Tracking and Estimation of Multiple Cross-Over Targets in Clutter

Area of Science:

• Signal Processing
• Data Fusion
• Target Tracking

Background:

• Tracking multiple targets with unknown numbers and uncertain motion in cluttered environments presents significant challenges.
• Existing tracking algorithms often struggle with low detection probabilities and estimating cross-over targets.

Purpose of the Study:

• To develop a novel fixed-interval smoothing JITS (FIsJITS) algorithm for enhanced multi-target tracking.
• To improve estimation accuracy for multiple cross-over targets in heavy clutter environments.

Main Methods:

• The FIsJITS algorithm initializes tracks using joint integrated track splitting (JITS) in both forward (fJITS) and backward (bJITS) directions.
• It utilizes backward tracking to incorporate future scan measurements for current time-step estimation.
• Smoothing multi-target data association probabilities are computed for improved estimation.

Main Results:

• The FIsJITS algorithm demonstrates significantly improved estimation accuracy for multiple cross-over targets.
• Numerical assessments show superior performance compared to existing algorithms in simulations.
• The method effectively handles challenges like uncertain target motion and heavy clutter density.

Conclusions:

• The proposed FIsJITS algorithm offers a robust solution for complex multi-target tracking scenarios.
• Backward tracking integration enhances the performance of fixed-interval smoothers.
• This approach provides a substantial advancement in accurately tracking multiple cross-over targets.