A Multi-Target State Extraction Method for the Probability Hypothesis Density Filter Based on Spectral Clustering

ZHANG Hui,HAN Chongzhao,YAN Xiaoxi
2012-01-01
Abstract:A spectral clustering target state extraction method is proposed to obtain multi-target state estimate in the sequential Monte Carlo(SMC) implementation of probability hypothesis density(PHD) filter.An affinity matrix is established based on the similarities among a large number of particles that are output of SMC implementation of PHD filter,and the Laplace matrix of the affinity matrix is obtained through Laplace transform.Then spectral clustering is performed through eigen-decomposition of the Laplace matrix,and the clustering center of each cluster is searched and is regarded as the estimation of multi-target states.Moreover,the Nystrm approximation method is used to generate eigenvectors so that the computation complexity of the algorithm is reduced.Simulation results show that the estimation accuracy of the proposed method is 60% higher than that of the k-means target state extraction method.
What problem does this paper attempt to address?