Fabric Texture Classification Based on Wavelet Packet

The main energy of texture image is concentrated on middle frequency regain, but ordinary image on low frequency regain. With increasing the resolution level, wavelet transform focuses on low frequency domain, but wavelet packet on any frequency domain. Therefore it is very effective to use wavelet packet algorithm to classify the fabric texture. In this study, we use wavelet packet and BP neural network together to classify the fabric texture. Experimental result shows that the classification rate can attain 98%.