Enhancing the Accuracy of XYZ Tristimulus Colorimeters in OLED Display Manufacturing through an Extended Unique Matrix = OLED디스플레이 제조라인에서 Extended Unique matrix XYZ 를 통한 삼자극치 색차계의 정합성 향상

다국어 초록 (Multilingual Abstract)

The inherent spectral mismatch between the XYZ filter’s spectral responsivities of tristimulus colorimeters and the Commission Internationale de l'Éclairage (CIE) 1931 Color-Matching Function (CMF) leads to measurement deviations whenever the Spectral Power Distribution (SPD) of the Device Under Test (DUT) differs from that of the reference display. To correct this problem, calibration matrices based on American Society for Testing and Materials (ASTM) E1455 have been widely adopted, which assumes that the tristimulus values XYZ increase linearly with light intensity. In Organic Light-Emitting Diode (OLED) displays, however, nonlinear spectral shift occurs. Moreover, SPD becomes narrower, the deviation caused by spectral shifts becomes more significant, resulting in substantial degradation of the ASTM matrix’s correction performance.
To address these limitations, this dissertation proposes a new calibration method termed the Extended Unique matrix which selects three reference SPDs for each primary color to cover the full range of spectral variations observed in OLED displays. To verify its performance, simulation studies were conducted using eighty- eight (88) RGB patterns from seven (7) commercial OLED smartphones, representing a variety of emission spectra. In addition, eight (8) XYZ tristimulus colorimeters with spectral mismatch indices (𝑓1 ′) ranging from 1.9% to 5.5% were employed to examine the influence of inter instrumental variations.
The simulation results demonstrated that the Extended Unique matrix substantially outperforms the ASTM and other methods. On average, it reduced ΔE₀₀ by around 50% compared with the ASTM matrix and maintained less Δu′v′ than the ASTM matrix. Even for colorimeters with large spectral mismatch (𝑓1 ′ > 5%), the Extended Unique matrix delivered consistent improvements, indicating that its calibration stability is largely independent of colorimeter’s filter accuracy. This robustness makes the method particularly valuable for production environments that operate with multiple colorimeter models exhibiting different spectral characteristics.
An additional optimization study investigated how the selection of reference DUTs affects calibration performance for Extended Unique matrix. Two groups were defined: a wide-distribution group, in which reference displays exhibit large chromatic separations in CIE 1976 u′v′ space, and a narrow-distribution group, in which SPDs are similar. The analysis confirmed that using wide-distribution references leads to smaller color errors after correction, as it ensures that the calibration matrix fully covers the range of possible spectral variations. Conversely, using narrow-distribution references limits the calibration space and yields poorer correction results. This finding provides practical guidance for selecting reference DUTs in industrial calibration systems.
To address the lack of standardization in selecting reference devices for ASTM matrix calibration, this study formulates the selection process as a Set Cover Problem. Utilizing a Greedy Set-Cover algorithm, fifteen potential reference displays across four distinct technologies were analyzed to identify the minimal subset required to maintain measurement accuracy within defined tolerances.

목차 (Table of Contents)

• Abstract i
• List of Tables v
• List of Figures vi
• Chapter 1. Introduction 1
• 1.1. Background and Motivation 1
• 1.2. Problem Statement 4
• 1.2.1. Failure of Linearity by Spectral Shift 4
• 1.2.2. Narrowband SPD 4
• 1.2.3. High f1′ 4
• 1.3. Contributions of the Dissertation 7
• Chapter 2. Background 10
• 2.1. OLED Emission and Spectral Behaviour 10
• 2.1.1. Narrowband SPD 10
• 2.1.2. Spectral Shift on OLED display 11
• 2.2. XYZ Tristimulus Colorimeter and Spectral Mismatch 16
• 2.2.1. XYZ Tristimulus Colorimeter 16
• 2.2.2. Spectral Mismatch 22
• 2.3. Existing Calibration Matrix 24
• 2.3.1. ASTM Matrix 24
• 2.3.2. Unique Matrix 25
• Chapter 3. Proposed Extended Unique Matrix 27
• 3.1. Concept and Formulation 27
• 3.2. Reference DUTs Selection Strategy 28
• Chapter 4. Validation 32
• 4.1. Simulation Set-up 32
• 4.1.1. Device Under Test 32
• 4.1.2. Light Measurement Device 39
• 4.2. Simulation Results 41
• 4.2.1. Extended Unique Matrix 41
• 4.2.2. ASTM Matrix affected by 𝑓1′ 45
• 4.2.3. Unique and Bodner Method 49
• 4.3. Optimization of Reference DUTs 54
• 4.3.1. Extended Unique Matrix 54
• 4.3.2. ASTM Matrix 60
• 4.4. Experimental Verification 69
• 4.4.1. Extended Unique Matrix 69
• 4.4.2. ASTM Matrix 72
• Chapter 5. Conclusion 79
• 5.1. Summary of Major Findings 79
• 5.2. Future Works 81
• References 83
• 국문초록 86
• Acknowledgement 88