Directional features are extremely important in the framework of computer vision. Recent neurophysiological evidence from the visual cortex of mammalian brains suggests that the lter response proles of the main class of linearly responding cortical neurons are best modeled as a family of affine coherent states generated by the affine group. The decomposition of an image into these states is the wavelet transform of the image. Furthermore, translation, rotation and scale invariance are highly desiderable features of any pattern recognition system. In this contribution, we propose a texture classification system using bi-dimensional Dyadic Frames of Directional Wavelet (DDWF) as texture descriptors. The local energy measures on the transformed coeÆcients are used as features. Due to the inherent multiresolution structure and the possibility to nely tune the angular selectivity, we expect DDWF to be an eective tool for texture classication. Furthermore, it is suited to make the classication system invariant with respect to translations and rotations. The feature vectors are extracted from the d subbands of the directional multiresolution decomposition introduced in the previous section. That is, the statistical properties of the texture are charecterized by the set of local energy measures computed at the output of the lter bank. Different distance measures and classication algorithms have been considered.

Dyadic frames of directional wavelets as texture descriptors

MENEGAZ, Gloria;
2000

Abstract

Directional features are extremely important in the framework of computer vision. Recent neurophysiological evidence from the visual cortex of mammalian brains suggests that the lter response proles of the main class of linearly responding cortical neurons are best modeled as a family of affine coherent states generated by the affine group. The decomposition of an image into these states is the wavelet transform of the image. Furthermore, translation, rotation and scale invariance are highly desiderable features of any pattern recognition system. In this contribution, we propose a texture classification system using bi-dimensional Dyadic Frames of Directional Wavelet (DDWF) as texture descriptors. The local energy measures on the transformed coeÆcients are used as features. Due to the inherent multiresolution structure and the possibility to nely tune the angular selectivity, we expect DDWF to be an eective tool for texture classication. Furthermore, it is suited to make the classication system invariant with respect to translations and rotations. The feature vectors are extracted from the d subbands of the directional multiresolution decomposition introduced in the previous section. That is, the statistical properties of the texture are charecterized by the set of local energy measures computed at the output of the lter bank. Different distance measures and classication algorithms have been considered.
9780819437648
Dyadic wavelets; textures
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11562/429569
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? 2
social impact