Pedestrian detection in video using shape features and mixture of SVMs

This paper proposes a real-time pedestrian detection system, which uses the point signature to make feature extraction of the object profile for the connected parts on the foreground image, it classifies the feature of the real-time pedestrians through the introduced mixture of SVMs classifier. It will divide the large scale training samples into many smaller sub-sets and it trains a SVM for each set. It uses a neural network to put out the weight sum of these SVMs results. Through the experimental verification, it can effectively improve the accuracy of pedestrian detection system. For large scale of training sample sets, the calculation time of the proposed method is shorter than that of using the single SVM, and the test accuracy also gets the better results than that of the single SVM.