Performance estimation of SAR imagery using NIIRS techniques

Using an information theoretic Global Image Quality Equation(GIQE), we map image formation parameters to National Image Interpretability Rating Scale(NIIRS) levels. These NIIRS levels are also projected onto PID derived from analysts and signature exploitation algorithms for various target sets. Imagery is generated randomly using a combination of truthed clutter background with measured and synthetic chips. Operating condition(OC) dimensions are stored for each classification enabling superior regression fits based on known OCs. BENEFIT: We have presented a set of tools, approaches, and algorithms which define a variety of performance bounds on SAR imagery. This is important to the sensor community because without it, there is no scientific justification for tasking sensors, defining new sensor specifications, and understanding the limitations of signature exploitation data algorithms. We envision these performance models getting used to bound confidence on Non Destructive Inspection algorithms used for airframe monitoring.