局部形状特征概率混合的半自动三维点云分类

doi:10.3785/j.issn.1008-9497.2017.01.001

摘要
图/表
参考文献
相关文章 (0)

全文: PDF (7218 KB) HTML( )
输出: BibTeX | EndNote (RIS)

摘要三维激光扫描获取的点云数据可用于数字城市建设、三维模型获取、场景分析与物体测量等领域.但因遮挡和噪声的影响，加之扫描场景复杂，采样精度受限，使得不能直接运用经典的曲面和三维空间理论对点云数据进行有效分析和处理.分类是点云数据预处理的重要方式之一.提取近邻四面体体积、近邻法向量差异度、主方向差异度和主曲率值4个局部形状特征，采用概率混合策略构建了一种点云数据的半自动分类方法，可实现平面点集、柱面点集和其他点集的有效区分.其中，概率混合策略是依据近邻点平均距离和单指标类别一致程度估计每个特征推断形状的概率，通过混合加权，依据概率赋权函数最大值准则进行局部形状推断.可实现用户交互，以便处理不同扫描尺度和精度的点云数据.采用本文方法对模拟生成的点云、单棵树木点云、街道场景点云、旷野自然场景扫描点云以及航空机载扫描点云等多组数据进行了实验，结果表明，基于局部形状特征的概率混合方法对各种点云数据均具有良好的分类效果.

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	李红军
	刘欣莹
	张晓鹏
	严冬明

关键词 ：点云分类, 局部形状推断, 概率混合, 法向量差异, 主曲率, 大规模场景分析

Abstract：Point clouds captured by three dimensional scanner have been used in many fields, including modeling of digital cities, acquisition of three dimensional shapes, scene analysis and object measuring. However, due to the limitation of the sampling process and the complexity of scanned scenes, most traditional methods of surface modeling and three dimensional space analysis cannot work effectively when dealing with the point cloud data. Classification is therefore an important way for point cloud preprocess. Four features, namely the volume of a tetrahedron constructed by 4 neighboring points, the deviation of normal directions of neighboring points, the deviation of principal directions of neighboring points, and the values of principal curvature, are mixed with probabilities for semi-automatic classification of the three dimensional point cloud data. With the new method, a point cloud is to be divided into three classes:plane points, cylinder points and other points. The initial classification result is labeled according to its single shape feature value. The probability mixture is completed by estimating the probability of inferring a shape from a local point set with respect to each feature, generating a mixture with weighted sum, and maximizing the mixture probability function, while the probability is estimated with the average distance between a point and its neighbor points together with the consistency ratio of initial labels of the point to its neighbors. User interactions are invoked to make the choice of classification thresholds and the setting of weights, which is helpful when dealing with point cloud with different space scale and scanning point resolution. Experiments show that the proposed method works well for various kinds of point cloud data sets, including point clouds generated by simulation, and those corresponding to a single pine tree, a street scene, a country scene, and an airborne big scene.

Key words： point cloud classification local shape inference probability mixture normal difference principal curvature large scale scene analysis

收稿日期: 2016-07-25

ZTFLH:

TP311

基金资助:国家自然科学基金资助项目（61372190，61372168，61572502，61571439）；国家863计划课题项目（2015AA016402）

作者简介: 李红军(1969-),ORCID:http://orcid.org/0000-0003-1102-6225,男,博士,副教授,主要从事计算几何、计算机图形学研究,E-mail:lihongjun69@bjfu.edu.cn.

引用本文:

李红军, 刘欣莹, 张晓鹏, 严冬明. 局部形状特征概率混合的半自动三维点云分类[J]. 浙江大学学报（理学版）, 2017, 44(1): 1-9.
LI Hongjun, LIU Xinying, ZHANG Xiaopeng, YAN Dongming. A semi-automatic 3D point cloud classification method based on the probability mixture of local shape features. Journal of ZheJIang University(Science Edition), 2017, 44(1): 1-9.

