基于离散余弦变换与Hessenberg分解的图像水印算法
|
摘要
为了增强水印系统的视觉隐秘性与抗几何攻击能力,设计了基于离散小波变换与Hessenberg分解的图像水印算法。将宿主图像实施分割,形成一系列的8×8的非重叠子块;随后,联合信息熵与边缘熵值,构建嵌入区域选择方法,从这些子块中确定出合适嵌入水印的子块;引入离散余弦变换DCT(Discrete Cosine Transform),对选择的子块进行分解,输出对应的低频与高频系数,将每个子块的低频系数组合成相应的矩阵;随后,利用Hessenberg分解方法处理这些低频系数矩阵,输出对应的上Hessenberg矩阵;根据结构相似度与位正确率,构建粒子群算法的适应度函数,通过对其迭代,获取最优的嵌入强度;根据优化的嵌入强度,设计水印嵌入机制,将水印信息隐藏到上Hessenberg矩阵中,形成水印图像;最后,构建水印检测方法,从水印图像中提取水印。实验数据表明:较当前鲁棒水印方案而言,所提技术具有理想的视觉不可感知性与鲁棒性。
In order to enhance the visual concealment and anti-geometric attack ability of the watermarking system, an image watermarking algorithm based on discrete wavelet transform and Hessenberg decomposition is proposed in this paper. Firstly, the host image is segmented to form a series of 8 * 8 non-overlapping blocks.Subsequently, by combining information entropy and edge entropy, the embedding region selection method is constructed to determine the appropriate embedding watermarking sub-blocks from these sub-blocks. Discrete cosine transform is introduced to decompose the selected sub-blocks for obtaining the corresponding low-frequency and high-frequency coefficients, and the low-frequency coefficients of each sub-block are combined into corresponding matrices. Then these low-frequency coefficient matrices are processed by using Hessenberg decomposition method to output the corresponding upper Hessenberg matrix. According to structural similarity and location accuracy, the fitness function of particle swarm optimization is constructed to obtain the optimal embedding strength through iteration. The watermarking embedding mechanism is designed base on the optimized embedding strength to hide the watermarking information into the upper Hessenberg matrix for obtaining the watermarking image. Finally, the watermark detection method is constructed to restore the watermark from the watermark image. The experimental data show that the proposed technique has ideal visual imperceptibility and robustness compared with the current robust watermarking schemes.
In order to enhance the visual concealment and anti-geometric attack ability of the watermarking system, an image watermarking algorithm based on discrete wavelet transform and Hessenberg decomposition is proposed in this paper. Firstly, the host image is segmented to form a series of 8 * 8 non-overlapping blocks.Subsequently, by combining information entropy and edge entropy, the embedding region selection method is constructed to determine the appropriate embedding watermarking sub-blocks from these sub-blocks. Discrete cosine transform is introduced to decompose the selected sub-blocks for obtaining the corresponding low-frequency and high-frequency coefficients, and the low-frequency coefficients of each sub-block are combined into corresponding matrices. Then these low-frequency coefficient matrices are processed by using Hessenberg decomposition method to output the corresponding upper Hessenberg matrix. According to structural similarity and location accuracy, the fitness function of particle swarm optimization is constructed to obtain the optimal embedding strength through iteration. The watermarking embedding mechanism is designed base on the optimized embedding strength to hide the watermarking information into the upper Hessenberg matrix for obtaining the watermarking image. Finally, the watermark detection method is constructed to restore the watermark from the watermark image. The experimental data show that the proposed technique has ideal visual imperceptibility and robustness compared with the current robust watermarking schemes.
引文
[1]董夙慧,孙中廷,徐永刚.基于YCoCg-R颜色空间与离散余弦变换的自适应彩色图像水印算法[J].包装工程,2018,39(13):181-187.
[2]Su Q T,Wang G,Zhang X F.A New Algorithm of Blind Color Image Watermarking Based on LUDecomposition[J].Multidimensional Systems and Signal Processing,2018,29(3):1055-1074.
[3]吴庆涛,施进发,曹再辉.基于多元频域变换与几何校正的彩色图像水印算法[J].光学技术,2018,44(4):435-442.
[4]曹再辉,吴庆涛,施进发,等.基于混合进制系统与像素差异的无损水印算法[J].计算机工程与设计,2018,39(6):1704-1710.
