Document Reader SDK 6.7 — February 14, 2022
- New authenticity check introduced for extended validation of OCR properties, such as size, lines and character positions, alignment, background, and others. Currently available for document reader scans only.
- New authenticity check introduced for portraits on the document. This check is verifying the number of faces present on the document and their exact position, including specific face landmarks, where applicable. Also, face absence is verified in specific places of the document where it should not be present.
- Native streamline integration of Regula Face SDK is extended for all platforms with automatic face match and search by a portrait taken from the document (visual and chip) or a live owner photo after document processing in the same request. Simplified KYC processing in just one single request.
- Added support for the Khmer (Cambodian) language.
- Added new fonts support for perforated document numbers in different countries documents.
- Updated Hebrew and Central European NNs with better recognition quality.
- Fixed the issue with recognizing Sex fields in Cyrillic when processParam.matchTextFieldMask is set to false.
- Fixed the issue that led to crash when trying to scan an ID card in the CreditCard scenario under some circumstances.
- New parsers:
- Rwanda Passport.
- Updated parsers:
- Andorra ePassport,
- Ecuador ID Card,
- Macau Work Permit,
- Morocco Registration Certificate,
- Spain Residence Permit.
- Fixed an issue with division by zero on some images without MRZ on x64.
- Added new parsers:
- Brunei Darussalam ID Card,
- Burkina Faso ID Card,
- Guatemala Driving License,
- Honduras Driving License,
- India Registration Certificate,
- Mexico Social Security Card,
- Mexico Driving License,
- Myanmar Temporary Border Crossing Permit,
- Pakistan Driving License,
- Pakistan Armed Forces Id Card,
- Pakistan Passport,
- Pakistan Registration Certificate,
- Panama ID Card,
- Russian Federation Work Permit,
- Sint Maarten ID Card.
- Updated parsers:
- Abkhazia Passport,
- Albania Driving License,
- Angola Driving License,
- Angola ID Card,
- Bahamas Residence Permit,
- Canada Driving License,
- Colombia ID Card,
- Philippines Driving License,
- South Africa ID Card,
- Suriname Passport,
- UAE Driving License.
Text data parsing and validation
- Added the capability to check static text in fields that may contain values from predefined lists, like country code, for example.
- Added validation of personal numbers for nine more different countries.
- Updated parsing of DG11/DG12 field values into correct field types and LCID for documents from multiple countries.
- Fixed the issue with postal code comparison between Viz and Barcode data.
- Fixed the issue with comparison between Viz and RFID data when it contains name reduced with point.
- Fixed the issue with transformation of Japanese dates to Gregorian calendar.
- Added a new security check for OCR. It allows extended checks of the Viz zone personalization text data on the image, including background, size, position, alignment, height, and other text parameters based on the document template. Currently works with images from document reader devices only.
- Added a new security check for portraits on the document. This check is verifying the number of faces present on the document and their exact position, including specific face landmarks, where applicable. Also face absence is verified in specific places of the document where it should not be present.
- Added a new security check for 2D printed barcode physical size (width, height in mm) based on the document template.
- Added a new security check of MLI/LEFI elements presence in axial light.
- Updated MRZ extended security check with vertical spacing of the symbols in the strings.
- Fixed the issue with Barcode format check result still available even if the check was disabled.
- Fixed the issue with IPI check where due to wrong resize detection quality suffered.
- Fixed the issue with fibers check that led to crash at some specific circumstances.
- Updated the screenshot check NN. Now image focus check results are analyzed before running the screenshot check.
- Added a possibility to make stop list check result counted in status.overallStatus result when processParam.respectStoplist is set to true.
- Added new eVisualFieldType enum members:
- Added a new eRPRM_Authenticity enum member RPRM_Authenticity_Ocr.
- Added new eCheckDiagnose enum members:
- Added new eRPRM_SecurityFeatureType enum members:
- Fixed the issue with serialization of Authenticity checks results into XML for RPRM_Authenticity_Mrz and RPRM_Authenticity_Photo_Embed_Type.
- Fixed the issue with false CAN detections on the image.
- Fixed the issue with face landmarks after results rotation.
- Fixed the issue with processParams.documentIdList parameter behavior for child documents.
- Fixed the issue with processing ICC base color profiles in a PDF file.
- Fixed the issue with incorrect image buffer size check when parsing a PDF file in some cases.
