On a general sensor fusion method by state space modeling approach using particle filters

Abstract

Sensor fusion aims at obtaining information, which cannot be obtained by single sensor, by combining signals from multiple sensors. The main problem of sensor fusion is that computational cost increases exponentially with the number of sensors because the combination number of association between signals and sensors is large. We propose a general method to solve this problem in nonlinear state space model to deal with the unknown associations. We adapt the model to a specific situation of a sensor fusion. The target states and the associations are simultaneously estimated through the state estimation. In the estimation of the association, we apply particle filters with clever proposal. The associations are estimated in probabilistic way, not deterministic way, to avoid falling into a bad solution. As an example of the sensor fusion, we deal with tracking problem of sound target using camera and two microphones. Keyword: Sensor fusion, unknown association, particle filters, clever proposal, target tracking.