Truck Peccancy Monitoring System Based On Vehicle Classification

With the developm ent of econom y, traffic problem s, such as traffic jam and management backward has been increasingl y sticking out. In order to solve these problems, every country develops ITS grea tly. This p aper designs a kind of truck peccancy monitoring system based on vehicle class ification, which could carry o ut efficient real-time monitoring on tru ck peccancy,with merits of high Precision and high automatization.Firstly,a proper scheme successfully advanced to collect real-time digital traffic video stream automatically,which is based on the cooper ation of video-collecting card and v idicon equipped with co lor CCD camera with high differentiation ratio. And then vehicle with yellow license is detected to get the large vehic le. The vehicle classification is based o n the perspective so as to obtain the whole im age of vehicle. According to the characteristic of our system ,mobile objects are distinguished by means of linear com posing between the dist ance and close shot, which is proved of simple. Then a new method which combines the whole and locally features is brought forward to. This paper improves the tradi tional method of s electing the whole veins other then the detail veins, and proposes to select the whole veins by Radon, which utilizes the difference b etween the truck and the bus and com putes fast. The detail features which are the am ount of blocks and colors and the ratio of window are distilled by the image segmentation. In the segmentation process, this paper improves Hough transfor m and proposes image segm entation based on the RGB m odel. It proves that the effect of segmentation is perfect.Finally, this paper shows the im plement of truck pecc ancy monitoring system based on vehicle classificati on, and through experim ental proof in algorithm s, this system satisfied the need of real time and veracity.