Combining wavelet and ridgelet transforms for texture classifications using support vector machines
Sluttao Li, Yi Li, Yaoman Wang · 2005
In this paper, we propose a method using combining features from discrete wavelet transforms and ridgelet transforms for texture classification. Typically, the 2D wavelet transform is good at capturing point singularities, while the ridgelet transform is good at capturing line singularities. Support vector machines (SVMs), which have demonstrated excellent performance in a variety of pattern recognition problems, were used as classifiers. The algorithm is tested on three different datasets, selected from Brodatz and VisTex databases. Experimental results demonstrated the combination of the two feature sets always outperformed each method individually. Compared to other methods, the proposed method produces more accurate classification results.