摘要
为解决虚假点迹对雷达探测和跟踪性能的影响,提出一种基于图像特征提取的雷达点迹识别方法.首先基于雷达原始视频数据提取出16个图像特征参数,其次利用主成分分析(PCA)方法对特征参数的应用价值进行研究和降维处理,最后结合BP神经网络算法进一步对目标和杂波进行真伪鉴别.实验结果表明,相对于支持向量机(SVM)算法和BP神经网络算法,本文方法的点迹识别准确率更高,虚假率和漏警率更低.
In order to solve the influence of false radar plot on radar detection and tracking performance,this paper proposes a radar plot identification method based on image feature extraction.In this method,16 image feature parameters are first extracted from raw radar video data.Secondly,principal component analysis(PCA)is used to conduct studies and dimensionality reduction on the application value of the feature parameters.Finally,the BP neural network algorithm is employed to further identify the authenticity of the target and clutter.Experimental results show that compared with support vector machine(SVM)algorithm and BP neural network algorithm,the proposed method has higher accuracy,lower false alarm rate and lower missed alarm probability in terms of radar plot identification.
作者
赵红梦
王刚
丁智青
ZHAO Hongmeng;WANG Gang;DING Zhiqing(Nanjing Changjiang Electronic Information Industry Group Co.,Ltd.,Nanjing 210039,China)
出处
《空天预警研究学报》
CSCD
2023年第4期274-278,284,共6页
JOURNAL OF AIR & SPACE EARLY WARNING RESEARCH
关键词
雷达点迹
图像处理
特征提取
主成分分析
BP神经网络
点迹识别
radar plot
image processing
feature extraction
principal component analysis(PCA)
BP neural network
plot identification
