Radar and optical data integration for land-use/land-cover mapping

This study evaluated the advantages of combining traditional spaceborne optical data from the visible and infrared wavelengths with the longer wavelengths of radar. East African landscapes, including areas of settlements, natural vegetation, and agriculture, were examined. For three study sites, multisensor data sets were digitally integrated with training data and ground-truth information derived from field visits. The primary methodology was standard image processing, including spectral signature extraction and the application of a statistical decision rule to classify the surface features. The relative accuracy of the classifications was established by comparison to ground-truth information. In all sites, the merger of optical and radar sensors improved the ability to map surface features over either sensor independently, although different manipulations of the radar data were necessary to obtain the most useful results. Those manipulations included measures of texture, spatial filtering, and despeckling prior to texture extraction.