Point Spread Function Obtaining Research Based On The Linear Space Variant Image Restoration

The existing method for image restoration mostly based on the assumption oflinear space invariant optical system. However, the actual optical system may not bestrictly linear space invariant, especially in large field, point spread function changea lot between center and edge. So when the method in dealing with the actualvariant optical system, the result is not ideal recovery commonly. Because of itsrestoration involves multiple even PSF extraction, mass storage and computing, andin fact we can’t measure all the point spread function, namely the point spreadfunction get not completely, so put forward effective feasible point spread functioninterpolation algorithm and processing algorithm is very meaningful, which willradically improve the quality of the image restoration.In this paper, first of all, we study the point spread function space variantnature based on the theory of aberration, mainly analyze the point spread functionspace variant nature cased by aberration.Secondly, we put forward the space variant point spread functioninterpolation method and processing methods. according to the aberration varianttheory, we suggest spread function interpolation method based on the view of field,and then apply pretreatment Fourier transform method and citing patternrecognition of the Karhumen-Loeve transformation processing spatial variant pointspread function, introduce CGLS and PCGLS algorithm restore space variant pointspread function image.Finally, in order to verify the effectiveness of the algorithm, digital simulationexperiment and experiment are carried out. Building the physical experimentalplatform, this paper gives two kinds of improvement methods of measuring spacevariant point spread function, and gets the blurred image, and verifies the usefulnessand effectiveness of the point spread function interpolation algorithm and imagerestoration algorithm.