Content-based Video Retrieval Technology Research And Implementation

With the rapid development of modern technology, multimedia technology hasbeen rapid development. In recent years, television technology and the gradualintegration of IT technology, it also signals that the realization of digital video and audioinformation management known as media asset management (Media Asset Management)is possible. But now most of the media asset management systems are the traditionalway of management information systems organization and management of video andaudio data, only solves the audio data storage and transmission of video informationretrieval remains to describe the video’s text label for the main Chuli object, this methodis simple, but there are two fundamental problems affecting the effective application ofthe system: First, the rich content of video information is difficult to fully express with atext label, and second, the information content of artificial mark on the video, very timeconsuming. Video information retrieval efficiency of the media asset managementsystem application and promotion of the core issues.According to the current video retrieval technology of the global newestdevelopment is dynamic, and the use of media asset management system to do somepractice content, focusing on the video information retrieval of conceptual and technicalstandard, based on text and content based on two major retrieval method has carried onthe analysis, thus forming the effective media asset management system in theapplication of video retrieval methods. As the video file is huge and complexinformation motion characteristics, so the retrieval target localization in media assetmanagement system in automatically extracting video key frame, this advantage is ableto dynamically image video retrieval, simplified to automatic extraction of video keyframe static image retrieval. In put to the realization of the process, we will be in mediaasset management system catalogue link, using MPEG-7standard multimediainformation describing thoughts, to increase the visual features of video key frameannotation, thus further expanding the search space.Finally, this paper is to studying the method of video retrieval applications work ofthe summary, and the next step research directions are discussed.