Knowledge-based classification of Landsat-TM satellite images for mapping land use in Bangkok within the context of urban planning

A large rapidly growing city like Bangkok requires effective means of extracting spatial information in order to develop a realistic understanding of the current state of the land and housing markets. This study evaluates the potentials of using high resolution satellite images as a source of information for land use mapping within the context of urban planning. Standard data acquisition methods, i.e. use of aerial photographs, are compared to satellite image analysis which is appropriate for mapping the overall spatial development of the urban areas. The digital representation of data permits fast computation of various statistics, including growth of different land use classes, degree of urbanization etc. A prototype for an automatic expert system-based classification method using texture and context information is described. The method is used to produce an urban land use map for a test area covering the central and western parts of Bangkok.