How to Extract Text From an Image (OCR)
A photo of a page isn't text - it's a picture of text. Here's how OCR turns it into something you can actually copy and search.
A photo of a printed page, a screenshot with text you want to reuse, a picture of a whiteboard - none of these are actually text, they're images that happen to contain text. OCR (optical character recognition) is what turns the picture into real, copyable, searchable text.
Extracting text from an image
- Upload the image to an OCR tool.
- The tool analyzes the image and identifies the text it contains.
- Review the extracted text - straightforward printed text is usually accurate; complex layouts may need light manual correction.
- Copy or download the extracted text to use elsewhere.
What OCR handles well
- Clear, well-lit photos of printed text - book pages, printed documents, signs
- Screenshots, which tend to produce very accurate results since there's no camera angle or lighting to account for
- Simple, single-column layouts without much visual clutter around the text
What OCR struggles with
- Handwriting, which standard OCR isn't designed to read reliably
- Blurry, poorly lit, or heavily angled photos
- Stylized or decorative fonts that don't resemble standard printed text
- Complex multi-column layouts, where the tool may read across columns instead of down one at a time
Getting better results from a photo
If you're photographing a document specifically to run OCR on it, a few small habits make a real difference: shoot it straight-on rather than at an angle, make sure the lighting is even without harsh shadows across the text, and get close enough that the text is sharp rather than relying on zoom, which tends to blur fine detail.
Frequently asked questions
Can OCR read handwriting?
Standard OCR is built for printed text and struggles significantly with handwriting - results on handwritten notes are usually unreliable compared to typed or printed text.
Does the photo need to be perfectly straight and well-lit?
It doesn't need to be perfect, but a clear, well-lit, reasonably straight-on photo produces noticeably more accurate results than a blurry, angled, or poorly-lit one.
Can I extract text from a screenshot instead of a photo?
Yes - a screenshot is generally easier for OCR to read accurately than a photo, since it doesn't have the lighting, angle, or focus issues a camera photo can introduce.