Automatic non-destructive UAV-based structural health monitoring of steel container cranes

• DOI: 10.1007/s12518-023-00542-7
• Publication Date: 2023-12-20T08:02:59Z
• Authors: Vanessa De Arriba...; Mehdi Maboudi; Pedro Achanccaray; Markus Gerke
• Keywords: Port (circuit theory)

Abstract Container cranes are of key importance for maritime cargo transportation. The uninterrupted and all-day operation these container cranes, which directly affects the efficiency port, necessitates continuous inspection massive hoisting steel structures. Due to large size current manual inspections performed by expert climbers costly, risky, time-consuming. This motivates further investigations on automated non-destructive approaches remote fatigue-prone parts cranes. In this paper, we investigate effectiveness color space-based deep learning-based separating foreground crane from whole image. Subsequently, three different ML-based algorithms (k-Nearest Neighbors, Random Forest, Naive Bayes) employed detect rust repainting areas detected body. Qualitative quantitative comparisons results were conducted. While evaluation pixel-based analysis reveals superiority k-Nearest Neighbors algorithm in our experiments, potential Forest Bayes region-based defect is highlighted.