Raman Spectrum Analysis Algorithm Research And System Realization

Raman spectroscopy technology has the characteristics of no damage to samples and convenient use,and has been widely used in many fields such as life sciences,materials science,medicine and environment.Portable Raman spectrometers have now entered a more mature stage of commercial application.Related foreign manufacturers have launched a variety of models of portable Raman spectrometers.However,the research and development of domestic related equipment and supporting facilities is still relatively behind foreign countries,so the development of my country The spectrometer equipment and Raman spectrum analysis software with independent intellectual property rights are of great significance to catch up with foreign technology and break through the stuck neck technology.This article has carried out related research on the data processing algorithm in Raman spectroscopy software.Since the actual software operators often only use software to do the qualitative identification of spectra,they generally do not have professional knowledge of chemistry and optical signal processing.In order to facilitate the use of the software,the article mainly studies that no manual intervention is required.Automated processing algorithm of Raman spectroscopy.This article mainly studies the flow of various Raman spectroscopy data processing algorithms,and on this basis,proposes a more efficient algorithm,and completes the design and implementation of the Raman spectroscopy analysis software.The main research work is as follows:(1)The related technology of Raman spectroscopy data preprocessing is studied.A variety of common spectrum preprocessing algorithms are implemented,and the focus is on the de-peaking,noise-reduction,and fluorescence-reducing background algorithms of Raman spectroscopy.For the research of de-spike methods,an improved Hampel filter algorithm is proposed,which can automatically de-spike without manually setting parameters.For the research on noise reduction methods,an improved cyclic three-point zero-order Savitzky-Golay filter method is proposed,and the performance is compared with the traditional mean filter method,median filter method and Savitzky-Golay filter method.A better effect.For the research on the method of de-fluorescence background,an improvement to the polynomial fitting algorithm is proposed.First,wavelet transform is used to suppress the low-frequency signal in the spectrum,so that the fluorescence background of the spectrum is more stable,and the fluorescence background can be removed automatically.(2)The related algorithm of Raman spectrum peak detection is studied.The paper mainly studies the continuous wavelet transform and the dual-scale correlation algorithm,and proposes an improvement to the dual-scale correlation algorithm.The experiment shows that the improved dual-scale correlation algorithm has better peak detection capabilities and does not require manual setting of parameters,so the algorithm Used in the software’s spectral peak detection.(3)The Raman spectroscopy discriminant analysis technology is studied.It mainly studies the discriminant analysis technology related to the three methods of distance measurement,similar shape function and spectrum peak matching,and proposes a spectrum discrimination algorithm that mixes the correlation coefficient and spectrum peak matching methods to improve the accuracy of spectrum identification Spend.(4)Based on the above-mentioned algorithm,developed a friendly interface Raman spectroscopy software,analyzed the requirements of Raman spectroscopy software,introduced the overall architecture of the software and the implementation process of each module,and independently designed the software interface layout.The functional test verifies that the software can automatically complete the discriminant analysis of Raman spectroscopy.