Fractal Geometry Segmentation of High Resolution Polarimetric Synthetic Aperture Radar Data
Abstract
This thesis investigated the potential of fractal dimension estimation for segmenting high resolution polarimetric synthetic aperture radar. The data used during this research were collected with the Advanced Detection Technology Sensor (ADTS) developed by Massachusetts Institute of Technology Lincoln Laboratory with Defense Advanced Research Projects Agency funding. ADTS is a fully polarimetric calibrated 35 GHz SAR with one foot impulse response. A method of applying the correlation dimension algorithm developed by Grassberger and Procaccia for estimating the dimension of time series data was implemented to estimate the correlation dimension of polarimetric SAR data. A threshold sensitivity study was performed to determine which combination of polarizations used to calculate the correlation dimension resulted in most accurately segmented image. Correlation dimension estimates were shown to be valid and robust features for segmenting ADTS imagery into culture, tree, field, and shadow regions. Simple thresholding and median filtering of correlation dimension estimates calculated from non-overlapping windows of ADTS imagery produced segmented imagery that was consistently over 90% accurate when using all four linear polarizations. An approach was implemented for automatically distinguishing between different classes of naturally occurring regions within the SAR image using correlation dimension estimates as input features to artificial neural networks. (RH)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1990
- Accession Number
- ADA230428
Entities
People
- Joseph L. Brickey
Organizations
- Air Force Institute of Technology