摘要
针对被动声呐多目标跟踪问题,通过研究目标连续谱特征表征方式和特征更新相似关联机制,提出了一种基于连续谱特征的被动目标跟踪方法。该方法利综合利用频带能量法、排序截断平均算法以及峰值提取实现目标的连续特征表征,利用表征出的特征谱作为输入,通过建立相似度搜索,交叉判断与模板更新机制,实现了多目标的跟踪。经仿真对比分析了该方法的跟踪性能,并利用海上试验数据验证了其有效性。结果表明,连续谱特征可作为辅助特征用于目标的跟踪分辨,该方法能够有效提高多目标交叉情况下的跟踪关联能力,并且具备较低的运算量和较高的精度。
In view of the multi-target tracking problem of passive sonar,a passive target tracking method based on continuous spectrum features is proposed by studying the characterization of target continuous spectrum features and the similarity association mechanism of feature renewal.In this method,the continuous feature characterization of the target is realized by comprehensive use of the frequency band energy method,order truncate average method and peak extraction method.The characteristic spectrum is used as input,and the multi-target tracking is realized by establishing similarity search,cross judgment and template updating mechanisms.The tracking performance of this method is compared and analyzed by simulation,and its effectiveness is verified by sea trial data.The results show that the continuous spectral features can be used as auxiliary features for target tracking resolution,and the method can effectively improve the tracking correlation ability in the case of multi-target crossing with lower computation and higher precision.
作者
王卓然
王鲁军
谢亮
王佳囡
WANG Zhuoran;WANG Lujun;XIE Liang;WANG Jianan(Science and Technology on Sonar Laboratory,Hangzhou 310012,Zhejiang,China;Hangzhou Applied Acoustics Research Institute,Hangzhou 310012,Zhejiang,China;Military Industry Department,China State Shipbuilding Corporation Limited,Shanghai 200011,China)
出处
《声学技术》
CSCD
北大核心
2023年第6期709-717,共9页
Technical Acoustics
关键词
被动声呐
连续谱
目标跟踪
特征提取
passive sonar
continuous spectrum
target tracking
feature extraction
