Statistical analysis of digital Side-Looking Airborne Radar (SLAR) data

Statistical characteristics of X-Band radar data taken with digital SLAR were analyzed. It is shown that gray value distributions of range strips (50 pixels wide) and of distributed targets are sufficiently described by a normal distribution. Errors of 1st and 2nd kind were evaluated to judge the possibilities of classification, to define the minimum characteristic area of a surface type, and to judge internal equality of different surface types and equality of surface data over range after radiometric correction. A method to derive a correction factor from the SLAR data is demonstrated. This factor comprises system characteristics and geometry parameters only. A calibration concept is described, and requirements concerning calibration normals and overall accuracy are derived from the results of the data analysis.