Landmarks based route planning and linear path generation for mobile navigation applications

In this study a landmarks based navigation model considering local context is designed and a prototype is implemented. A route planning model where both the distance of a route and the strength of landmarks along the route are considered when deciding the optimum route to follow is developed. The Dijkstra algorithm is used as the basis and if the route is having more strength related to landmarks, it is prioritize while comparing with the distance. The route’s strength is defined based on number of landmarks visible along it and their salience. The horizontal spread, the height and the visibility of landmarks at different time of the day and the social/cultural significance are considered as attributes which give strength to a landmark. To utilize the mobile screen in effective manner, a map generalization is used to show the optimum path. With landmarks the map becomes congested and difficult to read on mobile screen. Therefore the path is reduced to a linear map which shows the path reducing curves while emphasizing the turns. Douglas–Peucker algorithm is used to derive the linear path. The linear map, which is a schematic representation of the actual map, reduces the path between junctions to straight lines and thereby uses less screen area. The implementation is done using mobile web approach and web services to reduce the platform dependency. OpenStreetMap is used as the main data source with sample local landmarks. All technologies used are free and open source. The application is still being verified and tested in Sri Lanka. © 2016, Korean Spatial Information Society.