SNR_star – Covariance-based SNR measurement
Author: Motohiro TABUCHI
SNR_star is an ImageJ macro that estimates signal variance and noise variance using covariance between two repeated images acquired under identical imaging conditions.
It provides an unbiased and statistically optimal estimation of SNR.
Features
- Covariance-based signal variance estimation
- Noise variance estimation from difference of images
- Outputs SNR* [dB], ROI size, signal variance, and noise variance
- Simple workflow using ImageJ ROI tools
Requirements
- ImageJ 1.53 or later
- Exactly two observed images of identical dimensions acquired under identical imaging conditions
Download
Notes
- Assumes additive, zero-mean, independent noise between the two images
- Larger ROIs improve estimation stability
- Negative covariance may indicate unsuitable imaging conditions
