基于统计特征的数字图像重采样检测方法研究
|
摘要
先进数码设备和图像编辑工具的普及给人们是日常生活带来了极大的便利,同时,使得对原始生成的图像进行随意修改变得愈发容易。被恶意篡改的图像也被应用于新闻、司法、以及学术研究等领域,这严重影响到正常的社会秩序和安定,给人们的生活和工作带来了令人担忧的负面影响。因此数字图像取证技术为了判断图像的真实性逐渐受到人们的重视。本文运用图像内在统计特征的变化,针对数字图像重采样的被动取证展开研究,阐述重采样操作的基本原理,并利用二阶导数周期性的特性进行图像的重采样检测,本文主要工作如下:
本文首先研究现有图像取证技术的理论背景、研究意义及国内外研究现状,分析了数字图像被动取证技术的原理、方法和特点,讨论了图像被动取证技术的常用方法。通过对一维信号和图像重采样操作原理的分析,得出了进行重采样操作后图像像素间的相关性。讨论了重采样常用的两种基本方法,并确定本文的重采样检测思路。首先,通过对插值信号周期性的分析得到图像的频谱特征图并估测其缩放比例,从而检测图像是否经过拼接并确定具体的篡改区域。
针对现存的重采样算法主要适用于无损压缩图像检测的现状,提出一种算法克服了有损压缩与重采样操作在图像像素间产生的相似统计特性难以被检测的局限性,并保证了较高的准确性。
最后,总结本文中的工作,并展望数字图像被动取证技术的研究和发展。
The advent of digital equipment and sophisticated image editing software has not only brought large convenience,but also make some digital images being modified more and more easy.Those images which have been revised and misrepresented evilly may appear in various areas, such as news,judicial impartiality and scientific research. This kind of phenomenon can destroy the normal social order for some countries or governments,even can bring the worrisome adverse effects to our life and work. Therefore, the digital images evidence-collecting technology to obtain evidence for judging their originality, becomes a hot research field.
In this paper, according to research on resampling detection of digital image passive forensics. Introducing the basic principles of resampling algorithms and statistical properties for resampling images, the algorithm mainly use the second derivative periodicity of interpolation signal to detect resampling images. Analyzed the following five aspects of image detection:
Firstly, based on analyzing theoretical background, the significance, research status at home and abroad,discussing the principle and feature of digital image passive forensics. Some general methods were given. Periodic correlation of re-sampled pixels is employed to analyze the principle of signal resampling and image resampling.In order to analyze the best algorithm in this paper, two commonly methods of resampling were discussed. Moreover, according to analyze the periodicity of interpolating signal and then estimate scaling ratio of the image. Comparing the spectrum characteristics of each block after partitioning an image into small overlapping blocks for judging whether the image was spliced,then, tampered regions can be identified.
Due to the current resampling detection algorithms mainly applied to lossless compression, an algorithm which overcomes a detection limititation of similar statistical properties between pixels in lossy compression and resampling operation is proposed in this paper.Then it is capable to detect resampling efficiently in JPEG compressed images.
At the end of this paper, we summarize whole of the thesis and the prospective developing of digital image passive forensics in the future.
本文首先研究现有图像取证技术的理论背景、研究意义及国内外研究现状,分析了数字图像被动取证技术的原理、方法和特点,讨论了图像被动取证技术的常用方法。通过对一维信号和图像重采样操作原理的分析,得出了进行重采样操作后图像像素间的相关性。讨论了重采样常用的两种基本方法,并确定本文的重采样检测思路。首先,通过对插值信号周期性的分析得到图像的频谱特征图并估测其缩放比例,从而检测图像是否经过拼接并确定具体的篡改区域。
针对现存的重采样算法主要适用于无损压缩图像检测的现状,提出一种算法克服了有损压缩与重采样操作在图像像素间产生的相似统计特性难以被检测的局限性,并保证了较高的准确性。
最后,总结本文中的工作,并展望数字图像被动取证技术的研究和发展。
The advent of digital equipment and sophisticated image editing software has not only brought large convenience,but also make some digital images being modified more and more easy.Those images which have been revised and misrepresented evilly may appear in various areas, such as news,judicial impartiality and scientific research. This kind of phenomenon can destroy the normal social order for some countries or governments,even can bring the worrisome adverse effects to our life and work. Therefore, the digital images evidence-collecting technology to obtain evidence for judging their originality, becomes a hot research field.
