Deviation-based wavefront correction using the SPGD algorithm for high-resolution optical remote sensing

Abstract

Model-free image-based wavefront correction techniques, such as the stochastic parallel gradient descent (SPGD) algorithm, will be useful in achieving diffraction-limited optical performance in near-future optical remote sensing systems. One difficulty facing the image-based method is that the correction performance depends on the evaluation metric and the evaluated scene. We propose several evaluation functions and investigate the relationship between the optimization speed and the scene textures for each metric in the SPGD algorithm. Based on the simulation results, the study experimentally compared wavefront correction performance using four cost functions and two extended aerial images. Consequently, we found that the deviation-based cost function allowed efficient wavefront correction for versatile extended scenes. In addition, observing extended scenes with distinct structures can facilitate correction speed. Furthermore, we numerically validated this approach in a segmented-aperture imaging system for large telescopes. We believe that the presented approach allows us to realize spaceborne remote sensing with unprecedented high angular resolution.