基于小波域的数字水印方法研究
|
摘要
随着互联网的飞速发展以及多媒体资源的数字化,数字作品的侵权、盗版和随意篡改问题日益严重,使得数字水印成为重要的研究热点。
本文首先介绍了数字水印的原理、基本特性以及目前国内外的研究现状等基本问题,给出了数字水印各部分的基本框架,比较了现有水印算法的优缺点,并分析了当前数字水印技术中存在的问题。在水印嵌入算法方面重点介绍了离散小波变换域数字水印技术,由于离散小波变换域兼具有空间域和离散余弦变换域的优点,以它作为数据压缩、处理和分析工具在数字水印中有很好的应用。本文提出了两种变换域的水印算法,其中引入一种机器学习方法一支持向量机的方法来最大程度的提高水印的综合性能。
本文主要研究内容如下：
1.提出一种小波变换域的数字图像水印算法。该算法结合水印信息的置乱变换及人类视觉系统特性,在图像的小波变换不同尺度下的中频子带中嵌入经过预处理后的水印信息。提取过程需要原始图像参与,属于非盲水印。
2.提出一种基于支持向量机回归的数字图像水印算法。结合小波变换的多分辨率特点和支持向量机在理论和学习上的优势,提出了小波域中基于支持向量回归的数字图像水印算法。首先根据小波变换空频局域性特点,给出小波系数方向树结构,利用支持向量机回归建立方向树上根节点与其子孙节点之间的依赖关系模型,进而设计水印嵌入规则。在水印的提取过程中不需要原始图像参与,属于盲水印。
本文对所提出的水印算法均进行了大量的不可见性和鲁棒性仿真实验。结果表明两种图像水印算法均具有很好的不可见性,对常见的图像处理操作具有很强的鲁棒性。
With the rapid development boom of internet and digitization of multimedia resources, the problems of infringement, pirated illegal copy and malicious tampering of digital works are getting more serious, which makes digital watermarking an important research focus.
Firstly, the dissertation introduces the basic principle and characteristics of the digital watermarking as well as its research status domestic and abroad, and then gives the basic framework of every parts.compares the strengths and weaknesses of the existing watermarking algorithms and analyzed some existing questions of current digital watermark technology.
In the watermark embedding algorithm, this paper focuses on DWT transform domain digital watermarking embedding technique, which has been widely used as data compression, processing and analyzing tools in digital watermarking because the DWT transform domain possesses advantageous of space domain and DCT transform domain. In this dissertation, two kinds of transform domain watermarking algorithm has been proposed, and a machine learning method named supports vector machines also has been introduceed to maximize the overall performance improvement of the watermark.
The mainly research result is as follows:
1. A digital image watermarking algorithm in wavelet transform domain is proposed. With this algorithm, combined the watermark scrambling transformation and human visual system characteristics, the watermark information after pretreatment are embedded into middle frequency subband of the images under different scales of wavelet transform. The original image is required to take part in extraction, whose algorithm belongs to non-blind watermark.
2. A digital image watermarking algorithm based on the support vector machines is proposed. The watermarking schemes are proposed based on the coefficient direction-tree structure in the discrete wavelet transform domain. Firstly, the concept of wavelet coefficient direction-tree is defined. then the inherent relation between root node and its offspring in the direction-tree is discovered and modeled using SVM. Finally, the watermark embedding rules are designed. The original image is not required to take part in extraction, whose algorithm belongs to blind watermark.
In this dissertation, the proposed watermarking algorithm are carried out a lot of invisibility and robustness of the simulation experiment. The results show that two kinds of image watermarking algorithms do not all have good visibility of the common image processing operations are highly robust.
