Semantic Image Retrieval Based on Ontology and SPARQL Query

The main objective of this paper is how to use the ontology for semantic image annotation and search in huge collection of images. We have presentenced a framework for applying the semantics to enhance image retrieval. The entire problem is considered in two levels. First, An ontology is created to define the semantic space. Secondly, Natural language sentences are converted in to SPARQL statements and the relevant images are accessed using SPARQL query. The ontologies are represented in RDF form and these are based on existing data standard and knowledge corpura. Since the RDF structure provides the formal way of annotating the images, the image retrieval task is simplified as compared with earlier approaches. Retrieval is done by using the keyword (thesauri) description. We also show that we are able to retrieve desired images using the SPARQL query language (7).