基于Voigt函数拟合的拉曼光谱谱峰判别方法

发布时间：2018-04-27 13:56

本文选题：光谱学 + 拉曼光谱　；参考：《中国激光》2017年05期

【摘要】：提出了一种新的拉曼光谱谱峰判别算法,采用Voigt函数作为拟合函数,对小波变换脊线寻峰后得到的谱峰进行拟合处理,将拟合前后的谱峰相似度作为判别谱峰的依据,可有效识别与噪声幅值相近的小幅值拉曼信号。通过与已有的两种谱峰判别方法进行对比,在误判率为5%的情况下,该算法准确率提高了60%;在准确率高于90%的情况下,该算法误判率降低了10%。
[Abstract]:A new spectral peak discriminant algorithm for Raman spectrum is proposed. The Voigt function is used as a fitting function to fit the peaks of the peaks of the wavelet transform ridge. The similarity of the spectral peaks before and after the fitting can be used as the basis of the spectral peak, and the small amplitude Raman signal near the noise amplitude can be identified effectively. The two kinds of Raman signals have been identified. The spectral peak discrimination method is compared. With the error rate of 5%, the accuracy of the algorithm is increased by 60%, and the error rate of the algorithm is reduced by 10%. when the accuracy is higher than 90%.

【作者单位】：中国科学院上海光学精密机械研究所中国科学院空间激光信息传输与探测技术重点实验室;中国科学院大学;南京简智仪器设备有限公司;
【基金】：上海市自然科学基金(14ZR1445100)
【分类号】：O433.4

【相似文献】