How do I improve my OCR result?
If your document needs refining, Scan2x has features built-in during the configuration of a scan job for you to use to improve your OCR result. These include:
In the Scan Settings tab, as seen in the screenshot below, there are tools such as scan mode, text enhance, colour dropout, and colour enhance that can help one's OCR result.
For example, something as simple as choosing the scan mode Colour or Black & White can affect the outcome of an OCR result.
One should choose the scan mode Black & White when solely focusing on the written information required.
In other cases, one should choose colour when wanting to pick up other information besides text/numbers, such as a logo or a stamp.
For more information, please see the Scan Settings Tab in the Scan2x Handbook.
In the Post-Processing tab, there are several tools that one can use as seen in the screenshot below.
Some examples of them being used include an administrator checking the Contrast option and increasing it to enhance the black ink on the document.
The administrator can also check the Swap Color option and choose the colour of the background with the pipette icon in order to swap the colour of the background of the document to white so that the OCR will be more accurate.
For more information, please look at the Post-Processing Tab in the Scan2x Handbook.
In the Metadata tab, once the Manage OCR button is clicked, one can choose an OCR type depending on the information being gathered for a metadata field.
For example, as shown in the screenshot below, for the metadata field BoxNo, one would choose Numbers as the OCR type since that is the information required for that metadata field.
Whereas if one had a date metadata field, they could choose the Date (Default) OCR Type in order to make the OCR even more precise and accurate than before when searching for that information.
For more information, please look at the OCR Zones best practice Tab in the Scan2x Handbook.
In the Metadata tab, once the Manage OCR button is clicked, one can choose to anchor an OCR zone to a partcular word, barcode or qr code on a document.
For consistently structured documents, this is very effective as the Scan2x system will find the information needed in the OCR zone, even in scenarios where the information on the document moves up or down in the case of a bad scan.
For more information, please look at the Anchoring Tab in the Scan2x Handbook.