ViewNet - Context enhanced networked services by fusing mobile vision and location

The aim of the project is two fold to develop a flexible adhoc open communications network archetecture allowing plug-in of different carriers such as TETRA, WI-FI, GSM etc and novel sensor technology for personal localisation which provides greater contextual detail than existing systems. Localisation data, from GPS for example, only becomes useful when it is placed in context. Currently this is done mainly through integration with geographical data, locating users with respect to landmarks such as roads and buildings. Although effective for some applications, such as vehicle navigation, this is not sufficiently dynamic nor detailed enough for potential applications in personal localisation, which need to operate at far higher resolutions, e.g. in terms of rooms, objects, walkways, etc. The innovation in this project will to address this by fusing localisation with real-time vision based mapping. The latter will give a 'live' 3-D representation of space, enabling users to be simultaneously localised in absolute terms and relative to their immediate physical surroundings, e.g. within a room or with respect to an object. The use of vision based mapping also means that the derived map will be directly related to a users' perceived sense of location, allowing easy visualisation of the detailed context being provided and direct reference to the 3-D map. The implementation will be based within a scalable network architecture, with integrated communications and supporting multiple users. This will facilitate distributed mapping, allowing operational teams to share visual and position data relative to a common map. The approach will provide a powerful framework for personal localisation, particularly for applications in security or disaster incident management, in which rapid surveying and visualisation of sites is of critical importance.