版权保护和内容认证的全盲双功能数字水印算法

doi:10.11959/j.issn.1000-0801.2017017

摘要/Abstract

摘要：

针对现有大部分水印算法功能单一且无法实现完全盲检测的局限性，提出一种同时进行版权保护和内容认证的全盲双功能数字水印算法。算法采取子块区域分割的思想，水印产生和嵌入分别利用同一个子块的不同区域。通过判别两个最大奇异值均值的最高位奇偶性产生特征水印，然后自适应地调整相邻区域相同位置的两个离散余弦变换交流系数的大小，实现水印的嵌入。检测端通过产生的特征水印和盲提取的认证水印实现全盲的版权鉴别和内容认证。相比于已报道的一些水印方案，本文提出的水印算法对于常见的攻击具有更好的顽健性，可以实现版权保护功能，同时对于恶意的图像篡改又有较好的敏感性，可以准确定位篡改的区域。

关键词: 双功能水印, 特征水印, 全盲检测, 版权保护, 篡改检测

Abstract:

In view of the limitation that most of the watermarking algorithms are single in functions and unable to achieve perfectly blind detection,a bifunctional watermarking algorithm with perfectly blind detection for copyright protection and content authentication was proposed.The algorithm adopted the idea of region segmentation for sub-block,and the generation and embedding of watermark were respectively carried out in the different regions of the same sub-block.A feature watermark was generated by judging the parity of the first digit from the average of maximum singular values,and the watermark was embedded by adjusting two DCT coefficients of adjacent blocks in the same position.Both the image copyright protection and content authentication functions were achieved by comparing the generated feature watermark and blindly extracting authentication watermark in the detection.Compared to some of the reported watermarking schemes,the proposed algorithm not only has better robustness against the common attacks to realize the copyright protection,but also has a good sensitivity to malicious image tampering to accurately locate the tampered area.

Key words: dual function watermark, feature watermark, perfectly blind detection, copyright protection, tamper detection

中图分类号:

TN911.7

薛栋,周亚训,金炜. 版权保护和内容认证的全盲双功能数字水印算法[J]. 电信科学, 2017, 33(2): 79-89.

Dong XUE,Yaxun ZHOU,Wei JIN. Bifunctional watermarking for copyright protection and content authentication with perfectly blind detection[J]. Telecommunications Science, 2017, 33(2): 79-89.

图/表 16

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

图12

表1

表2

图像抗攻击的顽健性"

图像	编号	1	2	3	4	5	6	7	8
Lena	$\hat{W} 和 {\tilde{W}}^{'}$	0.986 3	0.974 6	0.974 0	0.960 7	0.993 5	0.942 8	0.985 0	0.986 0
	$\hat{W} 和 W$	0.986 3	0.974 6	0.985 7	0.979 7	0.993 5	0.977 8	0.997 4	0.992 6
	${\tilde{W}}^{'} 和 W$	1.000 0	1.000 0	0.987 3	0.980 1	1.000 0	0.962 1	0.987 6	0.992 0
Goldhill	$\hat{W} 和 {\tilde{W}}^{'}$	0.987 9	0.977 7	0.973 3	0.957 6	0.993 0	0.923 1	0.985 3	0.984 2
	$\hat{W} 和 {\tilde{W}}^{'}$	0.987 9	0.977 7	0.984 0	0.977 9	0.993 0	0.973 6	0.988 8	0.996 8
	${\tilde{W}}^{'} 和 W$	1.000 0	1.000 0	0.988 8	0.979 7	1.000 0	0.943 2	0.996 5	0.992 5
Peppers	$\hat{W} 和 {\tilde{W}}^{'}$	0.981 6	0.967 1	0.977 7	0.961 1	0.986 1	0.939 7	0.985 6	0.984 8
	$\hat{W} 和 {\tilde{W}}^{'}$	0.981 6	0.967 1	0.988 1	0.978 1	0.986 1	0.972 7	0.991 2	0.990 9
	${\tilde{W}}^{'} 和 W$	1.000 0	1.000 0	0.989 0	0.981 2	1.000 0	0.962 2	0.994 3	0.993 4

