摘要
为了解决雷达探测数据中噪点过多的问题,提出了结合基于密度的噪声聚类算法(DBSCAN)和拉依达准则(3σ)的去噪方法。以雷达实际测量的目标运动信息为实验数据,运用DBSCAN算法进行聚类,剔除数据中的离群噪点,再通过拉依达准则去除影响较大的奇异值。实验结果表明,去噪之后雷达测距的线性误差由12 mm减少到0.36 mm,性能优于经典的半径滤波算法,可为实际雷达测量提供参考。
In order to solve the problem of removal of noise in radar detection data,a de-noising method combining density-based spatial clustering of applications with noise(DBSCAN)and pauta criterion(3σ)was proposed.The target motion information was measured by radar as experimental data.DBSCAN algorithm was used to cluster firstly to remove outliers in the data,and then pauta criterion was used to remove more influential singularities.Theoretical analysis and experimental verification were carried out,and good results were obtained.The results showed that the linear error of radar ranging was reduced from 12 mm to 0.36 mm after de-noising,and the performance was better than the classical radius filtering algorithm,which had a certain application value in practical radar measurement.
作者
张浩
张荣福
ZHANG Hao;ZHANG Rongfu(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)
出处
《光学仪器》
2021年第4期55-62,共8页
Optical Instruments
