is provided for illustrative
purposes only and should not be construed as limiting the scope of the invention
in any way.
In a preferred implementation of the present invention, the identification unit
[204] is further configured to identify the one or more brand logos in real time
using a customized object detection technique. In an exemplary implementation
of the present invention, the system [200] may include a logo recognition module
configured specifically for identifying brand logos from the one or more frames of
the real-time media stream of the live sports event. The frames are further
processed by the logo recognition module for identifying the brand logos with the
help of the customized object detection technique from the one or more frames
of the real time media stream of the live sports event. Further, the logo
recognition module processes the result to compare the detected brand logos
with one or more pre-stored logos of one or more brands in a database to
accurately identify the brand logos. Further, in an implementation of the present
invention, the logo recognition module to improve logo recognition accuracy may
30
use one or more text recognition techniques to identify text-based brand logos
from the one or more pre-stored logos of the one or more brands in the database.
Further, the one or more text recognition techniques used may be one of the
known text recognition techniques, such as optical character recognition (OCR)
techniques or natural language processing (NLP) techniques etc. that may be
applied to one or more languages to implement the features of the present
invention. Further, in a preferred implementation of the present invention, the
processing unit [202] as disclosed by the present disclosure may also be configured
to analyse at least one of the one or more brand logos, and the real time
information of the live sports event. Further, in an implementation to analyse the
one or more brand logos, the processing unit [202] is further configured to
determine at least one of the logo name associated with the one or more brand
logos, the logo location associated with the one or more brand logos in the one or
more frames, and the logo score associated with the one or more brand logos by
comparing and matching the detected one or more brand logos with the one or
more pre-stored logos in the database to accurately identify the one or more
brand logos.
Further, once the one or more brand logos are identified, the logo recognition
module captures the result in a specific format, such as a format comprising logo
name, logo location, and logo score. Therefore, the logo recognition module
generates the output in a specific format, which includes the logo name, logo
location, and logo score for each identified logo. Now, let’s say a frame from the
real time media stream comprises two brand logos, i.e., AZ and BY. The
identification unit [204] identifies the logo of BY and shows the output as
{“Logo_name”: “BY”, “Logo_location”: {236, 324, 453, 565}, “Logo_score”: 0.65},
wherein the logo location" ("236, 324, 453, 565") depicts the location of the
identified brand logo in the frame and the logo score" ("0.65") depicts the match
31
score of the identified brand logo with the one or more pre-stored logo in the
database.
In a preferred implementation of the present invention, the identification unit
[204] is further configured to identify the real time information of the live sports
event in real time using an artificial intelligence based optical character
recognition technique, wherein the real time information of the live sports event
comprises at least one of a real time information of a critical event of the live
sports event and a context associated with the critical event. In an exemplary
implementation of the present invention, the system [200] may include a
scoreboard recognition module configured specifically for identifying the real time
information of the live sports event in real time. In an exemplary implementation
of the present invention, the system [200] may comprise the scoreboard
recognition module configured specifically for identifying the real time
information of the live sports event in real time, wherein the frames are further
processed by the scoreboard recognition module for extracting scoreboard
information with help of the optical character recognition technique. The optical
character recognition technique may be applied to detect text regions and extract
text from the detected text region from scoreboard of the real time media stream
of the live sports event in real time. Additionally, the scoreboard recognition
module may further evaluate the extracted texts by mapping it to required fields
such as Playing Team and opponent, Player Names, scores, etc. Further, the
mapping by the scoreboard recognition module further facilitates for identifying
match events or critical events of the live sports event and a context associated
with the critical event (such as for example Boundary/ Century runs, Wickets, and
so on indicating a happy moment in the live sports event, an emotional moment
in the live sports event and/or the like), passed to scoreboard recognition module
for extracting scoreboard information. In an implementation of present invention,
the system [200] may implement a known text detection technique on the frame/s
32
of the real time media stream of the live sports event to detect text regions
followed by the one or more text recognition technique to extract text/s from the
real time media stream of the live sports event. Further, the scoreboard
recognition module may generate the output in a specific format. For ease of
understanding, continuing from the above example of the cricket match, let’s say
the scoreboard recognition module detects the following details from the
scoreboard of the real time media stream of the live sports event: -
{“FieldingTeam”: “Whales”, “BattingTeam”: “Dolphin”, “Score”: 22, “Wicket”: 1,
“Overs”: 3.1, “Batsman1”: “John”, “Batsman2”: “Samuel”, “Score1”: 16, “Score2”:
0}
It should be noted that the scoreboard of different sports events may vary in terms
of format, layout, and the type of information displayed. The mapping of the
scoreboard by the scoreboard recognition module as described in the
implementation above is just an example, and the person skilled in the art would
appreciate that the scoreboard recognition module as disclosed in the present
invention is capable to map and extract the information from the scoreboard of
various sports event.
Further, the system [200] comprises the detection unit [206] connected at least to
the processing unit [202] and the identification unit [204], wherein the detection
unit [206] is configured to detect at least one of a content display priority of the
one or more player faces, a content display priority of the one or more brand
logos, and a content display priority of the real time information of the live sports
event, based on the analysis.
It should be noted that the content display priority of the one or more player faces,
the content display priority of the one or more brand logos, and the content
display priority of the real-time information of the live sports event, based on the
33
analysis by the processing unit [202] or by the one or more corresponding unit as
disclosed by the present disclosure may be a pre-defined content display priority,
or content display priority determined in real-time based on various factors such
as a popularity of a player, a brand value of a logo, a relevance of a real-time
information to an ongoing game, a highest face score associated with the
identified one or more player faces, a size of a face location associated with the
identified one or more player faces, a highest logo score associated with the
identified one or more brand logos, a size of a logo location associated with the
identified one or more brand logos, and so on. The person skilled in the art would
appreciate that the content display priority can be customized and adapted to suit
the particular requirements of each sports event and the advertiser's preferences.
Further, the detection unit [206] is configured to detect one or more contenttriggers, based on at least one of the content display priority of the one or more
player faces, the content display priority of the one or more brand logos, and the
content display priority of the real time information of the live sports event.
Further, in an event when the one or more content-triggers are detected by the
detection unit [206] based on the content display priority of the one or more
player faces, the one or more contextual contents comprises one or more content
associated with the one or more player faces. For ease of understanding,
continuing from the above example wherein the player face identified is John and
David, in an implementation, the detection unit [206] may detect content-trigger
based on content display priority associated with the one or more detected player
faces whereby a content of brand X that comprises John is preferred over a
content of David for the same brand i.e., brand X (one of the sponsor of the sports
event) which may be further based on a greater popularity of John than David.
Further, in another implementation, when the detection unit [206] may detect
content-trigger based on the content display priority of the one or more player
34
faces, the one or more contextual contents are associated with the detected
player face such that a content of the brand Y by the detected player John which
is a competitor brand of X (one of the sponsor of the sports event) and is preferred
over a content by David for the brand i.e., brand X which may also be further based
on the greater popularity of John than David.
Further, considering an event where the detection unit [206] is configured to
detect the one or more content-triggers, based on the content display priority of
the one or more brand logos. In an implementation of the present invention, when
the one or more content-triggers are detected based on the content display
priority of the one or more detected brand logos, the one or more contextual
contents comprises one or more content associated with at least one of one or
more native brand logos and one or more competitor brand logos. For ease of
understanding, continuing from the above example wherein the player face
identified is John and David and the brand logos identified are AZ and BY, wherein
brand BY is one of the sponsors of the currently delivered live sport event. Further,
in an implementation, the detection unit [206] may detect content-trigger based
on content display priority of the one or more brand logos whereby a content of
brand BY that comprises John is preferred over a content of David for the same
brand i.e., brand BY (one of the sponsors of the sport event) which may be further
based on the greater popularity of John than David and the detection brand logo
i.e., BY in the frame of the real time media stream of the live sports event. Further,
in another implementation, when the detection unit [206] may detect contenttrigger based on the content display priority of the one or more brand logos, the
contextual contents comprises content associated with the brand logo such as that
a content by John of brand AZ i.e., a competitor brand of BY (one of the sponsor
of the sports event) is preferred over a content by David for the sponsor brand i.e.,
brand BY which may also be further based on the greater popularity of brand AZ
with John in a 34egion than the popularity of BY with David in the 34egion and
35
which may further be based on the availability of brand or products of the brand
associated with the detected brand logo in the region of content delivery i.e., AZ
and BY in the frame of the real time media stream of the live sports event.
Further, considering an event where the detection unit [206] is configured to
detect the one or more content-triggers, based on the content display priority of
the real time information of the live sports event. In an implementation of the
present invention, when the one or more content-triggers are detected based on
the content display priority of the real time information of the live sports event,
the one or more contextual contents comprises one or more content associated
with at least one of the critical event and the context associated with the critical
event. For ease of understanding, continuing from above example wherein the
scoreboard recognition module detects the following details from the scoreboard
of the real time media stream of the live sports event {“FieldingTeam”: “Whales”,
“BattingTeam”: “Dolphin”, “Score”: 22, “Wicket”: 1, “Overs”: 3.1, “Batsman1”:
“John”, “Batsman2”: “Samuel”, “Score1”: 16, “Score2”: 0}. Now, the detection unit
[206] to detect one or more content-triggers for the the one or more contextual
contents based on the real time information of the critical event and the context
associated with the critical event which in the above scoreboard may be detected
by the scoreboard recognition module as fall of wicket based on the Score 2
associated with the Batsman2 i.e., Samuel and detection of wickets from
scoreboard may be detected by the scoreboard recognition module by the
“Wicket”: 1.
