Statistical analysis and segmentation of multi-look SAR imagery using partial polarimetric data

For terrain type classifications, when full polarimetric SAR data are not available, or when only selected discriminants are used, this paper presents customized maximum likelihood classification algorithms based on the probability density functions specifically developed for each case. It is found that in some cases, the use of partial information has actually improved the classification accuracy for some classes. The reason and its implications are discussed. NASA/JPL polarimetric SAR data are used for illustrations.