Target Tracking Using Various Filters in Synthetic Aperture Radar Data and Imagery
This thesis explores the use and accuracy of several discrete-time image filters for the purpose of target tracking in Synthetic Aperture Radar imagery. Both extended targets and point targets are used for tracking, showing the need for different types of filters for each target type. Monte Carlo an...
| Main Author: | |
|---|---|
| Format: | Others |
| Published: |
DigitalCommons@CalPoly
2009
|
| Subjects: | |
| Online Access: | https://digitalcommons.calpoly.edu/theses/121 https://digitalcommons.calpoly.edu/cgi/viewcontent.cgi?article=1132&context=theses |
| Summary: | This thesis explores the use and accuracy of several discrete-time image filters for the purpose of target tracking in Synthetic Aperture Radar imagery. Both extended targets and point targets are used for tracking, showing the need for different types of filters for each target type.
Monte Carlo analysis is performed on the results of the extended target filter results to determine the absolute mean-squared error between the filter prediction of the target centroid and the actual location of the target centroid. Two different filters were chosen for the extended target: Kalman and H Infinity.
Both the Kalman and H Infinity filters perform tracking by accurately estimating the state of the dynamic system, and in some cases it may be useful to simulate a situation when a target temporarily disappears from radar view. The ability of both filters to predict target location with no input measurements is investigated. A unique trait of the H Infinity filter is its ability to accurately and efficiently estimate the state of a dynamic system given no information about the noise environment.
To simulate more realistic targets, smaller circular and square targets are created and a sensitivity analysis is performed using the Kalman and H Infinity filters to determine the shortfalls of these filter techniques as targets become smaller and smaller. The results indicate that these tracking methods are no longer useful as the targets become so small that they approach being only a single pixel in size.
A new filter called the Prediction and Matching Detection (PAMD) filter is used for single-pixel point targets. This filter illustrates the importance of having very high frame rate images with little change in velocity over consecutive frames if choosing to use the PAMD algorithm. The PAMD filter is extended to track more than one target at a time.
Tracking of raw SAR data is preferred over post-processed images due to the decreased amount of processing time. The Kalman and H Infinity filters are implemented to track raw radar data during its first 3 seconds of motion in 2-dimensions by accounting for the measurements of two parameters: the squint angle and slant range. Noise is added to the measurements to simulate platform inaccuracies.
The project is a continuation of prior SAR research at Cal Poly under Dr. John Saghri with the sponsorship of Raytheon Space & Airborne Systems. |
|---|