In this paper, according to research on resampling detection of digital image passive forensics. Introducing the basic principles of resampling algorithms and statistical properties for resampling images, the algorithm mainly use the second derivative periodicity of interpolation signal to detect resampling images. Analyzed the following five aspects of image detection:
Firstly, based on analyzing theoretical background, the significance, research status at home and abroad,discussing the principle and feature of digital image passive forensics. Some general methods were given. Periodic correlation of re-sampled pixels is employed to analyze the principle of signal resampling and image resampling.In order to analyze the best algorithm in this paper, two commonly methods of resampling were discussed. Moreover, according to analyze the periodicity of interpolating signal and then estimate scaling ratio of the image. Comparing the spectrum characteristics of each block after partitioning an image into small overlapping blocks for judging whether the image was spliced,then, tampered regions can be identified.
Due to the current resampling detection algorithms mainly applied to lossless compression, an algorithm which overcomes a detection limititation of similar statistical properties between pixels in lossy compression and resampling operation is proposed in this paper.Then it is capable to detect resampling efficiently in JPEG compressed images.
At the end of this paper, we summarize whole of the thesis and the prospective developing of digital image passive forensics in the future.
引文
[1] Matthias Kirchner.Fast and reliable resampling detection by spectral analysis of fixed linear predictor residue.In:Proceedings of the 10th ACM workshop on multimedia and security. New York, NY, USA ,2008,11-20.
[2] A. C. Gallagher.Detection of Linear and Cubic Interpolation in JPEG compressed images.In:Proceedings of the 2nd Canadian Conference on Computer and Robot Vision (CRV’05),Washington,DC,USA,2005,65-72.
[3]魏瑗珍.低阶统计量改变的自然图像盲取证技术研究:[硕士学位论文].北京:北京交通大学,2009.
[4]周琳娜.数字图像盲取证技术研究:[博士学位论文].北京:北京邮电大学,2007.
[5]吴首阳.JPEG图像盲取证技术研究:[硕士学位论文].安徽:中国科学技术大学,2009.
[6] H. Farid. Exposing Digital Forgeries from JPEG Ghosts, IEEE Transactions on Information Forensics and Security, Mar,2009, 4(1):154-160.
[7]王波,孙璐璐,孔祥维等.图像伪造中模糊操作的异常色调率取证技术.电子学报,2006, 34(12A):2443-2446.
[8]周琳娜,王东明,郭云彪等.基于数字图像边缘特性的形态学滤波取证技术.电子学报,2008,36(6):1047-1051.
[9]朱秀明,宣国荣,姚秋明等.信息取证中图像重采样检测.计算机应用,2006, 26(11):2596-2597.
[10] C. Cruz, R. Reyes, M. Nakano. Image Content Authentication System Based on Semi-Fragile Watermarking.In:51st Midwest Symposium on Circuits and Systems, Knoxville, USA, 2008,306-309.
[11] D. Bashiri , A. Aghagolzadeh, J. Museviniya. A Novel Still Image Error Concealment Using Fragile Watermarking in Wireless Image Transmission and Packet-Switched Networks.In:International Symposium on Telecommunications, Tehran, Iran, 2008, 792-797.
[12] N. Chen, J. Zhu. A Multipurpose Audio Watermarking Scheme for Copyright Protectionand Content Authentication.In:IEEE International Conference on Multimedia and Expo, Hannover, Germany, 2008, 211- 224.
[13] X. Liu, X. Lv, Y. Wang. A semi-fragile digital watermarking algorithm based on integer wavelet matrix norm quantization for medical images.In:2nd International Conference on Bioinformatics and Biomedical Engineering, China, 2008,776-779.
[14] L. Chen, R. Zhang. Novel image authentication algorithm based on semi-fragile watermarking.Journal of Information and Computational Science, 2008, 5(3): 1069-1076.
[15] Tian-Tsong Ng,Shih-Fu Chang,Ching-Yung Lin,et a1.Passive-blind Image Forensics. In:Multimedia Security technologies for digital rights.Elsevier.2006, 18-21.
[16] A.Srivastava,A.Lee,E.Simoncelli, S.Zhu.On advances in statistical modeling of natural images.Journal of Mathematical Imaging and Vision,2003,18(1):17-33.
[17]吴琼.面向真实性检测的数字图像盲取证方法研究:[博士学位论文].湖南:国防科技大学,2008.
