Vehicle Tracking Method Using Background Subtraction and MeanShift Algorithm

As urban road intersections are prone to traffic congestion and traffic accidents, monitoring the crossing of vehicles and predicting the state is needed to reduce traffic congestion, regulate driver behavior and prevent accidents. Background subtraction and mean shift tracking are used to track vehicles. The whole monitoring process is as following. Firstly, secondary selected strategy is used to construct background model. Then vehicle tracking objects are built at the trigger area of detection by the background subtraction. Finally, the mean shift algorithm is utilized to track vehicles. The secondary selected strategy is a new algorithm designed in this article .It can reconstruct quickly the accurate background from the crowd video frames. Using background subtraction can eliminate the interference of background on the color probability density of target in mean shift algorithm. The whole algorithm achieves the real-time tracking in complicated situation in a high accuracy.