Point Spread Function Estimation Using Principal Component Analysis for a Double Diffractive Optical Element System

Spectral image (SI) acquisition relies on optical systems implementation; some systems as the shift-variant, require a calibration process where the point spread function (PSF) is acquired at each spatial point. Acquiring the complete PSF of an optical system is a complex and time-consuming process as it requires several snapshots, each one with a high acquisition time. For these reasons, this paper proposes a method to obtain an approximation of the PSF for a double diffractive optical element (DOE) system by capturing a set of calibration patterns designed using principal component analysis (PCA). Simulation results showed that by using 84% fewer snapshots, it is possible to obtain a PSF with an average peak signal-to-noise ratio (PSNR) of 40 dB and an average structural similarity index measure (SSIM) of 90%. Implementation results showed an average PSNR of 40 dB and an average SSIM of 71% with 97.7% fewer snapshots.