链接本文:

http://www.zjujournals.com/sci/CN/10.3785/j.issn.1008-9497.2017.01.001 或 http://www.zjujournals.com/sci/CN/Y2017/V44/I1/1

[1] 伍龙华,黄惠.点云驱动的计算机图形学综述[J].计算机辅助设计与图形学学报,2015,27(8):1341-1353. WU L H, HUANG H. Survey on points-driven computer graphics[J]. Journal of Computer-Aided Design & Computer Graphics, 2015, 27(8):1341-1353.
[2] 李慧盈,李文辉,陈圣波.一种机载雷达点云数据的快速分类方法[J].吉林大学学报:地球科学版,2010(5):1205-1210. LI H Y, LI W H, CHEN S B. A method for classify point clouds of airborne laser scanning[J]. Journal of Jilin University:Earth Science Edition, 2010(5):1205-1210.
[3] 李文宁,张爱武,王书民,等.地面激光点云阶层式分类方法[J].计算机辅助设计与图形学学报,2015,27(8):1555-1561. LI W N, ZHANG A W, WANG S M, et al. Hierarchical classification method for terrestrial laser point clouds[J]. Journal of Computer-Aided Design & Computer Graphics, 2015, 27(8):1555-1561.
[4] 吴永华,胡以华,顾有林,等.激光点云分类算法的探讨与展望[J].红外与激光工程,2010,39(增刊):147-151. WU Y H, HU Y H, GU Y L, et al. Discussion and prospect for classification algorithm of laser point cloud[J]. Infrared and Laser Engineering,2010,39(supp):147-151.
[5] HOUGH P. Method and means for recognizing complex patterns:3069654[P].1962-12-18.
[6] FISCHLER M A, BOLLES R C. Random sample consensus:A paradigm for model fitting with applications to image analysis and automated cartography[J]. Communications of the ACM,1981,24(6):381-395.
[7] BOLLES R C, FISCHLER M A.A ransack-based approach to model fitting and its application to finding cylinders in range data[C]//Proceedings of the 7th International Joint Conference on Artificial Intelligence, San Francisco:Morgan Kaufmann Publishers Inc,1981:637-643.
[8] CHAPERON T, GOULETTE F.Extracting cylinders in full 3-d data using a random sampling method and the gaussian image[C]//VMV01.Stuttgart:Aka Gmbh,2001:35-42.
[9] SCHNABEL R, WAHL R, KLEIN R. Shape detection in point clouds[J]. Computer Graphics Technical Reports,2006,CG-2006/2.
[10] SCHNABEL R, WESSEL R, WAHL R, et al. Shape recognition in 3d point-clouds[C]//The 16th International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision. Plzen:UNION Agency-Science Press,2008, 8.
[11] NIEUWENHUISEN M, STVCKLER J, BERNER A, et al. Shape-primitive based object recognition and grasping[C]//Proceedings of 7th German Conference on Robotics. Munich:VDE, 2012:1-5.
[12] LARI Z, HABIB A, KWAK E. An adaptive approach for segmentation of 3D laser point cloud[J]. ISPRS Workshop Laser Scanning, Remote Sensing and Spatial Information Sciences,2011,XXXVⅢ-5/W12:103-108.
[13] NING X, ZHANG X, WANG Y, et al. Segmentation of architecture shape information from 3D point cloud[C]//Proceedings of the 8th International Conference on Virtual Reality Continuum and Its Applications in Industry. New York:ACM, 2009:127-132.
[14] NAN L, XIE K, SHARF A. A search-classify approach for cluttered indoor scene understanding[J]. ACM Transactions on Graphics (TOG), 2012, 31(6):137.
[15] RABBANI T, VAN DEN HEUVEL F, VOSSELMANN G. Segmentation of point clouds using smoothness constraint[J]. International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2006, 36(5):248-253.
[16] LI H, ZHANG X, JAEGER M, et al. Segmentation of forest terrain laser scan data[C]//Proceedings of the 9th ACM SIGGRAPH Conference on Virtual-Reality Continuum and its Applications in Industry. New York:ACM, 2010:47-54.
