A novel multi-size block benford's law scheme for printer identification

Authors:
Weina Jiang;Anthony T. S. Ho;Helen Treharne;Yun Q. Shi
Affiliations:
Dept. of Computing, University of Surrey Guildford, UK;Dept. of Computing, University of Surrey Guildford, UK;Dept. of Computing, University of Surrey Guildford, UK;Dept. of Electrical and Computer Engineering, New Jersey Institute of Technology, Newwark, NJ
Venue:
PCM'10 Proceedings of the 11th Pacific Rim conference on Advances in multimedia information processing: Part I
Year:
2010

Citing 6
Cited 0

Digital halftoning

Digital halftoning
Survey of Scanner and Printer Forensics at Purdue University

IWCF '08 Proceedings of the 2nd international workshop on Computational Forensics
Geometric distortion signatures for printer identification

ICASSP '09 Proceedings of the 2009 IEEE International Conference on Acoustics, Speech and Signal Processing
Image forensics using generalised Benford's law for accurate detection of unknown JPEG compression in watermarked images

DSP'09 Proceedings of the 16th international conference on Digital Signal Processing
Digital camera identification from sensor pattern noise

IEEE Transactions on Information Forensics and Security
Determining Image Origin and Integrity Using Sensor Noise

IEEE Transactions on Information Forensics and Security

Quantified Score

Hi-index	0.00

Visualization

Abstract

Identifying the originating device for a given media, i.e. the type, brand, model and other characteristics of the device, is currently one of the important fields of digital forensics. This paper proposes a forensic technique based on the Benford's law to identify the printer's brand and model from the printed-and-scanned images at which the first digit probability distribution of multi-size block DCT coefficients are extracted that constitutes a feature vector as the input to support vector machine (SVM) classifier. The proposed technique is different from the traditional use of noise feature patterns appeared in the literature. It uses as few as nine numbers of forensic features representing each printer by leveraging properties of the Benford's law for printer identification. Experiments conducted over electrophotographic (EP) printers and deskjet printers achieve an average of 96.0% classification rate of identification for five distinct printer brands and an average of 94.0% classification rate for six diverse printer models out of those five brands.