本文首先介绍了数字水印的原理、基本特性以及目前国内外的研究现状等基本问题,给出了数字水印各部分的基本框架,比较了现有水印算法的优缺点,并分析了当前数字水印技术中存在的问题。在水印嵌入算法方面重点介绍了离散小波变换域数字水印技术,由于离散小波变换域兼具有空间域和离散余弦变换域的优点,以它作为数据压缩、处理和分析工具在数字水印中有很好的应用。本文提出了两种变换域的水印算法,其中引入一种机器学习方法一支持向量机的方法来最大程度的提高水印的综合性能。
本文主要研究内容如下：
1.提出一种小波变换域的数字图像水印算法。该算法结合水印信息的置乱变换及人类视觉系统特性,在图像的小波变换不同尺度下的中频子带中嵌入经过预处理后的水印信息。提取过程需要原始图像参与,属于非盲水印。
2.提出一种基于支持向量机回归的数字图像水印算法。结合小波变换的多分辨率特点和支持向量机在理论和学习上的优势,提出了小波域中基于支持向量回归的数字图像水印算法。首先根据小波变换空频局域性特点,给出小波系数方向树结构,利用支持向量机回归建立方向树上根节点与其子孙节点之间的依赖关系模型,进而设计水印嵌入规则。在水印的提取过程中不需要原始图像参与,属于盲水印。
本文对所提出的水印算法均进行了大量的不可见性和鲁棒性仿真实验。结果表明两种图像水印算法均具有很好的不可见性,对常见的图像处理操作具有很强的鲁棒性。
With the rapid development boom of internet and digitization of multimedia resources, the problems of infringement, pirated illegal copy and malicious tampering of digital works are getting more serious, which makes digital watermarking an important research focus.
Firstly, the dissertation introduces the basic principle and characteristics of the digital watermarking as well as its research status domestic and abroad, and then gives the basic framework of every parts.compares the strengths and weaknesses of the existing watermarking algorithms and analyzed some existing questions of current digital watermark technology.
In the watermark embedding algorithm, this paper focuses on DWT transform domain digital watermarking embedding technique, which has been widely used as data compression, processing and analyzing tools in digital watermarking because the DWT transform domain possesses advantageous of space domain and DCT transform domain. In this dissertation, two kinds of transform domain watermarking algorithm has been proposed, and a machine learning method named supports vector machines also has been introduceed to maximize the overall performance improvement of the watermark.
The mainly research result is as follows:
1. A digital image watermarking algorithm in wavelet transform domain is proposed. With this algorithm, combined the watermark scrambling transformation and human visual system characteristics, the watermark information after pretreatment are embedded into middle frequency subband of the images under different scales of wavelet transform. The original image is required to take part in extraction, whose algorithm belongs to non-blind watermark.
2. A digital image watermarking algorithm based on the support vector machines is proposed. The watermarking schemes are proposed based on the coefficient direction-tree structure in the discrete wavelet transform domain. Firstly, the concept of wavelet coefficient direction-tree is defined. then the inherent relation between root node and its offspring in the direction-tree is discovered and modeled using SVM. Finally, the watermark embedding rules are designed. The original image is not required to take part in extraction, whose algorithm belongs to blind watermark.
In this dissertation, the proposed watermarking algorithm are carried out a lot of invisibility and robustness of the simulation experiment. The results show that two kinds of image watermarking algorithms do not all have good visibility of the common image processing operations are highly robust.
引文
[1]Cox I J,Miller M L,,Bloom J A.数字水印[M].王颖,黄志蓓等译.北京：电子工业出版社,2003.
[2]王炳锡,陈琦,邓峰森.数字水印技术[M].西安电子科技大学出版社,2003.
[3]李媛姗,闫雪虎.国防科技工业信息安全标准应用与发展的探讨[J].航空标准化与质量,2009,2(229)：38-42.
[4]钟桦,张小华,焦李成.数字水印与图像认证-算法及应用[M].西安电子科技大学出版社,2006.
[5]钮心忻.信息隐藏与数字水印[M].北京邮电大学出版社,2004年6月,第1版,pp：6-8.
[6]苏育挺,张春田.关于信息隐藏信道模型的讨论,信息隐藏学术研讨会论文集[M],2001年9月第一版,306-311.