[18] Lanh T V,Chong K S,Emmanuel S,et a1.A survey on Digital Camera image forensic methods.In:Proceedings of IEEE International Conference on Multimedia and Expo (ICME07),Beijing,China,2007,16-19.
[19] Sencar H T,Memon N.Overview of state-of-the-art in digital image forensics.In:Part of Indian Statistical Institute Platinum Jubilee Monograph series titled Statistical Science and Interdisciplinary Research,World Scientific Press,2008,67-75.
[20] Swaminathan A,Wu M,Ray Liu KJ.Non-intrusive forensics analysis of visual sensors using output images.IEEE Transactions of Information Forensics and Security, 2007,2(1):91-106.
[21] Celiktutan O,Avcibas I,Sankur B,et a1.Source cell-phone identification.In:Proceedings of IEEE 14th Signal Processing and Communications Applications Conference,Antalya, 2006,l-3.
[22] Tsai MJ,Wu GH.Using image features to identify camera sources.In:Proceedings of IEEE International Coference on Acoustics,Speech and Signal Processing (ICASSP06), Toulouse,2006,2:297-300.
[23] Bayram S,SencarHT,Memon N.Source camera identification based on CFA interpolation.In:Proceedings of IEEE International Conference on Image Processing (ICIP05),Genoa, Italy,2005,3:69-72.
[24] Bayram S,Sencar H T,Memon N.Improvements on source camera-model identification based on CFA interpolation.In:Proceedings of WG 11.9 International Conference on Digital Forensics,Orlando,Florida,USA,2006,187-190.
[25] Choi KS,Lam E Y,Wong KKY.Source camera identification using footptints from lens aberration.In:Digital Photography II,Proceedings of SPIE,2006, 6069:172-179.
[26] Columbia DVMM Research Lab.Columbia image splicing detection evaluation dataset[EB/OL].http://www.ee.columbia.edu/ln/dvmm/downloads/AuthSplicedDataSet/AuthSplicedDataSet.htm,2004.
[27] Zhu X M,Xuan G R,Yao Q M, et a1.Resampling detection in information forensics.Computer Applications,2006,26(11):2596-2597.
[28]周琳娜,郭云彪,杨义先.充分利用视觉冗余的图像不同域信息掩密术.电子学报,2006, 34(12A):1048-1052.
[29]毕华,梁洪力,王珏.重采样方法与机器学习.计算机学报,2009,5(32):862-877.
[30] A.Popescu and H.Farid.Exposing digital forgeries by detecting traces of resampling.IEEE Transactions on Signal Processing,2005,53(2):758-767.
[31]尤玉虎,周孝宽.数字图像最佳插值算法研究.中国空间科学技术,2005,25(3):14-18.
[32]吴琼,李国辉,涂丹等.面向真实性鉴别的数字图像盲取证技术综述.自动化学报,2008,34(12):1458-1466.
[33]陈波.人像合成图像的取证算法研究:硕士学位论文.上海:上海大学,2008.
[34] S.Prasad and K.R.Ramakrishnan.On resampling detection and its application to image tampering.In:Proceedings of the IEEE International Conference on Multimedia and Exposition,Toronto,Canada,2006,1325-1328.
[35] B. Mahdian and S. Saic.On Periodic Properties of Interpolation and Their Application to Image Authentication.In:Third Information Symposium on Information Assurance and Security,2007,439-446.
[36] B . Mahdian and S . Saic.Blind Authentication Using Periodic Properties of Interpolation.IEEE Transactions on Information forensics and Security,2008,3(3):529-538.
[37] A. C. Popescu and H. Farid.Exposing Forgeries in Color Filter Array Interpolated Images.IEEE Transactions on Signal Processing,2005,53(10): 3948-3959.
[38]汪然,平西建.基于图像纹理复杂度和奇异值分解的重采样检测.计算机辅助设计与图形学学报,2010,22(9):1606-1612.
[39]李大锦.可控的连续多尺度纹理合成.计算机工程,2009,35(24):211-212.
[40]张震,康吉全,平西建等.用统计特征量实现的图像拼接盲检测.计算机应用,2008,28(12):3108-3111.
[41] Columbia DVMM Research Lab.Columbia image splicing detection evaluation dataset.http://www.ee.columbia.edu/ln/dvmm/downloads/AuthSplicedDataSet/AuthSplicedDataSet.htm,2004.
[42] Weihai Li,Yuan Yuan, Nenghai Yu.Detecting copy-paste forgery of JPEG image via block artifact grid extraction.In:2008 International workshop on local and non-local approximation in image processing,a satellite event of European signal processing confererence,2008,121-126.
