High Precision Piston Error Sensing of Segmented Telescope Based on Frequency Domain Filtering

Piston error is the main component of the co-phase errors of segmented telescopes. In this paper, we innovatively performed frequency domain filtering and processing on the focal plane image of the segmented telescopes with mask added, and obtained the image that only reflects the piston error between each submirror and the reference submirror. The representation of feature image that reflects each submirror's piston error which obtained by this method is the same.Therefore, regardless of the number or the arrangement of submirrors, the single shallow convolutional neural network trained by any of the extracted submirror interference image dataset can be used to achieve high-precision detection of different submirror piston errors.Finally, simulation experiment results show the effectiveness of the proposed method.