Information analyses and extraction techniques for synthetic aperture radar signals
Abstract
Techniques for extracting information from synthetic aperture radar images are examined. The properties of these images require the development of new models and processing procedures. The models are also needed to define the average information gained (in the classical Shannon sense) about the target from the observed radar image. A definition for radiometric resolution is proposed based on the average information gain and compared to other descriptions of radiometric resolution based on the probability of detection error for separating two targets as a function of system noise and noncoherent averaging. The information theory analysis also shows that the information per image sample is very small, indicating that data extraction procedures must work with many image samples. With this background, a procedure was developed to extract edge features from radar imagery. An adaptive minimum mean square filter for noise reduction/enhancement was also designed and successfully applied to SEASAT-A SAR imagery.
- Publication:
-
Ph.D. Thesis
- Pub Date:
- 1982
- Bibcode:
- 1982PhDT........16F
- Keywords:
-
- Image Processing;
- Information Theory;
- Radar Imagery;
- Radiometric Resolution;
- Synthetic Aperture Radar;
- Edges;
- Image Enhancement;
- Mathematical Models;
- Seasat 1;
- Signature Analysis;
- Communications and Radar