Object-oriented Image Classification for Urban Building Boundary Extraction from Ikonos Imagery

Detailed mapping of rooftops in urban environments requires high spatial resolution remotely sensed data. However, traditional pixel-based classifiers based on spectral classes are ineffective in high-resolution multispectral images due to large within-class spectral variations and between-class spectral confusions that characterize manmade features. In this study, a rule-based object-oriented classification algorithm for rooftop extraction is developed from an Ikonos urban scene of variable building morphology and patterning and spectral roof response. The algorithm includes the following steps: (1) fusion of 1m panchromatic and 4m multispectral bands to produce a pan-sharpened 1m multispectral image; (2) segmentation of the 1m dataset; (3) supervised object-based classification into broad spectral classes; and (4) spectral, spatial, textural and contextual parameters developed from sample rooftop objects are implemented in a fuzzy logic rule base to separate rooftops from the impervious surface class. The rule base identifies rooftops with 93% accuracy, demonstrating the applicability of fuzzy logic rule-based analysis to the discrimination of spectrally similar but semantically distinct classes.