Presentation Information
[17p-PA2-4]Beam shaping using Zernike coefficient estimation based on deep learning
〇Shota Onoguchi1, Yoshio Hayasaki1, Satoshi Hasegawa1 (1.Utsunomiya Univ.)
Keywords:
deep learning
The design of shaped beams with desired intensity distributions is a fundamental enabling technology that directly contributes to throughput enhancement in laser processing, optical manipulation, and microscopic imaging. In this study, we propose a method that directly estimates the required Zernike coefficients from a target intensity distribution by combining phase representation using Zernike polynomials with deep learning.