Contrast-limited Adaptive Histogram Equalization (CLAHE)¶

Back to MIB | User interface | Ribbon | Image

CLAHE enhances contrast locally in small rectangular tiles rather than globally. Each tile's histogram is redistributed to match the chosen Distribution, then neighboring tiles are blended with bilinear interpolation to avoid sharp boundaries. A clip limit caps contrast amplification in uniform areas to suppress noise.

See MATLAB's adapthisteq for details.

Dataset type: scope of the operation — Shown slice (2D), Current stack (3D), or Complete volume (4D).

Color channel: All, Displayed, or a specific channel.

Number of tiles Y / Number of tiles X: tile grid size (1–256 in each direction); more tiles = finer local adaptation.

Clip limit: contrast enhancement limit (0–1); higher values give stronger contrast but more noise amplification.

Number of bins: histogram bins used for the contrast transform (2–65536); more bins = greater dynamic range.

Distribution: target histogram shape — uniform, rayleigh, or exponential.

Alpha: distribution shape parameter for rayleigh and exponential (disabled for uniform).

Use Preview to check the result on the current slice before applying to the full dataset.

Back to MIB | User interface | Ribbon | Image