Lane Detection Algorithm based on Feature Constraint 

Hang AN1, Qingwen LI2

1Chongqing Jiaotong University, 400074, Chongqing, China

2Southwest JiaotongUniversity, 611756, Chengdu, China

Abstract: Based on the feature constraint, a good method of lane mark detection is proposed. According to the geometric characteristics, position characteristics and physical characteristics of the lane line, the lane line candidate area is obtained by connecting the regional analysis on the binary image. Then, the Hough trans-form is used to detect the straight line segment in the candidate area, The Then we calculate the image disap-pearance point and the lane line geometric characteristics according to the camera map. These characteristics are used as a priori knowledge constraint candidate lane line, and finally the filtered straight line is clustered and fitted to get the final lane line.

Keywords: lane line detection; Hough transform