Towards robust quality assessment of SAS imagery using the DPCA algorithm

We present a novel DPCA image quality metric based on beam-forming and evaluate it in light of navigation and topographic accuracy. We found the metric to be a simple, accurate, and robust predictor of image quality for the SAS. The metric is a natural by-product of image-formation and should always be calculated. In addition to the DPCA quality metric, we present a straightforward modification of DPCA using motion-compensation techniques to improve its performance in the presence of known geometrical errors. We also evaluate use of quality metrics based on the Normal Inverse Gaussian (NIG) distribution for selecting processing-parameters. The metrics give a parameter estimate close to but further from the manually selected optimum than those of traditional contrast and image-peak metrics.