Traffic Impact Assessment System using Yolov5 and ByteTrack

Monitoring software for traffic is not too much in this era of digital. Even cheaper is decent traffic monitoring software. You can gauge the quality of the software. It should be possible to assess the code's performance outside of a test environment. The most useful metrics are frequently those that support the program's ability to fulfil business requirements. Therefore, this project is planning to develop a traffic assessment system. The main purpose of development is to improve heavy traffic in this country – Malaysia. This system includes function vehicle detection using YOLOv5, vehicle counting with a different type (such as bus, car, truck), vehicle classification, vehicle idling time by each region, and vehicle counting for each junction. Users can draw regions and lines for each camera/video to count and record vehicles. After the traffic analysis, intelligent signal light systems that respond to loads and timing can be helpful in easing traffic congestion. Smart traffic lights may adapt to the patterns of bustle at junctions and other important road traffic places based on the number of cars, data from queue detectors, and images from cameras. Also, this report includes comparisons with StrongSORT, OC-SORT and ByteTrack and accuracy test for vehicle counting.