Furthermore, the detection unit [206] as disclosed by the system [200] is further
configured to retrieve from the storage unit [210], the one or more contextual
contents based on the one or more content-triggers, wherein the one or more
contextual contents are based on at least one of the one or more player faces, the
one or more brand logos and the real time information of the live sports event.
36
Further, in an exemplary implementation, the detection unit [206] of the system
[200] as disclosed by the present disclosure may retrieve the one or more
contextual contents from the storage unit [210] based on the one or more
content-triggers. In an implementation, the one or more content-triggers may be
based on the factors such as a region of player popularity, a region where a brand
or its competitor brand is available, and a region where a critical event and a
context associated with the critical event is relevant. The detection unit [206] may
retrieve the one or more contextual contents from the storage unit [210] based
on the analysis by the processing unit [202] or the corresponding unit as may be
implemented by the present system [200]. The one or more contextual contents
may be selected based on the identified player faces, brand logos, and real-time
information of the live sports event. The retrieved one or more contextual
contents may be delivered by the content-delivery unit [208] to the set of viewers
of the live sports event. The retrieval of contextual contents from the storage unit
[210] based on the content-triggers may enhance the effectiveness of the contents
by targeting specific regions and critical events where the contents are more
relevant to the audience.
Further, the system [200] encompasses the content-delivery unit [208] connected
at least to the processing unit [202] the identification unit [204] and the detection
unit [206], wherein the content-delivery unit [208] is configured to deliver the one
or more contextual contents to the set of viewers of the live sports event. Further,
in an implementation, delivering by the content-delivery unit [208] the one or
more contextual contents to the set of viewers of the live sports event are
determined based on various factors such as the region of player popularity, the
region where a brand or its competitor brand is available, and where the critical
event and the context associated with the critical event is relevant. In an
exemplary implementation of the present invention, the system [200] may further
comprise the processing unit [202] or one or more corresponding units configured
37
to analyze the real time media stream of the live sports event and determine the
content triggers. Based on the analysis, the content-delivery unit [208] may deliver
the contextual contents in real time.
Referring to Figure 3 an exemplary method flow diagram [300], for delivering one
or more contextual contents to a set of viewers of a live sports event, in
accordance with exemplary embodiments of the present invention is shown. In an
implementation the method is performed by the system [200]. Further, in an
implementation, the system [200] may be present in a server device to implement
the features of the present invention. Also, as shown in Figure 3, the method starts
at step [302].
At step [304] the method comprises receiving, at a processing unit [202], a real
time media stream of the live sports event, wherein the real-time media stream
comprises one or more frames. Further, the term "real time media stream" as used
in this patent specification includes but is not limited to images, videos, and one
or more frames captured continuously or at a time interval. It is important to note
that this definition is not intended to limit the scope of the invention in any way,
but rather to provide clarity and context to the reader. Therefore, any references
made to "real time media stream" in this patent specification should be
interpreted as encompassing all forms of media, whether captured continuously
or at a time interval, including but not limited to images and videos. For ease of
understanding, let's consider an example wherein the processing unit [202]
receives a real time media stream of a cricket match played between two teams
named for e.g., Dolphin vs. Whales. Further, the real time media stream may
comprise one or more frames encompassing at least one of one or more players
of either or both teams i.e., Dolphin and Whales, a scoreboard of the cricket match
and one or more brand logos.
38
At step [306] and step [308] the method comprises identifying in real time, by an
identification unit [204], at least one of one or more player faces, one or more
brand logos, and a real time information of the live sports event and analysing, by
the processing unit [202], at least one of the one or more player faces, the one or
more brand logos, and the real time information of the live sports event,
respectively. Further, as disclosed by the present disclosure, the analysing by the
processing unit [202] the one or more player faces further comprises determining
at least one of a player name associated with the one or more player faces, a face
location associated with the one or more player faces in the one or more frames,
and a player face score associated with the one or more player faces. Furthermore,
as disclosed by the present disclosure, the analysing by the processing unit [202]
the one or more brand logos further comprises determining at least one of a logo
name associated with the one or more brand logos, a logo location associated with
the one or more brand logos in the one or more frames, and a logo score
associated with the one or more brand logos.
Further, in a preferred implementation of the present invention, the identification
unit [204] identifies the one or more player faces in real time using at least one of
a face recognition technique and a face detection technique. In an exemplary
implementation of the present invention, the method as disclosed by the present
disclosure may include a face recognition module specifically for identifying player
faces in real time from the one or more frames of the real time media stream of
the live sports event. Further, the face recognition module used to identify player
faces in real time from the one or more frames of the real time media stream of
the live sports event may perform a two-step process consisting of face detection
followed by face recognition to accurately identify the players from the one or
more frames of the real time media stream of the live sports event. In an
exemplary implementation of the present invention, the method as disclosed by
the present disclosure may use one or more known face detection techniques to
39
detect the player faces from the one or more frames of the real time media stream
of the live sports event. Further, the face recognition module processes the result
to compare the detected faces with a one or more pre-stored faces in a database
of known player faces to accurately identify the player face. Furter, in an
implementation of the present invention, the processing unit [202] as disclosed by
the present disclosure may also analyses at least one of the one or more player
faces. Furthermore, to analyse the one or more player faces, the processing unit
[202] determines at least one of a player name associated with the one or more
player faces, a face by comparing the detected faces with a one or more pre-stored
faces in a database of known player faces to accurately identify the player face.
Once the faces are identified, the face recognition module captures the result in a
specific format, such as player name, face location, and face score. Further, the
face recognition module generates the output in a specific format, which includes
the player name, face location, and face score for each identified player. This
format makes it easy to analyse and interpret the output and enables the method
as disclosed by the present disclosure to seamlessly integrate with other
applications or systems [200]. For ease, of understanding continuing from the
above example, wherein the processing unit [202] receives the real time media
stream of the cricket match played between two teams named Dolphin vs.
Whales. Now, let’s say the frame from the real time media stream comprises two
players, John and David. The identification unit [204] identifies the face of John
and shows the output as "Face_name": "John," "Face_location": "657, 244, 728,
314," "Face_score": 0.75", wherein the face location" ("657, 244, 728, 314")
depicts the location of the identified face in the frame and the face score" ("0.75")
depicts the match score of the identified face with the one or more pre-stored
faces in the database of known player faces.
Further, in accordance with the above implementation it will be appreciated by
the person skilled in the art that the method as disclosed by the present disclosure
40
is capable of providing accurate real time identification of players, making it
suitable for use in sports applications where fast and reliable identification is
essential. Further, a person skilled in the art would appreciate that the specific
format mentioned in the present disclosure, includes the player name, location,
and accuracy of identification, can be one of a user-defined, pre-defined, or a
format common across all formats. Therefore, the specific format mentioned in
the previous paragraph is provided for illustrative purposes only and should not
be construed as limiting the scope of the invention in any way.
In a preferred implementation of the present invention, the identification unit
[204] identifiesthe one or more brand logos in real time using a customized object
detection technique. In an exemplary implementation of the present invention,
the method as disclosed by the present disclosure may include a logo recognition
module specifically for identifying brand logos from the one or more frames of the
real-time media stream of the live sports event. The frames are further processed
by the logo recognition module for identifying the brand logos with the help of the
customized object detection technique from the one or more frames of the real
time media stream of the live sports event. Further, the logo recognition module
processes the result to compare the detected brand logos with one or more prestored logos of one or more brands in a database to accurately identify the brand
logos. Further, in an implementation of the present invention, the logo recognition
module to improve logo recognition accuracy may use one or more text
recognition techniques to identify text-based brand logos from the one or more
pre-stored logos of the one or more brands in the database. Further, the one or
more text recognition techniques used may be one of the known text recognition
techniques, such as optical character recognition (OCR) techniques or natural
language processing (NLP) techniques etc. Further, in a preferred implementation
of the present invention, the processing unit [202] as disclosed by the present
disclosure may also analyses at least one of the one or more brand logos, and the
41
real time information of the live sports event. Further, in an implementation to
analyse the one or more brand logos, the processing unit [202] determines at least
one of the logo name associated with the one or more brand logos, the logo
location associated with the one or more brand logos in the one or more frames,
and the logo score associated with the one or more brand logos by comparing and
matching the detected one or more brand logos with the one or more pre-stored
logos in the database to accurately identify the one or more brand logos.
