Treffer: Algorithms for Wavefront Reconstruction in Adaptive Optics and Image Post-processing =: 自適應光學及圖像後處理中波前重構算法 ; 自適應光學及圖像後處理中波前重構算法 ; Algorithms for Wavefront Reconstruction in Adaptive Optics and Image Post-processing =: Zi shi ying guang xue ji tu xiang hou chu li zhong bo qian zhong gou suan fa ; Zi shi ying guang xue ji tu xiang hou chu li zhong bo qian zhong gou suan fa

Ph.D. ; In ground-based astronomical imaging systems, adaptive optics are commonly used to correct the wavefront distortion caused by the earth’s atmosphere. In this thesis, a method for real time wavefront reconstruction in adaptive optics system is proposed. The method relies on a decomposition of the wavefront and a truncation in the spatial domain, motivated by Kolmogorov turbulence statistics. The proposed method, which yields an accurate estimation of the wavefront, is efficient because the main computational workload is on solving a discrete Poisson equation. It is parameter free. ; Another wavefront reconstruction method is also considered. It combines the idea of searching for an estimation that minimizes the mean square error in a Krylov subspace and the approach of first reconstructing and then denoising. The method is demonstrated to have comparable accuracy to the minimum variance reconstructor with a low dimensional Krylov subspace. It can be carried out very fast as the creation of the Krylov subspace is cheap. ; Since both the wavefront measurements and the deformable mirror actuators are on coarse grids, the wavefront distortions cannot be eliminated entirely by adaptive optics system, and the observed images still suffer from the blur caused by the residual wavefront. We propose a model for reconstructing the residual wavefront in high resolution from a sequence of deformable mirror data. The high resolution residual wavefront can be used to compute the point spread function which is crucial in removing the blur and further improving the image quality. The model is based on the turbulence statistics and the Taylor frozen flow hypothesis with the knowledge of wind velocities. A tomography problem for the phase distortions from different altitudes is solved to get the reconstruction. We also consider inexact tomography operators resulting from the uncertainty in the wind velocities. This problem is tackled by discovering the wind velocities from the deformable mirror data itself via taking them as ...