Ship structure corrosion detection using advanced image processing, active contour algorithm, and parallel processing

Corrosion poses significant risks to ship infrastructure, causing safety hazards and financial losses. This study introduces a hybrid corrosion detection algorithm combining advanced image processing and a Chan-Vese active contour model with parallel processing. Proposed method achieves classification accuracy of 98 %, precision of 97 %, and segmentation accuracy of 93.83 %. It delivers a sensitivity of 96 % and specificity of 89 % with an 89.2 % lower cutoff. GPU-based acceleration reduces normalization time by 68.49 %, feature extraction by 81.25 %, and matching by 45.01 %. Therefore, proposed method offers a robust and efficient tool for the maritime industry, significantly enhancing the reliability of ship structure monitoring.