Sensor systems for urban operations - Annual report

This annual report gives an account of the activities carried out and the results produced in the three-year Armed Forces' project Sensor Systems for Urban Operations (SUO). The aim of the project is to study and evaluate how various combinations of sensor systems can contribute to an improved situation awareness in the urban environment, compared to the achievement of the single sensors. The research efforts in the project are focused towards the following important areas of needs: Continuous Surveillance, Crowd Surveillance and Red and Blue Force Tracking. The SUO project directs its research efforts towards solving problems in connection with sensor co-operation in the complex urban environment, where the focus is on the following areas: Detection/tracking/classification of human movement, Data fusion in multi-sensor networks, Sensors in co-operation, and Anomalies detection, i.e. the capability of detecting deviations from the normal picture. The project participates in several international collaborations. Through co-financing of the three EU FP7 projects Prometheus, ADABTS, and IMSK, the project gets access to the research and the results generated in these. Furthermore, co-operation is financed in the recently started NATO group SET-153 RTG-086 Multi-sensor integration in urban operations, whose aim and direction is in line with that of the SUO-project. In Continuous Surveillance an earlier developed general multi target tracker has been adapted to tracking of people and has been improved. This development is based on data collected in our own field tests and tests conducted within Prometheus. The introduction of colour histograms in the tracker has given a more robust tracking of temporarily concealed persons or when people move between different cameras. Fusion of auditory and visual data has also been tested with positive result. In a partial study an investigation is made of how good acoustic separation/resolution is achievable with and without, respectively, models of the environment in urban scenarios. Acoustic microphone arrays with exceptionally good directive properties are the primary sensor here. The sub-project Crowd Surveillance studies and develops methods for detection and tracking of the crowd but also single individuals being part of a crowd. A method has been developed which estimates the probability of normal behaviour of a crowd, and that detects deviation from the expected normal state. Data from simultaneous recordings with combinations of visual cameras, IR cameras and acoustic arrays have formed the basis of this development. In field tests in January, scenes of crowds were recorded with visual cameras, an IR camera and a 3D laser from the visiting ARL. A main objective of these tests was to evaluate how the distance information from the 3D laser can be used for improved detection, tracking and state estimation. In the sub-project Anomalies detection in urban environmet (AD) work has proceeded with the study and development of methods and system concepts for automatic detection of interesting, deviating events and courses of events in urban environment. The sub-project Red and blue force tracking investigates possible techniques for the identification and tracking of the movements of one's own troop in the urban scenario, especially in buildings without GPS coverage. In an initial experiment, we have pointed to the possibility of using hand-carried optical systems enabling a soldier to determine his or her own position, and simultaneously to construct a map of the interior of the building through the use of the SLAM method. A completely new "silent" method for positioning of one's own soldiers in buildings has been studied, based on a combination of RF/RFID. A test system has been developed and the method will be validated experimentally.