OCR – how information enters the system

Thursday, 25. March 2021 | 0 Comments

Users of GetMyInvoices know: Typing out invoice information, such as vendor and invoice amount, is a matter of the past. These are displayed automatically when the document is retrieved. But how is this possible? The OCR solution fintract is mainly responsible for this function. But let’s take it one step at a time.

Abbildung OCR-Technologie

What is OCR?

OCR stands for Optical Character Recognition and refers to automated text recognition. With the assistance of this technology, users can primarily convert images into text or extract text from images. It does not necessarily have to be installed on the end device, and users do not need their own OCR scanner. Such processes are often used in the background. They unfold their full potential when the information is not only read by machine, but also processed automatically. This is exactly the case with GetMyInvoices.

Challenges for fintract

Firstly, fintract supports the two standard formats for invoices – ZUGFeRD and XRechnung. Secondly, unstructured documents pass through the OCR tool. The challenges here are the different languages as well as the different document types. These range from fuel receipts to supplier invoices to outgoing invoices, with information located in different places, often in different spellings. There are also account and credit card statements. In addition to that, the documents are available in various formats – as PDF, TIFF, JPEG or PNG. fintract nevertheless recognizes relevant information based on optical character recognition. This is because our software uses an algorithm comprising approximately 100,000 rules, combined with artificial intelligence. As a result, recognition rates of more than 90 percent are achieved and fintract can reliably extract up to 50 different invoice information from invoice documents. 

  • Invoice number
  • Invoice date
  • Invoice items
  • Invoice currency
  • Net amount
  • Tax rate
  • Gross amount
  • Supplier with address
  • Payment method
  • Tax number
  • VAT number
  • Recipient with address
  • No error-prone typing of data
  • Documents can be processed within a very short time.

A OCR tool at the service of its users

If fintract should ever fail to find all the information, users can simply highlight the content in the invoice preview and extract it with one click. This way, our OCR learns and upcoming results become better. This means even less effort for users. 

The benefits:
  • Without processing, most documents and their information can be automatically transferred to other systems for direct processing.
  • No error-prone typing of data
  • Documents can be processed within a very short time.

For instance, invoices can be posted and paid within minutes.  Financial tractions are recorded within seconds and automatically assigned to the receipts.

Besides, fintract makes a significant contribution to combating the current pandemic. This is because OCR reads information from Corona tests and makes it available for further machine processing as required. But that is another story.

Try GetMyInvoices
now for free for 14 days

See the benefits of automated invoice management:
Better overview. Less accounting work. More time for your ideas.

Test for free

Leave a Reply