Sensor Data Fusion and Integration with Applications to Target Tracking and Robotics

Sensor integration and registration is a prerequisites to exploiting the inherent advantages of multi-sensor systems over single sensor systems. Using a single sensor, we can monitor objects with a precision and accuracy that depended on the sensor characteristics. By using multiple sensors to observe and object, we can obtain multiple viewpoints, extended coverage both spatially and temporally, reduce the ambiguity and obtain more precise estimate of object kinematics than that is possible through the best individual sensor. Engineers can replace a single very expensive sensor with many cheaper sensors in a tracking scenario or employ a variety of sensor to construct a complete view of the robot's environment. This certainly is the case with of a netted sensory system. For example, a single sensor may have a blind azimuth *screening angle) which an adjacent sensor may cover. In those areas where sensor coverage overlaps, the quality of which not only improve our estimate of object kinematics, but also help with its detection when the environment is changing. Form the military point of view, multiple sensors provide diverse information, which can be used by the decision-makers to derive an appropriate response to perceived threats. As the number of threats, or objects in robot's workspace, being monitored increases, the difficulty in maintaining techniques capable of functioning in a cluttered, dynamic environment containing the objects of interest is of fundamental importance to enhancing the survivability and usefulness of multi-sensor system.