Claims:WE CLAIM
1. A system (100) for monitoring highway assets, the system (100) comprising:
a video capturing unit configured to record a video;
a comparison unit (200) configured to detect fixed assets by comparing data from a master video with the video recorded by the video capturing unit;
an identification unit (300) configured to identify random assets in the captured video; and
a GPS unit configured to provide location co-ordinates for the fixed assets and the random assets, such that the system (100) uses this information to identify and track the highway assets.
2. The system (100) as claimed in claim 1, wherein a user provides the master video.
3. The system (100) as claimed in claim 2, wherein the comparison unit (200) extracts data from the master video.
4. The system (100) as claimed in claim 1, wherein the video capturing unit is captured while patrolling to monitor highways.
5. The system (100) as claimed in claim 1, wherein the GPS unit can be based on any positioning system.
6. The system (100) as claimed in claim 1, wherein the system (100) monitors the highway assets by detecting the fixed assets and identifying the random assets.
, Description:GPU BASED EMBEDDED VISION SOLUTION FOR HIGHWAYS ASSETS MANAGEMENT

FIELD OF INVENTION
[001] The present invention provides a system for identifying and analysing highways assets. More particularly, the present invention relates to a GPU based embedded computer vision technology for highway assets management.

BACKGROUND OF THE INVENTION
[002] The background description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is prior art or relevant to the presently claimed invention, or that any publication specifically or implicitly referenced is prior art.
[003] Computer Vision, a sub branch of Machine Learning has experienced a revolution in working with imagery data for the task of object classification and detection.
[004] The existing object detection choices include CNN’s, faster R-CNN, SSD, Retinanet, YOLO. Among the existing object detection choices YOLO is the promising one that require comparatively low computational power but still achieve good enough accuracy and it’s refreshingly simple to train and test.
[005] ‘YOLO’, which stands for ‘You Only Look Once’, is a deep learning-based object detection algorithm developed by Joseph Redmon and Ali Farhadi at the University of Washington in 2016. The rationale behind calling the system YOLO is that rather than pass in multiple sub images of potential objects, you only passed in the whole image to the deep learning system once. Then, you would get all the bounding boxes as well as the object category classifications in one go. This is the fundamental design decision of YOLO and is what makes it a refreshing new perspective on the task of object detection.
[006] Also, YoloV4 which is available in Public Domain developed by Alexey Bochkovskiy, Chien-Yao Wang, Hong-Yuan Mark Liao is the current state of the art model for Object recognition.
[007] The object detection technology is particularly helpful for companies managing highways and highway assets that keep track of all the highway assets. This particularly is a tedious task as there are thousands of assets along the highways. The highway assets could include fixed use-cases such as Streetlights, Signboards, Kilometre stones, Delineators, etc; Linear use-cases such as MBCB, Kerbs, etc; and other use-cases such as Pot Holes, Cracks, etc. Thus, it requires huge manual intervention and effort to keep a track of the asset conditions and at times it is practically impossible to manually keep track of all these assets.
[008] Thus, there is a need of a system that detects, tracks and monitors the condition of the highway assets present in the given site, with very minimal hardware requirements.

OBJECT OF THE INVENTION
[009] Some of the objects of the present disclosure, which at least one embodiment herein satisfies are as listed herein below.
[0010] The principal object of the present invention is to provide a system for highway assets management.
[0011] Another object of the present invention is to provide a system that detects, tracks and monitors the condition of the highway assets present in the given site, with very minimal hardware requirements and minimal human intervention.

SUMMARY OF INVENTION
[0012] A GPU embedded system to detect missing and damaged assets/use cases along the highway. These highway assets include streetlights, signboards, kilometre stones, delineators, cracks, potholes, etc.
[0013] The system includes a Comparison Unit and an Identification Unit for identifying and tracking the assets of the highway.
[0014] The Comparison Unit is configured to compare with a master video that the user has uploaded. Data extracted from the master video is used to compare with every day shift/patrol videos. The Comparison Unit works best for managing fixed assets such as Signboards, Kilometre stones, etc.
[0015] The Identification Unit does not require the master video. Thus, the Identification Unit works best for linear/random assets/use cases such as Potholes, Cracks, missing MBCB, damaged Kerbs, etc.
[0016] The system also works with Night videos. This can be used to know the working status of assets such as street lights, solar blinkers and reflective assets such as Hazard markers, Caution boards, etc.

