CrispUtils

Scan to PDF — a document scanner with nothing to install

Photograph a contract, a form or a receipt, drag the corners to straighten it, and download a clean multi-page PDF. It runs in your browser, so the document never leaves your phone.

Click to add photos of your document One or more images · nothing is uploaded
Your file is processed by your own browser. Nothing is uploaded, stored or logged.

The usual advice for scanning a document with a phone is to install an app. Most of those apps want an account, sync your scans to their cloud by default, put a watermark on the free tier, or all three — and the documents people scan are exactly the ones they should be careful with: contracts, ID, medical letters, receipts for expense claims.

This does the same job in a web page. Take the photos with your normal camera, open them here, drag the four corners onto the edges of the paper, and the page is geometrically straightened — the projective correction that turns a photo taken at an angle into a flat rectangle. A contrast filter then turns grey phone-camera output into something that reads like a real scan.

Add as many pages as you need, reorder or rotate them, and export a single PDF at A4, US Letter or the natural size of the photo. There is no OCR, deliberately: a usable text-recognition engine is a download of well over ten megabytes, and that cost is not worth it for a page that should load instantly.

How to use this tool

  1. Photograph each page. Lay the document flat, get even light and include all four corners — you do not need to line the shot up perfectly.
  2. Open the photos here, or tap Take a photo on a phone to use the camera directly.
  3. Drag the four round handles onto the corners of the paper. The area outside the shape is dimmed so you can see what will be cropped away.
  4. Pick a filter. High contrast suits printed text, greyscale suits photos and stamps, colour keeps everything as it was. Adjust brightness and contrast if the result looks washed out or too harsh.
  5. Add more pages, rotate or delete any that are wrong, and check the order in the thumbnail strip.
  6. Choose A4, Letter or fit-to-image, and download the PDF.

Worked example

A signed contract to email back

You print a two-page contract, sign it, and photograph both pages on the kitchen table at a slight angle. Here you drag the corners onto the paper on each page, choose high contrast, and export an A4 PDF. The result looks like it came off a flatbed rather than a phone, and the signed contract never touched anyone's server.

Receipts for an expense claim

Six receipts, photographed as you collected them. You add all six, straighten each one, use greyscale so the card terminal print stays readable, and export one PDF to attach to the claim.

Common use cases

  • Signed contracts and forms that have to go back by email
  • Receipts and invoices for expense claims or bookkeeping
  • Proof-of-address documents for a bank or landlord
  • Handwritten notes and whiteboards kept as a single file
  • School or medical forms returned as a PDF
  • Old letters and certificates being digitised for the record

Common mistakes

  • Shooting in your own shadow — Side lighting from a window beats overhead light with you standing over the page. A hard shadow across the paper survives every filter.
  • Cutting off a corner — The straightening needs all four corners of the paper visible. If one is outside the frame, the geometry cannot be worked out — retake the photo further back.
  • Using high contrast on everything — It is excellent for black printed text and terrible for photographs, coloured stamps and faint pencil. Switch to greyscale or colour for those pages.
  • Exporting one PDF per page — Add every page here first and export once. A reviewer receiving eight separate files will ask you to combine them anyway.

Frequently asked questions

Do I need to install anything?

No. It is a web page. Open it in your phone or desktop browser and use it. Once loaded it also works with no internet connection.

Is my document uploaded?

No. The photos are read by your browser, straightened and filtered in your device's memory, and assembled into a PDF locally. Nothing is transmitted, which is the whole reason to use this rather than an app that syncs to a cloud account.

Why is there no automatic edge detection?

Reliable automatic detection needs a computer-vision library that weighs around eight megabytes. Loading that on every visit would make the page slow for everyone, including people who just want to straighten one receipt. Dragging four handles takes a couple of seconds and is more accurate on awkward backgrounds.

Can it recognise the text so I can search the PDF?

No. OCR would require a recognition engine plus a language model — well over ten megabytes of download. If you need searchable text, export the PDF here and run it through a dedicated OCR tool afterwards.

What resolution does it export?

Pages are exported at up to about 1700 pixels on the long edge as JPEG, which keeps printed text crisp while producing a PDF small enough to email. Fit-to-image preserves the photo's own proportions if you would rather not force A4.

Can I add pages taken at different times?

Yes. Use Add pages as many times as you like; each page keeps its own corner positions and rotation, and the filter settings are applied at export.

Last updated 2026-09-15