Automatic Vehicle Detection and Tracking Strategy Using Deep Learning Model (YOLO v2 & R-CNN)
Year: 2024
Type: article
Abstract: Vehicle detection and tracking plays a crucial role in various smart mobility networkss, including traffic monitoring, security surveillance, and autonomous vehicles. Image processing techniques offer a powerful tool for achieving these tasks by analyzing visual information captured from cameras. Since Deep Learning (DL) is developing so quickly, the computer vision community has demanded that excellent, reliable, and efficient services be developed across a range of domains. The objective of vehicle detection and tracking is to automatically detect and track the movement of vehicles in real-time video sequences. The paper presents novel vehicle detection and tracking system using image processing techniques. The approach consists of the key stages: An assortment of image processing techniques was employed in the production of this data. Utilizing the most recent iteration of the YOLO model, YOLO v2, as well as the R-CNN model and Fast-RCN, have all been employed for detection purposes. Following the identification and tracking processes, the number of vehicles and their estimated speeds are calculated. Moreover, the proposed approach using Fast-RCNN has the best performance with precision of 98.94% and recall of 99.12% for the CDnet 2014 dataset respectively..
Institution: Hindustan Institute of Technology and Science
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7217029290