Further, once the one or more brand logos are identified, the logo recognition
module captures the result in a specific format, such as a format comprising logo
name, logo location, and logo score. Therefore, the logo recognition module
generates the output in a specific format, which includes the logo name, logo
location, and logo score for each identified logo. Now, let’s say a frame from the
real time media stream comprises two brand logos, i.e., AZ and BY. The
identification unit [204] identifies the logo of BY and shows the output as
{“Logo_name”: “BY”, “Logo_location”: {236, 324, 453, 565}, “Logo_score”: 0.65},
wherein the logo location" ("236, 324, 453, 565") depicts the location of the
identified brand logo in the frame and the logo score" ("0.65") depicts the match
score of the identified brand logo with the one or more pre-stored logo in the
database. In a preferred implementation of the present invention, the
identification unit [204] identifiesthe real time information of the live sports event
in real time using an artificial intelligence based optical character recognition
technique, wherein the real time information of the live sports event comprises at
least one of a real time information of a critical event of the live sports event and
a context associated with the critical event.
Further, in an exemplary implementation of the present invention, the method as
disclosed by the present disclosure may include the scoreboard recognition
module specifically for identifying the real time information of the live sports
42
event in real time. In an exemplary implementation of the present invention, the
method as disclosed by the present disclosure may further comprise a scoreboard
recognition module specifically for identifying the real time information of the live
sports event in real time, wherein the frames are further processed by the
scoreboard recognition module for extracting scoreboard information with help
of the optical character recognition technique. The optical character recognition
technique may be applied to detect text regions and extract text from the
detected text region from scoreboard of the real time media stream of the live
sports event in real time. Additionally, the scoreboard recognition module may
further evaluate the extracted texts by mapping it to required fields such as
Playing Team and opponent, Player Names, scores, etc. Further, the mapping by
the scoreboard recognition module further facilitates for identifying match events
or critical events of the live sports event and a context associated with the critical
event (such as for example Boundary/ Century runs, Wickets, and so on indicating
a happy moment in the live sports event, an emotional moment in the live sports
event and/or the like) passed to scoreboard recognition module for extracting
scoreboard information. In an implementation of present invention, the method
as disclosed by the present disclosure may implement a known text detection
technique on the frame/s of the real time media stream of the live sports event to
detect text regions followed by the one or more text recognition technique to
extract text/s from the real time media stream of the live sports event. Further,
the scoreboard recognition module may generate the output in a specific format.
For ease of understanding, continuing from the above example of the cricket
match, let’s say the scoreboard recognition module detects the following details
from the scoreboard of the real time media stream of the live sports event:
{“FieldingTeam”: “Whales”, “BattingTeam”: “Dolphin”, “Score”: 22, “Wicket”: 1,
“Overs”: 3.1, “Batsman1”: “John”, “Batsman2”: “Samuel”, “Score1”: 16, “Score2”:
0}
43
It should be noted that the scoreboard of different sports events may vary in terms
of format, layout, and the type of information displayed. The mapping of the
scoreboard by the scoreboard recognition module as described in the
implementation above is just an example, and the person skilled in the art would
appreciate that the scoreboard recognition module as disclosed in the present
invention is capable to map and extract the information from the scoreboard of
various sports event.
At step [310] the method comprises detecting, by the detection unit [206], at least
one of a content display priority of the one or more player faces, a content display
priority of the one or more brand logos, and a content display priority of the real
time information of the live sports event, based on the analysis. It should be noted
that the content display priority of the one or more player faces, the content
display priority of the one or more brand logos, and the content display priority of
the real-time information of the live sports event, based on the analysis by the
processing unit [202] or by the one or more corresponding unit as disclosed by the
present disclosure may be a pre-defined content display priority, or content
display priority determined in real-time based on various factors such as a
popularity of a player, a brand value of a logo, a relevance of a real-time
information to an ongoing game, a highest face score associated with the
identified one or more player faces, a size of a face location associated with the
identified one or more player faces, a highest logo score associated with the
identified one or more brand logos, a size of a logo location associated with the
identified one or more brand logos, and so on. The person skilled in the art would
appreciate that the content display priority can be customized and adapted to suit
the particular requirements of each sports event and the advertiser's preferences.
44
At step [312] the method comprises detecting, by the detection unit [206], one or
more content-triggers, based on at least one of the content display priority of the
one or more player faces, the content display priority of the one or more brand
logos, and the content display priority of the real time information of the live
sports event. Further, in an event the one or more content-triggers are detected
by the detection unit [206] based on the content display priority of the one or
more player faces, the one or more contextual contents comprises one or more
content associated with the one or more player faces. For ease of understanding,
continuing from the above example wherein the player face identified is John and
David, in an implementation, the detection unit [206] may detect content-trigger
based on content display priority associated with the one or more detected player
faces whereby a content of brand X that comprises John is preferred over a
content of David for the same brand i.e., brand X (one of the sponsor of the sports
event) which may be further based on a greater popularity of John than David.
Further, in another implementation, when the detection unit [206] may detect
content-trigger based on the content display priority of the one or more player
faces, the one or more contextual contents are associated with the detected
player face such that a content of the brand Y by the detected player John which
is a competitor brand of X (one of the sponsor of the sports event) and is preferred
over a content by David for the brand i.e., brand X which may also be further based
on the greater popularity of John than David.
Further, considering an event where the detection unit [206] detects the one or
more content-triggers, based on the content display priority of the one or more
brand logos. In an implementation of the present invention, when the one or more
content-triggers are detected based on the content display priority of the one or
more detected brand logos, the one or more contextual contents comprises one
or more content associated with at least one of one or more native brand logos
and one or more competitor brand logos. For ease of understanding, continuing
45
from the above example wherein the player face identified is John and David and
the brand logos identified are AZ and BY, wherein brand BY is one of the sponsor
of the currently delivered live sport event. Further, in an implementation, the
detection unit [206] may detect content-trigger based on content display priority
of the one or more brand logos whereby a content of brand BY that comprises
John is preferred over a content of David for the same brand i.e., brand BY (one of
the sponsor of the sport event) which may be further based on the greater
popularity of John than David and the detection brand logo i.e., BY in the frame of
the real time media stream of the live sports event.
Further, in another implementation, when the detection unit [206] may detect
content-trigger based on the content display priority of the one or more brand
logos, the contextual contents comprises content associated with the brand logo
such as that a content by John of brand AZ i.e., a competitor brand of BY (one of
the sponsor of the sports event) is preferred over a content by David for the
sponsor brand i.e., brand BY which may also be further based on the greater
popularity of brand AZ with John in a reigon than the popularity of BY with David
in the reigon and which may further be based on the availability of brand or
products of the brand associated with the detected brand logo in the region of
content delivery i.e., AZ and BY in the frame of the real time media stream of the
live sports event.
Further, considering an event where the detection unit [206] detects the one or
more content-triggers, based on the content display priority of the real time
information of the live sports event. In an implementation of the present
invention, when the one or more content-triggers are detected based on the
content display priority of the real time information of the live sports event, the
one or more contextual contents comprises one or more content associated with
at least one of the critical event and the context associated with the critical event.
46
For ease of understanding, continuing from above example wherein the
scoreboard recognition module detects the following details from the scoreboard
of the real time media stream of the live sports event {“FieldingTeam”: “Whales”,
“BattingTeam”: “Dolphin”, “Score”: 22, “Wicket”: 1, “Overs”: 3.1, “Batsman1”:
“John”, “Batsman2”: “Samuel”, “Score1”: 16, “Score2”: 0}. Now, the detection unit
[206] to detect one or more content-triggers for the the one or more contextual
contents based the real time information of the critical event and the context
associated with the critical event which in the above scoreboard may be detected
by the scoreboard recognition module as fall of wicket based on the Score 2
associated with the Batsman2 i.e., Samuel and detection of wickets from
scoreboard may be detected by the scoreboard recognition module by the
“Wicket”: 1.
At step [314] the method comprises retrieving, by the detection unit [206] from
the storage unit [210], the one or more contextual contents based on the one or
more content-triggers, wherein the one or more contextual contents are based on
at least one of the one or more player faces, the one or more brand logos and the
real time information of the live sports event. In an exemplary implementation,
the detection unit [206] of the method [300] as disclosed by the present disclosure
may retrieve the one or more contextual contents from the storage unit [210]
based on the one or more content-triggers. In an implementation, the one or more
content-triggers may be based on the factors such as a region of player popularity,
a region where a brand or its competitor brand is available, and where a critical
event and a context associated with the critical event is relevant. The detection
unit [206] may retrieve the one or more contextual contentsfrom the storage unit
[210] based on the analysis by the processing unit [202] or the corresponding unit
as may be utilized for such purpose by the present method. The one or more
contextual contents may be selected based on the identified player faces, brand
logos, and real-time information of the live sports event. The retrieved one or
47
more contextual contents may be delivered by the content-delivery unit [208] to
the set of viewers of the live sports event. The retrieval of contextual contents
from the storage unit [210] based on the content-triggers may enhance the
effectiveness of the contents by targeting specific regions and critical events
where the contents are more relevant to the audience.
At step [316] the method comprises delivering, by the content-delivery unit [208],
the one or more contextual contents to the set of viewers of the live sports event.
