Recognize text in images and PDFs

Make a scan or a photo searchable by its text, or turn it into a note you can edit. Recognition runs entirely on your device.

On this page
Documentation sections

A scanned letter, a photo of a receipt or a screenshot hold text that search cannot see. Filarr can read that text for you, in two ways:

  • Make searchable stores the text with the file, so the search box finds it;
  • Recognize text creates a note with the text, which you can edit and copy.

Recognition works with PNG, JPG, WebP, BMP, GIF, TIFF and AVIF images, and with PDF files.

Everything happens on your device. The file is neither sent nor decrypted anywhere else.

Make a file searchable

  1. Right-click the image or the PDF, then choose Make searchable.
  2. Under Text language, choose French, English or Both.
  3. Click Make searchable.

Filarr stores the text on the file record, encrypted like the rest. No note is created and the file itself does not change.

A PDF that already contains text is read instantly. A scanned page, which is only a picture, takes longer: Filarr shows the progress page by page, and Stop interrupts it.

From then on, the search box finds the file by the words it contains.

If the file changes, right-click it and choose Refresh searchable text to read it again.

Turn an image or a PDF into a note

  1. Right-click the file, then choose Recognize text.
  2. Under Text language, choose French, English or Both.
  3. Click Recognize text.

Filarr creates a note with one document page for each page of the file. The file itself stays where it was. Click Open the note when the reading is done.

For each page, the note shows where the text comes from: PDF text when the PDF already contained it, or scanned page with a confidence score when Filarr had to recognize it. Originals shows the original pages next to the text.

Inside a note, type / and choose Recognize text: pick an image or a PDF from your folders, and its pages are added to the note you are writing. See Images and files in notes.

Good to know

  • Filarr cannot take a picture or drive a scanner. Scan your pages with your usual tool, then import the files.
  • Recognition is set up for French and English text.