基于二元BKF统计建模的双树复数小波域数字水印检测算法
|
摘要
作为多媒体作品版权保护的有效手段,数字水印技术已成为国际学术界的研究热点之一.可保持不可感知性、鲁棒性、水印容量之间良好平衡的数字水印算法研究是一项富有挑战性的工作,而基于统计模型的变换域乘性水印思想为有效解决三者之间良好平衡问题提供了可能的解决方向.该文以双树复数小波变换(Dual-Tree Complex Wavelet Transform,DT CWT)及多元BKF(Multivariate Bessel K Form,MBKF)分布理论为基础,提出了一种新的统计模型数字音频盲水印算法.原始音频进行DT CWT,并选择重要的DT CWT高频系数嵌入水印.水印检测时,首先根据DT CWT系数的尺度间相关性,利用二元BKF分布对高频子带系数进行建模;然后结合DT CWT系数的子带内相关性,估计出二元BKF统计模型参数;最后,利用最大似然决策(Maximum Likelihood decision,ML)盲提取水印信息.仿真实验结果表明,本方法在保证良好不可感知性的同时,能够有效抵抗多种攻击,具有较好的鲁棒性.
With the development of Internet technology and multimedia tools,it becomes very easy to distribute and transmit the digital media like image,audio and video.Simultaneously,the protection of intellectual property has also become more and more attentive and important for the society.Digital watermarking has been proposed and widely studied during recent years for the purposes of copyright protection and content authentication.Whilst digital watermarking can be widely applied to various digital media,this paper focuses on digital audio watermarking.Imperceptibility and robustness are two main requirements of any watermarking systems to guarantee desired functionalities,but there is a tradeoff between them from the information-theoretic perspective.Imperceptibility denotes the ability of embedding the watermarks without significantly lowering the audio quality.Robustness denotes the capability of extracting the watermarks under various attacks.Above two requirements,respectively imperceptibility and robustness,are equally important for an audio watermarking algorithm,but there is an ambivalence between them.Improving the ability of robustness and imperceptibility at the same time has been a challenge for all audio watermarking algorithms.In order to solve the tradeoff between the imperceptibility and robustness of the watermark data,the advantage is usually taken of the statistical properties of the digital audio.Every digital audio has certain features and characteristics.In statistical modeling,it is intended to capture these characteristics using a small number of parameters.In the past decade,some statistical model-based digital audio watermarking schemes have been proposed,and it has been shown that the statistical model based multiplicative watermarking schemes are more robust and provide higher imperceptibility of the watermark than other approaches.Therefore,detection and extraction of the multiplicative watermarks have received a great deal of attention.Statistical audio modeling is mainly focused on transform domains in which the energy density has a more local structure,and the performance of a statistical model-based detection of a watermark is highly influenced by the multiresolution and compression properties of transform and the accuracy of the transform coefficients modeling.There are a number of transform and distributions that have been used in watermark detection;however,there is still scope to explore further the suitability of transform and distributions to improve the performance of watermark detectors.In this paper,a new blind detector is proposed for DT CWT(Dual-Tree Complex Wavelet Transform)-based multiplicative audio watermarking,wherein a PDF based on the multivariate BKF(Bessel K form)distribution is used.In the presented scheme,watermark data is embedded into the significant high-frequency coefficients in DT CWT domain.At the watermark receiver,DT CWT highpass coefficients are firstly modeled by employing multivariate BKF distribution according to the inter-scale coefficients dependencies,then the statistical model parameters of multivariate BKF are estimated using the intra-scale coefficients dependencies,and finally a blind watermark extraction approach is developed using the maximum likelihood decision rule.Experiments are performed to verify the imperceptibility and robustness of the proposed watermark decoder.The results show that the proposed blind watermark decoder is superior to other decoders in terms of providing a higher peak signal to noise ratio and lower bit error rate.It is also shown that the proposed decoder is highly robust against various kinds of attacks such as noise addition,MP3 compression,amplitude variation,random cropping,and jittering.
