📝 SummarXiv

Abstract

The size of sea ice floes in the marginal ice zone (MIZ) is a key factor influencing ice coverage, albedo, wave propagation through ice-covered waters, and ocean--atmosphere energy exchanges. Floe size can be observed by processing visual-range imagery from ships, aircraft, or satellites. However, autonomously capturing floe boundaries from imagery remains challenging, particularly due to the heterogeneity of sea ice, which impairs boundary definition and reduces image clarity. This study evaluates the accuracy of sea ice floe segmentation using the gradient vector flow (GVF) active contour method, the deep learning-based Segment Anything Model (SAM), and a hybrid approach combining GVF and SAM. These methods are evaluated on a representative subset of a large dataset of close-range, high-resolution sea ice imagery, collected from cameras aboard an icebreaker during an Antarctic winter expedition. Spanning a wide range of ice conditions and image clarity in the MIZ, the subset provides a rigorous test bed of segmentation approaches. Their performance is assessed in terms of floe detection accuracy, floe size distribution, and ice concentration, with results compared against a manually segmented benchmark. The findings indicate that SAM, when used in prompt-driven mode, offers the best balance between accuracy and computational efficiency. Its strong performance in estimating sea ice concentration and detecting floes, while maintaining close agreement with benchmark floe size distributions, makes it suitable for real-time applications and scalable analysis of large imagery datasets.