摘要： 提出了一种基于小波变换和模糊C均值(FCM)算法的两阶段纹理分割方法,各阶段采用不同的纹理特征。而且该特征结合了像素的空域和频域信息。实验表明该方法在分割错误率、边缘准确性及区域一致性方面均有较明显的改善,同时也有较高的分割效率。

关键词: 纹理分割, 小波变换, 模糊C均值聚类算法, 多特征

Abstract: A novel approach based on wavelet-transformation and FCM algorithm is proposed in this paper and it has two stages in which different texture features have been used.The features extracted in the algorithm have included not only spatial but also spatial frequency information.The present approach shows visible improvements both in diminishing segmentation error,and in increasing boundary precision and region harmony,and also it has a higher efficiency.

中图分类号: 

• TP391.41