Research On Key Techniques Of Chart Design Automation

Chart is the carrier of marine geo-information and the indispensable tool of marine activities. Also, it is an information channel between cartographers and chart users. The competence of chart design directly determines the quality and employment of the charts. Chart design, which is the most creative work, is the comprehensive definition for all the methods, processes and techniques in designing charts. It is not only includes producing a single piece of nautical chart, but also includes designing and organizing the whole chart production system. Currently, chart design's development still lags behind, though marine charts producing mode have gradually turned from manual operations into digital charting. As chart designing is still in the phase of relying on combination of traditional manual and computer-aided design, it significantly suspends advancing the progress of the marine charts auto-production system. Hence, this paper focuses on the thesis of chart design automation, standing on the current realistic situations and potential practical demands of chart design, and emphasizes on the research work of following key techniques, as material selection and processing, chart sheet design, automatic generalization of coastline and the construction of new production system. Main topics lie in following points:(1) Define the detailed concepts, contents and processes of chart design. Introduce the spacial cognition theory, map information transmission theory and map visual perception theory into the research of chart design. On the basis of analysis the chart design's current situations and potential development, the paper points out discrepancies in the current chart design process and hence educed the focusing research contents.(2) Accoring to the point of map information transmission theory, in the light of characteristics of the chart's information quantity changed resulting after the chart revision, specific measurement methods of chart elment semantics, positions and annotation information quantities can be induced. Also, the chart revision necessity assessment method which bases on changing information quantity variables is put forward. Hence, quantitative basis for decision making on current manual experience of chart revision is proposed. (3) In regards of the current situation using manual analysis all along in chart material selection, visualizing analysis platform and knowledge rule based material selection method is proposed on the summarization of the chart material evaluation standards and methods. The method extracts and summarizes the chart material selection knowledge, and makes specific quantization on the major knowledge rules in material selection process, then realizing the automatic selection of the basic material according to quantized indexes by using fuzzy comprehensive evaluation method.(4) In view of the discrepancies of tedious multiple work steps and poor multi-source data reusability in traditional chart design processing, multi-source data integration scheme for the purpose of digital chart design is presented. The specific integration method is proposed according to the scheme on following integration issues, as unification of spatial datum, data model and feature coding unification. Also, the integration experiment is carried out by means of data format conversion, which lays the foundation for digital chart production and further chart data fusion.(5) Considering the free chart framing features in chart design, relative framing mathematics basis is put forward based on the chart framing principles and methods. Additionally, computer-aided framing design is realized regarding the different factors (including geographical area, chart size, reference latitude and chart scale) and map forms is deduced.(6) Automatic generalization is the most creative and challenging process in charting, the experiment of cartographic generalization is also a very important part of chart design. On the basis of the thought breaking through elements, coastline generation method considering bend skeleton as simplised benchmarks is proposed. The method focuses on the discrepancies of the current automatic generalization and simulates the manual chart design thinking of cartographic generalization. First, identification on coastline bend based on curve monotone. Then, distilling bend skeleton based on triangular construction. Regarding principle of "enlarging land area and reducing sea area", automatic generalization experiment and analysis is carried out to verify the effectiveness and validity of the proposed method. (7) Regarding the issues as too many duplicate work steps, multifarious data update, element unification difficulty for the series charts and the different scale charts, draggled management system, new chart production system is proposed under the support of construction of major database. By studying the new system's software platform, data model and database and standards, new system's production process and framework is advanced. And specific analysis on key techniques to solve issues on new system is also brought forward.