A real-time lane detection and tracking algorithm

Abstract

Aimed at the problem of lane detection in intelligent vehicle safety aided system, fusing multiple features of the lane, we propose a real-time lane detection and tracking algorithm based on Gabor filter and edge extraction respectively. It can be suitable for fast lane detection on the structured road under the multi-lane and multi-scene. Firstly, in the initial lane detection module, the Gabor filter is used to enhance the characteristics of the lane, and the lane is detected under the restriction of polar angle range by the Hough transform, then using the geometric relationship between the lane and the vehicle to judge the current lane. Secondly, referring the tracking detection module, the lane location of the prior frame is used to establish the dynamic region of interest. The edge of lane is extracted by using the different edge detection operator under the different illumination conditions. Next, the lane is tracked by the least squares fitting in this region. Experimental results indicate that the proposed algorithm is real-time and efficient, and enhance the adaptability for the changing environment of road scene.