基于独立分量分析的数字水印算法研究
|
摘要
网络技术的发展使得网络传输的数字产品在没有授权的情况下容易被恶意篡改、拷贝。数字水印(Digital Watermarking)是目前国际信息安全学术界的一个前沿方向,为多媒体信息的安全和传送提供了新的方法。数字水印通过在原始数据中嵌入水印秘密信息来证实该数据的所有权,以实现版权保护。
独立分量分析(Independent Component Analysis,ICA)方法的研究起源于盲源分离技术,盲源分离是在不知道源信号和传输信道参数的情况下,根据输入原信号的统计特性,仅通过接收到的混叠信号恢复出各个源信号的过程。独立分量分析已经被广泛应用于实际数据的处理中,诸如图像处理、语音信号处理、生物医学信号处理、模式识别、数据挖掘和图像特征提取等领域。
本文从三个不同的角度来研究基于ICA的数字水印技术,其目的是在保证不可见性的前提下,提高数字水印的嵌入容量、隐蔽性及鲁棒性。主要完成了以下工作：
(1)研究了JADE(Joint Approximate Diagonalization of Eigenmatrices)盲信号分离算法失效的问题,从理论上找出算法失效的原因,获得了源信号相关性与算法失效的关系：源信号相关性越强,JADE盲信号分离算法失效问题越严重。并通过仿真实验证明了理论分析结果的正确性。
(2)针对常用的音频水印算法存在的鲁棒性较差、水印嵌入容量较少的问题,提出了一种基于独立分量分析的音频数字水印方法。使用该方法时,对原始音频文件进行离散余弦变换,在变换得到的音频文件中利用随机混合嵌入压缩编码的数字水印图像,再进行离散余弦逆变换；提取时,对音频文件进行离散余弦变换,再用独立分量分析方法进行水印检测。实验结果表明该方法具有较大的数字水印嵌入容量、强鲁棒性和可行性。
(3)针对数据库水印提出一种基于独立分量分析的公开数据库水印方法。给出了水印嵌入的混合概念,首先对水印图像经过独立分量分析(ICA)后得到若干个独立分量水印,利用迭代混合的方法嵌入水印,再利用盲源JADE算法提取水印。这种数据库水印方法不需原关系数据库,理论分析和仿真结果验证了该方法的有效性和可行性,显著提高了水印的隐蔽性及嵌入容量。
With the development of network, digital products are usually maliciously tampered or copied easily in the absence of authorization.At present, the technology of digital watermark keeps ahead in the field of international information security. It has been providing many new methods to transmit multimedia information. In order to prove the ownership of the data and protect the copyright, some secret information with watermark was embedded in the original data.
ICA (Independent Component Analysis, ICA) method originated from the blind source separation technology. Under the condiction of the unknown source signal and the transmission channel parameters, the blind source separation technology is such a technology that resumes each source signal only through analyzing the aliasing of signal, according to the the statistical properties of the input original signal.ICA has been widely used in the domain of data processing, such as image processing, speech signal processing, biomedical signal processing, pattern recognition, data mining, feature extraction of image, and so on.
In order to improve the embedding capacity, robustness and hiding of digital watermark, the technology of digital watermark is studied from three different perspectives based on ICA under the condiction of guarteeing the invisibility. Main work includes as follows:
(1)The failure problem of the JADE (Joint Approximate Diagonalization of Eigenmatrices) algorithm for blind signal separation is studied and the reason for the failure in theory is discussed.The relationship between the source signal's relativity and the algorithms's invalidation is summerized.The stronger the relativity of the source signal is, the more invalid the JADE algorithm for blind signal's separation is.The simulation results showed the correctness of the theory analysis.
(2) Aimming at the problem of the less robustness and the poor capacity of embedding watermark of the commonly used audio watermark algorithm,the audio digital watermark method based on independent components analysis is proposed.The audio file will be transformed by discrete cosine mean. After this transformation, some digital watermark images, which are coded mixedly and randomly, will be embeded in the audio file.Then, it is transformed by inverse discrete cosine.Finally, the independent component analysis is adopted to detect the watermark. The experimental results show that the proposed method has a greater embedding capacity of digital watermark, strong robustness and feasibility.
