Answers are grounded in this paper and cite their sources.
We present a high-speed deep learning-based phase retrieval approach for Shack–Hartmann wavefront sensors used in astronomical adaptive optics. It reconstructs the Zernike modal coefficients from the image captured by the wavefront sensor with a lightweight convolutional neural network. Compared to the traditional slope-based wavefront reconstruction, the proposed approach uses the image captured by the sensor directly as inputs for more high-order aberrations. Compared to the recently developed iterative phase retrieval methods, the speed is much faster with the computation time less than 1 ms for a 100-aperture configuration, which may satisfy the requirement of an astronomical adaptive optics system. Simulations have been done to demonstrate the advantages of this approach. Experiments on a 241-unit deformable-secondary-mirror AOS have also been done to validate the proposed approach.
No takes yet. Share an insight, caveat, or question.
Guo et al. (2021) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: