动态移动最小二乘多项式平滑的拉曼全自动基线校正算法

发布时间：2017-12-28 14:11

本文关键词：动态移动最小二乘多项式平滑的拉曼全自动基线校正算法　出处：《应用激光》2017年04期 　论文类型：期刊论文

【摘要】：基线漂移是目前光谱仪收集拉曼光谱时难以避免的现象。在应用拉曼光谱进行定性、定量分析过程中,基线漂移会给光谱分析带来不利影响,为增强分析结果的准确性和稳定性,需要对拉曼光谱进行基线校正预处理。为了校正基线漂移,提出一种动态移动最小二乘多项式平滑(DMSG)的方法拟合背景基线,自动识别原始光谱的谱峰区域和非谱峰区域的子区间,从而在每个拉曼光谱子区间中自动控制平滑迭代次数与窗口宽度,逐渐将拉曼谱峰数值逐渐拉低到与基线同等量级,最终得到整个光谱的基线估计。将该算法应用于模拟和实际拉曼光谱进行基线校正,结果表明该算法无论对模拟的拉曼光谱还是真实拉曼光谱,都有较好的基线校正效果,为进一步分析光谱数据提供更准确的信息,是一种可行的基线校正方法。
[Abstract]:Baseline drift is an unavoidable phenomenon in the collection of Raman spectra by spectrometers at present. In the qualitative and quantitative analysis of Raman spectroscopy, baseline drift will bring negative effects on spectral analysis. In order to enhance the accuracy and stability of the analysis results, we need to baseline correction for Raman spectra. In order to correct the baseline drift, this paper proposed a dynamic moving least squares polynomial smoothing (DMSG) method of fitting baseline, peak area automatic recognition of original spectral and non spectral peak area of the sub interval, thereby smoothing iterations and window width control automatically in each interval in the Raman optical spectrum, Raman spectrum peak value gradually will gradually pull with the same magnitude to low baseline, finally get the whole spectrum of baseline estimation. The algorithm is applied to simulated and real Raman spectra were baseline correction, the results show that the algorithm both on simulated and real Raman spectroscopy Raman spectroscopy, have good correction effect, to provide more accurate information for further analysis of spectral data, is a feasible baseline correction method.
【作者单位】：厦门理工学院电气工程与自动化学院;中国科学院海西研究院厦门稀土材料研究所;
【基金】：厦门市科技局国际科技合作资助项目(项目编号:3502Z2016125)
【分类号】：O433
【正文快照】： 0引言拉曼光谱是一种通过测量分子振动引起的散射光谱来表征物质内部分子结构的分析技术,具有快速简单以及无损伤和电离辐射的优点,因此在物理、化学、生物、考古等多个领域成为一种热门的物质分析与研究手段[1-4]。然而,在光谱采集过程中由于环境温度、外界振动以及仪器自身

【相似文献】