Optimization of a flatbed scanner acquisition and sample presentation parameters for sugar color measurement: comparison with spectrophotometry; a case study
Résumé
Abstract Accurate color measurement is essential for quality control in white powdered foods such as sugar. While spectrophotometers are the reference instruments, their cost limits routine use in resource limited settings. Flatbed scanners provide an affordable alternative, but their accuracy depends strongly on technical settings and sample presentation. This study systematically optimized a commercial flatbed scanner parameters (power supply, gamma correction, internal color management, image format, bit depth, and resolution) and physical variables (container dimensions, cover type, and substrate) for a sugar sample color measurement. Spectrophotometric measurements (D 65 illuminant, 2° observer) served as the reference, and color differences were quantified using ΔE 00 . Images captured via an Epson flatbed scanner were processed to derive mean RGB values. Results show that operating the scanner under mains power, using manual gamma correction ( γ = 2.2), 24-bit TIFF format, and 300 dpi resolution, minimized the color difference to ΔE 00 ≈ 4.07, significantly improving accuracy compared to internal color management profiles (ΔE 00 ≈ 6.9–7.2). Increasing bit depth to 48-bit or resolution beyond 300 dpi yielded negligible gains. Physically, a non-fluorescent white cover, quartz substrate, and sample container diameter ≥ 4 cm further improved repeatability. Three way ANOVA confirmed significant interactions (p < 0.001) between cover type and diameter. The study concludes that while the scanner does not yet meet strict industrial tolerances (ΔE 00 < 2), it provides a repeatable and cost-effective alternative for relative color measurement quality screening of white powdered foods when proper optimization protocols are followed in resource-limited settings.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|---|
| Licence |