Description:FORM – 2

THE PATENTS ACT, 1970
(39 OF 1970)

COMPLETE SPECIFICATION
(See Section 10; Rule 13)

“AN ACCIDENT DETECTION SYSTEM”

KasperTech
Having address at
Ojha Bhawan, Civil Lines,
Near Rail Manoranjan Kendra,
Bikaner, Rajasthan India-334001

The following specification particularly describes the nature of the invention and the manner in which it is performed:

FIELD OF INVENTION:
The present invention relates to an accident detection system. In particular the present invention provides an automated system that detects an accident and informs the authorities about the accident, thus enabling the authorities to provide immediate help to accident victims.

BACKGROUND OF INVENTION:
Road and traffic accidents are one the leading cause of death in India. And India ranks first in the number of road accident deaths across the 199 countries. Every year more than 2 lakhs of cases are registered for road accident and out of these nearly 40% victims submits to the injuries that do not receive immediate medical attention within the golden hour, that also leads to the victim death in critical situations. The first sixty minutes after the accident is called the “golden hour.” The golden hour play an important role in deciding the morbidity and the mortality of the patient. If the patient is given the right medical treatment in this golden hour the survival chance increases significantly, and the effects of trauma substantially decreases.
There are several organizations, along with government services that provide the ambulance under the cases of emergency where, the victim requires medical attention. However in majority of these cases such organizations fails to receive the information about the accidents and the accident locations, thus fail to provide medical attention to the victims in the golden hour.
Conventionally there are various accident report systems available, that utilize electronic system/software used by employees to report accidents in real-time. These systems help manage accident reports, and send the medical teams to the accident location. Also, these systems do not send any manually verifiable data with which a false alarm could be detected. This may often lead to authorities taking actions when it was actually not required.
Therefore, there is a need of an accident detection and reporting system that accurately detects and reports the accident, thus providing immediate help to accident victims that saves life of many victims.

PRIOR ART AND ITS DISADVANTAGES:

• US patent US10592990B1 relates to systems, apparatuses, computer-implemented methods, and computer-readable media for determining that vehicle accidents have occurred, collecting and analyzing accident characteristics and other related data, and providing customized accident recovery services. A mobile computing device within a moving vehicle may be configured to detect that an accident involving the vehicle has occurred. The mobile computing device may detect an accident using movement and location sensors on the device, or may establish communication with one or more vehicle-based devices (e.g., a vehicle control computer, on-board diagnostic system, telematics device, etc.) to receive accident indication data from the vehicle. After determining that an accident has occurred, the mobile device and/or vehicle-based.
However the prior art provides a device that is installed in vehicles by an individual, thus the risk of data getting deleted, false alarms, and the device getting damaged at the time of the accident increases substantially. Additionally the said device is costly that acts as a barrier in the installation of the device in vehicles.

? US patent US9758120B2 relates to an automatic cloud based notification system and method for vehicle accidents. The method of sending an alarm on the occurrence of a vehicle accident, comprising the steps of: a) a personal mobile device in a vehicle monitoring readings from sensors of the personal mobile device; b) the personal mobile device processing the readings from the sensors to detect a condition which indicates a possibility that the vehicle is about to have an accident; c) when the condition which indicates a possibility that the vehicle is about to have an accident is detected, the personal mobile device transmitting a provisional alarm to a remote server; d) the remote server receiving the provisional alarm from the personal mobile device; e) the remote server starting a timer upon receipt of the provisional alarm from the personal mobile device; f) if the remote server does not receive a cancellation of the provisional alarm from the personal mobile device before the timer reaches an end of a predefined time, the remote server forwarding an alarm to an emergency service center.
However the prior art relates to an accident awareness application that is installed individually on mobile devices, thus the prior art substantially increases the risk of false alarm or misreading of the incident to evaluate the seriousness of the accident. In addition the prior art fails to provide a device that verifies the generated alarm to evaluate the seriousness of the incident.
• US patent US9818239B2 relates to an accident notification system for a vehicle where, when a vehicle causes an accident, such an accident can automatically be notified to a related organ. The method including: receiving a movement dataset collected at least at one of a location sensor and a motion sensor arranged within the vehicle, during a time period of movement of the vehicle, extracting a set of movement features associated with at least one of a position, a velocity, and an acceleration characterizing the movement of the vehicle during the time period, detecting a vehicular accident event from processing the set of movement features with an accident detection model, and in response to detecting the vehicular accident event, automatically initiating an accident response action.
However the prior art fails to provide a device that accurately detects and reports an accident, and eliminates the risk of false alarms. Additionally the device disclosed in the prior art is installed individually in a vehicle that substantially increases the risk of the device getting damaged during the accident.

