About Image to Text Converter (OCR)
This image to text converter is free, needs no sign-up and runs in your browser, so your pictures are never uploaded. Drop, paste or photograph a page, a screenshot or a sign, press Read text, and the recognised text appears next to a confidence score for every word. Copy it or save it as a TXT file.
It uses Tesseract, the open-source OCR engine, running through Tesseract.js in a web worker so the page stays responsive. You can tick several languages at once, for example Bangla and English on the same page. The list includes English, Bangla, Hindi, Arabic, Urdu, Spanish, French, German, Portuguese, Russian, Turkish, Chinese, Japanese and Korean.
Something is downloaded, and the tool tells you what. Your picture stays on your device, but the OCR engine files and the data for each language you pick come from jsDelivr, English in the background when the page opens and other languages when you pick them, a few megabytes per language, and your browser may cache them. So you need a connection for that first run.
OCR works best on printed text that is sharp, straight and well lit. It is not reliable for handwriting, which Tesseract reads poorly, and it can confuse similar shapes such as 0 and O. Words with a low confidence are highlighted so you know where to look first.
How to use Image to Text Converter (OCR)
- Add a picture: drop it on the box, tap to choose a file, paste it, or press Take a photo on a phone. PDFs are read page by page, up to 10 pages.
- Tick the languages that appear in the picture. Pick only the ones you need, since each one is downloaded.
- If the picture is tilted, press Rotate 90 degrees. For a faint or noisy scan turn on Grayscale, and try Black and white to remove the background.
- For small text, such as a screenshot, choose Enlarge 2x.
- Press Read text. The card says when the engine is ready; English starts loading when the page opens.
- Check the amber and red words, then press Copy or Download TXT.
Features
- Tesseract OCR in a web worker, in your browser.
- Several languages at once, including Bangla, Hindi and Arabic.
- Input by file, drop, paste or camera, plus PDF pages.
- Grayscale, black and white threshold (Otsu) and rotation to improve a bad scan.
- Enlarge 2x for small text such as screenshots.
- Per-word confidence, with uncertain words highlighted.
- Copy the text or download a TXT file.
- Right-to-left text such as Arabic and Urdu shown in the right direction.
Worked examples
Three words with confidences 95, 88 and 52 (the last one read as $1,2O4.50) gives Average confidence 78%. Tones: Total is high, due: is high, $1,2O4.50 is low.
High is 85 and above, medium 60 to 84, low under 60. The low word is the one to check: it holds a letter O where a zero belongs.
A page that is 60% dark text pixels at level 40 and 40% paper at level 220 gives The black and white threshold is set automatically at level 40, so text turns black and paper turns white.
The tool uses Otsu’s method to pick the cut-off from the picture.
Languages you can read
| Code | Language |
|---|---|
| eng | English |
| ben | Bangla |
| hin | Hindi |
| ara | Arabic |
| urd | Urdu |
| spa | Spanish |
| fra | French |
| deu | German |
| por | Portuguese |
| rus | Russian |
| tur | Turkish |
| chi_sim | Chinese (Simplified) |
| jpn | Japanese |
| kor | Korean |
Good to know
Handwriting
Tesseract is built for printed text. It does poorly on handwriting, cursive and decorative fonts, so this tool does not promise to read handwritten notes. A very neat block-letter hand may partly work. Use the highlighted uncertain words as a guide, and expect to retype.
What gets downloaded
Tesseract needs its engine and a trained data file for each language. They are not built into the page. The first time you read text they are fetched from jsDelivr, a public file host, and your browser may keep them for later. Your picture is never part of those requests.
PDFs
A PDF is drawn page by page in your browser with pdf.js and each page is read as a picture, up to 10 pages. A PDF that already has selectable text does not need OCR; copy the text from the PDF directly.
Privacy: how it works
- Pictures and PDFs are read in your browser and are never uploaded to Ulkaa or to anyone else.
- The OCR engine and the language data are downloaded from jsDelivr (English as the page opens, other languages when you pick them). Those requests do not include your picture.
Tips
- Photograph the page straight on, in even light, with the text filling most of the frame.
- Choose only the languages in the picture. Extra languages slow the read and can add wrong words.
- If the result is poor, try Grayscale, then Black and white, then Enlarge 2x, one at a time.
- Check numbers, names and codes by eye. OCR often swaps 0 and O, 1 and l, 5 and S.
Frequently asked questions
Is it really free?
Yes. There is no price, no trial and no credit card. Ulkaa Tool does the work with your own device, so there is no server bill to pass on to you.
Do I need an account?
No. There is no sign-up and no login. Open the page and use it. An Ulkaa account is not needed for anything on this tool.
Does it run in my browser?
Yes. Your pictures are processed by code running in your browser tab, on your own device.
Is my data uploaded?
No. Your pictures are never sent to Ulkaa or to anyone else. You can check this yourself by watching the Network tab in your browser developer tools while you use the tool.
Is anything downloaded?
Yes. The OCR engine and the data for the languages you choose are downloaded from jsDelivr, English as the page opens and others when you choose them. Your picture is never uploaded.
Which languages does it support?
English, Bangla, Hindi, Arabic, Urdu, Spanish, French, German, Portuguese, Russian, Turkish, Chinese (Simplified), Japanese and Korean. You can tick several at once.
Can it read handwriting?
Not reliably. Tesseract is made for printed text and does poorly on handwriting.
How do I read a PDF?
Choose the PDF like any picture. Its first 10 pages are drawn in your browser and read one after another.
What do the highlighted words mean?
Amber words had a confidence from 60 to 84 percent and red words under 60 percent. They are the words most likely to be wrong.
Why is my result full of mistakes?
Blurry, tilted, low-contrast or small text is hard to read. Rotate the picture, try Grayscale or Black and white, or enlarge it, and make sure the right languages are ticked.
Can I use it on my phone?
Yes. Press Take a photo to use the camera, or choose a picture from your gallery. A long PDF is slow on a phone.
Popular Image to Text Converter (OCR) pages
- Extract Text From a ScreenshotPaste a screenshot and copy the text out of it.
- Bangla OCR OnlineExtract Bangla text from a picture or scanned page.
- Hindi OCR OnlineExtract Hindi text from a photo, scan or screenshot.
- Picture to Text ConverterTake a photo or choose a picture and convert it to text.
- OCR OnlineRun OCR on a scanned page or a PDF and copy the text.