Document forgery detection using distortion mutation of geometric parameters in characters

Abstract

Tampering related to document forgeries is often accomplished by copy-pasting or add-printing. These tampering methods introduce character distortion mutation in documents. We present a method of exposing document forgeries using distortion mutation of geometric parameters. We estimate distortion parameters, which consist of translation and rotation distortions, through image matching for each character. Detection of tampered characters with distortion mutation occurs based on a distortion probability, which is calculated from character distortion parameters. The introduction of a visualized probability map describes the degree of distortion mutation for a full page. The proposed method exposes the forgeries based on individual characters and applies to English and Chinese document examinations. Experimental results demonstrate the effectiveness of our method on low JPEG compression quality and low resolution. Downloaded From: http://electronicimaging.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms Journal of Electronic Imaging 023008-2 Mar∕Apr 2015 • Vol. 24(2) Shang, Kong, and You: Document forgery detection using distortion mutation of geometric parameters in characters Downloaded From: http://electronicimaging.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms Journal of Electronic Imaging 023008-4 Mar∕Apr 2015 • Vol. 24(2) Shang, Kong, and You: Document forgery detection using distortion mutation of geometric parameters in characters Downloaded From: http://electronicimaging.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms Journal of Electronic Imaging 023008-5 Mar∕Apr 2015 • Vol. 24(2) Shang, Kong, and You: Document forgery detection using distortion mutation of geometric parameters in characters Downloaded From: http://electronicimaging.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms Journal of Electronic Imaging 023008-6 Mar∕Apr 2015 • Vol. 24(2) Shang, Kong, and You: Document forgery detection using distortion mutation of geometric parameters in characters Downloaded From: http://electronicimaging.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms Journal of Electronic Imaging 023008-8 Mar∕Apr 2015 • Vol. 24(2) Shang, Kong, and You: Document forgery detection using distortion mutation of geometric parameters in characters Downloaded From: http://electronicimaging.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

VIPPrint: Validating Synthetic Image Detection and Source Linking Methods on a Large Scale Dataset of Printed Documents

“…After the letters are classified, the source of a document can be found by verifying the most voted class among all the individual letters “I” classification. Several other techniques used a similar pipeline with few modifications, such as considering the statistics of gray-level co-occurrence matrices [ 4 , 5 , 6 , 11 ], Distance Transform [ 8 ], Discrete Cosine Transform [ 10 ], statistics of gray-level co-occurrence matrices together with residual noise and sub-bands of wavelet transform [ 12 , 13 , 14 , 17 ], deep neural networks [ 18 , 20 ], ad hoc texture descriptors [ 19 , 22 ] among others [ 7 , 9 , 15 , 16 , 21 ].…”

“…Solutions for printer attribution are mostly based on the analysis of the extrinsic artifacts contained in printed documents, with the most popular ones for laser printers being banding, jitter, and skewed jitters. The presence of these artifacts has been exploited by several works to identify the sources of printed texts [ 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 ], color images [ 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ], or both [ 34 , 35 , 36 ]. Manipulation detection in printed documents has received some attention only recently [ 37 ] and mainly refers to unveiling post-processing operations that could alter the semantic meanings of the images.…”

Source Printer Classification Using Printer Specific Local Texture Descriptor

“…A recent method based on printer-specific local texture descriptor (PSLTD) [4] tries to address this challenge. A completely different category is characterized by geometric distortion-based approaches [20,21], which rely on features from translational and rotational distortions of printed text relative to its reference soft copy [22,23,24]. A detailed literature review has been covered in [12,2,3].…”

Forensik Digital Berdasarkan Citra Mikroskop untuk Autentikasi Arsip Tercetak

“…Teknik ini mengekstrak geometris huruf tercetak dengan memperkirakan posisi titik pada halftone di printer pada set pelatihan dan membandingkan, dengan korelasi, posisi poin dalam tes. Shang et al (2015) mengusulkan ekstraksi Vektor fitur 9d dari dokumen yang dipindai berdasarkan pada Benford's law. Fitur-fitur ini adalah distribusi probabilitas digit pertama Koefisien Discrete Cosine Transform (DCT) dari blok multi-ukuran.…”