Please use this identifier to cite or link to this item:
Title: A multi-agent system for generalization of features defined by isobaths in nautical chart construction
Authors: Zhang, Xunruo
Degree: M.Phil.
Issue Date: 2013
Abstract: Nautical charts provide a schematic representation of the seafloor and are used by navigators to plan their routes and to identify navigation hazards. Nautical charts are official documents issued by national hydrographic offices which are in charge of collecting the data and producing the charts. One of the most important stages in the production process is the chart generalization. It consists in abstracting the chart to emphasize important items and increase its legibility. The cartographer has not only to adapt the amount of information to the scale of the chart but also to select the information according to the types of features on the seabed and their relevance for navigation. Methods usually applied for contour generalization on topographic maps cannot be applied on isobaths directly because they do not consider the type of terrain features characterized by the isobaths. Therefore, a strategy is needed for the generalization of isobaths that fits nautical chart requirements. This dissertation focuses on a new generalization approach where features formed by groups of isobaths are identified and classified in a hierarchical structure based on their inclusion and elevation. Generalization constraints are defined according to the type of feature (pit or peak) and specific generalization operators are defined and applied according to the constraints. After exposing the specificities of nautical charts and reviewing the main techniques developed in line generalization, a list of generalization operators that apply to isobaths delineating features is defined. Aggregation and enlargement operations are based on a snake model where a system is stable when its energy is minimal. Violating a cartographic constraint adds energy to the line which has to be deformed in order to find a stable position. In order to automate the application of generalization operations, a strategy based on a Multi Agent System (MAS) is developed. Each feature and each isobath is represented by an agent. Each feature agent is able to evaluate its situation with regard to its own attributes and other agents. Based on this evaluation, a feature can set up different plans of action corresponding to the generalization operators that apply to the type of feature. An evaluation of the different plans is performed and the one minimizing constraint violations is selected. The method, combining the snake model and the MAS is applied on a set of isobaths provided by the French Hydrographic Office. Results are discussed and conclusions and directions for future works are presented in a last section.
Subjects: Nautical charts.
Nautical charts -- Data processing.
Hong Kong Polytechnic University -- Dissertations
Pages: 88 p. : ill. ; 30 cm.
Appears in Collections:Thesis

Show full item record

Page views

Citations as of May 15, 2022

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.