Spectrometer baseline correction method and system based on improved empirical mode decomposition

Abstract

The invention relates to the technical field of spectrum baseline correction, and provides a spectrograph baseline correction method and system based on improved empirical mode decomposition, and the method comprises the following steps: employing a sliding window smoothing method, and carrying out the preprocessing of obtained measurement spectrum data; an improved empirical mode decomposition algorithm is adopted, fitting is carried out on the preprocessed measurement spectrum data to obtain baseline data, and the improved empirical mode decomposition algorithm adopts least square fitting to construct upper and lower boundary envelopes of the spectrum data; subtracting the baseline data obtained by fitting from the obtained measurement spectrum data to obtain the measurement spectrum data after the baseline is removed; and sequentially carrying out abnormal point detection and removal, spectral power calibration and wavelength calibration on the measured spectral data after baseline removal to obtain spectral data after baseline correction. The spectral baseline is identified by improving empirical mode decomposition, the stability and boundary adaptability of mode decomposition are enhanced, and a more accurate baseline fitting result is obtained.

Description

Technical Field

[0001] The present disclosure relates to the technical field related to spectral baseline correction, and in particular to a spectrometer baseline correction method and system based on improved empirical mode decomposition. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] In spectral measurement systems, the acquired spectral data not only contains target signal information but is often superimposed with slowly varying background signals caused by instrument system characteristics, environmental changes, electrical noise, and other factors. This slowly varying component is commonly referred to as the spectral baseline. The spectral baseline does not reflect the true characteristics of the measured object and can significantly interfere with key parameters such as the amplitude, shape, and position of the spectral signal. Failure to effectively calibrate the baseline can lead to problems such as spectral integral area error, peak position shift, and spectral shape distortion, thereby affecting the accuracy of subsequent quantitative analysis, identification, and judgment. Spectral baseline calibration, which uses algorithms to identify and remove baseline components from spectral data, ensures that subsequent spectral analysis is based on the true signal. This is a crucial step in ensuring the accuracy and reliability of spectral measurements and is crucial in applications such as communications, sensing, molecular spectroscopy, and high-precision laser measurement.

[0004] During their research, the inventors discovered that existing spectral baseline calibration methods primarily include polynomial fitting, wavelet transform, and empirical mode decomposition (EMD). While the polynomial fitting method is simple and easy to use, it has limited ability to fit complex nonlinear baselines and is susceptible to overfitting due to interference from strong peaks. While the wavelet transform can separate different frequency components, it is sensitive to parameter selection and exhibits poor robustness. While the standard EMD method offers good adaptability, it is often affected by "modal aliasing" in actual spectral processing, resulting in low pole identification accuracy, unstable fitting, and envelope distortion, particularly at boundaries. Summary of the Invention

[0005] In order to solve the above problems, the present disclosure proposes a spectrometer baseline correction method and system based on improved empirical mode decomposition. By improving the empirical mode decomposition to identify the spectral baseline, the stability and boundary adaptability of the modal decomposition are enhanced, and a more accurate baseline fitting result is obtained.

[0006] In order to achieve the above objectives, the present disclosure adopts the following technical solutions: One or more embodiments provide a spectrometer baseline correction method based on improved empirical mode decomposition, comprising the following steps: The acquired measured spectral data are preprocessed using a sliding window smoothing method; Using an improved empirical mode decomposition algorithm to fit the preprocessed measured spectral data to obtain baseline data, the improved empirical mode decomposition algorithm uses least squares fitting to construct upper and lower boundary envelopes of the spectral data; Subtracting the fitted baseline data from the acquired measured spectrum data to obtain measured spectrum data after baseline removal; The measured spectral data after baseline removal are subjected to outlier detection and removal, spectral power calibration and wavelength calibration in sequence to obtain baseline-corrected spectral data.

[0007] One or more embodiments provide a spectrometer baseline correction system based on improved empirical mode decomposition, including: a sliding window smoothing module configured to preprocess the acquired measured spectral data using a sliding window smoothing method; An improved empirical mode decomposition module is configured to use an improved empirical mode decomposition algorithm to fit the preprocessed measured spectral data to obtain baseline data, wherein the improved empirical mode decomposition algorithm uses least squares fitting to construct upper and lower boundary envelopes of the spectral data; A baseline removal module is configured to subtract the fitted baseline data from the acquired measured spectrum data to obtain measured spectrum data after baseline removal; The correction module is configured to perform outlier detection and removal, spectral power calibration, and wavelength calibration on the measured spectral data after baseline removal to obtain spectral data after baseline correction.

