KOEN

Image to Text (OCR)

IMAGE · TO · TEXT

Extract text from screenshots and photos in English or Korean. See recognized areas as boxes, fix mistakes, then copy or save as TXT. Images never leave your browser.

NO IMAGE
HOW TO

How to use

  1. Drop an image, click to choose one, or paste a screenshot with Ctrl+V (⌘+V).
  2. Pick the language: English, or Korean + English for mixed text. The language data downloads once and is cached.
  3. Choose a layout if needed (for example Single line for signs or titles), then press Extract text.
  4. Check the boxes on the image and fix any mistakes in the result, then copy it or save it as a TXT file.
PAIRS WELL WITH

Pairs well with

FAQ

FAQ

Are my images uploaded?

No. Text recognition runs entirely in your browser with Tesseract.js. Only the language data is downloaded once from this site and cached.

Why is some text wrong?

OCR works best on sharp printed text and screenshots. Handwriting, tilted or blurry photos, tiny text and busy backgrounds cause errors. Try a larger, straighter image or turn on 2× upscaling.

What do the box colors mean?

Green means the engine is confident, yellow means it's unsure and red means you should double-check that part. Hover over a box to see the text and confidence score.

How do I fix broken line breaks?

Press Join lines to merge line breaks inside paragraphs while keeping blank lines between paragraphs.