Resolution Limits for Tracking Closely Spaced Targets

Abstract

In the field of target tracking, a tremendous amount of work has been performed on improving the ability of many different algorithms to detect and track a target in the presence of clutter and other interfering targets. However, to date, surprisingly little work has been performed on analyzing whether or not, for a given target in a given clutter/interfering target environment, it is even possible to track the target at all. Our recent work has started to explore this-we have developed a framework to assess "trackability"-that is, we assess if a given target can be statistically separated from the clutter in the given environment. Here, we expand this framework by adding a second target. Assuming that both the targets are individually trackable, we develop a mechanism for establishing how close the two targets can get to one another and still be resolvable as two distinct targets. Note that we answer the question from the perspective of the tracker: a resolution of measurements is a different (and simpler) matter to adjudicate. Generally resolution of measurements implies resolvability of targets, but the converse is not implied.