Getting the most out of PDFs
How to turn a scanned paper form into a fillable PDF
Open the scan in PDF Forms and choose Find fields on the scan: it measures the lines and boxes in the page image, reads the labels beside them, and places real form fields on top of the unchanged scan for you to accept, adjust and download.
Plenty of forms exist only on paper, or as a scan that has been emailed around for years. Retyping one in a word processor, lining up every box, can take an afternoon. You usually do not need to. The scan can become the form.
Why you cannot type into a scan
A scanned page is one picture. The lines, the boxes and the printed labels are all pixels. There is no field anywhere on it, so clicking a blank line does nothing.
Even recognising the text with OCR does not change that. OCR turns the printed words into text you can select; it does not create anywhere to type an answer. For that, the page needs form fields placed on top of it.
What happens when you open a scan in PDF Forms
When you open a scanned PDF in PDF Forms, the tool notices which pages are scans. The page stays exactly as scanned. Choose Find fields on the scan and, for each page, it measures the lines and boxes in the image and reads the labels beside them with text recognition.
It then suggests fields of the same kinds it finds on any other PDF: blank lines, empty boxes, check box squares, rows of letter boxes, empty table cells and rows of circles. Each suggestion is named from the label it sits next to.
The labels are read only to name the fields. The scan is not altered, and no text layer is added to it.
Step by step
- Get the form into a PDF. A scanner gives you one directly. If you only have phone photos, combine them into one PDF with Image to PDF.
- Clean it up if it is grey or speckled. Enhance Scan raises the contrast, which makes faint lines easier to find.
- Open it in PDF Forms and choose Find fields on the scan.
- Review the suggestions. The ones that look certain are ticked; the others are there to check on the page. Accept what is right and dismiss the rest.
- Place any missing fields by hand. Pick a type on the left, then click or drag on the page.
- Try the form, press Tab through it to check the order, and create the fillable PDF.
Getting a scan the detection can read
The detection is measuring lines, so it does best with lines that are clearly there:
- Straight. A page scanned at an angle gives fewer suggestions. Straighten it on the glass before scanning.
- Dark and even. Faint pencil lines, pale coloured boxes and grey photocopies all give fewer suggestions.
- Around 300 DPI. Low-resolution scans blur thin lines into the background.
- Flat and evenly lit. A phone photo of a curled page bends the lines and puts a shadow across half of them.
- Blank. Start from an empty copy of the form. A copy someone has already filled in by hand has writing sitting in the very spaces the detection is looking for.
Rows of boxes for one letter each
Many official forms ask for a code or a name in a row of touching squares, one character per box. On a scan these are recognised as a row of letter boxes and become one field that spreads the text evenly across them, with a character limit equal to the number of boxes, so each character typed lands in its own square.
What detection cannot do
It is tuned to be right rather than to find everything. A wrong suggestion costs you a click and makes you trust the rest less, so anything doubtful is left unticked or not suggested at all. Expect to place a few fields yourself.
A few things it cannot see at all:
- A label with empty space after it and no line or box. There is nothing drawn to measure. Place those fields by hand.
- Handwritten labels. Recognition is made for printed text. If a label cannot be read, the field gets a plain name you can change.
- Very faint or heavily skewed pages. If nothing clear is found, the tool says so, and you can still place every field by hand.
The finished file
The result is the original scan with real form fields on top. It opens in Acrobat, Reader, Chrome, Edge, Firefox and other PDF apps, and anyone can type into it and save it.
The printed words on the scan are still part of the image. Adding fields does not make them searchable or selectable.
For everything else about building the form, from field types to tab order and required fields, see how to make a PDF fillable.
Common questions
Can I use a photo of a form taken on my phone?
Yes. Turn the photo into a PDF with Image to PDF first. Photograph the page flat, straight on and in even light, because a curved or shadowed page gives far fewer suggestions.
Will finding the fields change how my scan looks?
No. The scan stays exactly as it is. The fields are placed on top of it.
It found nothing on my scan. Why?
Usually because the lines are too faint, the page is skewed, or the form has labels with blank space after them rather than drawn lines and boxes. Cleaning the scan with Enhance Scan often helps, and you can always place fields by hand.
Does the tool read handwriting on the form?
No. Labels are read with recognition made for printed text, and only to name the fields. Handwritten answers already on the page are left as part of the image.