摘要： 关键帧提取是基于内容检索的视频数据库中十分重要的一步，但在整个视频数据库数组和研究中，由于以前受到计算机处理能力的限制，一直被忽略。文章针对关键帧提取的问题，提出了时间自适应算法, 并用一系列的视频检索实验证明了其有效性。还讨论了在VC++平台上“视频分析及检索系统”软件包的开发问题。此软件包具有MPEG视频播放、视频文件分帧运算、帧图像直方图计算、镜头分割、基于关键帧的视频检索等功能。

关键词: 视频检索, 关键帧, 时间自适应算

Abstract: Key-frames extraction is a very important step in the organization of content-based video retrieval and browsing. In the whole procedure of the organization and studying in the video database, key-frames extraction is always neglected because of the limitation of the computer's ability in processing. This paper discusses on some new ways in allusion to key-frames extraction of video's characters in video database indexing. Adaptive time algorithm are proposed. The video database experiments are used to prove the effectiveness of the proposed method. It also presents an integrated system solution for computer-assisted video parsing and content-based video retrieval.

中图分类号: 

• TP391.3