시공간 관계 온톨로지 구축을 통한 이동 객체의 움직임 이해에 관한 연구

Abstract

These days, data size has been increasing and has grown exponentially not only in text data but also multimedia data of visual and auditory information due to the development of the IT environment. Image and video in multimedia data are typical representation methods which include various information such as color, shape, texture, pattern, and other characteristics. Especially in video data, information such as object movement is included. Objects have the passage of time and spatial features in spatio-temporal relations. In this connection, a lot of research studies which have been carried out over the time concentrate on information recognition by computer using low level data. Vocabulary for representing human thinking has a semantic gap between low level and high level information. A lot of research for reducing semantic gap are focus on the representation methods of logic such as, Propositional Logic, Predicate Logic, First-order Logic, Description Logic, Horn Logic and others. Other approach methods used also include Neural Network, Semantic Web, Knowledge base, and Heuristic approach among others. Ontology is the most important concept of semantic web among the mentioned methods and is a representation method for understanding the meaning of information using concept and relation of data, information, knowledge and others. Ontology is a representation of relation between concepts and it can be applied to find new concept relations using inference. Rules of Inference are to have sound, complete and tractable because it use premise relations. The goal in this paper is understanding of object movement and definition of spatio-temporal relation through mapping between vocabulary and object movement. Ontology mapping methods used between low level and high level information. In this case, spatio-temporal relation consist of temporal relation obedient to the passage of time, directional relation obedient to changes of object movement direction, changes of object size relation, topological relation obedient to changes of object movement position, and velocity relation using concept relations between topology models. This paper in the ontology building part defines the inference rules using proposed spatio-temporal relation and the use of Markov Logic Networks (MLNs) for probabilistic reasoning. Finally, the experiment and evaluation performs the verification recognition and understanding of object movement based on video data. This paper can be extended to retrieval and comparison between object movement, automatical annotation, and summarization in video.