Scan a contract on most copiers and you get a picture of it. It looks like the contract, but as far as your computer is concerned it’s a photograph: you can’t search it for a name, copy a paragraph out of it or find it later by typing a word that’s on page six. Optical character recognition (OCR) is what turns that picture into text.
That sounds like a small thing. For an office that scans a lot, it changes how documents get found, filed and used.
What OCR is
OCR is software that looks at an image of a page, works out which shapes are letters and numbers, and records them as text. The usual result is a searchable PDF: it still looks exactly like the scanned page, but behind the image there’s a layer of text your computer can search and copy.
OCR can run in several places:
- On the copier or scanner, as part of the scan. Many business devices can save a searchable PDF, sometimes as standard and sometimes as an option.
- In the scanning software on a computer, which is common with dedicated document scanners.
- In a document management system, which reads documents as they arrive, whether they came from a scanner, an email or a shared folder.
How it works, briefly
The software first cleans up the image: straightens a crooked scan, removes specks and sharpens the contrast between text and background. It then finds the blocks of text on the page, breaks them into lines, words and characters, and decides what each character is. Modern engines use trained models rather than simple pattern-matching, so they cope better with unusual fonts, low-quality scans and, increasingly, neat handwriting.
The quality of the scan matters more than most people expect. A clean page scanned at the resolution your software recommends will come out far better than a crumpled receipt photographed under a desk lamp.
Where it saves an office time
Finding documents
This is the benefit everyone notices first. Once your scanned files contain text, the search box in Windows, SharePoint or your document system can find them by anything that’s written on them: a client name, an invoice number, a property address. Nobody has to remember what a file was called or which folder it went into.
Using what’s in them
Text can be copied. Someone who needs a clause from a scanned agreement, or the figures from a scanned statement, can copy them instead of typing them out again.
Filing them
Combined with document management, OCR lets documents file themselves. The system reads the invoice number, vendor name and date from the page, uses them to name and index the file, and puts it in the right place. That’s the step that replaces most of the manual filing in an office.
Feeding other systems
For documents with a predictable layout, such as invoices or order forms, the recognized data can be passed on to an accounting system or another business application, so nobody keys it in by hand. This is where the real time savings tend to be, because typing data from paper into a screen is slow and error-prone.
OCR on its own, or as part of something bigger
Plain OCR reads text. It doesn’t know that one number on the page is an invoice total and another is a phone number. Intelligent document processing (IDP) adds that understanding: it works out what kind of document it’s looking at, picks out the fields that matter, checks them, and sends them where they need to go.
| OCR | Intelligent document processing | |
|---|---|---|
| What it does | Turns an image into searchable text | Reads, classifies and extracts the data that matters |
| Good for | Making archives and scans searchable | Invoices, forms and other documents in volume |
| Where it runs | Copier, scanner or software | Usually a document management or automation platform |
| Human checking | Spot checks | Exceptions and high-value fields |
If what you need is to find old documents, OCR on your scans may be all you need. If your staff spend hours a week keying invoices or forms into a system, IDP is the thing to look at. Our post on intelligent document processing goes into it further.
Typing data from paper into a screen is slow and error-prone.
Where OCR falls short
It’s worth being clear about the limits:
- It makes mistakes. A smudged 8 can become a 3, and a stylized font can confuse it. For figures that matter, such as invoice totals or account numbers, keep a human check.
- Handwriting is harder. Neat printing is often read well; cursive and rushed notes much less so.
- Bad scans give bad results. Faded thermal receipts, shadows, creases and low resolution all reduce accuracy.
- Search only finds what was recognized. If a word was misread, a search for it won’t find the document.
None of these is a reason to avoid OCR. They’re reasons to start with clean, typed documents and to check the numbers that matter.
A sensible way to start
- Pick one process that hurts. Accounts payable is the usual choice: lots of documents, similar layouts and a lot of typing.
- Count it. How many documents a month, how long each one takes, how often mistakes happen.
- Check what you already have. Your copier or scanner may already be able to produce searchable PDFs.
- Decide where the documents should end up. A shared folder, SharePoint, your accounting system or a document management system.
- Keep a human check on high-value fields while people get used to it.
- Measure again after a few months, then move to the next process.
Keeping digital documents safe
Scanned documents are only as secure as the place they’re kept. Limit who can open which folders, keep backups, and make sure the copier or scanner itself is set up securely, since it handles every page you scan. The NIST Cybersecurity Framework is a reasonable baseline for thinking about the rest. If your documents hold sensitive client information, decide who can see what before you start scanning, not after.
How we can help
We supply Ricoh fi Series and HP ScanJet Enterprise Flow document scanners and set up the copiers and scanners you already have so scans arrive as searchable files in the right place. When you’re ready to go further, we implement DocuWare document management, which captures documents from scanners, email and shared folders, reads them on the way in and routes them for approval. Most offices start with invoices. Call (386) 252-2292 and show us how one document moves through your office today. For the wider picture, see what business process automation is.