A vision-centered design strategy for enhancing color uniformity in diffractive AR-NEDs

Human visual perception is central to designing augmented reality near-eye displays (AR-NEDs), which must ensure consistent color fidelity and viewing comfort across spatial regions and angles. However, diffractive waveguide-based AR-NEDs often suffer from poor color uniformity due to their inherent wavelength and angular selectivity. Additional issues such as grating variations, optical path differences, and localized reflections exacerbate visual inconsistencies. This study introduces a vision-centered optimization framework that integrates optical design with perceptual uniformity. By using the MacAdam Ellipse concept, color variation is constrained within human just-noticeable-difference (JND) thresholds. A gamut contraction strategy further enhances consistency by limiting spectral components to a smaller chromaticity area. The method jointly optimizes the three primary light sources and grating parameters per color channel, coordinating their interaction to confine spectral mixtures within MacAdam bounds and achieve perceptually smooth, uniform color transitions—validated through simulation to enhance fidelity and viewing comfort in AR-NEDs.