A trace-before-detect algorithm based on expectation-maximization

This paper proposes a Trace-Before-Detect (TBD) algorithm based on Expectation- Maximization (EM) that achieves the conditional expectation estimation of unknown information by calculating the association of multi-frame pre-detection plots within a temporal window to distinct motion parameters. The parameters are then updated through likelihood function maximization until convergence. Simulation results demonstrate that under the signal-to-noise ratio(SNR) of 9 dB, the algorithm successfully detects two crossing moving targets, achieving stable multi-target detection in low-SNR environments.