Multiple Sub Ontology Frame Work for Comprehensive Video Database System

Video is most powerful, complex, voluminous, unformatted, and unstructured of all media used for conveying information. Hence, representing video information to enable effective and efficient video storage, semantic annotation, indexing and retrieval systems is an interesting problem. The current research is focusing on ontology based video annotation and retrieval, but most of the video systems focusing on one aspect of a particular domain. Not much research has been done on a comprehensive frame work for the identification of video semantic concept classes, classification of the video semantic concept classes and organizing concept hierarchies for generic video application. This paper proposes multiple sub ontologies for comprehensive video data base for any video applications. The proposed approach gives a comprehensive high level frame work for the identification of video concept classes, classification of the video concept classes and organizing concept hierarchies and classification of the concept instances. The beauty of the proposed approach is flexibility and generic which is applicable to any video application domain like sports, news, cinema, television programs, personalized video collections etc. multiple views of the video semantics are considered to build a comprehensive video system. Since a single ontology will not support the multiple views of the semantics of the video system. The multiple video concepts and the knowledge of the concepts and the concept hierarchies and the relationship between concepts are been addressed in multiple sub ontologies. The proposed model is been implemented using Protege tool. Finally conclusion and future work has been discussed.