A Lane Detection Technique Based on Adaptive Threshold Segmentation of Lane Gradient Image

In order to solve the problems of being difficult to adapt to the change of illumination conditions and the road shadow or other noises interference, an approach for real-time lane detection based on adaptive threshold segmentation of lane gradient image. Considering the feature that the lane line usually has a higher brightness than the surrounding road surface, that is, there is a larger grey value, we extract lane edge pixels by imitating Canny edge detection technique to extract edge pixels gradient and then segment it with a threshold obtained from OTSU algorithm. Finally, lane line detection and tracking is realized by Hough transform and Kalman filter. The experimental results show the effectiveness of the proposed methods, and the detection results are consistent with the actual situation. The processing speed is about 16 fps, which basically meets the real-time requirements.