DISADVANTAGES OF THE PRIOR ART:

The conventional accident detection systems suffers from all or at least one of the following disadvantages:
• Most or all of the prior art fails to provide an accident detection and alerting system.
• Most or all of the prior art fails to provide an automated system that accurately detects an accident.
• Most or all of the prior art fails to provide an accident detection system that lets the authorities gauge the severity of the accident.
• Most or all of the prior art fails to provide an accident detection system that sends an alert to the hospitals and the police station that are nearest to the accident site regarding the accident details.
• Most or all of the prior art fails to provide an accident detection system that stores the data about the accident in the database for future reference and investigation.
• Most or all of the prior arts fail to store the accident data, thus fails to study and identify the accident-prone areas.
• Most or all of the prior art fails to provide a system, that marks accident status to tackle the accidents efficiently.
Hence there is unmet need for an invention that overcomes the problems mentioned in above prior arts.

OBJECTS OF INVENTION:
• The primary object of the present invention is to provide an accident detection system.
• Another object of the present invention is to provide an accident detection system that is completely automated and works without any human intervention.
• Yet another object of the present invention is to provide an accident detection system that sends an alert to the hospitals and the nearest police station regarding the date, time and location of the accident to avail with the instantaneous action.
• Yet another object of the present invention is to provide an accident detection system that accurately detects an accident and does not send the reports in case of no accident.
• Yet another object of the present invention is to provide an accident detection system that captures and send multiple photographs and a video of the accident to check the severity of the accident.
• Yet another object of the present invention is to provide an accident detection that stores the data about the accident in the database for future reference and investigation.

BRIEF DESCRIPTION OF DRAWING:
Fig. 1 Shows the process of detection of an accident with the help of the present system (P)

Meaning of the reference numerals of the component parts of present system for the detection of an accident

P : Present system for the detection of accident

Cm : Capturing means
Pm : Processing means
AD : Analyzing and detection means
Bm : Buffer means
AB : Accident buffer means
Sm : Storage means
Dm : Displaying means
Am : Alerting means

SUMMARY OF THE PRESENT INVENTION:
The present invention provides an accident detection system. In particular the present invention provides an automated accident detection and reporting system that detects an accident and sends an alert to the nearby hospital and the police station about the details of the accident.
The embodiment of the present invention relates to an automated accident detection system that detects an accident and generates an alert. The present invention captures the pictures of the accident, processes the pictures of the accident and detects the occurrence of the accident, the data is subsequently stored in the storage means. The detected accident information is made available to the authorities through a central interface. This information is also sent to authorities, such as hospitals and police stations, as an alert containing crucial details about the accident, such as the date, time, location, image, and video.
The present system comprises:
A Capturing means (Cm),
A Processing means (Pm),
A Buffer means (Bm),
An Accident buffer means (AB),
An Analyzing and detection means (AD),
A Storage means (Sm),
A Displaying means (Dm),
An Alerting means (Am),

Said capturing means (Cm) is configured to monitor and capture the occurrence of an accident and transfer the captured feed to said processing means (Pm). The processing means (Pm) is connected to the capturing means (Cm) configured to process, sort the state of accident flag and read the captured feed, resize each frame and send resized accident flag means marked as 0 to Buffer (Bm) and the accident flag marked as 1 to the analyzing and detection means (AD). The analyzing and detection means (AD) connected to said processing means (Pm) is configured to detect the occurrence of an accident and classify each frame into “accident” and “no accident”; said accident buffer means (AB) connected to analyzing and detection means (AD) is configured to store the incremental accident count marked as 1 -7. The accident buffer means (AB) and said buffer means (Bm) comprises of accident picture frames configured to create video of accident. said storage means (Sm) connected to analyzing and detection means (AD) is configured to store data of the accident occurrence and transfer said data to the registered authorities via alerting means (Am) connected to the storage means (Sm); said displaying means (Dm) connected to the alerting means (Am) is configured to display the accident occurrence video simultaneously.

