Search Articles

Home / Articles

An Efficient Yolov7 and Deep Sort are Used in a Deep Learning Model for Tracking Vehicle and Detection

. Vinod Kumar Yadav, Dr. Pritaj Yadav and Dr. Shailja Sharma


Abstract

For tracking and detecting vehicles, use computer vision techniques. It is essential to traffic accident detection and intelligent transportation systems. On the highway an essential component of traffic surveillance is the detection, identification, and counting of vehicles. It takes a lot of effort to create a traffic monitoring model that performs well. Artificial intelligence-based traditional vehicle detection systems have weak detecting capability and robustness. A deep learning model for vehicle detection, tracking, and counting is proposed in this paper and is based on an efficient Yolov7 single shot detector and Deep- Sort of Multi Object Tracking algorithms. The suggested model examines the automobile detection algorithms and suggests proposed detection models using moving vehicle footage as survey data. When observed under a range of circumstances, such as high traffic, nighttime, many vehicles overlapping, and part of the vehicle missing, the suggested identification system exhibits excellent adaptability. The algorithm can accurately detect and identify automobiles based on their edge outlines, according to experimental data. YOLOv7-DeepSORT performs higher in tracking accuracy after experimental evaluation as compared to the earlier YOLOv5-DeepSORT.

 

Index Terms- Computer vision, deep learning, Deep-sort, Yolo, Yolov5, Yolov7 and MOT.

Download :