| With the maturation of automobile testing technology, the brightness of testing procedure, the straightforward and real-time showing the testing results have been a hot topic in automobile testing field. In this dissertation, storage, transmission and display of surveillance video in automobile testing station are reseached in detail. Finally an intelligent video surveillance system for automobile station is implemented.On video storage, a moving-object detection based video recording algorithm is proposed and implemented. After color conversion denoising, the Gaussian statistics model is applied to update background and moving object is detected by image difference. When the moving objects of interest are detected with threshold method, the storge devices are started up to record. The method effective solves the problem that video needs large storage.On data transmission, a hybrid signal transmission mode is adopted. Intelligent video surveillance data in automobile testing station consists of control signal and audio/video. For the scale of control data is small and it needs high reliability, they are transmitted with TCP protocol. For multimedia data are very large, UDP protocol is used. The hybrid transmission method is implemented with WinSocket.On diplaying data, a multi-source information fusion method is employeed. The testing data collected from testing equipments are superposed to the surveillance video. Tesing procedure and results are directly diplayed. Testing situation of different positons is recirculately displayed in multi-display panel, which can descrease the cost of automobile testing station.Experiments show the proposed system can not only improve the central control and management level, but also show the testing procedure and results to drivers in real-time, openness and brighteness, which improve the service and competitive capacity of the automobile testing station, and provide technology support to the administration department.
|
PDF Full Text Request