DETAILED DESCRIPTION OF THE PRESENT INVENTION:

The embodiment of the present invention is to provide an accident detection system. The present system (P) detects and alerts the police and hospital authorities about the date, time and location, of the accident, to avail the victim with quick medical attention. Further the present invention is an automated system that detects the accident through the pictures of the accident taken in real time, and shares a video and picture of the accident with the authorities for them to sense the severity of the accident and eliminating the chances of false alerts.

Referring to Fig 1. Present system (S) comprises of:
• Capturing means (Cm)
• Processing means (Pm)
• Analyzing and detection means (AD)
• Buffer means (Bm)
• Accident buffer means (AB)
• Storage means (Sm)
• Displaying means (Dm)
• Alerting means (Am)

Said capturing means (Cm) are the live CCTV cameras that are present on the roads for the live monitoring of roads and traffic. The live capturing of an accident in the form of picture frames helps in analyzing the detection of the accident. Therefore the present system (P) confirms the occurrence of the accident before generating the alerts, thus substantially reducing the risk of generating false alerts. The captured feed is sent to the processing means (Pm). Said processing means (Pm) processes and reads the live feed in the form of frame and resizes each frame before sending them for detection of the accident. The “accident flag”, “no accident flag”, “accident count”, “no accident count” are all set to 0 when the present system (P) starts. Said Processing Means (Pm) checks for the state of accident flag. If “accident flag” is 0, it stores the frame in the Buffer means (Bm) and thereafter sends the picture frame to the analyzing and detecting means (AD). If the accident flag is 1 the frame is sent directly to the analyzing and detecting means (AD) for the detection of accident.
The analyzing and detection means (AD) is the central part of the present system (P) that detects the occurrence of an accident. Said analyzing and detection means (AD) receives the accident pictures in the form of frames and observes each picture frame to detect the occurrence of the accident. After observing the picture frames, the pictures frames are classified either as “accident” or “no accident”. When an accident is detected said analyzing and detection means (AD) starts counting the frames for occurrence of the accident. In the present system (P) if the analyzing and detection means (AD) detects an accident, “accident count” is incremented, otherwise “no accident count” is incremented. With each picture frame that has captured an accident the accident count is incremented and if the accident count is greater than 7, the occurrence of accident is confirmed and the accident frame is stored in the said accident buffer means (AB). After this, the accident flag is marked 1. In the situations when the accident flag is 1 and the “no accident” count is greater than 4, the accident is considered to be over. The picture frames from the accident buffer means (AB) and buffer means (Bm) are combined to form a video of the accident, and the central image from the frames is taken as the accident image. Simultaneously this data is sent to the storage means (Sm). Whereas in the situations when the accident flag is 0 and the “no accident” count is greater than 4, such situation is considered a false alarm and no alert is generated. And the data related to the accident is deleted from the buffer means (Bm) and the “accident count”, “no accident” count and “accident flag” are set to 0.
Said accident buffer means (AB) is a temporary storage means that stores the pictures and the data related to the accident. After the confirmation of the occurrence of an accident, the data from the Accident Buffer means (AB) and the buffer means (Bm) is combined to form a video of the accident and the central image of combined buffering means is taken and considered as the picture of accident. This video and picture are sent to the Storage means (Sm) along with other crucial details like the time, location and the date of the accident are stored in the storage means (Sm) for future reference and investigation procedure. Said storage means (Sm) sends the accident related data, picture and the video of the accident to the registered authorities through a message.
Simultaneously the accident data and the video is sent to the police and hospital authorities called the central authorities that are located near to the accident site by the Alerting Means (Am). The sent data and the video is displayed through the displaying means (Dm) to the central authorities to assess the severity of the accident to make a decision on whether or not any help is required to be sent to the victim.
Said alerting means (Am) is configured to generate and send an alert to avail the victims with instant medical attention and investigation of the accident site. Accident status in the present device is ‘pending’ initially which is later updated as ‘resolved’ or ‘help on the way’ accordingly.
Under the situation, when the occurred accident is not severe, the authorities sends a signal to the storage means (Sm) to count the accident as “no accident”. After the signal is received by the storage means (Sm) the video and the photographs related to the incident are deleted automatically said storage means (Sm). Therefore the present invention provides an accident detection and alerting system that verifies the occurrence of the accident multiple times before sending help and eliminates the risk of false alarms.

