Sensor and data fusion are exploited in diverse applications such as Earth
resource monitoring, weather forecasting, vehicular traffic management, and
target classification and state estimation. The approach used in this chapter to
describe data fusion and its objectives is based on a model developed for the U.S.
Department of Defense. The model divides data fusion into low-level and high-level
processes. Low-level processes support preprocessing of data and target
detection, classification, identification, and state estimation. High-level processes
support situation and impact refinement and fusion process refinement. The
duality between the data fusion and resource management models of processing
levels can lead to improved insight into and utilization of resource management
assets. Various categories of algorithms are available to implement target
detection, classification, and state-estimation fusion. In addition, several data
fusion architectures exist for combining sensor data in support of data fusion.
The architectures are differentiated by the amount of processing applied to the
sensor data before transmission to the fusion process, resolution of the data that
are combined, and the location of the data fusion process. The chapter concludes
by addressing several concerns associated with the fusion of multi-sensor data.
These encompass dissimilar sensor footprint sizes, sensor design and operational
constraints that affect data registration, transformation of measurements from one
coordinate system into another, and uncertainty in the location of the sensors.