(3)A public database watermark method is proposed for database watermark based on the independent components analysis method. A mixed concept of embedding watermark is given.Firstly, a number of independent component watermark are obtained after the watermark image is handled by independent component analysis, and then an iterative mixture method is adopted to embed watermark. Finally, the blind source JADE algorithm is used to extract the watermark. This database watermark mothod do not need the original relational database. Theoretical analysis and simulation results showed that the proposed mothod is effective and feasible, and it has improved the invisibility and the capacity of database watermark greatly.
独立分量分析(Independent Component Analysis,ICA)方法的研究起源于盲源分离技术,盲源分离是在不知道源信号和传输信道参数的情况下,根据输入原信号的统计特性,仅通过接收到的混叠信号恢复出各个源信号的过程。独立分量分析已经被广泛应用于实际数据的处理中,诸如图像处理、语音信号处理、生物医学信号处理、模式识别、数据挖掘和图像特征提取等领域。
本文从三个不同的角度来研究基于ICA的数字水印技术,其目的是在保证不可见性的前提下,提高数字水印的嵌入容量、隐蔽性及鲁棒性。主要完成了以下工作：
(1)研究了JADE(Joint Approximate Diagonalization of Eigenmatrices)盲信号分离算法失效的问题,从理论上找出算法失效的原因,获得了源信号相关性与算法失效的关系：源信号相关性越强,JADE盲信号分离算法失效问题越严重。并通过仿真实验证明了理论分析结果的正确性。
(2)针对常用的音频水印算法存在的鲁棒性较差、水印嵌入容量较少的问题,提出了一种基于独立分量分析的音频数字水印方法。使用该方法时,对原始音频文件进行离散余弦变换,在变换得到的音频文件中利用随机混合嵌入压缩编码的数字水印图像,再进行离散余弦逆变换；提取时,对音频文件进行离散余弦变换,再用独立分量分析方法进行水印检测。实验结果表明该方法具有较大的数字水印嵌入容量、强鲁棒性和可行性。
(3)针对数据库水印提出一种基于独立分量分析的公开数据库水印方法。给出了水印嵌入的混合概念,首先对水印图像经过独立分量分析(ICA)后得到若干个独立分量水印,利用迭代混合的方法嵌入水印,再利用盲源JADE算法提取水印。这种数据库水印方法不需原关系数据库,理论分析和仿真结果验证了该方法的有效性和可行性,显著提高了水印的隐蔽性及嵌入容量。
With the development of network, digital products are usually maliciously tampered or copied easily in the absence of authorization.At present, the technology of digital watermark keeps ahead in the field of international information security. It has been providing many new methods to transmit multimedia information. In order to prove the ownership of the data and protect the copyright, some secret information with watermark was embedded in the original data.
ICA (Independent Component Analysis, ICA) method originated from the blind source separation technology. Under the condiction of the unknown source signal and the transmission channel parameters, the blind source separation technology is such a technology that resumes each source signal only through analyzing the aliasing of signal, according to the the statistical properties of the input original signal.ICA has been widely used in the domain of data processing, such as image processing, speech signal processing, biomedical signal processing, pattern recognition, data mining, feature extraction of image, and so on.
In order to improve the embedding capacity, robustness and hiding of digital watermark, the technology of digital watermark is studied from three different perspectives based on ICA under the condiction of guarteeing the invisibility. Main work includes as follows:
(1)The failure problem of the JADE (Joint Approximate Diagonalization of Eigenmatrices) algorithm for blind signal separation is studied and the reason for the failure in theory is discussed.The relationship between the source signal's relativity and the algorithms's invalidation is summerized.The stronger the relativity of the source signal is, the more invalid the JADE algorithm for blind signal's separation is.The simulation results showed the correctness of the theory analysis.
(2) Aimming at the problem of the less robustness and the poor capacity of embedding watermark of the commonly used audio watermark algorithm,the audio digital watermark method based on independent components analysis is proposed.The audio file will be transformed by discrete cosine mean. After this transformation, some digital watermark images, which are coded mixedly and randomly, will be embeded in the audio file.Then, it is transformed by inverse discrete cosine.Finally, the independent component analysis is adopted to detect the watermark. The experimental results show that the proposed method has a greater embedding capacity of digital watermark, strong robustness and feasibility.
