Local features for partial shape matching and retrieval
MM '11 Proceedings of the 19th ACM international conference on Multimedia
Discrete minimum distortion correspondence problems for non-rigid shape matching
SSVM'11 Proceedings of the Third international conference on Scale Space and Variational Methods in Computer Vision
Hierarchical matching of non-rigid shapes
SSVM'11 Proceedings of the Third international conference on Scale Space and Variational Methods in Computer Vision
A robust 3D interest points detector based on Harris operator
EG 3DOR'10 Proceedings of the 3rd Eurographics conference on 3D Object Retrieval
Learning the compositional structure of man-made objects for 3D shape retrieval
EG 3DOR'10 Proceedings of the 3rd Eurographics conference on 3D Object Retrieval
Evaluation of 3D interest point detection techniques
EG 3DOR'11 Proceedings of the 4th Eurographics conference on 3D Object Retrieval
Key-component detection on 3D meshes using local features
EG 3DOR'12 Proceedings of the 5th Eurographics conference on 3D Object Retrieval
3D point of interest detection via spectral irregularity diffusion
The Visual Computer: International Journal of Computer Graphics
Pose analysis using spectral geometry
The Visual Computer: International Journal of Computer Graphics
Key-components: detection of salient regions on 3D meshes
The Visual Computer: International Journal of Computer Graphics
| Hi-index | 0.00 |
This paper introduces a new method for extracting salient features from surfaces that are represented by triangle meshes. Our method extracts salient geometric feature points in the Laplace–Beltrami spectral domain instead of usual spatial domains. Simultaneously, a spatial region is determined as a local support of each feature point, which is correspondent to the “frequency” where the feature point is identified. The local shape descriptor of a feature point is the Laplace–Beltrami spectrum of the spatial region associated to the points which are stable and distinctive. Our method leads to the salient spectral geometric features invariant to spatial transforms such as translation, rotation, and scaling. The properties of the discrete Laplace–Beltrami operator make them invariant to isometric deformations and mesh triangulations as well. With the scale information transformed from the “frequency”, the local supporting region always maintains the same ratio to the original model no matter how it is scaled. This means that the spatial region is scale-invariant as well. Therefore, both global and partial matching can be achieved with these salient feature points. We demonstrate the effectiveness of our method with many experiments and applications.