IFT is known to underestimate floe area as a tradeoff to separating close-to-each-other floes. Since the area underestimate is generally a constant radius, we could recalculate the area to remove the bias. This could be done, for example, as a function of the major and minor axis lengths, or by dilating the shape and re-computing area.
IFT is known to underestimate floe area as a tradeoff to separating close-to-each-other floes. Since the area underestimate is generally a constant radius, we could recalculate the area to remove the bias. This could be done, for example, as a function of the major and minor axis lengths, or by dilating the shape and re-computing area.