v3.10.0

New features:

• added support for scanning Austrian passports - use AustrianPassportRecognizerSettings
• added support for scanning Swiss passports - use SwissPassportRecognizerSettings
• added support for scanning Mexican Voting Cards - use MRTDRecognizerSettings

Minor API changes:

• RegexParserSettings and RawParserSettings now work with AbstractOCREngineOptions, which is a base class of BlinkOCREngineOptions
  • default engine options returned by method getOcrEngineOptions for both parser settings return instance of BlinkOCREngineOptions

Improvements for existing features

• improved address parsing on Malaysian iKad documents
  • affects only iKad recognizer (represented by IKadRecognizerSettings)
• added support for scanning non-expiring Croatian ID documents
  • affects:
    • Croatian ID front recognizer (represented by CroatianIDFrontSideRecognizerSettings) - date of expiry is keyword TRAJNO
    • Croatian ID back recognizer (represented by CroatianIDBackSideRecognizerSettings) - date of expiry inside MRZ is 991231
    • Croatian ID combined recognizer (represented by CroatianIDCombinedRecognizerSettings)

Bug fixes:

• fixed occassional crash in MRTD detection algorithm
  • this affects both MRTD Recognizer (represented by MRTDRecognizerSettings) and MRTD Detector (represented by MRTDDetectorSettings)

v3.9.0

• bugfix in Croatian ID scanning:
  • ensured that OIB number is not returned for old ID cards, where it does not exist
• added Czech and Slovak translations
• Date fields in recognition results are returned as com.microblink.results.date.Date class which represents immutable dates that are consisted of day, month and year
• improved IbanParser:
  • improved extraction of IBANs without prefix and introduced setAlwaysReturnPrefix option to always return prefix (country code)
  • added support for french IBANs
• enabled reading of Pdf417 barcodes having width/height bar aspect ratio less than 2:1
• Added VinRecognizer for scanning VIN (Vehicle Identification Number) barcodes
• Added unified BarcodeRecognizer for scanning various tipes of barcodes
  • ZXingRecognizer, BarDecoderRecognizer and AztecRecognizer are deprecated, BarcodeRecognizer should be used for all barcode types that are supported by these recognizers
• OcrLine.getChars() method returns CharWithVariants array. OCR char is defined by all its parameters (value, font, position, quality, etc.) and for each resulting char it is possible to have multiple variants. For example it is possible to have same char value with different font.
• Fixed bug in SegmentScanActivity:
  • scan results are no longer hidden on shake event
• improved IKadRecognizer:
  • added support for long addresses and employer names
  • fixed returning of full document images as metadata
• improved Singapore ID Recognizers:
  • tuned reading positions
  • more accurate reading of name and blood type fields
• improved Slovak ID Recognizers:
  • tuned reading positions of ID elements
  • improved reading precision of address, place of birth, last name and issuing authority
  • added options to disable/enable extraction of certain fields in recognizer settings
  • result getters getPersonalIdentificationNumber() and getIssuingAuthority() are renamed to getPersonalNumber() and getIssuedBy()
• Renamed RomanianIDFrontSideRecognitionResult element getters for Sex and Nationality outside of the MRZ to getNonMRZNationality() and getNonMRZSex()
• For Austrian ID Recognizers added options to disable/enable extraction of certain fields in recognizer settings

v3.8.1

• improved CombinedRecognizers:
  • better handling of names containing dashes and extra long names
• improved TopUpParser:
  • added option to return USSD code without prefix
• by default MRTDRecognizer does not return results with incorrect check digits

v3.8.0

• introduced ability to create minimum-size AAR
  • a separate static library distribution now exists which contains a script that you can configure with features you need and it creates a AAR file which only contains features you need - this includes minimum-size native binary and only required assets. The rest (resources and java classes) can be thrown-away by ProGuard.
• LibBlinkID is now fully ProGuard-compatible, i.e. you no longer need to exclude com.microblink.** classes in your ProGuard configuration
• removed support for Android 2.3 and Android 4.0 - minimum required Android version is now Android 4.1 (API level 16)
  • devices with Android 4.0 and earlier take less than 2% of market share and is very costly to support them
• removed isItalic and isBold getters from OcrChar class
  • they always returned false, since OCR engine cannot accurately detect that
• removed setLineGroupingEnabled and isLineGroupingEnabled from BlinkOCREngineOptions because disabling line grouping completely destroyed the OCR accuracy
• added GermanIDCombinedRecognizer: scans front and back side of the German ID

v3.7.1

• prefixed custom attributes to avoid name collisions with attributes from other libraries:
  • CameraViewGroup: renamed animateRotation to mb_animateRotation, animationDuration to mb_animationDuration, rotatable to mb_rotatable
  • BaseCameraView: renamed initialOrientation to mb_initialOrientation, aspectMode to mb_aspectMode
• improved MRTDRecognizer:
  • WSA (World Goverment of World Citizens) added as valid country code when parsing MRZ
• added USDLCombinedRecognizer: scans face image and USDL barcode

v3.7.0

• updated German ID recognizers:

