Much work has been done in the area of qualitative spatial reasoning over the past years, with application in various domains. However, existing models only capture particular aspects of the spatial relations between objects and therefore are unable to represent these relations accurately. In this paper, we draw together two existing approaches to provide a calculus taking into account topology, orientation and distances, while keeping in mind cognitive considerations. Provided that some constraints are imposed on the spatial objects and the frame of references, our model has been successfully tested to infer implicit constraints from a knowledge base. With further improvements, it has applications to GISs and robot navigation.