Skip to content

Normalised Cross-Correlation (NCC)

Definition

A scalar statistic in the range -1 to 1 that measures how closely the noise residual of a query image matches a reference PRNU pattern. Values near 1 indicate a strong match; values near 0 indicate no relationship.

Value range
-1 to 1
Strong match
Values near 1
No relationship
Values near 0
Application
PRNU-based sensor fingerprinting of images

Common questions

How is a match threshold set when using NCC for source-camera identification?+

Labs typically establish a decision threshold empirically from a reference dataset of known-same and known-different image pairs, since a single fixed cut-off is not universally reliable across cameras, image content, and processing history.

What can produce a misleadingly low or high NCC score?+

Heavy JPEG compression, image resizing, cropping, or strong content correlation between unrelated scenes can distort the noise residual and shift the NCC value away from what pure sensor-pattern similarity would otherwise indicate.

Related terms

Flat-Field Image
A photograph of a uniformly illuminated surface, typically a white wall or diffuser, taken at moderate brightness to avoid saturation. Used to...
Noise Residual
The signal remaining after a denoising filter removes the scene component from an image. For PRNU estimation, the residual approximates the sum...
Peak-to-Correlation Energy (PCE)
An extension of NCC that computes the ratio of the squared correlation peak to the energy of the surrounding correlation map. PCE...
Photo Response Non-Uniformity (PRNU)
A fixed-pattern noise component caused by microscopic sensitivity differences between individual sensor pixels. PRNU is deterministic and camera-specific, making it a sensor...
PRNU Fingerprint
Photo Response Non-Uniformity: the unique pattern of pixel-level sensitivity variations in a camera sensor, which acts as a device fingerprint. Used to...

Explained in

Your journey to becoming a forensic professional starts here.

Practice with mock tests, learn from structured notes, and get your questions answered by a global forensic community, all in one place.