Abstract
A new feature-based method of stereo matching is proposed. The method matches curved segments in the images, using the minimum Hausdorff distance to find the best match for each curved segment. A neighborhood graph is used to propagate the disparity results to neighboring curved segments and to resolve ambiguous matchings. Finally, some experimental results obtained with the proposed method are presented.
| Original language | English |
|---|---|
| Pages | C-15-C-16 |
| State | Published - 1 Jan 1996 |
| Event | Proceedings of the 1996 12th Annual Symposium on Computational Geometry - Philadelphia, PA, USA Duration: 24 May 1996 → 26 May 1996 |
Conference
| Conference | Proceedings of the 1996 12th Annual Symposium on Computational Geometry |
|---|---|
| City | Philadelphia, PA, USA |
| Period | 24/05/96 → 26/05/96 |
ASJC Scopus subject areas
- Theoretical Computer Science
- Geometry and Topology
- Computational Mathematics