Semiparametric Statistical Stripmap Synthetic Aperture Autofocusing

Autofocusing synthetic aperture imagery by maximizing a statistical quality metric such as contrast or sharpness is a well-documented approach in both synthetic aperture radar (SAR) literature and synthetic aperture sonar literature. It is most successfully applied in spotlight-mode SAR applications, where the assumption of spatial invariance of the corrupting phase function is strong and expressions for the gradients of various quality metrics with respect to standard error models have been calculated. Examples of application to stripmap imagery often involve sectioning images into small blocks, allowing spotlight algorithms to be patchwise applied. This paper formulates the gradient of the cost function in a manner that is consistent with the stripmap error model, inherently providing solutions that compensate for the spatial variance while simultaneously bypassing the need for subdividing an image or aperture. This paper formulates the stripmap gradient expression in conjunction with a computationally efficient imaging approach to rapidly achieve metric-maximizing solutions. Demonstrations are shown on a widebeam wideband rail-based system experiencing random jitter and on an unmanned-underwater-vehicle-mounted sonar system exhibiting artificially injected crabbing and sway errors. Results indicate that the algorithm is particularly effective at compensating for random and rapidly oscillating navigation and jitter errors, as well as sidelobes introduced by crabbing in arrayed systems.