Point diffusion function estimation method in self-adaptive optical imaging

Abstract

The invention discloses a point diffusion function estimation method in self-adaptive optical imaging. The invention is based on a B-network method for studying a multivariate spline. Firstly, a distorted wavefront which has gone through atmosphere quality degradation is approximately represented by a bivariate spline in terms of pure subdivision, thereby obtaining a mathematic relation between wavefront sensor gradient measurement data and a spline function coefficient; secondly, in combination with a smooth splicing condition of a spline function in terms of adjacent triangulation, the gradient measurement data is employed to estimate the spline function coefficient by virtue of restricting a least square estimation, thereby reconstructing the distorted wavefront; and finally, the obtained distorted wavefront is combined with a lens pupil function, thereby estimating and obtaining a instantaneous point diffusion function in imaging through atmosphere turbulence. Compared with a traditional Zernike mode point diffusion function estimation method, the point diffusion function estimation method of the invention overcomes the defect of imperfection of a transformation matrix rank and increases the estimation precision; and the point diffusion function estimation method of the invention is applicable to wavefront sensors with shielded centers and various geometric assignment situations of micro-lens arrays.

Description

technical field [0001] The invention belongs to the technical field of space target detection and recognition, and relates to a space target adaptive optical image restoration theory and method, in particular to a point spread function estimation method in adaptive optical imaging. Background technique [0002] According to the characteristics of optical imaging detection of space targets, adaptive optics image restoration methods can be mainly divided into two categories: blind deconvolution and wavefront deconvolution. For blind deconvolution, a relatively accurate estimated value of the point spread function as the initial estimate of the iterative algorithm will speed up the convergence and reduce the amount of calculation; for wavefront deconvolution, the deviation between the estimated value of the point spread function and the true value is more It will directly affect the recovery effect. Therefore, a fast and accurate point spread function estimation method is of g...

AI Technical Summary

Problems solved by technology

The Zernike mode method ultimately boils down to solving linear equations, but due to the incomplete orthogonality of the partial derivatives of the Zernike function and the non-orthogonality of the function on limited sampling points, the rank of the transformation matrix D will appear when seeking the least square solution Incomplete, D Τ D is singular or near-singular, which will amplify the measurement error of the gradient vector, resulting in a decrease in the accuracy of the wavefront reconstruction results, which in turn affects the estimation accuracy of the point spread function; secondly, the centers of many large telescopes are often blocked, if still Using Zernike polynomials to expand the wavefront, because the expansion domain is circular, the Zernike expansion is no longer orthogonal, and still applying this rule requires very cumbersome mathematical derivations and calculations. These factors limit the application effectiveness of the Zernike mode method

In addition, although in recent years, driven by the rapid development of large-scale adaptive optics equipment, new wavefront reconstruction algorithms oriented to correction control have emerged, but the effectiveness of further applying these methods to point spread function estimation and image restoration is still limited. have not been effectively researched and verified

Images