Vehicle Color Recognition Using Vector Matching of Template

As an important feature of vehicle, vehicle color plays a significant role in Intelligent Transportation Systems. Traditional vehicle color recognition using experience-based artificial demarcation of threshold and voting method to identify, can´t meet the requirements of the recognition rate in complex environments of video traffic and subdivision of vehicle color. In this paper, we present a method inspired on vector matching of template and a color normalization operator for video pretreatment, and then select HSI color space and Relative error distance matching algorithm. After experimental analysis, we compare our method with various common parameters in color recognition in order to get optimal results. Experimental results show that our method obtains a higher recognition rate in a variety of external environment and multi-color classification.