摘要： 提出一种基于视频中运动目标的自嵌入水印方法，可以实现在视频信息接收端对运动目标的无损恢复。为保证在水印嵌入前后获取的目标完全一致，给出一种新的运动目标提取方式，即使用近似于原始视频帧的合成帧来提取视频中的运动目标。实验结果表明，在视频未被篡改或在一定程度上被篡改的情况下，该方法可以实现运动目标的完全恢复，具有一定的可行性。

关键词: 视频安全, 数字水印, 视频水印, 运动目标提取, 视频篡改

Abstract: This paper presents an object based self-embedding watermarking that can restore the moving objects completely at the receiver end. In order to guarantee consistent received objects, a new method of extracting the moving objects is adopted that using an artificial frame but not the original one. Experimental results show that the moving objects can be restored completely when the stego-video is un-tampered or the tampered area is not too extensive.

Key words: video security, digital watermarking, video watermarking, moving objects extraction, video tampering

中图分类号: 

• TP309