Characterising Space via Pattern Recognition Techniques: Identifying Patterns in Road Networks

Publisher Summary The chapter introduces the subject of pattern recognition in road networks. It describes the concepts of random and scale-free graphs, since scale-free graphs contain important information that can be used to control generalization processes. The legibility of maps and their accuracy with respect to content depends to a great extent on the quality of the original data. However, the presentation and thematic fidelity of the data can vary considerably depending on the scale and the theme. The generalization process seeks to highlight salient information while omitting information not pertinent to the particular theme. Different criteria are used to control elimination, presentation, and emphasis of geometrical elements—criteria such as measures of object density and shape of geometric elements. Approaches to the automatic detection of these patterns in vector data are reviewed and different patterns of graphs are presented. The chapter also presents methods for identifying the center of cities using a combination of various pattern characteristics. It is argued that pattern recognition techniques are critical to the automatic characterization and generalization of higher order structures in geographical data.