Using Pattern Noise of Imaging Sensors for Imaging Hardware Identification

Background:

A practical and important problem in digital forensics is to identify reliably the imaging device that acquired a particular electronic image or representation thereof. Prior approaches to identifying the device that acquired a particular image have significant limitations and/or limited reliability. Thus, as electronic images and digital video have replaced their analog counterparts, there is a need for reliable, inexpensive, and fast identification of a particular electronic image.

 Technology Overview:

The present technology provides a simple, reliable and automated method for image acquisition device identification that represents a forensics analysis tool to identify the physical camera used to create an image or video. The image identification process utilizes a pattern noise of the imaging sensor. Because every imaging sensor (CCD, CMOS, etc.) is slightly different due to limitations in manufacturing precision, there is pixel response non-uniformity noise (“PRNU”) in every picture a camera takes. This unique signal can be detected by proprietary filtering techniques, making  PRNU a stable, universally applicable, robust camera fingerprint. The confidence level of positive identification is extremely high and rises as the number of images or the length of the video sequence increases. Probability of false alarm (i.e., positive false identification) can be set as low as needed.  Identification is possible for all formats and at extremely low bit-rates.

 Advantages:

• High reliability and accuracy.
• Applicability to all sensor types and image acquisition devices.
• Ability to distinguish between cameras of the exact same brand.
• Simplicity and computational efficiency.

 Applications:

• Determining the origin of digital images
•  Matching an image or video sequence to a camera device
• Digital forgery detection
• Determining processing history
• Anti-piracy & counterfeit investigations

  Intellectual Property Summary:

  • U.S. 7,787,030
  • U.S. 7,616, 237