- Title
- Corners-based composite descriptor for shapes
- Creator
- Sajjanhar, Atul; Lu, Guojun; Zhang, Dengsheng; Zhou, Wanlei
- Date
- 2008
- Type
- Text; Conference paper
- Identifier
- http://researchonline.federation.edu.au/vital/access/HandleResolver/1959.17/70547
- Identifier
- vital:6535
- Identifier
-
https://doi.org/10.1109/CISP.2008.698
- Identifier
- ISBN:9780769531199
- Abstract
- In this paper, a composite descriptor for shape retrieval is proposed. The composite descriptor is obtained based upon corner-points and shape region. In an earlier paper, we proposed a composite descriptor based on shape region and shape contour, however, the descriptor was not effective for all perspective and geometric transformations. Hence, we modify the composite descriptor by replacing contour features with corner-points features. The proposed descriptor is obtained from Generic FourierDescriptors (GFD) of the shape region and the GFD ofthe corner-points. We study the performance of the proposed composite descriptor. The proposed method is evaluated using Item S8 within the MPEG-7 Still Images Content Set. Experimental results show that the proposed descriptor is effective.
- Publisher
- Sanya, China IEEE Computer Society
- Relation
- Proceedings of the First International Congress on Image and Signal Processing CISP2008 p. 714-718
- Rights
- This metadata is freely available under a CCO license
- Subject
- 0801 Artificial Intelligence and Image Processing; CBIR; Shape retrieval
- Reviewed
- Hits: 861
- Visitors: 892
- Downloads: 1
| Thumbnail | File | Description | Size | Format |
|---|