Residential areas extraction based on the spectral characteristics and GIS assistant information

Residential areas are the gathering of human activities. It has great significance to master the spatial distribution information timely and accurately by RS in disaster assessment, urban expansion and environmental change research. In this paper, taking Longkou city as an example, we find corresponding models to quickly extract residential areas by studying the image features, spatial features and spatial relations, furthermore, we use the spectral characteristics and GIS information to build knowledge base for the extraction.