图像	编号	9	10	11	12	13	14	15	16
Lena	$\hat{W} 和 {\tilde{W}}^{'}$	0.974 7	0.970 4	0.929 8	0.946 2	0.893 9	0.969 0	0.941 4	0.981 8
	$\hat{W} 和 {\tilde{W}}^{'}$	0.974 9	0.972 3	0.968 6	0.953 7	0.911 8	0.975 5	0.975 4	0.981 8
	${\tilde{W}}^{'} 和 W$	0.999 8	0.998 1	0.961 3	0.991 1	0.975 9	0.991 8	0.964 4	1.000 0
Goldhill	$\hat{W} 和 {\tilde{W}}^{'}$	0.977 9	0.974 7	0.951 5	0.942 1	0.892 7	0.967 8	0.915 8	0.982 6
	$\hat{W} 和 {\tilde{W}}^{'}$	0.977 9	0.976 3	0.970 5	0.949 3	0.911 5	0.974 4	0.967 3	0.992 6
	${\tilde{W}}^{'} 和 W$	1.000 0	0.998 4	0.979 4	0.990 5	0.911 4	0.992 5	0.943 7	1.000 0
Peppers	$\hat{W} 和 {\tilde{W}}^{'}$	0.970 3	0.955 3	0.908 9	0.943 4	0.884 8	0.986 1	0.937 6	0.974 4
	$\hat{W} 和 {\tilde{W}}^{'}$	0.970 5	0.959 7	0.956 8	0.947 9	0.894 8	0.986 1	0.965 4	0.974 7
	${\tilde{W}}^{'} 和 W$	0.999 3	0.994 8	0.943 7	0.993 6	0.981 3	1.000 0	0.968 4	0.999 8

