31 October 2016 Mean likelihood estimation of target micro-motion parameters in laser detection
Author Affiliations +
Abstract
Maximum Likelihood Estimation(MLE) is the optimal estimator for Micro-Doppler feature extracting. However, the enormous computational burden of the grid search and the existence of many local maxima of the respective highly nonlinear cost function are harmful for accurate estimation. A new method combining the Mean Likelihood Estimation(MELE) and the Monte Carlo(MC) way is proposed to solve this problem. A closed-form expression to evaluate the parameters which maximize the cost function is derived. Then the compressed likelihood function is designed to obtain the global maximum. Finally,the parameters are estimated by calculating the circular mean of the samples get from MC method. The high dependence of accurate initials and the computational complexity of the iteration algorithms are avoided in this method. Applied to the simulated and experimental data, the proposed method achieves similar performance as MLE but less computational amount. Meanwhile, this method guarantees the global convergence and joint parameter estimation.
© (2016) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Liren Guo, Yihua Hu, Yunpeng Wang, "Mean likelihood estimation of target micro-motion parameters in laser detection", Proc. SPIE 10021, Optical Design and Testing VII, 100211Z (31 October 2016); doi: 10.1117/12.2247697; https://doi.org/10.1117/12.2247697
PROCEEDINGS
10 PAGES


SHARE
Back to Top