Change Detection In Urban Land Cover Using Landsat Images Satellites, A Case Study In Algiers Town

In this paper three images of Landsat TM data of 1987, 2001 and 2010 being used to analyze the land cover and land use LC/LU changes in Algiers areas. The paper followed the expansion of urban tissue and its effect of decreasing agricultural and bare soil lands using difference soil adjusted vegetation index DSAVI, difference Greenness Tasseled Cap transformation DGTCT and post classification of multi-spectral and multi-temporal L5 and L7 Landsat satellite. The TM reflectance's images have been transformed to ETM+ reflectance's images using a regression method. The Maximum Likelihood algorithm is used to classify the reflectance images into the thematic urban, vegetation and bare soil map. For evaluating the classification and its assessment accuracy the neural net classification is carried out. The classified maps were then used as inputs to perform the post classification change detection. The Threshold changes are calculated for both DSAVI and DGTCT based on the means and standards deviation images. After that, the unchanged area, changes spaces are quantified and mapped. The proposed method is based on the study which mentioned that the urban tissue was faster growing with an annually growth of 0.5% and both DSAVI, DGTCT, post classification are useful for LC/LU change detection. Our proposed method is applied to Algiers town in North Africa.