Fusing mobile vision and location technologies to create context-enhanced networked services

The Problem

Emergency service personnel often enter a scene without adequate maps. They report movements and surroundings to a control centre and this is used to direct the response. In ViewNet, sensor networks will provide distributed localisation and mapping. By fusing visual mapping with RF localisation, each operative will be located absolutely and in relation to key features.

ViewNet aims to improve the effectiveness and dynamic cooperation between emergency services by enabling new models of collaboration between personnel and providing enhanced knowledge presentation and sharing in increasingly complex chains of command.

The Research

The objective of the project is to develop a demonstrator TETRA radio system that will integrate new wireless technologies and GPS absolute positioning with vision-based localisation and 3D mapping and data from inertial sensors.

The demonstrator will meet the customers’ requirements for enhanced location information so that an upgrade to the security services’ TETRA radios can be swiftly and efficiently introduced, improving efficiency, saving money and potentially lives.

The research will be concentrated in several key areas:

• the development of a network architecture for localisation and mapping capable of supporting multiple operatives remotely, and thus allowing distributed mapping
• vision-based simultaneous localisation and mapping, giving estimates of a camera’s 3D position and orientation in tandem with depth estimates for salient features of the scene
• developing stochastic fusion algorithms for integrating absolute position and relative localisation and mapping data, initially for single sensor use and then for multiple, distributed, mobile networks
• Sub-work packages will look at the interfaces to link with the HOLMES2 system used by the police and other issues that will be key to exploiting the IP.