Lane Detection on Autonomous Vehicle

Abstract

This research project proposed an upgraded lane detection function applied in autonomous vehicle in order to reduce the number of death accident recently. The main idea of this research project is to enhance an existing algorithm in detecting the continuous lane marking in highway, urban road or in country road with different scenario of road either straight road and winding road, and printed mark with a solid line or dash line, and applied as a real time application. For this purpose, the system simplify the process that groups the detected edge points and segmented as a straight line once being captured by an onboard camera. Additionally, before the process of estimating the road parameter in knowing the vehicle position and assist the driver to keep to the center of the departure lane, this system selected the best region of interest in purpose of removing the road surrounding and differentiate the image to get an accurate painted straight line counted as a lane boundaries. Thus, the method proposed is to improve with better and robustness system applied for intelligent vehicle in daytime with complicated geographical road background