Methodology of multichannel radar image processing and interpretation

The multichannel (i.e. multifrequency or multipolarization) radar images provide an opportunity to increase the probability of correct identification of remote sensing data and to solve some practical tasks when it is difficult to retrieve information from one-channel images. But, at the same time, many new applied problems arise dealing with the necessity to process and interpret the images jointly. They are the following: the image-to-image and/or image-to-topology map superimposing, geometric distortion correction, image filtering and information extraction. We analysed the main factors degrading the image quality and obstructing reliable data interpretation. It was proposed to remove them at different stages of multichannel image processing mainly by means of application of nonlinear algorithms.