mk:scia01

Summary

Binary pattern processing framework for perceptual fault detection. Mario Köppen, Aureli Soria Frisch and Thorsten Sy. In Proceedings of the 12th Scandinavian Conference on Image Analysis (SCIA2001?), pages 363-370, Bergen, Norway, 2001.

Abstract

This paper presents a framework for the binary processing of perceptual faults in texture analysis. The framework is composed of a chain of Binary Pattern Processing Modules (BPPM). Each BPPM is responsible for a class of perceptual faults. In a BPPM, several image operators are applied onto the preprocessed image. The result images are fused into a binary image, and a connected component analysis is performed for extracting fault indicating binary components. The whole image is described by a set of fuzzy features, and class memberships are derived by using the Choquet integral. Configurations of the framework for the perceptual fault classes strong-contrast faults (holes, cracks), broad-range faults and low-contrast faults are given, which are for all non-trivial cases based on the 2D-Lookup algorithm. The presented framework is able to work in the context of objective reproduction of subjective decisions about texture surface quality, and may be easily adapted to new technological needs by designing further components or by modifying existing ones.

Bibtex entry

@INPROCEEDINGS { mk:scia01,
    ABSTRACT = { This paper presents a framework for the binary processing of perceptual faults in texture analysis. The framework is composed of a chain of Binary Pattern Processing Modules (BPPM). Each BPPM is responsible for a class of perceptual faults. In a BPPM, several image operators are applied onto the preprocessed image. The result images are fused into a binary image, and a connected component analysis is performed for extracting fault indicating binary components. The whole image is described by a set of fuzzy features, and class memberships are derived by using the Choquet integral. Configurations of the framework for the perceptual fault classes strong-contrast faults (holes, cracks), broad-range faults and low-contrast faults are given, which are for all non-trivial cases based on the 2D-Lookup algorithm. The presented framework is able to work in the context of objective reproduction of subjective decisions about texture surface quality, and may be easily adapted to new technological needs by designing further components or by modifying existing ones. },
    ADDRESS = { Bergen, Norway },
    AUTHOR = { Mario Köppen and Aureli Soria Frisch and Thorsten Sy },
    BOOKTITLE = { Proceedings of the 12th Scandinavian Conference on Image Analysis (SCIA2001?) },
    MODIFIED = { 2008-02-28 15:47:02 +0900 },
    EDITOR = { Ivar Austvoll },
    HASABSTRACT = { Yes },
    PAGES = { 363--370 },
    PDF = { scia01.pdf },
    TITLE = { Binary pattern processing framework for perceptual fault detection },
    YEAR = { 2001 },
}