The Illegal Lane Change Monitoring System Based On Video Technique

With the development of economy, car ownership in China continues to rise. Although the increase of vehicles brings the convenience to people’s life and work, it makes traffic conditions get worse at the same time. Illegal act such as illegal lane change, drunk driving, speeding and overloading is the main reason of traffic accident, where illegal lane change is a dominant factor. Traditional detection of illegal lane change behavior is through artificial capture which is not only a waste of manpower, resources but also easy to produce the missed detection, resulting a low rate of accurate detection. To solve this problem, we use video detection technology to detect illegal lane change which not only saves resources, but also improves the accuracy of detection, and it has good prospects.This paper focuses on the core algorithm of illegal lane change detection based on video detection technology. We improve some algorithms through analysis and experimental verification, and complete the development of the illegal lane change monitoring system based on video detection technology. The main work of the paper is as follows:(1) Aiming at the necessity of system, the background subtraction algorithm is used to achieve the target background detection based on deeply analyzing the advantages and disadvantages of each target detection algorithm. The algorithm of color background modeling based on memory matrix is proposed in the stage of background modeling. It uses the color information of the video image effectively to eliminate the effect of noise and improve the accuracy of background modeling. To solve the problem of incorrect location, an improved region growing method that extends the seed point and neighborhood to super pixel is proposed and the experimental results show the effectiveness of the algorithm.(2) An illegal lane change detection algorithm based on trajectory is proposed. The algorithm judge whether the behavior of lane change occurs by the dispersion of distance between lane line and trajectories obtained by target tracking. Big dispersion Indicates the distance between lane line and trajectory obtained by target tracking changes greatly, and the behavior of lane change has occurred.(3) We have thoroughly studied license plate location, character segmentation and character recognition which are the three processes of the license plate recognition. A novel license plate location algorithm that combines edge detection and color features is proposed.(4) The design and implementation of illegal lane change monitoring system based on video detection technology is completed. The monitoring system in this paper is divided into three parts:image acquisition module; illegal lane change detection module and data storage communication module. Aiming at the necessity of system, we add some auxiliary functions such as PTZ control, image capture, and video recording, etc.