Academic journal article Cartography and Geographic Information Science

Building and Road Generalization with the CHANGE Generalization Software Using Turkish Topographic Base Map Data

Academic journal article Cartography and Geographic Information Science

Building and Road Generalization with the CHANGE Generalization Software Using Turkish Topographic Base Map Data

Article excerpt

Introduction

Computer technology has been widely used in cartography since the 1970s. In generalization, however, which is one of the major topics of cartography, there is still no adequate solution in a scientific and/or practical sense. New technologies have led to a rapid development in the field of data capture in cartography and related disciplines. In developed countries in particular, the era of data capture is almost complete. National mapping agencies have built large databases that include mostly large-scale data, i.e., in base map resolution. For applications at smaller scales, existing large-scale data should be used through digital generalization, to avoid redundancies in data capture and to ensure consistency among data sets.

The Institute for Cartography and Geoinformatics (IKG) of the University of Hanover (formerly the Institute of Cartography (IfK)) has dealt very, intensively with the generalization of large-scale maps. The Institute developed a software package, CHANGE, which is offered commercially under different operating systems. This software package includes generalization algorithms for roads and buildings. Some authors have already demonstrated that CHANGE delivers cartographically good results in the generalization of roads and buildings (Bucher 1998; Powitz 1993).

Recent scientific research on building and road generalization can be summarized as follows. Jones et al. (1995) developed an innovative approach for the generalization of large-scale data using a simplified data structure based on triangulation of map objects. They applied this approach to building and road generalization. Their approach is also appropriate to create centerlines of roads. Thomas (1998) developed another approach to centerline generation for roads, transforming vector data to raster data. Plazanet et al. (1995) focused on the generalization of linear data, aiming particularly at medium- to small-scale generalization of road data. Efforts are under way to solve spatial conflicts of road and building objects (displacement problem): Harrie (1999) uses least square adjustment and Hojholt (2000) tries to solve the problem with finite element methods.

This paper reports on a case study carried out at IKG with the support of a scholarship from the German Academic Exchange Service (DAAD). Based on this study, the author's dissertation was completed at the Technical University of Istanbul (Bildirici 2000). The aim of the study was to generalize data (buildings and roads at the scale of 1:1,000) from Turkey with CHANGE, to make improvements and suggestions if necessary, to examine the optimal utilization of CHANGE with computer-aided design (CAD) and geographic information system (GIS) software, AutoCAD MAP, to develop required support programs, and finally to implement methodological suggestions. The AutoCAD MAP can be described as a graphic interactive interface with exceptionally efficient query, topology, and line cleaning tools, especially for road objects. Support programs were developed for this study using the FORTRAN 90 programming language.

Using the approach proposed in this paper, building and road data required for such applications as city map production at scales of 1:15,000-1:25,000 could be automatically obtained from scales 1:1,000-1:5,000. These data can also be used for updating the settlement part of the national map set of Turkey at 1:25,000-scale.

Generalization Software: CHANGE

Brief History

For over 30 years the IKG has been focused on the field of automated generalization of large-scale topographic maps. Generalization of two feature types roads and buildings, was examined intensively. Staufenbiel (1973) developed algorithms for the simplification of building outlines and the comprehension of buildings (Figure 1). The term "comprehension" means in this context merging or grouping buildings. Bildirici and Ucar (2001) discuss Staufenbiel's algorithms extensively and suggest improvements. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.