METHOD FOR DISCRIMINATING BETWEEN SYNTHETIC APERTURE RADAR IMAGES OD TARGETS AND ARTIFICIAL CLUTTERS

摘要

Disclosed is a method of discriminating between targets and artificial clutter in synthetic aperture radar (SAR) images. The target and artificial clutter discrimination method of the present invention includes the steps of extracting images according to a plurality of different two-dimensional intrinsic modes, respectively, from a training target SAR image chip and a training artificial clutter SAR image chip. Each of the images according to the two-dimensional eigenmode of is combined to generate a plurality of unique characteristic enhancement images, each with enhanced characteristic characteristics of the target and artificial clutter, and a 2D kernel function is applied to the generated multiple eigen characteristic enhancement images. Thus, extracting feature values of the center and the periphery of the training target SAR image chip and the training artificial clutter SAR image chip, and generating weight vectors based on the extracted feature values, respectively, and using the generated weight vector. It includes the step of discriminating the target and artificial clutter from the test SAR image chip according to linear discrimination.