Abstract
In this paper we present new algorithms that exploit compressed image data to achieve coarse stereo reconstructions in a realtime environment. Others have shown large gains from processing image streams in compressed form, and we extend those results to address the stereo correspondence problem. We show that it is possible to obtain stereo matches between two frames of a compressed video stream in approximately the same time it takes to fully decompress just one image.
Original language | English |
---|---|
Title of host publication | Computer Vision - ACCV 1998 - 3rd Asian Conference on Computer Vision, Proceedings |
Editors | Roland Chin, Ting-Chuen Pong |
Pages | 558-565 |
Number of pages | 8 |
DOIs | |
State | Published - 1997 |
Event | 3rd Asian Conference on Computer Vision, ACCV 1998 - Hong Kong, China Duration: Jan 8 1998 → Jan 10 1998 |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 1351 |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Conference
Conference | 3rd Asian Conference on Computer Vision, ACCV 1998 |
---|---|
Country/Territory | China |
City | Hong Kong |
Period | 1/8/98 → 1/10/98 |
Bibliographical note
Publisher Copyright:© Springer-Verlag Berlin Heidelberg 1997.
Keywords
- Compresseddomain
- Compression
- Depth reconstruction
- Real-time
- Stereo
ASJC Scopus subject areas
- Theoretical Computer Science
- General Computer Science