利用多进制小波变换和模糊c-均值聚类进行纹理分割

Texture Segmentation with M-band Wavelet Transform and Fuzzy C-Means Clustering

提出了一种基于多进制小波变换的多纹理特征提取方法,通过对小波系数的标准差作为纹理测度以生成特征向量,利用模糊c-均值聚类算法进行纹理分割,获得了较好的实验结果。

A new method applying M-band wavelet transform to extract the muti-texture feature is presented and the feature vectors are calculated through standard difference with wavelet coefficient.Finally,textured image is segmented through fuzzy C-means clustering,and the experiments provide encouraging results.