Corrosion Detection in Ship and Offshore Structures: Part I – Deblurring for Image Preprocessing
Corrosion in ship and offshore structures presents significant threats to structural integrity and safety, leading to considerable maintenance costs. In maritime environments, where steel and other metals are continuously exposed to harsh conditions, early and accurate detection of corrosion is critical for effective asset management. However, the degradation of image quality due to factors such as motion, low light, and environmental interferences like vibration poses substantial challenges for precise corrosion detection. This paper proposes an advanced image preprocessing method utilizing the Wiener filter to address these challenges through efficient deblurring techniques. By transforming the images into the frequency domain and applying an adaptive deblurring process, the proposed method effectively reduces non-uniform blur while preserving critical high-frequency details, such as corrosion edges and boundaries. A comprehensive comparative analysis highlights the superior performance of the developed approach. Experimental results demonstrate significant improvements, with a Mean Absolute Error (MAE) of 1.01, a Peak Signal-to-Noise Ratio (PSNR) of 44.05, and a Mean Structural Similarity Index Measure (MSSIM) of 0.97. These advancements enhance the accuracy of corrosion detection and segmentation, offering a reliable foundation for further analysis in real-time applications and contributing to the overall efficiency of corrosion monitoring in maritime environments.