【摘要】 概述了图像的可视化特征如颜色、纹理、形状和运动信息,时空关系分析,以及多特征目标提取和相似度量度;分析了视频语义的提取,语义查询、检索;探讨了视频语义检索的性能评估,存在的问题和发展方向。更多还原

【Abstract】 Many information systems require a large multimedia storage and retrieval, including browsing, querying, etc. Image retrieval technique has gradually matured. Video semantic retrieval is yet a research hotspot. In this paper, we first discuss some visual features such as color, shape, texture and motion information, spatial temporal analysis, multi feature object extraction and similarity measure. Second, we analysis video semantic object extraction,retrieval via semantic query. At last, we discuss the performance evaluation of video semantic retrieval, open issues and promising directions.更多还原