No abstract

No abstract

No abstract

Establishing the pedigree of a digital image, such as the type of processing applied to it, is important for forensic analysts because processing generally affects the accuracy and applicability of other forensic tools used for, e.g., identifying the camera (brand) and/or inspecting the image integrity (detecting regions that were manipulated). Given the superiority of automatized tools called deep convolutional neural networks to learn complex yet compact image representations for numerous problems in steganalysis as well as in forensic, in this article we explore this approach for the task of detecting the processing history of images. Our goal is to build a scalable detector for practical situations when an image acquired by a camera is processed, downscaled with a wide variety of scaling factors, and again JPEG compressed since such processing pipeline is commonly applied for example when uploading images to social networks, such as Facebook. To allow the network to perform accurately on a wide range of image sizes, we investigate a novel CNN architecture with an IP layer accepting statistical moments of feature maps. The proposed methodology is benchmarked using confusion matrices for three JPEG quality factors.

Knowing the history of global processing applied to an image can be very important for the forensic analyst to correctly establish the image pedigree, trustworthiness, and integrity. Global edits have been proposed in the past for "laundering" manipulated content because they can negatively affect the reliability of many forensic techniques. In this paper, we focus on the more difficult and less addressed case when the processed image is JPEG compressed. First, a bank of binary linear classifiers with rich media models are built to distinguish between unprocessed images and images subjected to a specific processing class. For better scalability, the detector is not built in the rich feature space but in the space of projections of features on the weight vectors of the linear classifiers. This decreases the computational complexity of the detector and, most importantly, allows estimation of the distribution of the projections by fitting a mutlivariate Gaussian model to each processing class to construct the final classifier as a maximum-likelihood detector. Well-fitting analytic models permit a more rigorous construction of the detector unachievable in the original high-dimensional rich feature space. Experiments on grayscale as well as color images with a range of JPEG quality factors and four processing classes are used to show the merit of the proposed methodology.