Fuzzy evidental filter for detection and tracking of dim objects (Q1892054)

The paper presents analytical and computational aspects of detecting and tracking dim objects (targets). The detection and tracking of objects (targets) which have small signature is difficult and this can pose a real challenge to the testing and evaluation of systems. The authors define an evidential filter for detection and tracking that obviates dependence on arbitrary probabilistic assumptions, thresholding, and other artificial constructs. The filter processes an image from a detector array and updates a function that measures the degree of belief that there is a target present in the individual detectors of the array. The approach is based on concepts from the theory of artificial neural networks and associative memories, dynamic programming, fuzzy decision making, nonlinear filtering, and other optimization and estimation methods. The authors also present several theoretical results, which are verified by computer simulation.