中文摘要：

在最小化嵌入失真隐写算法框架的基础上，提出了一种基于小波域的非加性失真函数图像隐写算法。首先在离散小波域设计了加性失真函数，主要考虑了亮度、频率和纹理掩蔽因子对载体失真的影响；然后结合模型矫正的方法对设计的加性失真函数进行了非加性矫正；最后结合网格码设计了隐写算法，将嵌入信息对载体的修改最小化，并且集中在人眼不敏感区域。实验结果显示，方案具有良好的视觉不可见性，且能抵抗空域SPAM特征隐写分析的攻击。

英文摘要：

In the framework of minimization distortion steganography, an image steganography method based on minimizing embedding impact and non-additive distortion function is proposed. An additive distortion functionis first designed, considering mainly the influences of frequency, luminance and texture masking factors on the cover distortion. Then the Non-additive distortion function based on the model correction is obtained. The Syndrome-Trellis Codes is combined with the distortion function to propose the steganography method to make sure that the embed- ding impact on the cover is minimized and centralizes in in-sensitivie domain. Experiment results show that the proposed approach maintains a good visual quality of stego-image and has a high security against steganalysis with SPAM features.