摘要
针对基于减法聚类的Hough变换航迹起始法存在的虚假航迹起始率高、目标航迹点缺失严重的问题,提出了一种基于运动二步约束聚类的Hough变换航迹起始法;该算法首先采用直观法滤除部分杂波点;随后利用Hough变换进行低阈值筛选;然后利用减法聚类得到多个聚类中心;最后由最近邻法判断出每个样本点所归属于聚类中心,得到目标航迹的数目以及参数;仿真表明,在密集杂波环境下,文章提出的算法的航迹起始概率近似是多级Hough变化航迹起始算法的1.2倍,同时虚假航迹起始概率是多级Hough变化航迹起始算法的三分之一,结果表明该算法改善了航迹起始的性能,特别适合于密集杂波下低信噪比目标的目标检测问题。
Aiming to solve the problem of high false alarm and seriously missing target tracking from the Track Initiation Algorithm based on Hough Transform and Subtractive Clustering, a new Hough Transform track initiation method based on two--stage clustering with motion constraints is proposed. In this algorithm, first, the Object Method is used to filtrate parts of clutter, then the Hough Transform is adopted and filtrate the data by a low threshold values, following, the subtractive clustering is used to obtain the number and parameters of the target tracking, at last, the Nearest Neighbor method is used to decide which cluster center each sample point belong to. The simulation shows that under the density clutter environment, the initial probability of the proposed algorithm is 1.2 time than that of multi-- Hough track initiation and the false track initiation probability of the present method is one third of the multi-- Hough track initiation algorithm, which show that the present algorithm can improve the performance of track Initiation effectively, especially detect the targets in density clutter and low SNR environment.
出处
《计算机测量与控制》
CSCD
北大核心
2011年第11期2759-2762,共4页
Computer Measurement &Control
基金
国家自然基金重点项目(60434030)
航空基金(20090853013)
关键词
HOUGH变换
减法聚类
航迹起始
最近邻法
直观法
运动约束
Hough transform
subtractive clustering
track initiation
nearest neighbor
object method
motion constraints
