Multisensor multiple-attribute data association

A multisensor system can provide a variety of target information (including dynamic parameters and attribute parameters). Target dynamic parameters are regarded as a kind of target kinematic attribute. A probability assignment method is explained in two cases both of process noise and measurement noise being Gaussis statistics and of being non-Gaussis statistics. A multisensor multiple-attribute data association method is presented based on Dempster and Shafer (D-S) evidence theory, and this approach is illustrated by simulations involving multisensor multiple targets in a dense clutter environment. Comparison with the NN (nearest-neighbour) method which only uses target dynamic parameters shows that the approach has an improved tracking accuracy and resolved correlation ambiguity.