How to Extract Text from a Receipt or Invoice

OCR guide · 6 min read

A shoebox of receipts and a folder of PDF invoices are two of the most stubborn kinds of paperwork to deal with. The information you actually need — a total, a date, a vendor name, a line item — is trapped inside an image. You can see it, but you cannot copy it, search it, or drop it into a spreadsheet without retyping every character by hand.

Optical character recognition (OCR) solves exactly this problem. It reads the text in a photo or scan and hands you back editable, selectable characters. This guide walks through turning a receipt or invoice into usable text, with tips aimed specifically at the quirks that make these documents hard to read.

Why receipts and invoices are so hard to capture

Receipts and invoices fight back in ways ordinary documents do not:

• Thermal print fades. Most cash-register receipts are printed on heat-sensitive paper. Within months the ink browns and the contrast drops, so the numbers you care about are the first thing to disappear.
• Paper curls and creases. A receipt that has lived in a wallet is rarely flat. Curl and folds bend the text lines, which confuses recognition.
• Fonts are tiny and dense. Line items, SKUs, and tax breakdowns are often printed in small, condensed type to fit narrow paper.
• Photos are uneven. A quick phone snap adds glare, shadow, and skew that a clean flatbed scan would not.

OCR can handle all of these, but the cleaner the input, the fewer corrections you make afterward. A little care when you capture the image pays off later.

The basic workflow: image to text in four steps

Whether you are reconciling expenses, keeping books, or filing a reimbursement, the shape of the task is the same. Capture the image, extract the text, pull out the parts you need, and export.

1. Capture the receipt or invoice

You have two common starting points. For a paper receipt, photograph it or scan it. For an invoice that already lives on your screen — a PDF, an email, an online order confirmation — a screenshot is faster than printing and re-scanning.

With Textquill you can grab an on-screen invoice directly by selecting the area with Alt+Shift+S and dragging a box around it. For a paper receipt, take the photo first, then upload, drag, or paste it into the extension.

2. Extract the text

Once the image is in hand, run OCR on it. If you saved a photo of the receipt as a file, right-click the image and choose Extract text, or drop the file into Textquill. The extension reads the characters and returns them as plain, selectable text you can edit.

3. Copy the totals and line items

Now the useful part. From the extracted text you can select just the numbers and rows you need — the subtotal, tax, grand total, invoice number, or the individual line items — and copy them. This is far quicker than transcribing a long itemized receipt by hand, and it removes the typos that creep in when you retype figures.

4. Export to a spreadsheet or expense tool

When you want the whole thing in a structured form, export rather than copy. Textquill can save results as TXT or Markdown, both of which paste cleanly into a spreadsheet, a bookkeeping app, or an expense report. A Markdown table, in particular, drops neatly into tools that understand it, while plain TXT works everywhere.

Accuracy tips for receipts specifically

Receipts reward a few seconds of setup. These habits noticeably improve what OCR returns:

• Flatten the paper. Press a curled receipt flat under a book or against a table edge before you photograph it. Straight text lines read far better than bowed ones.
• Use even, diffuse light. Avoid a single harsh lamp or direct sun, which cause glare and hard shadows. Soft, even light across the whole receipt is ideal.
• Shoot straight on. Hold the camera parallel to the paper, not at an angle. Skew and perspective distortion are common causes of misread characters.
• Crop to the receipt. Trim away the table, your hand, and background clutter so the recognizer only sees the document. The area-select shortcut makes this easy for on-screen items.
• Reach for Accurate mode on faint thermal print. Textquill offers Fast and Accurate modes with an Auto option. For crisp, high-contrast invoices, Fast is fine; for faded thermal receipts, Accurate mode spends a little more time and usually recovers more of the text.
• Pick the right language. If your receipt is not in English, set the matching language from the 16 supported so numbers, accented characters, and currency symbols are interpreted correctly.

Getting the numbers right

On a receipt the numbers matter more than anything, and they are also where OCR is most likely to trip. A few checks save you grief:

• Confirm the totals against the paper. Glance at the subtotal, tax, and grand total in the extracted text and compare them to the receipt before you trust them.
• Watch 0 versus O and 1 versus l. In a monetary figure it should be a zero, not a letter O; these look nearly identical in some fonts.
• Check decimal points and separators. A misplaced decimal turns 12.50 into 1250. Confirm the position of the point, and note whether the receipt uses a comma or a period as the decimal mark.
• Re-shoot rather than guess. If a digit is genuinely unreadable in the original, a better photo beats a hopeful guess.

Why on-device OCR matters for financial documents

Receipts and invoices are not neutral scraps of paper. They routinely carry your name, a billing address, the last four digits of a card, an account number, or a vendor relationship you would rather not broadcast. Uploading them to a web OCR service means handing that information to a third-party server you do not control.

Textquill runs OCR on your own device. The image is processed locally, nothing is uploaded, and it works offline — so the sensitive details on a receipt stay with you. Extracted results are kept in a searchable local history, which also means you can find that one invoice from three months ago without it ever leaving your machine.

Digitizing your receipts this way turns a pile of fading paper into text you can search, edit, and file. How you categorize or report those figures for expenses, bookkeeping, or reimbursement is up to you and your own process — OCR simply gets the numbers out of the image and into a form you can work with.

FAQ