【摘要】 为了实现长文本信息的隐藏和保密通信,算法先对一个文本文件中的文本内码进行二值化和冗余编码,得到二值水印序列;然后在载体图像分块的Hilbert扫描序列中,每次选取两相邻分块分别进行一层离散小波变换,并利用一个视觉模型计算图像分块的双临界噪声阈值;再根据每两位二值水印编码的不同形式,分别选择双临界噪声阈值之一,来调整两相邻分块各对应细节子带均值之间的大小关系;最后,进行一层离散小波逆变换,以实现信息的隐藏。实验证明,被该算法隐藏的信息具有很好的透明性,且该算法对几种常见图像攻击具有较好的鲁棒性。故该算法是有效和实用的。更多还原

【Abstract】 For realizing of long text information hiding and secrecy communication,a binary watermark was obtained firstly from a text file and encoded by a redundant encoding method.Then,two neighboring blocks were selected each time from the Hilbert scanning sequence of the carrier image blocks,and transformed by 1-level DWT.And then the double block based JNDs(just noticed difference) were calculated with a human visual model.According to the different codes of each two bits watermark,the average values of the two corresponding detail sub-bands were modified by use of one of JNDs.Lastly,the 1-level inverse DWT was executed to hide the text information into the carrier image.The experimental results show that the hidden information is invisible to human eyes,and the algorithm is robust to some common image processing operations.The conclusion is that the algorithm is effective and practical.更多还原