Quantitative description of the quality of daylight electroluminescense (dEL) images against dark room EL images

Abstract

In this study, we present a robust comparison between EL/dEL images taken with an InGaAs camera, and “golden images” taken under optimal conditions with a Si camera in the dark, which serves as the benchmark. We study the correlation of EL/dEL data quality correlated with the golden image, considering different acquisition parameters and electrical current panel modulations. A key contribution of this work is the correlation of the SNR25 metric, recently introduced by our group, with pixel-by-pixel metrics used to assess image similarity using the Structural Similarity Index (SSIM). Our findings indicate that the quality of dEL images is reliable, showing a satisfactory correlation with data obtained through dark room EL. This analysis was conducted by comparing the structural correlation pixel by pixel (SSIM), demonstrating that dEL data has a very high correlation, indicating that both methods provide the same information about the panel. Furthermore, we found a correlation between the SNR25 metric and the SSIM, allowing us to use this metric as a proxy to assess the quality of the dEL images taken on the field. The results of this research validate the use of dEL imaging for practical applications.