(3)A public database watermark method is proposed for database watermark based on the independent components analysis method. A mixed concept of embedding watermark is given.Firstly, a number of independent component watermark are obtained after the watermark image is handled by independent component analysis, and then an iterative mixture method is adopted to embed watermark. Finally, the blind source JADE algorithm is used to extract the watermark. This database watermark mothod do not need the original relational database. Theoretical analysis and simulation results showed that the proposed mothod is effective and feasible, and it has improved the invisibility and the capacity of database watermark greatly.
引文
[1]Voyatzis G,Nikolaidis N, Pitas I.Digital watermarking:An overview[J]. Eur.Signal Processing Conf. EUSIPCO, Rhodes, Greece.1998,9:9-12
[2]Barnett R. Digital watermarking:Applications, techniques, and challenges[J]. Electronics and Communication Engineering Journal.1999,11(4):173-183
[3]尹浩,林闯,邱锋,等.数字水印技术综述[J].计算机研究与发展,2005,42(7)：1093-1099
[4]Su J.K, Hartung F, Girod B.Digital watermarking of text, image, and video documents [J].Computers Graphics.1998,22(6):687-695
[5]Jean Z, Koch E. A generic digital watermarking model[J]. Computers&Graphics.1998,22(4):397-403
[6]Gross-Ambler D. Query-preserving watermarking of relational data-bases and XML documents[C].Proceedings of PODS 2003,San Diego, US,2003: 191-201
[7]Petitcolas F A P, Anderson R J,Kuhn M G.Information hiding-a survey[J], Proc.of the IEEE Special Issue on Protection of Multimedia Content,1999, 87(7):1062-1078
[8]Johnson N F, Jadodia S, Exploring steganography:seeing the unseen[J],IEEE Computer,1998,31(2):26-34
[9]Cox J, Miller M L, Bloom J A. Digital Watermarking[M].San Francisco, CA: Morgan Kaufmann,2001
[10]Eggers J J, Girod B.Informed Watermarking[M].MA:Kluwer Academic Publishers,2002
[11]王志雄,王慧琴,李人厚.数字水印应用中的攻击和对策综述[J].通信学报,2002,23(11)：74-79
[12]刘据,聂开宝,何振亚.非线性混叠信号的可分离性及分离方法研究[J].电子与信息学报,2003,25(1)：54-61
[13]Yu D, Sitar F, et al.Watermark detection and extraction using independent component analysis method[J].EURASIP Journal on Applied Signal Processing,2002,1:92-103
[14]Sheen M, Huang J, Beadle P J.Application of ICA to the digital image watermarking[J].IEEE International Conferences on Neural Networks and Signal Processing, Naming, China,2003,1485-1488
[15]Gonzalez-Serrano F J, Molina-Bulla H Y, et al. Independent component analysis applied to digital image watermarking[C].Proceedings of ICASSP'2001,Salt Lake City, Utah,2001,1997-2000
[16]Mebane, Bartolini F, Watermark Systems Engineering[M].New York:Marcel Decker,2004
[17]Wang H M, Su P C,et al. Wavelet-based digital image watermarking[J]. OPTICS EXPRESS,1998,3(12):491-496
[18]Detangle J F, et al.Watermarking algorithm based on a human visual model[J]. Signal Processing[J],1998,66(3):319-335
[19]Stefan Katzenbeisser, Fabian A P,吴秋新等泽.Information Hiding Techniques for Steganography and Digital Watermarking[M],北京：人民邮电出版社,2001
[20]汪小帆,戴跃伟,茅耀斌.信息隐藏技术：方法和应用[M],北京：机械出版社,2001
[21]Fabian A, Colas P, Anderson R J, Markus G K. Information Hiding-Survey[J], Proceedings of the IEEE, Special issue on protection of multimedia content, July 1999,87(7):1062-1078
[22]Ingmar J C, Kalian J, Leighton T,et al. Secure Spread Spectrum Watermarking for Multimedia[J].IEEE Transaction on Image Processing,1997,12(6): 1673-1687
[23]王炳锡主编.数字水印技术[M],西安：西安电子科技大学出版社,2003
[24]Thomas M. An Information-Theoretic Approach to Steganography and Watermarking[J].Lecture Notes in Computer Science,1999,1768:1-16
[25]傅朝阳,翟玉庆.一种用于数字权限管理安全审计的日志描述模型[J].计算机应用研究.2006(1)：125-128
[26]Cox L J, Killian J, Leighton Information Hiding, Lecture T. and Shimon T. A Secure, Robust Watermark for Multimedia[J],Notes in Computer Science, 1996,1174:183-206
[27]杜鹏超,唐通林.数字水印研究中常用的测试指标(上)[J].测试技术与应用.2002(11)：7-10.
