OCR-Processing a Document

Optical character recognition (OCR) is the process of extracting the textual content of a scanned image or PDF file and converting it into something that can be processed by a computer.

Whenever you receive a PDF document in Continia Document Capture, the document is automatically OCR-processed, which enables it to be imported into Document Capture for registration and further processing. For the actual OCR-processing, Document Capture uses technology provided by the official Continia partner ABBYY, one of the world's leading providers of OCR and document-scanning services. The actual OCR method depends on your environment, as outlined below.

The technology behind

ABBYY provides the technology that's used to OCR-process incoming documents in Document Capture, but the method depends on whether you have an online or an on-premises deployment:

If you’re using Microsoft Dynamics 365 Business Central online, your default OCR method is Continia Cloud OCR. The Cloud OCR uses ABBYY technology to OCR-process incoming documents by calling the ABBYY Cloud OCR SDK.

If you’re using Microsoft Dynamics NAV/Business Central on-premises, you can choose to use either on-premises OCR or Continia Cloud OCR. If you choose on-premises OCR, the process is carried out by the Continia OCR service, which consists of the Document Capture service (downloading emails and monitoring for incoming files for OCR-processing) and the ABBYY FineReader Engine (carrying out the actual OCR-processing of imported documents).

For more information on Continia Cloud OCR and how to set it up, see Configuring Cloud OCR.

For details on on-premises OCR and the Continia OCR service, as well as information on how to set these up, see Configuring On-Premises OCR and Installing ABBYY and Document Capture Services. Relevant minimum requirements can be found here and here.

The overall process

For every PDF document that's imported into Document Capture, the ABBYY engine will attempt to capture all characters and their positions. Using this OCR data, Document Capture can then locate all words and phrases in the document, which enables it to search for captions and their corresponding values. In order to be able to display the scanned PDF document optimally in the user interface, along with all identified captions and values, Document Capture converts it into a TIFF file, as this is more manageable and easier to render. However, you can always retrieve the original PDF document via the action bar by selecting Document > PDF File.

For each identified caption, Document Capture will initially look for the value to the right of the caption, and if no value is found there, it will search for it immediately below the caption. If Document Capture fails to locate a value, or if it finds the wrong one, you can help it by identifying the correct value manually. Once a value has been found – whether by Document Capture or by you – Document Capture registers the relative position of the caption and the value, and this is then used to identify the values of this caption in all future documents sent by the same vendor. The position of the caption itself is unimportant; only the relative position between the caption and the value is used.

See also

The ABBYY website
Configuring Cloud OCR
Configuring On-Premises OCR
Installing ABBYY and Document Capture Services
Minimum Requirements for Using Continia Document Capture On-Premises
Working with Paper and PDF Documents