| Automatic passenger counting system is mainly applied to some public places as bank, shopping mall, bus station, gymnasium, airport and so on. It can detect the number of people going in and out in real-time for managers. So that they can operate the management mechanism more effective to realize intelligent office.With the development of video image processing technology, the automatic passenger counting technology based on image processing is widely used. But two-dimensional image quality is easily affected by outside conditions such as sunshine. In addition, it is difficult to separate targets from crowds Therefore, the passenger counting system based on binocular stereo vision is designed, improving on existing technology. Three-dimensional information is obtained from grayscale image scene through binocular stereo vision.In this paper, System algorithm is divided into three main sections:1) The stereo matching algorithm is designed based on the theory of division, but it is not exactly the same. So we called it "the theory of analogy division". It can realize stereo matching of image with little damage to the image. It has high matching accuracy according to the experimental results2) Human head can be identified quickly by the algorithm based on image depth information processing. It is applied to real-time monitoring system due to its higher precision and fast speed. And also the amount of calculation is small.3) The head target tracking algorithm is designed by the algorithm of Linear prediction based on adaptive gate. In order to make multiple wave gate work separately, the Priority control strategy is designed to make the gate work more effectively.The algorithm of the automatic passenger counting system. In this paper not affected by light and meet the requirements of real-time detection with simple algorithm of high precision and fast speed. According to the experiment results, the accuracy is more than 95 percents and the speed is 30Hz. |
PDF Full Text Request