4.3
Low-cost urban data sets through LIDAR feature extraction
PAPER WITHDRAWN
Kevin S. Williams, SAIC, Chantilly, VA
As weather models become increasingly detailed and computers become more capable, the lack of urban data becomes more and more noticeable as the limiting factor in today's urban weather models. There are very few high-resolution urban databases to use in models because digitizing urban features has been a time-consuming and expensive process in the past. The advent of LIDAR data along with SAIC's new automated feature-extraction technology promises to revolutionize this process.
The RTV/SAIC LIDAR Toolkit is a set of powerful algorithms useful in extracting features from LIDAR data. Originally developed for military purposes, its scope and functionality has expanded to encompass a broad range of applications. It now includes a variety of fully automated tools that can perform a wide range of tasks: extraction of bare earth surface models, mapping of high-detail building structures, delineation and characterization of individual trees and shrubs, detection of road networks, and numerous other analyses.
The LIDAR Toolkit has been used operationally to extract urban features for several projects. Most recently, it was used to support DTRA’s dispersion modeling effort over Salt Lake City in preparation for the Winter Olympics. In this project alone, over 60,000 individual buildings were extracted for a fraction of the time and cost of traditional extraction methods. This presentation will describe the existing LIDAR Toolkit functions as well as near-future development efforts to improve extraction capabilities.
Session 4, Urban Surface Databases
Monday, 20 May 2002, 3:30 PM-5:15 PM
Previous paper Next paper