Name
Abstract | ||
|---|---|---|
A fast algorithm for similarity registration for shapes with various topologies is put forward in this paper. Fourier transform and Geometric moments are explored here to calculate the rotation, scaling and translation parameters to register two shapes by minimizing a dissimilarity measure introduced in the literature. Shapes are represented by signed distance functions. In comparison with the algorithms in the literature, the algorithm proposed here demonstrates superior performance for the registration of two shapes with various topologies as well as two shapes, each containing various and different numbers of shape components. The registration process using this algorithm is robust in comparison with the shape registration algorithms in the literature and is as fast as a couple of FFTs. |
Year | Venue | Field |
|---|---|---|
2012 | ADVANCES IN VISUAL COMPUTING, ISVC 2012, PT I | Computer vision,Pattern recognition,Computer science,Signed distance function,Network topology,Fourier transform,Artificial intelligence,Scaling,Geometric moments,Shape analysis (digital geometry) |
DocType | Volume | ISSN |
Conference | 7431 | 0302-9743 |
Citations | PageRank | References |
1 | 0.37 | 10 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Sasan Mahmoodi | 1 | 94 | 17.37 |
| Muayed S. Al-Huseiny | 2 | 18 | 2.17 |
| Mark S. Nixon | 3 | 3080 | 304.45 |