Lane Detection for Intelligent Vehicle under Different Illumination Conditions

Lane detection is an important component for intelligent vehicle. A novel method for lane detection under complex illumination conditions is presented. An adaptive image segmentation technology based on OTSU algorithm was used to segment lane markings by combining the global and the local threshold. The improved Sobel operator was adapted to abstract the edges of the lane markings. The lane was finally identified using the improved Hough transform by limiting the range of lane parameters. On road experiments show that the proposed method has good accuracy and real-time performance under different illumination conditions. Keywords-intelligent vehicle; lane detection; linear lane model; Hough transform