Answers are grounded in this paper and cite their sources.
Square windows are usually used to estimate the parameters of local background statistics for the constant false alarm rate detection of anomalies/targets in multispectral and synthetic-aperture radar imagery. It is shown that more isotropic windows offer improved detection performance in Monte Carlo simulations. Separable and nonseparable realizations are discussed and a novel recursive realization is presented.
No takes yet. Share an insight, caveat, or question.
Hugh L. Kennedy (2018) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: