Decision-theoretic model to identify printed sources

Abstract

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.

Printer source identification by feature modeling in the total variable printer space

“…Various approaches have been proposed for identifying the print source based on intrinsic signatures. These approaches are divided into methods based on: 1-banding effect [13,14], 2-identifying paper-ink interaction [15], 3-applying Benford's law [16], 4printing textures [17,18], 5-distortion traces in documents [10,19],…”

Source Printer Classification Using Printer Specific Local Texture Descriptor

“…This work reported both groupof-character (GOC) level and page level accuracies. In contrast, a very recent method proposes a decision-fusion model based approach for source printer classification [20]. In particular, it applies feature selection approach on a variety of features including local binary pattern (LBP), GLCM and DWT.…”

“…The features include gray-level co-occurrence matrices (GLCM) [11], GLCM extended in multiple directions, and scales [12], convolutional texture gradient filter (CTGF) based on filtering textures with a specific gradient [12], a combination of the discrete wavelet transform (DWT) and GLCM based features [13], a combination of features obtained after applying a spatial filter, Wiener filter, and Gabor filter [14], and GLTrP-based features [3]. A recent method proposes a decision-fusion model-based approach for source printer classification [15]. All these methods extract features that learn a model for a specific letter type except [3], which introduces a singleclassifier approach that learns a single model for all letter types printed on a document.…”

Source Printer Identification from Document Images Acquired using Smartphone

“…Tsai et al [22] used a combination of discrete wavelet transform (DWT) and GLCM based features extracted from all occurrences of a specific Chinese character followed by classification using SVM. The same authors extended this method to a decisionfusion model-based approach for source printer identification, which applies a feature selection approach on a variety of features, including local binary pattern (LBP), GLCM and DWT [23]. Ferreira et al [8] used statistical features extracted from GLCM extended in multiple directions and scales as well as convolutional texture gradient filter (CTGF) which is a new feature descriptor based on filtering textures with a specific gradient.…”