Classification is an important part of the Level 1 Fusion problem. It provides necessary information to the user to make decisions about the object in question. In the military problem, this decision can be the difference between prosecution of a target and declaration as a noncombatant. A number of approaches use different sources of information address classification. Many of these techniques are probabilistic. Some are linguistic interpretations by expert analysis. These classifiers can provide different views of the same object. The key is to combine these disparate pieces of information, each with different levels of quality and/or confidence. We propose a technique to combine these various classifiers using the concept of evidence accrual. Information can affirm a class' hypothesis, refute it, or have no bearing upon it. Also, each source of evidence has a degree of quality or uncertainty about it. Finally, the sources may be represented as numeric or nonnumeric information. To address these issues, we utilized a set of decoupled fuzzy Kalman filters. This bank of filters, similar in nature to first-order observers, estimates the degree of evidence that each known class achieves. The paper outlines the development of our approach, its implementation to emulate a Bayesian classifier, and a set of examples.
Feature object extraction - a fuzzy logic approach for evidence accrual in the Level 1 Fusion classification problem
2003
5 Seiten, 9 Quellen
Conference paper
English
A Feature Level Fusion Approach for Object Classification
British Library Conference Proceedings | 2007
|Adaptive UAS Route Planner Based Upon Evidence Accrual
IEEE | 2019
|Adaptive Evidence Accrual for Context-Sensitive Situation Understanding
British Library Conference Proceedings | 2010
|