High‐resolution image processing for monitoring bubble dynamics in green hydrogen electrolysis cells

Abstract This paper introduces a high‐resolution image processing framework for the monitoring of hydrogen bubble dynamics in green hydrogen electrolysis cells. The proposed system accurately detects, segments, and quantifies bubble behavior under diverse electrical conditions, thereby enhancing the analysis of electrolyser efficacy. The study employs sophisticated imaging and preprocessing techniques to visualize and analyze hydrogen bubble formation. Electrochemical conditions are correlated with quantitative bubble parameters to facilitate performance optimization. Gas bubble accumulation during electrolysis reduces the efficacy of the system, despite the fact that green hydrogen is a critical sustainable energy carrier. Therefore, it is imperative to comprehend bubble dynamics in order to enhance the design and operation of electrolysers. Electrochemical modeling and optical tomography were employed in previous research to investigate bubble behavior, but these methods were frequently hindered by noise and overlapping bubbles. Recent image‐based and machine‐learning methodologies demonstrate enhanced detection capabilities; however, they necessitate superior preprocessing. The proposed methodology captures high‐resolution electrolysis images, which are subsequently normalized, contrast enhanced, bubble detected, and segmented. The growth, motion, and detachment behavior of bubbles are assessed through the examination of their features. The use of effective noise reduction and precise bubble edge detection with enhanced Peak Signal‐to‐Noise Ratio (PSNR) and minimal Mean Square Error (MSE) values is illustrated by the experimental outcomes. The system effectively quantified bubble size, frequency, and movement across multiple current densities.