A Modified Fourier-Mellin Approach for Source Device Identification on Stabilized Videos

TL;DR

This paper introduces a frequency domain method based on a modified Fourier-Mellin transform to efficiently realign camera noise patterns in stabilized videos, improving device identification accuracy.

Contribution

It presents a novel Fourier-Mellin based approach for geometric realignment of noise patterns in stabilized videos, reducing computational complexity compared to existing methods.

Findings

Effective realignment of noise patterns demonstrated on real videos.

Improved accuracy in source device identification.

Promising results on a standard dataset.

Abstract

To decide whether a digital video has been captured by a given device, multimedia forensic tools usually exploit characteristic noise traces left by the camera sensor on the acquired frames. This analysis requires that the noise pattern characterizing the camera and the noise pattern extracted from video frames under analysis are geometrically aligned. However, in many practical scenarios this does not occur, thus a re-alignment or synchronization has to be performed. Current solutions often require time consuming search of the realignment transformation parameters. In this paper, we propose to overcome this limitation by searching scaling and rotation parameters in the frequency domain. The proposed algorithm tested on real videos from a well-known state-of-the-art dataset shows promising results.