Title :
Texture segmentation method based on incomplete tree structured wavelet transform and fuzzy Kohonen clustering network
Author :
Xiaodan, Wang ; Hua, Jin ; Rongchun, Zhao
Author_Institution :
Dept. of Comput. Eng., Northwestern Polytech. Univ., Xian, China
Abstract :
A new texture segmentation method based on incomplete tree structured wavelet transform and fuzzy Kohonen clustering network (FKCN) is proposed in this paper. It consists of three main steps: 1) feature extraction using incomplete tree structured wavelet transform; 2) feature coarse classification using fuzzy Kohonen clustering network which involves initializing the weight vector of the network according to the density function of input patterns and training the network using the reduced set of feature vectors to get the coarse segmentation result; 3) the refinement of the coarse segmentation result. Texture segmentation experiments give the effectiveness of this method
Keywords :
fuzzy neural nets; image segmentation; image texture; self-organising feature maps; trees (mathematics); wavelet transforms; FKCN; feature coarse classification; feature extraction; fuzzy Kohonen clustering network training; incomplete tree structured wavelet transform; texture segmentation method; weight vector; Classification tree analysis; Computer networks; Data mining; Energy measurement; Feature extraction; Filters; Fourier transforms; Humans; Wavelet packets; Wavelet transforms;
Conference_Titel :
Intelligent Control and Automation, 2000. Proceedings of the 3rd World Congress on
Conference_Location :
Hefei
Print_ISBN :
0-7803-5995-X
DOI :
10.1109/WCICA.2000.862541