[7]孙圣和,陆哲明,牛夏牧.数字水印技术和应用[M].北京科学出版社,2004.
[8]Zhao Jian, K.Eckhard.A generic digital watermarking model[J]. Computer&Graphics,1998,22(4): 397-403.
[9]Schyndel R..Tirkel A.,Oshorne C.A digital watermark [J]. In:Proceedings of the IEEE International Conference on Image Processing.Austin,Texas,1994,2:86-90.
[10]J Fridrich,M Goljan.On Estimation of Secret Message Length in LSB Steganography in Spatial Domain [J]. Proceedings of SPIE. Bellingham:Internationgal Society for Optical Engineering,2004,530 6:23-34.
[11]W Bender.Techniques for Data Hiding [J]. IBM Systems Journal,1996,35:313-336.
[12]Cox I J,Kilian J.Leighton F T,et al.Secure spread spectrum watermarking for multimedia[J].IEEE Transactions on Image Processing.1997,6(12):1673-1687.
[13]Adams M D,Kossentini F.Jasper:a software-based JPEG-2000 codec implenmentation[J].IEEE International Conference on Image Processing,Vancouver, CA,2000,9:1204-1211.
[14]张军,王能超,用于图像认证的基于神经网络的水印技术[J],计算机辅助设计与图形学学报,2003,15(3)：307-312.
[15]王宏霞,何震,丁科,基于混沌映射的鲁棒性公开水印[J],软件学报,2004,15(8)：1245-1251.
[16]Shieh C.S.,Genetic watermarking based on transform domain Technique[J]. Pattern recognition,2004, 37:555-565.
[17]张新法,秦茂玲.基于人类视觉特性HVS和DWT的数字水印算法[J].计算机技术与发展,2007,17(7)：135-138.
[18]Xia-Mu Niu.Zhe-Ming Lu, Sheng-He Sun. Digital image watermarking based on multiresolution decomposition[J]. Electronics Letters,2002,36(14):1108-1110.
[19]李春,黄继武.一种抗JPEG压缩的半脆弱图像水印算法[J].软件学报,2006,17(2)：315-324.
[20]Zhang Jun, Wang Nengchao, Xiong Feng. A Novel Watermarking for Images Using Neural Networks[J]. International conference on Machine Learning and Cybernetics,2002,11(3): 1405-408.
[21]Mingshing Hsieh. Dinchang Tseng. Yong Huai Huang. Hiding Digital Watermarks using multiresolution wavelet transform[J]. IEEE Transactions on Industrial Electronics,2001, 48(5):875-882.
[22]张旭东,卢国栋,冯健编.图像编码基础和小波压缩技术[M].北京：清华大学出版社,2004.
[23]刘芳,贾成,冯雁,刘歆.图像置乱在数字水印中的应用研究[J].通信技术,2008,41(19)：165-172.
[24]彭军,刘勇国.基于Logistic映射的图像保密通信[J].计算机学报.2003,26(1)：102-107.
[25]易开祥.数字图像加密与数字水印技术研究[D].博士学位论文,浙江大学,2001.7.
[26]孙铁锋,温泉,王树勋.基于人类视觉的混沌阵列在图像上的水印算法[J].电子学报.2003,31(1)：149-154.
[27]沈柯.光学中的混沌[M].长春：东北师范大学出版社,2000.
[28]Vapnik V N. Statistical Learning Theory[M]. NewYork:John Wiley and Sons,1998.
[29]邓乃扬,田英杰.数据挖掘中的新方法-支持向量机[M].科学出版社,2006：224-235.
[30]Steve R Gunn.Support vector machines for classification and regression[R].Technical report. Southampton University of Southampton, Faculty of Engineering,Science and Mathemstics Scod of Electronics and Computer Science,1998,5.10:1-28.
[31]Vapnik V N, Chervonenkis A. On the uniform convergence of relative frequencies of events to their probabilities. Theory of Probability and its Applications[J],1991,16(2):264-280.