With the development of Internet technology and multimedia tools,it becomes very easy to distribute and transmit the digital media like image,audio and video.Simultaneously,the protection of intellectual property has also become more and more attentive and important for the society.Digital watermarking has been proposed and widely studied during recent years for the purposes of copyright protection and content authentication.Whilst digital watermarking can be widely applied to various digital media,this paper focuses on digital audio watermarking.Imperceptibility and robustness are two main requirements of any watermarking systems to guarantee desired functionalities,but there is a tradeoff between them from the information-theoretic perspective.Imperceptibility denotes the ability of embedding the watermarks without significantly lowering the audio quality.Robustness denotes the capability of extracting the watermarks under various attacks.Above two requirements,respectively imperceptibility and robustness,are equally important for an audio watermarking algorithm,but there is an ambivalence between them.Improving the ability of robustness and imperceptibility at the same time has been a challenge for all audio watermarking algorithms.In order to solve the tradeoff between the imperceptibility and robustness of the watermark data,the advantage is usually taken of the statistical properties of the digital audio.Every digital audio has certain features and characteristics.In statistical modeling,it is intended to capture these characteristics using a small number of parameters.In the past decade,some statistical model-based digital audio watermarking schemes have been proposed,and it has been shown that the statistical model based multiplicative watermarking schemes are more robust and provide higher imperceptibility of the watermark than other approaches.Therefore,detection and extraction of the multiplicative watermarks have received a great deal of attention.Statistical audio modeling is mainly focused on transform domains in which the energy density has a more local structure,and the performance of a statistical model-based detection of a watermark is highly influenced by the multiresolution and compression properties of transform and the accuracy of the transform coefficients modeling.There are a number of transform and distributions that have been used in watermark detection;however,there is still scope to explore further the suitability of transform and distributions to improve the performance of watermark detectors.In this paper,a new blind detector is proposed for DT CWT(Dual-Tree Complex Wavelet Transform)-based multiplicative audio watermarking,wherein a PDF based on the multivariate BKF(Bessel K form)distribution is used.In the presented scheme,watermark data is embedded into the significant high-frequency coefficients in DT CWT domain.At the watermark receiver,DT CWT highpass coefficients are firstly modeled by employing multivariate BKF distribution according to the inter-scale coefficients dependencies,then the statistical model parameters of multivariate BKF are estimated using the intra-scale coefficients dependencies,and finally a blind watermark extraction approach is developed using the maximum likelihood decision rule.Experiments are performed to verify the imperceptibility and robustness of the proposed watermark decoder.The results show that the proposed blind watermark decoder is superior to other decoders in terms of providing a higher peak signal to noise ratio and lower bit error rate.It is also shown that the proposed decoder is highly robust against various kinds of attacks such as noise addition,MP3 compression,amplitude variation,random cropping,and jittering.
引文
[1]Liu X L,Lin C C,Yuan S M.Blind dual watermarking for color images’authentication and copyright protection.IEEETransactions on Circuits and Systems for Video Technology,2016,DOI:10.1109/TCSVT.2016.2633878
[2]Hua G,Huang J,Shi Y Q,et al.Twenty years of digital audio watermarking-A comprehensive review.Signal Processing,2016,128:222-242
[3]Verma V S,Jha R K.An overview of robust digital image watermarking.IETE Technical Review,2015,32(6):479-496
