隨選網路地圖系統之設計考量

Design Considerations of an On-Demand Web Mapping System

目前國內外已有很多機關及組織利用網路，透過建立網路地圖系統將地圖資料提供給使用者使用。而網路地圖系統大多以多解析地圖資料庫做為地圖資料庫的核心架構。為了提供使用者“流暢”及“無縫”的地圖瀏覽品質，多解析地圖資料庫必需預先儲存大量已處理好的地圖資料，這勢必增加資料的儲存空間及導致資料庫維護的困難。再者，資料庫內所儲存的大量資料亦未必能滿足所有使用者的應用需求。相對地，單純利用即時縮編技術提供網路地圖資料，雖可節省資料儲存空間及增加系統的彈性，但目前線上地圖縮編仍無法做到完全自動化。本研究即探討如何結合即時縮編及多解析資料庫技術來建立一個兼具執行效率與使用彈性的隨選網路地圖系統，研究重點著重在系統架構及功能層面上的設計考量。最後，本文歸納出系統建置的要點，可做為後續系統發展之基礎。 At present, many organizations and institutions have established web-mapping systems to provide geo-information services by the internet. Most of them used multi-representation databases as the technical kennel. In order to provide “smooth” and “seamless” map browsing quality, a large amount of pre-processed data should be saved in the database. This would increase the need of memory space and cause the difficulty in database maintenance. Furthermore the data to be saved in the database may not satisfy the requirements of all users. On the contrary, web-mapping systems which use on-the-fly map generalization technique to provide multi-scale map services can reduce the need of memory space and improve the flexibility of system use. However, at present on-the-fly map generalization still cannot be fully automated. This paper explores the integration of multi-representation database and on-the-fly map generalization to overcome the drawback of existing web-mapping systems and concludes the key factors to develop an on-demand web mapping system.

At present, many organizations and institutions have established web-mapping systems to provide geo-information services by the internet. Most of them used multi-representation databases as the technical kennel. In order to provide “smooth” and “seamless” map browsing quality, a large amount of pre-processed data should be saved in the database. This would increase the need of memory space and cause the difficulty in database maintenance. Furthermore the data to be saved in the database may not satisfy the requirements of all users. On the contrary, web-mapping systems which use on-the-fly map generalization technique to provide multi-scale map services can reduce the need of memory space and improve the flexibility of system use. However, at present on-the-fly map generalization still cannot be fully automated. This paper explores the integration of multi-representation database and on-the-fly map generalization to overcome the drawback of existing web-mapping systems and concludes the key factors to develop an on-demand web mapping system.