基于小波包变换和蚁群算法的纹理分割方法

为了解决纹理图像分割中所存在的区域一致性差、像素分割错误率较大的问题，提出基于小波包变换并且利用蚁群算法来对纹理图像进行分割的方法。考虑到传统聚类算法依赖于初始值的选取，选用了具有较强鲁棒性的蚁群算法来进行聚类分割。同时考虑到同种纹理的像素往往在空间上具备连续性，所以在进行再次分割时，结合了像素临域的其它像素的分割结果，来抑制纹理分割中的孤立点的出现。

Texture Segmentation Based on Wavelet Packet Transform and Ant Colony Optimization

In order to solve the problems of poor regional consistency and the large error rate in texture image segmentation,we proposed an approach based on wavelet packet transform and ant colony optimization in this paper.Considering the traditional clustering algorithm depends on the selection of the initial value,we use the ant colony algorithm instand,which has better robustness.The pixels which belong to the same texture are often continuity in space,therefore,we take the surrounding pixels’ last segment results into consideration,to reduce the presence of isolated points.