MATLAB-based toolkit for an introductory course in sar image processing

Abstract

This paper documents the continuation of the work that started with a student paper presented at Radar04. The earlier paper was a description of a graphic user interface (GUI) for SAR data processing that was developed for classroom use, based on a previously published algorithm. This GUI and associated MATLAB toolbox has now been extended to include several other capabilities and has grown into a toolkit that can be used in an introductory course in SAR image processing. Besides the processing of raw SAR data, the GUI and toolbox now include additional teaching and research capabilities, to include additional types of speckle filters and image compression capability. Apart from this the user can choose from two different algorithms that help to extract the single look image from the raw SAR data. These algorithms were implemented in MATLAB, because of its wide availability, and self documenting capability. Thus, the resulting package is suitable for classroom use to illustrate the principles of SAR data processing. Results of the project is made available to the radar community for use by the other schools and institutions that desire a simple SAR processing capability. The MATLAB code has been extensively commented, so that it provides useful baseline for other class projects. Although the material presented is not cutting edge research, it does provide a much-needed tutorial introduction to SAR processing in the classroom. We have also begun developing C and Python based versions of this package so that users can make comparative performance measurements and understand how various implementations of the algorithms affect the end results. We would like to acknowledge the help received from Alaska Satellite Facility (ASF) and the Space Dynamics Laboratory (SDL) at Utah State University for this project.