| 576 | 44 | 26 |
| 下载次数 | 被引频次 | 阅读次数 |
在原始测量获取的点云数据中,除了目标数据外,还有大量的噪声数据。噪声往往无规律地分布在目标物体周围,难以用统一数学模型区分。基于密度的聚类算法将簇定义为密度相连的点的最大集合,能发现任意形状、大小的类簇,将该算法应用在点云去噪中,能将密度分布连续点进行聚类,从中提取出目标点云。
Abstract：There are lots of noise data in the raw data except the target data. And the noise data always distribute around the target object irregularly, it is impossible to build a math model to make a distinction between the noise data and target data. The cluster is defined as the maximum set of density-collected in Density-based clustering algorithms, it can discover arbitrary shaped or sized cluster. To apply this algorithm it the noise removing of point clouds, can make the continuous distributed points as a cluster, and then extract the target point clouds.
[1]徐慧朴.逆向工程中基于离散点云的噪声预处理及模型重构技术研究[D].大连:大连海事大学,2009
[2]毛国君,段立娟,王实,等.数据挖掘原理与算法[M].北京:清华大学出版社,2005
[3]Leandro BORNAZ,Chiara PORPORATO.3D High AccuracySurvey and Modeling of One of AOSTA'S AnthropomorphicStelae[J].XXI International CIPA Symposium,2007,(10):176-181
[4]Huipu Xu,Ying Hu.Unilateral Left-side Quartile Algorithm Bas-ed Data Processing Scheme for 3D Scattered Point Data[J].Inter-national Journal of the Physical Sciences,2010,(5):1641-1646
[5]A.Adami,L.Fregonese,F.Guerra,et al.Digital Representationsand Analysis of Deformations Induced in Map Supporting Mater-ials[J].XXI International CIPA Symposium,2007,(10):7-11
[6]O.Etxaniz,E.Solaberrieta.Digital Modelling of a Human Skull[J].Journal of Achievements in Materials and Manufacturing En-gineering,2008,(27):55-58
[7]刘艳丰.基于kd-tree的点云数据空间管理理论与方法[D].长沙:中南大学,2009
基本信息:
中图分类号:P225.2
引用信息:
[1]张巧英,陈浩,朱爽.密度聚类算法在连续分布点云去噪中的应用[J].地理空间信息,2011,9(06):101-104+2+1.
2011-12-28
2011-12-28