Multi-traffic objects classification using support vector machine

In order to classify the traffic objects in multi-traffic scenes, six classes were divided firstly, then eight features base on shape and motion information are extracted. The eight features of traffic objects will be the input of the support vector machine (SVM) classifier which is contrasted with RBF neural network classifier. The object type is classified according to the output of the SVM. Experimental results based on actually scene video indicate that the algorithm could classify the traffic objects in multi-traffic scenes at a high recognition ratio.