The Role Of Texture Analysis In Extracting Urban Blocks From Radar Images Sentinel 1 SAR_ SLC

(Cace Study: Selected Areas In The Idleb Region)

Abstract

The automatic extraction of urban blocks from satellite images has become necessary at the present time to obtain ready-made digital plans that are automatically digitization instead of manual digitization, which requires long hours of office and field work. However, multispectral images may not achieve this easily due to the similarity in the reflectance characteristics between urban areas and some other features (such as areas of un covered rocky). From this standpoint, the research is working on developing a methodology for detecting urban areas from Sentinel1 radar images, type SLC, due to the ability of these rays to show the textural characteristics of urban blocks, in the hope that they can separate urban features from other features. A radar image of the study area was downloaded Sentinel 1 SAR type: SLC from the open date copernicus website which belong to European Space Agency (ESA), and was processed so that the image was segmented to cut the survey lines covering the study area, then the orientation file was applied, and the survey lines were removed. To correct the distortion resulting from the tilt of the imaging axis, terrain correction and projection selection were applied. After that, the study began with the improvement and classification steps, starting with the special filtering, applying the texture analysis, and then the supervised classification, to obtain an image classified into urban and non-urban areas.