Metrics for level 2 fusion association
Stephen Craig Stubberud, Peter J. Shea, Dale M. Klamer · 2003
In Level 2 Fusion, the concept of data association is much broader than in the Level 1 case. Apart from the concept of associating objects to one another, in Level 2, how the combination of lower level objects associate with interpretations of the scene is also important. One interpretation of this view is fhat the association among objecfs is performed through nonlinear and symbolic jimctions and that the solution con be assigned a symbolic infeipretation. To this end, we must create metrics or measures among objects and among objects and abstraction of the scene. In this paper, we introduce several mathematically based metrics for fhe association in the Level 2 problem. These measures have been developed using mathematical rigor.