Review on Semantic Content Extraction in Videos

In recent years the need for extracting contents from video in applications such as medical, forensics has increased. Raw data representing the elementary physical video attributes and low-level features representing audio, video, text alone are not sufficient to fulfill the user’s needs resulting in a deeper understanding of the content at the semantic level. Hence digital video databases have come to be more pervasive and finding video clips quickly in video databases have become a major challenge. Manual techniques were used which were inefficient, subjective and costly in time and limit the querying capabilities. Here Video Semantic Content Model is presented which helps in the automatic extraction of objects, events and concepts which is based on domain ontology. In proposed framework, viewing angle of camera to improve extraction capabilities is been considered.