A General Approach for Egomotion Estimation with Omnidirectional Images
OMNIVIS '02 Proceedings of the Third Workshop on Omnidirectional Vision
Head movement estimation for wearable eye tracker
Proceedings of the 2004 symposium on Eye tracking research & applications
Modelling and accuracy estimation of a new omnidirectional depth computation sensor
Pattern Recognition Letters
| Hi-index | 0.00 |
In this paper, we demonstrate that all single viewpoint catadioptric projections are equivalent to the composition of central projection to the sphere followed by a point projection from the sphere to an image plane. Special cases of this equivalence are parabolic projection, for which the second map is a stereographic projection, and perspective projection, for which the second map is central projection.We also show that two projections are dual by the mapping, which takes conics to their foci. The foci of line images are points of another, dual, and catadioptric projection; and vice versa, points in the image are foci of lines in the dual projection. They are dual because the mapping preserves incidence relationships.Finally, we show some applications of the theory presented above. We present a general algorithm for calibration of a catadioptric system with lines in a single view.