Image Processing for the Analysis of an Evolving Broken-Ice Field in Model Testing

Dynamic positioning (DP) experiments in model ice were carried out in the ice tank at the Hamburg Ship Model Basin (HSVA) in the summer of 2011. In these experiments the behavior of two different ships in a broken-ice field were studied. One of the main parameters characterizing a broken-ice field is the ice concentration, defined as the fraction of the total water area covered by ice. In this paper, image processing techniques are applied to derive the ice concentration in the model basin. Several points in time are analyzed in order to describe the evolution of the ice field. The applied techniques include methods for identifying individual ice floes and calculating the ice concentration in the vicinity of the model ship. Ice floe boundaries are then obtained, and the ice floe size distribution and shape factor may further be extracted from the images. The image processing methods applied in this work are object extraction and edge detection algorithms, which are further customized to ice identification. The obtained results can be used for relating the ice field characteristics to the model test results, such as the vessel’s displacements and the corresponding ice forces.