基于tophat分割和曲线模型的三车道检测方法.pdfVIP

第42卷 第8期 北 京 工 业 大 学 学 报 Vol.42 No.8 2016年 8月 JOURNAL OF BEIJING UNIVERSITY OFTECHNOLOGY Aug. 2016 基于TopHat分割和曲线模型的三车道检测方法 段建民,战宇辰,刘冠宇 (北京工业大学北京市交通工程重点实验室,北京摇 100124) 摘摇 要:为了解决传统三车道检测过程中算法易受干扰、车道线拟合不准确、两侧车道误判等问题,提出了一种基 于顶帽算法(TopHat)分割和曲线模型的三车道检测方法. 利用车道线的形状和颜色特征,在图像预处理阶段提出 了一种变内核TopHat 的车道线分割算法. 在车道线识别阶段,首先,提出了一种基于加权最小二乘法(weighted least squares,WLS)的消失点拟合方法以约束霍夫变换;其次,在极坐标中以DBSCAN(density鄄based spatial clustering of applicationswith noise)聚类法对直线聚类并匹配三车道模板;再次,以该模板为基础建立车道线感兴趣区,在每 个感兴趣区内有哪些信誉好的足球投注网站并以三次曲线模型拟合车道线;最后,对于不确定的边侧车道,提出了一种基于随机投种法的边 侧车道可行驶性判定方法. 算法检测率以及漏检率结果显著优于传统三车道识别算法. 实验结果表明:该算法具 有良好的准确性及稳定性,更适用于三车道环境. 关键词:三车道检测;TopHat算法;消失点;车道线聚类;最小二乘法 中图分类号:TP391 文献标志码:A 文章编号:0254-0037(2016)08-1174-08 doi:10.11936/ bjutxb2015070090 Three鄄lane Detection Method Based on TopHat Segmentation and Curve Models DUANJianmin,ZHAN Yuchen,LIU Guanyu (Beijing Key Laboratory of Traffic Engineering,Beijing University of Technology,Beijing 100124,China) Abstract:In order to solve problems in the process of traditional three鄄lane detection,such aslow anti鄄 interference capacity, inaccuracy in lane fitting, and error identification of side lane, a three鄄lane detection method was proposed in this paper based on TopHat segmentation and curve models. A lane segmentation algorithm using variable鄄kernel TopHat was proposed as image pre鄄processing, by using shape features and color features of lane markings. For lane detection,firstly, a vanish鄄point fitting method based on WLS (weighted least squares) was proposed as a constraint of Hough transform. Secondly,straight lines were clustered in polar coordinates by