Decipher IDP 2.2: March 2023

Active Directory authentication through SAML 2.0 using Microsoft® Active Directory Federation Services (AD FS) is now supported, enabling users to log in to Decipher IDP using their Active Directory credentials. This enhancement is currently limited to credential authentication only.

As part of this feature, a new SsoStandardLoginEnabled parameter has been added to the Web client config file. This parameter is automatically set to false when single sign-on (SSO) is enabled, and the native login options are disabled on the login page. This ensures that users log into Decipher IDP using their Active Directory credentials. If required, the parameter can be configured by updating the Web client config file.

Learn more...

Active Directory authentication through SAML 2.0 using Microsoft® Active Directory Federation Services (AD FS) is the only identity provider (IdP) verified and supported for this release. However, this functionality is not limited to AD FS and could be applied to other SAML IdP providers.

Signatures can now be detected via a new Signature format in document form definitions. During data verification, signature fields can be recognized as Empty, Signed, or Not Signed depending on the RFT (region fullness threshold) of the field. The RFT miscellaneous parameter can be updated to change the threshold required to recognize a field as signed. Learn more...

When restarting batches, users can now choose to restart a batch at the Input or Export stage of the Decipher IDP process. Learn more...

A new DisableTableCapture miscellaneous parameter has been added, which determines if the Capture client processes data from tables. The parameter is unselected by default, meaning that table data will be captured by the client. Learn more...

Region selection during data verification has been improved for overlapping fields, making it easier for users to select and resize elements if they overlap with other fields.

Regions selected during data verification are now highlighted with an animated effect to help users locate the region in the document.

Detection of text regions which are merged, or not all vertical, has been improved during data verification.

Table data captured during data verification is now shown with a fixed table header, ensuring users can easily identify rows if the table spans beyond the page length.

The functionality assigned to the F4 keyboard shortcut during data verification has been changed, and now selects the next soft validation, low confidence, or invalid field. The existing functionality, which only selected the next invalid field, has been moved to the shortcut Windows + F4.

To improve security, the Decipher IDP web client now requires secure cookies by default.

A new English (India) locale has been added, and can be selected when configuring batch types and document types. The locale provides support for the processing of Indian documents with the handling of Indian characters and currency.

To support improvements made in this release, Decipher IDP now supports the use of the Tesseract 5 OCR engine.

The number of rows that can be exported as part of Accuracy and Automation Rates reports has been increased from 100 to 1,000.

Users can now configure different decimal separators, as required for specific locales and currencies. On export, Money or Decimal type fields will be exported with the chosen decimal separator. This is configured via a new DecimalSeparator setting in the export client config file, which can be set to:

• 0 – The configured locale settings are used (default).

• 1 – A period (.) separator is used.

• 2 – A comma (,) separator is used.

Learn more...

The following improvements to user management have been added for enhanced security:

• Administrators can now force users to reset the admin password after the first login. This is done by adding the adminPasswordChangeRequired setting to the Web.config file, which is disabled by default. Learn more...

• The following redundant default users have been removed to improve security: 

  • InputUser

  • OcrUser

  • VerifyUser

  • ReportUser

  If upgrading from an existing Decipher IDP installation, the users will only be removed if they have been inactive for three months.

Additional configuration options have been added to batch types to improve the detection of blank pages. Users can now define the left, right, top, and bottom page margins to be ignored by the blank page detection algorithm, this is useful in situations where large page borders are visible after scanning. In addition, users can also set a percentage threshold of black content, above which the page will not be classified as blank. Learn more...

New options, Automatically skip Class Verify and Automatically skip Data Verify, have been added to batch type configuration. Previously, these options could only be defined at system level. This gives users more control to skip verification for batches with high confidence. Learn more...

Vector content can now be extracted from PDF documents containing both scanned and vector data. Vector content is from digitally created PDFs, and differs from content that originates from a scanned PDF file. Previously, vector data would not be extracted if the document also contained scanned content, and the whole document would be processed using optical character recognition (OCR). Now, such documents are processed with partial OCR, meaning vector data is extracted, and the remaining scanned content is captured with OCR.

If it is detected that scanned text areas are more than 125% larger than vector text areas, the whole document will be processed using OCR, as per the previous functionality. This feature can be disabled by updating the image processing client config file to set the VectorPdfOcrMode parameter to 0. Learn more...

In addition, this feature addresses an issue where incorrect values were captured during data verification for scanned documents.

A new miscellaneous parameter has been added, RegexMode, which determines how regular expression formulas match and search fields. This functionality gives more flexibility around how regex can be used for validation and/or filtering. The parameter can be set to:

• 0 – Direct match. Matches the entire contents of the fields, and can be used for validation.

