Optimize & secure guide
Make a Scanned PDF Searchable with OCR
Updated
Short answer: Drop the scan into OCR PDF, pick the document's language, and download a PDF with a searchable text layer behind the original page images.
A scanned PDF looks like a document, but to a computer it is a set of photos. Search finds nothing, you can't copy a sentence, screen readers are silent, and document systems can't index it. OCR fixes all of that without changing how the pages look.
Step by step
Step 1: Make sure pages are upright; OCR is far more accurate on correctly rotated text.
Open Rotate PDF(runs on your device)Step 2: Open OCR PDF and drop the scanned file.
Open OCR PDF(runs on our server)Step 3: Choose the language: English, Spanish, French, German, Italian or Portuguese.
Step 4: Download the result and press Ctrl+F (Cmd+F on a Mac) to search for a word on the page.
Step 5: If the file is large, compress it afterwards; the text layer is tiny, the images aren't.
Open Compress PDF(runs on our server)
Why this works, and what to watch for
OCR (optical character recognition) examines each page image, finds the shapes of letters, and works out the words. OCR PDF uses OCRmyPDF with the Tesseract engine and places the recognised words as invisible text exactly over the matching spot on the image. You still see the original scan, but selecting, searching and copying work on the hidden text underneath. Slightly crooked scans are straightened as part of the process.
Choosing the right language makes a real difference. It tells Tesseract which characters and accents to expect and which word lists to use, so a German scan recognised as English loses its umlauts and misreads more words.
The scan's quality sets the ceiling. 300 DPI is the sweet spot for normal-sized print; 150 DPI or lower causes noticeably more errors, especially with small type. Even lighting, flat pages and good contrast matter more than colour, so grayscale scans work fine. Handwriting, stylised fonts and text over photos are where OCR struggles.
Once a PDF has a text layer, everything else gets better: PDF to Word, PDF to Excel and PDF to Text can extract the content, your computer's search can index it, and assistive technology can read it aloud. OCR doesn't make a PDF fully accessible, which also needs tags and reading order, but it's the necessary first step.
OCR is a Pro feature because it is by far the most processor-intensive thing we do. Large scans can take a minute or two.