Least average residual algorithm (LARA) for tracking the motion of Arctic sea ice

Suppose we have two images ℐ and 𝒥 of a collection of ice floes taken at times t₀ and t₁, respectively. Given an ice floe I in ℐ, the authors develop a new automated tracking algorithm to track the position of I in image 𝒥. The proposed least average residual algorithm (LARA) performs a constrained search for matching ice floes by determining an appropriate search space for each floe to be tracked. LARA takes a stepwise approach with suitable decision rules at various stages. It takes into account various distinguishing characteristics such as: i) the geometry of the ice field; ii) size of an ice floe; and iii) geometry of the floe. LARA also attempts to detect broken floes and amalgamated floes