基于离散余弦变换的大容量信息隐藏盲提取算法

High capacity blind information hiding algorithm based on DCT

基于图像离散余弦变换(DCT)交流系数的拉普拉斯分布模型,提出了一种盲提取数字图像信息隐藏算法。该算法利用一种新颖的数据嵌入规则,调整绝对值较小的DCT系数的正负值,分别用特定DCT系数的正负值来代表1和0。提取隐藏信息的时候不需要载体图像。实验结果表明,该算法具有容量大的特性,适合于隐秘通信的各种应用。

The AC coefficients of an image＇s Discrete Cosine Transform（DCT） are best approximated by Laplacian distribution model. Based on this model, the blind information hiding algorithm was proposed. Using a novel data embedding rule, this algorithm adjusted coefficients of DCT, whose absolute value is relative smaller than others. Positive numbers were used to denote one and negative numbers were used to represent zero respectively. Furthermore, the original image was not needed to extract secret data. The experiment results show that the algorithm has high capacity attribute, which is suitable for applications of secret communication.