Violation vehicle automated snap and road congestion detection

Violation vehicle automated snap and road congestion detection play an important role in intelligent transportation system. This article discusses the methods of target detection and recognition in mobile videos. Firstly, a practical method of lane detection based on Hough transforms is put forward. Then a method of target localization under the complex conditions in videos is studied, which is based on rear lights detection. Lastly, a standard how to set traffic congestion into 5 levels according to traffic conditions is proposed. The experiments show the algorithm is effective in reducing the false detection rate and improving the detection speed.