Target recognition and tracking is an important research filed in the surveillance industry.Traditional target recognition and tracking is to track moving objects, however, for the detected moving objects the specific...Target recognition and tracking is an important research filed in the surveillance industry.Traditional target recognition and tracking is to track moving objects, however, for the detected moving objects the specific content can not be determined.In this paper, a multi-target vehicle recognition and tracking algorithm based on YOLO v5 network architecture is proposed.The specific content of moving objects are identified by the network architecture, furthermore, the simulated annealing chaotic mechanism is embedded in particle swarm optimization-Gauss particle filter algorithm.The proposed simulated annealing chaotic particle swarm optimization-Gauss particle filter algorithm(SA-CPSO-GPF) is used to track moving objects.The experiment shows that the algorithm has a good tracking effect for the vehicle in the monitoring range.The root mean square error(RMSE), running time and accuracy of the proposed method are superior to traditional methods.The proposed algorithm has very good application value.展开更多
This paper presents the derivation of Gauss-Newton filter in linear cases and an analysis of its properties. Based on the minimum variance theorem, the Gauss-Newton filter is constructed and derived, including its sta...This paper presents the derivation of Gauss-Newton filter in linear cases and an analysis of its properties. Based on the minimum variance theorem, the Gauss-Newton filter is constructed and derived, including its state transition equation, observation equation and filtering process. Then, the delicate relationship between the Gauss-Aitken filter and the Kalman filter is discussed and it is verified that without process noise the two filters are equivalent. Finally, some simulations are conducted. The result shows that the Gauss-Aitken filter is superior to the Kalman filter in some aspects.展开更多
基于状态空间模型的许多传统滤波算法都基于Rn空间中的高斯分布模型,但当状态向量中包含角变量或方向变量时,难以达到理想的效果。针对J.T.Horwood等提出的nS?R流形上的Gauss Von Mises(GVM)多变量概率密度分布,扩展了狄拉克混合逼近方...基于状态空间模型的许多传统滤波算法都基于Rn空间中的高斯分布模型,但当状态向量中包含角变量或方向变量时,难以达到理想的效果。针对J.T.Horwood等提出的nS?R流形上的Gauss Von Mises(GVM)多变量概率密度分布,扩展了狄拉克混合逼近方法,给出了联合分布的GVM逼近方法,推导了后验分布的GVM参数计算公式,设计了量测更新状态估计算法。将J.T.Horwood等的时间更新算法与所提出的量测更新算法相结合,可实现基于GVM分布的递推贝叶斯滤波器(GVMF)。仿真结果表明,当状态向量符合GVM概率分布模型时,GVMF对角变量的估计明显优于传统的扩展卡尔曼滤波器。展开更多
Recently there have been researches about new efficient nonlinear filtering techniques in which the nonlinear filters generalize elegantly to nonlinear systems without the burdensome lineafization steps. Thus, truncat...Recently there have been researches about new efficient nonlinear filtering techniques in which the nonlinear filters generalize elegantly to nonlinear systems without the burdensome lineafization steps. Thus, truncation errors due to linearization can be compensated. These filters include the unscented Kalman filter （UKF）, the central difference filter （CDF） and the divided difference filter （DDF）, and they are also called Sigma Point Filters （SPFs） in a unified way. For higher order approximation of the nonlinear function. Ito and Xiong introduced an algorithm called the Gauss Hermite Filter, which is revisited in [5]. The Gauss Hermite Filter gives better approximation at the expense of higher computation burden, although it＇s less than the particle filter. The Gauss Hermite Filter is used as introduced in [5] with additional pruning step by adding threshold for the weights to reduce the quadrature points.展开更多
