Study On The Detection Method In Moving Vehicle Streams Based On Video Monitor

Intelligent transport system (ITS) is currently being studied and the subject of widespread concern in the area of the world transport. These years, the great economic benefit has been gained through ITS, At the same time, it plays an important role in road design, traffic surveillance and highway charge. Video Vehicle Detection Technology discussed in this paper plays an important part in the ITS. Compared with traditional vehicle detection method with piezoelectricity sensor, Video Vehicle Detection Technology not only has lots of advantages, i.e., easy installation and maintenance, wide monitoring areas, and also through image analysis and processing it can obtain much useful traffic information, so it has been widely used in ITS.Now existing video vehicle detection technology in the complex weather changes, large-scale, multi-objective circumstances, the movement of vehicles segmentation and testing results are not very satisfactory, therefore an efficient and robust detection and tracking system is needed eagerly for these all. So, in this paper we use a means of monitoring cameras on the bridge to detect the movement of vehicles and statistics vehicular traffic. our researches are mainly as follows:1. In the video image preprocessing of improvement, we have given a value based on an improved gray similarity of the weighted mean filtering method, Experiments show that application of this method to image filtering has reached very good results, both noise filtering, and to overcome the traditional median filtering fuzzy details of the shortcomings. It provides a high quality image for the latter part of the background to update, target detection, analysis of the characteristics.2. After in-depth study in the context of the existing model, According to the statistical characteristics of static background, This paper improves and proposes a new method of statistical modeling and using dynamic and ternary group to carry out a maintenance update, At the same time it has improved detection accuracy and stability, achieved a rapid generation and real-time updates of the background. 3. In the round-the-clock video vehicle detection process, this paper provides two vehicles detection method. Daytime uses virtual coil vehicle detection based on the background, the difference in the evening inter-frame vehicle detection method. And then in the shadows on the Elimination we use edge detection operator and part of the shadow of the method of combining color to detect and remove the shadow. Finally, based on the movement of vehicles pixel count methods to statistics traffic flow information, and to provide the necessary information for divert traffic control.There results of the paper indicates that the detection method in moving vehicles based on video monitor is effective and feasible. This method is not only lower cost and more automatic, but also more chips. The application prospect is bright. Of course, in this paper, there are still some problems need a further research, such as the cooperation of multi-loops under the virtual loops based framework and Night and light of the objectives mutation detection.