Ontological description of image content using regions relationships
Zurina Muda · ePrints Soton (University of Southampton) · 2008
Much initial research on image annotation pays little attention to the spatial relationships between regions. Annotations are most frequently assigned globally and even when assigned locally the extraction of such relations is often neglected. For rich semantic description of visual information, it is essential to capture such relations. The aim of this research is to attempt and enhance annotation systems, either through automatic or semi-automatic means, by capturing the spatial relationships between regions or objects in images and incorporating such knowledge in a knowledge base such as ontology. As part of a preliminary experiment, a comparative analysis of three existing tools has been performed and the result obtained. This research will contribute a new approach to automate image linking to an ontology based on regions with automatic spatial relationships extraction between the composite regions or objects in images, which reliable with images in the real world.