Road Traffic Parameter Monitoring Systems Based On Video Detection Technique

With the sustained and rapid development of urbanization, car ownership and road traffic mileage increase continuously in China. Some of new problems like road traffic congestion, frequent traffic accidents, environment pollution and so on become prominent. Development of Intelligent Transportation System (ITS) has a good effect to alleviate these problems, and traffic parameters extraction is one of the important research content of ITS. Traditional traffic parameters detectors are installed in the road with high cost, small detection range, single acquisition parameter and so on shortcomings. Traffic parameters extraction method based on video is developing rapidly with the advantages of flexible system setup, rich information, and wide detection range. It becomes a research hotspot in the field of intelligent transportation, and has a broad application prospect.This paper focus on the video image dehazing algorithm and moving object detection algorithm. We improve part of algorithms through the experiment analysis and validation, and complete traffic parameters monitoring system based on video detection technology with haze. The main work of the paper is as follow:(1) A novel fast image dehazing method using guided filter is proposed. Before guided filter processing, the rough transmission maps and the guided image employed nearest neighbor interpolation down-sampling, to zoom out on the input images of guided filter and reduce the processing of guided filter time complexity. After guided filter processing, the guided filter output image employed bilinear interpolation up-sampling, to zoom in on the output images of guided filter and get the refining transmission maps. Results demonstrate shows that the algorithm is effective and real-time.(2) In further analyzes the advantages and disadvantages of all kinds of moving object detection algorithm, and the research is mainly focused on the background subtraction method. Finally through comparing various background extraction method, using the based-pixel adaptive moving object detection algorithm. Results demonstrate shows that the algorithm is effective.(3) On the basis of understanding physical meaning of the traffic parameters, studying the method of extracting traffic parameters and building the road traffic parameter extraction process.(4) On the basis of mastery the video image dehazing algorithm, moving object detection algorithm and parameter extraction process, designing the road traffic parameters monitoring system based on video detection technology. And implementing the monitoring system function of video image acquisition, video image dehazing, moving vehicle detection, traffic parameters extraction and traffic parameters storage. The experimental results show that the system has achieved high accuracy on traffic parameters monitoring, and has reached the design requirements.