Make a scanned PDF searchable with OCR

When your scan has no text layer

Open the PDF and try to select a single word. If you cannot, and Ctrl+F finds nothing, the page is only an image: the scanner took a photograph, not text. It happens most with signed contracts, older invoices and anything that arrives through an office copier or a phone camera. If selecting does work, there is a text layer in there and the PDF to text tool gets you done faster. Only when that layer is missing does the text have to be read out of the picture first — that is what this tool is for.

Searchable PDF or plain text

Once recognition finishes, the text appears on screen: copy it or save it as .txt, handy when you need to lift a passage into a quote or a case file. Choose the searchable PDF and the scan keeps exactly the look it had, with the text laid over it as an invisible layer. Ctrl+F works, selecting works, and the signed copy itself is left untouched. That layer covers the page rather than sitting millimetre-perfect on every word: searching and copying are fine, pointing at a word with the mouse is not. The file stays as heavy as the scan; too heavy for email, and you can send it as a download link.

A sharp scan makes for reliable text

The result depends on what goes in. A page scanned straight at 300 dpi reads almost flawlessly; a phone photo taken at an angle, with a shadow across half the sheet, produces the most errors. Every page is scaled down before recognition, so small footnotes and stamps are the first to go. Columns end up one after another and tables come back as text aligned with spaces, so check amounts and reference numbers yourself before you reuse them. Each page travels to the recognition step as an image and is not kept afterwards; no copy of your contract stays behind.