GSST/CBERS-5 International Symposium on Machine Learning in Heliophysics and Space Weather

A study on the use of optical neural networks for adaptive optics applications

Aug 20, 2026, 4:10 PM

1h 20m

Fernando de Mendonça - LIT (National Institute for Space Research, São José dos Campos, SP, Brazil)

Poster Instrumentation & Observational Systems Poster Session

Speaker

Mr Hugo Bernardes (INPE - Instito Nacional de Pesquisas Espaciais)

Description

This work presents a study on optical neural network implementations aimed at contributing to adaptive optics applications for ground-based telescopes. The tests and proposals focus on solar telescopes. The approach offers processing-time advantages, as the systems involve optical stages for image interpretation that operate at the speed of light. Optical systems are proposed to assist in identifying Zernike coefficients, thereby aiding in the estimation of atmospheric turbulence.