Many vehicle targets of interest to military Automatic Target Recognition (ATR) possess articulating components: that is, they have components that change position relative to the main body. Many vehicles also have multiple configurations wherein one or more devices or objects may be added to enhance specific military or logistical capabilities.As the expected target set for military ATR becomes more comprehensive, many additional articulations and optional components must be handled. Mobile air defense units often include moving radar antennae as well as turreted guns and missile launchers. Surface-to-Surface Missile (SSM) launchers may be encountered with or without missiles, and with the launch rails raised or lowered. Engineer and countermine vehicles have a tremendous number of possible configurations and even conventional battle tanks may vary items such as external reactive armor, long-range fuel tanks, turret azimuth, and gun elevation.These changes pose a significant barrier to the target identification process since they greatly increase the range of possible target signatures. When combined with variations already encountered due to target aspect changes, an extremely large number of possible signatures is formed. Conventional algorithms cannot process so many possibilities effectively, so in response, the matching process is often made less selective. This degrades identification performance, increases false alarm rates, and increases data requirements for algorithm testing and training. By explicitly involving articulation in the detection and identification stages of an ATR algorithm, more precise matching constraints can be applied, and better selectivity can be achieved. Additional benefits include the measurement of the position and orientation of articulated components, which often has tactical significance.In this paper, the results of a study investigating the impact of target articulation in ATR for military vehicles are presented. 3-D ladar signature data is used. An algorithmic solution is proposed and directions for further research are noted.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.