Further, in an implementation, delivering by the content-delivery unit [208] the
one or more contextual contents to the set of viewers of the live sports event are
determined based on various factors such as the region of player popularity, the
region where a brand or its competitor brand is available, and where the critical
event and the context associated with the critical event is relevant. In an
exemplary implementation of the present invention, the method may further
encompass the processing unit [202], or one or more corresponding units to
analyze the real time media stream of the live sports event and determine the
content triggers. Based on the analysis, the content-delivery unit [208] may deliver
the contextual contents in real time.
The method ends at step [318].
Further an aspect of the present disclosure may relate to a user device for
delivering one or more contextual contents to a set of viewers of a live sports
event, the user device comprises a system [200]. The system [200] comprises a
processing unit [202], configured to receive a real time media stream of the live
sports event, wherein the real-time media stream comprises one or more frames.
The system [200] further comprises an identification unit [204] connected at least
to the processing unit [202], wherein the identification unit [204] is configured to
identify in real time, at least one of one or more player faces, one or more brand
logos, and a real time information of the live sports event. The processing unit
48
[202] is further configured to analyze at least one of the one or more player faces,
the one or more brand logos, and the real time information of the live sports
event. Further the system [200] comprises a detection unit [206] connected at
least to the processing unit [202] and the identification unit [204], wherein the
detection unit [206] is configured to: detect at least one of a content display
priority of the one or more player faces, a content display priority of the one or
more brand logos, and a content display priority of the real time information of
the live sports event, based on the analysis; detect one or more content-triggers,
based on at least one of the content display priority of the one or more player
faces, the content display priority of the one or more brand logos, and the content
display priority of the real time information of the live sports event; and retrieve
from a storage unit [210], the one or more contextual contents based on the one
or more content-triggers, wherein the one or more contextual contents are based
on at least one of the one or more player faces, the one or more brand logos and
the real time information of the live sports event. The system [200] further
comprises a content-delivery unit [208] connected at least to the processing unit
[202], the identification unit [204] and the detection unit [206], wherein the
content-delivery unit [208] is configured to deliver the one or more contextual
contents to the set of viewers of the live sports event.
Further, another aspect of the present disclosure relates to a non-transitory
computer readable storage medium storing instructions for delivering one or
more contextual contents to a set of viewers of a live sports event, the storage
medium comprising executable code which, when executed by a processor, causes
the processor to: receive a real time media stream of the live sports event,
wherein the real-time media stream comprises one or more frames; enable an
identification unit [204] to identify in real time, at least one of one or more player
faces, one or more brand logos, and a real time information of the live sports
event; analyse at least one of the one or more player faces, the one or more brand
49
logos, and the real time information of the live sports event; enable a detection
unit [206] to: 1) detect at least one of a content display priority of the one or more
player faces, a content display priority of the one or more brand logos, and a
content display priority of the real time information of the live sports event, based
on the analysis, 2) detect one or more content-triggers, based on at least one of
the content display priority of the one or more player faces, the content display
priority of the one or more brand logos, and the content display priority of the
real time information of the live sports event, and 3) retrieve from a storage unit
[210], the one or more contextual contents based on the one or more contenttriggers, wherein the one or more contextual contents are based on at least one
of the one or more player faces, the one or more brand logos and the real time
information of the live sports event; and enable a content-delivery unit [208] to
deliver the one or more contextual contents to the set of viewers of the live sports
event.
In one of the implementations of the present disclosure the system [200] is
configured providing one or more notifications to the content- delivery unit [208],
with the help of the interconnection between the components/units of the system
[200]. In such implementation the content delivery unit [208] may reside within
the system [200] or may reside outside the system [100] but is in a connection
with the system [200].
Further in such implementation the processing unit [202] is configured to receive
a real time media stream of the live sports event, wherein the real-time media
stream comprises one or more frames. As disclosed above the term "real time
media stream" as used in this patent specification includes but is not limited to
images, videos, and one or more frames captured continuously or at a time
interval. It is important to note that this definition is not intended to limit the
50
scope of the invention in any way, but rather to provide clarity and context to the
reader. Therefore, any references made to "real time media stream" in this patent
specification should be interpreted as encompassing all forms of media, whether
captured continuously or at a time interval, including but not limited to images
and videos. For ease of understanding, let's consider an example wherein the
processing unit [202] receives a real time media stream of a cricket match played
between two teams named Dolphin vs. Whales. Further, one or more frames of
the real time media stream comprises at least one of one or more players of either
or both teams i.e., Dolphin and Whales and one or more brand logos.
Further, the processing unit [202] is configured to receive scoreboard proportions,
wherein the scoreboard proportions are determined by the processing unit [202]
based on a manual input and wherein the scoreboard proportions comprise at
least one of a scoreboard position and a scoreboard layout. A scoreboard region
of each sport or sporting event may have a different scoreboard format and
position. Therefore, a person skilled in the art would appreciate that the system
[200] disclosed herein must account for this variation to accurately detect and
crop out the scoreboard area from the real-time media stream of the live sports
event. Further, in an exemplary implementation, the processing unit [202]
analyzes the position and layout of the scoreboard and identifies the proportions
for a scoreboard region from the one or more frames of the real-time media
stream of the live sports event. These pre-identified scoreboard proportions are
then applied on the one or more frames of the real-time media stream of the live
sports event to crop out a scoreboard area. Further, the person skilled in the art
would also appreciate that in an exemplary implementation the analyzing and
identifying the proportions for the scoreboard region may be performed based on
the manual input to ensure accurate detection of the scoreboard area. Further, in
an implementation the scoreboard area from the one or more frames of the real-
51
time media stream of the live sports event, once analyzed by the processing unit
[202], may be stored in the storage unit [210] and applied on a frame of another
real-time media stream of the live sports event to crop out the scoreboard area
from the frame of the another real-time media stream of the live sports event.
Therefore, the processing unit [202] is configured to crop the scoreboard area
from the one or more frames based on the scoreboard proportions. It should be
noted that the scoreboard of different sports events may vary in terms of format,
layout, and the type of information displayed.
Further, the processing unit [202] is configured detect using one or more text
detection techniques, one or more region of interests (ROIs) from the scoreboard
area. In a preferred implementation of the present invention, the processing unit
[202] is further configured to detect the one or more region of interests (ROIs)
from the scoreboard area of the live sports event in real time using at least one of
one or more artificial intelligence based optical character recognition technique
and one or more text detection techniques.
Further, the processing unit [202] is configured extract using one or more text
recognition techniques, one or more texts from the one or more ROIs. Further, the
processing unit [202] as disclosed by the present invention may further extract
text from the region of interests (ROIs) identified from the scoreboard area of the
one or more frames of the real time media stream of the live sports event, in real
time. Further, in an exemplary implementation the optical character recognition
technique may be applied to detect text regions and extract text from the
detected region of interests (ROIs) from the scoreboard area of the real time
media stream of the live sports event in real time.
Further, the processing unit [202] is configured to map the one or more texts with
one or more pre-defined texts, wherein the one or more pre-defined texts
52
comprises one or more fields associated with the live sports event. In an
exemplary implementation of the present invention, the processing unit [202] is
configured to evaluate the extracted text(s) by mapping the one or more texts
with the one or more pre-defined textssuch as Playing Team and opponent, Player
Names, scores, etc. Further, according to the present disclosure the mapping the
one or more texts with one or more pre-defined texts by the processing unit [202]
further facilitate for identifying sports events or critical events of the live sports
event and a context associated with the critical event such as Boundary/ Century
runs, Wickets, and etc., for extracting scoreboard information from the
scoreboard area of the real time media stream of the live sports event in real time.
Further, in an implementation of present invention, the system [200] may
implement a known text detection technique on the frame/s of the real time
media stream of the live sports event to detect text regions followed by the one
or more text recognition technique to extract text/s from the real time media
stream of the live sports event.
Further, the processing unit [202] is configured determine a scoreboard-based
information (or may be referred as the scoreboard information) based on the
mapping. Further, the processing unit [202] is configured detect at least one of
one or more players and one or more scoreboard-based events based at least on
the scoreboard-based information, wherein the one or more scoreboard-based
events comprises one or more scoreboard-based contextual events. In an
implementation, the processing unit [202] is configured determine the
scoreboard-based information based on the mapping and generate an output in a
specific format, for ease of understanding lets consider an example wherein the
processing unit [202] determine a scoreboard-based information based on the
mapping and detects details such as: -
53
{“FieldingTeam”: “Whales”, “BattingTeam”: “Dolphin”, “Score”: 22, “Wicket”: 0,
“Overs”: 3.1, “Batsman1”: “John”, “Batsman2”: “Chris”, “Score1”: 16, “Score2”: 0}
Further, in the above example, in a next frame of the the real time media stream
of the live sports event the processing unit [202] determines the scoreboard-based
information based on the mapping and detects:
{“FieldingTeam”: “Whales”, “BattingTeam”: “Dolphin”, “Score”: 22, “Wicket”: 1,
“Overs”: 3.1, “Batsman1”: “John”, “Batsman2”: “Samuel”, “Score1”: 16, “Score2”:
0}.