[0008] An electronic device includes a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of the above-mentioned spectrometer baseline correction method based on improved empirical mode decomposition are completed.

[0009] A computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the steps in the above-mentioned spectrometer baseline correction method based on improved empirical mode decomposition are completed.

[0010] Compared with the prior art, the present invention has the following beneficial effects: The improved empirical mode decomposition algorithm disclosed in the present invention constructs upper and lower boundary envelopes through the least squares fitting method, avoiding the problem of unstable interpolation of extreme points in traditional EMD, significantly alleviating the modal aliasing phenomenon, and improving the fitting accuracy and baseline extraction accuracy in the boundary area. This improved algorithm can achieve more accurate intrinsic mode function decomposition, enhance the sensitivity of baseline identification to low-frequency drift, and make the entire spectral data processing process more stable and reliable. The advantages of the present invention and the advantages of additional aspects will be described in detail in the specific embodiments below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings, which constitute a part of the present disclosure, are used to provide a further understanding of the present disclosure. The exemplary embodiments of the present disclosure and their descriptions are used to explain the present disclosure but do not constitute a limitation of the present disclosure.

[0012] Figure 1 is a flowchart of the baseline correction method of Example 1 of the present disclosure; Figure 2 is a block diagram of a baseline correction system according to Embodiment 2 of the present disclosure; DETAILED DESCRIPTION The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0013] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present disclosure belongs.

[0014] It should be noted that the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof. It should be noted that, in the absence of conflict, the various embodiments in the present disclosure and the features in the embodiments can be combined with each other. The embodiments will be described in detail below with reference to the accompanying drawings.

[0015] Example 1 In the technical solutions disclosed in one or more embodiments, Figure 1 As shown in FIG, the spectrometer baseline correction method based on improved empirical mode decomposition includes the following steps: Step 1: Preprocess the acquired measured spectral data using a sliding window smoothing method; Step 2: using an improved empirical mode decomposition algorithm to fit the preprocessed measured spectral data to obtain baseline data, wherein the improved empirical mode decomposition algorithm uses least squares fitting to construct upper and lower boundary envelopes; Step 3: Subtract the fitted baseline data from the acquired measured spectrum data to obtain the measured spectrum data after baseline removal; Step 4: Perform outlier detection and removal, spectral power calibration, and wavelength calibration on the measured spectral data after baseline removal to obtain baseline-corrected spectral data; This method uses a sliding window smoothing method to perform preliminary processing on the measured spectral data to reduce high-frequency noise interference and improve the stability of subsequent processing. After preprocessing, an improved empirical mode decomposition (EMD) algorithm based on least squares fitting is used, replacing the traditional EMD method based on extreme point interpolation. This algorithm constructs the upper and lower boundary envelopes of the spectral data through global minimum error optimization, thereby enhancing the stability and boundary adaptability of the modal decomposition and obtaining more accurate baseline fitting results. This baseline data is then subtracted from the original measured spectrum to obtain a pure spectrum with the baseline removed. To further improve the quality of the spectral data, this method also includes steps such as outlier detection and removal, power intensity normalization, and wavelength coordinate correction. The final output is spectral data that has undergone high-fidelity baseline correction.

[0016] The method of the present embodiment significantly improves the accuracy and stability of spectral baseline correction, and has stronger robustness and adaptability than traditional polynomial fitting, wavelet transform or standard EMD methods. Specifically, the improved empirical mode decomposition algorithm adopted constructs the upper and lower boundary envelopes by the least squares fitting method, avoids the problem of instability of extreme point interpolation in traditional EMD, significantly alleviates the modal aliasing phenomenon, and improves the fitting accuracy and baseline extraction accuracy in the boundary area. The improved algorithm can achieve more accurate intrinsic mode function decomposition, enhances the sensitivity of baseline identification to low-frequency drift, and makes the entire spectral data processing process more stable and reliable. Since the entire baseline correction process is completed at the software processing level, there is no need to change any hardware structure, so it has good versatility and low cost advantages. In addition, the method can be widely used in precision fiber spectrometers with different principles such as stimulated Brillouin spectrometers, heterodyne coherence spectrometers, Fourier spectrometers, etc., and can significantly improve the dynamic range, resolution and sensitivity of the instrument, thereby enhancing its performance in complex measurement environments.

