Scan Documents to PDF

How to scan a document with your camera — 3 steps

1. 1

   Allow camera access

   Grant the page permission to use your camera. The video feed is handled entirely in the browser and is neither recorded nor transmitted.

2. 2

   Capture each page

   Hold the camera square above the page and capture. Work through a multi-page document in order and the pages are collected as you go.

3. 3

   Export the PDF

   Review the captures, drop any that came out poorly, and download the finished document.

Technique makes more difference than equipment

A phone camera is capable of producing genuinely good page scans, and the difference between a clean result and an unusable one is almost entirely in how the shot is taken. Light is the biggest factor. Indirect daylight is ideal — near a window but not in direct sun. What ruins captures is uneven light: a desk lamp to one side produces a bright patch and a dark corner, and overhead lighting puts the shadow of your own head and phone across the page. If you cannot get even light, moving the page rather than the light is usually quicker.

Position is the second factor. Hold the camera parallel to the page and directly above its centre. Shooting at an angle produces keystoning, where the page comes out as a trapezoid with one edge longer than the other, and text on the far edge is smaller and softer. Put the page on a flat surface, flatten any fold or curl — a curved page from a book throws the text near the spine out of focus — and fill the frame with the page so you are not wasting resolution on the desk.

A background that contrasts with the paper helps the edges of the page be found cleanly, so white paper on a dark surface works better than white on white. Finally, hold still: the most common defect in phone captures is slight motion blur, which looks acceptable on a small screen and turns out to be the reason text recognition fails later. Brace your elbows, or rest the phone against something.

What a camera scan is good for, and what it is not

For everyday purposes a camera capture is entirely adequate and vastly more convenient than finding a scanner. Receipts for an expense claim, a signed form to email back, a page of a book for reference, a letter to file, an ID document for an application, notes from a whiteboard — in all of these the point is a legible record, and a well-taken capture provides one within seconds.

It is less suitable where fidelity or formality is required. Documents for legal or official submission are sometimes required to be scanned properly at a specified resolution. Archival copies, anything where colour accuracy matters, and material to be reproduced in print are better handled by a flatbed. Large-format documents, thick bound volumes, and glossy pages that reflect light are all awkward with a camera, and a stack of fifty sheets is genuinely faster through a feeder.

One follow-on step is worth knowing about. A camera capture produces a picture of the page, so the resulting PDF contains no searchable text — you cannot find a word in it or copy a sentence out. Running the document through OCR afterwards recognises the characters and makes it searchable, which is what turns a photographed archive into something you can actually use. That works far better when the capture was sharp, evenly lit, and square, which is the practical reason to get the technique right at the point of capture rather than hoping to fix it later.

Related tools