Texture Classification Using Discrete Shearlet Transform
M. Kanchana, P. Varalakshmi Dr. P. Varalakshmi · International Journal of Scientific Research · 2012
In this paper, an efficient approach for texture classification system based on Discrete Shearlet Transform (DST) is proposed. The classification of texture is achieved by extracting block based energy form the shearlet decomposed image. First, the texture image is decomposed by 2-level DST with 2-direction. Then the proposed block based energy is extracted from each sub-band of the decomposed image. The block size used in the proposed system varies from 4 to 64. These block based energies are used as features to classify the given texture image using robust nearest neighbor classifier. Brodatz database tex- ture images are used to evaluate the proposed system. Experimental results show that the proposed method produces more accurate classification rate for block size 64 is over 94%.