[28]崔晓瑜,数字水印技术的研究[D],[博士学位论文],北京：北京大学,2001.
[29]Gonzalez S, Molina Y, Murillo B,et al.Independent Component Analysis Applied to Digital Image Watermarking[C].Proceeding of the IEEE International Conference on Acoustics, Speech, and Signal Processing,2001, 3(3):1997-2000.
[30]Zhou W, Jiao L. An Iterative Algorithm of Independent Component Analysis and the Experiment Results[J].Acta Biophysical Sinica.2002,18(1):57-60.
[31]Hyvarinen A, Erkki O. Independent component analysis:algorithms and applications[C], Finland Neural Networks,2000,13(4-5):411-430
[32]盛骤,谢式千,潘承毅.概率论与数理统计[M],北京：高等教育出版社,1989.
[33]Zhang X X, Qiu T S.The study on the principle of Kurtosisbased ICA method[J].Neural Networks and Signal Processing,2003,(2):1346-1349.
[34]Hyaline A. Fast and Robust Fixed-Point Algorithms for Independent Component Analysis[C].IEEE Trans on Neural Networks,1999,10(3): 626-634.
[35]Heeyoul C, Seungiin C.Relative Trust-region learning for ICA[C]. Proceedings of Acoustics, Sppech, and Signal Processing,2005,5:261-264
[36]郑春厚.独立分量分析算法及其应用研究[D]：[博士学位论文],合肥：中国科学技术大学,2006.
[37]Lin C Y, Chang S F. Watermarking capacity of digital images based domain-specific masking effects[C].Proceedings of IEEE International Conference Information Technology, Coding and Computing,2001:90-94.
[38]张勇.水印关系数据库研究[D]：[博士学位论文],南京：中国人民解放军理工大学,2004
[39]武荣,曹加恒,黄敏,等.关系数据库的数字水印新技术[J].武汉大学学报(理学版),2005,51(5)：589-593
[40]孙圣和,陆哲明,牛夏牧,等.数字水印技术及应用[M].北京：科学出版社,2004
[41]Rakish A J, Jerry K. Watermarking Relational Databases[C].Proceeding of the 28th VLDB Conference, Hongkong, China,2002:155-156
[42]Holliman M,Memo N. Counterfeiting attacks for block-wise independent watermarking techniques[J],IEEE Trans.on Image Processing,2000,9(3): 432-441
[43]Jutted C,Herald C,Blind Separation of Sources. Part I:An Adaptive Algorithm Based on Necromantic architecture[J].Signal Processing,1994, (24):1-10
[44]Common P. Independent Component Analysis, A New concept[J].Signal Processing,1994, (36):287-314
[45]Armani S, Cardoon J F, Blind Source Separation Semipro metric Statistical Approach[J].IEEE Trans. On Signal Processing,1997,45(11):2692-2700
[46]Lee T, Georama M, Ssejnowski.T J. Independent Component Analysis Using an Extended Informix Algorithm for Mixed Subgaussian-supergaussian Sources[J],Neural Comp,1999,(11):417-441
[47]Cardoon J F, Leafed B.Equivariant Adaptive Source Separation[J],IEEE Trans, Signal Processing,1996, (44):3017-3029
[48]Cardoon J F, soulful A. Blind beam beam forming for no Gaussian signals, IEEE Proceedings-F,1993,140(6):362-370
[49]Zhao J, Koche, Luo C.In business today and tomorrow. Communications of the ACM,1998,41(7):62-67
[50]Swanson M D, Zhu B, Tewfik AH. Current state of the art, challenges and future directions for audio watermarking[C].Proceeding of International Conference on Multimedia Computing and Systems,1999,(1):19-24
[51]Kutter M,Winkler S.A Vision-based Masking Model for Spread-spectrum Image Watermarking[J]. IEEE Transactions on Image Processing,2002,11(1): 16-25
