OCR and scanned documents: turning pictures of words back into words
By Nasir Uddin Shamim ยท Founder, PrintxPDF
Short answer: OCR converts the pixels of a scanned page into machine-readable characters and stores them as an invisible text layer aligned with the image. The page still looks identical, but it becomes searchable, selectable and copyable. Accuracy runs above 98 percent on clean 300 DPI print and far lower on handwriting.
A scanned page is a photograph. Your computer sees a grid of grey pixels where you see a sentence, which is why Ctrl+F finds nothing and copy-paste returns an empty clipboard. Optical character recognition reads those pixels and works out which letters they represent, then writes the result back into the file as an invisible text layer sitting exactly on top of the image. These guides cover running OCR on a PDF, pulling text out of photos and screenshots, testing whether a file is searchable, and rescuing bad scans.
Every guide in this topic
- How to OCR a scanned PDF so you can search it, Add a searchable text layer to a scanned PDF with OCR. How the invisible layer works, which language packs to pick, and what accuracy you should expect.
- How to extract text from an image or photo, Copy text out of a photo or screenshot on any device. Live Text on Mac and iPhone, Snipping Tool on Windows, Google Lens, and what to do with a whole folder.
- How to tell if a PDF is searchable, and make it so, Three quick tests that tell you whether a PDF contains real text or just pictures of text, what causes each failure, and how to fix an image-only document.
- How to get better OCR results from bad scans, Fix the input and OCR fixes itself. Scan at 300 DPI, deskew, raise contrast, split columns, pick one language, and know where handwriting recognition stops.