<title>Application of the H/A/alpha polarimetric decomposition theorem for land classification</title>

Classification of Earth terrain components within a full polarimetric SAR image is one of the most important applications of Radar Polarimetry in Remote Sensing. Unsupervised classification procedure, based around neural networks with competitive architecture, is applied to the full polarimetric SAR images of San Francisco Bay from the NASA/JPL AIRSAR data base (1988) for segmentation and clustering of different Earth terrain components. The linear feature vector used during the classification procedure is defined from a new scheme for parameterizing polarimetric scattering problems, which has application in the quantitative analysis of polarimetric SAR data. The method relies on an eigenvalue analysis of the coherency matrix and employs a 3-level Bernoulli statistical model to generate estimates of the average target scattering matrix parameters from the data.