20.07.2013 Views

Download the report (PDF, 54.4 MB)

Download the report (PDF, 54.4 MB)

Download the report (PDF, 54.4 MB)

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

INTRODUCTION<br />

MAP GENERALIZATION:<br />

THEORY, PRACTICE AND ECONOMICS<br />

Joel L. Morrison<br />

The University of Wisconsin-Madison<br />

The <strong>the</strong>ory of map generalization is discussed with particular reference to <strong>the</strong><br />

processes of simplification and classification. Both of <strong>the</strong>se processes are used<br />

extensively in computer-assisted cartography; however, a comprehensive, flexible<br />

software package has not yet been organized from <strong>the</strong> myriad of algorithms available.<br />

Fur<strong>the</strong>r, systematic use of algorithms, based on <strong>the</strong> reasons for generalizing in car<br />

tography, is not common practice. ¥hen viewed in respect to <strong>the</strong> reasons for gener<br />

alization, <strong>the</strong> costs of using a <strong>the</strong>oretically based algorithm can easily be absorbed<br />

in most computer-assisted cartographic situations.<br />

Robinson and Sale in <strong>the</strong> third edition of Elements of Cartography define carto<br />

graphic generalization as being composed of four elements. These elements or pro<br />

cesses are simplication, classification, symbolization, and induction. They do not<br />

claim that <strong>the</strong>se four processes are independent of one ano<strong>the</strong>r, only that <strong>the</strong>y rep<br />

resent facets of <strong>the</strong> total process of cartographic generalization which cartographers<br />

can conveniently isolate and discuss. The definitions of <strong>the</strong>se four processes by-<br />

Robinson and Sale are not rigorous, but <strong>the</strong>y are <strong>the</strong> best available.<br />

Non-rigorous definitions are usually ignored when <strong>the</strong> cartographer interacts<br />

with automated systems to produce a map. Therefore, it is not out-of-line to sug<br />

gest that a non-rigorous definition of generalization and its elements is of little<br />

or no use for <strong>the</strong> purpose of making a map with computer assistance. One can use<br />

non-rigorous definitions only to discuss <strong>the</strong> map after a software manipulation se<br />

quence has been selected and/or <strong>the</strong> map plotted. What is basically needed is to de<br />

fine generalization so that <strong>the</strong> reason or purpose for making <strong>the</strong> map can aid <strong>the</strong><br />

cartographer in selecting <strong>the</strong> most appropriate software sequence for map production<br />

given <strong>the</strong> structure of stored data files and <strong>the</strong> available software manipulation<br />

sequences. To date, this luxury is not available.<br />

The intent of this paper is to offer more rigorous definitions for <strong>the</strong> purpose<br />

of classifying <strong>the</strong> processes involved in map generalization, to show how <strong>the</strong>se pro<br />

cesses have been and/or may be implemented in practice, and hopefully to give some<br />

insight into <strong>the</strong> economics of generalization algorithms relative to <strong>the</strong> total map<br />

preparation costs.<br />

EPITOR'S NOTE: Mi giaphicA<br />

A c.at& £01 publication<br />

99<br />

have, been reduced <strong>the</strong>.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!