Document Reader SDK 9.7

Document Reader SDK version 9.7 is available for immediate download and can be obtained from the links below: 

• Separate installation executable SDK x86 

• Separate .msi installation packages x86

• Separate installation executable SDK x64

• Separate .msi installation packages x64

It is also available for download using our Regula Downloads Manager tool.

Please review the release notes and change list since 9.6 prior to installing this release:

IMPORTANT changes

• Added new neural networks for more accurate detection and recognition of Simplified Chinese text.
• Added a new neural network for more accurate detection of glares.
• Added document occlusion detection to identify cases where parts of a document are covered by fingers, hands, or other objects during image capture. This helps prevent incomplete or obstructed document images from being accepted and notifies users to retake the photo before continuing. The feature improves document visibility, data extraction accuracy, and the overall reliability of the verification process.
• Added a capability to determine whether the document holder’s handwritten signature is present on the document.

Core SDK

OCR

• Added new neural networks for more accurate detection and recognition of Simplified Chinese text.
• Improved perforation text recognition for a wider range of fonts.

MRZ

• Fixed an issue with distinguishing between 0 and O in Ireland driving licenses.
• Updated parsers:
  • Afghanistan ID Card

Barcode

• Added new parsers:
  • Benin Consular ID Card
  • India eVisa
  • Moldova Registration Certificate
  • Sri Lanka ID Card
• Updated parsers:
  • Argentina ID Card
  • Canada Quebec Health Insurance Card
  • Dominican Republic ID Card
  • Honduras Resident Card
  • USA ID Card and DL

Text data parsing and validation

• Added personal number verification for documents from the following countries:
  • Eswatini
  • Saudi Arabia
• Updated parsing of DG11/DG12 field values into correct field types and LCID for documents from multiple countries.
• Fixed an issue with address parsing when document design elements affected recognition accuracy.
• Fixed an issue with incorrect validity status for gender X.
• Fixed an issue where the DL restriction code NONE could be returned with spaces.
• Fixed an issue with parsing a surname value that contains marital status.

Authenticity

• Added a capability to determine whether the document holder’s handwritten signature is present on the document, use processParam.authParams.propertiesParams.checkHoldersSignature to enable the Signature presence security check.
• Added a capability to use a liveness transaction ID instead of a selfie photo for face matching and search when integration with Face API is enabled, use the processParam.faceApi.livenessTransactionId parameter. This simplifies the workflow by removing the need to download the photo from storage and helps reduce network traffic when the person’s face data has already been collected. Works with Face SDK 8.3 and later.
• Added the LetterScreen++ check for Bulgarian passports.
• Updated the neural network for hologram detection check with greater accuracy for Turkey passports.
• The Security text check is now performed only when working with Document Reader devices.
• Fixed an issue resulting in an incorrect liveness status after session reprocessing when data for a specific page had not been collected on the mobile device.

Image QA

• Added a new neural network for more accurate detection of glares.
• Added document occlusion detection to identify cases where parts of a document are covered by fingers, hands, or other objects during image capture. This helps prevent incomplete or obstructed document images from being accepted and notifies users to retake the photo before continuing. The feature improves document visibility, data extraction accuracy, and the overall reliability of the verification process:
  • Occlusion check is disabled by default. To enable the occlusion check, use processParam.imageQa.occlusionCheck parameter.
  • Added the real-time processing message to notify users about document occlusion. The occlusion check is disabled in processParam.imageQa.expectedPass by default.
  • The occlusion check result is written to Image Quality Results → Occlusion.

Enums

• Added new eVisualFieldType enum members:
  • ft_DLClassCode_B3_From
  • ft_DLClassCode_B3_To
  • ft_DLClassCode_B3_Notes
• Added new eRPRM_Authenticity enum member:
  • RPRM_Authenticity_Property
• Added new eRPRM_SecurityFeatureType enum member:
  • SecurityFeatureType_SignaturePresence
• Added new eCheckDiagnose enum members:
  • chd_Property_NoSignature
  • chd_Property_TextAsSignature
  • chd_Property_FingerprintAsSignature

Other

• Memory usage optimization.
• Code quality has been improved with minor issues fixed.

Desktop API

• Fixed an issue where attempting to reprocess a saved session on the Regula 7017D.110 resulted in the RPRM_Error_NotAvailable error.
• Fixed an issue with document processing when using the SeparateOpticalReading and VideoDetectionMode parameters.
• Updated document presence detection logic in video detection mode: moving or repositioning a document on the reader glass no longer triggers a false document detection event, and the event is returned only after the document has been completely removed and no movement is detected above the glass.

Mobile API

• Added support for Traditional Chinese (zh-Hant) localization.
• Updated sample projects on GitHub to align with the latest changes.
• Code quality has been improved with minor issues fixed.

Android

• Fixed an issue where reinitialization was incorrect after deinitialization with no delay between them.
• Fixed an issue with camera frame size in the Capture scenario.

iOS

• Added support for mutual TLS, which requires both the server and the client device to authenticate each other when establishing a connection.
• Fixed an issue where the forced landscape orientation might reset if users locked the device orientation to portrait.

Hybrid

• Added Swift Package Manager support for Flutter.

Web Components

• Added support for Traditional Chinese (zh-Hant) localization.
• Added the documentReaderElement.settings.allDevicesSwitchToMobileButton property to allow mobile devices and tablets to switch to another device when using the Switch to Mobile feature.
• Fixed an issue where the Kurdish locale was not set correctly and fell back to English.
• Updated dependencies to the latest versions to address security vulnerabilities.

UI Components

• Added support for Traditional Chinese (zh-Hant) localization.
• Added visualization of glare and occlusion results on the document image when the corresponding checks are selected.
• Added a capability to expand text field results by clicking anywhere in the cell.
• Fixed an issue where the scrollbar was not active when scrolling in Safari.
• Updated dependencies to the latest versions to address security vulnerabilities.

Web Service API

• Added license country restriction information to the countryFilter object in the /api/healthz endpoint.
• Improved stability in outOfProc mode, where the Core runs in a separate worker process detached from the web server.

Demo Website

• Code quality has been improved with minor issues fixed.
• Updated dependencies to the latest versions to address security vulnerabilities. Known issue: the PostCSS library is affected by vulnerabilities CVE-2026-45623 and GHSA-r28c-9q8g-f849. The issue will be addressed in a patch release once an updated version of the library is available.

Licensing

• Fixed an issue where the Tenant and Environment properties were not registered if their length exceeded 128 characters. Values longer than 128 characters are truncated to 128 characters.

Installation Packages

• Windows
  • New driver version 10.5.10.26 included in Regula Reader Drivers package 1.8.34

Database

• Updated multiple document templates to include OVI, MLI, Contact chip, Holograms, Extended MRZ, Extended OCR, Photo embedding type, Patterns checks where appropriate.
• Includes 254 countries and territories, 16807 documents.
• Added 79 new documents.

For details, see the Supported documents list.