中红外光谱信号去噪方法研究  

Research on denoising method of mid-infrared spectral signal

作　　者：赵媛 杨萱蔓 王星 董建娥 Zhao Yuan;Yang Xuanman;Wang Xing;Dong Jian'e(Faculty of Mathematics and Physics Southwest Forestry University,Kunming,Yunnan 650000,China;School of Big Data and Intelligent Engineering Southwest Forestry University)

机构地区：[1]西南林业大学数理学院,云南昆明650000 [2]西南林业大学大数据与智能工程学院

出　　处：《计算机时代》2023年第9期142-145,150,共5页Computer Era

基　　金：国家级大学生创新创业训练计划平台项目(202010677011);云南省农业联合专项项目:二维相关光谱结合残差神经网络的云南牛肝菌可追溯质量评价体系研究(202101BD070001-038)。

摘　　要：红外光谱已成为现代许多领域常用的工具。在中红外光谱信号的分析中,由于其包含噪声及各种外界干扰因素,导致信号分析的准确度降低。为减小误差,依次将滑动平均法、小波分析法、傅里叶变换法、SG平滑滤波器法、处理离群值法与HHT法依次运用到人工构造信号及绒柄牛肝菌中红外光谱的实际信号中去噪,通过比较其信噪比及均方根误差,选出较为优质的方法。实验结果表明,HHT法的信噪比为最大、均方根误差为最小,说明去噪效果最好。Infrared spectrum has become a common tool in many modern fields.However,in the analysis of mid-infrared spectroscopy signal,the accuracy of signal analysis is reduced due to its inclusion of noise and various external interference factors.In order to reduce the error,the moving average method,wavelet analysis method,Fourier transform method,SG smoothing filter method,outlier processing method and HHT method are applied to the artificial signal and the actual signal of the mid-infrared spectrum of boletus velutipes,and the better method is selected by comparing its signal-to-noise ratio and root-meansquare error.The experimental results show that the HHT method has the largest signal-to-noise ratio and the smallest root-meansquare error,which shows that the denoising effect is the best.

关 键 词：红外光谱 去噪 HHT法 信噪比 均方根误差 

分 类 号：TN919.4[电子电信—通信与信息系统]