Camera Model Identification Based on the Heteroscedastic Noise Model

• Published in: IEEE transactions on image processing Vol. 23; no. 1; pp. 250 - 263
• Main Authors: Thanh Hai Thai, Cogranne, Remi, Retraint, Florent
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.01.2014; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Applied sciences; camera model identification; Cameras; Context; Data Interpretation, Statistical; Detection, estimation, filtering, equalization, prediction; digital forensics; Digital images; Engineering Sciences; Equipment Failure Analysis - methods; Exact sciences and technology; Hypothesis testing; Image Enhancement - methods; Image Interpretation, Computer-Assisted - methods; Image processing; Information, signal and communications theory; Mathematical model; Multimodal Imaging - methods; natural image model; Noise; nuisance parameters; Pattern Recognition, Automated - methods; Photography - instrumentation; Photography - methods; Reproducibility of Results; Sensitivity; Sensitivity and Specificity; Signal and communications theory; Signal and Image processing; Signal processing; Signal, noise; Signal-To-Noise Ratio; Telecommunications and information theory; Testing; Performance evaluation; Hypothesis testing; digital forensics; High performance; Nuisance parameter; Forensic aspect; False alarm rate; Hypothesis test; Statistical test; Model matching; Likelihood ratio test; Numerical simulation; natural image model; System identification; camera model identification; nuisance parameters
• ISSN: 1057-7149; 1941-0042
• DOI: 10.1109/TIP.2013.2290596

The goal of this paper is to design a statistical test for the camera model identification problem. The approach is based on the heteroscedastic noise model, which more accurately describes a natural raw image. This model is characterized by only two parameters, which are considered as unique fingerprint to identify camera models. The camera model identification problem is cast in the framework of hypothesis testing theory. In an ideal context where all model parameters are perfectly known, the likelihood ratio test (LRT) is presented and its performances are theoretically established. For a practical use, two generalized LRTs are designed to deal with unknown model parameters so that they can meet a prescribed false alarm probability while ensuring a high detection performance. Numerical results on simulated images and real natural raw images highlight the relevance of the proposed approach.