Abstract
This disclosure describes the training and use of a deep neural network model that can be used to perform color calibration of display devices. The color calibration model employs an encoder-decoder architecture and is trained on primary color calibration data. RGBW color measurement data is obtained for multiple devices and clustered based on the color difference(s) between devices. Multiple iterations based on CGC data from device clusters are performed to train the model until an error between the estimated CGC output by the model and true CGC converges to a minimum specified error. CGC LUTs generated for the devices are utilized to train the model. At the time of use, the trained network model (GCN) is provided the RGBW color patch measurement for a particular display device. The model outputs the CGC LUT for the display device. The described techniques can reduce per-panel color calibration time and improve production cycle (TAKT) times as well as calibration accuracy.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Kim, Peter; Wen, Vicky; Foo, Ken; and Bai, Xiaoping, "Deep Neural Network Model for Display Color Calibration", Technical Disclosure Commons, (July 17, 2024)
https://www.tdcommons.org/dpubs_series/7207