中文摘要：

目标跟踪中的一个核心部分就是滤波算法。经典的卡尔曼滤波算法对于具有高斯分布噪声的线性系统,可得系统状态的最小均方差估计,α-β-γ滤波器比卡尔曼滤波器计算量小且易于实现。基于贝叶斯滤波的目标跟踪原理,在线性高斯环境下对比分析了标准卡尔曼滤波器和α-β-γ滤波器的估计精度,并给出两种方法的适用条件。介绍了标准卡尔曼滤波器和α-β-γ滤波器的原理及算法实现,分别对匀速直线运动和匀加速直线运动的目标进行跟踪仿真分析。仿真表明,在高斯背景环境下,卡尔曼滤波器、α-β滤波器和α-β-γ滤波器都具有很好的跟踪精度和很强的实时性。

英文摘要：

One of the most important parts in target tracking is the filtering algorithm. The typical Kalman filter can get the recursive minimum mean-square estimation under the linear and white Gaussian noise circumstance. Compared to the Kaman filter, the α-β-γ filter reduces the computation complexity and can be easily implemented in engineering applications. Based on the principle of Bayesian filtering method, this paper analyzes the estimation accuracies of Kalman filter and a-,8-y filter, and gives the applicative conditions of the two different methods. The fundamental theories and filtering algorithm implementations of Kalman filter and α-β-γ filter are studied. Simulations of tracking the targets with uniform velocity and acceleration in the straight line are presented. The simulation results show that the Kalman filter, α-βfilter and α-β-γ filter are all with better tracking capabilities and good real-time performances in linear and Gaussian environment.