Interactions of dynamic geospatial objects with static landmarks

Analyzing the behavior of moving objects has multitude of applications e.g. in the area of transportation. Each application might require identification of different behavior patterns and their relationships to different landmarks. Machine learning algorithms can help in automatic recognition of spatiotemporal patterns. However this is still a largely unsolved problem, especially identification of the relationships of moving point objects with stationary objects or landmarks on a map. In our project we considered dynamic objects such as cars and humans on a terrain with static elements such as road networks and buildings e.g. airports, bus stops etc. We created application specific ontologies of patterns of moving objects in relation to static landmarks. Based on ontologies we built machine learning models to classify trajectories of moving objects.