Reduction scheme of spatial relationship for efficient image retrieval system

Content-based image retrieval systems often rely on global image characteristics such as color and texture histograms, but not the spatial relationships among objects of an image. Reasoning spatial relationships is very expensive since it requires massive geometric computation. Especially reading and comparing information to reason directional and topological relationships could be repeated as they are processed separately in previous works. In order to avoid such repetition and improve the performance of reasoning spatial relationships, we propose a set of rules that permits reducing relationships from a given set of relationships. A reduction mechanism using these ndes can be used in query processing systems that retrieve images by content. This could give better precision and flexibility in image retrieval.