[32]Vapnik V N, Chervonenkis A. The necessary and sufficienct conditions for consistency in the emprirical risk minimization method[J]. Pattern Recognition and Image Analysis,1991,1(3):283-305.
[33]Vapnik著,张学工译.统计学习理论的本质.北京：清华大学出版社.2000.
[34]Li C.H..Lu Z.D.,Zhou K.SVR-Parameters Selection for Image watermarking[J].In:Proceedings of 17th IEEE International Conference on Tools with Artificial Intelligence,2005,466-470.
[35]Xiang-Yang Wang.Hong-Ying Yang,Chang-Ying Cui,An AVM-based robust digital image atermarking against desynchronizationg attacks[J],Signal Processing,88(2008)2193-2205, September 2008.
[36]Li C.H.,Lu Z.D.,Zhou K.An image watermarking technology based on support vector regression[J]. In:Proceedings of the IEEE International Symposium on Communications and Information Technologies. 2005,177-180.
[37]Li Chunhua, Ling hefei, Lu Zhengding. Adaptive Spatial Domain Image Watermarking Based on Support Vector Machine[J]. Journal of Computer Research and Development,2007,44(8):1399-1405
[38]阎辉.基于核函数的机器学习方法研究[D].清华大学.2000.
[39]阎威武.支持向量机理论、方法和应用研究[D],上海交通大学.2003.
[40]SakrN, Zhao Jiying, GrozaV. A Dynamic Fuzzy Logic Approach to Adaptive HVS-Based Watermarking[J], Proc of the IEEE International Workshop on Haptic Audio Visual Environments and Their Applications.Ottawa, Canada,2005:121-126.
[2]王炳锡,陈琦,邓峰森.数字水印技术[M].西安电子科技大学出版社,2003.
[3]李媛姗,闫雪虎.国防科技工业信息安全标准应用与发展的探讨[J].航空标准化与质量,2009,2(229)：38-42.
[4]钟桦,张小华,焦李成.数字水印与图像认证-算法及应用[M].西安电子科技大学出版社,2006.
[5]钮心忻.信息隐藏与数字水印[M].北京邮电大学出版社,2004年6月,第1版,pp：6-8.
[6]苏育挺,张春田.关于信息隐藏信道模型的讨论,信息隐藏学术研讨会论文集[M],2001年9月第一版,306-311.
[7]孙圣和,陆哲明,牛夏牧.数字水印技术和应用[M].北京科学出版社,2004.
[8]Zhao Jian, K.Eckhard.A generic digital watermarking model[J]. Computer&Graphics,1998,22(4): 397-403.
[9]Schyndel R..Tirkel A.,Oshorne C.A digital watermark [J]. In:Proceedings of the IEEE International Conference on Image Processing.Austin,Texas,1994,2:86-90.
[10]J Fridrich,M Goljan.On Estimation of Secret Message Length in LSB Steganography in Spatial Domain [J]. Proceedings of SPIE. Bellingham:Internationgal Society for Optical Engineering,2004,530 6:23-34.
[11]W Bender.Techniques for Data Hiding [J]. IBM Systems Journal,1996,35:313-336.
[12]Cox I J,Kilian J.Leighton F T,et al.Secure spread spectrum watermarking for multimedia[J].IEEE Transactions on Image Processing.1997,6(12):1673-1687.
[13]Adams M D,Kossentini F.Jasper:a software-based JPEG-2000 codec implenmentation[J].IEEE International Conference on Image Processing,Vancouver, CA,2000,9:1204-1211.
[14]张军,王能超,用于图像认证的基于神经网络的水印技术[J],计算机辅助设计与图形学学报,2003,15(3)：307-312.
[15]王宏霞,何震,丁科,基于混沌映射的鲁棒性公开水印[J],软件学报,2004,15(8)：1245-1251.
[16]Shieh C.S.,Genetic watermarking based on transform domain Technique[J]. Pattern recognition,2004, 37:555-565.
