Overview
ISO/TR 17321-5:2021 - "Graphic technology and photography - Colour characterization of digital still cameras (DSCs) - Part 5" describes sample methods to generate colour targets including highly saturated colours for evaluating the colorimetric image-capture capability of digital still cameras (DSCs). The technical report provides procedural approaches and example datasets to produce spectral distributions that stress camera spectral responses, helping assess how closely a camera’s overall sensor spectral sensitivities (OSSS) approximate colour-matching functions for accurate colour reproduction.
Key technical topics and requirements
- Two target-generation methods
- Eigenvector extension of real spectra: statistical extension and spectral reconstruction using eigenvectors computed from a selected spectral database. Includes procedures for generating boundary colours and highly saturated approximations constrained by chroma, hue and smoothness.
- Artificial LED‑based spectra: selection and synthesis of spectra whose wavelength peaks coincide with colour-difference-sensitive wavelengths (CDSW) to reveal camera sensitivity gaps.
- Spectral sampling guidance: aligns with the ISO 17321 series (e.g., spectral range typically 380–730 nm with ≤10 nm sampling) for dataset selection and reconstruction.
- Evaluation of OSSS: introduces a figure-of-merit (FOM) framework and compares evaluation metrics (e.g., Vora’s μ‑factor, Sharma’s FOM, ΔE-based approaches), discusses advantages/disadvantages, and provides worked examples showing how 17321-5 datasets can be used for camera assessment.
- Supporting materials: informative annexes with eigenvector selection, boundary-gamut analysis, worked examples (e.g., Pointer’s surface colours), CDSW definitions, target spectral datasets and downloadable spectral distributions.
Practical applications and who should use this standard
ISO/TR 17321-5 is practical for professionals who need rigorous, colorimetric camera evaluation:
- Camera and sensor manufacturers - to optimize filter and sensor designs and validate colorimetric capture performance.
- Color scientists and imaging labs - for laboratory evaluation, FOM calculations and research into spectral sensitivity.
- Digital imaging integrators - for validating color accuracy in applications such as medical imaging, cosmetics imaging, e‑commerce, catalogue photography, fine-art reproduction and archiving.
- Test-lab and QA engineers - to develop test charts and LED-based targets that expose colour-difference weaknesses in imaging pipelines.
Related standards and resources
- Part of the ISO 17321 series on colour characterization of DSCs. See ISO 17321-1 for spectral sampling recommendations and the ISO website for related parts.
- Annex H of ISO/TR 17321-5 provides downloadable spectral distribution datasets for immediate use in evaluation and FOM calculations.
Keywords: ISO/TR 17321-5:2021, colour characterization, digital still cameras, DSCs, colour targets, saturated colours, CDSW, eigenvector spectral reconstruction, OSSS, camera FOM, colorimetric image capture.