Title :
A solution for eigen fuzzy sets of adjoint max-min composition and its application to image analysis
Author :
Nobuhara, Hajime ; Hirota, Kaoru
Author_Institution :
Dept. of Comput. Intelligence & Syst. Sci., Tokyo Inst. of Technol., Yokohama, Japan
Abstract :
As a principal component analysis of the images based on an ordered structure, the greatest eigen fuzzy set (GEFS) and the smallest eigen fuzzy set (SEFS) of max-min composition and adjoint one, are proposed. In the case of the proposed method, an original image is regarded as a fuzzy relation by intensity normalization, and the GEFS and the SEFS are obtained as the transitive closure of the fuzzy relation. Through experiments using 91 test images extracted from ´view sphere database´, it is confirmed that the GEFS and SEFS are obtained during 10 iterations, under the condition that the size of the original image is 256 x 256 pixels.
Keywords :
eigenvalues and eigenfunctions; fuzzy set theory; image processing; minimax techniques; principal component analysis; adjoint max-min composition; eigen fuzzy sets; fuzzy relation; image analysis; principal component analysis; view sphere database; Computational intelligence; Fuzzy sets; Fuzzy systems; Image analysis; Image coding; Image databases; Image processing; Medical diagnosis; Principal component analysis; Testing;
Conference_Titel :
Intelligent Signal Processing, 2003 IEEE International Symposium on
Print_ISBN :
0-7803-7864-4
DOI :
10.1109/ISP.2003.1275808
