图像纹理的灰度共生矩阵计算问题的分析

图像的灰度共生矩阵(GLCM)已知被理论证明并且实验显示它在纹理分析中是一个很好的方法,广泛用于将灰度值转化为纹理信息.然而,由于GLCM是像素距离和角度的矩阵函数,因此完整的GLCM的计算,其参数的选取范围很广,这样GLCM的计算量很大,通常是不能这样用的.为了解决这个问题,本文应用马尔可夫链的性质,从理论上证明了GLCM的计算结果,当像素距离足够大的时候趋于一致性.这样只需较少的参数值就可以完整的描述图像的纹理特征.最后,通过对Brodatz纹理库中自然纹理图像和几幅SAR图像进行仿真,仿真结果验证了上述结论.

Research on Computation of GLCM of Image Texture

The Gray Level Co-occurrence Matrix (GLCM) has been proved to be a promising method for image texture analysis.However,the parameters in computing the GLCM of an image can be selected from a wide range,which results in a large amount of computation to be needed and makes it difficult to analyze the image textures.To simplify the computation of GLCM,we analyze the computation of GLCM by using Markov Chain property.We prove that GLCM is independent of distance and angle when distance is large enough.According to our analysis,the computation of GLCM is simplified by reducing the number of selected values of distance and angle.Finally,we give simulation results on natural texture images and SAR images to validate the theoretical analysis.