22‐1: A Novel Gamma Prediction Algorithm for FDC Region of AMOLED Panel Based on CNN Model

Abstract

To ensure the uniformity of panel image quality, it is necessary to gamma correct the normal and FDC（full display camera） regions of the panel, but the gamma TT（tack time） is particularly long, especially the FDC takes about 110 seconds, resulting in lower production efficiency. Currently, the FDC correction algorithm uses the gamma value of the previous panel as the initial value, and its average error is about 60, and the error is large. To reduce the FDC gamma TT, this paper proposes a new algorithm based on CNN, which inputs normal gamma and predicts FDC gamma. The predicted results are taken as the starting value of the gamma in the FDC region, the starting value error is small and the FDC gamma TT is greatly shortened. Experimental results show that compared with the existing algorithm, the accuracy of the initial value of the FDC region is increased by 10 times. FDC gamma TT was reduced by about 3.5 times.