Crucial video content extraction using ontology rule-based technology and decision making algorithm

Interpreting and extracting the content from video is fascinating issue to focus in many video based applications especially in mission critical situations. Many facilities exist in video processing that generates raw data and low level features namely color, texture and format which are not enough in complex data analysis of video application. Semantic level features from the video need to be extracted to meet the requirement of high level video processing. This paper explores the extraction technique of semantic level features such as object, event and concepts automatically. Crucial video content extraction model exploits Spatio-temporal relations in addition to concept and event definitions. Meta-Ontology presents a domain-independent rule generation standard that enables the user to generate ontological model for a specific domain. Using decision making algorithm, one can derive a decision of the occurrence of particular event. Rule definitions assist in lowering the computation cost of spatial relation and define the complex situation effectively. Efficacy of the proposed technique is evaluated in terms of precision and recall rates of the extraction of semantic level features.