结合小波框架和灰度共生矩阵的纹理分割算法

Texture Segmentation Algorithm Based on Wavelet Frame and Gray Level Co-occurrence Matrix

提出了一种结合小波框架和灰度共生矩阵的纹理分割算法,首先利用强壮的小波框架提取纹理的频域特征;然后利用灰度共生矩阵提取纹理的空间特征;最后将提取到的频域特征和空间特征结合起来,利用模糊C均值算法对结合后的纹理特征进行分类。实验结果表明与只利用纹理频域特征或空间特征进行分割相比,该方法的分割错误率显著降低。

A texture segmentation algorithm based on wavelet frame and gray level co-occurrence matrix is studied in this paper.Firstly frequency feature is extracted in different scale using modified discrete wavelet frame transform.Secondly,space feature ofa texture image is extracted based on gray level co-occurrence matrix.And then,the fuzzy c-means algorithm is applied to themerged features for texture segmentation.At last,simulations are performed on the presented algorithm,and the simulation resultshows that the presented algorithm has good performance in time.