Video-based Vehicle Detection And Tracking Method

In recent years, accompanied by social and economic development, the road transport has been developed rapidly, resulting in a relatively complex traffic environment in order to solve the problems which cased by the development of ground traffic, intelligent transportation system has been referred to an important position, the use of computer vision technology to detect traffic conditions and information collection has become the Intelligent Transportation Systems (ITS) in the area of an important topic . That is, without damaging the road surface combined with the use of computer image processing, pattern recognition and communication of scientific knowledge, to get some traffic information, to solve a series of traffic problems, such as to alleviate traffic congestion and improve road through rates, and apply to Automatic toll collection system of highway, etc. So the exactitude of vehicle detection and tracking in the video is directly related to the correctness of the decision-making of. ITS. Therefore, this article conducts an in-depth research on the key technologies on the video vehicle detection and tracking method.In this article, the work carried out includes the following: the traditional methods of detection of the target vehicle, innovative ways of the Shadow removal and innovative ways to remove the vehicle tracking. As follows:1. The detection of the target vehicle. Detection of the target vehicle used the commonly finite difference method to obtain the background of moving targets. in order to make use of the background difference method to obtain the background, We make use of the existing adaptive background extraction and updating of methods to access and update the background; the next use of the background difference method, threshold-segmentation and morphological operations to detect a moving target.2. The shadow of the removal. In this article, proposed an adaptive Gaussian method for the weighted method to remove the shadow. This method from Imaging results of test images is effective than ever before.3. Vehicle tracking. This article presents a tracking method of moving targets based on random sampling and Color gradient .Test-image data show that the method is more robust, real-time.