A Comparison of Texture Features Based on SVM and SOM

Experimental results of texture features derived from Gabor and other four wavelet transforms classified and clustered based on Support Vector Machine (SVMs) and Self-Organizing Maps (SOMs) are reported in this paper. A comparison of SVM and SOM in texture classification is illustrated. The results show that these texture sets with appropriate classifiers perform reasonably well.

• Type: Conference
• Year: 2006
• Where: ICPR
• Authors: Chaur-Chin Chen, Chien-Chang Chen, Chih-Ming Chen
• Keywords: Computer Vision, ICPR 2006, Texture Classification, Texture Features, Texture Sets