paper

Three Mathematical Formalisms of Multiple Hypothesis Tracking

Volume Number:
14
Issue Number:
2
Pages:
176
–
198
Publication Date:
1 December 2019
Return To Issue

paper Menu

Abstract

This paper describes three mathematical formalisms, each of which provides a solid foundation for developing multiple hypothesis tracking (MHT) theories and algorithms, as solutions to detection-based multiple target tracking (MTT) problems. The three formalisms are 1) random finite sequence (RFSeq), 2) finite point process (FPP), and 3) random finite set (RFSet) formalisms. We will discuss equivalencies and some subtle differences among them. In addition, we will discuss theoretical consequences of various assumptions on MHT hypothesis evaluation, as well as recent RFSet-based MTT algorithm developments claiming relationship to MHT.