Multiscale Shape Description with Laplacian Profile and Fourier Transform

Evanthia Mavridou 1, 2 James L. Crowley 1, 2 Augustin Lux 2, 1
1 LIG - Laboratoire d'Informatique de Grenoble
2 PRIMA - Perception, recognition and integration for observation of activity
Inria Grenoble - Rhône-Alpes, UJF - Université Joseph Fourier - Grenoble 1, INPG - Institut National Polytechnique de Grenoble , CNRS - Centre National de la Recherche Scientifique : UMR5217
Abstract : We propose a new local multiscale image descriptor of vari-able size. The descriptor combines Laplacian of Gaussian values at dif-ferent scales with a Radial Fourier Transform. This descriptor provides a compact description of the appearance of a local neighborhood in a manner that is robust to changes in scale and orientation. We evaluate this descriptor by measuring repeatability and recall against 1-precision with the Affine Covariant Features benchmark dataset and as well as with a set of textureless images from the MIRFLICKR Retrieval Evalu-ation dataset. Experiments reveal performance competitive to the state of the art, while providing a more compact representation.
Keywords : robust image description scale invariance local appearance description compact descriptor variable vector length
Document type :
Conference papers
Complete list of metadatas

Cited literature [17 references]  Display  Hide  Download
https://hal.inria.fr/hal-01079658
Contributor : Evanthia Mavridou <>
Submitted on : Monday, November 3, 2014 - 1:59:48 PM
Last modification on : Friday, October 25, 2019 - 2:01:05 AM
Long-term archiving on: Wednesday, February 4, 2015 - 10:35:44 AM

File

iciar2014 (1).pdf
Files produced by the author(s)

Identifiers

Collections

CNRS | INRIA | LIG | UGA | LIG_SIC_PRIMA

Citation

Evanthia Mavridou, James L. Crowley, Augustin Lux. Multiscale Shape Description with Laplacian Profile and Fourier Transform. International Conference on Image Analysis and Recognition, ICIAR 2014, Oct 2014, Vilamoura, Algarve, Portugal. pp.46 - 54, ⟨10.1007/978-3-319-11758-4_6⟩. ⟨hal-01079658⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

343

Files downloads

319

 

 

Contact

archive-ouverte@inria.fr