Research On Target Classification Technology In Situation Assessment And System Testing

Situation assessment(SA) is the key technique for high level processing ofinformation fusion，which is the level2fusion in JDL model. Situation assessment is theongoing process of inferring relevant information about forces of concern in abattlefield situation to achieve situation awareness, which is needed by the campaigncommanders or analysts to support decision-making. The analysis is made of data fromsensors and other sources in SA for building an accurate and timely picture of thebattlefield situation, including location, movements, deployment and intension ofenemy force.In the fast-paced, modern battlefield, target classification will greatly reduce thefocus of decision-makers. On the other hand, as the combat group is a reflection of thecollaborative relationship between entities. Therefore, it also reflected the intention andthreat of situation information.First, this paper gives a brief introduction to target classification, aero naval targetclassification base on nearest-neighbor+template is proposed to overcome theshortcoming of the nearest neighbor method, and the accuracy of classification isimproved.Land forces clustering and command post recognition is researched, aims atmeeting the demand of land situation analysis. The problem to recognize the commandpost and force deployment is solved by the military knowledge template. Meanwhile,correlative technologies realized ground force classification is discussed.Finally, three scenes are created to test the algorithms for ground forceclassification, track and track fusion method, event detection, target classification andplan recognition. Through the simulation experiments, algorithm models used in thesystem are feasibility and effectiveness.