Multimedia Tools and Applications

, Volume 77, Issue 20, pp 27543–27587 | Cite as

Decision-theoretic model to identify printed sources

  • Min-Jen TsaiEmail author
  • Imam Yuadi
  • Yu-Han Tao


When trying to identify a printed forged document, examining digital evidence can prove to be a challenge. Over the past several years, digital forensics for printed document source identification has begun to be increasingly important which can be related to the investigation and prosecution of many types of crimes. Unlike invasive forensic approach which requires a fraction of the printed document as the specimen for verification, noninvasive forensic technique uses the optical mechanism to explore the relationship between the scanned images and the source printer. To explore the relationship between source printers and images obtained by the scanner, the proposed decision-theoretical approach utilizes image processing techniques and data exploration methods to calculate many important statistical features, including: Local Binary Pattern (LBP), Gray Level Co-occurrence Matrix (GLCM), Discrete Wavelet Transform (DWT), Spatial filters, the Wiener filter, the Gabor filter, Haralick, and SFTA features. Consequently, the proposed aggregation method intensively applies the extracted features and decision-fusion model of feature selections for classification. In addition, the impact of different paper texture or paper color for printed sources identification is also investigated. In the meantime, the up-to-date techniques based on deep learning system is developed by Convolutional Neural Networks (CNNs) which can learn the features automatically to solve the complex image classification problem. Both systems have been compared and the experimental results indicate that the proposed system achieve the overall best accuracy prediction for image and text input and is superior to the existing approaches. In brief, the proposed decision-theoretical model can be very efficiently implemented for real world digital forensic applications.


Decision fusion Scanner Feature filters Feature selection Support Vector Machines (SVM) Deep learning Convolutional Neural Networks (CNNs) 



This work was partially supported by the National Science Council in Taiwan, Republic of China, under NSC104-2410-H-009-020-MY2 and NSC106-2410-H-009-022-.

The authors would like to thank the anonymous reviewers with their valuable comments to improve the quality of this manuscript. Special thanks to Jin-Sheng Yin and Goang-Jiun Wang at National Chiao Tung University who help the revision and the software experiments.