- Fixed the issue with incorrect rotation and flip of the images when parsing a PDF file in some cases.
- Code quality improved with lots of small issues fixed.
- Documentation updated.
Core RFID SDK
- Added a new eRFID_Commands enum member RFID_Command_Set_Default_Params.
- Fixed the issue with determining data group file size led to incorrect behavior on some chips.
- Fixed the issue with not setting necessary parameters when reprocessing captured RFID session on the server, that lead to incorrect results for some documents (China).
- Fixed the issue with the Gender field incorrectly serialized into JSON from eDL DG1.
- Code quality improved.
- Documentation updated.
- Fixed the issue with face detection results after rotation and cropping.
- Fixed the issue with image rotation by MRZ in some cases.
Desktop web service
- Implemented LED control logic to comply with the Windows version of the service.
- Implemented face match and search capability by adding necessary endpoints to comply with the Windows version of the service.
- Fixed the issue with incorrect video detection behavior in the dark environment for all reader models.
- Added the processParam.convertCase property that allows output text case transformation. No changes are applied by default to the original values.
- Changed user-initiated QOS for scanningProcessQueue in RGLDocumentReaderCameraViewController to avoid priority inversion.
- Fixed the issue with voicing of Status and Result Status messages in the VoiceOver accessibility mode.
- Fixed the issue with the camera on iPhone 12 Pro Max when it didn't focus if a document was placed too close to the camera frame.
- Fixed the issue with correct displaying of the camera frame and preview. Before, they may be displayed incorrectly if the orientation was set explicitly.
- Added the possibility to pass the path to the custom database. Before, the only option of passing the database as bytes was available.
- Fixed the issue when trying to use RFID activity without having NFC permission in AndroidManifest.
- Fixed the issue with database downloading. In rare cases, a crash appeared.
- Fixed the issue with canceling database download. A crash may happen when the method cancelDBUpdate is invoked in the onDestroy lifecycle hook.
- Fixed the issue with the animation of the RFID chip processing on 7310 devices. Now it stops when the RFID processing is started.
- Added an example of the usage of an Ionic Capacitor.
- Changed the naming of methods that are responsible for getting results, like getTextFieldValueByType. See the migration guide on which methods to use instead.
- Updated built-in methods for getting results.
Document Reader Application
- Transferred all status calculation logic from an application into the SDK. Now all statuses are coming directly from the SDK Status result. This means that now all decision-making logic is contained only in one place in SDK for all platforms and deployment variants.
- Added the IMAGE_FILE_LARGE_ADDRESS_AWARE flag for x86 version to allow more RAM to be used for support of 13MP and 18MP devices.
- Fixed the issue with generating a PDF report on multiple pages scan.
- Fixed the issue with unexpected behavior when scanning fingerprints with Jenetric fingerprint scanners via TOUCHLAB SDK.
- Added a property RFIDDefaultParams for setting/getting default parameters used for RFID reading and processing in JSON format (variant, read/write). For example, PA sensitive notifications list can be set using this property in runtime.
- Added the possibility to configure processing parameters.
- Added the possibility to set any property of procesParam available in Core SDK.
- Added the possibility to customize localization.
- Added new actions:
- CAMERA_PROCESS_STARTED triggers when processing using the camera is started.
- FILE_PROCESS_STARTED triggers when processing from files is started.
- Changed the procesParam.returnUncroppedImage property default value to true.
- Updated values of main, hover, active, background, and border colors to fit brand colors.
- Fixed the issue with camera launching in the Edge browser.
- Fixed the issue with the component reloading after document processing.
- Fixed the issue with incorrect naming of codes in Japanese, Chinese, Czech, Estonian, Greek, Tamil locales.
- Fixed the issue with displaying the camera change button. Now if a user has more than two cameras available, the camera change button will be displayed.
Web Service API
- Added new environment variable to support the config file name definition DOCREADER_CONFIG_PATH. This variable can be set to provide a path to the config.yaml file with all necessary parameters set in one place.
- OpenAPI specification updated on GitHub.
- Clients and packages updated.
- Windows Desktop API
- Regula Licensing x64 Package included for x64 build.
- Updated base image to the latest.
- ICAO PKD master list version 219 included.
- ICAO Health master list version 5 included.
- Includes 248 countries and territories, 12200 documents.
- Added 398 new documents.
For details, see the Supported documents list.