This paper proposesSathyabama, B. an intelligent traffic surveillance system thatDevpura, Ashutosh detects the vehicle and its speed, colour, direction,Maroti, Mayank and type of vehicle using computer vision and deepRajput, Rishabh Singh learning. This information can be used to find the traffic violator using automatic number plate recognition. Research shows that over-speeding accounts for 60% of total accidents in India, which raises a serious concern. The proposed approach uses TensorFlow object detection API for vehicle detection, cumulative Vehicle counting, and colour detection of the vehicle using colour histogram integrated with the KNN machine learning algorithm in a real-time environment and a robust approach using deep learning and computer vision for speed estimation and direction detection. This study will effectively monitor traffic usage and help officials track, detect, and lay a floor plan to effectively stop speeding and wrong-side driving vehicles from getting into accidents. This paper proposed an efficient and robust approach for detecting moving vehicles along with their speed and other attributes. The proposed approach can be integrated with a pre-installed traffic monitoring camera system without significant adjustments.
Vehicle Speed Estimation and Tracking Using Deep Learning and Computer Vision
Lecture Notes on Data Engineering and Communications Technologies
24.02.2022
12 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Vehicle Tracking and Speed Estimation Using Deep Sort
IEEE | 2022
|Hybrid lane estimation using both deep learning and computer vision
Europäisches Patentamt | 2023
|Vision-Based Vehicle Speed Estimation
IEEE | 2024
|