Model-driven extended scene piston sensing for synthetic aperture telescopes

Efficient piston estimation is a critical factor in preserving the image quality in synthetic aperture telescopes. When the light source or observation scene is an extended object, the spatial properties of the target and the point spread function (PSF) will undergo convolution effects on the scientific image plane, posing a significant challenge to numerous developed point-source piston sensing methods. In this paper, we investigate a model-driven-based piston sensing strategy capable of high-accuracy piston measurement for extended scenes. Firstly, a mathematical model of the feature vector is constructed to respond to the piston accurately and subsequently we characterize its nonlinear relationship with the piston, termed the frequency secondary-peak piston extraction (FSPE) algorithm. Furthermore, an optimization framework is designed to automatically generate the non-redundant configuration, avoiding the potential baseline crosstalk that can cause the misalignment of feature vector extraction in FSPE. Since the decoupled feature vector contains the analytic properties, through sequentially placing the non-redundant mask and performing the FSPE algorithm, the pistons can be directly retrieved without iterations and any additional instruments. Both numerical simulation and experimental results demonstrate the effectiveness of the proposed method.Given the efficiency and superiority, we believe that the proposed method might find wide applications in future extremely large synthetic aperture telescopes.