中文摘要：

隐蔽性和隐藏容量是信息隐藏系统的两个重要特性，同时信息隐藏系统还必须具有一定的安全性。本文充分考虑以上特性，利用混沌映射和人类视觉模型，提出一种大容量的离散余弦变换域信息隐藏算法。首先将载体图像分成互不重叠的8×8图像块，并对各图像块分别进行离散余弦变换，然后将密钥作为混沌系统初值映射生成混沌序列，以此序列确定嵌入秘密信息的图像块位置，并利用人类视觉模型计算的JND值进行图像块中可用DCT中频系数对的选取，最后通过调整中频系数相对幅值策略自适应地嵌入秘密信息。仿真结果表明该算法不仅具有大的隐藏容量，而且秘密信息具有非常好的视觉不可见性，同时对JPEG压缩、剪切、中值滤波和高斯噪声等常见的攻击具有较好的鲁棒性。

英文摘要：

The payload and imperceptibility are the two most important properties of the information hiding system. Intrinsically, these two properties conflict with each other. In addition, security features in the information hiding system. Giving full consideration to all these characteristics, a high capacity information hiding algorithm is proposed for the discrete cosine transform domain based on chaotic mapping and the human visual model. The carrier image is decomposed into nonoverlapping 8 × 8 image blocks, and discrete cosine transform is implemented independently with each image block. The key is taken as the initial value to generate the orderly sequence which is used to identify the location of the image block embedded with secret messages. The secret messages are embedded adaptively into the usable middle frequency coefficient pairs judged by Just Noticeable Difference（JND）. Simulation results show that the algorithm has a high capacity and good invisibility, and that it is robust for the common image processing like JPEG compression and cropping and so on.