Detection Method For Intersection Vehicle Queue Length Based On Image Processing

Vehicle queuing length at signalized intersections is an important indicator for evaluating the capacity of intersections.Traditional vehicle queuing detection methods require special detection equipment,which has a small detection range and is expensive to repair.Video-based vehicle queuing length detection not only can be enriched Traffic parameters,and it can realize the backtracking of information,which has a broad application prospect.Existing video-based queuing length detection is difficult to meet the detection requirements of queuing queues in actual scenarios due to low accuracy or poor applicability.Because of challenges mentioned above,the research of vehicle queue length detection method based on video is of great theoretical and practical significance.This paper focuses on the specific scene of intersections.After analyzing the problems and difficulties of vehicle-based queue length detection based on video,we focus on the camera partition calibration and vehicle queue detection technology based on lane line information.Two parts are used for vehicle target extraction and queuing tail detection,and a set of video-based vehicle queuing length detection methods are finally formed.It mainly includes the following two points:(1)In terms of camera calibration,first use image graying and image filtering to remove noise interference in the road image.Then,through the detection of the solid and dashed lines of the lane,a single lane is separated as the region of interest,and finally the lane is used.The geometric features of the line,combined with the idea of zoning,propose a zoning calibration algorithm based on lane lines.The experimental results show that the method can improve the calibration accuracy of the camera and can lay the foundation for the subsequent calculation of the vehicle queue length.(2)In this paper,an algorithm of queue length detection based on double detection frames is proposed to address the problem of vehicle queuing images in intersection scenarios.First,a fixed frame is set at the head of the queue,and vehicle motion detection is performed in the frame to determine vehicle queuing State,combined with background difference and texture difference to extract vehicle foreground information,and then use variable threshold line scanning method to search out the queue tail,at the same time according to the vehicle width information in the foreground information,a simple classification of vehicles to solve the camera angle and The error problem caused by the height of the body,so as to calculate the actual queue length of the vehicle.Finally,combined with the vehicle queuing length detection method proposed in this paper,the video data in the intersection scenario is used to conduct experimental verification in the Visual Studio2013 environment.Experimental results show that the proposed method can effectively improve the detection accuracy of vehicle queue length.