A scanned page or a screenshot looks like text, but to a computer it is just a picture. OCR (optical character recognition) reads the letters in the picture and turns them into real text that you can search, copy and edit. This guide shows three ways to use OCR on Tryyourselfs, including for Hindi and other Indian languages.
Which tool should I use?
- A photo or screenshot with text: use Image to Text. You get the text on screen and as a TXT file.
- A scanned PDF you want to keep as a PDF, but searchable: use OCR PDF. The pages look the same, but you can now search and copy their text.
- A PDF where you only want the text: use Extract Text from PDF. If the PDF is a scan, OCR is applied automatically.
- A scanned PDF you want to edit in Word: use PDF to Word.
Step by step: image to text
- Open Image to Text.
- Upload a JPG, PNG, GIF or WebP picture.
- Choose the language of the text. This matters a lot: choosing Hindi + English for a Hindi document is far more accurate than leaving English.
- Press the button, then copy the text or download the TXT file.
Which languages are supported?
English, Hindi, Bengali, Tamil, Telugu, Marathi, Gujarati, Kannada, Malayalam, Punjabi, Urdu, Spanish, French, German, Arabic, Portuguese, Italian, Russian, Japanese and Chinese (simplified). Indian languages are combined with English because most documents also contain English words, numbers and brand names.
How to get accurate results
OCR quality depends mostly on the picture.
- Use a sharp, well-lit image with dark text on a light background.
- Keep the page straight. For a tilted scan in OCR PDF, turn on "Straighten tilted pages".
- Use enough resolution: small text needs roughly 300 dpi when scanning.
- Crop away borders, hands and shadows before uploading.
- Avoid heavy compression, which blurs letters. If a photo is dark or low-contrast, improve it first with Enhance Image, then run OCR.
What OCR cannot do well
- Handwriting, especially cursive.
- Very small print, decorative fonts and text over busy pictures.
- Tables and columns: the text may come out in the wrong order. For these, try the "keep the page layout" option in Extract Text from PDF. Always check names, amounts, dates and phone numbers before you rely on the result. One wrong digit can matter.
Privacy
Uploaded files are scanned for malware, processed on our server and deleted automatically after about an hour.