Image to Text (OCR)

How to use Image to Text (OCR)

1. Add one or more images containing printed or typed text.
2. Select the Language, or English + Hindi for mixed pages.
3. Press Extract text; the engine and language data download on first use.
4. Edit the result in the text box, then use Copy or Download extracted-text.txt.

Getting good results from OCR

The page loads tesseract.js, a WebAssembly build of the open-source Tesseract engine, together with trained data for the chosen language. Each image is analysed for lines and characters, and the recognised text appears in an editable box. When several images are processed, each section is headed with its file name.

Accuracy depends heavily on the source. A straight, sharp, well-lit page of printed text gives the best output; tilted phone photos, low resolution, shadows and decorative fonts reduce it. Students often use it to lift paragraphs from printed notes or newspaper clippings, but the output should always be proofread.

Limitations

• Handwriting is generally recognised poorly; the engine is designed for printed text.
• Layout is not preserved: tables, columns and formatting come out as plain lines of text.
• PDF files are not accepted here, and the first run needs internet to fetch the engine and language files.

Frequently asked questions

Can it read Hindi text?

Yes. Choose Hindi, or English + Hindi for bilingual pages. Clear printed Devanagari works best; matras can be misread on blurry images.

How can I improve accuracy?

Photograph the page flat and straight in good light, crop away margins, and try the brightness and contrast tool to make the text darker.

Is my image processed online?

No. Only the engine and language files are downloaded; the recognition itself runs in your browser.

Does it work on screenshots?

Yes, screenshots of clear text usually give very accurate results.