A Lane Detection Algorithm under Complex Scenes

Chun-yang CHENG, Min LI, Xue-wu ZHANG, Yu-bo XIE, Yan XIANG, Jin-bao SHENG


A lane detection algorithm under complex road conditions is proposed in this paper, which will be helpful to improve the accuracy and stability of lane detection. In the image preprocessing stage, a rapid temporal blur algorithm is employed to connect the dashed lanes. Besides, based on the traditional gamma correction method, a new specific method combined with the tangent function is designed to achieve a good threshold segmentation effect in all pixel of the image. In the lane detection period, the detection algorithm based on Hough transform and heuristic search is applied to the straight lane and curved lane respectively, and then the two parts are fitted by B-spline into a smooth curve. Experimental results show that the lane detection algorithm under complex road condition proposed by this paper has a good performance of real-time capability and robustness.


Lane detection, Temporal blur, Tan-Gamma correction, B-spline


Full Text:



  • There are currently no refbacks.