[0017] In step 1, the method for preprocessing the measured spectral data using a sliding window smoothing method includes the following steps: Step 1.1, calculating the total number of sampling points based on the wavelength measurement range of the precision fiber optic spectrometer for baseline correction and the resolution of the spectrum, and setting the window width of the sliding window based on the total number of sampling points; For example, the wavelength measurement range of a precision fiber optic spectrometer performing baseline correction is 1525 nm to 1615 nm, and the spectral resolution is 0.08 pm. According to the Nyquist sampling theorem, the spectral sampling rate is half the spectral resolution, that is, the spectral sampling rate is 0.04 pm. The formula for calculating the total number of sampling points is: ; In order to balance the fitting accuracy, fitting efficiency and fitting spectrum distortion, the window width was selected as 100 times the spectral resolution based on the empirical value, and the window width was set to 200 spectral sampling points; Step 1.2, calculate the average value of the sliding window: Process the acquired spectral data using the set sliding window. For each data point i, take the data points half the size of the sliding window before and after the data point i, and calculate the average value of the data points in the sliding window as the value of the data point i; Specifically, when the sliding window is set to 200, the size of the data before and after the data point i is taken, and the window area is expressed as [i-99,i+100]; Furthermore, the window size is automatically reduced when there are less than 200 points at the beginning or end of the spectrum; Step 1.3: Repeat the above steps for all points until all points are processed, and obtain the spectral data after sliding window smoothing; The sliding window averaging process described in this embodiment provides the foundation for subsequent least squares boundary fitting and EMD decomposition, eliminating unstructured noise from spectral fluctuations. Removing high-frequency noise and distortion improves the accuracy of subsequent envelope fitting and the physical interpretability of IMFs, avoids initial extreme point errors, and improves the smoothness and accuracy of baseline fitting.

[0018] In step 2, the improved empirical mode decomposition algorithm is used to fit the preprocessed measured spectral data to obtain baseline data, including the following steps: Step 21, identifying extreme points: identifying all extreme points of the pre-processed measured spectrum data Sp(n), including the maximum and minimum values ​​of the data points; Step 22: Using least squares fitting, obtain the upper boundary u0(n) and the lower boundary d0(n) as the upper and lower boundary envelopes of the construction; Specifically, between two adjacent maximum (or minimum) extreme points, the upper boundary u0(n) and the lower boundary d0(n) are obtained using least squares fitting. The process is as follows: Step 221: Select data between two adjacent extreme points and use least squares fitting to obtain fitting data; The data between adjacent extreme points may be the extreme points corresponding to the maximum value or the extreme points corresponding to the minimum value; Step 222, fitting matching judgment: If the deviation between the fitting data and the original data exceeds the set deviation threshold, the point is identified as an abnormal point; the set deviation threshold can be set to 20%. Step 223: using least squares fitting to obtain corrected fitting data for the data between two adjacent extreme points after removing the abnormal points; Step 224: Take the minimum point of the corrected fitting data as the lower boundary d0(n), and the maximum point as the upper boundary u0(n).

[0019] Traditional EMD typically uses cubic spline interpolation to construct upper and lower envelopes for extreme points. This interpolation method may cause envelope jitter or "overfitting" at boundaries or in areas with sparse extreme point distribution. This embodiment performs least squares fitting between two adjacent extreme points and uses least squares straight line or polynomial fitting instead of interpolation between each pair of adjacent maxima (or minima). The fitting function does not force every extreme point to pass through, but rather "approximates" the entire interval, optimizing the overall fitting error. This prevents fluctuation amplification at the edges and avoids boundary instability. It also prevents significant deviations of the entire envelope curve due to abnormalities at individual extreme points. It reduces sensitivity to local distortion or errors, significantly reduces boundary oscillations, improves fitting stability, and avoids physically inconsistent fitting behavior.

