Continuous Road Network Generalization

This paper presents our research efforts to obtain a sensible vario-scale representation of a road network. A vario-scale representation makes it possible to get maps at an arbitrary map scale, being more flexible with how much detail is shown during the use of geographic data. For this, an offline process creates a suitable vario-scale data structure (called tGAP), that then can be used for generating the maps at arbitrary map scale. Our intent is to design a generalisation process from (very) large scale data (e.g. 1:2,500,where roads are present as areas) to small scale data (where roads are present as lines, e.g. 1:1M). In this work we focus on the scale range where roads are present as lines (mid scale), but our intention is to start with large scale data where roads are represented as areas. At the heart of our approach we work with the area objects that are defined by the bounding road segments (the ‘cycles’ of the road network), but for determining the feature class of a merged edge, we fall back to an approach that is inspired by strokes. This together gives a solution for continuous road network generalization that aligns with our earlier attempts for creating vario-scale data structures. 1Section of GIS technology, OTB Research for the Built Environment, Delft University of Technology, The Netherlands e-mail: [email protected], e-mail: [email protected], e-mail: [email protected], e-mail: [email protected] 2School of Resource and Environmental Science, Wuhan University, Wuhan, 430079 China,