NGA University Research Initiative (NURI) - Automated Scene Modeling From Multi-Sensor Modality Inputs

Sponsor: National Geospatial-Intelligence Agency (NGA)

Project Scope: This NGA University Research Initiative (NURI) project is aimed at speeding up the construction of DIRSIG scenes by automating aspects of the process. In particular, algorithms are being developed to extract the necessary geometry and material characteristics of a scene from multiple modality remote sensing data including 3D LIDAR, stereo, video, and hyperspectral imagery.

Project Status: This is a three-year project initiated in the summer of 2005; this report covers activity during its first year. Three teams consisting of a Ph.D. student and their faculty advisor are performing research under this grant with overall leadership provided by the Principal Investigator Harvey Rhody (LIAS). The activity is summarized under the following three tasks.

Subtopics

Multi-modality Image Registration

Stereo Mosaic Extraction from Airborne Video Imagery

Semi-Automated DIRSIG Scene Modeling from 3D LIDAR and Passive Imaging Sources