  • instead of GermanIDMRZSideRecognizer, which was used for scanning front side of the older ID cards and back side of the new ID cards, there are two specialised recognizers: GermanIDBackSideRecognizer and GermanOldIDRecognizer
  • improved scanning accuracy of the GermanIDFrontSideRecognizer (name and surname)
  • splitting address from back side of the new German ID (GermanIDBackSideRecognizer) to ZIP code, city, street and house number

• better support for German passports: GermanPassportRecognizer reads passport data from MRZ and fields outside of the MRZ

• splitting address in MyKadRecognizer to street, ZIP code, city and state

• improved Croatian ID recognizers:

  • multiple scans are used for better confidence

• TopUpParser improvements

• DateParser can parse dates with month names in English (either full or abbreviated), if this option is enabled

• added support for polish IBAN without PL prefix to IBANParser

• fixed returning of images inside TemplatingAPI for frames when document was not correctly detected

• introduced combined recognizers:

  • AustrianIDCombinedRecognizer: scans front and back side of the Austrian ID
  • CroatianIDCombinedRecognizer: scans front and back side of the Croatian ID
  • CzechIDCombinedRecognizer: scans front and back side of the Czech ID
  • MRTDCombinedRecognizer: scans face image from any type of the document and Machine Readable Zone
  • SerbianIDCombinedRecognizer: scans front and back side of the Serbian ID
  • SingaporeIDCombinedRecognizer: scans front and back side of the Singapore ID
  • SlovakIDCombinedRecognizer: scans front and back side of the Slovak ID
  • SlovenianIDCombinedRecognizer: scans front and back side of the Slovenian ID

  Combined recognizers can be used for scanning multiple parts/sides of the document in predefined order. They combine results from individual scans to boost accuracy and merge them into the final result.

• added VerificationFlowActivity which is designed for scanning multiple parts/sides of the document by using provided combined recognizers

• optimised native binary size

  • 15% size reduction for arm64-v8a ABI
  • 14% size reduction for x86 ABI
  • 13% size reduction for armeabi-v7a and x86_64 ABIs

v3.6.0

• added AztecRecognizer for scanning Aztec barcodes
• improved TopUpParser:
  • added suport for 14 digits long sim numbers in addition to existing lengths (12, 19, 20)

v3.5.0

• MobileCouponsParser is renamed to TopUpParser
• improved TopUpParser:
  • added option to enable all prefixes at the same time (generic prefix)
• fixed layouting of provided SegmentScanActivity and RandomScanActivity in multi-window mode
• added SimNumberRecognizer which scans SIM numbers from barcodes
• improved MRTDRecognizer:
  • better support for arab MRZ
• updated CroatianIDFrontSideRecognizer:
  • returning sex as written on front side of a document
• added support for scanning front side of Romanian ID cards

v3.4.1

• improved multi-window camera support
• fix layouting of default activities in multi-window mode
• fixed problem with front side Singapore recognizer

v3.4.0

• improved quality of german ID address recognition
• added support for extracting place of birth on old German IDs
• added support for scanning IBAN from Georgia in Segment Scan
• added support for cancelling ongoing DirectAPI recognition call
• added option to allow unverified results for MRTDRecognizer:
  • by using method setAllowUnverifiedResults in MRTDRecognizerSettings, it is possible to allow obtaining of results with incorrect check digits
• Singapore ID recognizer is split in two recognizers - one for front and one for back side
• added Belgian account number check to IBAN parser
• added support for Android 7 multi-window mode
• fixed autofocus bug on Huawei Honor 8
• fixed black camera on Motorola Moto Z
• made camera focusing more stable on some devices
  • stable means less "jumpy" when searching for focused image
• added support for receiving of GlareMetadata which informs user that there is too much glare for performing recognition of ID document