基于改进小波阈值的量子点荧光光谱降噪研究
发布时间：2018-11-13 17:11
【摘要】：使用光谱仪采集到的信号难免受到不同噪声源的影响。为了提高光谱信号解析的精准度,通过分析小波应用于信号降噪的原理以及经典的软、硬阈值降噪法存在的缺陷,提出了一种改进的阈值降噪法。该方法既克服了硬阈值法产生间断点,软阈值法产生恒定偏差的缺陷,又尽量地保留了有用信号。实验选用的小波基函数为SymletsA,分解层数为4,结合Birge-Massart策略模型确定的分层阈值对硒化镉量子点荧光光谱信号进行降噪处理。结果表明,与经典的软、硬阈值降噪法相比,通过改进阈值降噪法得到重构信号的信噪比(SNR=47.550 2)、能量占比(PER=0.973 3)和均方误差(MSE=149.421 3)均有提高和改善。
[Abstract]:The signals collected by spectrometer are inevitably affected by different noise sources. In order to improve the accuracy of spectral signal resolution, an improved threshold denoising method is proposed by analyzing the principle of wavelet application in signal denoising and the defects of classical soft and hard threshold de-noising method. This method not only overcomes the defect that hard threshold method produces discontinuity point, soft threshold method produces constant deviation, but also preserves useful signal as far as possible. The wavelet basis function of the experiment is SymletsA, decomposition layer number is 4, combined with the delamination threshold determined by Birge-Massart strategy model, the fluorescence spectrum signal of cadmium selenide quantum dot is de-noised. The results show that the SNR (SNR=47.550 _ 2), PER=0.973 _ 3 and mean square error (MSE=149.421 _ 3) of reconstructed signal are improved by the improved threshold denoising method compared with the classical soft-hard threshold denoising method.
【作者单位】： 贵州大学大数据与信息工程学院;
【基金】：国家国际科技合作专项项目(2014DFA00670)
【分类号】：TH744.1
本文编号：2329798
[Abstract]:The signals collected by spectrometer are inevitably affected by different noise sources. In order to improve the accuracy of spectral signal resolution, an improved threshold denoising method is proposed by analyzing the principle of wavelet application in signal denoising and the defects of classical soft and hard threshold de-noising method. This method not only overcomes the defect that hard threshold method produces discontinuity point, soft threshold method produces constant deviation, but also preserves useful signal as far as possible. The wavelet basis function of the experiment is SymletsA, decomposition layer number is 4, combined with the delamination threshold determined by Birge-Massart strategy model, the fluorescence spectrum signal of cadmium selenide quantum dot is de-noised. The results show that the SNR (SNR=47.550 _ 2), PER=0.973 _ 3 and mean square error (MSE=149.421 _ 3) of reconstructed signal are improved by the improved threshold denoising method compared with the classical soft-hard threshold denoising method.
【作者单位】： 贵州大学大数据与信息工程学院;
【基金】：国家国际科技合作专项项目(2014DFA00670)
【分类号】：TH744.1
【相似文献】
相关期刊论文 前6条
1 程晓锋,李泽仁,江孝国;一种高能X光转换屏的荧光光谱实验研究[J];光电工程;1996年S1期
2 凌明胜;钱志余;梁超英;;血糖浓度荧光光谱检测研究[J];量子电子学报;2007年05期
3 黄珍;;ARL9800XP荧光光谱议故障分析及维修[J];农村经济与科技;2009年10期
4 吕飞;石春;刘晓光;秦琳琳;吴刚;赵春江;;基于FPGA和DSP的时间分辨荧光光谱测硼仪[J];电子测量技术;2012年06期
5 陈小强;赖万昌;张震;陈亨贵;于新华;;小波分析在X荧光光谱去噪声中的应用[J];核电子学与探测技术;2009年04期
6 ;[J];;年期
相关硕士学位论文 前1条
1 李艳春;基于CCD的农药光纤荧光光谱测量系统的研究[D];燕山大学;2006年
，本文编号：2329798
本文链接：https://www.wllwen.com/kejilunwen/yiqiyibiao/2329798.html
