High Speed Front-Vehicle Detection Based on Video Multi-feature Fusion

Abstract

Aiming at the scene where the speed of the vehicle on the highway is fast, this paper proposes a method to detect the front-vehicle in the video. Firstly, the method detects the lanes based on Canny edge detection and Hough transform to determine the vehicle’s driving area. Secondly, we use the multi-feature obtained by the combination of the histogram of oriented gradient (HOG) feature, the color feature and the Harr feature of the vehicle to train support vector machine (SVM) classifier, and then the classifier detects the vehicles in the driving area. The experimental results show that the SVM classifier trained by the multi-feature fusion method has a better detection result than a single feature, and compared with the detection of vehicles in the entire image, the detection time can be greatly shortened in the driving area.