Modeling urban landscapes with integrated datasets

Abstract

Three‐dimensional (3D) modeling of city landscapes has became affordable since 3D mapping and visualization technologies matured. However, a variety of data sources may be involved in a project of landscape modeling due to the complexity of urban landscapes. Under this circumstance, one may face the problem of how different data sources are used and integrated. This paper introduces useful datasets for landscape modeling and proposes a scheme for the integration of multi‐source datasets. In general, the terrain, buildings and other objects are the three major components to model for an urban landscape. This study develops a coarse‐to‐fine modeling procedure with respect to the three components by taking advantage of various data sources. The data integration was achieved by transforming the coordinate systems of applied datasets into a unified frame. In this study, the datasets of topographic maps, aerial images, airborne LiDAR (Light Detection and Ranging), close‐range images and terrestrial LiDAR were employed to model the main campus of NCKU. The quality of data integration was investigated and the results of intermediate phases are presented to demonstrate the feasibility of the proposed method. Issues related to manpower requirements are discussed as well.