Gabor Filters and Neural Networks for Segmentation of Synthetic Aperture Radar Imagery.

摘要

This research investigates Gabor filters and artificial networks for autonomous segmentation of 1 foot by 1 foot) high resolution polarimetric synthetic aperture radar (SAR). Processing involved frequency correlation between the SAR imagery and biologically motivated Gabor functions. Methods for selecting the Gabor tuning parameters from the endless choices of frequency, rotation, standard deviation and bandwidth are discussed. Using these parameters, resulting Gabor correlation images were reduced in speckle, and more detailed. This research used cosine Gabor functions and operated on single polarization HH magnitude data. Following selection of the appropriate Gabor features, multiple Gabor representations were generated and converted for ANN training. Networks investigated were the Kohonen and radial basis function (RBF) algorithms. Provided are results demonstrating a Kohonen network calibration technique and how combination of Gabor processing and RBF networks provide scene segmentation.