What is invoice OCR?
Invoice OCR is optical character recognition applied to invoices: it reads the characters on a PDF, scan or photo and turns them into machine-readable text.
Invoice OCR converts an invoice document into text a computer can work with. On its own it does not know which number is the GST amount, so it is normally paired with invoice data extraction, which maps that text into named fields. InvoiceHaven, an AI-powered invoice-to-data platform for New Zealand businesses, runs OCR first and extraction second, then validates the numbers before you export to Excel.
How invoice OCR works
- The document is received as a PDF, JPG or PNG, including scans and phone photos.
- Each page is analysed and the characters on it are recognised as text.
- The recognised text keeps its position on the page, which helps identify labels and their values.
- That text is handed to an extraction step that assigns values to invoice fields.
The full InvoiceHaven sequence is Upload → Extract → Review → Export, described in detail on how InvoiceHaven works.
What information can be extracted from an invoice
- Supplier name.
- Invoice number.
- Invoice date and due date.
- Currency, including NZD.
- Subtotal, GST amount and total.
- Line items with their descriptions and amounts.
The field list and the checks applied to it are covered on invoice data extraction.
Invoice OCR versus manual invoice data entry
Manual entry
A person reads each invoice and types the values into a spreadsheet. The work scales linearly with the number of invoices and every value depends on the typist's attention.
OCR-based processing
The document is read automatically and the values are proposed for you. Your job changes from typing to checking, and arithmetic checks run on every invoice rather than only on the ones someone happens to re-add.
Invoice OCR versus structured invoice data extraction
- OCR answers: what characters are on this page?
- Extraction answers: which of those characters are the supplier, the invoice number, the GST amount and the total?
- Validation answers: do the subtotal, GST and total actually add up, and has this invoice been seen before?
More on the reading step itself on the invoice OCR page.
Why human review can still be useful
OCR can misread a character, and a misread digit changes a total. InvoiceHaven recalculates the relationship between subtotal, GST and total, reports a missing GST amount as not applicable rather than as a pass, and flags possible duplicates. Every value stays editable on the Review screen, and what you export is what you approved.
Invoice OCR and Excel
Once the fields are reviewed, InvoiceHaven exports the selected invoices as a multi-sheet Excel (.xlsx) workbook with header fields and line items. See invoice to Excel and PDF invoice to Excel.
Frequently asked questions
What is invoice OCR?
Invoice OCR is optical character recognition applied to invoices. It reads the characters on a PDF, scan or photo of an invoice and turns them into machine-readable text that software can process.
Is invoice OCR the same as invoice data extraction?
No. OCR produces text. Invoice data extraction is the following step that assigns that text to named fields such as supplier, invoice number, invoice date, subtotal, GST, total and line items.
Does invoice OCR work on photographed invoices?
Yes. Photos taken with a phone camera go through the same OCR and extraction workflow as uploaded PDF or image files, though a sharp, well-lit and flat image reads more cleanly.
Can invoice OCR output go straight into Excel?
Raw OCR text is not spreadsheet-ready on its own. InvoiceHaven runs extraction and validation after OCR, then exports the reviewed fields and line items as an Excel (.xlsx) workbook.
Is human review still needed after invoice OCR?
Review is still useful. OCR can misread characters, so InvoiceHaven checks GST and total arithmetic, flags possible duplicates, and lets you correct any value on the Review screen before exporting.
Try InvoiceHaven
Create an account and process your first 10 pages free — no card required.