[20a-P01-1~21] 3.7 Optical measurement, instrumentation, and sensor

[20a-P01-21] Change in recognition rate of white interference fringes detection using convolutional neural network due to change in display scale of figures

〇(B)Ryohei Takeishi¹, Taketo Miura¹, Dong Wei¹ (1.NUT)

Keywords:

convolutional neural network,interference fringes,interferometer

Analysis of interference fringes is essential for measurement using interferometers. The process of extracting interferometric fringes for analysis has conventionally been performed by human eyes, which requires a lot of time and labor. Therefore, we aim to automate the extraction process by using a convolutional neural network. The obtained time-series data are converted into images, and discriminations are attempted using a supervised clustering method. The purpose of this study is to investigate the effect of the difference between linear and logarithmic scales on the detection accuracy.