Exploiting the Rolling Shutter Read-Out Time for ENF-Based Camera Identification

Ericmoore Ngharamike, Li Minn Ang*, Kah Phooi Seng, Mingzhong Wang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

The electric network frequency (ENF) is a signal that varies over time and represents the frequency of the energy supplied by a mains power system. It continually varies around a nominal value of 50/60 Hz as a result of fluctuations over time in the supply and demand of power and has been employed for various forensic applications. Based on these ENF fluctuations, the intensity of illumination of a light source powered by the electrical grid similarly fluctuates. Videos recorded under such light sources may capture the ENF and hence can be analyzed to extract the ENF. Cameras using the rolling shutter sampling mechanism acquire each row of a video frame sequentially at a time, referred to as the read-out time ((Formula presented.)) which is a camera-specific parameter. This parameter can be exploited for camera forensic applications. In this paper, we present an approach that exploits the ENF and the (Formula presented.) to identify the source camera of an ENF-containing video of unknown source. The suggested approach considers a practical scenario where a video obtained from the public, including social media, is investigated by law enforcement to ascertain if it originated from a suspect’s camera. Our experimental results demonstrate the effectiveness of our approach.

Original languageEnglish
Article number5039
JournalApplied Sciences (Switzerland)
Volume13
Issue number8
DOIs
Publication statusPublished - Apr 2023

Keywords

  • ENF extraction
  • camera forensics
  • camera sensor types
  • electric network frequency (ENF)
  • read-out time, T
  • rolling shutter mechanism

Fingerprint

Dive into the research topics of 'Exploiting the Rolling Shutter Read-Out Time for ENF-Based Camera Identification'. Together they form a unique fingerprint.

Cite this