传统上有些 研究工作利用场景变换侦测(shot change detection)來达到资料缩减的目的，但 成果并非特别好。鼓勵有志者一同來思考这个问题，使快速视讯撷取的美梦能 早日成真。
基于28个网页-相关网页
In this article an abrupt shot change detection method based on statistical model is proposed.
提出了一种基于统计模型的镜头突变检测方法。
The framework of video abstraction system in compressed domain is described, and main algorithms such as shot change detection, key-frame extraction and face detection are discussed.
给出了压缩域视频摘要系统的结构框架，并对其中的压缩域视频镜头变换检测、关键帧提取和压缩域人脸检测等主要算法进行了讨论。
Our experimental results demonstrate the performance of scene change detection based on shot is better than that of shot clustering based on multi-features.
实验表明，基于镜头的场景边界检测性能优于基于多特征的镜头聚类分析。
应用推荐