UNIRI-SeaState

The UNIRI-SeaState dataset contains visual-range sea surface images labelled according to eight sea state categories (classes 1–8), corresponding to sea states defined by the Beaufort wind force scale. Images are provided in both RGB and grayscale formats, and are organized into a fixed test set and multiple training splits for benchmarking deep learning image classification models. This dataset accompanies the following publication: M. Pobar, G. Paulin, M. Ivasic-Kos and A. Vorkapic, "Automated Real-Time Localized Sea State Estimation During Navigation Based on the Beaufort Scale," in IEEE Open Journal of Intelligent Transportation Systems, vol. 7, pp. 136-154, 2026, doi: 10.1109/OJITS.2025.3650057.