摘要
在JPEG图像中隐写会引起图像DCT系数直方图变化和分块效应比,分析者可以根据这些特征变化察觉到秘密信息的存在。为了提高隐写的安全性,提出了一种基于混合蛙跳优化的JPEG图像隐写算法。对图像进行分块处理后,在每个块中采用混合蛙跳优化算法,寻找最优的置换矩阵来替换秘密信息。在隐写过程中,将置换矩阵和置换后的秘密信息隐藏于修改了量化表的DCT系数中,算法不需传递其他边缘信息。仿真实验表明,所提出的隐写算法与修改量化表隐写算法,基于粒子群优化隐写算法相比,在隐写容量增大的同时,具有更高的安全性。
JPEG steganography can cause the changes in the histogram of the image DCT coefficients and the ratio of blocks.Steganalysis can perceive the existence of secret information based on the changes in these characteristics.In order to improve the security of steganography,this paper proposes a JPEG image steganography algorithm based on SFLA(Shuffled Frog Leaping Algorithm).First,the image is divided into some blocks.Then,SFLA is used in each block to look for the optimal permutation matrix to replace the secret information.In the steganography process,the permutation matrix and the secret information are hidden in the DCT coefficients of the modified quantization table,so the proposed steganography does not need to transmit other edge information.Simulation experiments show that,compared with JQTM(JPEG and Quantization Table Modification)and the JQTM steganography optimized by PSO(Particle swarm optimization),the proposed steganography has higher security while increasing the steganography capacity.
作者
熊勇
欧阳迎春
欧阳春娟
XIONG Yong;OUYANG Ying-chun;OUYANG Chun-juan(School of Electromechanical,Jinggangshan University,Ji'an,Jiangxi 343009,China;Guangdong Xin'an Technical Institute,Shenzhen,Guangdong 518000,China;School of Electronic and Information Engineering,Jinggangshan University,Ji'an,Jiangxi 343009,China)
出处
《井冈山大学学报（自然科学版）》
2021年第5期57-62,共6页
Journal of Jinggangshan University (Natural Science)
基金
国家自然科学基金项目(61462046,42061055)
江西省自然科学基金项目(20192BAB207021,20202BABL202047)
江西省教育厅科技计划项目(GJJ201008)
江西省高校人文社科项目(YS20129)。
