Title :
Fixed lag smoothing for target tracking applications based on IMM
Author :
Sundkvist, Mari ; Bergman, N.
Author_Institution :
SaabTech Syst., Jarfalla, Sweden
Abstract :
Interactive multiple model (IMM) filtering is the most common approach to handle maneuvering targets in real time filtering. In this work we present an approach to extend IMM filtering to fixed lag smoothing. The presented smoother operates on the estimates produced by a regular real time IMM filter. This configuration makes it easy to implement the smoother as an extra feature in a commercial tracker based on the IMM concept. The structure of the fixed lag smoother takes full advantage of the computations performed in the IMM filters. Each fixed lag estimate is determined by a "backward" IMM operation based on the estimates produced in each mode of the real time IMM filter.
Keywords :
digital filters; smoothing methods; state estimation; target tracking; IMM filtering; backward IMM operation; fixed lag smoothing; interactive multiple model filtering; maneuvering targets; real time filtering; target tracking applications;
Conference_Titel :
Target Tracking: Algorithms and Applications (Ref. No. 2001/174), IEE
DOI :
10.1049/ic:20010237