Adaptive maps for mobile applications

The rapid development in the field of mobile devices has led to high acceptance and an increase of its application. These devices, namely Personal Digital Assistants (PDAs) and Smartphones, are used in private as well as in business sectors. The applications range from simple organiser applications to navigation software for car drivers as well as for pedestrians. Additional hardware like GPS-modules can be connected to the devices and make them more flexible in use. Therefore, these devices should be taken into account as a basis for digital cartography as well. But their input and output options significantly differ from desktop-PCs and web-applications. They are subjected to limitations like limited memory capacity and small displays including low resolution. Therefore techniques and experiences of designing useful user interfaces as well as the design of readable maps can’t be transferred easily from existing applications. The limitations mentioned have to be taken into account. In this paper we give a general overview of the status quo of mobile devices, their possibilities and restriction in terms of mobile cartography as well as their variations. We then describe an approach to address the restrictions of the small display and the limited memory capacity. As the amount of information the device can handle is limited, the load of information has to be reduced. For instance a clear map view can be kept by simplifying or leaving out unimportant information. To generate such a map, the system has to know, which information are relevant for the user and which are not. This decision is made by analysing the user’s request and by exploiting user profiles stored in a database. Finally, to provide detailed and less detailed geometries for generating maps containing objects with different levels of details, online generalisation techniques must be available to simplify the shape of the less important objects. But generalising geometries is time consuming and can’t be done online in all the cases. We propose an alternative way based on a Multiple Representation Database (MRDB). This kind of database includes several representations of the same real world objects. In our case it has objects of different levels of details derived by generalisation procedures. High resolution geometries will be used to visualise objects of interest, low resolution geometries reflect the other objects. Not only geometries but also attribute information can be accessed from the database with a flexible Level of detail (LoD).