What is OCR?

OCR — Optical Character Recognition — is the technology that turns a picture of text into actual, machine-readable text. It's what makes a scanned document searchable, selectable, and editable.

Why scans need OCR

When you scan a page, you get an image — a grid of pixels. To a computer there are no letters in it, just light and dark dots. That's why you can't select text in a scanned PDF, or convert it to Word, until OCR has recognised the characters.

How OCR works

1. Pre-processing — the image is cleaned up: straightened, contrast-boosted, converted to black-and-white.
2. Layout analysis — the engine finds blocks of text, lines, and words.
3. Recognition — each character shape is matched to a letter, increasingly using neural networks (like the LSTM models in modern engines) that read whole lines in context.
4. Output — recognised text, often with confidence scores and position data.

What affects accuracy

• Scan quality — 300 DPI, high contrast, straight pages read best.
• Clean typefaces beat handwriting and decorative fonts.
• Language and layout — the right language model and simple columns help.

OCR and privacy

OCR can run entirely on your device — the document image never has to be uploaded. FreshPDF runs OCR in your browser with a local recognition model, so your file stays private. You can try it from the Tools menu in the editor.