Documents
OCR
OCR (Optical Character Recognition) converts text inside images and scanned documents into real, selectable and searchable characters.
Updated 12 Aug 2026
From pixels to characters
A scan looks like a document but contains no text — only a picture of one. That is why searching a scanned PDF returns nothing. OCR analyses the image, locates lines and glyphs, classifies each shape as a character, and outputs text that can be copied, searched and indexed.
- 1Pre-processing: deskew, straighten and increase contrast
- 2Binarisation: separate ink from background
- 3Layout analysis: find columns, paragraphs, lines and tables
- 4Recognition: classify glyphs, historically by template matching, now with neural networks
- 5Post-processing: apply a dictionary and language model to fix likely misreads
What drives accuracy
| Factor | Effect |
|---|---|
| Resolution | 300 DPI is the practical minimum for reliable results |
| Contrast and lighting | Shadows and grey backgrounds cause dropped characters |
| Skew | Even a few degrees of rotation hurts line detection |
| Typeface | Clean printed serif and sans text beats decorative or handwritten script |
| Language | Recognition must be told the correct language and script |
OCR output is never guaranteed correct. Digits and lookalike characters — 0/O, 1/l, 5/S — are the usual failures, so verify anything financial or legal before relying on it.
Compress PDF
Compress, merge and split scanned documents privately in the browser.
Characteristics
- Turns images of text into machine-readable characters
- Modern engines use neural networks rather than template matching
- Can produce a searchable PDF by layering invisible text over the scan
- Accuracy is probabilistic and depends heavily on input quality
Common uses
- Making scanned archives searchable
- Extracting data from invoices, receipts and forms
- Digitising books and historical records
- Reading text from photographs for accessibility tools
Advantages
- Unlocks decades of paper archives for search
- Removes manual re-typing from data entry workflows
- Improves accessibility for screen-reader users
- Enables automated indexing and classification
Limitations
- Accuracy collapses on low-resolution or skewed scans
- Handwriting recognition remains unreliable
- Complex tables and multi-column layouts are often mangled
- Output needs human review for critical data
Examples
Printed: Invoice #10583 — Total £1,240.00
OCR output: lnvoice #IO583 — Total £1,24O.00
^ l for I ^ O for 0Frequently asked questions
What is OCR used for?
Turning scans and photographs of text into searchable, copyable characters for archives, invoices and forms.
Why can't I search my scanned PDF?
It contains images rather than text. Running OCR adds a text layer that search can index.
How accurate is OCR?
Well over 98% on clean 300 DPI printed text, and far lower on poor scans or handwriting.
Does OCR work on handwriting?
Only partially. Handwriting recognition is a harder, separate problem with much lower accuracy.
How can I improve OCR results?
Scan at 300 DPI or higher, straighten the page, raise contrast and select the correct language.
Related terms
PDF Compression
PDF compression reduces the file size of a PDF by re-encoding its images, removing redundant objects and compressing content streams, ideally without visible loss of quality.
Image Compression
Image compression reduces the byte size of a picture by removing redundant or visually unimportant data, either losslessly or by discarding detail the eye barely notices.
PDF (Portable Document Format) is a file format that preserves the exact layout, fonts and graphics of a document so it renders identically on any device.
Character Encoding
Character encoding is the mapping between the characters people read and the bytes computers store, defining how text is turned into binary and back again.
XML
XML (Extensible Markup Language) is a text format that describes data using nested, self-labelled tags, designed for documents and long-lived system-to-system messaging.
Base64
Base64 is an encoding scheme that represents binary data using 64 printable ASCII characters so it can travel safely through text-only channels.
Related tools
Related guides
How to Compress a PDF Without Wrecking the Quality
Understand what actually makes a PDF large, which compression settings matter, how much size you can realistically save, and how to compress a PDF privately in your browser.
Compress PDF Without Losing Quality
Honest guide to shrinking a PDF without wrecking it: Lossless (small safe savings), Compress images (re-encode photos, keep text selectable), and Rasterize (biggest savings, flattens pages). Pick the mode that matches your file.
Protect PDF: User Password vs Owner Password (What Permissions Actually Mean)
User (open) password vs owner (permissions) password in PDFs: what Protect PDF sets with AES encryption, how permission toggles map to allowed actions, and why owner locks are etiquette — not real secrecy.
How to Merge PDF Files Without Uploading Them
Combine several PDFs into one document: how merging works, how to control page order and orientation, what breaks bookmarks and forms, and how to do it without uploading files.