Multidimensional SAR data analysis based on binary partition trees and the covariance matrix geometry

In this paper, we propose the use of the Binary Partition Tree (BPT) as a region-based and multi-scale image representation to process multidimensional SAR data, with special emphasis on polarimetric SAR data. We also show that this approach could be extended to other types of remote sensing imaging technologies, such as hyperspatial imagery. The Binary Partition Tree contains a lot of information about the image structure at different detail levels. At the same time, this structure represents a convenient vehicle to exploit both the statistical properties, as well as the geometric properties of the multidimensional SAR data given by the covariance matrix. The BPT construction process and its exploitation for PolSAR and temporal data information estimation is analyzed in this work. In particular, this work focuses on the speckle noise filtering problem and the temporal characterization of the image dynamics. Results with real data are presented to illustrate the capabilities of the BPT processing approach, specially to maintain the spatial resolution and the small details of the image.