In this paper, we compare the characters of pyramid-structured and tree-structured wavelet transform. According to the difference between these two, we propose a texture image segmentation approach which uses tree-structured wavelet transform to extract texture features and then be classified by fuzzy neural network. The experiment shows that tree-structured wavelet transform is effective in texture analysis%分析了小波变换的特点,将传统的塔型小波变换和树型小波变换进行了比较,给出了一种用树型小波来提取纹理特征进行纹理图象分割的方法。经过实验结果表明,用树型小波变换可以在纹理分析中获得较好的效果。
展开▼