基于灰度共生矩阵的纹理分析的统计方法

Statistical Method of Texture Analysis Based on Gray Co-occurrence Matrix

由于影像图像纹理性强,这使得对于压缩后图像纹理质量的要求显得更加严格。然而,目前的影像图像压缩技术对压缩后图像的纹理效果处理不佳。本文将图像的灰度信息加进灰度共生矩阵,使得灰度矩阵能包含图像的纹理基元及其排列信息,并给出了统计分析方法。结果表明,与传统的压缩方法相比,文中提出的新方法更好地保留了原始图像的纹理效果及其相关信息。

The strong texture characteristics of image produce more stringent requirements for the texture qualities of the compressed image. However, the treatment effect of the compressed image texture is poor based on the current video compression technology. The experiment puts the gray information of images into gray co-occurrence matrix and the gray matrix can contain the image texton and arrangement information, and the method of statistical analysis is given. The results show that compared with the traditional compression methods, the new method proposed in this paper is better in retaining the texture effect and relevant information of original images.