[4]Tang Xin,Ma Zhao-Feng,Niu Xin-Xin,Yang Yi-Xian.Blind audio watermarking mechanism based on variational Bayesian learning.Journal on Communications,2015,36(1):121-128(in Chinese)(唐鑫,马兆丰,钮心忻,杨义先.基于变分贝叶斯学习的音频水印盲检测方法.通信学报,2015,36(1):121-128)
[5]Rahman S M M,Ahmad M O,Swamy M N S.A new statistical detector for DWT-based additive image watermarking using the Gauss-Hermite expansion.IEEE Transactions on Image Processing,2009,18(8):1782-1796
[6]Singh J,Garg P,De A.Multiplicative watermarking of audio in DFT magnitude.Multimedia Tools and Applications,2014,71(3):1431-1453
[7]Wang X Y,Shi Q L,Wang S M,et al.A blind robust digital watermarking using invariant exponent moments.AEU-International Journal of Electronics and Communications,2016,70(4):416-426
[8]Ngo N M,Unoki M.Method of audio watermarking based on adaptive phase modulation.IEICE Transactions on Information and Systems,2016,99(1):92-101
[9]Dong L,Yan Q,Lv Y,et al.Full band watermarking in DCT domain with Weibull model.Multimedia Tools and Applications,2017,76(2):1983-2000
[10]Amini M,Ahmad M O,Swamy M N S.A new locally optimum watermark detection using vector-based hidden Markov model in wavelet domain.Signal Processing,2017,137:213-222
[11]Amini M,Ahmad M,Swamy M.A robust multibit multiplicative watermark decoder using vector-based hidden Markov model in wavelet Domain.IEEE Transactions on Circuits and Systems for Video Technology,2016.DOI 10.1109/TCSVT.2016.2607299
[12]Kalantari N K,Ahadi S M,Vafadust M.A robust image watermarking in the ridgelet domain using universally optimum decoder.IEEE Transactions on Circuits and Systems for Video Technology,2010,20(3):396-406
[13]Sadreazami H,Ahmad M O,Swamy M N S.Multiplicative watermark decoder in contourlet domain using the normal inverse Gaussian distribution.IEEE Transactions on Multimedia,2016,18(2):196-207
[14]Wang Xiang-Yang,Liu Yu-Nan,Xu Huan,et al.Blind optimum detector for robust image watermarking in nonsubsampled Shearlet domain.Information Sciences,2016,372:634-654
[15]Kwitt R,Meerwald P,Uhl A.Lightweight detection of additive watermarking in the DWT-domain.IEEE Transactions on Image Processing,2011,20(2):474-484
[16]Bian Y,Liang S.Image watermark detection in the wavelet domain using Bessel K densities.IET Image Processing,2013,7(4):281-289
[17]Sadreazami H,Ahmad M O,Swamy M N S.A study of multiplicative watermark detection in the contourlet domain using alpha-stable distributions.IEEE Transactions on Image Processing,2014,23(10):4348-4360
[18]Bian Y,Liang S.Locally optimal detection of image watermarks in the wavelet domain using Bessel K form distribution.IEEE Transactions on Image processing,2013,22(6):2372-2384
[19]Khawne A,Attachoo B,Hamamoto K.Optimum watermark detection of ultrasonic echo medical images.IEEJ Transactions on Electrical and Electronic Engineering,2015,10(2):149-156
[20]Bhinder P,Singh K,Jindal N.Image-adaptive watermarking using maximum likelihood decoder for medical images.Multimedia Tools&Applications,2018,77(8):10 303-10328
[21]Akhaee M A,Sahraeian S M E,Marvasti F.Contourletbased image watermarking using optimum detector in a noisy environment.IEEE Transactions on Image Processing,2010,19(4):967-980
[22]Amirmazlaghani M,Rezghi M,Amindavar H.A novel robust scaling image watermarking scheme based on Gaussian mixture model.Expert Systems with Applications,2015,42(4):1960-1971
[23]Akhaee M A,Sahraeian S M E.Scaling-based watermarking with universally optimum decoder.Multimedia Tools and Applications,2015,74(15):5995-6018
[24]Akhaee M A,Kalantari N K,Marvasti F.Robust audio and speech watermarking using Gaussian and Laplacian modeling.Signal Processing,2010,90(8):2487-2497
[25]Asikuzzaman M,Alam M J,Lambert A J,et al.Imperceptible and robust blind video watermarking using chrominance embedding:A set of approaches in the DT CWT domain.IEEE Transactions on Information Forensics and Security,2014,9(9):1502-1517
[26]Boubchir L,Nait-Ali A,Petit E.Multivariate statistical modeling of images in sparse multiscale transforms domain//Proceedings of the International Conference on Image Processing(ICIP 2010).Hong Kong,China,2010:1877-1880
[2]Hua G,Huang J,Shi Y Q,et al.Twenty years of digital audio watermarking-A comprehensive review.Signal Processing,2016,128:222-242
