AbstractThis thesis describes automated cartographic name placement using rule-based systems. In particular it describes the problem involved with designing a system which is flexible enough to place names on a variety of maps. This is demonstrated using logic programming techniques written in PROLOG.
Most previous name placement systems are either map specific or have demonstrated only a few aspects of name placement. However two of these systems, which use the rule-based approach for solving the name placement problem, do show greater flexibility. Nevertheless all known results from these seem unsophisticated when compared to many manually produced maps. This thesis describes further research into the use of rule-based systems. The systems described have the capability to handle a wider range of maps of greater complexity. Also described is a procedural program which implements an iterative strategy for name placement on the Ordnance Survey Route Planner map.
The research attempts to classify label positions and configurations used on a wide range of maps and discusses ways of implementing these in an automated name placement system. A range of name placement rules are also studied in order to decide what type of data a flexible automated name placement system must be able to access. A combined vector and raster data structure approach is adopted. This supplies the necessary "visual" information needed to apply most of the name placement rules. Name placement and database primitives are used to construct the high level rules which make up the rule-based systems.
This work has been undertaken in collaboration with the Ordnance Survey. The procedural name placement program, capable of placing names on the 1:625000 Route Planner map, has been implemented at their headquarters.
|Date of Award||Sep 1988|
- Digital mapping