| Autor: |
Pan, Tzu-Lung, Chao, Paul C.-P., Nguyen, Duc Huy, Lin, Ching-Chun, Pai, Feng-Ting, Tsai, Yung-Cheng, Tsung, Wan-Nung, Chang, Yao-Jen |
| Předmět: |
|
| Zdroj: |
SID Symposium Digest of Technical Papers; Jun2024, Vol. 55 Issue 1, p793-796, 4p |
| Abstrakt: |
A new efficient color correction method, utilizing neural network models, is proposed by this study for OLED panels in production, offering a transformative approach to achieve accurate and rapid color calibration. Utilizing transfer learning with a limited dataset, it creates color conversion matrices for AI‐driven color correction. The process involves selecting a similar baseline model from prior batches with a ΔE00 below 1, then refining it using a representative dataset. This model corrects white point and overall color, achieving a ΔE00 less than 2. It corrects 300 data points in less than 4 seconds, demonstrating both speed and accuracy. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
Plný text