Attributes in tracking and classification with incomplete data

摘要: An obvious use for feature and attribute data is classification in combat identification. The term used broadly here to include discrimination, detection, target typing, identification, pattern recognition. additional the (or track) association process reduce misassociations, often called aided tracking. Previous papers discussed integration of features attributes into track processing addition their multiple classification. distinction made between because they are processed differently. applies from continuous sample space discrete space. primary concern this paper address when there incomplete data. For example fusion sensor distributed sensors, a processor might make hard decision on whether exhibits specific class value) or not (without providing information such as likelihoods probabilities). In another example, distribute likelihood most likely property) no other possibilities. Both tracking addressed these examples purpose show feasibility simple approach dealing with data.© (2004) COPYRIGHT SPIE--The International Society Optical Engineering. Downloading abstract permitted personal only.