Fuzzy Logic-based Decision Fusion and Fuzzy Filtering
Jitendra R. Raol, Sudesh Kumar Kashyap, Lakshmi Shrinivasan · 2024
The objective of decision fusion is to take final a course of decision/action in the entire surveillance volume at any instant of time using outputs from different levels, e.g. Level 1-object refinement and Level 2-situation refinement, of an MSDF system. The accuracy of outputs from decision fusion depends not only on the architectures/algorithms involved in it but also on the different fusion levels. Hence, it becomes a necessity to explore the MSDF system first and then follow by a decision fusion philosophy. MSDF is the process of combining measured information originating from different sources, e.g. active or passive sensors, to produce the most specific and comprehensive unified data or model about an entity or event of interest. This technique achieves improved prediction-accuracy and more specific inferences than could be achieved by the use of a single sensor alone. Various applications of MSDF are i) target tracking, ii) automated target recognition, iii) guidance for autonomous vehicles, iv) remote sensing, v) battlefield surveillance and vi) automatic threat recognition systems, etc. The techniques employed in MSDF are drawn from diverse disciplines: digital signal processing, statistical estimation and control theory and classical numerical methods. In principle, fusion of data from multiple sources provides significant advantages over single source data. In addition to the statistical advantage gained by combining same-source data, the use of multiple types of sensors may increase the accuracy with which a quantity can be observed and characterized. The benefits of fusion are: i) Robust operational performance, ii) Extended spatial coverage, iii) Extended Temporal coverage, iv) Increased confidence, v) Reduced ambiguity, vi) Improved detection, and vii) Enhanced spatial resolution. In this chapter the application of fuzzy logic to decision fusion and filtering is discussed.