基于椭圆方向和半轴长度的多伯努利扩展目标跟踪

Multi-Bernoulli extended target tracking based on orientation and half axes lengths of an ellipse

针对杂波环境下扩展目标外形难以估计、新生目标先验信息未知等问题,在扩展目标势均衡多目标多伯努利(ET-CBMeMBer)滤波器的基础上,开展基于椭圆方向和半轴长度(OAL)的参数化多扩展目标跟踪研究。借助乘性噪声构造考虑目标空间信息的显式量测方程,并推导OAL-CBMeMBer滤波器高斯混合实现的封闭形式解;基于显式的量测方程,利用已知的量测数据自适应构造考虑了质心位置和外形状态的新生目标先验信息,并提出自适应新生OAL-CBMeMBer滤波器。仿真实验结果表明:所提OAL-CBMeMBer滤波器提高了目标数目和状态的估计精度,能够有效地对多扩展目标进行跟踪。

The problem that the shape of the extended target is hard to estimate and the newborn target's prior information is unknown in the clutter environment is solved in this paper by using the extended targets-cardinality balance multi-target and multi-Bernoulli(ET-CBMeMBer)filter.This parameterized multi-extended target tracking algorithm is based on the orientation and half axes lengths of an ellipse(OAL).Considering the spatial information of the target,an explicit measurement equation is constructed by multiplicative noise,and the closed-form solution implemented by the Gaussian mixture of the OAL-CBMeMBer filter is deduced.Based on the explicit measurement equation,the prior information of the newborn target that considers the position of the centroid and the shape state is adaptively constructed by employing the known measurement data,and the adaptive OAL-CBMeMBer filter is proposed.The proposed OAL-CBMeMBer filter can effectively track numerous extended targets and increases the estimation accuracy of target number and state,according to simulation findings.