Practical Considerations in Restoring Images from Phase‐Diverse Speckle Data

Abstract

We discuss practical problems and solutions associated with the "phase-diverse" speckle (PDS) imaging technique for reconstruction of high-resolution images from ground-based telescope data. The PDS method uses an ensemble of short-exposure images obtained simultaneously from multiple cameras to jointly estimate the object and the sequence of instantaneous phase distortions observed by each camera. Different known static phase aberrations are imposed on each optical channel; the known aberrations then function as constraints in the estimation problem. In this paper, we extend a previously derived model and present new data calibration techniques to compensate for various undesirable effects present in real-world PDS data. These effects include relative tip-tilt and image magnification changes between the diversity channels. The extended model and its associated (modified) objective functions lead to a new nonlinear optimization problem. Solution methods, software, and results using real telescope data are discussed.