[52]Bassia P, Pitas I,Nikolaidis N. Robust Audio Watermarking in the time Do-main[J].,IEEE Transacation on Multimedia,2001,3(2)232-241
[53]Ikeda M, Takeda K, Itakura F. Audio Data Hiding by use of Band-limited Random Sequences[J].ISASSP.1999,4:2315-2318
[54]Moriya T, Takashima Y, Nakamura T, et al.Digital Watermarking Schemes Based on Vector Quantization[J].IEEE Workshop on Speech Coding for Telecommunication.1997:95-96
[55]Chung P W. Techniques for Audiovisual data Confidentiality[J],Integrity and Copyright Protection. Doctoral Thesis.2001:130-142
[56]Chook K P, Lee K Y. An Efficient Audio Watermarking by Using Spectrum Warping[J].IEICE Trans on Fundamentals.2002,85(6):1257-1264
[57]Li H, Sun Y L. The Study and Test of ICA Algorithms[C].Wireless Communication, Networking and Mobile Computing,2005,1:602-605
[58]Hyaline A. Fast and Robust Fixed-Point Algorithms for Independent Component Analysis[J].IEEE Trans on Neural Networks,1999,10(3): 626-634
[59]Cuixiang F. Extract all Real-Roots of Polynomial Equation with Newton Method and Define Initial Value of Iteration[J].HLJ.Hydr.Eng,1994,2:53-55
[60]Jun A Y,Zhen Q Z.An Improved Fast ICA Algorithm Based on Negentropy Maximization[J].Journal of Circuits and Systems.2002,7,4:37-40
[61]Arrayal K J.Watermarking relational databases[C].Proceeding of the 28thVLDB Conference, Hong Kong, China,2002:155-166
[62]Sion R, Atallah M, Sunil P.Watermarking relational databases[J].Indiana:the Center for Education and Research in Information Assurance and Security of Purdue University,2002
[63]Pham D T, Cardoso J.Blind separation of instantaneous mixtures of no stationary sources[J].IEEE Trans on Signal Processing,2001,49(9): 1837-1848
[64]牛夏牧,赵亮,黄文军,等.利用数字水印技术实现数据库的版权保护[J].电子学报,2003,31(12)：2050-2053
[2]Barnett R. Digital watermarking:Applications, techniques, and challenges[J]. Electronics and Communication Engineering Journal.1999,11(4):173-183
[3]尹浩,林闯,邱锋,等.数字水印技术综述[J].计算机研究与发展,2005,42(7)：1093-1099
[4]Su J.K, Hartung F, Girod B.Digital watermarking of text, image, and video documents [J].Computers Graphics.1998,22(6):687-695
[5]Jean Z, Koch E. A generic digital watermarking model[J]. Computers&Graphics.1998,22(4):397-403
[6]Gross-Ambler D. Query-preserving watermarking of relational data-bases and XML documents[C].Proceedings of PODS 2003,San Diego, US,2003: 191-201
[7]Petitcolas F A P, Anderson R J,Kuhn M G.Information hiding-a survey[J], Proc.of the IEEE Special Issue on Protection of Multimedia Content,1999, 87(7):1062-1078
[8]Johnson N F, Jadodia S, Exploring steganography:seeing the unseen[J],IEEE Computer,1998,31(2):26-34
[9]Cox J, Miller M L, Bloom J A. Digital Watermarking[M].San Francisco, CA: Morgan Kaufmann,2001
[10]Eggers J J, Girod B.Informed Watermarking[M].MA:Kluwer Academic Publishers,2002
[11]王志雄,王慧琴,李人厚.数字水印应用中的攻击和对策综述[J].通信学报,2002,23(11)：74-79
[12]刘据,聂开宝,何振亚.非线性混叠信号的可分离性及分离方法研究[J].电子与信息学报,2003,25(1)：54-61
[13]Yu D, Sitar F, et al.Watermark detection and extraction using independent component analysis method[J].EURASIP Journal on Applied Signal Processing,2002,1:92-103
