基于粒子滤波的无人机实时在线目标跟踪算法研究  

Research on Real-Time Online Target Tracking Algorithm for Unmanned Aerial Vehicle Based on Particle Filter

作　　者：季陈倍 田梦楚[2] 吴云[3] 吴岑 马韬 JI Chenbei;TIAN Mengchu;WU Yun;WU Cen;MA Tao(School of Automation,Nanjing University of Science and Technology,Nanjing 210094,China;School of Intelligent Manufacturing,Nanjing University of Science and Technology,Wuxi 210014,China;China Satellite Maritime Tracking and Control Department,Wuxi 214431,China)

机构地区：[1]南京理工大学自动化学院,南京210094 [2]南京理工大学智能制造学院,江苏无锡210014 [3]中国卫星海上测控部,江苏无锡214431

出　　处：《智能物联技术》2025年第1期13-20,共8页Technology of Io T& AI

基　　金：国家自然科学基金青年基金项目(62103190)。

摘　　要：无人机的迅速发展催生出许多非法飞行现象,及时发现和实时跟踪无人机尤为重要。无人机在远距离或低信噪比环境下难以被传统雷达系统有效探测。红外成像技术由于其高精度和强抗干扰能力,加上拥有出色非线性非高斯处理能力的粒子滤波算法,成为理想的无人机检测跟踪手段。针对复杂背景和低信噪比环境下无人机弱小目标的实时在线检测跟踪问题,提出一种改进粒子滤波的无人机实时在线检测前跟踪算法。在图像预处理阶段,通过结合非局部均值滤波改进了基于形态学滤波的背景抑制算法。针对粒子重采样,融合3种传统重采样算法的思想改进得出一种混合重采样算法。经过一系列的实验测试与分析,验证了改进算法在应对低信噪比及动态多变背景下的无人机目标追踪难题时显著提升了跟踪精度和稳定性。The rapid development of unmanned aerial vehicle has given rise to many illegal flight phenomena,and timely detection and real-time tracking of unmanned aerial vehicle is particularly important.It is difficult for unmanned aerial vehicle to be effectively detected by traditional radar system in long distance or low signal to noise ratio environment.Infrared imaging technology is an ideal method for unmanned aerial vehicle detection and tracking because of its high precision,strong anti-interference ability and excellent nonlinear non-Gaussian particle filter algorithm.Aiming at the problem of unmanned aerial vehicle real-time online detection and tracking of dim targets in complex background and low signal to noise ratio environment,an improved particle filter algorithm is proposed.In the stage of image preprocessing,the background suppression algorithm based on morphological filtering is improved by combining non-local mean filtering.Aiming at particle resampling,a hybrid resampling algorithm is proposed by combining the ideas of three traditional resampling algorithms.After a series of experimental tests and analysis,it is verified that the improved algorithm can significantly improve the tracking accuracy and stability of unmanned aerial vehicle target tracking in the context of low signal to noise ratio and dynamic changes.

关 键 词：粒子滤波 检测前跟踪 扩展卡尔曼滤波 目标跟踪 非局部均值滤波 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]