  1. 1.
    Ali GN, Mikkilineni AK, Chiang PJ, Allebach GT, Delp EJ (2003) Intrinsic and extrinsic signatures for information hiding and secure printing with electrophotographic devices. In International Conference on Digital Printing Technologies. New Orleans, LA, USA; 28 Sept–3 Oct, 511–515Google Scholar
  2. 2.
    Bekhti MA, Kobayashi Y (2016) Prediction of vibrations as a measure of terrain traversability in outdoor structured and natural environments. In: Image and video technology, Vol. 9431 of the series lecture notes in computer science. Springer International Publishing, Auckland 282–294. CrossRefGoogle Scholar
  3. 3.
    Bulan O, Mao J, Sharma G (2009) Geometric distortion signatures for printer identification. International conference on acoustics, speech and signal processing (ICASSP), Taipei, 1401–1404.
  4. 4.
    Burger W, Burge MJ (2018) Digital image processing: an introduction algorithmic using Java. Springer Science Business Media, New YorkGoogle Scholar
  5. 5.
    Choi JH, Lee HY, Lee HK (2013) Color laser printer forensic based on noisy feature and support vector machine classifier. Multimed Tools Appl 67:363–382. CrossRefGoogle Scholar
  6. 6.
    Costa AF, Humpire-Mamani G, Traina AJM (2012) An efficient algorithm for fractal analysis of textures. SIBGRAPI conference on graphics, patterns and images, August, Ouro Preto, 39–46.
  7. 7.
    Daugman JG (1988) Complete discrete 2D Gabor transforms by neural networks for image-analysis and compression. IEEE Trans Acoust Speech Signal Process 36(7):1169–1179. CrossRefzbMATHGoogle Scholar
  8. 8.
    Ferreira A, Navarro LC, Pinheiro G, Santos JAD, Rocha A (2015) Laser printer attribution: exploring new features and beyond. Forensic Sci Int 247:105–125. CrossRefGoogle Scholar
  9. 9.
    Gonzales RC, Woods RE (2008) Digital image processing, 3rd edn. Prentice Hall, New JerseyGoogle Scholar
  10. 10.
    Gonzales RC, Woods RE, Eddins SL (2009) Digital image processing using MATLAB, 2nd edn. Gatesmark, United StatesGoogle Scholar
  11. 11.
    Haghighat M, Zonout S, Abdel-Mottaleb M (2015) CloudID: trustworthy cloud-based and cross-enterprise biometric identification. Expert Syst Appl 42(21):7905–7916. CrossRefGoogle Scholar
  12. 12.
    Haralick RM, Shanmugam K, Dinstein I (1973) Textural features for image classification. IEEE Trans Syst Man Cybernet SMC 3(6):610–621CrossRefGoogle Scholar
  13. 13.
    He K et al (2016) Deep residual learning for image recognition. IEEE conference on computer vision and pattern recognition (CVPR) 770–778Google Scholar
  14. 14.
    Hinton G, Salakhutdinov R (2006) Reducing the dimensionality of data with neural networks. Science 313(5786):504–507MathSciNetCrossRefGoogle Scholar
  15. 15.
    Hsu CW, Chang CC, Lin CJ (2003) A practical guide to support vector classification. National Taiwan University, Taipei Accessed 8 Apr 2017Google Scholar
  16. 16.
  17. 17.
  18. 18.
    Hubel D, Wiesel T (1962) Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. J Physiol 160(1):106–154CrossRefGoogle Scholar
  19. 19.
    Jurič I, Ranđelović D, Karlović I, Tomić I (2014) Influence of the surface roughness of coated and uncoated papers on the digital print mottle. J Graph Eng Des 5(1):17–23Google Scholar
  20. 20.
    Kawasaki M, Ishisaki M (2009) Investigation into the cause of print mottle in halftone dots of coated paper: effect of optical dot gain non-uniformity 63(11):1362–1373. Accessed 7 April 2017
  21. 21.
    Kee E, Farid H (2008) Printer profiling for forensics and ballistics. ACM Workshop on Multimedia and Security, 3–10Google Scholar
  22. 22.
    Khanna N, Delp EJ (2010) Intrinsic signatures for scanned documents forensics: effect of font shape and size. Proceedings of 2010 I.E. international symposium on circuits and systems (ISCAS), 30 May– 2 June.
  23. 23.
    Kim DG, Lee HK (2014) Color laser printer identification using photographed halftone images. Proc. of EUSIPCO. September, IEEE, Lisbon, 795–799Google Scholar
  24. 24.
    Kim KI, Jung K, Park SH, Kim HJ (2002) Support vector machines for texture classification. IEEE Trans Pattern Anal Mach Intell 24(11):1542–1550. CrossRefGoogle Scholar
  25. 25.
    Krizhevsky A, Sutskever I, Hinton G (2012) ImageNet classification with deep convolutional neural networks. Process Int Conf Neural Inf Process Syst (NIPS) 1:1097–1105Google Scholar
  26. 26.
    Lecun Y, Bottou L, Bengio Y, Haffner P (1998) Gradient-based learning applied to document recognition. Proc IEEE 86(11):2278–2324CrossRefGoogle Scholar
  27. 27.
    Lewis JA (2014) Forensic document examination: fundamentals and current trends. Elsevier, Oxford. CrossRefGoogle Scholar
  28. 28.
    Lin CJ (2007) A tutorial of the wavelet transforms. National Taiwan University, Accessed 3 Apr 2017
  29. 29.
    Lopez FM, Martins DC, Cesar RM (2008) Feature selection environment for genomic applications. BMC Bioinformatics 9:451. CrossRefGoogle Scholar
  30. 30.
    Mäenpää T, Pietikäinen M (2004) Texture analysis with local binary patterns. In: Chen CH, Wang PSP (eds) Handbook of pattern recognition & computer vision, 3rd edn. World Scientific, Singapore, pp 115–118Google Scholar
  31. 31.
    Markoff J (2012) How many computers to identify a cat? 16,000. The New York. Retrieved June 22, 2012, from
