Texture Classification Using Tree-structured Wavelet Transform and Neural Networks

In this paper a new approach for texture classification problem is presented. Tree-structured wavelet transform for the pre-processing task while the neural network is used for the classification task. The energy values obtained at the subbands are used for the input vector of the neural network system. The subbands where the energy values are greater than a threshold are determined and it is tried to establish tthe texture classification problem. The back-propagation algorithm is used to test a performance of the system.