Segmentation methods for smooth point regions of conventional engineering objects

被引:58
作者
Benko, P [1 ]
Várady, T [1 ]
机构
[1] Comp & Automat Res Inst, Geometr Modeling Lab, H-1111 Budapest, Hungary
关键词
reverse engineering; segmentation; local indicators; statistical similarity tests;
D O I
10.1016/S0010-4485(03)00159-3
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
The purpose of reverse engineering is to convert a large point cloud into an accurate, fair and consistent CAD model. The class of conventional engineering objects comprises, by a priori assumption, objects bounded only by simple, analytic surfaces, linear extrusion surfaces and surfaces of revolution. In this case it is possible to create a CAD model from point clouds with a minimal amount of user assistance. The key issue is segmentation, i.e. how to subdivide the point Cloud into smaller regions, where each of which can be approximated by a single surface. Segmentation is relatively simple, if regions are bounded by sharp edges and small blends problems arise when smoothly connected regions need to be separated. The direct segmentation method described in this paper is based on a particular hierarchy of tests, by means of which a large point Cloud can be robustly split into smaller and smaller subregions until no further subdivision is needed. Tests are given using different types of indicators. These are based on local estimates of various geometric and statistical quantities several examples are given to show their effect. The structure of the smooth, multiple regions obtained provides the basis for constrained surface fitting, which completes the final model building phase. (C) 2003 Elsevier Ltd. All rights reserved.
引用
收藏
页码:511 / 523
页数:13
相关论文
共 19 条
[1]  
ALBOUL L, 1997, MATH SURFACES, V7, P309
[2]   Direct segmentation of smooth, multiple point regions [J].
Benkö, P ;
Várady, T .
GEOMETRIC MODELING AND PROCESSING: THEORY AND APPLICATIONS, PROCEEDINGS, 2002, :169-178
[3]   Constrained fitting in reverse engineering [J].
Benko, P ;
Kós, G ;
Várady, T ;
Andor, L ;
Martin, R .
COMPUTER AIDED GEOMETRIC DESIGN, 2002, 19 (03) :173-205
[4]   Algorithms for reverse engineering boundary representation models [J].
Benko, P ;
Martin, RR ;
Várady, T .
COMPUTER-AIDED DESIGN, 2001, 33 (11) :839-851
[5]  
BENKO P, 2000, MATH SURFACES, V9, P70
[6]  
Besl P.J., 1988, Surfaces in Range Image Understanding, V1th
[7]   SEGMENTATION THROUGH VARIABLE-ORDER SURFACE FITTING [J].
BESL, PJ ;
JAIN, RC .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1988, 10 (02) :167-192
[8]  
CSAKANY P, 2000, MATH SURFACES, V9, P19
[9]   High level CAD model acquisition from range images [J].
Fitzgibbon, AW ;
Eggert, DW ;
Fisher, RB .
COMPUTER-AIDED DESIGN, 1997, 29 (04) :321-330
[10]   EXTRACTING A VALID BOUNDARY REPRESENTATION FROM A SEGMENTED RANGE IMAGE [J].
HOOVER, A ;
GOLDGOF, D ;
BOWYER, KW .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1995, 17 (09) :920-924