Awantika Singh, Akhilesh Sharma
Abstract: Building extraction from very high resolution (VHR) symbolism assumes a significant job in urban arranging, fiasco the executives, route, refreshing geographic databases, and a few other geospatial applications. Contrasted and the conventional structure extraction approaches, profound learning systems have as of late demonstrated remarkable execution in this undertaking by utilizing both abnormal state and low-level component maps. Extraction information from satellite pictures is considered as a fundamental field of research in remote recognizing and machine vision. Various estimations for extraction of structures from satellite pictures have been shown up until this point. These computations generally have considered radiometric, geometric, edge area and shadow criteria approaches to manage play out the structure extraction. In this paper basically we present a system which is able to get the information from the HD satellite images. As per our propose approach we are able to capture the building information in very less time & accuracy of our proposed approach is far better than previous existing approach.
Keywords: GIS, building Extraction, Satellite image processing, Corner, Line