Description / 描述

The Multichannel Gray Level Co-Occurrence Matrix (MGLCM) is proposed as an upgrade to the standard Gray Level Co-Occurrence Matrix (GLCM) for extracting textural information from multispectral/hyperspectral images. The remote sensing image is first processed with the proposed multichannel gray level quantization algorithm, after which the various GLCM textural measures are computed using tunable window sizes and displacement vectors. The clustering-based GLCM, namely the K-Means GLCM, is demonstrated here. The five images shown are the original remote sensing image, the MGLCMs computed using the 'contrast', 'dissimilarity', 'energy', and 'homogeneity' measures (from left to right respectively).

多通道灰度共生矩阵（MGLCM）是标准灰度共生矩阵（GLCM）的升级版，主要用于从多光谱/高光谱图像中提取纹理信息。 遥感图像首先使用提议的多通道灰度量化算法进行处理，之后使用可调窗口大小和位移向量计算各种 GLCM 纹理度量。 这里实现了基于聚类计算出的GLCM，即 K-Means GLCM。 左边展示的五张图像分别是原始遥感图像、使用“对比度”、“不相似度”、“能量”和“均匀性”度量计算的 MGLCM（从左到右）。