基于蜜蜂进化算法的JPEG隐写方法

JPEG steganography based on Bee Evolutionary Genetic Algorithm

摘要:

为提高嵌入效率，缩短算法运行时间，提出了一种基于蜜蜂进化算法的变换域隐写方法。该方法首先利用由蜜蜂进化算法搜索到的最优置换矩阵将秘密信息进行置换；接着修改了JPEG标准量化表以便能够嵌入较多的信息；将置换结果嵌入到量化后DCT系数的DC和中低频AC区域中；最后利用JPEG熵编码得到隐藏了信息的JPEG格式文件。实验结果表明，与直接嵌入方法相比较，该方法提高了含密图像的质量和安全性，与基于遗传算法的信息隐藏方法相比，该算法所花费时间更少。

Abstract:

In order to improve the embedding efficiency and reduce the operating time，a novel steganographic method，based on JPEG and Bee Evolutionary Genetic Algorithm(BEGA)，was proposed. The proposed method first derived an optimal substitution matrix by BEGA for transforming the secret messages. By means of the substitution strategy，the quality of stego-images was improved. Next，it modified the standard JPEG quantization table for the purpose of accommodating more secret messages. The transformed messages were then hidden in the cover-image with its DC-to-middle frequency components of the quantized DCT coefficients modified. Finally，a JPEG file was generated through JPEG entropy coding. The experimental results showed that the proposed method had achieved a better image quality and security than steganographic method without optimal substitution，and it had costed less computation time than the genetic algorithm-based information hiding method.