BDBComp
Parceria:
SBC
Texture Analysis with the Spot Noise Model

ANTONIO CESAR GERMANO MARTINSALEXANDRE DA SILVA SIMÕES

Identification, segmentation and classification of texture can be used for image analysis indifferent computer vision applications. Since texture can only be defined over an area, Fourier domain methods are well suited for texture analysis, once it takes advantage of the fact that each frequency component contains information from all the pixels in that area. Although much work has been done with this approach, the use of a texture generation model can lead to a better interpretation of the power spectrum and the definition of the necessary measurements to distinguish image characteristics. This paper discusses texture analysis in the Fourier domain with the Spot Noise model as a guideline to interpret the power spectrum and shows the obtained results by the application of measurements that can characterize the spot underling the generation of a given random texture. The presented application is based on automatic analysis of cereal grains in a loading belt.

http://www.lbd.dcc.ufmg.br/colecoes/wvc/2006/0030.pdf

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato: bdbcomp@lbd.dcc.ufmg.br
     Mantida por:
LBD