Symptoms

Some fields in the document are not being extracted correctly during the recognition process. This means that the characters within these fields are either being extracted incorrectly or are missing entirely. As a result, the extracted values do not accurately reflect the content of the document.

Cause

The FlexiCapture 12 extracts data from the text layer of a document by default. Such text layer may contain characters that cannot be accurately interpreted by FlexiCapture 12. These characters may include symbols, special characters, or non-standard fonts.

Resolution

In order to resolve the issue and extract values from the page image instead of the text layer, the Processing Mode should be set to OCR only instead of Auto or Prefer PDF text layer. This change should be made either on the batch type settings if a custom batch type is being used, or on the project level if the Default batch type is being used.

To change the processing mode on the batch type level, please follow these steps:

1. Click Project > Batch types > Select used batch type > Edit…
2. Select the Image Processing tab
3. Check the OCR only mode
   

To modify the processing mode for the entire project (default batch type), please follow these steps:

1. Click Project > Project Properties...
2. Select the Image Processing tab
3. Check the OCR only mode