[14]Sheen M, Huang J, Beadle P J.Application of ICA to the digital image watermarking[J].IEEE International Conferences on Neural Networks and Signal Processing, Naming, China,2003,1485-1488
[15]Gonzalez-Serrano F J, Molina-Bulla H Y, et al. Independent component analysis applied to digital image watermarking[C].Proceedings of ICASSP'2001,Salt Lake City, Utah,2001,1997-2000
[16]Mebane, Bartolini F, Watermark Systems Engineering[M].New York:Marcel Decker,2004
[17]Wang H M, Su P C,et al. Wavelet-based digital image watermarking[J]. OPTICS EXPRESS,1998,3(12):491-496
[18]Detangle J F, et al.Watermarking algorithm based on a human visual model[J]. Signal Processing[J],1998,66(3):319-335
[19]Stefan Katzenbeisser, Fabian A P,吴秋新等泽.Information Hiding Techniques for Steganography and Digital Watermarking[M],北京：人民邮电出版社,2001
[20]汪小帆,戴跃伟,茅耀斌.信息隐藏技术：方法和应用[M],北京：机械出版社,2001
[21]Fabian A, Colas P, Anderson R J, Markus G K. Information Hiding-Survey[J], Proceedings of the IEEE, Special issue on protection of multimedia content, July 1999,87(7):1062-1078
[22]Ingmar J C, Kalian J, Leighton T,et al. Secure Spread Spectrum Watermarking for Multimedia[J].IEEE Transaction on Image Processing,1997,12(6): 1673-1687
[23]王炳锡主编.数字水印技术[M],西安：西安电子科技大学出版社,2003
[24]Thomas M. An Information-Theoretic Approach to Steganography and Watermarking[J].Lecture Notes in Computer Science,1999,1768:1-16
[25]傅朝阳,翟玉庆.一种用于数字权限管理安全审计的日志描述模型[J].计算机应用研究.2006(1)：125-128
[26]Cox L J, Killian J, Leighton Information Hiding, Lecture T. and Shimon T. A Secure, Robust Watermark for Multimedia[J],Notes in Computer Science, 1996,1174:183-206
[27]杜鹏超,唐通林.数字水印研究中常用的测试指标(上)[J].测试技术与应用.2002(11)：7-10.
[28]崔晓瑜,数字水印技术的研究[D],[博士学位论文],北京：北京大学,2001.
[29]Gonzalez S, Molina Y, Murillo B,et al.Independent Component Analysis Applied to Digital Image Watermarking[C].Proceeding of the IEEE International Conference on Acoustics, Speech, and Signal Processing,2001, 3(3):1997-2000.
[30]Zhou W, Jiao L. An Iterative Algorithm of Independent Component Analysis and the Experiment Results[J].Acta Biophysical Sinica.2002,18(1):57-60.
[31]Hyvarinen A, Erkki O. Independent component analysis:algorithms and applications[C], Finland Neural Networks,2000,13(4-5):411-430
[32]盛骤,谢式千,潘承毅.概率论与数理统计[M],北京：高等教育出版社,1989.
[33]Zhang X X, Qiu T S.The study on the principle of Kurtosisbased ICA method[J].Neural Networks and Signal Processing,2003,(2):1346-1349.
[34]Hyaline A. Fast and Robust Fixed-Point Algorithms for Independent Component Analysis[C].IEEE Trans on Neural Networks,1999,10(3): 626-634.
[35]Heeyoul C, Seungiin C.Relative Trust-region learning for ICA[C]. Proceedings of Acoustics, Sppech, and Signal Processing,2005,5:261-264
[36]郑春厚.独立分量分析算法及其应用研究[D]：[博士学位论文],合肥：中国科学技术大学,2006.
[37]Lin C Y, Chang S F. Watermarking capacity of digital images based domain-specific masking effects[C].Proceedings of IEEE International Conference Information Technology, Coding and Computing,2001:90-94.
