Video Content Analysis And Retrieval Technology And Its Educational Applications

With the fast development of Internet and multimedia technology, there are more and more multimedia resources. They are more and more widely used in education. Multimedia resources, especially the video information, including much lively and abundant information, was paid more and more attention by the scholar and users. In the field of multimedia research, content-based video analysis and retrieval technology is one of the hot topics, it has wide application prospect in the fields of electronic library, abundant data storage and retrieval and so on, and there's still much valuable technology content to be intensively researched and probed. This thesis researches content-based video analysis and retrieval technology to serve for education and teaching. We hope that the technology can help the education workers in video retrieval and video management.Following the thread of content-based retrieval, the thesis carries out a deeper research in content-based video analysis and retrieval technology, especially the video used in education. The main jobs and innovation are as follows:First, we researched and realized several shot segmentation method, and on the basis of the methods, we raised a self-adaptive method of shot detection based on neighbor frame difference and non-neighbor frame difference in allusion to the characteristics of abrupt shot and gradual shot. The method can detect almost all abrupt and gradual shot and achieve fine general detection effects.Second, we realized a simple and effective content-based shot extraction method, on the basis of summarizing existing key frame extraction methods, and indicated the color of shot by computing the average color of key frames of the shot.Third, in the aspect of video semantic information analysis, we gave shot classification schemes of general video and instruction video. General video shot can be classified as creative shot and photograph shot; Instruction video shot can be classified as platform shot and student shot.Fourth, we researched web semantic information extraction of video resources and got some achievements.Fifth, we designed and realized a content-based video retrieval system and carried out experiments on elementary education relevant web video resources and achieved good experiment effects.The general video shot segmentation system, video library analysis system and content-based video retrieval system are designed and developed based on Microsoft Direct Show technology, and then implemented using Visual C++ 6.0.In the course of our research, we try to track the academic advances of relevant field in technology research, and implement new algorithm and make some practical improvements, and discuss its application in management and search of educational resources from the visual angle of education and the goal of serving for education. Although our research is junior and simple, it is innovative to research content-based video retrieval technology in the field of educational technology. It not only expands the research angle of educational technology, but also enriches the application domain of content-based video analysis and retrieval technology. We research technology for serving education and hope more and more people that engage in technology pay attention to technology application in education.