Abstract

The lane detection and recognition is very important in unmanned driving technology. In order to improve the accuracy and robustness of lane detection and overcome the influence of changes in illumination, curvature, and road interference, a lane detection algorithm based on reverse perspective mapping is established. The binarization image of a lane with less noise is obtained by the global optimal threshold method. Then through the reverse perspective mapping, the binary lane image was converted into the top view to overcome the shortcomings of different resolution and geometric deformation of the image caused by the perspective effect. Then, the lane images transformed by reverse perspective were clustered and fitted by k-mean algorithm, and the clear lane detection results were obtained. Finally, by analyzing the lane detection of road images under different imaging conditions, the robustness of the lane detection algorithm under the conditions of high curvature, large change in brightness, and multiple interference factors were verified.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call