KNOWLEDGE-BASED CLASSIFICATION OF POLARIMETRIC SAR IMAGES

In preparation for the flight of the Shuttle Imaging Radar-C (SIR-C) o n board the Space Shuttle in the spring of 1994, a Level-1 automatic c lassifier was developed on the basis of polarimetric SAR images acquir ed by the JPL AirSAR system. The classifier uses L- and C-Band polarim etric SAR measurements of the imaged scene to classify individual pixe ls into one of four categories: tall vegetation (trees), short vegetat ion, urban, or bare surface, with the last category encompassing water surfaces, bare soil surfaces, and concrete or asphalt-covered surface s. The classifier design uses knowledge of the nature of radar backsca ttering from surfaces and volumes to construct appropriate discriminat ors in a sequential format. The classifier, which was developed using training areas in a test site in Northern Michigan, was tested against independent test areas in the same test site and in another site imag ed three months earlier. Among all cases and all categories, the class ification accuracy ranged between 91% and 100%.