Analysis of Texture Classification By Wavelet Transform And Curvelet Transform

Main Article Content

M. Santhanalakshmi
Dr.K. Nirmala


In this paper, the task of texture image classification is analyzed by using Discrete Wavelet Transform (DWT) and Discrete Curvelet Transform (DCT). The wavelet and curvelet coefficients are used to describe the textures in the given image. These coefficients are obtained by the decomposition process. First, the texture image is decomposed by using DWT and DCT at multiscale. As the sub-bands in the decomposed image carries diverse information about the texture, predefined number of coefficients is selected in each sub-band image. Before selecting the coefficients, sub-band coefficients are sorted in order to account high energy coefficients. The results show that the classification accuracy of DWT based features outperforms the DCT energies. The classification accuracy of DWT is 5% higher than DCT features at 2-level decomposition with 50% of coefficients used.


Keywords: Texture, texture classification, discrete wavelet transform, discrete curvelet transform, nearest neighbor classifier


Download data is not yet available.

Article Details