Video-based Ontology And Mpeg-7 Semantic Retrieval Technology

As the computer technology develops, network technology and semanticweb technology are developing very quickly and the multimedia informationis increasing rapidly. Now, how to denote and obtain the multimediainformation in the semantic web has been the research hot point and thenew research topic in computer area and other study areas.In this paper, we base on the description schema of the MPEG-7 andthe theory of the ontology, using the Resource Description Framework (RDF)and the Web Ontology language (OWL) to organize the semantic informationof the video. In this way, we construct a model to describe the semanticinformation of video. At the same time, we describe the semanticinformation of the video shot from the video object, video screen and videoevent. Combining with the hiberarchy of the semantic information of thevideo, we describe the high-level semantic information of the video andhave the connection with the low semantic feature of the video describedin the MPEG-7. In order to make the semantic searching in this descriptionschema more convenience, we transform the video description data in XMLformat to the ontology description data in OWL format.To realize the semantic searching, we use the tools Jena to parse theOWL data, and set up a set of semantic rules. We realize some descriptionlogic reseaning algorithm and realize the reasoner. In this way, we realizethe semantic searching in the semantic web based on the ontology. What ismore, we introduce a relating-feed back technology based on the networkkeyword. Through studying the feed back information constantly, thesystem adjusts the weight between the keyword and the keyframe to improvethe semantic seaching result.