基于模糊集理论的车道标志线精确检测  

Lane Marking Line Precision Detection Based on Fuzzy Theory

作　　者：李庆武[1] 石丹[1] 孟凡[1] 徐立中[1] 

机构地区：[1]河海大学,南京210098

出　　处：《交通与计算机》2007年第6期1-3,共3页Computer and Communications

基　　金：国家863计划项目资助(批准号:2007AA11Z227);江苏省社会发展科技项目资助(批准号:BS2007058);常州国家高新区科技攻关项目资助(批准号:XE120060408)

摘　　要：为了解决复杂路面条件下车道标志线的检测问题,提出了一种基于模糊集理论的车道标志线精确检测方法。采用阈值分割、腐蚀、边缘检测等技术对交通图像进行处理;通过Hough变换粗略提取位于各车道标志线及其附近的所有亮点像素作为待分类的像素点,构成样本集;按照模糊聚类分析准则得到直线的斜率和截距,实现车道标志线的精确检测。实验证明了该方法的精确性和有效性。In order to realize the detection of lane marking line in complex road conditions, a detection method of lane markings based on fuzzy theory was introduced. The method first utilized different image pre-processir_g techniques to process the road images, and then used Hough Transform to get the initial lane marking line, and finally used fuzzy clustering analysis to get the slope and intercept of lane marking line. The experiment results of lane detection from images of real road scenes validate the acccracy and reliability of the proposed method.

关 键 词：车道标志线 HOUGH变换 模糊集 模糊聚类 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]