一种DCT域的灰度级数字水印算法研究

A gray level digital watermarking algorithm based on the DCT transform

针对灰度级数字水印图像中,常因信息量大而影响嵌入质量与稳定性的缺陷,提出了一种新的嵌入灰度水印信息的算法。该算法先把灰度水印信息转换成二进制矩阵,再按每两位结合转化成含0、1、2、3的矩阵,然后用两个不相关的伪随机序列分别代表0、1和2、3,随后形成两个伪随机矩阵。嵌入过程中,按分块方差升序的方法嵌入到中频系数中。实验证明,这种算法具有很好的抗剪切、噪音、滤波与JPEG压缩失真的能力。

In response to the weak embedding quality and stability of gray level digital image watermarking which is embeded with too much information, the authors put forward a new digital watermarking algorithm for imbedding gray level digital watermarking, information of the gray digital watermarking is converted into binary matrix , the matrix containing 0, 1,2 and 3 is converted in accordance with each combination of the two numbers. With two independent pseudo- random sequence substitute the ‘0＇ ,‘1＇ and ‘2＇ ,‘ 3＇ respectively, finally,combins into two new pesudo-- random sequence. In the process of embedding, the modulated signal is embedded in the middle frequency band of carrier image with ascending order according to variance. The experiments show the proposed algorithm is effective and robust to many common image processing operations ,especially to cutiing,noise,median filters and JPEG compression.