[43]张静,吉建华,苏育挺.一种基于小波系数正则性的JPEG2000图像篡改检测算法.计算机研究与发展,2009,146(47),239-243.
[44] Shih-Fu Chang,Tian-Tsong Ng.welcome to Trust Fotohompage. http://www.ee. columbia. edu/ dvmm/ trustfoto/,2004.
[45] Jerome Buzzi,Frederic Guichard.Automatic Detection of Digital Zooms.In:Paper presented at the Proceedings of the 2004 International Conference on Image Processing(ICIP2004), Singapore, 2004,24-27.
[2] A. C. Gallagher.Detection of Linear and Cubic Interpolation in JPEG compressed images.In:Proceedings of the 2nd Canadian Conference on Computer and Robot Vision (CRV’05),Washington,DC,USA,2005,65-72.
[3]魏瑗珍.低阶统计量改变的自然图像盲取证技术研究:[硕士学位论文].北京:北京交通大学,2009.
[4]周琳娜.数字图像盲取证技术研究:[博士学位论文].北京:北京邮电大学,2007.
[5]吴首阳.JPEG图像盲取证技术研究:[硕士学位论文].安徽:中国科学技术大学,2009.
[6] H. Farid. Exposing Digital Forgeries from JPEG Ghosts, IEEE Transactions on Information Forensics and Security, Mar,2009, 4(1):154-160.
[7]王波,孙璐璐,孔祥维等.图像伪造中模糊操作的异常色调率取证技术.电子学报,2006, 34(12A):2443-2446.
[8]周琳娜,王东明,郭云彪等.基于数字图像边缘特性的形态学滤波取证技术.电子学报,2008,36(6):1047-1051.
[9]朱秀明,宣国荣,姚秋明等.信息取证中图像重采样检测.计算机应用,2006, 26(11):2596-2597.
[10] C. Cruz, R. Reyes, M. Nakano. Image Content Authentication System Based on Semi-Fragile Watermarking.In:51st Midwest Symposium on Circuits and Systems, Knoxville, USA, 2008,306-309.
[11] D. Bashiri , A. Aghagolzadeh, J. Museviniya. A Novel Still Image Error Concealment Using Fragile Watermarking in Wireless Image Transmission and Packet-Switched Networks.In:International Symposium on Telecommunications, Tehran, Iran, 2008, 792-797.
[12] N. Chen, J. Zhu. A Multipurpose Audio Watermarking Scheme for Copyright Protectionand Content Authentication.In:IEEE International Conference on Multimedia and Expo, Hannover, Germany, 2008, 211- 224.
[13] X. Liu, X. Lv, Y. Wang. A semi-fragile digital watermarking algorithm based on integer wavelet matrix norm quantization for medical images.In:2nd International Conference on Bioinformatics and Biomedical Engineering, China, 2008,776-779.
[14] L. Chen, R. Zhang. Novel image authentication algorithm based on semi-fragile watermarking.Journal of Information and Computational Science, 2008, 5(3): 1069-1076.
[15] Tian-Tsong Ng,Shih-Fu Chang,Ching-Yung Lin,et a1.Passive-blind Image Forensics. In:Multimedia Security technologies for digital rights.Elsevier.2006, 18-21.
[16] A.Srivastava,A.Lee,E.Simoncelli, S.Zhu.On advances in statistical modeling of natural images.Journal of Mathematical Imaging and Vision,2003,18(1):17-33.
[17]吴琼.面向真实性检测的数字图像盲取证方法研究:[博士学位论文].湖南:国防科技大学,2008.
[18] Lanh T V,Chong K S,Emmanuel S,et a1.A survey on Digital Camera image forensic methods.In:Proceedings of IEEE International Conference on Multimedia and Expo (ICME07),Beijing,China,2007,16-19.
[19] Sencar H T,Memon N.Overview of state-of-the-art in digital image forensics.In:Part of Indian Statistical Institute Platinum Jubilee Monograph series titled Statistical Science and Interdisciplinary Research,World Scientific Press,2008,67-75.
[20] Swaminathan A,Wu M,Ray Liu KJ.Non-intrusive forensics analysis of visual sensors using output images.IEEE Transactions of Information Forensics and Security, 2007,2(1):91-106.
