Methodology for Automating the Cartometric Evaluation of Urban Topographic Maps of Brazil

Abstract

Cartometric evaluation is part of the cartographic generalization process and can indicate which operator should be applied and which features should be generalized. Those decisions are based on the occurrence of geometric conditions and spatial and holistic measures. Those aspects can be understood as legibility and visibility problems, and this knowledge can be described as rules of an expert system. The automation of cartographic generalization is a challenge partly because of subjective decisions made during the process. Cartometric evaluation show this subjectivity. The formalization and automated identification of the representation problems can make this process more holistic and less dependent on human control and influence and, therefore, more efficient. The objective of this paper is to present a method for developing a cartometric evaluation for topographic maps at 1:5000 derived from 1:2000 scale that is appropriate for the Brazilian landscape. Our approach consists of designing an expert system based on a decision tree. We present the first results of the expert system developed using ArcGIS and its application ModelBuilder to detect automatically map representation problems.