Straightness measurement with laser beam and deep learning

Support for working distance change with a contrast learning

Abstract

Straightness measurement that using a laser beam is affected by wavefront aberration due to air refractive index fluctuation (air fluctuation). To compensate wavefront aberration, we propose the deep learning method. In this method, deep learning model estimate air fluctuation uncertainty and subtract it from the measurement value. In this study, we focused the change of working distance, which is the propagation distance of laser beam. The goal is to enable estimating the uncertainty by using deep learning the same as previous study.