DCT域基于块分类的不可感知加性数字水印

Additive Insertion Imperceptual Watermarking Based on Block Classification for DCT-domain

该文根据Watson感知模型提出了DCT域基于块分类不可感知加性水印模型。算法计算亮度块方差,将图像块分为五类,利用亮度块特征来加性控制嵌入服从正态分布随机水印序列的区域和长度,提高了水印图像的鲁棒性。在检测端可根据水印图像的统计特性确定水印的检测阀值,不使用原始图像,就可进行水印的相关检测。实验结果表明,MPEG视频水印图像质量的PSNR平均下降4.1dB,但主观视觉质量没有明显下降。

This paper presents a new technique of how to insert additively imperceptual watermarking according to Watson's model based DCT-domain and the block classification.With the approach proposed in this paper,where image block is classed as five classification by computing the luminance block variance,region and length of the random watermarking sequence with normal distribution is embedded additively by exploiting the luminance characteristic,which greatly improves the robustness of watermarked image.Through analyzing the statistical property of the watermarked image,the detection threshold is determined according to the false alarm and missing rejection rate,where the original image is not required for watermark detection.Experiment results in video quanlity degrading averagely with PNSR 4.1dB but subjective quanlity remaining.