作者: Gongjian Zhou , Ding Ma , Taifan Quan
DOI: 10.1109/ICOSP.2014.7015334
关键词: Probability density function 、 Doppler effect 、 Estimator 、 Density estimation 、 Constant false alarm rate 、 Computer science 、 Clutter 、 Radar tracker 、 Artificial intelligence 、 Moving target indication 、 Computer vision
摘要: This paper investigates the problem of incorporating Doppler measurements to improve multi-target tracking performance. The clutter density in both spatial domain and direction is assumed be unknown non-homogeneous. Instead modeling as product probability function (pdf), this introduces a hyper-spatial concept, where considered an additional pseudo position dimension. sparsity estimator extended estimation combined with Joint Integrated Probabilistic Data Association (JIPDA) tracker deal non-parametric measurements. Monte Carlo simulation results demonstrate effectiveness proposed approach.