Troubleshooting
How to Make a Scanned PDF Searchable
Turn a stack of page images into a document you can search, copy from and index — what OCR does, and how to get good results.
An unsearchable archive is a filing cabinet. Everything is in there and finding anything requires knowing where you put it. OCR is what turns a scanned collection into something you can actually query.
Make a scan searchable
- 1Open the OCR tool and add your scanned PDF.
- 2Run it. Each page image is analysed and recognised text is written invisibly over it.
- 3Download the result and search for a word you know is in the document.
- 4Spot-check the recognition on a page with numbers or names.
What the text layer is
The scanned image stays exactly as it was. The recognised characters are placed on top, invisibly, each positioned over the word it was read from. The page looks identical; selecting text highlights in the right place; searching finds words; a screen reader can read it.
This is why OCR does not degrade a scan. It adds rather than replaces.
Getting good accuracy
| Factor | Target |
|---|---|
| Resolution | 300 DPI — below 200 accuracy falls sharply |
| Skew | Straight; even a few degrees hurts line segmentation |
| Contrast | Clean black on white; faint originals lose thin strokes |
| Colour mode | Greyscale is fine and smaller than colour |
| Text type | Printed. Handwriting is generally not recognised |
What it will get wrong
Confusions between similar shapes — 0 and O, 1 and l, 5 and S, rn and m. In prose these are visible and harmless; in a column of figures or a reference number they are invisible and consequential. Check anything numeric against the image.
Why it is worth doing on everything you archive
- You can find documents by content instead of by filename.
- Text can be copied instead of retyped.
- Screen readers can read the document, which is often a legal requirement.
- The document can be converted to Word or Markdown.
- Search indexes can crawl it — including your own desktop search.
A scan without OCR is a document you have kept but cannot use.
Do it now
OCR PDF
Extract text from scanned documents. It runs in this browser tab — your file is not uploaded anywhere.
Open OCR PDFFrequently asked questions
Does OCR change how my scan looks?+
No. The image is untouched; the text layer is invisible.
How accurate is it?+
Very high on clean 300 DPI print. Poor on faint, skewed or handwritten material.
Can OCR read handwriting?+
Generally no. It is built for printed characters.
Should I OCR before or after compressing?+
Before. Compression degrades the image the recognition depends on.
Does OCR work in other languages?+
Yes, with the right language data selected. Recognition is much worse with the wrong language set.