Abstract

Adaptive optics (AO) is crucial for extreme Large Telescopes (ELTs), and its core of operation lies in the use of wavefront sensors. Although the Shack Hartmann and Pyramid Wavefront Sensors are more common, the axicon wavefront sensor (AxWFS) is a less-explored alternative, where the light is projected onto the detector over a doughnut-shape area. This study introduces a groundbreaking enhancement, employing a state-of-the-art deep neural network to perform wavefront estimation from the intensity changes within the ring produced by the axicon under different turbulence conditions, without requiring any optical modulation.