The image series forgery detection algorithm based on the camera pattern noise analysis

Abstract

In the paper, the image series forgery detection algorithm based on the analysis of camera pattern noise is proposed. Distribution characteristics of the camera pattern noise are obtained by extracting the noise component of images from the non-tampered image series. A noise residual of a forgery image is compared with the camera pattern noise. We compare various noise filtering algorithms to choose the one that achieves the best performance of the proposed method. The proposed algorithm is tested both on examples of copy-move forgeries and forgery fragments which were inserted from an image not included in the image series.