[38]张勇.水印关系数据库研究[D]：[博士学位论文],南京：中国人民解放军理工大学,2004
[39]武荣,曹加恒,黄敏,等.关系数据库的数字水印新技术[J].武汉大学学报(理学版),2005,51(5)：589-593
[40]孙圣和,陆哲明,牛夏牧,等.数字水印技术及应用[M].北京：科学出版社,2004
[41]Rakish A J, Jerry K. Watermarking Relational Databases[C].Proceeding of the 28th VLDB Conference, Hongkong, China,2002:155-156
[42]Holliman M,Memo N. Counterfeiting attacks for block-wise independent watermarking techniques[J],IEEE Trans.on Image Processing,2000,9(3): 432-441
[43]Jutted C,Herald C,Blind Separation of Sources. Part I:An Adaptive Algorithm Based on Necromantic architecture[J].Signal Processing,1994, (24):1-10
[44]Common P. Independent Component Analysis, A New concept[J].Signal Processing,1994, (36):287-314
[45]Armani S, Cardoon J F, Blind Source Separation Semipro metric Statistical Approach[J].IEEE Trans. On Signal Processing,1997,45(11):2692-2700
[46]Lee T, Georama M, Ssejnowski.T J. Independent Component Analysis Using an Extended Informix Algorithm for Mixed Subgaussian-supergaussian Sources[J],Neural Comp,1999,(11):417-441
[47]Cardoon J F, Leafed B.Equivariant Adaptive Source Separation[J],IEEE Trans, Signal Processing,1996, (44):3017-3029
[48]Cardoon J F, soulful A. Blind beam beam forming for no Gaussian signals, IEEE Proceedings-F,1993,140(6):362-370
[49]Zhao J, Koche, Luo C.In business today and tomorrow. Communications of the ACM,1998,41(7):62-67
[50]Swanson M D, Zhu B, Tewfik AH. Current state of the art, challenges and future directions for audio watermarking[C].Proceeding of International Conference on Multimedia Computing and Systems,1999,(1):19-24
[51]Kutter M,Winkler S.A Vision-based Masking Model for Spread-spectrum Image Watermarking[J]. IEEE Transactions on Image Processing,2002,11(1): 16-25
[52]Bassia P, Pitas I,Nikolaidis N. Robust Audio Watermarking in the time Do-main[J].,IEEE Transacation on Multimedia,2001,3(2)232-241
[53]Ikeda M, Takeda K, Itakura F. Audio Data Hiding by use of Band-limited Random Sequences[J].ISASSP.1999,4:2315-2318
[54]Moriya T, Takashima Y, Nakamura T, et al.Digital Watermarking Schemes Based on Vector Quantization[J].IEEE Workshop on Speech Coding for Telecommunication.1997:95-96
[55]Chung P W. Techniques for Audiovisual data Confidentiality[J],Integrity and Copyright Protection. Doctoral Thesis.2001:130-142
[56]Chook K P, Lee K Y. An Efficient Audio Watermarking by Using Spectrum Warping[J].IEICE Trans on Fundamentals.2002,85(6):1257-1264
[57]Li H, Sun Y L. The Study and Test of ICA Algorithms[C].Wireless Communication, Networking and Mobile Computing,2005,1:602-605
[58]Hyaline A. Fast and Robust Fixed-Point Algorithms for Independent Component Analysis[J].IEEE Trans on Neural Networks,1999,10(3): 626-634
[59]Cuixiang F. Extract all Real-Roots of Polynomial Equation with Newton Method and Define Initial Value of Iteration[J].HLJ.Hydr.Eng,1994,2:53-55
[60]Jun A Y,Zhen Q Z.An Improved Fast ICA Algorithm Based on Negentropy Maximization[J].Journal of Circuits and Systems.2002,7,4:37-40
[61]Arrayal K J.Watermarking relational databases[C].Proceeding of the 28thVLDB Conference, Hong Kong, China,2002:155-166
[62]Sion R, Atallah M, Sunil P.Watermarking relational databases[J].Indiana:the Center for Education and Research in Information Assurance and Security of Purdue University,2002
[63]Pham D T, Cardoso J.Blind separation of instantaneous mixtures of no stationary sources[J].IEEE Trans on Signal Processing,2001,49(9): 1837-1848
[64]牛夏牧,赵亮,黄文军,等.利用数字水印技术实现数据库的版权保护[J].电子学报,2003,31(12)：2050-2053