Nowadays, video media data is already facilitates generation, transmission, storage and circulation on the global scale. Video data is geometrically fast as the rate of growth, the video data processing and analysis have lagged behind the pace of development in the growth of data, resulting in large amounts of data is wasted. Therefore, it becomes an urgent need for efficient retrieval of video data content. In this paper, firstly, starting from the color feature, the color space of digital mapping and semantic color space conversion technology is proposed according to the problem of Semantic concepts for video does not match with the perceived characteristics. And then we realize the mapping from the low-level features to high-level semantic. Finally, semantic rules of uncertainty reasoning based on cloud model established to complete video content retrieval.