A Method for Point Spread Function Estimation of Single Lens Based on Corner Point Detection

In this paper, the existing Point Spread Function (PSF) estimation method is improved, and the spatial variation PSF of aspherical single-lens imaging system is estimated. Most of the existing PSF estimation methods for spatial variation do not consider the fusion of adjacent regions, and the accuracy of edge regions is poor. In this paper, a PSF estimation method based on corner detection is proposed, which makes full use of the similarity of adjacent fuzzy kernels of asphere lens and the spatial symmetry of non-curved lens PSF, and improves the estimation accuracy of PSF with spatial variation.