What is Histogram Equalization?
Histogram equalization remaps image intensities according to their cumulative distribution. By spreading frequently occurring values, it can increase global contrast across the image.
How Histogram Equalization works
Histogram equalization derives a monotonic lookup function from the cumulative frequency of input intensity values. Applying that function redistributes occupied levels across more of the available output range, making some low-contrast structures easier to distinguish. For color images, equalizing RGB channels separately can disturb hue, so luminance-oriented processing is often preferable. It serves as an enhancement or preprocessing stage before inspection, feature extraction, or segmentation.
Key facts
- 1The mapping is global and monotonic, so pixel ordering by intensity is largely preserved, but multiple input levels may collapse to one output level when the image or numeric range is discrete.
- 2Applying equalization independently to red, green, and blue histograms can alter their ratios and create color casts; operating on a luminance or lightness component better preserves chromatic relationships.
- 3Contrast-limited adaptive histogram equalization works on local regions and clips histogram peaks, addressing uneven illumination while limiting the amplification of noise in nearly uniform areas.
When Histogram Equalization matters
Developers use it to expose detail when useful intensities occupy a narrow range. Global equalization can overemphasize noise or produce unnatural contrast in regions with different lighting.