摘要
为了在移动车载激光扫描点云数据中尽可能完整准确地提取道路边线,本文根据道路边界在点云场景中分布特征,提出了一种自动化的道路边线提取方法。首先,为消除无用点对道路边线提取的影响,在原始渐进式形态学滤波算法的基础上提出了一种改进的点云滤波算法,用于非地面点滤波,从而提高了点云滤波的运算效率和精度;其次,将地面点投影至二维图像上,并通过直线段检测算法(LSD)得到道路边线直线段;最后,通过直线连接和直线特征匹配实现最终的道路边线提取。通过对两组道路点云数据进行算法验证,结果表明,本文算法提取道路边线的完整率、准确率以及F-Measure均达到90%以上,验证了本文方法的有效性与适应性。
In order to extract the road edge from the mobile vehicle-borne laser scanning point cloud data as completely and accurately as possible,an automatic road edge extraction method is proposed according to the distribution characteristics of the road edge in the point cloud scene.Firstly,in order to eliminate the influence of useless points on road edge extraction,an improved point cloud filte-ring algorithm is proposed based on the original progressive morphological fltering algorithm to realize non-ground point filtering,which improves the operation efficiency and accuracy of point cloud filtering;secondly,the ground points are projected onto the two-dimensional image and Line Segment Detector(LSD)algorithm is carried out to obtain the straight line segment of the road edge;fi-nally,the final road edge is extracted by line connection and line feature matching.Two groups of road point cloud data are used to verify the algorithm.The results show that the integrity,accuracy and F-measure of the road edge extraction results of the two groups of experimental data are more than 90%,which verifies the effectiveness and adaptability of this method.
作者
何慧
杜国政
HE Hui;DU Guozheng(Zhejiang Institute of Surveying and Mapping Science and Technology,Hangzhou 310012,China)
出处
《测绘与空间地理信息》
2024年第7期207-210,214,共5页
Geomatics & Spatial Information Technology
关键词
车载激光扫描
点云数据
改进渐进式形态学滤波
矢道路边线
直线特征匹配
vehicle-borne laser scanning
point cloud data
improved progressive morphological filtering
arrow road edges
line fea-ture matching
