基于改进Hough变换的车道线检测技术  

Research on Lane Detection Based on Improved Hough Transform

作　　者：杨喜宁[1] 段建民[1] 高德芝[1] 郑榜贵[1] 

机构地区：[1]北京工业大学智能测控研究室,北京100124

出　　处：《计算机测量与控制》2010年第2期292-294,298,共4页Computer Measurement &Control

基　　金：北京市教委重点项目;北京市自然科学基金项目资助(KZ20041000501)

摘　　要：为提高车道线识别的实时性和可靠性,提出了一种基于改进Hough变换的车道线检测方法;在图像预处理时对不同光照图像进行分类处理,得到二值化图像;利用极角约束Hough变换进行车道线初始定位;根据前一帧图像信息使用基于动态ROI的Hough变换进行车道跟踪;算法加入了车道线检测失效判别模块,以提高检测的可靠性;由于该方法减少了图像空间中被投票的目标点数,缩小Hough变换的投票空间,在一定程度上提高了车道检测的实时性和稳定性;实验结果表明,在结构化道路上,对于不同的路况,算法均具有较好的实时性和鲁棒性。In order to enhance the real time and stability of lane detection, a method based on improved Hough Transform （HT） is proposed. The image classified by illumination is processed by different algorithm during image pretreatment. Lane is initially detected by HT in the area of polar angle constraints. In lane tracking, region of interest （ROD is established according to last image information and then using HT in the ROI. A judgment module that is responsible for estimating tracking results was used to improve the reliability. The improved HT reduces the total edge points, as a result the computational complexity is decreased and the efficiency is increased. For various kinds of lanes on most structural road, experiment results indicate that the method has good robustness and stability.

关 键 词：车道线检测 HOUGH变换 图像分类 感兴趣区域 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]