BRIEF DESCRIPTION OF DRAWINGS
[0017] In the figures, similar components and/or features may have the same reference label. Further, various components of the same type may be distinguished by following the reference label with a second label that distinguishes among the similar components. If only the first reference label is used in the specification, the description is applicable to any one of the similar components having the same first reference label irrespective of the second reference label.
[0018] Fig. 1 illustrates a block diagram of a system of the present invention, according to an embodiment herein.
[0019] Fig. 2 illustrates a block diagram of Comparison Unit of the system, according to an embodiment herein.

DETAILED DESCRIPTION OF INVENTION
[0020] Exemplary embodiments will now be described more fully hereinafter with reference to the accompanying drawings, in which exemplary embodiments are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. These embodiments are provided so that this invention will be thorough and complete and will fully convey the scope of the invention to those of ordinary skill in the art. Moreover, all statements herein reciting embodiments of the invention, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents as well as equivalents developed in the future (i.e., any elements developed that perform the same function, regardless of structure).
[0021] While embodiments of the present invention have been illustrated and described, it will be clear that the invention is not limited to these embodiments only. Numerous modifications, changes, variations, substitutions, and equivalents will be apparent to those skilled in the art, without departing from the scope of the invention, as described in the claim.
[0022] As illustrated in Fig. 1, the present invention provides a system 100 for managing the highway assets. The system includes a Comparison Unit 200 and an Identification Unit 300.
[0023] In another aspect the system 100 includes a video capturing unit to record the daily patrolling videos.
[0024] In an aspect when a user uploads a master video the Comparison Unit 200 compares the data extracted from the master video to get information on fixed assets such as Signboards, Kilometre stones, etc. in the highway with every day shift/patrol videos of the video capturing unit.
[0025] In another aspect the Identification unit 300 is used to identify the linear/random assets/use cases such as Potholes, Cracks, missing MBCB, damaged Kerbs, etc. in the highway.
[0026] In another aspect the present invention includes GPS unit 400 which can be any positioning system known to the person skilled in the art. The GPS unit 400 associates the fixed assets detected by the Comparison Unit 200 and the random assets identified by the Identification unit 300 with location co-ordinates. Thus, the system 100 of the present invention is used for monitoring the highway assets.
[0027] The present invention provides a combination of YOLOACT and YoloV4 algorithms that are customised to run on GPU as a multi-threaded process to detect Highway Assets.
[0028] The present invention includes inhouse developed modules such as Tracking & Mapping, Localizing, etc. in addition to the YoloV4 model. The present invention also combines other proven segmentation models such as YOLACT to create a unique solution which addresses a wider range of use cases. The system 100 of the present invention is customized such that the entire framework to work specifically for Highways.
[0029] More than 35 use cases have to be identified and classified. However out of them, most cases have a lot of similarity. Default YOLO model assigns equal priority to every class, but road assets are of various sizes and shapes. Thus, the loss functions were modified according to the dataset and tweaked anchor boxes to get much better and accurate results. The hyper parameters were finalized after several designs of experiments.
[0030] Having a lot of trackers also takes a hit on the processing time. Any conventional tracker requires at least 3 frames to assign tracker and track its path. Any occlusion which is very common on Highways, leads to multiple count which will affect the whole solution. But in the present invention of tracking road assets, various occlusions were anticipated and possibilities of an asset being seen could even be just one frame. The present invention creates an inhouse tracking algorithm which not only works on single frame but also hits the right balance of performance and processing time. The present invention only consumes less than half the computational power than the existing available tracker algorithms. Also, as a lot of assets are located nearby and assets looking similar, it is difficult to use existing trackers like Kalman filter or IOU for tracking the objects.
[0031] As the solution requires to map assets to its location, the present invention uses geo coordinates & speed from the dash camera on the video to locate and mark assets location.
[0032] On the whole, the present invention runs multiple process simultaneously using multithreading approach, for identifying & classifying objects (GPU process), for tracking the objects (for counting), etc.
[0033] While the foregoing describes various embodiments of the invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof. The scope of the invention is determined by the claims that follow. The invention is not limited to the described embodiments, versions or examples, which are included to enable a person having ordinary skill in the art to make and use the invention when combined with information and knowledge available to the person having ordinary skill in the art.