BS ISO IEC 19794 — Biometric Data Interchange Formats Standard Series

Standardised exchange framework for biometric recognition data

📐 BS ISO/IEC 19794 is the UK adoption of the ISO/IEC 19794 series on Biometric Data Interchange Formats. Developed by the ISO/IEC JTC 1 biometrics subcommittee, it specifies standardised formats, encoding rules and conformance requirements for data from many biometric modalities (fingerprint, face, iris, signature, vascular, hand geometry, DNA), enabling capture devices, template stores and recognition systems from different vendors to exchange and interoperate. It underpins national biometric databases, e-passports, credential verification, border control and identity-authentication systems.

⚠️ This article concerns a scanned version, written based on public-domain knowledge. Parts 1, 10 and 11 are extracted from readable text; the remaining parts are summarised from the public-domain ISO/IEC 19794 body of knowledge. Specific clauses should be verified against the official BSI publication.

1. Series Overview and Scope 📌

ISO/IEC 19794 uses a “framework + per-modality” structure: Part 1 gives the general framework — definitions, conceptual model, data-block structure and conformance requirements; Parts 2 to 14 define record formats for specific modalities (fingerprint, face, iris, signature, vascular, hand geometry, DNA). Each part defines a CBEFF-compatible Biometric Data Block (BDB) with record structure and encoding details.

The series emphasises “interoperability” and “extensibility”: formats use TLV (Tag-Length-Value) or fixed-field structures so different systems can parse them; extension bits and vendor-defined fields allow private information while remaining standard-compatible. Most parts also specify conformance-test methods to verify that generated BDBs match the record structure.

2. Composition of the Series 📚

StandardModalityCore content
BS ISO/IEC 19794-1:2011 + A2:2015FrameworkConceptual model, data-block structure, conformance
BS ISO/IEC 19794-2:2011 + A2:2015Finger minutiaeMinutia (ridge ending/bifurcation) coordinates, direction, quality
BS ISO/IEC 19794-3:2006Finger pattern spectralSpectral pattern representation
BS ISO/IEC 19794-4:2011 + A2:2015Finger imageGreyscale/binary image, resolution, compression
BS ISO/IEC 19794-5:2011 + A2:2015Face imageFace image specification (size, illumination, pose)
BS ISO/IEC 19794-6:2011 + A1:2015Iris imageIris image, annulus and quality
BS ISO/IEC 19794-7:2014 + A1:2015Signature time seriesDynamic signature pen trajectory time series
BS ISO/IEC 19794-8:2011 + A1:2014Finger pattern skeletalMinutiae and texture skeleton representation
BS ISO/IEC 19794-9:2011 + A2:2015Vascular imageFinger/palm vein image specification
BS ISO/IEC 19794-10:2007Hand geometry silhouetteHand geometry silhouette data
BS ISO/IEC 19794-11:2013 + A1:2014Signature processed dynamicProcessed dynamic signature features
BS ISO/IEC 19794-14:2013DNA dataDNA markers and genotype data format

3. Key Technical Requirements 🔧

1. Framework and data block (Part 1)

Part 1 establishes the unified conceptual model: biometric data is carried by a Biometric Data Block (BDB) following the CBEFF structure, including modality type, quality, format identifier and subject information. It defines common fields, extension mechanisms and conformance declarations — the umbrella framework all other parts must follow.

2. Fingerprint data (Parts 2, 3, 4, 8)

Fingerprint minutiae represent ridge endings and bifurcations by coordinates, direction and quality, typically used for fingerprint matching; fingerprint image data defines greyscale or binary image size, resolution (e.g. 500 dpi) and compression; spectral and skeletal data provide richer texture representation for fast comparison in large template databases.

3. Face, iris and vascular images (Parts 5, 6, 9)

Face images specify facial localisation, minimum size, illumination and pose (frontal, uniform light, unobstructed) for face recognition; iris images define pupil/iris annulus extraction parameters; vascular images specify finger/palm vein imaging conditions. Each part defines image-quality indicators to ensure captured data meets recognition requirements.

4. Signature, hand geometry and DNA (Parts 7, 10, 11, 14)

Signature dynamic data records pen X/Y coordinates, pressure and time series (dynamic signature), with Part 11 adding processed features; hand geometry silhouette (Part 10) records geometric features of the hand silhouette; DNA data (Part 14) specifies encoding of loci and alleles for kinship and identity determination.

💡 Practical tip: the key to cross-system interoperability is “quality identifier” and “version management”. The quality field and format identifier in the BDB determine whether the receiving system can parse and match correctly; when upgrading versions (e.g. +A1/+A2 amendments) keep backward compatibility or define a clear version-switch strategy.

4. Engineering Practice 🛠️

  • 🔍 System selection: confirm the capture and recognition systems support the format of the corresponding ISO/IEC 19794 part and that versions/amendments match, ensuring interoperability.
  • 🔧 Template management: store biometric templates in the standard BDB format to avoid lock-in and migration difficulties from vendor-private formats.
  • 📊 Quality control: accept captured data per each part’s image/quality indicators (face pose, iris annulus, fingerprint quality); reject non-conforming data.
  • 📋 Compliance and privacy: biometric data are sensitive personal data; control access, retention and encryption per GDPR / personal-information protection law.

5. Frequently Asked Questions ❓

❓ Q: How do ISO/IEC 19794 and CBEFF relate?
A: CBEFF (ISO/IEC 19785) defines the general framework for biometric data exchange (block wrapping and header); ISO/IEC 19794 defines the record content per modality. Together, 19794 data blocks are wrapped as CBEFF biometric data blocks for exchange.
❓ Q: Why are there multiple parts for the same modality (e.g. fingerprint)?
A: Because representations differ — minutiae, image, spectral and skeletal each suit different use cases (comparison speed, accuracy, storage) and are separate parts for selection, extension and independent revision.
❓ Q: What does the face image format (Part 5) require of a photograph?
A: Frontal pose, minimum pixel size, eye spacing, uniform illumination and no occlusion (glasses, hat brim, etc.), to ensure consistent and accurate feature extraction and matching.
🔍 Q: Is DNA data also “biometric”?
A: Yes. DNA is a highly unique biometric modality; ISO/IEC 19794-14 specifies loci and allele encoding for forensic identification and kinship, following the same exchangeable framework.
🔒
正在加载下载地址,请稍候 10 秒

Leave a Reply

Your email address will not be published. Required fields are marked *