Now, in the above example the processing unit [202] of the system [200] as
disclosed by the present disclosure is configured detect a scoreboard-based event
based at least on the scoreboard-based information a as fall of wicket based on
the Score 2 associated with the Batsman2 i.e., Samuel and detection of wickets
from scoreboard as “Wicket”: 1.
Further, the processing unit [202] is configured to provide the one or more
notifications to the content delivery unit [208] based at least on the at least one
of the detected one or more players and the detected one or more scoreboardbased events, such as wicket as discussed in above example. Further, the
processing unit [202] to provide the one or more notifications to the content
delivery unit [208] based on the detected one or more scoreboard-based events,
is further configured to analyze, the one or more scoreboard-based events based
on a historical sports data associated with the live sports event. Further, the
processing unit [202] is configured to: determine at least one of a winning
probability associated with the live sports event and a priority of the one or more
scoreboard-based events based on the analysis, and provide, the one or more
notifications to the content delivery unit [208] based on at least one of the winning
probability and the priority. Further, in an exemplary implementation of the
54
present invention the processing unit [202] is further configured to generate one
or more avatars of at least one of one or more commentators, the one or more
players and the one or more scoreboard-based events based on the scoreboardbased information. In an exemplary implementation, the processing unit [202] is
configured to generate avatars of commentators, players, and scoreboard-based
events based on the scoreboard-based information. Further, the avatars may be
further processed by the content delivery unit [208]. The processing unit [202]
may use a variety of techniques to generate the avatars, such as image processing,
machine learning, or animation. By generating avatars based on scoreboard-based
information, the processing unit [202] provides a unique and engaging experience
for users who can interact with the avatars in real-time.
Referring to Figure 4, an exemplary method flow diagram [400], for providing one
or more notifications to a content- delivery unit [208], in accordance with
exemplary embodiments of the present invention is shown. In an implementation
the method [400] is performed by the system [200]. Further, in an
implementation, the system [200] may be present in a server device to implement
the features of the present invention. Also, as shown in Figure 4, the method [400]
starts at step [402].
Next, at step [404] the method [400] comprises receiving, at a processing unit
[202], a real time media stream of the live sports event, wherein the real-time
media stream comprises one or more frames. As disclosed above the term "real
time media stream" as used in this patent specification includes but is not limited
to images, videos, and one or more frames captured continuously or at a time
interval. It is important to note that this definition is not intended to limit the
scope of the invention in any way, but rather to provide clarity and context to the
reader. Therefore, any references made to "real time media stream" in this patent
specification should be interpreted as encompassing all forms of media, whether
55
captured continuously or at a time interval, including but not limited to images
and videos. For ease of understanding, let's consider an example wherein the
processing unit [202] receives a real time media stream of a cricket match played
between two teams named Dolphin vs. Whales. Further, one or more frames of
the real time media stream at least one of one or more players of either or both
teams i.e., Dolphin and Whales and one or more brand logos.
At step [406] the method [400] comprises receiving, by the processing unit [202],
scoreboard proportions, wherein the scoreboard proportions are determined by
the processing unit [202] based on a manual input and wherein the scoreboard
proportions comprise at least one of a scoreboard position and a scoreboard
layout.Further at step [408] the method [400] comprises the cropping, by the
processing unit [202], a scoreboard area from the one or more frames based on
the scoreboard proportions. A scoreboard region of each sport or sporting event
may have a different scoreboard format and position. Therefore, a person skilled
in the art would appreciate that the method [400] as disclosed herein must
account for this variation to accurately detect and crop out the scoreboard area
from the real-time media stream of the live sports event. Further, in an exemplary
implementation, the processing unit [202]as disclosed herein analyzes the
position and layout of the scoreboard and identifies the proportions for a
scoreboard region from the one or more frames of the real-time media stream of
the live sports event. These pre-identified scoreboard proportions are then
applied on the one or more frames of the real-time media stream of the live sports
event to crop out a scoreboard area. Further, the person skilled in the art would
also appreciate that in an exemplary implementation the analyzing and identifying
the proportions for the scoreboard region may also be performed manually to
ensure accurate detection of the scoreboard area. Further, the scoreboard region
from the one or more frames of the real-time media stream of the live sports
event, once analyzed by the processing unit [202], may be stored in the storage
unit [210] and applied on a frame of another one or more frames of the real-time
56
media stream of the live sports event to crop out the scoreboard area from the
one or more frames of the real-time media stream of the live sports event.
Thereafter, the processing unit [202] crops a scoreboard area from the one or
more frames based on the scoreboard proportions. It should be noted that the
scoreboard of different sports events may vary in terms of format, layout, and the
type of information displayed.
At step [410], the method [400] comprises detecting, by the processing unit [202]
using one or more text detection techniques, one or more region of interests
(ROIs) from the scoreboard area. In a preferred implementation of the present
invention, the processing unit [202] detects the one or more region of interests
(ROIs) from the scoreboard area of the live sports event in real time using at least
one of oner or more artificial intelligence based optical character recognition
technique and one or more text detection techniques.
At step [412], the method [400] comprises extracting, by the processing unit [202]
using one or more text recognition techniques, one or more texts from the one or
more ROIs. In an implementation, the processing unit [202] as disclosed by the
present invention may further extract text from the region of interests (ROIs)
identified from the scoreboard area of the one or more frames of the real time
media stream of the live sports event, in real time. Further, in an exemplary
implementation the optical character recognition technique may be applied to
detect text regions and extract text from the detected region of interests (ROIs)
from the scoreboard area of the real time media stream of the live sports event in
real time.
At step [414], the method [400] comprises mapping, by the processing unit [202],
the one or more texts with one or more pre-defined texts, wherein the one or
more pre-defined texts comprises one or more fields associated with the live
sports event. In an exemplary implementation of the present invention, the
processing unit [202] evaluates the extracted texts by mapping the one or more
57
texts with the one or more pre-defined texts such as Playing Team and opponent,
Player Names, scores, etc. Further, according to the present disclosure the
mapping the one or more texts with one or more pre-defined texts by the
processing unit [202] further facilitate for identifying sports events or critical
events of the live sports event and a context associated with the critical event such
as Boundary/ Century runs, Wickets, and etc. for extracting scoreboard
information from the scoreboard area of the real time media stream of the live
sports event in real time. Further, in an implementation of present invention, the
method [400] as disclosed herein may implement a known text detection
technique on the frame/s of the real time media stream of the live sports event to
detect text regions followed by the one or more text recognition technique to
extract text/s from the real time media stream of the live sports event.
At step [416], the method [400] comprises determining, by the processing unit
[202], a scoreboard-based information (or may be referred as the scoreboard
information) based on the mapping. Further, and at step [418], the processing unit
[202] is comprises determining, by the processing unit, a scoreboard-based
information based on the mapping, wherein the one or more scoreboard-based
events comprises one or more scoreboard-based contextual events. In an
implementation, the processing unit [202] determines the scoreboard-based
information based on the mapping and generate an output in a specific format,
for ease of understanding lets consider an example wherein the processing unit
[202] determine a scoreboard-based information based on the mapping and
detects details such as:
{“FieldingTeam”: “Whales”, “BattingTeam”: “Dolphin”, “Score”: 22, “Wicket”: 0,
“Overs”: 3.1, “Batsman1”: “John”, “Batsman2”: “Chris”, “Score1”: 16, “Score2”: 0}
Further, in the above example, in a next frame of the the real time media stream
of the live sports event the processing unit [202] determines the scoreboard-based
information a scoreboard-based information based on the mapping and detects:
58
{“FieldingTeam”: “Whales”, “BattingTeam”: “Dolphin”, “Score”: 22, “Wicket”: 1,
“Overs”: 3.1, “Batsman1”: “John”, “Batsman2”: “Samuel”, “Score1”: 16, “Score2”:
0}.
Further, in the above example the processing unit [202] as disclosed by the present
disclosure detects a scoreboard-based event based at least on the scoreboardbased information a as fall of wicket based on the Score 2 associated with the
Batsman2 i.e., Samuel and detection of wickets from scoreboard as “Wicket”: 1.
At step [420], the method [400] comprises providing, by the processing unit [202],
the one or more notifications to the content delivery unit [208] based at least on
the at least one of the detected one or more players and the detected one or more
scoreboard-based events. such as wicket as discussed in above example. Further,
processing unit [202] as disclosed herein provides the one or more notifications to
the content delivery unit [208] based on the detected one or more scoreboardbased events, and also the processing unit [202] analyzes the one or more
scoreboard-based events based on a historical sports data associated with the live
sports event. Further, the processing unit [202] determines at least one of a
winning probability associated with the live sports event and a priority of the one
or more scoreboard-based events based on the analysis, and provide, the one or
more notifications to the content delivery unit [208] based on at least one of the
winning probability and the priority.
The method [400] terminates at step [422].
