Cropmat / Image to Text (OCR)

Image to Text (OCR)

Turn the text in a photo, screenshot or scan into text you can copy and edit. Works with English and Urdu.

Drop a photo or screenshot here JPG, PNG, WEBP · English and Urdu text · paste with Ctrl+V
Choose image
Language in the image

How to copy text from an image

  1. Choose the language in the image: English, Urdu, or both.
  2. Drop a photo or screenshot on the mat, or paste one with Ctrl+V.
  3. Wait a few seconds while the text is read. The first time, the reader and language data download once (about 8–10 MB).
  4. Check and correct the text, then press Copy text or Download .txt.

Getting the best results

Urdu OCR

Urdu is written in Nastaliq, which is much harder for computers to read than English because letters join and sit on a sloping line. Cropmat uses the open-source Tesseract Urdu model, which handles clear printed Urdu (books, newspapers, typed documents) reasonably well. Expect to correct some words, especially in small or decorative text.

PDFs and scans

For a scanned PDF, first turn its pages into images with PDF to JPG at High quality, then drop the images here. If the PDF already contains real text, PDF to Word is faster and more accurate.

Everything is read inside your browser; the image is never uploaded.

Questions

Can it read Urdu text from an image?

Yes. Choose Urdu or English + Urdu. Clear printed Urdu works best; check the result for mistakes.

Is my image uploaded?

No. The text reader runs inside your browser, so your image stays on your device.

Can it read handwriting?

Only neat block letters, and not reliably. It is made for printed text and screenshots.

Related tools