Integration of Domain-Specific and Domain-Independent Ontologies for Colonoscopy Video Database Annotation

Flexible and effective indexing of video data so as to render it useful in automated inference and datadriven knowledge acquisition is a key problem in knowledge engineering. In this paper, we present an approach to annotation of a video database using a domain specific ontology, a domain independent video ontology that encodes the structure and attributes of video data. The two ontologies are integrated using domain-specific semantic linkage. The result, an integrated ontology for video annotation (IOVA) is represented in OWL, a description logic based ontology language. We describe a user-friendly platform for video database annotation (VIDAI) using IOVA. We demonstrate the use of IOVA and VIDAI in annotating a database of colonoscopy videos. Video entities are annotated with a combination of domain-independent features and the domain-specific entities. Annotation of video data using IOVA constitutes an important first step towards flexible and fully automated indexing, retrieval, inference, and data-driven knowledge discovery using video data in a broad range of applications including colonoscopy video analysis.