Optimal Image Denoising for In Situ X-ray Tomographic Microscopy of Liquid Water in Gas Diffusion Layers of Polymer Electrolyte Fuel Cells

Abstract

Improvements in synchrotron based operando X-ray tomographic microscopy (XTM) of polymer electrolyte fuel cells (PEFCs) have paved the way for 4D imaging studies of the water distribution in the gas diffusion layer (GDL). In order to capture the full water dynamics in 4D, a decrease of the scan time towards 0.1 s is aspired, posing significant challenges in image processing for quantitative water detection. In this work, ex situ and in situ X-ray tomographic microscopy experiments were conducted to study the influence of imaging parameters and image denoising settings on image quality and water detectability in the GDL. The image quality is quantified for scan times between 50 ms and 12.8 s at the TOMCAT beamline of the Swiss Light Source. Denoising strategies for a broad range of image qualities were identified, which enable high in situ water detectability rate of 96% at a scan time of 1.6 s and 88% at subsecond scan time as short as 0.4 s. The presented methodology can be transferred to other PEFC or similar XTM imaging setups and image processing pipelines to verify their water detection capabilities.