An aspect of the present disclosure relates to a user device for providing one or
more notifications to a content delivery unit [208], the user device comprising: a
system [200], wherein the system [200] comprises: a processing unit [202],
configured to: receive a real time media stream of the live sports event, wherein
59
the real-time media stream comprises one or more frames; receive scoreboard
proportions; crop a scoreboard area from the one or more frames based on the
scoreboard proportions; detect using one or more text detection techniques, one
or more region of interests (ROIs) from the scoreboard area; extract using one or
more text recognition techniques, one or more texts from the one or more ROIs;
map the one or more texts with one or more pre-defined texts; determine a
scoreboard-based information based on the mapping; detect at least one of one
or more players and one or more scoreboard-based events based at least on the
scoreboard-based information; and provide the one or more notifications to the
content delivery unit [208] based at least on the at least one of the detected one
or more players and the detected one or more scoreboard-based events.
Another aspect of the present disclosure relates to a non-transitory computer
readable storage medium storing instructions for providing one or more
notifications to a content delivery unit [208], the storage medium comprising
executable code which, when executed by a processor, causes the processor to:
receive a real time media stream of the live sports event, wherein the real-time
media stream comprises one or more frames; receive scoreboard proportions;
crop a scoreboard area from the one or more frames based on the scoreboard
proportions; detect using one or more text detection techniques, one or more
region of interests (ROIs) from the scoreboard area; extract using one or more text
recognition techniques, one or more texts from the one or more ROIs; map the
one or more texts with one or more pre-defined texts; determine a scoreboardbased information based on the mapping; detect at least one of one or more
players and one or more scoreboard-based events based at least on the
scoreboard-based information; and provide the one or more notifications to the
content delivery unit [208] based at least on the at least one of the detected one
or more players and the detected one or more scoreboard-based events.
60
In an implementation of the present disclosure, the system [200] is configured for
automatically generating one or more scoreboard templates for a live sports
event, with the help of the interconnection between the components/units of the
system [200].
In said implementation the processing unit [202] is configured to receive a media
stream of the live sports event, wherein the media stream comprises one or more
frames. For ease of understanding, let's consider an example wherein the
processing unit [202] receives a real time media stream of a cricket match played
between two teams named Dolphin vs. Whales. Further, one or more frames of
the real time media stream comprises one or more players of either or both teams
i.e., Dolphin and Whales and/or one or more brand logos.
Further, the processing unit [202] is configured to crop a pre-defined portion of
the one or more frames. In one implementation of the invention, the processing
unit [202] of the present disclosure assumes that a scoreboard may be present at
a bottom of the one or more frames. It is to be noted that in most cases, the
scoreboard will be present at the bottom of the frame(s), and the present
invention may further process the frame to detect the pre-defined region of the
one or more frames for cropping the pre-defined portion. To avoid processing an
entire frame, the invention initially crops out the bottom 1/4th of the frame for
further processing and learning of the more accurate position of fields of the
scoreboard. The person skilled in the art will appreciate the significance of this
approach in saving processing time and improving accuracy. The invention as
disclosed in this specification provides a novel solution for cropping and
processing video frames in a more efficient manner.
Further, the system [200] comprises the detection unit [206] connected at least to
the processing unit [202]. The detection unit [206], configured to detect using one
or more text detection techniques, one or more region of interests (ROIs) in the
cropped pre-defined portion, wherein the detection of the one or more region of
61
interests (ROIs) in the cropped pre-defined portion is based on a standard
deviation in the pre-defined portion of the one or more frames. In one
implementation of the invention, the system [200] uses a standard deviation
method across the one or more frames of the media stream to identify a portion
of the one or more frames where a text is displayed more frequently and with
some variations. More specifically, the system [200] identifies the portion based
on an identification of the standard deviation in a pre-defined range, such as for
example between 5 to 20. Portions of the one or more frames that are having
almost zero standard deviation are not considered as a part of the scoreboard.
This is because, if the standard deviation is close to zero this is some static part of
the one or more frames with no variation so that means it won’t have any relevant
scoreboard information that changes as well. The system [200] then applies text
recognition in that portion of the one or more frames to identify the one or more
scoreboard templates. The use of the standard deviation technique for text
detection allows for efficient and accurate detection of ROIs in the one or more
frames, thereby improving the overall performance of the system [200].
The detection unit [206] is further configured to detect using one or more text
recognition techniques, one or more texts and a position associated with each of
the one or more texts from the one or more ROIs, wherein the position associated
with each of the one or more texts is detected further based on the one or more
text detection techniques. Further, the text recognition techniques can be
implemented either sequentially or in parallel to detect the texts and their
associated positions. The system [200] can use optical character recognition (OCR)
techniques or deep learning-based approaches to recognize the texts in the ROIs.
Once the texts and their positions are identified, the system [200] can extract
relevant information such as scores, player names, team names, and other
relevant details. The use of text recognition techniques allows for accurate
extraction of information from the video frames, providing valuable insights for
sports enthusiasts, broadcasters, and other stakeholders. The present invention
62
offers an efficient and accurate approach to extract relevant information from
video frames, improving the overall performance of the system [200].
The detection unit [206] is further configured to detect a set of template
coordinates based on the one or more texts, the position associated with each of
the one or more texts, one or more pre-defined texts and a target template mask
value associated with the one or more pre-defined texts. The one or more predefined texts comprise one or more fields associated with the live sports event.
The target template mask value associated with the one or more pre-defined texts
is determined based on a detection of one or more components of the scoreboard
in the one or more texts for a predefined threshold time, wherein the one or more
components are corresponding to the one or more pre-defined texts. In an
implementation of the present invention, the system [200] maintains an array of
same dimensions as an input frame for the one or more fields, hereinafter referred
to as a template mask. In the beginning, all values of the template mask are
initialised to 0. Further, once a position of each field is identified, the values of the
template mask associated with said each field is increment by 1, and reaches to
the target template mask value associated with each field corresponding to the
one or more pre-defined texts. E.g., for a team say Team1, template_mask
["Team1"]+=1. This is done for every field. Further, once the max value for the
template mask in each field crosses the predefined threshold i.e., reaches the
target template mask value, the detected max value for the template mask is a set
of coordinates of corresponding each field. Further, in another exemplary
implementation of the present invention, coordinates for an area having a
maximum value are detected as the set of template coordinates. Further, once all
coordinates for the set of template coordinates are detected, the present
invention marks a boundary for the scorecard area such that it contains all the
coordinates identified along with some extra buffer space.
63
Further, the processing unit [202] is further configured to automatically generate
the one or more scoreboard templates based at least on the set of template
coordinates, wherein each of the one or more scoreboard templates comprises at
least one of a position of the one or more components of the scoreboard of the
live sports event and an order of the one or more components. Further in an
implementation, the present invention receives a real time media stream of the
live sports event, wherein the real-time media stream comprises one or more new
frames. Further, the system [200] as disclosed by the present disclosure is
configured to predict, in the one or more new frames, a position of one or more
target components of a target scoreboard of the live sports event and an order of
the one or more target components based on the one or more scoreboard
templates. Further, to predict the position of one or more target components of
the target scoreboard of the live sports event and the order of the one or more
target components based on the one or more scoreboard templates, the present
invention as disclosed further determines, using the one or more text detection
techniques, one or more new regions of interest (ROIs) in the one or more new
frames. Further, the present invention as disclosed determines an overlap
between the set of template coordinates and the one or more existing ROIs, and
thereafter the present invention predicts, in the one or more new frames, the
position of the one or more target components of the target scoreboard of the live
sports event and the order of the one or more target components based on the
overlap. Further, the system [200] as disclosed herein detects at least one of one
or more players and one or more scoreboard-based events based on the one or
more target components. Then, the system [200], as disclosed, would deliver, via
the content delivery unit [208], one or more contextual contents to a set of
viewers of the live sports event based on at least one of the one or more players
and the one or more scoreboard-based events.
Also, in an implementation the processing unit [202] is further configured to
update the one or more scoreboard templates based on at least one of the
64
position of the one or more target components of the target scoreboard of the live
sports event and the order of the one or more target components.
Referring to Figure 5 an exemplary method [500] flow diagram [400], for
automatically generating one or more scoreboard templates for a live sports
event, in accordance with exemplary embodiments of the present invention is
shown. In an implementation the method [500] is performed by the system [200].
Further, in an implementation, the system [200] may be present in a server device
to implement the features of the present invention. Also, as shown in Figure 5, the
method [500] starts at step [502].
Next, at step [504], the method [500] comprises receiving, at a processing unit
[202], a media stream of the live sports event, wherein the media stream
comprises one or more frames.
For ease of understanding, let's consider an example wherein the processing unit
[202] receives a real time media stream of a cricket match played between two
teams named Dolphin vs. Whales. Further, one or more frames of the real time
media stream comprises one or more players of either or both teams i.e., Dolphin
and Whales and/or one or more brand logos.
