An approach to shape matching based on forming a multiresolution set of shape signatures derived from full resolution signatures is described. The matching is performed by establishing associations between corresponding peaks in pairs of derived signatures at comparable resolutions and computing the peak matching error for these. This error is the weighted sum of differences in positions and other simple parameters for associated peaks. This matching process provides a computationally efficient initial match for subsequent refinement by localised conventional feature matching, for purposes such as image fusion. A sample set of shapes is matched using the approach to demonstrate its effectiveness.
|Number of pages||4|
|Journal||Proceedings - International Conference on Image Processing, ICIP|
|Publication status||Published - Dec 1994|
|Event||Proceedings of the 1994 1st IEEE International Conference on Image Processing. Part 3 (of 3) - Austin, TX, USA|
Duration: 13 Nov 1994 → 16 Nov 1994