[17]张新法,秦茂玲.基于人类视觉特性HVS和DWT的数字水印算法[J].计算机技术与发展,2007,17(7)：135-138.
[18]Xia-Mu Niu.Zhe-Ming Lu, Sheng-He Sun. Digital image watermarking based on multiresolution decomposition[J]. Electronics Letters,2002,36(14):1108-1110.
[19]李春,黄继武.一种抗JPEG压缩的半脆弱图像水印算法[J].软件学报,2006,17(2)：315-324.
[20]Zhang Jun, Wang Nengchao, Xiong Feng. A Novel Watermarking for Images Using Neural Networks[J]. International conference on Machine Learning and Cybernetics,2002,11(3): 1405-408.
[21]Mingshing Hsieh. Dinchang Tseng. Yong Huai Huang. Hiding Digital Watermarks using multiresolution wavelet transform[J]. IEEE Transactions on Industrial Electronics,2001, 48(5):875-882.
[22]张旭东,卢国栋,冯健编.图像编码基础和小波压缩技术[M].北京：清华大学出版社,2004.
[23]刘芳,贾成,冯雁,刘歆.图像置乱在数字水印中的应用研究[J].通信技术,2008,41(19)：165-172.
[24]彭军,刘勇国.基于Logistic映射的图像保密通信[J].计算机学报.2003,26(1)：102-107.
[25]易开祥.数字图像加密与数字水印技术研究[D].博士学位论文,浙江大学,2001.7.
[26]孙铁锋,温泉,王树勋.基于人类视觉的混沌阵列在图像上的水印算法[J].电子学报.2003,31(1)：149-154.
[27]沈柯.光学中的混沌[M].长春：东北师范大学出版社,2000.
[28]Vapnik V N. Statistical Learning Theory[M]. NewYork:John Wiley and Sons,1998.
[29]邓乃扬,田英杰.数据挖掘中的新方法-支持向量机[M].科学出版社,2006：224-235.
[30]Steve R Gunn.Support vector machines for classification and regression[R].Technical report. Southampton University of Southampton, Faculty of Engineering,Science and Mathemstics Scod of Electronics and Computer Science,1998,5.10:1-28.
[31]Vapnik V N, Chervonenkis A. On the uniform convergence of relative frequencies of events to their probabilities. Theory of Probability and its Applications[J],1991,16(2):264-280.
[32]Vapnik V N, Chervonenkis A. The necessary and sufficienct conditions for consistency in the emprirical risk minimization method[J]. Pattern Recognition and Image Analysis,1991,1(3):283-305.
[33]Vapnik著,张学工译.统计学习理论的本质.北京：清华大学出版社.2000.
[34]Li C.H..Lu Z.D.,Zhou K.SVR-Parameters Selection for Image watermarking[J].In:Proceedings of 17th IEEE International Conference on Tools with Artificial Intelligence,2005,466-470.
[35]Xiang-Yang Wang.Hong-Ying Yang,Chang-Ying Cui,An AVM-based robust digital image atermarking against desynchronizationg attacks[J],Signal Processing,88(2008)2193-2205, September 2008.
[36]Li C.H.,Lu Z.D.,Zhou K.An image watermarking technology based on support vector regression[J]. In:Proceedings of the IEEE International Symposium on Communications and Information Technologies. 2005,177-180.
[37]Li Chunhua, Ling hefei, Lu Zhengding. Adaptive Spatial Domain Image Watermarking Based on Support Vector Machine[J]. Journal of Computer Research and Development,2007,44(8):1399-1405
[38]阎辉.基于核函数的机器学习方法研究[D].清华大学.2000.
[39]阎威武.支持向量机理论、方法和应用研究[D],上海交通大学.2003.
[40]SakrN, Zhao Jiying, GrozaV. A Dynamic Fuzzy Logic Approach to Adaptive HVS-Based Watermarking[J], Proc of the IEEE International Workshop on Haptic Audio Visual Environments and Their Applications.Ottawa, Canada,2005:121-126.