Change Over Time in Grey Levels of Multispectral Landsat 5TM/8OLI Satellite Images

Change detection (CD) from the Earth’s surface has developed into an important issue throughout the world. CD methods have found applicability in socio-economic assessments, environmental monitoring, exploring resources, etc., and till today, various numbers of CD methodologies utilising remotely sensed data have been developed. Researchers have categorised different CD techniques on the basis of distinct viewpoints. A common one is to associate different techniques in two categories pre-classification CD and post-classification CD. This paper examines the efficiency of post-classification, texture-based CD method. Changing conditions are investigated for the three cases of civic urbanisation, depleted lake’s water and surging rivers and lake. Changes in the image parameters are first analysed with first-order statistics, i.e. (non-zero ‘mean’ and ‘standard deviation’), and then changes in the visual characteristics are quantitatively measured through the grey level by second-order statistical parameters. Contrast, correlation, energy and homogeneity are the grey level co-occurrence matrix (GLCM) features which predict change in the ‘visual’ appearance of the texture. Texture gets converted from one form to another form represented by pre- and post-grey level image, respectively. Finally, we have established a novel changing ‘pattern’ of the texture from the aforesaid three cases.