Robust vehicle detection at large distance using low resolution cameras

This paper describes a robust vehicle detection system which detects vehicles in the rear view of the host car. Two cameras with 12mm focal length mounted behind the back pane are used. Reliable vehicle recognition up to 150m and the exact track localization of the vehicles are the key features of the system. The robust stereo algorithm generates object hypotheses which are tracked with Kalman filter. Using the steering angle and the detected obstacles, the driven way of the host vehicle is reconstructed. This information is used for a lane change assistant. Experimental results in real traffic scenarios confirm the robustness of the proposed algorithms.