摘要
针对多目标检测与跟踪任务,提出了基于实时递推最小二乘的多目标编批方法,采用轨迹预测的方式解决了目标遮挡、目标交叉等多目标相互干扰问题,并综合考虑了空间距离、长度近似度、宽度近似度以及区域框交并比等多方面因素来实现目标检测值与已存在目标队列间的数据关联。实验结果表明,该方法可以在目标机动的情况下,实时快速地实现目标编号建立、编号维持以及编号销毁,并能够同时对多个感兴趣目标的运动轨迹进行记录和维持,具有运算简单、易于工程实现等特点。
For multi-target detection and tracking tasks,this paper proposes a multi-target serial-number method based on least square principle in real-time recursively.In this method,trajectory prediction is used to solve the problem of multi-target mutual interference,such as target occlusion,target crossover and so on.The data association between the target detection value and the existing target queue is achieved by taking into account multiple factors such as space distance,length approximation,width approximation and area frame intersection over union.The experimental results show that method can realize the establishment,maintenance and destruction of the target number in real-time when the target is maneuvering.It can record and maintain the trajectory of multiple interested objects simultaneously.It has the characteristics of simple operation and engineering implementation easily.
作者
姜鑫
陈武雄
朱明
郝志成
高文
Jiang Xin;Chen Wuxiong;Zhu Ming;Hao Zhicheng;Gao Wen(Changchun Institute of Optics,Fine Mechanics and Physics,Chinese Academy of Sciences,Changchun 130033,China;Chongqing Jialing Huaguang Photoelectric Technology CO.LTD,Chongqing 400700,China)
出处
《国外电子测量技术》
2020年第2期59-64,共6页
Foreign Electronic Measurement Technology
关键词
编批
数据关联
最小二乘
显著性检测
serial-number
data association
least square
saliency detection