Title page for ETD etd-10222009-124949


Type of Document Master's Thesis
Author Mainguy, Yves
URN etd-10222009-124949
Title A robust variable order facet model for image data
Degree Master of Science
Department Computer Science and Applications
Advisory Committee
Advisor Name Title
Birch, Jeffrey B. Committee Co-Chair
Watson, Layne T. Committee Co-Chair
Ehrich, Roger W. Committee Member
Keywords
  • Computer vision
Date of Defense 1991-12-05
Availability restricted
Abstract

The underlying piecewise continuous surface of a digital image can be estimated through robust statistical procedures. This thesis contains a systematic Monte Carlo study of M-estimation and LMS estimation for image surface approximation, an examination of the merits of postprocessing and tuning various parameters in the robust estimation procedures, and a new robust variable order facet model paradigm. Several new goodness of fit measures are introduced, and systematically compared. It is shown that the M-estimation tuning parameters are not crucial, postprocessing is cheap and well worth the cost, and the robust variable order facet model algorithm (using M-estimation, new statistical goodness of fit measures, and postprocessing) manages to retain most of the statistical efficiency of Mestimation yet displays good robustness properties, and preserves the main geometric features of an image surface: step edges, roof edges and corners.

Files
  Filename       Size       Approximate Download Time (Hours:Minutes:Seconds) 
 
 28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access 
[VT] LD5655.V855_1991.M358.pdf 2.73 Mb 00:12:39 00:06:30 00:05:41 00:02:50 00:00:14
[BTD] next to an author's name indicates that all files or directories associated with their ETD are accessible from the Virginia Tech campus network only.

Browse All Available ETDs by ( Author | Department )

dla home
etds imagebase journals news ereserve special collections
virgnia tech home contact dla university libraries

If you have questions or technical problems, please Contact DLA.