Change detection of buildings in suburban areas from high resolution satellite data developed through object based image analysis

Abstract

The aim of this paper was to investigate the development of a fuzzy knowledge base within an object based image analysis (OBIA) system for automatic change detection of buildings. A multitemporal analysis of very high resolution satellite data (QuickBird and IKONOS) was performed. Two case studies for the Keratea suburb of Athens, Greece were selected. For each dataset, primitive image objects were created through multi-resolution segmentation, in five hierarchical levels, following a mixed top-down strategy established by a trial-and-error procedure. Subsequently each object was assigned by fuzzy classification to one of the classes representing the land cover/use categories of each level. The aim of the classification procedure was to separate the image objects into buildings and not buildings, extract the changes occurring between the two dates, and perform qualitative and quantitative evaluation.