A Review and Conceptual Framework for Generalization of Maps

Generally, National Mapping Agencies (NMAs) produce and maintain topographic and cadastral maps. Regular update of map products is the most critical problem faced by NMAs. Therefore, implementation of new technology to automate map generalization has become crucial in updating smaller-scale maps from frequently updated larger-scale maps. Additionally, customized data products can be produced. Implementation of single database and derivation (generalization) of multiple-scale maps from one database can improve the efficiency and maintenance of derived data products. However, consensus between NMAs on a common generalization technique has not been achieved due to incompatibility and complexity of data models. Definitions of terms, technology and organizational structures of the NMAs across the globe are different. This chapter reviews the pertinent generalization efforts made by a number of European NMAs. A framework is proposed based on two correlated models: process-oriented generalization models and object-level-oriented generalization models, which is related to level of map objects. A portion of this framework is implemented and tested on datasets of Kelantan state in Malaysia to produce generalized maps. Generalization operations (e.g., simplification, aggregation, smoothing, etc.) are applied with the aim to generalize polygons and transportation network (e.g., roads, etc.). The proposed framework could be adopted, implemented and tested on datasets maintained by Malaysian National Mapping Agency known as JUPEM (Portal Rasmi Jabatan Ukur Dan Pemetaan Malaysia).