靶场测量中跟踪目标的多尺度多级模糊聚类分类算法

Multi-scale and Multi-level Fuzzy Clustering Algorithm for Tracking Targets in Range

摘要:

靶场测量中的连发弹目标数量较多,需要对此目标进行分类跟踪,文章提出了一种多尺度多级模糊聚类的分类算法,首先对靶场测量中的多连发目标的特性进行分析,定义了3种不同的类目标群序列,其次针对模糊聚类算法必须先验确定聚类初始点以及聚类中心的缺陷,在此基础上加入尺度因子,并根据目标群的位置信息以及运动方向信息进行多级聚类,最后通过多次聚类实验证明分类结果能够满足靶场测量中的多连发目标的分类要求。

Abstract:

The number of the tracking targets is so big as to do classifying for better tracking in the range. the article proposes multi-scale-level classify algorithm based fuzzy clustering algorithm. First, we give different conception for three various target groups by analyzing the motion characteristics of the target group.Then, the fuzzy clustering algorithm must has the clustering initial location and the clustering centre,In accordance with the problem,we add the scale factor,then making multi-level clustering according to the location and moving direction of the targets. Finally, Many experiments indicate that, the algorithm proposed in the article can content the classifying demand for the multi-continuous targets in the range. 