OpenCV的车道线检测方法

基于开源计算机视觉库(OpenCV),提出一种轻量级的车道线检测方法.首先,对输入的原始图像进行灰度化处理,紧接着使用双边滤波滤除噪声,大幅度保留原始图像的边缘信息;然后,用Canny边缘检测提取图像边缘;最后,使用速度更快的渐进概率Hough变换(PPHT)识别车道线.仿真结果表明:预期检测车道线的效果较好.

Lane Line Detection Method Based on OpenCV

Based on the open source computer vision library(OpenCV), a lightweight lane line detection method was proposed. First, the input original image was grayed, and next the noise was filtered out with bilateral filtering to greatly preserve the edge information of the original image. Then,the method of Canny edge detection was used to extract the edge of the image. Finally, a faster progressive probability Hough transform(PPHT)was adopted to recognize lane lines. The simulation results show that the effect of expected lane detection is better.