集群目标跟踪任务中密集机动个体跟踪存在编目难度大、算力资源要求高、航迹建模复杂的问题,在分析区分密集多目标和集群目标的运动特性及跟踪要点的基础上,提出了基于个体目标和群包络特征融合规划的集群目标跟踪策略,通过群组结构及数目建模,结合群结构的动态演化,对群组进行聚类、规划和分类,并在此基础上融合计算集群目标包络线的形状、中心、目标数量以及单个目标的灰度、位置等特征信息,实现集群目标跟踪位置的鲁棒提取,提高集群稳定跟踪能力和航迹准确生成能力。重点解决了群目标的聚类问题、群包络的位置提取和多个单目标位置融合问题、多个群组之间或群组超视场跟踪位置切换策略和平稳过渡问题,实验结果验证了所提方法能够有效提高跟踪性能,跟踪准确率提高了12%,跟踪稳定性提高了4倍,同时减少了计算量,工程应用前景好。

The dense maneuvering individual tracking in cluster target tracking tasks has the problems of difficult cataloging,high computational resource requirements,and complex trajectory modeling.Based on the analysis and differentiation of the motion characteristics and tracking points of dense multi-target and cluster targets,a cluster target tracking strategy based on the fusion planning of individual targets and group evolution features is proposed.By modeling the group structure and number,combined with the dynamic evolution of the group structure,the group is clustered,planned,and classified.On this basis,the shape,center,number of targets,as well as the grayscale,position and other feature information of individual targets are fused and calculated to achieve robust extraction of the tracking position of cluster targets,improve the stable tracking ability and accurate trajectory generation ability of the cluster.It mainly solves the clustering problem of group targets,the location extraction of group envelope,the location fusion problem of multiple single targets,the position switching strategy and the smooth transition problem of tracking between multiple groups or group beyond the field of view.The experimental results verify that the proposed method can effectively improve the tracking performance,the tracking accuracy is increased by 12%,the tracking stability is increased by 4 times,and the calculation amount is reduced.