基于插值检测的扫描图像源辨识方法
|
摘要
针对扫描图像真实性的问题,提出了应用EM算法与扫描仪传感器噪声的自然统计特性相结合的识别方法。该识别方法一共生成63维特征,用SVM进行训练,通过大量实验图像,文中对取证方法效果进行了分析研究。结果显示,扫描图像虽然在生成过程中不存在CFA插值过程,但是在被显示前,仍经过了插值运算,傅立叶频谱图仍具有周期性。该算法对SG图像(扫描图像)与CG图像、SG图像与DG图像(数码相机生成图像)之间的识别,达到了较理想的效果。
Since scanned images are not generated by CFA process,a part of researchers have studied scanned image forensic around sensor pattern noise. A method is proposed to contain the statistical features of the noise image and features derived from EM algorithm for the authenticity of the scanned image.This method generates a total of 63 dimensional feature classified by SVM. By analyzing the evidence effectives and testing a large number of images,the work results show that although scanned images is not generated with CFA process,scanned images also has experienced interpolation on its Fourier spectrum map. Moreover,for the purpose of distinguish between scanned images and other images( CG,DG),this method achieves the desired results.
Since scanned images are not generated by CFA process,a part of researchers have studied scanned image forensic around sensor pattern noise. A method is proposed to contain the statistical features of the noise image and features derived from EM algorithm for the authenticity of the scanned image.This method generates a total of 63 dimensional feature classified by SVM. By analyzing the evidence effectives and testing a large number of images,the work results show that although scanned images is not generated with CFA process,scanned images also has experienced interpolation on its Fourier spectrum map. Moreover,for the purpose of distinguish between scanned images and other images( CG,DG),this method achieves the desired results.
引文
[1]GLOE T,FRANZ E,WINKLER A,Forensics for flatbed scanners[C]∥Proc of SPIE Security,Steganography,Watermarking of Multimedia Contents IX,2007:65051I.
[2]DIRIK A E.,SENCAR H T,MEMON N,Flatbed scanner identification based on dust and scratches over scanner platen[C]∥IEEE International Conference on Signal Processing,2009:1385-1388.
[3]KHANNA NITIN,MIKKILINENI A K,DELP Edward J.Scanner identification using sensor pattern noise[C]∥Proc of SPIE Security,Steganography,and Watermarking of Multimedia Contents IX,2007:65051K.
[4]KHANNA NITIN,GEORGE T C,DELP Edward J.Forensic techniques for classifying scanner,computer generated and digital camera images[C]∥IEEE International Conference on Signal Processing,2008:1653-1656.
[5]GOU H,SWAMINATHAN A,Robust scanner identification based on noise features[C]∥Proc of SPIE Security,Steganography,Watermarking of Multimedia Contents IX,2007:65050S.
[6]张桂兰.扫描仪的原理及其使用方法[J].印刷技术,1994,11:14.
[7]HOLST G C,LOMHEIM T S,CMOS/CCD Sensors and Camera Systems[C]∥The International Society for Optical Engine,2007.
[8]黄颖,李伟,EM算法与K-Means算法的比较[J].计算机与现代化,2007,9:12.
[9]POPESCU A C,FARID Hany.Exposing digital forgeries by detecting traces of re-sampling[J].IEEE Tran on Signal Processing,2005,53(2):758-767.
[10]LUKAS J,FRIDRICH J,Digital camera identification from sensor pattern noise[J].IEEE Trans on Information Forensics and Security,2006,1(2):205–214.
[11]GOU H,SWAMINATHAN A,WU Min.Intrinsic sensor noise features for forensic analysis on scanners and scanned images[J].IEEE Trans on Information Forensics and Security,2009,4(3):476-491.
[12]CRISTIANINI N,SHAWE-Taylor J,An introduction to support vector machines(and other kernel-based learning methods)[M].Cambridge University Press,2000.
[13]CHOI Chang-Hee,LEE Min-Jeong.Scanner identification using spectral noise in the frequency domain[C]∥IEEE International Conference on Image Processing,2010:2121-2124.
