GENERALIZATION OF MAP DETAILS WITHIN
COMPUTER ASSISTED CARTOGRAPHY
by
Matharuddin b. Yusuf
Jabatan Fotogrametrl dan Kartografi,
Fakulti Ukur, Unlversiti
Teknologi Malaysia,
Abstract
Throughout the century many cartographers have addressed the issue of
subjectivity in generalization, the metamorphasis from manual to digital techniques
in cartography has resulted in the development of computer algorithms designed to
replicate many tasks previously performed by humans. In manual generalization
cartographers have taken an intrisically hollistic approach where, when
simpilifying manuscript line, many decisions a r e applied al once: the important
characteristics are selected, these important characteristics might be exaggerated,
and the unwanted detail is elimited all in one swoop of a pen. In contrast to manual
generalization the computer approach requires separate algorithm to complete each
component in t h e process: simplification, classification, simbolization, smoothing,
diplacement and enhancement. Most of the research effort on automated
generalization has been directed towards areas of linear simplification.
This paper begins with a review of the principles and concepts of Generalization and
proceeds with the problems inloved in automating the generalization process.
1.0 Principles And Concept of Generalisation
1.1 General
The producing a map at a smaller scale than the existing map or in
construction a map from raw data, the cartographer traditionally employs
subjective judgement to perform the necessary generalisation This subjectivity is an
individual characteristic of generalisation will also vary among various map
products. {See Steward, 1974).
Generalisation belongs to the most important decisive concepts of
cartography. It is a fundamental ingredient in the preparation of maps because of the
necessity of retaining legibility as the scale is reduced. In most cases the method
adopted has been empirical and subjective but there is the increasing interest in
objective and theoritical approaches especially since introduction of automation
(computer) in cartography. Most advances have been acheved in the estimation of the
number or length of features to be retained but the choice of which individual
features which would make up the number is harder to define.