chapter  2
Concepts and Theory of Data Fusion
Pages 52

Data fusion (DF) or multisensor data fusion (MSDF) is the process of combining or integrating measured or preprocessed data or information originating from different active or passive sensors or sources to produce a more specifi c, comprehensive, and unifi ed dataset or world model about an entity or event of interest that has been observed [1-8]. A conceptual DF chain is depicted in Figure 2.1, wherein the fusion symbol is indicative of the fusion process: addition, multiplication (through operations involving probabilities, e.g. Bayesian or Dempster-Shafer [DS] fusion rule), or logical derivation; in addition, the hierarchy is indicated by two circles. This implies that fusion is not just an additive process. If successful, fusion should achieve improved accuracy (reduce the uncertainty of predicting the state or declaring the identity of the observed object) and more specifi c inferences than could be achieved using a single sensor alone. Multiple sensors can be arranged and confi gured in a certain manner to obtain the desired results in terms of sensor nodes or decision connectivity. Occasionally, the arrangement is dictated by the geometrical or geographical disposition of the available sensors, such as radars.