运动矢量聚类的视频水印方法

Video watermarking technique based on motion vector clustering

基于运动矢量视频水印算法,结合人的视觉特性(HVS),根据运动矢量值及运动矢量块区域聚类的特征,利用FCM算法实现运动矢量的分类,以运动矢量块为单位,将水印信息嵌入到运动矢量上。运动矢量不需要完全解码和再编码过程,所提出的水印算法简单、有效,从而实现了水印盲检。与嵌入水印前的原视频图像相比,嵌入水印后的视频图像信噪比损失很小,对帧删除、帧插入和噪声等攻击方式具有极强的鲁棒性。

In the motion vector block-based algorithm,the characteristics of motion vector,the Human Visual System（HVS） and the clustering features of motion vector block are considered.The FCM algorithm is used to classify the motion vectors.Watermarking information is embedded into the motion vector block units.Therefore complete decoding and re-encoding are unnecessary. The proposed algorithm is simple and efficient.Moreover,watermark can be extracted blindly without original images.Compared with the original video sequences before watermarking,PSNR of watermarked frames is reduced slightly.At the same time,it has good robustness to withstand frame-deleting,frame-inserting and noise attack.