成像目标质心跟踪算法适应性研究  

Study on Adaptability of Centroid Tracking Algorithm of Imaging Targets

作　　者：穆晓敬 杨振声 刘红梅 黄子露 Mu Xiaojing;Yang Zhensheng;Liu Hongmei;Huang Zilu(AVIC-HONGDU,Nanchang,Jiangxi,330024)

机构地区：[1]航空工业洪都,江西南昌330024

出　　处：《教练机》2022年第2期43-47,共5页Trainer

摘　　要：本文对质心跟踪算法进行适应性研究,主要解决不同质心跟踪算法的应用场景问题,通过原理分析和仿真验证,对基于种子点生长和阈值分割的质心跟踪算法适应性进行研究。基于种子点生长的质心跟踪算法适用于简单背景下的无纹理目标,且对视场中目标大小或当前目标距离有一定要求;基于阈值分割的质心跟踪算法对简单背景下的低对比度目标能够稳定跟踪,不仅能够实现规则均匀目标区域的跟踪,而且对远距离小目标也有一定适应性,对复杂场景中的低对比度目标在加严阈值条件的情况下有一定适应性。This paper studies the adaptability of the centroid tracking algorithm with the main purpose to resolve the issue of application scenarios for different centroid tracking algorithms.The adaptability of the centroid tracking algorithms based on seeds growing and threshold segmentation is investigated by principle analysis and simulation verification.The centroid tracking algorithm based on seeds growing is suitable for targets without texture under simple backgrounds.Some requirements are specified for the target size in the field of view and the present target range.The centroid tracking algorithm based on threshold segmentation is able to achieve stable tracking of the low contrast targets under simple backgrounds,which can not only track targets in regular and uniform areas,but also adapt to the long-range small targets and the low contrast targets in complex scenarios under stricter threshold conditions.

关 键 词：质心跟踪 种子点生长 阈值分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]