一种基于事件检测的视频取证方法

Digital video forensics method based on event detection

目前计算机视频取证的一个重要目标是如何快速准确地在海量视频中定位犯罪事件发生的时刻和地点,最终形成视频证据。针对复杂背景条件下丢弃或拾起等事件的监控视频,提出一种基于光流特征和形状特征结合的事件检测方法。通过实验证明了该方法在视频事件分析取证中的有效性。

A most important goal of computer video forensics is how to find out the crimes events from the huge numbers of video data rapidly and accurately, and then form the video evidence. This paper focused on the discard and pickup event surveillance video of complexity background, proposed a video event detection approach based on flow feature and shape feature match. The experiments on human actions in a real entertainment of video event analysis forensics is effective.