Multi-target linear converter - MLC

The Multi-target Linear Converter (MLC) is a general procedure for converting certain classes of single target tracking algorithms into equivalent multi-target tracking algorithms which have linear complexity both in the number of tracks and the number of measurements. In multi-target situations, multiple tracks may share the same measurement(s). Dedicated multi-target tracking algorithms usually form complex hypotheses based on all possible combinations of measurement to track origins, which results in the number of operations growing exponentially with the number of tracks and the number of measurements. Starting from any single target tracking algorithm which provides a-posteriori probability of measurement origin (IPDA, IPDA-DLL, IMM-PDA.. .) based on a single target assumption, MLC corrects the a-posteriori probabilities of measurement origin to allow tracks with common measurements to influence each other. From these corrected probabilities of measurement origin, MLC recalculates probabilities of individual track target existence and data association. Simulations compare IPDA converted by MLC (MLIPDA) with IPDA, JIPDA, and LJZPDA in a crossing target situation in a non-homogenous clutter environment. © 2003 IEEE.