[17] DAI M, LI H, ZHANG X. Tree modeling through range image segmentation and 3d shape analysis[C]//Advances in Neural Network Research and Applications. Springer/Berlin/Heidelberg:Springer, 2010:413-422.
[18] HEBEL M, STILLA U. Pre-classification of points and segmentation of urban objects by scan line analysis of airborne LIDAR data[J]. International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2008, 37(B3a):105-110.
[19] YU F, XIAO J, FUNKHOUSER T. Semantic alignment of LIDAR data at city scale[C]//Computer Vision and Pattern Recognition (CVPR). Los Alamitos:IEEE Computer Society,2015:1722-1731.
[20] DEMANTKE J, MALLET C, DAVID N, et al. Dimensionality based scale selection in 3D lidar point clouds[J]. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2011, 38(Part 5):W12.
[21] 李海亭,肖建华,李艳红,等.机器学习在车载激光点云分类中的应用研究[J].华中师范大学学报:自然科学版,2015,49(3):460-464. LI H T, XIAO J H, LI Y H, et al. Application of machine learning in the vehicle-borne laser point cloud extraction[J].Journal of Central China Normal University:Natural Sciences, 2015, 49(3):460-464.
[22] TOPONOGOV V A. Differential Geometry of Curves and Surfaces:A Concise Guide[M]. Boston:Springer Science & Business Media,2006.
[23] MOUNT D M. ANN Programming Manual[EB/OL].[2009-12-14]. http link:http://www.cs.umd.edu/~mount/ANN/Files/1.1,2006,1.
[24] CHENG Z L, ZHANG X P. Estimating differential quantities from point cloud based on a linear fitting of normal vectors[J]. Science in China(Ser F):Information Sciences,2009,52(3):431-444.
[25] WEINMANN M, JUTZI B, HINZ S, et al. Semantic point cloud interpretation based on optimal neighborhoods, relevant features and efficient classifiers[J]. ISPRS Journal of Photogrammetry and Remote Sensing,2015,105:286-304.
[26] YAN D M, WINTZ J, MOURRAIN B, et al. Efficient and robust reconstruction of botanical branching structure from laser scanned points[C]//CAD/Graphics, Los Alamitos:IEEE Computer Society,2009:572-575.
[27] MUNOZ D, BAGNELL J A, VANDAPEL N, et al. Contextual classification with functional max-margin markov networks[C]//Computer Vision and Pattern Recognition(CVPR2009).Los Alamitos:IEEE Computer Society,2009:975-982.
[28] AIJAZI A K, CHECCHIN P, TRASSOUDAINE L. Segmentation based classification of 3D urban pointclouds:A super-voxel based approach with evaluation[J]. Remote Sensing,2013,5(4):1624-1650.
[29] BRODU N, LAGUE D. 3D terrestrial lidar data classification of complex natural scenes using a multi-scale dimensionality criterion:Applications in geomorphology[J]. ISPRS Journal of Photogrammetry and Remote Sensing,2012,68:121-134.
[30] SHAPOVALOV R, VELIZHEV A, BARINOVA O. Non-associative markov networks for point cloud classification[C]//Photogrammetric Computer Vision and Image Analysis. Paris:出版社,2010.
[31] SHAPOVALOV R, VELIZHEV A. Cutting-plane training of non-associative markov network for 3D point cloud segmentation[C]//3D Imaging, Modeling, Processing, Visualization and Transmission (3DIMPVT). Paris:ISPRS,2011:1-8.
[32] YAN D M, LIU Y, WANG W. Quadric surface extraction by variational shape approximation[M]//Geometric Modeling and Processing-GMP 2006. Berlin:Springer,2006:73-86.