Document Scanner & Cleanup

Scan and Clean Up Documents Online — Free, in Your Browser

A phone photo of a page rarely looks like a real scan. One corner is darker, the lighting falls off across the paper, the background is grey instead of white, and a little sensor grain sits over every line. This document scanner fixes those problems with classic pixel algorithms that run entirely on your own device. You drop in the photo, choose color, grayscale or black and white, and the tool evens out the lighting, lifts the contrast and removes speckle before you download the cleaned page.

There is no server in the middle. The image is decoded into a canvas in your browser tab, the shadow, contrast, denoise and threshold steps are applied locally, and the finished file is generated on your machine — the document is never uploaded, which matters for contracts, receipts, medical papers and anything else you would rather not send away.

Key features

• Remove page shadows and uneven lighting by dividing out the blurred background
• Auto contrast: whiter paper, darker and more readable text
• Three outputs — color, grayscale, or crisp black and white
• Median despeckle clears dust, grain and salt-and-pepper noise
• Runs fully in your browser; no upload, no account, no watermark

How to use

1. Drop in a photo of a document or an existing flat scan (JPG, PNG, WebP).
2. Pick the output: color, grayscale or black and white.
3. Leave shadow removal and auto contrast on, and choose a despeckle level if needed.
4. Press Clean up document and download the result.

What this document scanner fixes

The tool targets the specific defects that make a photographed page look amateur. Each one is a separate, well-understood image-processing operation, so you can switch them on and off and see exactly what is changing.

• Shadows and vignetting — a dark corner or a gradient across the page caused by a lamp or by the camera not being perfectly level.
• Grey or coloured paper — the white of the page pulled toward grey or a warm tint, fixed by stretching the tonal range.
• Low contrast text — faint grey letters that become dark and legible after the contrast pass.
• Speckle and grain — dust spots, sensor noise and the salt-and-pepper dots common in low-light photos.
• Flat, muddy scans — existing scans that look washed out and benefit from the same contrast balancing.

Color, grayscale, or black and white — which to choose

The three output modes serve different documents, and picking the right one makes the difference between a page that looks natural and one that looks over-processed.

Color keeps the original hues and is the right choice for anything where color carries information — a form with colored fields, a letterhead, a receipt with stamps, a magazine page, or a child’s worksheet. Grayscale strips the color while keeping smooth shades, which is the most reliable default for plain typed pages and gives a clean, professional look without the harshness of a hard threshold. Black and white converts every pixel to either pure black or pure white using an automatic threshold, producing the sharpest, smallest file and the best basis for fax-style archiving and OCR.

• Color — forms, receipts, letterheads, highlighted notes, anything with meaningful color.
• Grayscale — typed pages, notes and book text; smooth, natural, and the safest default.
• Black and white — pure two-tone pages for archiving, OCR and the smallest file size.
• Avoid black and white on photos or pages with faint pencil, which the hard threshold can drop entirely.

How shadow removal and binarization actually work

Shadow removal is a background-normalization trick. The tool builds a heavily blurred version of the page that keeps only the slow-changing lighting — the shadow and the falloff — and loses the sharp text. It then divides the original by that blurred background, so a pixel that was dark only because the background was dark there is lifted back up, while genuinely dark ink stays dark. The blur is computed with a summed-area table, which returns the average of any rectangle in constant time and keeps the step fast even on a large, high-resolution page.

The black-and-white mode uses Otsu’s method, which inspects the histogram of the grayscale image and chooses the single threshold that separates the pixels into two groups with the greatest separation between them. Paper pixels fall on one side and become white; ink pixels fall on the other and become black. Because the threshold is chosen per image rather than fixed, it adapts to a faint photocopy or a bold print without manual tuning.

How to photograph a document so it scans cleanly

The cleanup can rescue a lot, but a well-captured starting photo always looks better. A few seconds of setup remove the defects the software has to fight.

1. Put the page on a flat, dark-free surface and light it evenly from two sides if you can, so no single lamp throws a shadow.
2. Hold the camera straight above the page, not at an angle, so the text is not skewed and the lighting stays even.
3. Turn off the flash to avoid a bright hotspot and glare on glossy paper.
4. Fill the frame with the page, tap to focus, and keep your hands steady to avoid blur.
5. Run the scanner, then crop the border if the surface around the page is still visible.

After cleanup: turn pages into a PDF and share

A single cleaned page is often enough, but a multi-page document is easier to send as one PDF. After cleaning each page, run the images through the image to PDF converter, which packs them in order and, like this tool, works entirely in your browser. If the cleaned file is still large for email, the image compressor can shrink it without a visible loss, and you can convert between formats with the image converter.

Related tools and guides

• Image to PDF — combine cleaned pages into one PDF in order
• Compress image — shrink the cleaned page before sharing
• Grayscale image — convert a color page to smooth black and white
• Crop image — trim the border around the page
• Convert image — change between JPG, PNG and WebP

Frequently Asked Questions