DocumentCode :
2491396
Title :
Principal patches-a viewpoint-invariant surface description
Author :
Sinha, Sarvajit S. ; Besl, P.J.
Author_Institution :
Artificial Intelligence Lab., Michigan Univ., Ann Arbor, MI, USA
fYear :
1990
fDate :
13-18 May 1990
Firstpage :
226
Abstract :
The authors deal with the principal patch representation of visible surfaces in a scene, an algorithm for creating a stable surface model in this representation, and an algorithm for extending this model into subsequent views. The representation is based on the fact that the lines of minimum and maximum curvature on a generic smooth surface are orthogonal at almost every point. An engineering solution is presented for creating global principal patch networks. By computing the lines of curvature from a range image, it will be possible to construct a mesh data structure and repatch the surface with tensor product surfaces that are virtually free of tensor product artifacts and allow application of intersurface continuity algorithms. The authors focus on the computational difficulties encountered in building a quadrilateral mesh based on integrated lines of curvature
Keywords :
computer graphics; computerised picture processing; data structures; generic smooth surface; global principal patch networks; mesh data structure; principal patch representation; scene; stable surface model; tensor product surfaces; viewpoint-invariant surface description; visible surfaces; Artificial intelligence; Buildings; Computer science; Computer vision; Data structures; Laboratories; Layout; Libraries; Merging; Tensile stress;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Robotics and Automation, 1990. Proceedings., 1990 IEEE International Conference on
Conference_Location :
Cincinnati, OH
Print_ISBN :
0-8186-9061-5
Type :
conf
DOI :
10.1109/ROBOT.1990.125977
Filename :
125977
Link To Document :
بازگشت