Pedestrian Classification Based on Improved Support Vector Machines

Abstract

We presented a pedestrian classification method based on improved support vector machine in order to solve non-rigid objects are difficult to identify in intelligent monitoring system. The video activity in the prospect is represented by a series of spatio-temporal interest point. Since human posture has the characteristics of uncertainty and illegibility, the clustering centers of each class are computed by fuzzy clustering technique. Then a full-SVM decision tree is constructed based on conventional decision tree. At last, the method is evaluated on the Weizmann action dataset and received comparative high correct recognition rate.