Expand description
Metrics for scalar intensity images.
stats summarizes one image. All other functions compare a reference
image with an estimate; signed residuals are estimate - reference.
Implementation modules are private so callers use this stable facade.
Structs§
- Intensity
Comparison Metrics - Aggregate residual statistics comparing an estimate with a reference image.
- Intensity
Stats - Summary statistics calculated from one intensity image.
Enums§
- Intensity
Metric Error - Errors returned by intensity comparison metrics.
Functions§
- amplitude_
nrmse - Compare square-root intensities, normalized by the reference amplitude L2 norm.
- bias
- Return the mean signed residual,
estimate - reference. - compare_
intensity - Calculate aggregate residual statistics for a reference/estimate pair.
- correlation
- Return the Pearson correlation coefficient of valid pixels.
- fitted_
gain - Fit the least-squares scalar
gaininestimate ≈ gain × reference. - mae
- Return the mean absolute error.
- mean_
poisson_ deviance - Return mean Poisson deviance, flooring estimate intensities at
epsilon. - mse
- Return the mean squared error.
- nrmse
- Return
||estimate - reference||_2 / ||reference||_2. - poisson_
deviance - Return summed Poisson deviance, flooring estimate intensities at
epsilon. - psnr
- Return peak signal-to-noise ratio in dB for an explicit intensity range.
- relative_
l1 - Return
sum(abs(estimate - reference)) / sum(abs(reference)). - rmse
- Return the root mean squared error.
- ssim
- Return canonical single-scale SSIM using an 11×11 Gaussian window (σ=1.5).
- stats
- Calculate summary statistics for a non-empty intensity image.