자율주행 차량을 위한 주행방향 추정 및 도로 노면표시 검출 방법에 대한 연구

This paper presents a method to detect the road marking and estimates road following direction by a front camera in vehicle for an autonomous navigation. An algorithm is proposed which is based on the following processing steps. In the preprocessing step, the proposed method sets the vehicle lane as a region of interest. Estimation of the vehicle direction is estimated as the coordinate conversion and the estimated parameter of the road system. The processing for detecting the lane mark on the road detects the lane as obtained vanishing point using connected straight line component according to Hough transform, application of Bird’s eye view algorithm. Road marking detection of template matching using normalized correlation of coefficient was detected. In experiment, we detect the road mark such as arrows mark, crosswalk notice mark, and so on. We think that the proposed algorithm can be extended easily to solve the problem of the identification of all types of road markings.