A Segmented aperture space telescope modeling tool and its application to remote sensing as understood through image quality and image utility

The remote sensing community is constantly pushing technology forward to achieve bet- ter system performance, this is often done by improving signal-to-noise ratio and spatial and spectral resolution. However, improving one design parameter (such as spatial resolution) could detract from another (such as signal-to-noise). A flexible imaging system simulation tool capable of modeling the effects of changes in system parameters would be a great asset to design engineers. In words, this tool would manipulate a perfect image and produce an output image identical to one physically created by the imaging system. Having such a tool available would make it possible to fully understand a design\u27s potential. In addition, this tool can be used to understand the importance of changes in system parameters. Modern space based remote sensing systems are taking on new forms using sparse and segmented apertures with lightweight mirrors. The driving force for this is that systems are constrained by the size and weight tolerances of the launch vehicles. The new designs come with new problems, many of which are related to the geometry and aberrations of the aperture. The tool developed in this effort will be able to examine the effects of different amounts and types of aperture aberrations. The task is to build an imaging system simulation tool, based on linear systems and standard radiometry, capable of accurately displaying the performance of a plausible design. Using this tool, several designs will be tested using image quality analysis and image utility. Image quality/utility will be determined using three techniques. The first is an image quality prediction technique called the Generalized Image Quality Equation (GIQE) which relates system characteristics to the National Imagery Interpretability Rating Scale (NIIRS). How- ever, due to the unusual aperture geometry of the sparse aperture systems Fiete et al. (2002) showed that the GIQE is unable to accurately predict image quality. The other two approaches are therefore somewhat unorthodox. These approaches do not actually define an image quality but allow systems to be ranked by their performance in a test of motion detection and a test of spatial target detection. A multispectral motion detection algorithm developed and implemented by Adams (2008) combined with motion truth show a given imaging system\u27s ability to track motion. A similar experimental design is evaluated using the spatial target detection algorithm. The tests reveal how changes in parameters such as GSD; SNR; spectral band selection; piston, tip, and tilt aberrations; and light weight optic aberrations affect a system\u27s NIIR\u27s estimate, ability to detect motion, or ability to detect objects