How to get Hindi text out of a photo

Many tools read English well and struggle with Hindi. Devanagari has joined letters, small marks above and below the line, and many look-alike shapes. Luvkaam has a Hindi setting so you do not have to retype a page by hand.

Steps

1. Open Hindi OCR. The language is already set to Hindi.
2. Choose your photo or screenshot.
3. If the page has both Hindi and English words, change the language to Hindi and English mixed.
4. Press Extract text.
5. Copy the result or download it as a .txt file.

The picture stays on your device. Only the language data is downloaded, once.

What works well

Clear printed Hindi works best: newspaper columns, book pages, forms, notices, printed letters and screenshots from the phone. Larger text works better than very small text.

What does not work well

• Handwriting.
• Very decorative or stylised fonts, like those on shop boards and posters.
• Blurry or dark photos.
• Text on a coloured or patterned background.

Why is my result full of wrong letters?

Most mistakes come from the picture, not from you. Try these in order:

1. Take the photo again in better light, holding the phone straight above the page.
2. Crop the picture so only the text is left.
3. Choose the right language. If you picked English for a Hindi page, the result will be nonsense.
4. If the page mixes both languages, use the mixed setting.

Even a good result needs a read-through. The small marks, such as the dot (bindu) and the matras, are the usual place for slips.

Scanned PDFs in Hindi

If the Hindi is inside a scanned PDF, use OCR scanned PDF and choose Hindi. It reads up to 20 pages at a time.

Saving Hindi text in a file

Paste the text into Text to Word and PDF and save it as a Word document. Choose Word, not PDF. The PDF option in that tool supports English and other Latin letters only, so Hindi will not show properly there.

Free tries

You can try it up to three times a day without registering. Registration is free and removes the limit.