New sensor technologies providing image information of increasing detail and complexity require adequate methods for processing and information extraction. The Z_GIS research field ?Integrated object-based GIS/image processing? investigates and applies methods of semi-automated image analysis utilizing GIS functionality.

Herein we explore how image data can be transformed into aggregated groups of pixels that represent geographical objects in a specific scale domain. These can be characterized by spectral behaviour, geometrical properties and spatial arrangement. Our research activities draw on an object-centred world view that builds upon the notion of a hierarchical organization of real world objects. Making expert knowledge explicit for image understanding means to construct cognition networks that utilize class schemes based on spectral, spatial and semantic descriptions. An integrated approach of object based GIS/image analysis communicates image content in object-related terms.