激光雷达信号的可变间隔阈值经验模式分解去噪法  

Denoising method using empirical mode decomposition with switchable interval threshold for lidar signals

作　　者：李猛[1] 蒋立辉[1] 熊兴隆[1] 冯帅[1] 

机构地区：[1]中国民航大学天津市智能信号与图像处理重点实验室,天津300300

出　　处：《强激光与粒子束》2014年第11期8-12,共5页High Power Laser and Particle Beams

基　　金：国家自然科学基金项目(41075013);国家重点基础研究发展计划项目;中央高校基金资助项目(ZXH2010D020;3122013P009)

摘　　要：针对采用经验模式分解直接阈值(EMD-DT)和经验模式分解间隔阈值(EMD-IT)在激光雷达回波信号的去噪应用中会产生的模态混叠现象,采用一种可变间隔阈值的经验模式分解(EMD-SIT)的去噪方法。首先,对信号进行经验模式分解。然后,采用过零率方法将分解出的含有噪声的固有模态函数分离。最后,应用过零点阈值,设立一个新的可变阈值,将EMD-IT和EMD-DT有效融合对信号进行去噪。通过与多种阈值的仿真对比以及激光雷达的回波信号去噪实验,结果表明该方法可以有效地去除噪声,抑制模态混叠,较EMD-IT和EMD-DT更具有优越性,因此有着很好的应用前景。Using the empirical mode decomposition direct threshold （EMD-DT） and empirical mode decomposition interval threshold （EMD-IT） to de-noise lidar return signals,aiming to solve the problem of mode-mixing,a method of EMD switch interval threshold （EMD-SIT） is applied.Firstly,the signal is processed by the use of EMD.Then the decomposed intrinsic mode functions （IMF） with noise are removed by zero-crossing rate method.Finally,by the use of the zero-crossing threshold,a switch interval threshold is obtained,which combines the EMD-IT and EMD-DT to de-noise the noise signals.The comparisons of different thresholds and the de-noise experiment of lidar shows that noise can be removed effectively by the proposed method,meantime the mode mixing is restrained.Compared with EMD-DT and EMD-IT methods,EMD-SIT method has more advantages Therefore,the demonstrated method has a promising future.

关 键 词：经验模式分解 模态混叠 可变间隔阈值经验模式分解 激光雷达 去噪 

分 类 号：TN958.98[电子电信—信号与信息处理]