[14]KHANNA Nitin,MIKKILINENI A K,DELP Edward J,Scanner identification using feature-based processing and analysis[J].IEEE Trans on Information Forensics and Security,2009,4(1):123-139.
[15]张春梅,张太镒.基于双密度双树复数小波变换的图像去噪[J].西安交通大学学报,2006,40(2):4-8.
[16]CHANG C C,LIN C J.LIBSVM:A library for support vector machines[EB/OL].(2001-04-17)[2013-03-13].Available:http://www.csie.ntu.edu.tw/~cjlin/libsvm.
[17]KEETTHI Sathiya.Training and testing low-degree polynomial data mappings via Linear SVM[J].Journal of Machine Learning Research,2010(11):1471-1490.
[18]尹京,方艳梅.数码翻拍图像取证算法[J].中山大学学报:自然科学版,2011,50(6):48-52.
[2]DIRIK A E.,SENCAR H T,MEMON N,Flatbed scanner identification based on dust and scratches over scanner platen[C]∥IEEE International Conference on Signal Processing,2009:1385-1388.
[3]KHANNA NITIN,MIKKILINENI A K,DELP Edward J.Scanner identification using sensor pattern noise[C]∥Proc of SPIE Security,Steganography,and Watermarking of Multimedia Contents IX,2007:65051K.
[4]KHANNA NITIN,GEORGE T C,DELP Edward J.Forensic techniques for classifying scanner,computer generated and digital camera images[C]∥IEEE International Conference on Signal Processing,2008:1653-1656.
[5]GOU H,SWAMINATHAN A,Robust scanner identification based on noise features[C]∥Proc of SPIE Security,Steganography,Watermarking of Multimedia Contents IX,2007:65050S.
[6]张桂兰.扫描仪的原理及其使用方法[J].印刷技术,1994,11:14.
[7]HOLST G C,LOMHEIM T S,CMOS/CCD Sensors and Camera Systems[C]∥The International Society for Optical Engine,2007.
[8]黄颖,李伟,EM算法与K-Means算法的比较[J].计算机与现代化,2007,9:12.
[9]POPESCU A C,FARID Hany.Exposing digital forgeries by detecting traces of re-sampling[J].IEEE Tran on Signal Processing,2005,53(2):758-767.
[10]LUKAS J,FRIDRICH J,Digital camera identification from sensor pattern noise[J].IEEE Trans on Information Forensics and Security,2006,1(2):205–214.
[11]GOU H,SWAMINATHAN A,WU Min.Intrinsic sensor noise features for forensic analysis on scanners and scanned images[J].IEEE Trans on Information Forensics and Security,2009,4(3):476-491.
[12]CRISTIANINI N,SHAWE-Taylor J,An introduction to support vector machines(and other kernel-based learning methods)[M].Cambridge University Press,2000.
[13]CHOI Chang-Hee,LEE Min-Jeong.Scanner identification using spectral noise in the frequency domain[C]∥IEEE International Conference on Image Processing,2010:2121-2124.
[14]KHANNA Nitin,MIKKILINENI A K,DELP Edward J,Scanner identification using feature-based processing and analysis[J].IEEE Trans on Information Forensics and Security,2009,4(1):123-139.
[15]张春梅,张太镒.基于双密度双树复数小波变换的图像去噪[J].西安交通大学学报,2006,40(2):4-8.
[16]CHANG C C,LIN C J.LIBSVM:A library for support vector machines[EB/OL].(2001-04-17)[2013-03-13].Available:http://www.csie.ntu.edu.tw/~cjlin/libsvm.
[17]KEETTHI Sathiya.Training and testing low-degree polynomial data mappings via Linear SVM[J].Journal of Machine Learning Research,2010(11):1471-1490.
[18]尹京,方艳梅.数码翻拍图像取证算法[J].中山大学学报:自然科学版,2011,50(6):48-52.