A Novel Self-Adaptive Vehicle Segmentation Method in Traffic Video with Symmetric Frame Difference Constraints
Hongtao Wu, Ying Meng, Bingqing Niu, Junyi Ren, Mingshu Shen · 2023
Vehicle target detection technology for traffic video is the technical foundation and key bottleneck technology of traffic monitoring. Aiming at the high-definition video image collected by the roadside monitoring equipment, this paper explores the vehicle detection method based on the traffic video. In the traffic scene, moving and stopping are the vehicle target participating states that affect the road safety. In this paper, a vehicle self-adaptive threshold segmentation method for traffic video under the constraint of symmetrical frame difference is proposed. By the idea of symmetrical difference, the adjacent three consecutive images are differentiated, the background of the high-definition monitoring video frame is extracted and updated, the moving vehicle target in the result image is segmented and detected by self-adaptive threshold, and the stopped vehicle is segmented by background updating and filtering. The experimental results show that the proposed method can segment vehicle targets with different traffic states completely and accurately. In addition, the proposed method effectively improve the detection accuracy of vehicle targets with different situations.