ISO/IEC 29794-5:2025

Overview

ISO/IEC 29794-5:2025 - "Information technology - Biometric sample quality - Part 5: Face image data" defines requirements and methods for assessing face image quality for single-image inspections. The standard specifies how implementations should quantify a face image’s properties against canonical face images (for example, those in ISO/IEC 39794-5:2019) for three primary use-cases:

• collection of reference samples for ID documents,
• sample system enrolment, and
• probes for instantaneous response.

It includes terms, definitions and algorithmic computations for face image quality assessment (FIQA), while explicitly excluding multi-image comparison, 3D capture assessment, quality-data encodings and algorithm performance evaluation.

Key topics and technical requirements

ISO/IEC 29794-5:2025 focuses on measurable, reproducible quality components and provides required computations and mappings. Major technical topics include:

• Common computations: image conversions, face detection, landmark estimation, alignment, face parsing, occlusion segmentation, head-pose estimation, luminance and color space conversions, histograms and entropy.
• Quality measures: a unified quality score model plus individual components grouped as:
  • Capture-related: background uniformity, illumination uniformity, dynamic range, sharpness, compression artifacts, natural colour.
  • Subject-related: single-face presence, eyes open/visible, mouth closed, occlusions, inter-eye distance, head size and crop, head pose, expression neutrality, no head covering.
• Quality block and encodings: guidance on binary and XML quality blocks and identifiers for organization/algorithm provenance.
• Conformance and testing: conformance assertions and informative annexes covering quantitative goals, applications, testing reports and sequential use guidance.

The document mandates that quality assessment software inspects exactly one image; many component computations and mappings are standardized to ensure interoperable quality outputs.

Applications and users

Practical applications of ISO/IEC 29794-5:2025 include:

• ID document capture and issuance (passport, driver’s license)
• Biometric enrolment systems and self-service kiosks
• Border control and biometric exit/entry systems
• Access control and identity verification services (mobile apps, banking)
• Camera and imaging system design where face capture quality is critical

Primary users: biometric engineers, FIQA algorithm developers, system integrators, device manufacturers, government agencies, testing laboratories and product managers who need standardized face image quality metrics to reduce recognition errors and improve capture usability.

Related standards

• ISO/IEC 39794-5:2019 (canonical face image specifications) - referenced for conformity targets.
• Other parts of the ISO/IEC 29794 series - for broader biometric sample quality guidance.

Keywords: ISO/IEC 29794-5, face image quality, biometric sample quality, face image data, face image quality assessment, FIQA, biometric capture.