Online Batch Image Histogram Equalization Tool
Equalize luminance histograms for up to 30 images in your browser. Preview each result, then download separate PNG files. No strength controls.
Start using
Online Batch Image Histogram Equalization Tool
Batch histogram equalization accepts up to 30 selected images and has no adjustable processing controls. Generate applies the same luminance-CDF method independently to every loaded image; loading alone does not alter the previews.
For each image, the browser builds a 256-bin histogram from rounded weighted luminance Y=0.299R+0.587G+0.114B. The cumulative distribution maps the first occupied luminance to 0 and the last to 255. The mapped-minus-original luminance delta is added equally to R, G and B, then each channel is clamped to 0-255.
Alpha and image dimensions are preserved. A constant-luminance image is returned unchanged instead of being forced to black. Because equalization uses one luminance distribution per image, two files in the same batch do not share a histogram or influence each other's result.
Verified examples: a two-pixel grayscale image containing 50 and 150 maps to 0 and 255 while retaining alpha values 40 and 80. A one-pixel RGB(100,150,200) image remains unchanged because it has only one occupied luminance level.
Worked use: a low-contrast scan with most tones clustered in a narrow range can gain separation after CDF mapping. Review colored material carefully because adding the same luminance delta preserves channel differences only until a channel clips at 0 or 255; this is not color-managed Lab processing.
Each input must have an image MIME type, be at most 25 MB and decode to no more than 24,000,000 pixels. The selected batch is capped at 100 MB and 40,000,000 decoded pixels; selecting more than 30 files rejects the whole batch.
Every generated result is encoded as PNG and named from the source base name plus -equalize.png. Download all starts separate downloads about 180 ms apart; it does not create a ZIP, retain the source format or preserve metadata. Processing is local in the browser and is not AI enhancement, local adaptive equalization, CLAHE, denoising or detail recovery.
Histogram bins use rounded Y, including pixels whose alpha is zero. If every pixel falls in the same rounded bin, the decoded image is unchanged even when its RGB values differ. Transparent background RGB can therefore affect visible results. Each Generate starts from the loaded source, not the last result.
Per image: ≤25 MiB and ≤24,000,000 pixels. Per batch: ≤100 MiB and ≤40,000,000 pixels. Selecting over 30 files or any invalid image rejects the whole batch.
Processing takes place locally in your browser.