Name
Abstract | ||
|---|---|---|
In this paper, we propose an a-ne-invariant method for describing and matching curves. This is important since a-ne transformations are often used to model perspective distortions. More speciflcally, we propose a new deflnition of the shape of a curve that character- izes a curve independently of the efiects introduced by a-ne distortions. By combining this deflnition with a rotation-invariant shape descriptor, we show how it is possible to describe a curve in an intrinsically a-ne- invariant manner. To validate our procedure we built a database of shapes subject to perspective distortions and plotted the precision-recall curve for this dataset. Finally an application of our method is shown in the context of wide baseline matching. imaged object is planar and the camera optical center is far enough from such plane (3). This explains why we want to develop a methodology to compare curves related by an a-ne transformation. In this paper, we propose a method that will lead us to a compact and intrinsically a-ne invariant shape de- scriptor that is shown to be efiective in matching curves that are a-ne equivalent. Previously, the curvature scale space (CSS) descriptor (4) of MPEG-7 has been shown to be robust under perspective distortion. How- ever, this descriptor is not intrinsically a-ne invariant. In order to achieve robustness to a-ne transformation, it uses the a-ne length curve parametrization method that requires derivatives up to the fourth order. In a digital implementation, higher-order derivatives (a) are di-cult to implement, and (b) may cause instability in the presence of noise. Our method merits from not using any derivatives. |
Year | DOI | Venue |
|---|---|---|
2004 | 10.1109/ICIP.2004.1421754 | Image Processing, 2004. ICIP '04. 2004 International Conference |
Keywords | Field | DocType |
image matching,visual databases,affine transformation,affine-invariant-method,matching curve,perspective distortion,precision-recall curve,rotation-invariant shape descriptor,wide baseline matching | Affine transformation,Computer vision,Affine shape adaptation,Pattern recognition,Curve matching,Computer science,Image matching,Affine invariant,Artificial intelligence | Conference |
Volume | ISSN | ISBN |
5 | 1522-4880 | 0-7803-8554-3 |
Citations | PageRank | References |
19 | 0.91 | 6 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Marco Zuliani | 1 | 94 | 7.67 |
| Sitaram Bhagavathy | 2 | 114 | 9.82 |
| B. S. Manjunath | 3 | 7561 | 783.37 |
| Charles S. Kenney | 4 | 292 | 33.83 |