[21] Celiktutan O,Avcibas I,Sankur B,et a1.Source cell-phone identification.In:Proceedings of IEEE 14th Signal Processing and Communications Applications Conference,Antalya, 2006,l-3.
[22] Tsai MJ,Wu GH.Using image features to identify camera sources.In:Proceedings of IEEE International Coference on Acoustics,Speech and Signal Processing (ICASSP06), Toulouse,2006,2:297-300.
[23] Bayram S,SencarHT,Memon N.Source camera identification based on CFA interpolation.In:Proceedings of IEEE International Conference on Image Processing (ICIP05),Genoa, Italy,2005,3:69-72.
[24] Bayram S,Sencar H T,Memon N.Improvements on source camera-model identification based on CFA interpolation.In:Proceedings of WG 11.9 International Conference on Digital Forensics,Orlando,Florida,USA,2006,187-190.
[25] Choi KS,Lam E Y,Wong KKY.Source camera identification using footptints from lens aberration.In:Digital Photography II,Proceedings of SPIE,2006, 6069:172-179.
[26] Columbia DVMM Research Lab.Columbia image splicing detection evaluation dataset[EB/OL].http://www.ee.columbia.edu/ln/dvmm/downloads/AuthSplicedDataSet/AuthSplicedDataSet.htm,2004.
[27] Zhu X M,Xuan G R,Yao Q M, et a1.Resampling detection in information forensics.Computer Applications,2006,26(11):2596-2597.
[28]周琳娜,郭云彪,杨义先.充分利用视觉冗余的图像不同域信息掩密术.电子学报,2006, 34(12A):1048-1052.
[29]毕华,梁洪力,王珏.重采样方法与机器学习.计算机学报,2009,5(32):862-877.
[30] A.Popescu and H.Farid.Exposing digital forgeries by detecting traces of resampling.IEEE Transactions on Signal Processing,2005,53(2):758-767.
[31]尤玉虎,周孝宽.数字图像最佳插值算法研究.中国空间科学技术,2005,25(3):14-18.
[32]吴琼,李国辉,涂丹等.面向真实性鉴别的数字图像盲取证技术综述.自动化学报,2008,34(12):1458-1466.
[33]陈波.人像合成图像的取证算法研究:硕士学位论文.上海:上海大学,2008.
[34] S.Prasad and K.R.Ramakrishnan.On resampling detection and its application to image tampering.In:Proceedings of the IEEE International Conference on Multimedia and Exposition,Toronto,Canada,2006,1325-1328.
[35] B. Mahdian and S. Saic.On Periodic Properties of Interpolation and Their Application to Image Authentication.In:Third Information Symposium on Information Assurance and Security,2007,439-446.
[36] B . Mahdian and S . Saic.Blind Authentication Using Periodic Properties of Interpolation.IEEE Transactions on Information forensics and Security,2008,3(3):529-538.
[37] A. C. Popescu and H. Farid.Exposing Forgeries in Color Filter Array Interpolated Images.IEEE Transactions on Signal Processing,2005,53(10): 3948-3959.
[38]汪然,平西建.基于图像纹理复杂度和奇异值分解的重采样检测.计算机辅助设计与图形学学报,2010,22(9):1606-1612.
[39]李大锦.可控的连续多尺度纹理合成.计算机工程,2009,35(24):211-212.
[40]张震,康吉全,平西建等.用统计特征量实现的图像拼接盲检测.计算机应用,2008,28(12):3108-3111.
[41] Columbia DVMM Research Lab.Columbia image splicing detection evaluation dataset.http://www.ee.columbia.edu/ln/dvmm/downloads/AuthSplicedDataSet/AuthSplicedDataSet.htm,2004.
[42] Weihai Li,Yuan Yuan, Nenghai Yu.Detecting copy-paste forgery of JPEG image via block artifact grid extraction.In:2008 International workshop on local and non-local approximation in image processing,a satellite event of European signal processing confererence,2008,121-126.
[43]张静,吉建华,苏育挺.一种基于小波系数正则性的JPEG2000图像篡改检测算法.计算机研究与发展,2009,146(47),239-243.
[44] Shih-Fu Chang,Tian-Tsong Ng.welcome to Trust Fotohompage. http://www.ee. columbia. edu/ dvmm/ trustfoto/,2004.
[45] Jerome Buzzi,Frederic Guichard.Automatic Detection of Digital Zooms.In:Paper presented at the Proceedings of the 2004 International Conference on Image Processing(ICIP2004), Singapore, 2004,24-27.