microblink/blinkid-android v8001.0.0 on GitHub

What's new

Accelerated support for newly issued documents
- We’ve dramatically optimized our document support and extraction pipeline to drastically accelerate our time-to-production for newly issued identity documents worldwide, as well as unobtrusively support improvements related to document rules and knowledge.
- Zero-day readiness: Depending on design availability and when a new document begins circulating, BlinkID can now deliver 0-day support—and at a maximum, within 4 weeks of the document's release.
- Future-proof compliance: This allows your workflows to seamlessly adapt to global document updates faster than ever before.
Enhanced frame selection & image quality
- We’ve upgraded our image capture engine to ensure only the highest-quality frames are processed, boosting extraction accuracy while reducing friction during scanning
  - Smart quality selection: BlinkID now automatically analyzes multiple frames in real time and picks the crispest, most stable image before extracting data.
  - Higher first-pass success: By filtering out blurry or unstable frames early, the SDK significantly reduces scan failures and false reads.
  - Flexible speed vs. quality tuning: You can easily tune the scanner’s balance between ultra-fast results and maximum image precision to fit your specific onboarding flow.
  - Intelligent app feedback: Clear state updates inform your application whenever more stable images are needed, making it easy to provide smooth, real-time guidance to end users.
Enhanced photo mode intelligence
- We’ve added a new configuration setting specifically designed for workflows utilizing Photo Mode:
  - Smart image handling: If you are unsure whether the images being fed into the SDK are already cropped or uncropped, this new toggle optimizes how the SDK handles the input. It ensures maximum extraction accuracy regardless of the initial image state.
  - Explicit image signals: The SDK now returns an explicit signal informing the caller whether the input image was cropped. This allows the host application to intelligently determine the required next steps in complex extraction and verification processes.
Expanded & improved barcode capabilities
- We’ve broadened our barcode recognition capabilities and boosted overall scanning performance
  - Aztec barcode support: You can now enable scanning for Aztec barcodes, expanding the types of documents and codes BlinkID can read.
  - Enhanced 2D & 1D reading: Substantially improved speed and accuracy when scanning QR codes, DataMatrix codes, and select 1D barcode formats, ensuring faster capture even on complex or damaged codes.
Parsing & extraction improvements
- Persian digit recognition: We have significantly improved our parser’s accuracy when reading and processing Persian numbers, ensuring more reliable data extraction for regional documents.
As part of the mandatory data redaction we have expanded the list with Netherlands DL QR code being redacted

Bug fixes

Document swap data caching: Fixed an issue in continuous-video mode where data from a previously scanned document could persist after a new document was introduced. The SDK now immediately detects document swaps—either when a document is removed from the frame or when a new document is introduced—and automatically clears cached images (such as cropped faces, signatures, and barcodes) to prevent data cross-contamination.

New documents support

Bailiwick Of Jersey - Driver's License
Bailiwick Of Jersey - Paper Passport
Bailiwick Of Jersey - Polycarbonate Passport
Botswana - Driver's License
Brunei - Polycarbonate Passport
Democratic Republic Of The Congo - Polycarbonate Passport
Dominican Republic - Polycarbonate Passport
Eswatini - Driver's License
Gambia - Driver's License
Georgia - Residence Permit
Iceland - Identity Card
Iceland - Residence Permit
Japan - Specified Residence Card
Liechtenstein - Residence Permit
Liechtenstein - Resident ID
Mali - Polycarbonate Passport
Mauritania - Resident ID
Monaco - Identity Card
Monaco - Residence Permit
Nepal - Identity Card
Palestine - Identity Card
San Marino - Identity Card
Sudan - Driver's License
UK, Northern Ireland - Voter ID
USA, Arkansas - Medical Marijuana ID
USA, Massachusetts - Medical Marijuana ID
USA, Michigan - Medical Marijuana ID
USA, New Jersey - Medical Marijuana ID
Zambia - Residence Permit

New document versions for supported documents

Bolivia - Driver's License
Burkina Faso - Identity Card
Central African Republic - Paper Passport
Estonia - Identity Card
Dominican Republic - Identity Card
Guyana - Identity Card
Israel - Identity Card
Luxembourg - Polycarbonate Passport
Mexico, Baja California - Driver's License
Mexico, Hidalgo - Driver's License
Oman - Identity Card
Oman - Resident ID
Uruguay - Identity Card
USA, Mississippi - Identity Card
USA, Nebraska - Driver's License
USA, Nebraska - Identity Card
USA, North Carolina - Identity Card
USA, Oklahoma - Driver's License
USA, Oklahoma - Identity Card
USA, Texas - Weapon Permit

New extracted fields from documents

Brazil national identity card and regional (22 regions) identity cards now extract parent names (parentInfo).
Croatia identity card: added support for cyrillic values in sex, address, issuingAuthority, lastName and firstName, and latin sex.
Virgin Islands of the United States identity card and driver's license: added documentSubtype and specificDocumentValidity.
Tunisia identity card: added arab values for lastName, firstName, placeOfBirth; added dateOfBirth.
Argentina identity card and alien identity card: instead of documentAdditionalNumber, cardAccessNumber should now be used.

