Lane departure identification on highway with searching the region of interest on Hough space

An updated expressway lane departure identification which based on a combination of the lane boundaries detection method and lane following with searching of the region of interest on Hough domain were proposed in this paper. After preprocessing captured road images, we employed mathematical morphology technique to obtain edge map and Hough Transform (HT) to detect lane candidates. To estimate lane boundaries among these candidates, we use some information such as lane orientation (angle), the symmetrical property, boundary length, etc. on Hough domain. And then, lane following with prior of knowledge of lane detection showed a good performance. The robustness and the low compute cost of the proposed algorithm had been shown with experiment results.