Refractometer based on phase measuring deflectometry using smartphone and machine learning assisted analysis

Measuring refractive index values in transparent liquids has broad applications across industrial, scientific, and technological domains for their identification and characterization. Here a phase measuring deflectometry technique with an artificial fringe system is demonstrated to measure the refractive indices of transparent liquids. It works by projecting a line pattern displayed on a smartphone screen through a test chamber with a unique geometry containing the test solution, which a smartphone camera records. The technique detects changes in refractive index by analyzing phase changes resulting from fringe shifts due to the test solution. The phase difference is determined using Fourier transform-based fringe analysis, and the refractive index is measured by extracting features from the computed phase difference profile and training a regression machine learning algorithm. The developed system is compact, simple, low-cost and accurate. It can measure refractive index with a root mean squared error (RMSE) of 8.5375 × 10⁻⁴, a mean absolute error (MAE) of 7.9 × 10⁻⁴, and a precision of 3.175