WORKING OF THE INVENTION:
The present invention provides an automated accident detection and alerting system. Said accident detection system substantially reduces the time required to provide medical attention to the accident victim. Further the present invention captures the picture and video of the accident, making it easier to investigate the accident site.
Referring to fig. 1
1) When the present system (P) starts, “the accident flag”, “no accident flag”, “accident count”, “no accident count” are all set to 0. The capturing means (Cm) of the present system (P) captures the pictures of the accident in the form of live feed. Thus the present invention records and captures the accident in real-time and eliminates the chances of an action being taken on false alerts. Said capturing means captures series of frames continuously and substantially reduces the risk of false detection of an accident.
2) The pictures captured from the capturing means (Cm) are thereafter sent to the processing means (Pm). Said processing means (Pm) processes and reads the live feed in form of frame and resizes them. Said processing means (Pm) checks the accident flag of the incident, if the accident flag is 0, the picture frames are sent to the buffer means (Bm) first and then to the analyzing and detection means (AD). However if the accident flag is 1, the data of the accident is sent directly to the analyzing and detection means (AD).

3) The processed picture frames are sent to the analyzing and detection means (AD) that detects the occurrence of the accident, by going through each picture frame. After receiving the picture frames, the analyzing and detection means (AD) reads each picture frame thoroughly and classifies the incidents into “accident” and “no accident”. If an accident is detected “accident count” is incremented, and if there is no accident “no accident count” is incremented.
Thus said analyzing and detection means (AD) starts counting the occurrence of the accident in reference to the picture frame. After each picture frame that has captured the occurrence of the accident, the accident count is incremented by 1. Such that the first picture frame that has captured the accident is given a count 1, the next picture frame is given a count 2 and so on. If the accident count is 7 or greater than 7, the occurrence of accident is confirmed and the accident frame is stored in the said accident buffer means (AB) that temporarily retains the accident data. And the accident flag is marked 1.
In the situation when the accident flag is 1 and the “no accident” count is greater than 4, the accident is considered as over and the picture frames from accident buffer means (AB) and buffer means (Bm) are combined together to form a video of the accident. The central picture from the combined buffer is taken as the picture of the accident. The video and image along with other crucial details like date, time and location of the accident are sent to the Storage means (Sm)
In the situation when the accident flag is 0 and no accident count is greater than 4, it is considered to be a false alarm and the data related to the accident is deleted from the buffer means (Bm) and the “accident count”, “no accident” count and “accident flag” are set to 0.
4) Simultaneously said accident buffer means (AB) temporarily stores the accident related data and creates a video of the accident by combining the frames from accident buffer means (AB) and local buffer means (Bm) and sends the video, image and the data to the storage means (Sm). Also the central image the combined buffering means is considered as the picture of accident.
5) Said storage means (Sm) stores the video and the crucial data of the accident that includes date, time and the location of the accident for investigation and future reference.
6) Said storage means (Sm) sends the accident data and the video to the registered authorities through a message via Alerting means (Am). Simultaneously the data is sent to the police and hospital authorities that are located near the accident site. Said hospital and police authorities are registered during the time of installation of the system.

