Dim Target Track-Before-Detect based on Particle Filtering

This paper addresses dim target detection of multiband infrared data. First, the dynamic model of the target in the multi-spectral infrared image sequence and the measurement model of the multi-spectral infrared image sequence are established. Then, the particle filtering algorithm for single target tracking before detection is studied. Based on this, combined with sequential measurement fusion strategy, the algorithm is proposed for detection and tracking of dim moving targets from multi-spectral infrared image sequences. Finally, the validity of the proposed algorithm is verified by simulation experiments combined with the posterior probability and root mean square error of the target.