A Novel Method for UV Measurement Using PLA-based Photochromic Material and Machine Learning Models

Abstract

A new method has been developed that can accurately measure the relationship between color change under ultraviolet (UV) light and the amount of UV using a mobile phone camera to observe the color changes of a polylactic acid-based photochromic material (PLA-PM). Videos of the color change of the PLA-PM were recorded using a mobile phone camera. The obtained data were used for machine learning, and the amount of UV light causing the color change was predicted using machine learning models. The target variable in the models is the voltage values read from the UV sensor via Arduino. Data set preparations process different high-resolution and fps images through six image processing stages. Image processing stages end with the normalization of values for the final image's three-color space channels. This work, which enhances traditional colorimetric measurements with artificial intelligence technologies, demonstrates potential applicability in studies such as UV index measurement, medical measurements, and UV-sensitive textile products. The study experimentally shows the feasibility of measuring UV with mobile phones without relying on expensive devices.