ABSTRACT
With the continuous development of Internet technology, Digital video has become an important way of network information recording and transmission. The video data is increasing at a tremendous speed in every day and the huge amounts of data bring great challenge for video classification and video retrieval. However, the traditional artificial way to analyze, manage and retrieve the video data, which not only consume large amounts of manpower, but also cannot adapt to the rapid growth of video data for classification and retrieval. In this case, there is an urgent need for a kind of content-base video retrieval system (CVBR) automatically. Because of the semantic gap between low-level features and high-level semantic concepts in video[5],how we bridge the semantic gap and extract semantic concept in video have attract more and more attention.
