Performance of Fast Algorithms of Multi-target Tracking

Hamad, A. M.; Yousef, S.; Matar, M. A.; Nofal, H.

doi:10.21608/asat.1991.25833

Performance of Fast Algorithms of Multi-target Tracking

Document Type : Original Article

Authors

¹ Staff members in Avionics Dept..MTC.

² Staff members in Avionics Dept.,MTC.

³ Ph.D. applicant in Avionics Dept..MTC.

10.21608/asat.1991.25833

Abstract

The tracking properties of fast tracking algorithms were considered and analyzed (1-57 assuming a single target in clean and in cluttered environment with various values for probability of detectionCP D. In the present work.we extend our study to tracking environments where typical multi-target tracking cases arise,e.g.,two closely spaced targets Ctypical for a target in formation)and the case of two crossing targets.The obtained the results reveal that with a 100% PD and zero clutter density.perfect data association at smaller targets spacings is provided with fast Kalman filter of order N=2 compared to N=3 for the gradient lattice filter.At higher clutter densities, even with a 100% PD, the fast Kalman filter has shown unstable behavior at target spacings below some threshold values, In case of tracking crossing tracks a nondiverging fast Kalman tracking filter with Nm2 has almost the same capability as a gradient lattice filter having N=4 at various values of P and at various clutter densities. Conditions that guarantee a stable performance of the fast Kalman filter are given.