【摘要】 本文分析了信息隐藏变换域方法鲁棒性强的原因 ,指出其实质就是将机密信息多处隐藏。将这一思想应用于信息隐藏的空间域方法 ,提出MLSB算法。该算法把载体图像分为n×n个区域块 ,用LSB方法将机密信息隐藏到图像的每一个区域块。当图像遭受攻击后 ,图像像素值发生改变 ,通常这种改变只能发生在图像像素值的某些位。所以 ,仍然可以从未改变的像素值中提取嵌入信息。文中给出了MLSB信息嵌入和提取算法 ,并通过多方面的实验检测该算法。实验表明 ,和LSB算法相比 ,该算法的鲁棒性全面提高 ,也超过了很多信息隐藏的变化域方法。该算法对滤波亦有较好的鲁棒性。但该算法抗压缩能力 ,特别是在较低位平面时较差更多还原

【Abstract】 The scheme of transform domain has better robustness than the scheme of space domain.Why so,because it embeds information into each pixel of an image.In other words,it hides information in multiarea of a host image.Applied the idea within space domain,this paper presents MLSB(Multiarea Least Significant Bit)algorithm for information hiding.The host image is divided into n×n areas.The secret information is embedded into the each area.When the attack happen,the value of the pixel may changed.It must be noticed,this change only takes place in some bits on partly pixels,so it is still possible to pick out the information from the bits which has not yet been changed on some pixels in other areas.The experimental results show that MLSB is much better than LSB.It has very good robustness to cropping,which is better than related papers published,and good robustness to noise adding,which is even better than some scheme in transform domain.It is also has good robustness to filter,but not good to compressing,in lower bitplane especially.更多还原