Skip to Main Content
Frequently Asked Questions
Submit an ETD
Global Search Box
Need Help?
Keyword Search
Participating Institutions
Advanced Search
School Logo
Files
File List
osu1101240670.pdf (99.77 MB)
ETD Abstract Container
Abstract Header
Reconstruction and segmentation of 3D objects from point samples
Author Info
Goswami, Samrat
Permalink:
http://rave.ohiolink.edu/etdc/view?acc_num=osu1101240670
Abstract Details
Year and Degree
2004, Doctor of Philosophy, Ohio State University, Computer and Information Science.
Abstract
Computers often represent geometric shapes by discrete approximations. To capture such approximations, devices, such as range scanners, CMMs, sample the surface of the physical objects. These devices generate a set of unorganized points. The first part of this thesis describes a method of creating a replica of the original shape from a set of sample points even when the "quality'' of the point set is not good. The second part of the thesis addresses the problem of analyzing the "feature's" of the shape from point sample and using it to determine the similarity of two shapes. The problem of reconstructing a surface from a set of points dates back to 1980s. Since then, several algorithms have been proposed to solve this problem. Some of those algortihms provide theoretical guarantees about the quality of the reconstructed surface, in terms of topology and geometry. The applicability of these methods is limited in practice because the set of unorganized points, gathered through the sampling process, does not always follow the conditions assumed to prove the theoretical correctness of the algorithms. This problem is addressed in the first part of this thesis. Two methods are proposed to reconstruct a piecewise linear surface even when the "quality'' of the point set is not good. The second part of the thesis addresses the problem of segmenting an object into its salient features. First, we define "feature'' for continuous shapes and then we translate that definition into a discrete setting so that it can be computed efficiently from a set of points, sampled from the object boundary, using some well-known geometric structures. The feature-preserving segmentation of the object from its point sample gives a compact handle to represent it, which, in turn, is used to measure the similarity of two objects. The effectiveness of the proposed algorithms is demonstrated through numerous examples and experimental results.
Committee
Tamal Dey (Advisor)
Keywords
Delaunay
;
OCONE
;
Voronoi
;
POINT SAMPLES
Recommended Citations
Refworks
EndNote
RIS
Mendeley
Citations
Goswami, S. (2004).
Reconstruction and segmentation of 3D objects from point samples
[Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1101240670
APA Style (7th edition)
Goswami, Samrat.
Reconstruction and segmentation of 3D objects from point samples.
2004. Ohio State University, Doctoral dissertation.
OhioLINK Electronic Theses and Dissertations Center
, http://rave.ohiolink.edu/etdc/view?acc_num=osu1101240670.
MLA Style (8th edition)
Goswami, Samrat. "Reconstruction and segmentation of 3D objects from point samples." Doctoral dissertation, Ohio State University, 2004. http://rave.ohiolink.edu/etdc/view?acc_num=osu1101240670
Chicago Manual of Style (17th edition)
Abstract Footer
Document number:
osu1101240670
Download Count:
638
Copyright Info
© 2004, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.