• 1 – Regular search. Matches strings, and sub strings, without character replacement, and can be used for validation and filtration. This is the default behavior, even if the parameter is not defined.

• 2 – Fuzzy search. Matches strings, and sub strings, with character replacement, and can be used for fuzzy validation and filtration.

Learn more...

Log files for the Decipher IDP installers are now retained after the installation has completed, and are stored in the local temp folder. Previously, log files could not be accessed after the installation wizard was closed.

During document form definition (DFD) configuration, the Sample Headers column has been renamed Keywords to be more intuitive to users.

Users can now add or remove items from validation lists during data verification, enabling them to update validation lists without requiring intervention from a database administrator.

To support this functionality, a new AppendableListDirect miscellaneous parameter has been added, which when enabled will automatically add new items to the validation list when the field is populated. Learn more...

Training data is no longer created for pages where no data has been extracted. Where previously training data was created for every page in a batch, it is now only generated for pages containing at least one field of extracted data.

The export client has been updated to improve the handling of failed exports. Now, if an error occurs during export, the data which has already been exported is deleted to prevent partially exported data sets.

Documents containing QR codes are now processed correctly by Decipher IDP. Previously, table data was not captured from documents with QR codes.

A new strict mode table capture mechanism has been added, improving the handling of complex tables during data verification. The new mode can be manually enabled using a new miscellaneous parameter, StrictMode. Previously, the existing table capture model was limited for documents containing multiple tables and inconsistent layouts. The new mode can learn a large number of table line item layouts, and is less likely to capture extra line items which are not required. Decipher IDP now automatically compares the results achieved by the existing mode and the new strict mode. If the new mode returns better results three times in a row, it is automatically enabled for the template. Learn more...

Decipher IDP is now supported on Microsoft Edge. Learn more...

The handling of how values are captured during data verification when using keywords (previously sample headers) configured in the document form definition (DFD) has been improved. Previously, fields were not captured completely as values preceding the defined keyword were removed. Now, the entire field is captured.

An issue has been resolved where text on a black background was not detected during data verification. Now, text on black backgrounds is captured correctly.

Configuration changes to DFDs with specific versions are now saved correctly. Previously, the changes were not retained when making updates which involved switching multiple times between the master DFD and the specific version.

A loading message has been added to the edit document type dialog and the Capture Models page in the Admin Panel. Previously, no message displayed and the pages appeared unresponsive while the data was loaded, which may take several minutes.

An issue has been resolved where the first row of data was not captured for new multi-line regions. Now, regions with multi-line fields are captured in their entirety.

Decipher IDP no longer presents exceptions when classification model data is deleted or removed from the user's temp folder. Previously, batches failed with exceptions if classification model data was deleted.

An issue has been resolved where the existing version of the Decipher Server Plugin was not automatically uninstalled when upgrading to the latest version, and had to be manually removed.

Caching issues no longer impact the display of document form definitions, document types, and batch types across multiple users following an upgrade. Previously, the cache needed to be manually cleared after upgrading for data to be displayed correctly.

The Decipher Licensing Service is now completely removed when uninstalled. Previously, if the component could not connect to RabbitMQ, it would remain in a hung state and could not be uninstalled without manual intervention.

In addition, the Decipher Licensing Service now has a default RabbitMQ connection timeout of 5 seconds. This value can be updated by updating the RabbitMqConnectionTimeoutInSeconds attribute in the Licensing Service config file. Learn more...

An issue has been corrected where certain fields weren't captured on the document's second processing, even after being assigned and trained when first processed. Now, such fields are successfully captured after they have been successfully assigned when first processed.

An issue has been resolved during data verification where Decipher IDP captured and returned data from another document in the batch. Now, only data from the currently loaded document is captured.

The error messages displayed to users when there is no valid license found in the SS&C | Blue Prism database, or when Decipher Server is unable to connect to the SS&C | Blue Prism database, have been improved. Previously, generic error messages displayed and did not inform users to the cause of the issue.

An issue has been corrected where training data loaded into a Decipher IDP system which had been upgraded from 1.2 to 2.1 did not function as expected. Now, training data imported into upgraded systems works correctly.

Documents marked as exceptions during data verification are now consistently exported as part of the batch. Previously, if a batch was returned to the system before it was reloaded and submitted, exception documents were incorrectly deleted when the batch was exported.

If all documents in a batch are marked as exceptions, the batch is now also set as an exception and retained. Previously, the batch was incorrectly deleted. In addition, a Bad request error no longer displays if a batch only contains one document that is then set as an exception.

An issue has been fixed when pages did not reload when setting documents as exceptions during data verification. This was caused by the system sending multiple requests at the same time. The page now reloads successfully after a document is set as an exception.