Robust Source Camera Identification of Image Using Mean Subtracted Contrast Normalization for Digital Forensic Investigation

Abstract

This paper aims to investigate source camera identification (SCI), one of the challenges in image forensics. Besides SCI, research on the robustness of the SCI algorithm for practical applications is necessary because images are altered due to JPEG compression, Gaussian white noise, and rescaling on social networking platforms, where the fingerprints of the images may be contaminated. In this study, we explore robust SCI by extracting sensor pattern noise (SPN) for each camera model using mean subtracted contrast normalization (MSCN). Firstly, MSCN is extracted for every camera model. In this study, it is termed basic sensor pattern noise (BSPN). We further enhance the basic sensor pattern in the Fourier domain to obtain the final fingerprint, termed SPN. To attribute an unknown image to its source camera, the SPN of the image is extracted. The SPN of the unknown image is then correlated with the reference SPN of all camera models, respectively. If the correlation is greater than the particular threshold, then it leads to the camera model identification. Experiment results confirm that the proposed method effectively attributes an unknown image with its source camera and can resist JPEG compression, Gaussian white noise, and rescaling attacks more efficiently than the state-of-the-art SCI methods. Furthermore, the time required to extract SPN from the query image is also low.

Received: 20 May 2024 | Revised: 27 February 2025 | Accepted: 28 March 2025

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

The Dresden image database that supports the findings of this study is openly available at https://doi.org/10.1145/1774088. 1774427, reference number [43].

Author Contribution Statement

Pabitra Roy: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization. Shyamali Mitra: Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision. Nibaran Das: Validation, Formal analysis, Resources, Data curation, Writing – review & editing, Visualization, Supervision.