Prior knowledge in Synthetic Aperture Radar (SAR) processing

Application of information theory concepts of the relation between model and data interpretation to SAR processing is discussed. A model with correlated gamma-distributed cross section fluctuations fits the clutter which is observed in SAR data. Such a model can be augmented by the inclusion of collateral map data of the scene which the SAR is surveying. Thus the map may be used to determine the clutter model parameters in advance of receiving the SAR image data. Regions of the SAR image inconsistent with this clutter model (targets) require a different model. A general method of interpreting such target data is provided by superresolution. Thus prior knowledge of the general properties of targets is introduced to extract the maximum amount of target structural detail from the image. The effect of such processing is to enhance the resolution of the target image.