Published Paper Details:

Remote sensing satellite images are developed rapidly regarding quantity, quality, and applications to detect and extract various natural and artificial features on earth's surface like(vehicles, buildings, trees, roads, water, aircraft, ships). These satellite images provide important information in various applications like urban planning, disaster management, and environmental management. Different algorithms and methods are introduced by researchers to extract the specified features from satellite images. In an urban scenario, buildings are one among the foremost elementary structures and play an important role in the field of urban development, urban planning, climate studies, disaster management, map making, land use analysis, and change detection. The study aims are to extract building footprints employing a machine learning algorithm. various methods for extraction of building footprints are discussed in various studies However, extraction of rooftop of buildings from an metropolitan place has been a difficult task, since building rooftops have different shape, size, and spectral properties. In addition to that, other urban features, such as, road, barren land, etc. exhibit spectral properties similar to that of building rooftops. so that building extraction technique has become a vital and tough research problem receiving better recognition. The proposed technique uses a Machine Learning algorithm for classifying the building and non-building pixels. To eliminate false-detected building pixel use median filtering, morphological operator, and connected component labelling are used .the technique has been evaluating by pixel and object-based criteria with the consideration of precision, recall, quality of building(urban object).