Basic methods and progress of information fusion (ⅡⅡ)
PAN Quan,WANG Zeng-fu,LIANG Yan,YANG Feng,LIU Zhun-ga
2012-01-01
Abstract:Because of the demands from military technology,automation and intelligence,information fusion has been broadly concerned in academia and industry.Since a number of methods and algorithms have been emerging in recent years,it is necessary to analyze and review their recent developments.We start by analyzing the inherent problems and challenges in information fusion,including fusion architecture,uncertainty of information,multimodal information,conflict information,correlative information,networked information and nonlinearity of information;and then,we review various methods and algorithms developed in the past ten years for implementing the fusion of the above types of information.In addition,several future directions of research are highlighted and described,including the joint processing for information,human-centered information fusion,joint optimization for information gathering and fusion,architecture design for complex multisensor information fusion systems,simulation and performance evaluation for information fusion and the application of more mathematical theories.