表2

图13

图14

参考文献 17

[1]	PREDA R O . Semi-fragile watermarking for image authentication with sensitive tamper localization in the wavelet domain[J]. Measurement, 2013,46(1):367-373.
[2]	叶天语 . 双重功能图像水印算法[J]. 光电子·激光， 2011,22(4): 612-617.
	YE T Y . An image watermarking algorithm with dual purpose[J]. Journal of Optoelectronics·Laser, 2011,22(4): 612-617.
[3]	吕建平，刘齐 . 基于混沌理论和量化方法的双功能水印算法[J]. 西安邮电大学学报， 2014,19(3): 54-57.
	LV J P , LIU Q . Bifunctional watermarking algorithm based on chaos theory and quantitative methods[J]. Journal of Xi'An University of Posts and Telecommunications, 2014,19(3): 54-57.
[4]	叶天语，钮心忻，杨义先 . 多功能双水印算法[J]. 电子与信息学报， 2009,31(3): 546-551.
	YE T Y , NIU X X , YANG Y X . A multi-purpose dual watermark algorithm[J]. Journal of Electronics ＆ Information Technology, 2009,31(3): 546-551.
[5]	陈光喜，尹柳，易招师 . 基于整数提升小波变换的多功能数字水印[J]. 计算机工程与应用， 2010,46(11): 115-118.
	CHEN G X , YIN L , YI Z S . Multi-purpose digital watermark based on integer lifting wavelet transformation[J]. Computer Engineering and Application, 2010,46(11): 115-118.
[6]	叶天语 . 自嵌入双功能图像水印算法[J]. 光子学报， 2012,41(7): 859-867.
	YE T Y . A self-embedding image watermarking scheme with dual purpose[J]. Acta Photonica Sinica, 2012,41(7): 859-867.
[7]	朱光，师文，朱学芳，等. 结合SURF特征的多功能彩色图像水印算法[J]. 中国图象图形学报， 2013,18(12): 1694-1702.
	ZHU G , SHI W , ZHU X F , et al. Multi-purpose watermark of color image based on SURF[J]. Journal of Image and Graphics, 2013,18(12): 1694-1702.
[8]	赵春晖，刘巍 . 基于压缩感知的交互支持双水印算法[J]. 电子学报， 2012,40(4): 681-987.
	ZHAO C H , LIU W . Mutual support dual watermark algorithm based on compressive sensing[J]. Acta Electronica Sinica, 2012,40(4): 681-987.
[9]	孟倩，王希常，刘江 . 基于DCT变换的数字水印算法研究[J]. 微型机与应用， 2012,29(23): 37-40.
	MENG Q , WANG X C , LIU J . Research on the digital watermark based on DCT[J]. Image Processing and Multimedia Technology, 2012,29(23): 37-40.
[10]	PARAH S A , SHEIKH J A , LOAN N A , et al. Robust and blind watermarking technique in DCT domain using inter-block coefficient differencing[J]. Digital Signal Processing, 2016,53(C):11-24.
[11]	DAS C , PANIGRAHI S , SHARMA V K , et al. A novel blind robust image watermarking in DCT domain using inter-block coefficient correlation[J]. AEU-International Journal of Electronics and Communications, 2014,68(3):244-253.
[12]	QI X , XIN X . A singular-value-based semi-fragile watermarking scheme for image content authentication with tamper localization[J]. Journal of Visual Communication ＆ Image Representation, 2015,30(C):312-327.
[13]	RANI A , BHULLAR A K , DANGWALD , et al. A zero-watermarking scheme using discrete wavelet transform[J]. Procedia Computer Science, 2015,70:603-609.
[14]	王友卫，申铉京，吕颖达，等. 基于DCT和SVD变换的盲数字水印算法[J]. 计算机工程与应用， 2011,47(21): 157-161.
	WANG Y W , SHEN X J , LV Y D , et al. Blind digital watermarking algorithm based on DCT and SVD transform[J]. Computer Engineering and Applications, 2011,47(21): 157-161.
[15]	翟依依，王光义 . 基于Tent混沌序列的数字图像加密方法[J]. 现代电子技术， 2014,37(12): 73-77.
	ZHAI Y Y , WANG G Y . Digital image encryption algorithm based on tent chaotic sequence[J]. Modern Electronics Technique, 2014,37(12): 73-77.
[16]	叶天语 . 基于子块区域分割和自嵌入技术的全盲多功能图像水印算法[J]. 通信学报， 2013,34(3): 148-156.
	YE T Y . Perfectly blind image watermarking scheme with multi-purpose based on region segment for sub-block and self-embedding technology[J]. Journal on Communications, 2013,34(3): 148-156.
[17]	叶天语 . DWT-SVD域全盲自嵌入顽健量化水印算法[J]. 中国图象图形学报， 2012,17(6): 644-650.
	YE T Y . Perfectly blind self-embedding robust quantization-% based watermarking scheme in DWT-SVD domain[J]. Journal of Image and Graphics, 2012,17(6): 644-650.

编号	攻击类型
1	添加高斯噪声（方差为0.000 1）
2	添加高斯噪声（方差为0.000 3）
3	添加椒盐噪声（噪声密度为0.003）
4	添加椒盐噪声（噪声密度为0.005）
5	3×3高斯低通滤波（σ=0.5）
6	3×3中值滤波
7	剪切（左上角1/16）
8	剪切（中心1/16）
9	JPEG压缩（Q=60）
10	JPEG压缩（Q=50）
11	JPEG压缩（Q=40）
12	重采样（先缩小0.8倍再扩大到1.25倍）
13	重采样（先缩小到0.5倍再扩大到2倍）
14	JPEG70+中心剪切1/16
15	JPEG70+3×3中值滤波
16	JPEG70+3×3高斯低通滤波（σ=0.5）