An ontology-driven multi-agent system for nautical chart generalization

Abstract

On nautical charts, undersea features are portrayed by sets of soundings (depth points) and isobaths (depth contours) from which map readers can interpret undersea features. Different techniques were developed for automatic sounding selection and isobath generalization. These methods are mainly used to generate a new chart from the bathymetric database or from a larger scale chart through selection and simplification. However a part of the process consists in selecting and emphasizing undersea features formed by groups of soundings and isobaths on the chart according to their relevance to maritime navigation. Hence automation of the process requires classification of features and their generalization through the application of a set of operators according not only to geometric constraints but also to their meaning.The objective of this work is to propose a multi-agent system for nautical chart generalization that is driven by the knowledge on the generalization process and the undersea features and their relationships. Firstly, this work provides a featurecentered ontology modeling the generalization process. Then, the MAS structure is introduced where agents access cartographic knowledge stored in the ontology. The MAS makes use of measure algorithms to evaluate constraint violations on the chart in order to decide which generalization operators to apply. The whole model has been implemented to provide generalization plans on a real case study.

Automatic (Tactile) Map Generation—A Systematic Literature Review

“…A "micro" agent is a single geographic object (e.g., building), while a "meso" (macro) agent is a composition of "micro" or "meso" agents, as for generalization purposes they have to be considered together (e.g., a set of buildings in a neighborhood). This model is best used in dense [54] and well-structured urban areas [39]. On the other hand, the GAEL model (Generalisation based on Agents and Elasticity) is an extension of the Agent-based model for management of background themes such as relief.…”

“…The balance between them is required for successful generalization [55]. GAEL is often used for terrain models such as relief or land cover [54]. The CartACom model relies on communicating agents and was designed to handle unstructured geographic spaces with no clear borders between groups of objects [56].…”

“…There are three types of them: legibility constraints, constraints of preservation and of geographic coherence [58]. This model is best used for low density [54], heterogeneous rural areas [39]. In Touya and Duchêne [59] the authors argue that it is impossible to solve the generalization problem with a single model.…”

“…The authors of identified primary studies used various software. Some authors mentioned their own dedicated software or system [54,[69][70][71]. In most cases, however, researchers used commercial software and, if necessary, modified it to suit their needs.…”

A Review of Geospatial Semantic Information Modeling and Elicitation Approaches

“…Gould and Mackaness [74] also include generalization algorithms in their ontology in order to have a knowledge base that could be used in an automated system. Yan et al [75] propose an ontology for the generalization of isobaths on charts. In their ontology, they not only store cartographic elements such as soundings and isobaths, but also include features formed by groups of isobaths and soundings representing submarine features as perceived by the readers.…”