Abstract
We provide a complete analysis of the geometry of N points in 1 image, employing a formalism in which multi-frame and multi-point geometries appear in symmetry: points and projections are interchangeable. We derive bilinear equations for 6 points, trilinear equations for 7 points, and quadrilinear equations for 8 points. The new equations are used to design new algorithms for the reconstruction of projective shape from many frames. Shape is represented by shape descriptors, which are sufficient for object recognition, and for the simulation of new images of the object. We further propose a linear shape reconstruction scheme which uses all the available data- all points and all frames - simultaneously. Unlike previous approaches, the equations developed here lead to direct and linear computation of shape, without going through the cameras' geometry.
Original language | English |
---|---|
Title of host publication | Computer Vision – ECCV 1996 - 4th European Conference on Computer Vision, Proceedings |
Editors | Bernard Buxton, Roberto Cipolla |
Publisher | Springer Verlag |
Pages | 217-227 |
Number of pages | 11 |
ISBN (Print) | 3540611231, 9783540611233 |
DOIs | |
State | Published - 1996 |
Event | 4th European Conference on Computer Vision, ECCV 1996 - Cambridge, United Kingdom Duration: 15 Apr 1996 → 18 Apr 1996 |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 1065 |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Conference
Conference | 4th European Conference on Computer Vision, ECCV 1996 |
---|---|
Country/Territory | United Kingdom |
City | Cambridge |
Period | 15/04/96 → 18/04/96 |
Bibliographical note
Publisher Copyright:© Springer-Verlag Berlin Heidelberg 1996.