Multiscale Texture Enhancement

  • The ideas of texture analysis by means of the structure tensor are combined with the scale-space concept of anisotropic diffusion filtering. In contrast to many other nonlinear diffusion techniques, the proposed one uses a diffusion tensor instead of a scalar diffusivity. This allows true anisotropic behaviour. The preferred diffusion direction is determined according to the phase angle of the structure tensor. The diffusivity in this direction is increasing with the local coherence of the signal. This filter is constructed in such a way that it gives a mathematically well-funded scale-space representation of the original image. Experiments demonstrate its usefulness for the processing of interrupted one-dimensional structures such as fingerprint and fabric images.

Download full text files

Export metadata

Additional Services

Search Google Scholar
Metadaten
Author:Joachim Weickert
URN:urn:nbn:de:hbz:386-kluedo-5398
Series (Serial Number):Berichte der Arbeitsgruppe Technomathematik (AGTM Report) (139)
Document Type:Article
Language of publication:English
Year of Completion:1995
Year of first Publication:1995
Publishing Institution:Technische Universität Kaiserslautern
Date of the Publication (Server):2000/04/03
Tag:anisotropy; image enhancement; nonlinear diffusion; structure tensor; texture
Source:V. Hlavac, R. Sara (Eds.), Computer analysis of images and patterns, Lecture Notes in Computer Science, Vol. 970, Springer, Berlin, 230-237, 1995
Faculties / Organisational entities:Kaiserslautern - Fachbereich Mathematik
DDC-Cassification:5 Naturwissenschaften und Mathematik / 510 Mathematik
MSC-Classification (mathematics):35-XX PARTIAL DIFFERENTIAL EQUATIONS / 35Qxx Equations of mathematical physics and other areas of application [See also 35J05, 35J10, 35K05, 35L05] / 35Q80 PDEs in connection with classical thermodynamics and heat transfer
Licence (German):Standard gemäß KLUEDO-Leitlinien vor dem 27.05.2011