[5]汤永利,高玉龙,于金霞.基于DCT域的增益不变量化的数字图像水印算法[J].重庆邮电大学学报:自然科学版,2018,29(2):223-231.
[6]Zhang Z W.Wu L F,Xiao S Z.Adaptive Reversible Image Watermarking Algorithm Based on IWT and Level Set[J].EURASIP Journal on Advances in Signal Processing,2017,102(1):2-15.
[7]Musrrat A,Chang W A.An Optimal Image Watermarking Approach Through Cuckoo Search Algorithm in Wavelet Domain[J].International Journal of System Assurance Engineering and Management,2018,9(3):602-611.
[8]韩绍程,王蕊,贾桂敏.基于Hessenberg分解的彩色图像盲水印算法[J].计算机工程与设计,2017,38(12):3354-3360.
[9]张旭,郭东恩.基于DCT与自适应人类学习优化算法的图像匹配算法[J].电子测量与仪器学报,2018,32(6):148-154.
[10]Aana B.Artificial Bee Colony Approach for Enhancing LSB Based Image Steganography[J].Multimedia Tools and Applications,2018,77(20):27491-27504.
[11]晁妍,王诗兵,王慧玲.基于奇异值分解和粒子群优化算法的图像水印算法[J].吉林大学学报:理学版,2018,56(5):1163-1169.
[12]肖根福,刘欢,李东洋.一种求解大规模问题的自学习协同粒子群算法[J].井冈山大学学报:自然科学版,2018,39(3):32-37.
[13]耿卫江.结构相似度索引耦合最优稀疏表示的大规模损坏图像动态修复[J].科学技术与工程,2014,14(25):107-114.
[14]Agilandeeswari L,Ganesan K.A Bi-Directional Associative Memory Based Multiple Image Watermarking on Cover Video[J].Multimedia Tools and Applications,2016,75(12):7211-7256.
[2]Su Q T,Wang G,Zhang X F.A New Algorithm of Blind Color Image Watermarking Based on LUDecomposition[J].Multidimensional Systems and Signal Processing,2018,29(3):1055-1074.
[3]吴庆涛,施进发,曹再辉.基于多元频域变换与几何校正的彩色图像水印算法[J].光学技术,2018,44(4):435-442.
[4]曹再辉,吴庆涛,施进发,等.基于混合进制系统与像素差异的无损水印算法[J].计算机工程与设计,2018,39(6):1704-1710.
[5]汤永利,高玉龙,于金霞.基于DCT域的增益不变量化的数字图像水印算法[J].重庆邮电大学学报:自然科学版,2018,29(2):223-231.
[6]Zhang Z W.Wu L F,Xiao S Z.Adaptive Reversible Image Watermarking Algorithm Based on IWT and Level Set[J].EURASIP Journal on Advances in Signal Processing,2017,102(1):2-15.
[7]Musrrat A,Chang W A.An Optimal Image Watermarking Approach Through Cuckoo Search Algorithm in Wavelet Domain[J].International Journal of System Assurance Engineering and Management,2018,9(3):602-611.
[8]韩绍程,王蕊,贾桂敏.基于Hessenberg分解的彩色图像盲水印算法[J].计算机工程与设计,2017,38(12):3354-3360.
[9]张旭,郭东恩.基于DCT与自适应人类学习优化算法的图像匹配算法[J].电子测量与仪器学报,2018,32(6):148-154.
[10]Aana B.Artificial Bee Colony Approach for Enhancing LSB Based Image Steganography[J].Multimedia Tools and Applications,2018,77(20):27491-27504.
[11]晁妍,王诗兵,王慧玲.基于奇异值分解和粒子群优化算法的图像水印算法[J].吉林大学学报:理学版,2018,56(5):1163-1169.
[12]肖根福,刘欢,李东洋.一种求解大规模问题的自学习协同粒子群算法[J].井冈山大学学报:自然科学版,2018,39(3):32-37.
[13]耿卫江.结构相似度索引耦合最优稀疏表示的大规模损坏图像动态修复[J].科学技术与工程,2014,14(25):107-114.
[14]Agilandeeswari L,Ganesan K.A Bi-Directional Associative Memory Based Multiple Image Watermarking on Cover Video[J].Multimedia Tools and Applications,2016,75(12):7211-7256.