GIF图像掩密算法的统计安全性比较

Statistical Security Comparison of Steganographic Algorithm on GIF Images

利用一维统计特征的检测方法,高阶统计特征的检测方法对几种常用的信息隐藏方法的安全性进行了比较研究。实验结果表明,只改动图像索引值的信息隐藏方法的安全性优于同时改动调色板及索引值的信息隐藏方法。Fridrich信息隐藏算法由于采用了最优奇偶值分配的方法,其安全性优于EZStego。

The principle of steganography on GIF images is introduced in this paper. Some familiar steganographic methods based on GIF images are compared by the detecting methods on one dimension statistical character and multiple dimension statistical character. The results of experiments indicate that data hiding methods on modifying indices of images have better security than methods on modifying indices and palette of images together. The steganographic method presented by Fridrich is more secure than EZ Stego because it uses the optimal parity assignment.