API changes

Breaking changes

In DocumentCaptureModuleSettings, cropType replaces the boolean inputImageCropped.
DocumentClassInfo changed:
- type is now documentType.
- country, region, and documentType are now nested objects, each of which contains an id and rawValue property. The rawValue property returns information delivered by the OTA mechanism.
- The classification enums were renamed: Country → CountryId, Region → RegionId, Type → DocumentTypeId.
- countryName, isoNumericCountryCode, isoAlpha2CountryCode, and isoAlpha3CountryCode are now optional, and documentClassInfo is now nullable on BlinkIdScanningResult and InputImageAnalysisResult.

Additive changes

inputImageSelectionStrategy added to DocumentCaptureModuleSettings.
aztecScanningEnabled added to BarcodeModuleSettings.
AwaitingMoreStableInputImages added to ProcessingStatus.
ethnicity added to BlinkIdScanningResult and VizResult.
fullName added to ParentInfo.
ParentFullName and Ethnicity added to FieldType.
inputImageCropAnalysis and vizExtractionType added to InputImageAnalysisResult.

New OTA (over-the-air) feature

We've introduced OTA support to ensure your application can scan the latest documents without requiring a full SDK update; to support this, BlinkIdSdkSettings has been refactored to separate base resources from OTA resources
Resource-related variables in BlinkIdSdkSettings have been moved into two new configuration classes: ResourcesConfig and OtaResourcesConfig
OLD:

public data class BlinkIdSdkSettings(
    val licenseKey: String,
    val licensee: String? = null,
    val downloadResources: Boolean = true,
    val resourceDownloadUrl: String = defaultResourceDownloadUrl,
    val resourceLocalFolder: String = defaultResourcesLocalFolder,
    val resourceRequestTimeout: RequestTimeout = RequestTimeout.DEFAULT,
    val microblinkProxyUrl: String? = null
)

NEW - resource related variables have been moved to ResourcesConfig:

public data class BlinkIdSdkSettings(
    val licenseKey: String,
    val licensee: String? = null,
    val resourcesConfig: ResourcesConfig = ResourcesConfig(),
    val otaResourcesConfig: OtaResourcesConfig = OtaResourcesConfig(),
    val microblinkProxyUrl: String? = null
)

public data class ResourcesConfig(
    val download: Boolean = true,
    val serviceUrl: String = defaultResourceDownloadUrl,
    val localFolder: String = defaultResourcesLocalFolder,
    val requestTimeout: RequestTimeout = RequestTimeout.DEFAULT
)

Understanding OtaResourcesConfig: this behaves similarly to ResourcesConfig, but introduces a strict parameter and update checks.

public data class OtaResourcesConfig(
    val checkForUpdates: Boolean = true,
    val strict: Boolean = false,
    val serviceUrl: String = defaultOtaDownloadUrl,
    val localFolder: String = defaultOtaResourcesLocalFolder,
    val requestTimeout: RequestTimeout = RequestTimeout.DEFAULT
)

strict - when set to true, the SDK is forced to fetch the latest OTA resources before proceeding, ensuring support for the newest documents
Other settings mirror the behavior of standard base resources
Important to note:
First-run downloads are unavoidable for OTA when resources are missing locally - checkForUpdates: false only suppresses update checks, not the initial fetch.
strict = true changes init to a throwing failure path. Make sure your initialization error handling accounts for OTA download failures if you enable it.
Two distinct hosts: base resources default to https://models.cdn.microblink.com/resources; OTA resources default to https://blinkid-ota.microblink.com. Don't cross-wire the service URLs.
Distinct cache folders: blinkid (base) vs ota (OTA). Keep them separate to avoid collisions.
OTA resources can be prebundled just like base resources to avoid additional downloads while using SDK:
1. Use script from OTA resources downloader repository to download OTA resources - recognizer version for this release is 24.0.4
2. Bundle downloaded OTA resources to the assets directory of your Android app build (e.g. assets/microblink/blinkid/ota)
3. Set checkForUpdates to false and localFolder to match the location of your bundled OTA resources

Other platform features

Beep sound which is activated on a successful side-scan and document scan has been added
Increased default ResourceTimeout from 10 to 30 seconds when downloading resources