Gaussian mixture model based on the number of moving mehicle detection algorithm
Weiqi Yuan, Ji Wang · 2012
Video surveillance is a modern city in an important way to monitor traffic, it can real-time, reflecting the effective operation of vehicles on the road. In a fixed scene, in order to detect moving vehicles on the road to the city the number of proposed algorithms using the Gaussian mixture model in the foreground video image to extract information on the use of regional markers in each frame the number of vehicles for identification. The algorithm first use of Gaussian mixture model for statistical analysis of video images, to make judgments on the current frame image obtained after the current frame in the foreground information; and morphological processing of information with prospects, the binary image obtained after easy machine readable binary image; last through the foreground image in the region marked the vehicles to do, get the city moving vehicles on the road number.