Thursday, May 2025
11:00 AM - 11:20 AM
Room: LL20A
Session: Automotive HUDs and Transparent Displays
See-Through Image Quality Evaluation Index for Transparent Displays Considering Human Visual Sensitivity
Description:
As transparent displays become more prevalent, assessing the quality of see-through images is increasingly important. Currently, no standard metrics exist for this purpose. We introduce a novel method that generates a human visual simulation image by modifying the PSF with the CSF and convolving it with the background. The simulated image is compared to the original using the SSIM Index. Our method's effectiveness is validated by human visual perception scores, with a Kendall correlation coefficient of 0.929, confirming its reliability.