An entropy method for multisource data fusion

Abstract

-The present paper proposes a generic model of the multisource data fusion in the framework of the theory of information, with closer attention being given the di erent nature of data processed in common cases. This model that we have called entropy model is then used to elaborate processing methods able to face s p eci c problems that may arise when multisource systems are implemented to achieve functions like classi cation and pattern recognition, matching of ambiguous observations, estimation, detection or tracking. Crucial practical problems to data fusion are more specifically dealt with, such as information representation, appropriate combination processing and decision making. Some clues are given on the practical use and implementation of such an approach, for example, in the distributed estimation problem.

A new probabilistic and entropy fusion approach for management of information sources

“…The decision-making phase is based on the entropy rule of decision whose expression is given by Eq. (22). It permits to search the most probable hypothesis type through the frame of discernment, which is a nice solution to the information fusion problem.…”

“…The goal of the fusion method using entropy approach, that we call Entropy Fusion Model (EFM) [22,23], is to process the set of input vectors X ¼ fX s g S s¼1 in such a way the resulting subset of input vectors e X Ã contains as much information as possible about the treated problem and minimizes the uncertainty of the fused information. The fusion system is to be optimized by maximizing the mutual information IðX ; Y Þ between all its input vectors X s of X and its output vector Y .…”

Sensor selection for target tracking based on single dimension information gain

“…In 1991, information theory was first applied in a problem related to sensor management and state estimation [5, 6]. Subsequently, expected information gain was applied in sensor management and data‐fusion problems [1, 7]. Sensor selection method by computing the expected Rényi divergence between the prior and posterior probability density was presented [1, 8], and the KL divergence has been applied in various sensor management scenes [3, 4, 9, 10].…”