What OCR is
OCR, or optical character recognition, is software that finds letters in a picture and turns them into text you can copy, search and edit. It is what lets you pull a quote from a photographed page, an error message from a screenshot or the details from a receipt.
How to get the best results
- Light it well. Even daylight with no shadows across the page.
- Shoot straight on. Tilted pages distort letters.
- Use high resolution. Small, blurry text gives errors.
- Crop to the text. Remove pictures, borders and background.
- Use contrast. Dark text on a plain light background is easiest.
Step by step
- Open the image to text tool for photos and scans, or the screenshot to text tool for screen captures.
- Add your picture.
- Wait while the text is recognised in your browser.
- Read through the result and fix mistakes.
- Copy the text.
Always proofread
No OCR is perfect. Characters that look alike are often confused, such as the letter O and the number 0, a lowercase l and the number 1, or the letters r and n together and the letter m. Look closely at names, numbers, codes and email addresses.
What about handwriting?
Handwriting is far harder than print, because every person writes differently. You can try the handwriting to text tool and you may get a useful first draft, but expect to correct many words. Neat block letters in dark ink on plain paper give the best chance.
Privacy
Many online OCR services upload your picture to their servers. Lightbench reads the text in your browser, so a private letter, ID or invoice stays on your device.
Which tool should I use?
Use image to text for photographs of printed pages, books, signs and documents. Use screenshot to text for captures from your screen, which are cleaner and usually read more accurately. Use handwriting to text only for handwritten notes, and treat the output as a rough draft.
Handy uses
- Copying error messages to search for a fix.
- Turning a printed page into an editable document.
- Pulling the details from a receipt or business card.
- Quoting from a book photographed in a library.
Frequently asked questions
Why does OCR make mistakes?
Blur, shadows, unusual fonts, low resolution and tilted pages all make letters harder to recognise.
Can it read text in other languages?
Recognition quality depends on the language and the script. Clear printed text in common languages works best.
Can OCR read a table?
It reads the text, but the table structure may not survive. Check the layout and fix it by hand.
Is it safe for private documents?
The text is read in your browser and the picture is not uploaded.