Image OCR

Drop a photo or scan, pick a language, and get the recognised text.

Advertisement

How to image ocr

  1. 1

    Drop an image.

  2. 2

    Pick the language.

  3. 3

    Run OCR — download the .txt file.

Understanding image ocr

Image OCR reads text out of a photo or screenshot. It is the everyday version of the technology: point it at a picture of a page, a sign, a slide or a receipt and get the words back as editable text.

The same local Tesseract engine used for scanned PDFs handles this, so the image never leaves your device — which matters when the photo is of an ID card or a bank letter.

Screenshots give near-perfect results because their text is already rendered pixel-perfect. Photographs vary with focus, lighting and perspective.

When you would use it

  • —Copying text from a screenshot where selection is impossible.
  • —Reading a receipt photo into an expense note.
  • —Capturing text from a conference slide photographed on a phone.
  • —Extracting a serial number or reference from a product photo.

Practical tips

  • Shoot straight on: perspective distortion is the single biggest cause of bad recognition.
  • Crop to the text area before running OCR — surrounding clutter creates spurious characters.
  • Even lighting beats bright lighting; hard shadows across a page confuse the engine.

Troubleshooting

Output is gibberish
Check that the selected language matches the text, and that the image is upright and in focus.
Handwriting is not recognised
Tesseract targets printed type. Handwriting recognition needs a different class of model.

Frequently asked questions

Related tools

View all tools →

Merge PDF

Combine multiple PDF files into a single document. Reorder, preview, then download.

Open tool

Split PDF

Extract a page range or split every page into its own PDF file. Fully private.

Open tool

Organize PDF

Rearrange or remove pages visually and download a clean new PDF.

Open tool