Normalization and Standardization¶
Min-max normalization rescales intensities to [0, 1]:
\[f_{\text{norm}}(x,y) = \frac{f(x,y) - f_{\min}}{f_{\max} - f_{\min}}\]
Z-score standardization removes the mean and scales to unit variance:
\[f_{\text{std}}(x,y) = \frac{f(x,y) - \mu}{\sigma}\]
Both operations are essential when combining images from different acquisition sessions or instruments.
See also
Image Transformations — the general point-wise intensity transform these two operations are special cases of
Contrast Enhancement — adjusting contrast rather than range
Image Preprocessing — every other preprocessing operation