Indian roads incorporate a combination of crossroads and intersections. The traffic situation in India is very complex due to the ever-increasing vehicle density. There is a need for an autonomous system to detect and analyze the vehicles on the road. However, designing an accurate object/entity detection mechanism is not easy because of the need for high dependency factors. This paper aims to construct a system that can detect, count, and classify vehicles accurately in real time with limited resources. The paper proposes a Semi-Automatic Vehicle Detection System (SAVDS) framework to recognize objects using outdoor CCTV footage by combining strategies such as optical flow, background subtraction, and convolutional neural networks. The optimization achieved as a result of a reduced detection area and a detection grid system is the work’s major feature. Various background removal methods and HSV color model processing were utilized to determine the optimal approach to perform in various ambient situations. The results show that the proposed work classifies vehicles with accuracy up to 85% for benchmark datasets.
Semi-automatic Vehicle Detection System for Road Traffic Management
Algorithms for Intelligent Systems
Proceedings of 3rd International Conference on Artificial Intelligence: Advances and Applications ; Kapitel : 23 ; 303-314
15.04.2023
12 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Road intelligent traffic system for vehicle automatic driving
Europäisches Patentamt | 2022
|Road traffic: Vehicle detection and classification
IEEE | 2017
|Europäisches Patentamt | 2025
|