[3]Verma V S,Jha R K.An overview of robust digital image watermarking.IETE Technical Review,2015,32(6):479-496
[4]Tang Xin,Ma Zhao-Feng,Niu Xin-Xin,Yang Yi-Xian.Blind audio watermarking mechanism based on variational Bayesian learning.Journal on Communications,2015,36(1):121-128(in Chinese)(唐鑫,马兆丰,钮心忻,杨义先.基于变分贝叶斯学习的音频水印盲检测方法.通信学报,2015,36(1):121-128)
[5]Rahman S M M,Ahmad M O,Swamy M N S.A new statistical detector for DWT-based additive image watermarking using the Gauss-Hermite expansion.IEEE Transactions on Image Processing,2009,18(8):1782-1796
[6]Singh J,Garg P,De A.Multiplicative watermarking of audio in DFT magnitude.Multimedia Tools and Applications,2014,71(3):1431-1453
[7]Wang X Y,Shi Q L,Wang S M,et al.A blind robust digital watermarking using invariant exponent moments.AEU-International Journal of Electronics and Communications,2016,70(4):416-426
[8]Ngo N M,Unoki M.Method of audio watermarking based on adaptive phase modulation.IEICE Transactions on Information and Systems,2016,99(1):92-101
[9]Dong L,Yan Q,Lv Y,et al.Full band watermarking in DCT domain with Weibull model.Multimedia Tools and Applications,2017,76(2):1983-2000
[10]Amini M,Ahmad M O,Swamy M N S.A new locally optimum watermark detection using vector-based hidden Markov model in wavelet domain.Signal Processing,2017,137:213-222
[11]Amini M,Ahmad M,Swamy M.A robust multibit multiplicative watermark decoder using vector-based hidden Markov model in wavelet Domain.IEEE Transactions on Circuits and Systems for Video Technology,2016.DOI 10.1109/TCSVT.2016.2607299
[12]Kalantari N K,Ahadi S M,Vafadust M.A robust image watermarking in the ridgelet domain using universally optimum decoder.IEEE Transactions on Circuits and Systems for Video Technology,2010,20(3):396-406
[13]Sadreazami H,Ahmad M O,Swamy M N S.Multiplicative watermark decoder in contourlet domain using the normal inverse Gaussian distribution.IEEE Transactions on Multimedia,2016,18(2):196-207
[14]Wang Xiang-Yang,Liu Yu-Nan,Xu Huan,et al.Blind optimum detector for robust image watermarking in nonsubsampled Shearlet domain.Information Sciences,2016,372:634-654
[15]Kwitt R,Meerwald P,Uhl A.Lightweight detection of additive watermarking in the DWT-domain.IEEE Transactions on Image Processing,2011,20(2):474-484
[16]Bian Y,Liang S.Image watermark detection in the wavelet domain using Bessel K densities.IET Image Processing,2013,7(4):281-289
[17]Sadreazami H,Ahmad M O,Swamy M N S.A study of multiplicative watermark detection in the contourlet domain using alpha-stable distributions.IEEE Transactions on Image Processing,2014,23(10):4348-4360
[18]Bian Y,Liang S.Locally optimal detection of image watermarks in the wavelet domain using Bessel K form distribution.IEEE Transactions on Image processing,2013,22(6):2372-2384
[19]Khawne A,Attachoo B,Hamamoto K.Optimum watermark detection of ultrasonic echo medical images.IEEJ Transactions on Electrical and Electronic Engineering,2015,10(2):149-156
[20]Bhinder P,Singh K,Jindal N.Image-adaptive watermarking using maximum likelihood decoder for medical images.Multimedia Tools&Applications,2018,77(8):10 303-10328
[21]Akhaee M A,Sahraeian S M E,Marvasti F.Contourletbased image watermarking using optimum detector in a noisy environment.IEEE Transactions on Image Processing,2010,19(4):967-980
[22]Amirmazlaghani M,Rezghi M,Amindavar H.A novel robust scaling image watermarking scheme based on Gaussian mixture model.Expert Systems with Applications,2015,42(4):1960-1971
[23]Akhaee M A,Sahraeian S M E.Scaling-based watermarking with universally optimum decoder.Multimedia Tools and Applications,2015,74(15):5995-6018
[24]Akhaee M A,Kalantari N K,Marvasti F.Robust audio and speech watermarking using Gaussian and Laplacian modeling.Signal Processing,2010,90(8):2487-2497
[25]Asikuzzaman M,Alam M J,Lambert A J,et al.Imperceptible and robust blind video watermarking using chrominance embedding:A set of approaches in the DT CWT domain.IEEE Transactions on Information Forensics and Security,2014,9(9):1502-1517
[26]Boubchir L,Nait-Ali A,Petit E.Multivariate statistical modeling of images in sparse multiscale transforms domain//Proceedings of the International Conference on Image Processing(ICIP 2010).Hong Kong,China,2010:1877-1880