中文摘要：

针对目前主流的图像自适应隐写算法假定像素之间彼此独立,但忽略了像素之间的相关性这一问题,在最小化嵌入失真的模型下,提出了一种基于整数小波和马尔科夫链的图像自适应隐写术。首先利用整数小波对图像进行小波变换,然后利用马尔科夫链来表征邻近像素小波系数之间的相关性;并据此设计了新的失真代价函数。最后结合网格码（syndrome trellis code,STC）对秘密信息进行嵌入。实验表明该隐写术能够有效抵抗通用隐写分析,在安全性上相比已有的三种隐写术有明显提升。

英文摘要：

The current mainstream image adaptive steganography assumes that the pixels are independent of each other,but this hypothesis ignores the correlation between pixels.For this problem,an adaptive steganography was proposed based on the integer wavelet and Markov Chain for digital images under the model of minimizing a suitable defined distortion.Firstly wavlet transforming to the image by integer wavelet was done.Secondly first-order Markov chain was used to represent the correlation between adjacent pixels,and designed a new distortion cost function.Finally,the secret message is embedded by syndrome trellis Code（STC） according to the pixel＇s embedding distortion.Experiment results show that the new adaptive steganography proposed can effectively resist the common steganalysis detection,and achieves better security comparing to other three popular adaptive steganography.