Integration of hyperspectral and IFSAR data for improved 3D urban profile reconstruction

In this paper hyperspectral (AVIRIS) and radar (AIRSAR) aerial data acquire d over urban environments are considered. The information available from ea ch sensor was extracted and merged to improve the 3D profile reconstruction of builtup areas. Two classification schemes were evaluated for AVIRIS dat a clustering, while the effect of the radar view angle was considered in as sessing the quality of the associated digital elevation models. A detailed analysis of what is possible to extract and to what extent these data are u seful was also produced, considering precise 2D and 3D ground truth of the UCLA campus.