Inspection of Underwater Hull Surface Condition Using the Soft Voting Ensemble of the Transfer-Learned Models

Abstract

In this study, we propose a method for inspecting the condition of hull surfaces using underwater images acquired from the camera of a remotely controlled underwater vehicle (ROUV). To this end, a soft voting ensemble classifier comprising six well-known convolutional neural network models was used. Using the transfer learning technique, the images of the hull surfaces were used to retrain the six models. The proposed method exhibited an accuracy of 98.13%, a precision of 98.73%, a recall of 97.50%, and an F1-score of 98.11% for the classification of the test set. Furthermore, the time taken for the classification of one image was verified to be approximately 56.25 ms, which is applicable to ROUVs that require real-time inspection.

Semantic Segmentation in Underwater Ship Inspections: Benchmark and Data Set

“…Data sets with underwater imagery such as SUIM [9] or Seagrass [10] exist that aim at the semantic segmentation task or the classification of fish [11] or marine growth [12] species. Although works related to the detection and segmentation of relevant classes and objects in the domain of visual surface inspections as marine growth, corrosion, and cracks exist, the underlying data sets remain undisclosed or are inaccessible [13], [14], [15], [16], [17], [18], [19].…”