Transformation Designer

A new locator method is available to process handwritten checks. This locator replaces the A2iA Zone Locator configured to use the "CheckReader" functionality.

A separate volume license is required to use this locator.

A new recognition engine is available so that the Advanced Zone Locator can recognize handwritten text. This new profile is available for several languages and can extract various types of field data. Some of the field types have specific settings to further refine extraction.

Additional image cleanup settings enable users to clean images to improve recognition results.

Classification online learning has been augmented to include separation online learning when a project is configured for trainable document separation. All labels that reference classification online learning now include separation as well as classification online learning.

When a document is manually separated or classified during production, that document is collected for document separation and classification online learning.

A new Timeout property is available in the MpsPageRecogProfileOmniPage and can be set via script. The Timeout value uses milliseconds for its unit.

Private Sub Application_InitializeScript()
Dim profile As MpsPageRecogProfileOmniPage
Set profile = Project.RecogProfiles.ItemByName("OmniPage")
Set timeout to 5 minutes
profile.Timeout = 300000
End Sub

Google normally uses service account keys and tokens for authentication. However, for Kofax Clarity using Google Vision, authentication is handled by Kofax licensing services. This is true for both Google Vision on-premise as well as Google Vision in the cloud.

Google Vision in the cloud requires no further configuration. However, Google Vision on-premise requires some additional configuration. For on-premise, use your own Google project that contains the service account key. Once the key is available to Kofax licensing, the client uses this information to retrieve the token.

You can use Google Vision on-premise through Kofax Clarity by following these steps:

1. Download the Google Service Account key for your Google project.

   This is a JSON file.

2. Save the JSON file in a local folder.

   For example, C:\Temp\Google Vision.

3. Create a registry key under HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Kofax\Mobile\AccessTokenClient.

   1. New string value: "ServiceAccountCredentialPath" and provide the path to the JSON file saved in the previous step.

   2. Optionally, configure debug log output by adding an additional string value under this registry key called "DebugLogFile" and provide a path for the log file as the value data.

      For example: C:\Temp\Google Vision\AccessTokenClient.log

4. If you are using Google Vision on-premise, it is necessary for you to achieve the recognition processing speed that you require without relying on any Kofax software. This is because the performance of the on-premise Google environment is outside of the control of Kofax.

Handling of large training sets is improved.