At step [506], the method [500] comprises cropping, by the processing unit [202],
a pre-defined portion of the one or more frames. In one implementation of the
invention, the processing unit [202] of the present disclosure assumes that a
scoreboard may be present at a bottom of the one or more frames. It is to be
noted that in most cases, the scoreboard will be present at the bottom of the
frame(s), and the present invention may further process the frame to detect the
pre-defined region of the one or more frames for cropping the pre-defined
portion. To avoid processing an entire frame, the invention initially crops out the
bottom 1/4th of the frame for further processing and learning of the more
65
accurate position of fields of the scoreboard. The person skilled in the art will
appreciate the significance of this approach in saving processing time and
improving accuracy. The invention as disclosed in this specification provides a
novel solution for cropping and processing video frames in a more efficient
manner.
At step [508], the method [500] comprises detecting, by the detection unit [206]
using one or more text detection techniques, one or more region of interests
(ROIs) in the cropped pre-defined portion, wherein the detecting, by the detection
unit [206] using the one or more text detection techniques, the one or more region
of interests (ROIs) in the cropped pre-defined portion is based on a standard
deviation in the pre-defined portion of the one or more frames. In one
implementation of the invention, the method [500] as disclosed by the present
invention uses a standard deviation method across the one or more frames of the
media stream to identify a portion of the one or more frames where a text is
displayed more frequently and with some variations. More specifically, the system
[200] identifies the portion based on an identification of the standard deviation in
a pre-defined range, such as for example between 5 to 20. Portions of the one or
more frames that are having almost zero standard deviation are not considered as
a part of the scoreboard. This is because, if the standard deviation is close to zero
this is some static part of the one or more frames with no variation so that means
it won’t have any relevant scoreboard information that changes as well. The
method [500] as disclosed by the present invention then applies text recognition
in that portion of the one or more frames to identify the one or more scoreboard
templates. The use of the standard deviation technique for text detection allows
for efficient and accurate detection of ROIs in the one or more frames, thereby
improving the overall performance of the method [500] disclosed herein.
At step [510], the method [500] comprises detecting, by the detection unit [206]
using one or more text recognition techniques, one or more texts and a position
66
associated with each of the one or more texts from the one or more ROIs, wherein
the position associated with each of the one or more texts is detected further
based on the one or more text detection techniques. Further, the text recognition
techniques can be implemented either sequentially or in parallel to detect the
texts and their associated positions. The method [500] as disclosed by the present
invention may use optical character recognition (OCR) techniques or deep
learning-based approaches to recognize the texts in the ROIs. Once the texts and
their positions are identified, the method [500] as disclosed by the present
invention may extract relevant information such as scores, player names, team
names, and other relevant details. The use of text recognition techniques allows
for accurate extraction of information from the video frames, providing valuable
insights for sports enthusiasts, broadcasters, and other stakeholders. The present
invention offers an efficient and accurate approach to extract relevant information
from video frames, improving the overall performance of the method [500] as
disclosed herein.
At step [512], the method [500] comprises detecting, by the detection unit [206],
a set of template coordinates based on the one or more texts, the position
associated with each of the one or more texts, one or more pre-defined texts and
a target template mask value associated with the one or more pre-defined texts.
The one or more pre-defined texts comprise one or more fields associated with
the live sports event. The target template mask value associated with the one or
more pre-defined texts is determined based on a detection of one or more
components of the scoreboard in the one or more texts for a predefined threshold
time, wherein the one or more components are corresponding to the one or more
pre-defined texts. In an implementation of the present invention, the method
[500] maintains an array of same dimensions as an input frame for the one or more
fields, hereinafter referred to as a template mask. In the beginning, all values of
the template mask are 66nitialized to 0. Further, once a position of each field is
identified, the values of the template mask associated with said each field is
67
increment by 1, and reaches to the target template mask value associated with
each field corresponding to the one or more pre-defined texts. Eg, for a team say
Team1, template_mask [“Team1”]+=1. This is done for every field. Further, once
the max value for the template mask in each field crosses the predefined
threshold, i.e., reaches the target template mask value the detected max value for
the template mask is a set of coordinates of corresponding each field. Further, in
another exemplary implementation of the present invention, coordinates for an
area having a maximum value are detected as the set of template coordinates.
Further, once all coordinates for the sets of template coordinates are detected,
the present invention marks the boundary for the scorecard area such that it
contains all the coordinates identified along with some extra buffer space.
At step [514], the method [500] comprises automatically generating, by the
processing unit [202], the one or more scoreboard templates based at least on the
set of template coordinates wherein each of the one or more scoreboard
templates comprises at least one of a position of the one or more components of
the scoreboard of the live sports event and an order of the one or more
components. Further, in an implementation the method [500] further comprises
receiving, at the processing unit [202], a real time media stream of the live sports
event, wherein the real-time media stream comprises one or more new frames.
Further, the method [500] comprises predicting, by the processing unit [202] in
the one or more new frames, the position of one or more target components of a
target scoreboard of the live sports event and an order of the one or more target
components based on the one or more scoreboard templates, Further, the
predicting, by the processing unit [202] in the one or more new frames, the
position of the one or more target components of the target scoreboard of the live
sports event and the order of the one or more target components further
comprises determining, by the processing unit [202] using the one or more text
detection techniques, one or more new region of interests (ROIs) in the one or
more new frames. Further, the method [500] comprises determining, by the
68
processing unit [202], an overlap between the set of template coordinates and the
one or more ROIs, and thereafter the method [500] comprises predicting, by the
processing unit [202] in the one or more new frames, the position of the one or
more target components of the target scoreboard of the live sports event and the
order of the one or more target components based on the overlap. Further, the
method [500] as disclosed by the present invention comprises detecting, by the
processing unit [202], at least one of one or more players and one or more
scoreboard-based events based on the one or more target components.
Thereafter, the method [500], as disclosed comprise delivering, by the processing
unit [202] via the content delivery unit [208], one or more contextual contents to
a set of viewers of the live sports event based on at least one of the one or more
players and the one or more scoreboard-based events.
Also, in an implementation of the method [500], the method further comprises
updating by the processing unit [202] the one or more scoreboard templates
based on at least one of the position of the one or more target components of the
target scoreboard of the live sports event and the order of the one or more target
components.
Another aspect of the present disclosure may relate to a user device for
automatically generating one or more scoreboard templates for a live sports
event. The user device further comprises a system [200], wherein the system [200]
comprises a processing unit [202], configured to receive a media stream of the live
sports event, wherein the media stream comprises one or more frames. The
processing unit [202] is further configured to crop a pre-defined portion of the one
or more frames. Further, the system [200] comprises a detection unit [206],
configured to detect using one or more text detection techniques, one or more
region of interests (ROIs) in the cropped pre-defined portion. Further, the
detection unit [206] is configured to detect using one or more text recognition
techniques, one or more texts and a position associated with each of the one or
69
more texts from the one or more ROIs. The detection unit [206] is further
configured to detect a set of template coordinates based on the one or more texts,
the position associated with each of the one or more texts, one or more predefined texts and a target template mask value associated with the one or more
pre-defined texts. The processing unit [202] is further configured to automatically
generate, the one or more scoreboard templates based at least on the set of
template coordinates wherein each of the one or more scoreboard templates
comprises at least one of a position of one or more components of a scoreboard
of the live sports event and an order of the one or more components.
Yet another aspect of the present disclosure may relate to a non-transitory
computer readable storage medium storing instructions for automatically
generating one or more scoreboard templates for a live sports event, the storage
medium comprising executable code which, when executed by a processor, causes
the processor to: receive a media stream of the live sports event, wherein the
media stream comprises one or more frames; crop a pre-defined portion of the
one or more frames; enable a detection unit [206] to: 1) detect using one or more
text detection techniques, one or more region of interests (ROIs) in the cropped
pre-defined portion, 2) detect using one or more text recognition techniques, one
or more texts and a position associated with each of the one or more texts from
the one or more ROIs, wherein the position associated with each of the one or
more texts is detected further based on the one or more text detection
techniques, and 3) detect a set of template coordinates based on the one or more
texts, the position associated with each of the one or more texts, one or more predefined texts and a target template mask value associated with the one or more
pre-defined texts; and automatically generate, the one or more scoreboard
templates based at least on the set of template coordinates wherein each of the
one or more scoreboard templates comprises at least one of a position of one or
more components of a scoreboard of the live sports event and an order of the one
or more components.
70
The present invention represents a significant improvement over the prior known
solution that displays contents targeted to a specific audience based solely on the
viewer's search or browsing history, location, language, and other relevant data
collected from the viewer's device. Unlike these prior solutions, the present
invention provides a more sophisticated and effective approach to delivering
contextual contents during live sports events. Specifically, the present solution
identifies and analyzes the player faces, brand logos, and critical events from the
live media stream of the live sports event, and thereafter detects the contextual
content for delivering based on a content display priority. This approach allows
advertisers to better target their contents to specific events or individuals during
the live sports event, thereby increasing the relevance and effectiveness of the
advertising campaign. Further, the present invention uses a novel approach to
detect and extract real-time match/live sports event information through the use
of at least one of advanced face recognition, brand logo detection, multilingual
text recognition techniques and scoreboard recognition techniques. By analyzing
the live media stream of the sports event in real-time, the invention is able to
detect critical events as they occur, and extract the relevant context associated
with these events. This enables the delivering of contextual contents that are
highly targeted and relevant to the viewer based on the real-time information of
a critical event of the live sports event and the context associated with it. The
patent specification provides further details regarding the technical
implementation of these advanced techniques, as well as their potential
applications in the field of contextual advertising. Overall, the present invention
represents a significant advancement in the ability to detect and extract real-time
information from live sports events, and greatly enhance the effectiveness of
contextual advertising during these events.
