Figure 2 The original image and result after amalgamation has been applied on 
two objects in real raster data base. 
The newly created areas can also in some cases be 
kept for the purpose of performing simplification of 
the form of the original objects. An acceptance rule 
for this could be that more than 75 % of the borders 
pixels of the new area are neighbours with one of the 
original objects. Another rule also has to be used 
here to ensure that the new area is not covering 
something of importance. 
The method presented in this paper could also be 
used for exaggerating one object toward another 
object, e.g., if a wetland area that is close a river, but 
has a narrow strip of open fields in between. In this 
case the amalgamation could be carried out between 
the wetland and water objects. The newly created 
area between the objects can be reclassified to wetland 
after amalgamation and acceptance check. 
The methodology shown here is general in the sense 
that it could be used in several application areas. The 
system design is aimed at generalizing topographic 
area data. However, this strategy could also be used 
as a post processing step after classification of satellite 
images. 
ACKNOWLEDGEMENTS 
The support from the National Land Survey of 
Sweden who financed this work is greatly acknow- 
ledged. I would especially like to thank Dr. Bengt 
Rystedt and my colleges at the National Land Survey 
in Gávle for inspiring discussions about digital 
cartography. Ialso like to thank Dr. Hans Hauska for 
the comments on the paper. 
138 
  
REFERENCES 
Monmonier, M.,1983. Raster Area Generalization for 
Land Use and Land Cover Maps. Cartographica, Vol 
20, No. 4, pp. 65 - 91. 
Person, J and Jungert, E., 1991. Generalization of 
Multi-Resolution Maps. FOA Report C 30609-3.4, 
ISSN 0347-3708, 66 p. 
Shea K.S., and McMaster R., 1989. Cartographic 
Generalization In a Digital Environment: When and 
How to Generalize. Technical Papers 1989 
ASPRS/ACSM Annual Convention, April 2-7. pp. 
56-67. 
Weber, W., 1982. Automationsgestützte Generalisie- 
rung. Nachrichten aus dem Karten- und Ver- 
messungswesen, Heft Nr. 88 ISSN 0469-4236 Verlag 
des Instituts für Angewandte Geodäsie Frankfurt 
A.M. pp 77-109. 
Weibel, R., 1991. Specifications for a platform to 
support research in map generalization. . Department 
of Geography, University of Zürich, Switzerland, 
Discussion paper presented at the ICA conference 
Bournemouth, 12p. 
Westervelt, J., 1991, (Editor). Grass 4.0 Users Manual, 
U.S. Army of Construction Engineering Research 
Laboratory, Champaign, Illinois, USA. 
ABSTRA 
Research 
This pape 
The seco 
photograr 
generalize 
Special € 
generalize 
attribute c 
KEY WOI 
Computer 
topic for | 
appeared 
restriction 
of these | 
became c 
stages. 
Late in th 
times note 
grate thes 
gress on 
applicatior 
Carl Zeiss 
generaliza 
veloped b 
and is fully 
cartograpt 
2. COMP 
2.1 Relev 
The use c 
mental che 
forms of p 
what rele 
have. Is i 
Systems? 
obvious: 1 
siderably ir 
* Graphi 
require 
data (fi 
* The us 
hardwe 
to be k 
* The co 
acquisi 
data be 
use of 
Furthermor 
maps form 
higher the 
need for th 
2.2 Exemy 
The theory 
ferent appr 
be distingui