RETRACTED ARTICLE: Multi-dimensional image edge localization method based on edge symmetry algorithm

Abstract

Video vehicle detection is the premise of vehicle tracking, video speed measurement, traffic statistics, vehicle identification and other technologies. Its accuracy and algorithm complexity directly affect the overall efficiency and final results of intelligent transportation system. Aiming at the problems that existing video vehicle detection algorithms are greatly influenced by illumination, shadow and other environmental factors, such as high rate of missed detection and false detection, a video vehicle detection algorithm is proposed. Unlike traditional algorithms, which use motion features for vehicle detection, the algorithm uses edge features and pairwise features to locate vehicles. In this paper, video vehicle detection technology is deeply studied. In this paper, firstly, the difficulties and problems of video vehicle detection are briefly introduced. Then, a video vehicle detection algorithm based on edge symmetry is proposed according to the characteristics of video vehicles. Finally, the detailed experimental results and analysis are given. In this paper, a video vehicle detection algorithm based on edge symmetry is proposed, which integrates the edge features and symmetry characteristics of vehicle images, improves the reliability and accuracy of the video vehicle detection algorithm, and has fast detection speed, and has strong adaptability to outdoor environment changes. It is suitable for vehicle detection under simple background.