Advanced Vehicle Tracking and Detection System using ByteTrack, Supervision, and YOLO Algorithms

With the assistance of ByteTrack, supervision, and YOLO v8 algorithms, I have developed a system that efficiently tracks and counts vehicles, including cars, buses, and bikes. Moreover, the system maintains separate counters for incoming and outgoing vehicles, as clearly demonstrated in the attached video.

The ByteTrack algorithm plays a crucial role in object tracking, as it not only tracks the objects but also determines their direction of movement. This information is invaluable in determining whether a vehicle is travelling in the incoming or outgoing direction.

The supervision library greatly assists us in determining which vehicles are crossing the designated line, whether it be in the inward or outward region. It enables us to accurately monitor and identify the vehicles' movements with respect to the defined boundary.

For object detection, we have utilised YOLO v8, renowned for its speed and high accuracy in detecting vehicles on the road. Furthermore, in the upcoming update, I will introduce the system with the latest version of YOLO, namely YOLO NAS. This advanced version is not only faster but also offers superior accuracy compared to YOLO v8.

In conclusion, our developed system incorporates the powerful ByteTrack, supervision, and YOLO v8 algorithms to efficiently track, count, and detect vehicles on the road. By leveraging the ByteTrack algorithm, we can accurately track objects and determine their directional movement, distinguishing between incoming and outgoing vehicles. The supervision library further enhances our system by enabling precise identification of vehicles crossing the designated line, regardless of whether they are in the inward or outward region. Utilizing YOLO v8 for object detection ensures rapid and accurate identification of vehicles. However, our commitment to innovation leads us to announce an upcoming update featuring YOLO NAS, the latest version known for its superior speed and accuracy in vehicle detection. This advancement will further elevate the capabilities of our system, enhancing its performance and reliability.

For code and git repository ::

https://github.com/sankalpvarshney/Track-And-Count-Object-using-YOLO

For checking the video of this system ::

https://user-images.githubusercontent.com/41926323/237469357-bbeb35b4-3f0f-49cd-b222-2bf92ac001f7.mp4