* This research was supported in part by the Defense Advanced Research Projects Agency under contract DACA76-97-K-0001 and monitored by the Topoggraphic Engineering Center of the U.S. Army.

An overview of a system for building detection and description using multiple views is given and its performance is evaluated. Use of cues from sensors such as IFSAR is included and helps improve the performance substantially. An option for user assisted operation is also described which is able to produce highly accurate models with relatively small efforts from a user. 1 Introduction and Overview The goal of this project is to develop feature extraction methods for automated population of geospatial databases (APGD) with particular focus on buildings. The task includes detection, delineation and description of 3-D buildings. Buildings are objects of obvious importance in urban environments and accurate models for them are needed for a number of battlefield awareness tasks such as mission planning, mission rehearsal, tactical training and damage assessment. Other applications include intelligence analysis for site monitoring and change detection. The problem of 3-D feature extractio..