Training method for vehicle detection

Recently, vehicle detection methods have been popularly used in the field of intelligent vehicles. The performance and processing time of vehicle detection is very important because it is associated with the life of a driver. However, all vehicle detection methods generate missing detections and false detections because of different vehicle appearances. However, in a general road environment, the appearance of most of these vehicles has a front and a rear. In this paper, we propose a training method to detect the front and rear of the vehicle. Our vehicle detection integrates state-of-the-art feature-based detection.