Reducing Semantic Gap in Video Retrieval with Fusion: A Survey

Abstract: Multimedia provides a rich content of information and huge amount of data's are available in the field of video retrieval. Now day's enormous videos are available on web and online to accessible from internet or retrieve videos from smart phones, digital cellular assistants. There is the drastic growth in the amount of multimedia field of improving data storage, acquisition and communication technologies, which are all supported by major improvements in processing of video and audio. Researches focused on more efforts in video retrieval that contain certain visual information rather than image of their interest. Such a search is facilitated by Content Based Video Retrieval (CBVR) methods. Specifically segmentation of video is the most prominent step as the retrieved results are based on the segmentation boundaries. The shot boundary detection can be performed using various different techniques like Motion/hybrid DCT, edge tracking, histogram, HSV Model, Motion vector and Block matching methods. This paper mainly presents a study of different methods/algorithm that has been proposed in literature for video retrieval to reduce the semantic gap between low and high level features. Semantic gap between these two feature level is improving by its efficiency with the help of advanced algorithms and techniques using machine learning with fusions.