适合水声隐蔽通信的图像隐写算法  

Image Steganography Algorithm Suitable for Hydroacoustic Communication

作　　者：王文锦 李庆忠[2] Wang Wenjin;Li Qingzhong(School of Management Science and Engineering,Shandong University of Finance and Economics,Jinan 250014;College of Engineering,Ocean University of China,Qingdao 266100)

机构地区：[1]山东财经大学管理科学与工程学院,济南250014 [2]中国海洋大学工程学院,青岛266100

出　　处：《计算机辅助设计与图形学学报》2024年第8期1263-1271,共9页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家重点研发计划(2017YFC1405202).

摘　　要：为了实现水声信道场景中的快速、安全信息隐藏,提出一种基于杜鹃搜索(CS)与整数小波变换的水声隐蔽通信下图像隐写算法.首先对待嵌入的水下图像进行交织打包处理,生成秘密信息码流,提高传输差错检测和恢复能力;然后将载体图像划分为不重叠的块,对每个子块应用整数小波变换进行处理,既可以节省CS搜索空间,又能提高算法的寻优性能;再对高频子带系数的嵌入位置矩阵进行混沌置乱,使用CS选择最优密钥获得秘密信息的最优匹配位置,产生最佳质量的隐秘图像;最后对提取出的秘密信息采用传输差错检测与恢复措施,提高水声信道传输后水下图像的重建质量.在标准测试图像数据集SIPI上进行实验结果表明,当信息嵌入容量为0.5 bit/像素时,所提算法生成隐秘图像的平均峰值信噪比(PSNR)达到52.80 dB,平均数据结构相似性指数度量为0.9999,优于现有算法,满足水下通信安全性、实时性的要求;当丢包率高达20%时,重建图像的PSNR超过35.9 dB,在0.8%的混合误码率下,重建图像的PSNR超过36 dB,具有应对恶劣水下噪声环境下传输差错的检测与恢复能力.To realize high-speed and secure information steganography in hydroacoustic communication scenar-ios,an image steganography algorithm based on cuckoo search(CS)and integer wavelet transform under hydroacoustic communication is presented.At first,the underwater images to be embedded are interleaved and packed to generate the secret information stream to improve transmission error detection and recovery;then the cover image is partitioned into non-overlapped blocks,and the integer wavelet transform is implemented for each sub-block,in order to save CS search space and enhance the search performance of the algorithm;further,the embedding location matrix of high-frequency subband coefficients is chaotically dislocated,and the optimal matching location of the secret information is obtained by using CS to select the optimal key to generate the best quality stego image;finally,transmission error detection and recovery measures are applied to the extracted secret information to improve the reconstruction quality of underwater images after hydroacoustic transmission.Ex-perimental results on the standard test image dataset SIPI show that the proposed algorithm achieves an average peak signal to noise ratio(PSNR)of 52.80 dB and an average structure similarity index measure of 0.9999 when the information embedding capacity is 0.5 bit/pixel,meeting the requirements of underwater communication se-curity and real-time.Moreover,the PSNR of reconstructed images reaches over 35.9 dB when the packet loss rate is up to 20%,and over 36 dB at a mixed bit error rate of 0.8%,providing the ability to detect and recover from transmission errors under harsh underwater environment.

关 键 词：整数小波变换 杜鹃搜索 水声通信 图像隐写 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]