How to recognize text in a scanned PDF using OCR

You receive a PDF with a scanned contract, certificate, or form. You try to select some text and nothing happens — the cursor behaves as if the page is a picture. That is exactly what it is: a scanned image with no text layer underneath. OCR (optical character recognition) adds that layer and turns the image into a working document where text can be selected, copied, and searched.

When you need OCR

OCR is useful when you have a scanned PDF and want to:

• copy a passage without retyping it by hand;
• search for a word in the document with Ctrl+F;
• extract content into another tool without manual entry;
• make an archived document searchable by keyword.

If the PDF already has a text layer — meaning you can select words — OCR is not needed.

What affects recognition quality

The quality of the output is almost entirely determined by the source scan. The key factors:

• Resolution: good OCR needs at least 150–200 dpi, ideally 300 dpi. Phone photos are often sufficient if the page is flat and the lighting is even.
• Alignment: the page should be straight. Heavily tilted text recognizes poorly.
• Lighting and contrast: shadows from the spine, yellowed paper, or faded ink all reduce accuracy.
• Language: the algorithm needs to know which language the text is in to correctly interpret letters and words.

How to run OCR, step by step

1. Open ocr-pdf and upload your scanned PDF. 2. Select the document language if the tool prompts for it. 3. Wait for processing — multi-page scans take longer. 4. Download the result. 5. Open the file and verify: can you select text? Are key details — names, dates, numbers — recognized correctly?

What can go wrong

• Many errors in names and numbers. The scan quality is too low. Try rephotographing or rescanning at a higher resolution.
• Text is not recognized at all. The PDF may contain vector graphics that look like a scan but technically are not. OCR cannot help with that.
• The file got much bigger. OCR adds a text layer and slightly increases size. If the result is too heavy to send, compress it with compress-pdf.
• Handwriting is not readable. Standard OCR targets printed text. Handwritten signatures and margin notes recognize poorly or not at all.

What to check after OCR

• Select a few lines and paste them into a text editor — confirm the text came through correctly.
• Search for key words with Ctrl+F: a name, an organization, a date.
• Compare numbers and dates in the recognized text against the original — those are the most common source of errors.

To make a scanned PDF searchable and selectable, upload it to ocr-pdf. For long documents, split the file with split-pdf first, process the parts separately, and combine the results with merge-pdf.