Image analysis for semantic image databases

In this paper the problem of accessing a database of images, giving conditions on the semantic image content, is addressed. The approach proposed here, unlike most research in spatial data management, is deeply connected for the image analysis process. The focus is on the composition of objects, in terms of other objects, taking into account different interpretations and different levels of recognition. The image analysis process can be applied for images belonging fo specilic domains, described in advance for the system. Image analysis is performed in two steps: in the firsf step, simple objects are recognized starting from basic graphical pictorial elements; in the second step, complex objects are recognized as compositions of simpler objects and different interpretations are generated. Heuristics are adopted fo control the complexity of the algorithms. Finally, an image query language is proposed. It allows the user for express complex conditions on the image content (i.e. selections on objects, their composition and degree of recognition), resulting from the image analysis process.