Therefore, the present invention provides a novel solution for delivering one or
more contextual contentsto a set of viewers of a live sports event, which increases
71
the chances of engagement and leads to a positive impact on revenue. The
invention leverages the use of face recognition, brand logo detection, and
scoreboard recognition techniques to determine and deliver relevant contextual
contents during live sports events. By analyzing the live media stream of the sports
event in real-time, the invention is able to identify specific players, teams, and
brands that are present, as well as detect critical events as they occur. This
information is then used to determine the most relevant contextual contents to
deliver, based on factors such as the real-time context of the event, the interests
of the viewer, and the priorities of the advertisers. The use of these advanced
techniques allows for highly targeted and effective advertising during live sports
events, as well as the ability to adapt to changing circumstances in real-time.
Further, the present invention provides a technical solution that not only
enhances viewers experience but also enhance the impact of contextual
advertising during live sports events, which in turn can lead to increased revenue
for brands and greater user engagement with the contents. By delivering highly
relevant and targeted contextual contents based on real-time information from
the live sports event, the invention enables advertisers to better capture the
attention and interest of viewers. This can result in increased customer
conversions, and greater brand awareness and loyalty. Additionally, the use of
advanced face recognition, brand logo detection, and scoreboard recognition
techniques allows for a more seamless and non-intrusive integration of contextual
contents into the live sports event, further enhancing the overall viewing
experience for the viewer. Overall, the present invention represents a significant
advancement in the ability to leverage real-time information from live sports
events to enhance the impact of contextual advertising, increase revenue for
brands, and improve user engagement with contents.
72
While considerable emphasis has been placed herein on the disclosed
embodiments, it will be appreciated that many embodiments can be made and
that many changes can be made to the embodiments without departing from the
principles of the present disclosure. These and other changes in the embodiments
of the present disclosure will be apparent to those skilled in the art, whereby it is
to be understood that the foregoing descriptive matter to be implemented is
illustrative and non-limiting.

WE CLAIM:
1. A method [500] for automatically generating one or more scoreboard
templates for a live sports event, the method [500] comprising:
- receiving, at a processing unit [202], a media stream of the live sports
event, wherein the media stream comprises one or more frames;
- cropping, by the processing unit [202], a pre-defined portion of the one
or more frames;
- detecting, by a detection unit [206] using one or more text detection
techniques, one or more region of interests (ROIs) in the cropped predefined portion;
- detecting, by the detection unit [206] using one or more text
recognition techniques, one or more texts and a position associated
with each of the one or more texts from the one or more ROIs;
- detecting, by the detection unit [206], a set of template coordinates
based on the one or more texts, the position associated with each of
the one or more texts, one or more pre-defined texts and a target
template mask value associated with the one or more pre-defined
texts; and
- automatically generating, by the processing unit [202], the one or more
scoreboard templates based at least on the set of template coordinates
wherein each of the one or more scoreboard templates comprises at
least one of a position of one or more components of a scoreboard of
the live sports event and an order of the one or more components.
2. The method [500] as claimed in claim 1 wherein the detecting, by the
detection unit [206] using the one or more text detection techniques, the
one or more region of interests (ROIs) in the cropped pre-defined portion
is based on a standard deviation in the pre-defined portion of the one or
more frames.
74
3. The method [500] as claimed in claim 1, wherein the target template mask
value associated with the one or more pre-defined texts is determined
based on a detection of the one or more components in the one or more
texts for a predefined threshold time.
4. The method [500] as claimed in claim 1, wherein the one or more predefined texts comprise one or more fields associated with the live sports
event.
5. The method [500] as claimed in claim 1, the method [500] further
comprises:
- receiving, at the processing unit [202], a real time media stream of the
live sports event, wherein the real-time media stream comprises one
or more new frames,
- predicting, by the processing unit [202] in the one or more new frames,
a position of one or more target components of a target scoreboard of
the live sports event and an order of the one or more target
components based on the one or more scoreboard templates,
- detecting, by the processing unit [202], at least one of one or more
players and one or more scoreboard-based events based on the one or
more target components, and
- delivering, by the processing unit [202] via a content delivery unit, one
or more contextual contents to a set of viewers of the live sports event
based on at least one of the one or more players and the one or more
scoreboard-based events.
6. The method [500] as claimed in claim 5, wherein the predicting, by the
processing unit [202] in the one or more new frames, the position of the
one or more target components of the target scoreboard of the live sports
event and the order of the one or more target components further
comprises:
75
- determining, by the processing unit [202] using the one or more text
detection techniques, one or more new region of interests (ROIs) in the
one or more new frames,
- determining, by the processing unit [202], an overlap between the set
of template coordinates and the one or more ROIs, and
- predicting, by the processing unit [202] in the one or more new frames,
the position of the one or more target components of the target
scoreboard of the live sports event and the order of the one or more
target components based on the overlap.
7. The method as claimed in claim 5, the method further comprises updating
by the processing unit [202] the one or more scoreboard templates based
on at least one of the position of the one or more target components of
the target scoreboard of the live sports event and the order of the one or
more target components.
8. A system [200] for automatically generating one or more scoreboard
templates for a live sports event, the system [200] comprising:
- a processing unit [202], configured to:
receive a media stream of the live sports event, wherein the media
stream comprises one or more frames, and
crop a pre-defined portion of the one or more frames;
- a detection unit [206], configured to:
detect using one or more text detection techniques, one or more
region of interests (ROIs) in the cropped pre-defined portion,
detect using one or more text recognition techniques, one or more
texts and a position associated with each of the one or more texts
from the one or more ROIs, and
detect a set of template coordinates based on the one or more
texts, the position associated with each of the one or more texts,
one or more pre-defined texts and a target template mask value
76
associated with the one or more pre-defined texts, wherein the
processing unit [202] is further configured to:
automatically generate, the one or more scoreboard
templates based at least on the set of template coordinates
wherein each of the one or more scoreboard templates
comprises at least one of a position of one or more
components of a scoreboard of the live sports event and an
order of the one or more components.
9. The system [200] as claimed in claim 8 wherein the detection of the one or
more region of interests (ROIs) in the cropped pre-defined portion is based
on a standard deviation in the pre-defined portion of the one or more
frames.
10. The system [200] as claimed in claim 8, wherein the target template mask
value associated with the one or more pre-defined texts is determined
based on a detection of the one or more components in the one or more
texts for a predefined threshold time.
11. The system [200] as claimed in claim 8, wherein the one or more predefined texts comprise one or more fields associated with the live sports
event.
12. The system [200] as claimed in claim 8, wherein the processing unit [202]
is further configured to:
- receive a real time media stream of the live sports event, wherein the
real-time media stream comprises one or more new frames,
- predict in the one or more new frames, a position of one or more target
components of a target scoreboard of the live sports event and an
order of the one or more target components based on the one or more
scoreboard templates,
- detect at least one of one or more players and one or more scoreboardbased events based on the one or more target components, and
77
- deliver via a content delivery unit, one or more contextual contents to
a set of viewers of the live sports event based on at least one of the
one or more players and the one or more scoreboard-based events.
13. The system [200] as claimed in claim 12, wherein to predict in the one or
more new frames, the position of the one or more target components of
the target scoreboard of the live sports event and the order of the one or
more target components, the processing unit [202] is further configured
to:
- determine using the one or more text detection techniques, one or
more new region of interests (ROIs) in the one or more new frames,
- determine an overlap between the set of template coordinates and the
one or more ROIs, and
- predict in the one or more new frames, the position of the one or more
target components of the target scoreboard of the live sports event
and the order of the one or more target components based on the
overlap.
14. The system as claimed in claim 12, wherein the processing unit [202] is
further configured to update the one or more scoreboard templates based
on at least one of the position of the one or more target components of
the target scoreboard of the live sports event and the order of the one or
more target components.
15. A user device for automatically generating one or more scoreboard
templates for a live sports event, the user device comprising:
- a system, wherein the system comprises:
a processing unit, configured to:
receive a media stream of the live sports event, wherein the
media stream comprises one or more frames, and
crop a pre-defined portion of the one or more frames;
a detection unit, configured to:
78
detect using one or more text detection techniques, one or
more region of interests (ROIs) in the cropped pre-defined
portion,
detect using one or more text recognition techniques, one
or more texts and a position associated with each of the one
or more texts from the one or more ROIs, and
detect a set of template coordinates based on the one or
more texts, the position associated with each of the one or
more texts, one or more pre-defined texts and a target
template mask value associated with the one or more predefined texts, wherein the processing unit is further
configured to:
automatically generate, the one or more scoreboard
templates based at least on the set of template
coordinates wherein each of the one or more
scoreboard templates comprises at least one of a
position of one or more components of a scoreboard
of the live sports event and an order of the one or
more components.