基于运动目标检测的视频水印算法研究  

ON VIDEO WATERMARKING SCHEME BASED ON MOVING OBJECT DETECTION

作　　者：周萍[1] 刘岩[2] 陈希[2] 景新幸[3] 

机构地区：[1]桂林电子科技大学电子工程与自动化学院,广西桂林541004 [2]桂林电子科技大学计算机科学与工程学院,广西桂林541004 [3]桂林电子科技大学信息与通信学院,广西桂林541004

出　　处：《计算机应用与软件》2014年第2期157-159,187,共4页Computer Applications and Software

基　　金：国家自然科学基金项目(60961002);广西自然科学基金项目(2012GXNSFAA053221)

摘　　要：深入研究视频序列的基本特性,并在此基础上提出一种改进的DCT(离散余弦变换)域的分块视频水印方案。提出的方案中,通过结合人类视觉感知系统和运动目标检测技术,从视频图像序列中实时地发现并提取运动目标,确定运动目标所在区域,选取视频图像中的运动区域和高细节纹理复杂度区域作为视频水印的嵌入区域,并在水印的嵌入的过程中,通过细分像素块来减少像素分类带来的分块效应。通过仿真实验结果显示,该算法具有较好的不可感知性和鲁棒性。We make deep research on basic characteristics of the video sequence,and propose an improved block video watermarking scheme of discrete cosine transform （DCT）domain on the basis of the research.In the scheme,by combining the human visual system with the moving object detection technology,the moving object is found timely from a sequence of video images and is extracted,the area where the moving object located is determined,the moving areas and the areas with highly detailed texture complexity in video image set are chosen as the embedding area of video watermarks.Moreover,in the process of watermarks embedment,the scheme uses the sub-pixel blocks to reduce the block effects of pixel classification.Simulation experimental results show that the algorithm proposed in the paper has better imperceptibility and robustness.

关 键 词：离散余弦变换 人类视觉系统 运动目标检测 帧间差分法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]