DMCA
Attentive visual motion processing: computations in the log-polar plane (1995)
Cached
Download Links
- [www.cis.upenn.edu]
- [www.ks.informatik.uni-kiel.de]
- DBLP
Other Repositories/Bibliography
| Venue: | Accepted in "Computing, Archives for Informatic and Numerical Computation |
| Citations: | 5 - 1 self |
BibTeX
@INPROCEEDINGS{Daniilidis95attentivevisual,
author = {Kostas Daniilidis},
title = {Attentive visual motion processing: computations in the log-polar plane},
booktitle = {Accepted in "Computing, Archives for Informatic and Numerical Computation},
year = {1995},
pages = {1--20}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Attentive vision is characterized by selective sensing in space and time as well as selective processing with respect to a speci c task. Selection in space involves the splitting of the visual eld in a high resolution area- the fovea- and a space-variant resolution area-the periphery. Both in neurobiology and in robot vision, models of the resolution decrease towards the image boundaries have been established. The most convincing model is the theory of log-polar mapping where very high data compression rates are achieved. In combination with the complexity reduction we believe that the log-polar mapping has further computational advantages which we elaborate in this study. Based on the optical ow we study the computation of 3D motion and structure globally and locally. We present a global method to compute the Focus of Expansion in the case of pure translation. By xating on an object we show how to estimate ego motion in the presence of translation and rotation of the observer from the ow in the log-polar periphery. Then, we turn to local di erential computations and we establish both approximate and exact expressions for the time to collision. 1
Keyphrases
log-polar plane attentive visual motion processing log-polar mapping high resolution area ego motion pure translation convincing model image boundary complexity reduction space-variant resolution area-the periphery selective processing global method selective sensing attentive vision resolution decrease optical ow high data compression rate visual eld local di erential computation log-polar periphery speci task exact expression computational advantage robot vision
