An example of using ontologies and symbolic information in automatic target recognition
Mieczyslaw M. Kokar, Jiao Wang · 2002
The goal of th i paperi s to show an approach to target recogn iogn (ATR) that allows for e#ci t updati g of the recogn iogn algorio m of a fusi agent when new symboli i formati becomes avai le. Th i i formati may, for i stance, provi de addi ti onal characteri zati on of a known type of target, or supply a descri pti of a new type of target. The new symboli i formati can beei er posted on a web page or prov i ed by another agent. The sensoryi nformati can be obtai ed from twoi magi g sensors. In our scenar the fusi agent, after noti g such an event, processes the new symboli i formati and i corporatesi i to irecogn iogn rules. To achi e th i goal the fusi agent needs to "understand" the symboli i formati . Th i capabi yi s achi ed through the use of an ontology. Both the fusi on agent and the knowledge provi der (i t may be another software agent or a human annotator) know the ontology, and the web basedi nformati i s annotated usi ng that ontology. In thi s paper we descri be the approach, provi de examples of symboli target descri pti s, descr i e an ATR scenar and show some i i results of si ulati s for the selected scenar i- The di scussi i th i paper shows the advantages of the proposed approach over thati n whi ch the recogn iogn algorig m i fixed. 1