Advancements and Challenges in Convolutional Neural Networks for Marine Corrosion Detection and Classification

Abstract

Marine corrosion significantly undermines the structural integrity of maritime infrastructure, necessitating the development of sophisticated techniques for its early detection and classification. This paper offers an exhaustive critical review of Convolutional Neural Networks (CNNs) applied in marine corrosion detection and classification, covering research from 2018 to 2023. It compiles insights from various scholarly articles, elucidating the progression of CNN methodologies in tackling the intricate challenges associated with corrosion in marine, offshore and oil & gas sectors. This review meticulously examines the deployment of CNN technologies in evaluating corrosion across a myriad of maritime assets, including ships, marine structures, offshore platforms and oil & gas pipelines, also construction materials. It explores a broad spectrum of methodologies, underscoring the advancements in CNN-based strategies for corrosion monitoring. Importantly, the review pinpoints key obstacles, innovative strategies and forthcoming trends in the field, offering a comprehensive summary of current research on marine corrosion detection and classification through CNNs. The insights gained from this thorough analysis are instrumental in deepening the understanding of technological and methodological progress, serving as a guide for future research endeavours in the crucial field of maritime asset integrity management.