A tabu search based fuzzy c-means algorithm for VQ codebook design

Author

Lu, Zhe-Ming ; Pan, Jeng-Shyang ; Sun, Sheng-he

Author_Institution

Dept. of Autom. Test & Control, Harbin Inst. of Technol., China

Volume

2

fYear

2000

fDate

2000

Firstpage

1049

Abstract

The fuzzy c-means algorithm (FCM) has been presented for vector quantization (VQ) codebook design. In this paper, we present a tabu search (TS) based fuzzy c-means algorithm for VQ codebook design. Two key problems of the tabu search algorithm are how to define a solution and how to generate neighboring solutions from the current solution. A codebook-based method is presented to describe a solution for tabu search approach, and the corresponding codebook design algorithm is called CB-TSFCM. Test results show that the CB-TSFCM generates better codebooks than the traditional FCM algorithm and the c-means algorithm

Keywords

fuzzy systems; image coding; search problems; vector quantisation; CB-TSFCM; VQ codebook design; codebook-based method; fuzzy c-means algorithm; tabu search algorithm; vector quantization; Algorithm design and analysis; Clustering algorithms; Distortion measurement; Fuzzy sets; Iterative algorithms; Partitioning algorithms; Simulated annealing; Speech coding; Testing; Vector quantization;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing Proceedings, 2000. WCCC-ICSP 2000. 5th International Conference on

Conference_Location

Beijing

Print_ISBN

0-7803-5747-7

Type

conf

DOI

10.1109/ICOSP.2000.891711

Filename

891711