Raman spectrum denoising method based on deep learning

Abstract

The invention provides a Raman spectrum denoising method based on deep learning, which solves the problem that pure Raman spectrum information is easy to lose in the denoising process in the prior art, and input parameters are set by means of human intervention. The invention provides a deep learning network-based Raman spectrum denoising method, which comprises the following steps of: generating Raman spectrum data required by model training, and dividing the Raman spectrum data into a training set, a verification set and a test set; then, a U-Net Raman spectrum denoising model is established, the model is divided into an encoder network module and a decoder network module, feature extraction is performed on Raman spectrum data by using one-dimensional convolution in the encoder network, and an extracted feature spectrum is reconstructed and a pure spectrum is output in the decoder network. According to the method, noise of the Raman spectrum can be effectively removed, Raman spectrum information is well reserved, and accurate and reliable information is provided for further qualitative and quantitative analysis of the Raman spectrum. The method can be widely applied to the optical field.

Description

technical field [0001] The present invention relates to the field of optics, in particular to the field of optical detection technology, and specifically refers to a deep learning-based Raman spectrum denoising method. Background technique [0002] Raman spectroscopy is a technology proposed by Indian scientist Raman in the 1920s. The birth of lasers in the 1960s brought room for development in the application of Raman spectroscopy. Raman spectroscopy has been widely used as an effective detection and analysis method due to its rich material property information, non-destructive testing, and no need for sample preparation. [0003] Raman spectroscopy is an inelastic scattering spectrum produced by the molecular vibration of substances, which can perform qualitative and quantitative analysis of substances, so it has been widely used in many fields such as medical treatment and chemical industry. Since Raman spectroscopy is an extremely sensitive spectrum, which is easily af...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention includes two: (1) solve the problem that the current method cannot accurately and efficiently denoise the Raman spectrum; (2) solve the problem that the current method cannot automatically set the denoising parameters

Images