Interpolation of PSF based on compressive sampling and its application in weak lensing survey

We propose a new point spread function (PSF) interpolation method based on compressive sampling (CS). Complex-Valued fractional Brownion motion (CV-fBm) field is used as a model of the PSF spatial distribution. The 1/f property of the fBm implies that it is a compressible field; therefore, CS will be able to fully reconstruct the field based on a small number of random samples. Performance evaluation shows the advantages of the proposed method over boxcar filtering, polynomial fitting, inverse distance weighting, and thin-plate methods. Potential applicability of the proposed method in weak gravitational lensing survey, particularly for interpolating fast varying PSF that represent distortion by turbulent field is also discussed.