Available at: https://digitalcommons.calpoly.edu/theses/345
Date of Award
MS in Electrical Engineering
The goal of this Master’s thesis is to develop a more realistic simulation of Synthetic Aperture Radar (SAR) that has the ability to image detailed targets, and that can be used for Automatic Target Recognition (ATR). This thesis project is part of ongoing SAR ATR research at California Polytechnic State University (Cal Poly) sponsored by Raytheon Space & Airborne Systems and supervised by Dr. John Saghri.
SAR is a form of radar that takes advantage of the forward motion of an antenna mounted on a moving platform (such as an airplane or spacecraft) to synthetically produce the effect of a longer antenna. Since most SAR images used for military ATR are classified and not available to the general public, all academic research to date on ATR has been limited to a small data set of Moving and Stationary Target Acquisition and Recognition Radar (MSTAR) images. Due to the unavailability of radar equipment or a greater range of SAR data, it has been necessary to create a SAR image generation scheme in which the parameters of the radar platform can be directly modified and controlled to be used for ATR applications.
This thesis project focuses on making several improvements to Matthew Schlutz’s ‘Synthetic Aperture Radar Imaging Simulated in Matlab’ thesis. First, the simulation is optimized by porting the antenna pattern and echo generator from Matlab to C++, and the efficiency of the code is improved to reduced processing time. A three-dimensional (3-D) graphics application called Blender is used to create and position the target models in the scene imaged by the radar platform and to give altitude, target range (range of closest approach from the platform to the center area of the target) and elevation angle information to the radar platform. Blender allows the user to take pictures of the target as seen from the radar platform, and outputs range information from the radar platform plane to each point in the image. One of the major advantages of using Blender is that it also outputs range and reflectivity information about each pixel in the image. This is a significant characteristic that was hardcoded in the previous theses, making those simulations less realistic.
For this thesis project, once the target scene is created in Blender, an image is rendered and saved as an OpenEXR file. The image is rendered in orthographic mode, which is a form of projection whereby the target plane is parallel with the projection plane. This parameter means that the simulation cannot image point targets that appear and disappear during the platform motion. The echo generation program then uses the range and reflectivity obtained from the OpenEXR file, the optimized antenna pattern, and several other user defined parameters to create the echo (received signal). Once the echo is created in the echo generation program, it is then read into Matlab in order for it to go through the Range Doppler Algorithm (RDA) and then output the final SAR image.