  32. 32.
    McAndrew A (2016) A computational introduction to digital image processing. CRC Press, Boca RatonzbMATHGoogle Scholar
  33. 33.
    Mikkilineni AK, Chiang PJ, Ali GN, Chiu GT, Allebach JP, Delp EJ (2004) Printer identification based on textural features. Intl. conference on digital printing technologies 306–311Google Scholar
  34. 34.
    Mikkilineni AK, Chiang JP, Ali GN, Chiu GT, Allebach JP, Delp EJ (2005) Printer identification based on graylevel co-occurrence features for security and forensic applications. Intl. Conference on Security, Steganography and Watermarking of Multimedia Contents VII, Proc. SPIE. 5681, 430–440, March 21.
  35. 35.
    Mikkilineni AK, Arslan O, Chiang PJ, Kumontoy RM, Allebach JP, Chiu GT (2005) Printer forensics using SVM techniques. Intl. conference on digital printing technologies, 223–226Google Scholar
  36. 36.
    Mikkilineni AK, Khanna N, Delp EJ (2010) Texture based attacks on intrinsic signature based printer identification. Proceedings SPIE 7541, Media Forensics and Security II, 28 January.
  37. 37.
    Netzer Y, Wang T, Coates A, Bissacco A, Wu B, Ng A (2011) Reading digits in natural images with unsupervised feature learning. NIPS workshop on deep learning and unsupervised feature learningGoogle Scholar
  38. 38.
    Ojala T, Pietikäinen M, Harwood D (1996) A comparative study of texture measures with classification based on featured distributions. Pattern Recogn 29(1):51–59. CrossRefGoogle Scholar
  39. 39.
    Ojala T, Pietikäinen M, Mäenpää T (2002) Multiresolution gray-scale and rotation invariant texture classification with LBP. IEEE Trans Pattern Anal Mach Intell 24(7):971–987. CrossRefzbMATHGoogle Scholar
  40. 40.
    Pudil P, Ferry FJ, Novovicova J, Kittler J(1994) Floating search methods for feature selection with nonmonotonic criterion functions. IEEE, 1051-465U9, Accessed 3 Apr 2017
  41. 41.
    Pudil P, Novovicova J, Kittler J (1994) Floating search methods in feature selection. Pattern Recogn Lett 15:1119–1125CrossRefGoogle Scholar
  42. 42.
    Qiu Z, Jin J, Lam HK, Zhang Y, Wang X, Cichocki A (2016) Improved SFFS method for channel selection in motor imagery based BCI. Neurocomputing. CrossRefGoogle Scholar
  43. 43.
    Rumelhart E, Geoffrey E, Ronald J (1986) Learning representations by back-propagating errors. Nature 323:533–536CrossRefGoogle Scholar
  44. 44.
    Russakovsky O et al (2015) Imagenet large scale visual recognition challenge. Int J Comput Vis 115(3):211–252MathSciNetCrossRefGoogle Scholar
  45. 45.
    Ryu SJ, Lee KY, Im DH, Choi JH, Lee HK (2010) Electrophotographic printer identification by halftone texture analysis. In: IEEE Intl. conference on acoustics speech and signal processing (ICASSP), 1846–1849.
  46. 46.
    Say OT, Sauli Z, Retnasamy V (2013) High density printing paper quality investigation. IEEE Regional Symposium on Micro and Nano electronics (RSM), Langkawi, 273–277.
  47. 47.
    Schalkoff RJ (1989) Digital image processing and computer vision. Wiley, AustraliaGoogle Scholar
  48. 48.
    Simonyan K. and Zisserman A. (2015) Very deep convolutional networks for large-scale image recognition. IEEE conference on computer vision and pattern recognition (CVPR), arXiv preprint arXiv:1409.1556Google Scholar
  49. 49.
    Su R, Pekarovicova A, Fleming PD, Bliznyuk V (2005) Physical properties of LWC papers and Gravure Ink Mileage, Accessed 3 Apr 2017
  50. 50.
    Szegedy, C. (2015) Going deeper with convolutions. IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
  51. 51.
    The Electron Microscope (2017) Accessed 11 Apr 2017
  52. 52.
    Tong S, Koller D (2001) Support vector machine active learning with applications to text classification. J Mach Learn Res 45–66. Accessed 7 Apr 2017
  53. 53.
    Tsai MJ, Liu J (2013) Digital forensics for printed source identification. In: IEEE international symposium on circuits and systems (ISCAS), May, 2347–2350.
  54. 54.
    Tsai MJ, Liu J, Wang CS, Chuang CH (2011) Source color laser printer identification using discrete wavelet transform and feature selection algorithms. IEEE international symposium on circuits and systems (ISCAS), May, Rio de Janeiro, 2633–2636.
  55. 55.
    Tsai MJ, Yin JS, Yuadi I, Liu J (2014) Digital forensics of printed source identification for Chinese characters. Multimed Tools Appl 73:2129–2155. CrossRefGoogle Scholar
  56. 56.
    Tsai MJ, Hsu CL, Yin JS, Yuadi I (2015) Japanese character based printed source identification. IEEE International Symposium on Circuits and Systems (ISCAS), May, Lisbon, 2800–2803.
  57. 57.
    Vega LR, Rey H (2013) A rapid introduction to adaptive filtering. Springer-Verlag, BerlinCrossRefGoogle Scholar
  58. 58.
    Wu Y, Kong X, You X, Guo Y 2009 Printer forensics based on page document’s geometric distortion. Intl. conference on image processing (ICIP), Cairo, 2909–2912.
  59. 59.
    Zhou H, Wu J, Zhang J (2010) Digital image processing: part 1. Ventus Publishing ApS, DenmarkGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Institute of Information ManagementNational Chiao Tung UniversityHsin-ChuTaiwan, Republic of China
  2. 2.Department of Information and Library ScienceAirlangga UniversitySurabayaIndonesia

Personalised recommendations