Normalize Layers¶

Normalizes image intensities slice-by-slice across Z or time to reduce illumination variation.

Demonstration

Target: what to normalize:

Z stack: normalize each Z-slice to a common mean and standard deviation.
Time series: normalize each time frame; controlled by Time series normalization — either Based on current 2D slice or Based on complete 3D stack.
Masked area: compute normalization coefficients only from the masked region (see Mask layer).
Background: shift each slice so that the mean intensity of the masked background area matches the dataset mean.

Mode:

Automatic: coefficients (mean and std) are computed from the data.
Manual: use the fixed target Mean and Std values.
BasedOnSlice: use the slice specified by Reference slice No as the normalization reference.

Color channel: channels to normalize (All channels, Shown channels, or a specific channel).

Exclude: pixels to exclude from mean/std calculations — Whole range, Exclude blacks, or Exclude whites.

Mask layer (Masked area / Background only): selection or mask layer used to define the region.

Normalization algorithm

Calculate mean intensity and standard deviation (std) for the whole dataset.
Calculate mean and std for each slice.
Shift each slice by the difference between its mean and the dataset mean; stretch by the ratio of dataset std to slice std.

For 4D datasets, normalization can be done across time.
For Z-stacks, black or white pixels can be excluded from the statistics.