Determine the relationship between AIN, refractive indices, and haze in transparent wood

Determine the functional relationship between the Average Interface Number (AIN), the refractive indices of the cell wall and polymer phases, and the haze in transparent wood composites; and evaluate the contributions of other microstructural parameters—specifically lumen volume ratio, average fiber cross-section area, standard deviation of fiber cross-section area, sample thickness (in pixels), sample width, vessel number, and fiber number—to haze.

Background

The paper introduces an analytical model for angular light scattering in transparent composites, identifying the Average Interface Number (AIN) and the refractive-index mismatch as dominant parameters. For transparent wood composites, the authors propose an Equivalent Average Interface Number (EAIN) to simplify haze prediction, and then use deep neural networks to assess parameter importance. Prior to presenting the DNN analyses, the authors explicitly state that the relationship between AIN and refractive indices with haze is still unknown and that the impact of other structural parameters has not been evaluated. This motivates their subsequent machine learning approach to appraise feature importance and validate the central role of AIN in governing optical scattering and haze.