Vehicle Classification using IPCP and EsKNN Algorithm for Surveillance Camera

Abstract

This paper presents a new approach vehicles detection and classification. In these works, we are creating a system to detect the vehicles and automatically classifying them into three desired classes as Car, Suv, and Truck. Datasets from changedetect.net were used for testing the method. The proposed approach combine Incremental Principle Component Pursuit (IPCP) background subtraction for vehicle detection and it complies with Ensemble subspace -Nearest Neighbor (EsKNN) classifier. A combination of background subtraction and classification of filtering background is used in order to focus on vehicle’s features extraction using HOG corner detections. Then, this feature extraction is classified using an ensemble-based classifiers technique. The choice of classifier depends on features data characteristics, which they were formed in certain unique distance shape stationary relative between classes. The proposed method is evaluated using three datasets of common highway surveillance video. Comparing with other direct detection and classification technique our method has achieved outstanding result. The proposed approach delivers the accuracy 96.5%, the highest among the tested methods. Experimental results show the outstanding performance of the proposed method.