24.3: Understanding and Achieving Reproducibility for the Evaluation of Display Sparkle Contrast

Abstract

The reproducible quantification of anti‐glare layer based display sparkle with Imaging Luminance Measurement Devices (ILMD) is essential for testing and conformity assessment of many displays aiming at outdoor applications. This study systematically researches relevant setup and system influences of a Fourier space‐based sparkle evaluation. These include 13 different antiglare layers, two displays with different PPI, and several different system setups featuring 12 different camera/lens combinations and more than 10000 individual luminance images. Based on the resulting sensitivities, the measurement procedure is optimized with respect to the achieved reproducibility. The resulting procedure serves as basis to define a new automotive specification for reproducible sparkle measurements and may do so for other applications, that need to quantify sparkle in a reproducible way.

Sparkle measurements for an automotive specification: Evaluation method for flexible setup conditions

“…To have a large database for the setup dependency and reproducibility experiments, we measured and evaluated combinations of the 224 and 183 ppi automotive displays with 6 of the 12 different AGL per display. We selected the AGL samples to cover the complete range of perceived sparkle levels 6 . We used four ILMDs from TechnoTeam (LMK5‐1, LMK5‐5, LMK6‐5, and LMK6‐12) with two ILMD pixel pitches (6.45 and 3.45 μm) and different standard lenses with focal lenses ranging from 16 to 50 mm.…”

“…Several separation methods have been proposed. These include defocusing the pixel matrix within a low depth of focus (DOF) configuration, 3 spatial filtering, 4 frequency filtering, 5,6 undersampling, 7 and defocusing by diffraction blurring with a high DOF configuration 8 …”

“…However, the final sparkle value is falsely increased when using Equation () due to non‐periodic defects (not filtered) such as defective pixels or dust and ringing artifacts. A simple approach to minimize this effect is a local sparkle evaluation, according to our previous work 6 . In this approach, the ROI that shall be evaluated is separated into local regions according to the following equation.…”

“…We reported on the influence of a non‐optimal focus position in prior contributions 6,13 . The focus position is especially critical for manually focusable lenses with a lower DOF, as explained by a distance focus scan.…”

“…2 Several separation methods have been proposed. These include defocusing the pixel matrix within a low depth of focus (DOF) configuration, 3 spatial filtering, 4 frequency filtering, 5,6 undersampling, 7 and defocusing by diffraction blurring with a high DOF configuration. 8 Defocusing the pixel matrix within a low DOF configuration can lead to reproducibility problems of the measured sparkle values between different setups.…”