基于小波变换的拉曼光谱仿真分析

Simulation Analysis of Based on Wavelet Transform for Raman Spectra

工业生产中实现质量目标控制,通过在线拉曼光谱分析,噪声的存在将影响确定拉曼光谱的谱峰位置和峰高数值等,并影响拉曼光谱定性和定量分析的准确性。为了消除混杂在拉曼光谱信号中的噪声,并保留信号的真实细节,提出了基于小波变换空域相关法的在线拉曼光谱信号消噪方法,对含噪的邻二甲苯原始拉曼光谱信号进行了消噪处理,进行仿真实验。结果表明,小波变换空域相关滤波法能较好的保留拉曼光谱信号的边缘特征,取得了较好的滤波效果,具有较高的实用价值。

Signals in online Raman spectra analysis often have noise,which greatly influences the Raman spectra peak position and the peak height,so as to influence the accuracy of the quantitative and qualitative analysis result.In order to eliminate the noise in the Raman spectra and hold details of the signa1,spatial correlation filtering based on wavelet transform for Raman spectra de-noising is put forward in this paper,and is used to remove the noise in the Raman spectra of o-xylene.The experimental results show that the fringe characters of Raman spectra are reserved effectively through spatial correlation filtering,obtain ideal result,and have a high practical value.