Statistical information fusion criteria for multi-sensory systems
Hong-Dar Lin, Chaowen Chang · 2002
Most current information fusion techniques focus on adding all information sources, and then observe the operation of a fused system. These methods generally do not concern whether a new sensor source would enhance system performances beforehand. The statistical meta-analysis offers a set of quantitative techniques that permit synthesizing a variety of independent information sources. Using the statistical meta-analysis, this paper presents a method that can provide fusion criteria to foretell the effect of adding a new information source in terms of statistical type I errors before the source is actually combined. This method is then implemented as an illustration.