use hough transform to detect line(lane)

最新推荐文章于 2024-09-13 08:18:37 发布

原创最新推荐文章于 2024-09-13 08:18:37 发布 · 210 阅读

0 ·

CC 4.0 BY-SA版权

course 专栏收录该内容

47 篇文章

订阅专栏

博客介绍可通过参考链接获取霍夫变换基础知识，阐述处理方法，包括将RGB图像转为灰度图像、边缘检测和霍夫变换。还指出原处理中右车道缺失问题，添加高斯滤波到边缘检测前，表明高斯滤波可处理停车标志噪声。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

you can read

ref:https://zhuanlan.zhihu.com/p/29012549 ;ref:https://blog.youkuaiyun.com/happy_stars_2016/article/details/52691255

to get the basic knowldge of hough transform.

method:

rgb image to grayimage-->edge detect--->hough transform

edge detect

hough transform

as you can see, the right lane was missing.

After add gassian filter before dedge detector

   %% preprocessing of image
    % apply Gaussian Filter
    sigma=5;
    window=double(uint8(3*sigma)*2+1);
    H=fspecial('gaussian', window, sigma);
    grapy_pic=imfilter(grapy_pic,H,'replicate');
    imshow(grapy_pic)
    
edge_pic=edge(grapy_pic,'canny',[0.05 0.35]);
imshow(edge_pic)