脉冲激光雷达信号降噪方法对比  

Comparison of backscattering lidar signal denoising methods

作　　者：陈冬[1] 王江安[1] 康圣[1] 

机构地区：[1]海军工程大学电子工程学院光电研究所,湖北武汉430033

出　　处：《舰船科学技术》2011年第4期93-97,共5页Ship Science and Technology

基　　金：国家高技术发展计划项目及海军预先研究项目(401060702)

摘　　要：研究了脉冲激光雷达大气后向散射回波信号特性,介绍了一种以H ilbert-Huang变换(HHT)为基础的滤波新方法——固有模态(EMD)分解滤波,并与扩展卡尔曼滤波(EKF)、小波阈值滤波算法进行对比。针对典型大气条件下回波信号特点进行仿真,对各种方法在不同信噪比信号条件下滤波效果进行评价,分析它们的优缺点。仿真结果证明:EKF法仅在均匀大气高信噪比条件下滤波效果较好;小波阈值方法具有更强的适应性,是中低信噪比条件下的最优算法;EMD分解滤波具有很高研究价值,该方法在高信噪比条件下处理效果最好,是现有方法的必要补充。综合运用3种方法才能使滤波效果达到最佳。The property of lidar backscattering signal in atmosphere was discussed in the paper.A new filter method——empirical mode decomposed method（EMD）which is based on Hilbert-Huang transplantation（HHT）was applied.The extend Kalman filtering method（EKF）,wavelet threshold method were compared with.The simulation based on typical atmosphere conditions,showed effect of each method,including its advantages and disadvantages.It is proved by simulation that EKF method works well only when the backscattering signal SNR is high,the wavelet threshold method is the best method in most case,and the EMD method has best result when high SNR.The EMD method that is proved to be valid is worthy of further researching on.As a result,integrating three methods is the best way for lidar signal filtering.

关 键 词：激光雷达 后向散射 小波阈值滤波 经验模态分解 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置]