摘要
现针对目前的视频水印算法在篡改检测方面存在不足,提出一种基于压缩传感的半脆弱水印的视频篡改检测算法。通过压缩传感算法提取I帧的图像特征,生成半脆弱的内容认证水印,嵌入到I帧图像的中高频系数中,实现帧内篡改检测。通过对P帧的序号进行哈希运算,生成完整性水印,嵌入到P帧的运动矢量中,实现帧间篡改检测。仿真实验表明,算法对视频帧间篡改和帧内篡改具有较好的检测能力,能够精确到图像帧的子块。
For the shortcomings of current video watermarking algorithms on tamper detection,this paper proposes a video tamper detection algorithm based on semi-fragile watermark with Compressive Sensing.The image features of I-frame are extracted by Compressed Sensing algorithm,and semi-fragile content authentication watermarks are generated.The watermarks are embedded into the middle-high-frequency coefficients of I-frame,so to implement tamper detection inner frame.Through hash operation on the index of P-frames,integrity watermarks are generated.The watermarks are embedded into the motion vectors of P-frame,so to implement tamper detection among frames.Simulation results show that the algorithm has good detection ability for video tamper among frames and inner frame,whose accuracy can reach to the sub-block of image.
出处
《电路与系统学报》
CSCD
北大核心
2011年第4期86-93,共8页
Journal of Circuits and Systems
基金
广东省自然科学基金资助项目(10452800001004185)
佛山科技发展专项基金(2005010031)
关键词
压缩传感
半脆弱水印
帧间篡改检测
帧内篡改检测
compressive sensing
semi-fragile watermark
tamper detection among frames
tamper detection inner frame