Personalized Visual Simulation and Objective Validation of Low-Order Aberrations of the Human Eye

We present a practical approach for personalized simulation and objective validation of the effects of low-order aberrations (i.e., myopia, hyperopia and astigmatism) on optical systems, such as the human eye. We characterize the optical aberrations using a wavefront aberration function derived from one's eyeglasses or contact lenses prescription data, expressed in terms of Zernike polynomials. Given the prescription data, a target image at known distance from the subject, and a specified pupil size, the simulation efficiently computes the corresponding aberrated view. We validate the quality of our simulations using ground truth data captured by a camera with the same kind of aberrations. For validation, we use three objective metrics: structural similarity (SSIM), peak signal-to-noise ratio (PSNR), and absolute pixelwise differences (AD) between the simulated images and the ones captured by the camera with induced aberrations. Our results achieved a SSIM mean value of 0.93 (minimum of 0.91) and a PSNR mean value of 35.50 dB (minimum of 29.50 dB), showing that they are structurally similar and virtually indistinguishable from the ground truth images captured by the camera.