通过将模型的状态噪声和观测噪声均表示成高斯和的形式,推导出非线性非高斯状态空间模型的高斯和递推算法,进一步提出了对应的扩展卡尔曼和滤波器(extended Kalman sum filter,EKSF)和高斯厄密特和滤波器(Gauss-Hermite sum filter,GHSF...通过将模型的状态噪声和观测噪声均表示成高斯和的形式,推导出非线性非高斯状态空间模型的高斯和递推算法,进一步提出了对应的扩展卡尔曼和滤波器(extended Kalman sum filter,EKSF)和高斯厄密特和滤波器(Gauss-Hermite sum filter,GHSF)。EKSF和GHSF分别用扩展卡尔曼滤波器(extended Kalman filter,EKF)和高斯厄密特滤波器(Gauss-Hermite filter,GHF)作为高斯子滤波器。分析的结果表明,现有的高斯和滤波算法是本文算法的特例;仿真结果表明,EKSF和GHSF能有效处理非线性非高斯模型的状态滤波问题,与高斯和粒子滤波器(Gaussian sum particle filter,GSPF)相比,EKSF和GHSF在保证精度的同时,大大降低了计算量,仿真时间分别约为GSPF的5%和6%。展开更多
基金Supported by the National Key R&D Plan of China (2021YFE0105000)the National Natural Science Foundation of China (52074213)+1 种基金Shaanxi Key R&D Plan Project (2021SF-472)Yulin Science and Technology Plan Project (CXY-2020-036)。
文摘Target recognition and tracking is an important research filed in the surveillance industry.Traditional target recognition and tracking is to track moving objects, however, for the detected moving objects the specific content can not be determined.In this paper, a multi-target vehicle recognition and tracking algorithm based on YOLO v5 network architecture is proposed.The specific content of moving objects are identified by the network architecture, furthermore, the simulated annealing chaotic mechanism is embedded in particle swarm optimization-Gauss particle filter algorithm.The proposed simulated annealing chaotic particle swarm optimization-Gauss particle filter algorithm(SA-CPSO-GPF) is used to track moving objects.The experiment shows that the algorithm has a good tracking effect for the vehicle in the monitoring range.The root mean square error(RMSE), running time and accuracy of the proposed method are superior to traditional methods.The proposed algorithm has very good application value.
文摘This paper presents the derivation of Gauss-Newton filter in linear cases and an analysis of its properties. Based on the minimum variance theorem, the Gauss-Newton filter is constructed and derived, including its state transition equation, observation equation and filtering process. Then, the delicate relationship between the Gauss-Aitken filter and the Kalman filter is discussed and it is verified that without process noise the two filters are equivalent. Finally, some simulations are conducted. The result shows that the Gauss-Aitken filter is superior to the Kalman filter in some aspects.
文摘基于状态空间模型的许多传统滤波算法都基于Rn空间中的高斯分布模型,但当状态向量中包含角变量或方向变量时,难以达到理想的效果。针对J.T.Horwood等提出的nS?R流形上的Gauss Von Mises(GVM)多变量概率密度分布,扩展了狄拉克混合逼近方法,给出了联合分布的GVM逼近方法,推导了后验分布的GVM参数计算公式,设计了量测更新状态估计算法。将J.T.Horwood等的时间更新算法与所提出的量测更新算法相结合,可实现基于GVM分布的递推贝叶斯滤波器(GVMF)。仿真结果表明,当状态向量符合GVM概率分布模型时,GVMF对角变量的估计明显优于传统的扩展卡尔曼滤波器。
文摘Recently there have been researches about new efficient nonlinear filtering techniques in which the nonlinear filters generalize elegantly to nonlinear systems without the burdensome lineafization steps. Thus, truncation errors due to linearization can be compensated. These filters include the unscented Kalman filter （UKF）, the central difference filter （CDF） and the divided difference filter （DDF）, and they are also called Sigma Point Filters （SPFs） in a unified way. For higher order approximation of the nonlinear function. Ito and Xiong introduced an algorithm called the Gauss Hermite Filter, which is revisited in [5]. The Gauss Hermite Filter gives better approximation at the expense of higher computation burden, although it＇s less than the particle filter. The Gauss Hermite Filter is used as introduced in [5] with additional pruning step by adding threshold for the weights to reduce the quadrature points.
文摘通过将模型的状态噪声和观测噪声均表示成高斯和的形式,推导出非线性非高斯状态空间模型的高斯和递推算法,进一步提出了对应的扩展卡尔曼和滤波器(extended Kalman sum filter,EKSF)和高斯厄密特和滤波器(Gauss-Hermite sum filter,GHSF)。EKSF和GHSF分别用扩展卡尔曼滤波器(extended Kalman filter,EKF)和高斯厄密特滤波器(Gauss-Hermite filter,GHF)作为高斯子滤波器。分析的结果表明,现有的高斯和滤波算法是本文算法的特例;仿真结果表明,EKSF和GHSF能有效处理非线性非高斯模型的状态滤波问题,与高斯和粒子滤波器(Gaussian sum particle filter,GSPF)相比,EKSF和GHSF在保证精度的同时,大大降低了计算量,仿真时间分别约为GSPF的5%和6%。
