车载激光点云中杆状目标自动提取
发布时间：2018-11-19 07:25
【摘要】：针对车载激光扫描数据中杆状目标点云自动提取易受临近地物干扰的问题,提出了一种基于扫描线聚类判别的杆状目标自动提取方法。该方法根据不同地物在扫描线数据中的形态,设计了从单条扫描线到多条扫描线的数据处理流程。首先对单条扫描线上的点进行距离聚类,得到不同的地物横剖面点集;然后根据点集内点的数量和杆状目标在扫描线上的分布形态,对聚类点集进行杆状目标种子点集判别;最后在疑似杆状目标种子点集的平面投影位置处,对多条扫描线进行点集聚类和去噪处理得到杆状目标点云。实验分析表明,针对不同的激光扫描数据,该方法能有效降低临近地物干扰,提取出道路两侧的杆状目标。
[Abstract]:In order to solve the problem that point cloud extraction from vehicle laser scanning data is easy to be interfered with by adjacent objects, a method of automatic extraction of rod objects based on scanning line clustering discrimination is proposed. According to the shape of different ground objects in scanning line data, the data processing flow from single scan line to multiple scanning line is designed. Firstly, distance clustering of points on a single scan line is carried out to obtain different points set of cross-section of ground objects, and then, according to the number of points in the point set and the distribution form of the rod object on the scanning line, the cluster point set is distinguished from the seed point set of the rod object. Finally, at the plane projection position of the seed point set of the suspected rod target, the point cloud of the rod target is obtained by the point clustering and de-noising processing of several scanning lines. The experimental results show that the proposed method can effectively reduce the interference of adjacent objects and extract the rod-shaped targets on both sides of the road according to the different laser scanning data.
【作者单位】： 山东科技大学测绘科学与工程学院;
【基金】：海洋公益性行业科研专项(201305034-1) 国家重大仪器设备开发专项(2013YQ120343) 测绘公益性行业科研专项(201512034) 山东科技大学人才引进科研启动基金(2016RCJJ004)
【分类号】：P235
本文编号：2341543
[Abstract]:In order to solve the problem that point cloud extraction from vehicle laser scanning data is easy to be interfered with by adjacent objects, a method of automatic extraction of rod objects based on scanning line clustering discrimination is proposed. According to the shape of different ground objects in scanning line data, the data processing flow from single scan line to multiple scanning line is designed. Firstly, distance clustering of points on a single scan line is carried out to obtain different points set of cross-section of ground objects, and then, according to the number of points in the point set and the distribution form of the rod object on the scanning line, the cluster point set is distinguished from the seed point set of the rod object. Finally, at the plane projection position of the seed point set of the suspected rod target, the point cloud of the rod target is obtained by the point clustering and de-noising processing of several scanning lines. The experimental results show that the proposed method can effectively reduce the interference of adjacent objects and extract the rod-shaped targets on both sides of the road according to the different laser scanning data.
【作者单位】： 山东科技大学测绘科学与工程学院;
【基金】：海洋公益性行业科研专项(201305034-1) 国家重大仪器设备开发专项(2013YQ120343) 测绘公益性行业科研专项(201512034) 山东科技大学人才引进科研启动基金(2016RCJJ004)
【分类号】：P235
【相似文献】
相关期刊论文 前4条
1 杨崇俊;任应超;李津平;;基于单调链的Red/Blue扫描线求交算法[J];武汉大学学报(信息科学版);2006年09期
2 潘俊;王密;李琪;;基于扫描线填充的快速镶嵌算法[J];测绘信息与工程;2006年05期
3 汪志章;;SPOT系统及制图应用(三)[J];铁路航测;1982年03期
4 管海燕;张剑清;邓非;钟良;;基于扫描线的城区机载激光扫描数据滤波算法研究[J];测绘通报;2007年12期
相关会议论文 前1条
1 任应超;李建平;;基于单调链的Red/Blue扫描线求交算法[A];中国地理信息系统协会第八届年会论文集[C];2004年
，本文编号：2341543
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2341543.html
