Probabilistic Data Association (PDA) and Joint PDA (JPDA) algorithms a
re approaches for target tracking which have received considerable att
ention. It has been observed for some years that they both yield biase
d tracks in a multitarget environment. However, most work assumes no f
alse alarms and the rejection phenomenon of the JPDA algorithm has not
been reported. In this paper, the general procedure of multitarget tr
acking and the PDA/JPDA algorithms are first described. Their bias phe
nomenon is simulated and investigated. It is observed that (1) the JPD
A algorithm has less bias than the PDA algorithm in a clean environmen
t. Both of them yield coalescence, (2) the JPDA algorithm has coalesce
nce and rejection bias phenomenon while the PDA algorithm has only coa
lescence phenomenon in a clutter environment. Bias compensated algorit
hms are then presented using the polynomial regression method. Simulat
ions are carried out to select the order of polynomial regression. Mon
te Carlo simulations also demonstrate the effectiveness of the compens
ated PDA/JPDA algorithms. (C) 1998 Elsevier Science Ltd. All rights re
served.