Detecting Spliced Image Based on Simplified Statistical Model

The goal of this paper is to design an effective forensic detector for image splicing forgery detection. The detector is designed based on a simplified noise model of JPEG image under the framework of hypothesis testing theory, where the model parameters can serve as camera fingerprints for image forgery detection. By specifying the false alarm rate, the designed detector can achieve high detection performance. Meanwhile, the decision threshold can be set independently of image content. In addition, with the support of solid theoretical explanations, the statistical performance of detector can be analytically established. Numerical results validate the effectiveness of our detector and the superiority compared with the current comparison detectors.