| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 536297 | Pattern Recognition Letters | 2015 | 9 Pages |
Abstract
•3D rotation invariants of Gaussian–Hermite moments are presented.•The new invariants can be indirectly constructed from the geometric invariants.•Better numerical stability of Gaussian–Hermite moments is demonstrated.
3D rotation invariants based on orthogonal Gaussian–Hermite moments are proposed in this paper. We present an elegant and easy theoretical derivation of them. At the same time we prove by experiments that the Gaussian–Hermite invariants have better numerical stability than the traditional invariants composed of geometric moments.
Related Topics
Physical Sciences and Engineering
Computer Science
Computer Vision and Pattern Recognition
Authors
Bo Yang, Jan Flusser, Tomáš Suk,