| The sports activity has attracted people all over the world with its own charm.The development of TV and sattlite technologies have contributed to the prosperity of sports activities,there are live game events every week or even every day .Besides, the sports events videos are being widely spread on the internet with the developement of internet , high-capacity storage, and the compression standards such as MPEG-2,MPEG-4,MPEG-7.Most people tend to only enjoy their favorite games and fragments and they wish to find out the game fragments which attract them most from the mountains of game videos. The effective overview and retrival give an urgent call for the establishment of catalogue and indexing information for sports videos. The slow-motion replay, namely the relpaying of the fantastic parts of the game, can be used to form the summaries of the games and help to build the catalogue and the indexing information.With the aim to provide a better analysis of the content of sports videos and better retrieving and indexing services, this article, pre-equipped with the knowlege of sports-video producing and the rules of preception, based on the research for sports video content analysis framework, has researched and implemented the technology for detection of slow-motion replay in sports video. The main jobs in this paper lie in:1. It has proposed a framework based on Basic Semantic Unit (BSU) for sports video content analysis and has studied slow-motion replay analysis in the framework.2. It has studied how to use HMM to detect slow motion replay, including how to model slow-motion replay, how to extract features, how to train HMM and how to get the optimal state sequence.3. It has proposed an algorithm for detection of slow motion replay by detecting frames containing logos in the special scene transitions that sandwich replays.4. It has designed and implemented a sub-system for detection of slow-motion replay. The experiment shows that the sub-system can help sports video content analysis.
|
PDF Full Text Request