A Comparative Study on Urban Visualization Using LIDAR Data in GIS

Light Detection and Ranging (lidar) is a remote sensing technique utilizing laser technology that has found applications in a wide variety of fi elds. This article focuses on the application of lidar data and digital imagery to three-dimensional urban visualization in a geographic information system (GIS) environment. An ArcView GIS package with advanced 3-D analysis and modeling extensions is used for the study. Both interactive and automatic approaches are investigated and compared. The problems encountered are analyzed. An integration of both approaches provided satisfactory results. This study shows that existing powerful GIS tools can be effectively used to process urban lidar data so that most urban buildings can be modeled and visualized satisfactorily through an automated process. Manual delineation is needed to precisely defi ne building footprints and separate the buildings from the ground. Results from the built-up urban area of downtown Baltimore, Maryland are presented to support the analyses. independent, is relatively weather independent, and is extremely precise (3Di, LLC 2000). In addition, because lidar operates at much shorter wavelengths, it has higher accuracy and resolution than microwave radar (Jelalian 1992). A number of publications have addressed the use of lidar data in many different fi elds. Urban planning has been identifi ed as a major benefactor of realistic visualization. Lidar has been used for topographic mapping of forested terrain and other areas not suitable for aerial photography (Wever and Lindenberger 1999). As each data point is georeferenced, the lidar data can also be easily merged with other data sources (Kletzli and Peterson 1998). Hug (1997) stated that laser scanners are the best choice for obtaining digital surface models, especially for dense urban areas. Haala and Brenner (1997) reported on similar work that uses airborne lidar data for the generation of 3-D city models. Kim et al. (2000) provided a concise examination of using photogrammetric imagery and lidar data for obtaining a DTM in urban areas. Förstner (1999) presented a thorough and informative discussion of the problems encountered in acquiring and establishing the building models. Although fully automatic techniques are improving, a review of extant acquisition systems has revealed that, up to the date of writing, the systems have not proved to be reliable enough to be used alone. Another interested party is the telecommunication industry that uses information in 3-D city models for planning the locations of antennas (Brenner 1999, Kirtner 2000). The desired ultimate outcome of urban modeling is realistic visualization (Danahy 1999, Bhagawati 2000). Fritsch (1999) stated that the overlay of laser scan data with digital aerial imagery delivered the fi rst virtual 3-D model. It has been pointed out that as many available data sources as possible ought to be used to obtain a truly virtual urban model through the data fusion process of ground plans, aerial photographs, and laser scanning (Förstner 1999, Fritsch 1999, Toth and Grejner-Brzezinska 2000) The objective of this work is to study the methodology and effi ciency of urban modeling and visualization by integrating