7) The data is displayed to the authorities through the displaying means (Dm), to judge the severity of the accident and to decide whether or not any help is required to be sent to the victim.
8) If the accident is severe said registered authorities send a signal through the alerting means (Am) to generate an alert to avail the victims with instant medical attention and investigation, also once an accident is detected, its status is initialized to ‘Pending’. The authorities then update the accident status to ‘Help on the way’ or ‘Resolved’ accordingly.
• In an embodiment of the present invention said capturing means (Cm) captures live video feed of the roads and sends the video feed to the processing means (Pm). Said processing means (Pm) processes and reads the feed in form of frames, resizes each frame and sends the frames to the analyzing and detection means (AD). Said analyzing and detection means (AD) detects the occurrence of the accident and classifies the incident as ‘accident’ and ‘no accident’. In the situations when accident is detected, the “accident count” is incremented. If no accident is detected, the “no accident count” is incremented. If the “accident count” is greater than 7 the accident flag is marked as 1 and the picture frames related to the accident are stored in the accident buffer means (AB). After a certain point when the accident flag is still 1 and the “no accident count” is greater than 4 then the accident is considered to be over and the picture frames are stored in the buffer means (Bm). In such situations the picture frames from the accident buffer means (AB) and the buffer means (Bm) are combined to form an accident video. The central image of this video is taken as the picture of the accident. This picture and video along with other crucial details like date, time and location are sent to the Storage means (SM). The accident data is then sent to the registered authorities as an alert through a message by the Alerting Means (Am). Simultaneously the Alerting Means (Am) sends the data to the police and hospital authorities that are located near the accident site. The accident data is also displayed to the authorities through displaying means (Dm).
• In another embodiment of the present invention said capturing means (Cm) captures live pictures of and accident and sends the pictures to the processing means (Pm). Said processing means (Pm) processes and reads the pictures in form of frames and sends the pictures to the analyzing and detection means (AD). Said analyzing and detection means (AD) detects the occurrence of the accident and classifies the incident as ‘accident’ and ‘no accident’. In situations where accident flag is 0 and ‘no accident count’ is greater than 4, the accident is considered as a false alarm and all the data from the local buffer means (Bm), and the accident buffer means (AB) is deleted and the “accident count”, “no accident” count and “accident flag” are set to 0.

ADVANTAGES OF INVENTION:

• The present invention provides an automated accident detection and alerting system.
• The present invention provides an automated accident detection system that sends an alert to the hospitals and the police station, regarding the time, date and location of the accident for instant medical attention and investigation.
• The present invention captures the video of the accident from different frames to accurately detect the occurrence and the severity of the accident, and does not send the alert report in case of the accident is not sever.
• The present invention stores the data of the accident in the database, for future reference.
• By utilizing the stored accident data the present invention identifies accident prone areas to take appropriate safety measures.
• The present invention lets the authorities constantly updates the status of the accident to manage the accidents efficiently.
• Most or all of the prior art fails to provide a system, that marks accident status to tackle the accidents efficiently.

Claims:We claim:

1. An accident detection system (P), wherein said system comprises:
A Capturing means (Cm),
A Processing means (Pm),
A Buffer means (Bm),
An Accident buffer means (AB),
An Analyzing and detection means (AD),
A Storage means (Sm),
A Displaying means (Dm),
An Alerting means (Am),

Wherein;

Said capturing means (Cm) is configured to monitor and capture the occurrence of an accident and transfer the captured feed to said processing means (Pm); said processing means (Pm) connected to the capturing means (Cm) is configured to process, sort the state of accident flag and read the captured feed, resize each frame and send resized accident flag means marked as 0 to Buffer (Bm) and the accident flag marked as 1 to the analyzing and detection means (AD); said analyzing and detection means (AD) connected to said processing means (Pm) is configured to detect the occurrence of an accident and classify each frame into “accident” and “no accident”; said accident buffer means (AB) connected to analyzing and detection means (AD) is configured to store the incremental accident count marked as 1 -7; said accident buffer means (AB) and said buffer means (Bm) comprises of accident picture frames configured to create video of accident; said storage means (Sm) connected to analyzing and detection means (AD) is configured to store data of the accident occurrence and transfer said data to the registered authorities via alerting means (Am) connected to the storage means (Sm); said displaying means (Dm) connected to the alerting means (Am) is configured to display the accident occurrence video simultaneously;

Wherein the present system (P) utilizes existing CCTV cameras installed on roads as the Capturing Means (Cm).