High Fidelity Algorithm to Model the Statistical Variations of Ground Target Signatures in Scene Generator Systems

ABSTRACT: Scene generators are commonly used to create synthetic images of ground targets, such as tanks and battlefield wheeled vehicles, to facilitate seeker algorithm development.These scene generators can produce realistic output for sensor types such as infrared or radar, but are limited by the quality of the target models.Typically, the target models show little or no variation so that the seeker algorithms are not exercised thoroughly against variations in targets such as different paint schemes, minor damage, missing components, etc.We propose to develop a tool to automatically create these types of target variations with minimal user interaction.This tool will be able to generate target model variations for scene generators (such as Irma, FLITES, and CSG) while producing variations that will appear in visible, infrared, and radar wavebands.; BENEFIT: The tools developed in this SBIR will be allow the various scene generators used for seeker algorithm development to be fed target models with random variations so that the algorithms may be more thoroughly tested against targets likely to be seen in the real world.