Paper forms don't disappear just because everything else went digital. A client hands you a printed intake sheet, a school sends home a permission slip, a government office gives you a photocopy. The usual workaround is to fill it out by hand and scan it back, which produces a document nobody can search, copy from, or reuse. There is a better path: photograph it once, let OCR read it, and fill it as a document.
To fill a paper form from a photo, photograph each page in good light, upload the image, and let OCR convert the picture into readable text. Filly AI runs OCR on image uploads in your browser, detects the blank lines and boxes, and places your saved answers on them so the result prints as a clean, completed form.
Why does a photo of a form need OCR before it can be filled?
A photograph is pixels. To a computer, the words "Date of Birth" printed on the page are indistinguishable from a smudge — there is no text to read and therefore no way to work out what the blank line beside them is asking for. Optical character recognition converts those pixels into actual characters, which is what makes automatic field detection possible at all. Skip OCR and any tool is reduced to letting you drag text boxes around by hand.
How should I photograph a form so it actually reads?
OCR accuracy is decided at capture time, not afterwards. A few things matter far more than the tool:
- Flat and square. Shoot from directly above. Perspective skew is the single biggest cause of garbled output.
- Even light, no shadow. Your own shadow across the page eats whole lines. Daylight near a window beats overhead light.
- Fill the frame. One page per photo, edge to edge — resolution spent on your desk is resolution not spent on the text.
- Avoid glare. Glossy paper under a lamp produces a white blowout exactly where the fields usually are.
If your phone has a document-scan mode, use it — it deskews and flattens automatically, which is doing half the work for you.
How do I turn several photos into one form?
A four-page form photographed as four separate images is four documents as far as any tool is concerned. Combine them into a single PDF in page order first, then treat that PDF as the form. Filly's photos-of-documents-into-one-PDF tool does this in the browser without an account, and the resulting file keeps the pages together through filling, signing, and export.
Can a photographed form be filled automatically, or only by hand?
Automatically, once the text is readable. After OCR, the labels on the page become searchable text, and the blank lines and boxes beside them can be matched to what they're asking for — name, address, date, signature. Your saved profile details then land on those positions. The workflow is the same one described in auto-filling a PDF form online; the photo step just gets a paper original into that pipeline.
Expect to review, not to trust blindly. OCR on a photograph is meaningfully less accurate than on a clean digital file, and a misread label can put the right answer next to the wrong question. Check the placements before exporting — this is a minute of work, not a rewrite.
When should I look for the official digital version instead?
If the paper in your hand is a standard government form, the issuing agency almost certainly publishes a digital original, and starting from that is always better than starting from a photocopy of a photocopy. A form like the SSA Form SS-5 for a Social Security card is bundled in Filly's library at its official layout, so there is nothing to photograph. Reserve the photo route for forms that genuinely only exist on paper: a landlord's own application, a clinic's in-house intake sheet, a club membership form.
What do I do with the finished document?
Export it as a PDF and it prints exactly as filled, or send it for signature as a link so the other party completes their part without printing anything. Either way, the form now exists as a reusable digital document — the next time that same in-house intake sheet comes around, it fills from a saved profile instead of being photographed again.
Frequently asked questions
Does the OCR happen on my device or on a server?
Filly runs OCR in your browser on the uploaded image, so the recognition step happens on your device before anything is analysed further.
What image formats can I upload?
Standard photo formats — JPEG and PNG among them — as well as PDFs and Word documents. A phone photo straight from the camera roll works.
Will handwriting on the form be recognised?
OCR is built for printed type. Handwritten entries are recognised unreliably at best, so treat handwriting on a photographed form as something to re-enter rather than something to import.
Is a filled photographed form as good as the original?
For most everyday uses, yes — the export is a normal PDF that prints cleanly. Where an authority requires the official form on its own stationery, download the official version and fill that instead.
How much does this cost?
The PDF tools are free and need no account. Form filling itself is free up to 10 forms, 5 clients, and 10 fills a month, with paid plans from $19/month.