OCR is a technology utilized to recognize and extract key data fields from image files, and a productive tool for invoice automation. Invoices are generally received via pdf files, or are scanned in from a paper invoice, and manually input. This can lead to multiple errors and tedious work. By using automatic extraction of key fields, a lot of time and effort can be saved.
For AP invoices, OCR technology can be separated into two main methods for data extraction, traditional zonal OCR and intelligent word recognition. The time-tested zonal process maps invoices to templates that identify key fields based on their location on the page, while the word recognition process looks for key terms such as invoice number, date, total, etc. Word recognition is more convenient to setup, since there are no required templates that must be adjusted any time a vendor changes their invoice format, however, the results are generally not as reliable as zonal OCR.
As each method has its advantages and drawbacks, a new market of cloud-based OCR engines has emerged that combines both of these technologies to provide reliable results without the need for vendor template configuration. DynamicPoint has integrated these technologies with its Office 365 based Invoice Automation solution, EasyAP365. Test your own invoices directly from our product page: EasyAP365. Scroll down to the Frequently Asked Questions section and follow the link found in “How well does OCR work?”