Real-time lane line detection method

Abstract

The invention provides a real-time lane line detection method. According to the real-time lane line detection method, an interested area is determined through the position of a current frame vanishing point, and an upper half image where no lane lines exist is removed, and therefore, the processing time of each frame of image can be decreased; edge point scanning is performed at two directions from inside to outside of the interested area, and what is detected every time is the inner side of a lane line which is most adjacent to vehicles on a lane, and except for the interference of other edge points, jigger will not be generated because of the width of the lane line, and since an image is divided into two half parts of detection lane lines from the position of the vanishing point, a situation that two detected straight lines are located at one side of the vanishing point when the entire image is detected can be avoided, and thus, the accuracy of the detection can be improved; and further, when Hough transform lane line detection is performed, only straight lines in a range of straight lines which can form a certain included angle with a horizontal line are selected, and therefore, the accuracy of the detection can be improved, and time for calculating Hough weight of straight lines that are not in the range of the straight lines can be saved. With the real-time lane line detection method of the invention adopted, lane line detection at highways and rural areas where roads are in good condition can be realized rapidly, accurately and stably.

Description

Technical field [0001] The invention belongs to the field of computer vision, mainly relates to straight line detection technology in image processing, and is used in a lane departure warning system. technical background [0002] At present, many useful explorations have been carried out in preventing lane departure at home and abroad. After a lot of research and practice over a long period of time, people have gradually realized that the use of monocular vision technology and the use of only one camera can achieve the functions of road environment, vehicle detection, and vehicle distance monitoring to a certain extent. Machine vision images are rich in information, and can simultaneously perform lane line detection, traffic signal recognition, and obstacle recognition on multiple lanes. It can also provide real-time video for subsequent analysis. Most importantly, the application of machine vision technology to realize vehicle lane departure warning and collision warning system...

AI Technical Summary

Problems solved by technology

This will cause large errors in the subsequent calculation of the relative position of the vehicle to the lane line and the angle between the vehicle's driving direction and the lane line

Images