The vision of GeoDec is to enable an information-rich and realistic 3-dimensional visualization and/or simulation of a geographical location (e.g., a city), rapidly and accurately. The idea is not just to allow navigation through a 3-D model, but to be able to ask queries and get information about the area seamlessly and effortlessly. The main challenge is: How to quickly and cheaply integrate, visualize, and simulate all aspects of a geographic region. The ability to create high-fidelity information-rich models of cities is critical for a wide variety of decision makers. For example, in the United States, GeoDec can be used by city managers, city planners, emergency response planners, and first responders.
In this talk, I will first motivate GeoDec and explain its vision. Next, I will briefly discuss our current activities in the following areas towards the realization of GeoDec:
Rapid construction of accurate 3-D models of buildings from two aerial images with minimum user’s interactions.
Visualization and accurate rendering of static and dynamic (i.e., [live] video) textures on the constructed 3-D models.
Automatic integration and accurate fusion of various geospatial (e.g., maps, road-networks) and geospatially-related (e.g., web yellowpages) data into the models.
Finally, I will conclude with a demonstration of the current prototype of GeoDec.