[0020] Step 23: Calculate the average boundary based on the constructed upper and lower boundary envelopes: (1) Where, represents the average boundary.

[0021] Step 24: Subtract the average boundary from the pre-processed measured spectrum data Sp(n) to obtain the preliminary oscillation component ; (2) Step 25: Determine whether the obtained preliminary oscillation component meets the requirements of the eigenmode function. If so, use the preliminary oscillation component as the eigenmode function component. Otherwise, use the preliminary oscillation component instead of the preprocessed measured spectrum data to execute steps 21 to 24 until the eigenmode function component is obtained. Determine the initial oscillation component The judgment conditions for meeting the requirements of the eigenmode function include the following two criteria: 1) The total number of poles and zero crossings does not differ by more than one point; 2) At any point, the sum of the upper and lower bounds of the local extreme envelope is zero; Meeting the above two criteria h1(n) means meeting the requirements of the intrinsic mode function IMF.

[0022] Step 26: Minimize the residual generated by subtracting the intrinsic mode function component IMF1 obtained in the previous step from Sp(n), and perform multiple iterations to extract all IMF components until the remaining residual meets the set stopping condition. The remaining residual is the spectral baseline; Specifically, the residual calculation formula is: (3) Where k is the number of empirical mode decompositions. The residual is calculated by the empirical mode function component IMF1, and the second empirical mode decomposition is performed. This is repeated until the stopping condition of the iterative decomposition is met: the remaining residual satisfies the condition of becoming a constant, a slope, or a function with only one extreme point. Use the residual r1(n) instead of Sp(n) and repeat steps 21 to 26 to obtain other IMFs, such as IMF2, IMF3, ..., IMFN until r N (n) becomes a constant, a slope, or a function with only one extreme point, and rN(n) is the baseline data obtained by fitting.

[0023] The final baseline can be expressed as: (4) The traditional EMD suffers from subjective judgment when determining whether an IMF is "valid"; its stopping criteria are unclear, which can easily lead to overfitting or incomplete decomposition. The eigenmode function judgment criteria set in this embodiment ensure that each extracted IMF has good physical interpretability and numerical stability. The stopping criteria for the multiple iterative extraction of all IMF components are more clearly defined, ensuring that the final extracted baseline has typical low-frequency "trend term" characteristics, truly meeting the physical definition of a "baseline."

[0024] Complete the baseline fitting of the measured spectral data. In step 3, the baseline can be removed by directly subtracting the fitted baseline from the measured spectral data. In step 4, outlier detection and removal are performed on the measured spectral data after baseline removal. For the measured spectral data after baseline removal, outlier detection is performed point by point starting from the first data point. If a point is less than or equal to zero, it is replaced by a set small constant value to ensure that all sampling points of the measured spectral data after baseline removal are greater than zero. The small constant value can be set to 1e-20; In step 4, a spectral power calibration method is sequentially performed on the measured spectral data after baseline removal, including the following steps: Step 41: Select a standard light source to provide standard spectral power; Preferably, a narrow-linewidth single-frequency laser having a linewidth much smaller than that of the precision fiber spectrometer outputting the measurement spectrum of this embodiment is selected as the light source; Step 42: Integrate the spectrum peak area of ​​the measured spectrum data after baseline removal according to the set spectrum area size to obtain the measured power; Considering that the spectral resolution in this embodiment is 0.08 pm, the spectral peak integration region can be selected as a spectral region with 20 times the spectral resolution. Therefore, the spectral peak integration region is selected as the spectral region centered on the spectral peak wavelength and around ±0.8 pm, and the integrated power is obtained as the measured power; Step 43: Subtract the integrated measured power from the standard spectrum power to obtain a power calibration coefficient; Step 44: calibrate the spectral power of the measured spectral data after baseline removal based on the power calibration coefficient; In step 4, a wavelength calibration method is sequentially performed on the measured spectral data after baseline removal, including the following steps: Step 4.1: Select a standard light source to provide a standard spectrum wavelength. Preferably, a narrow-linewidth single-frequency laser having a linewidth much smaller than that of the precision fiber spectrometer outputting the measurement spectrum of this embodiment is selected as the light source; Step 4.2: Fit the spectrum peak of the measured spectrum data after baseline removal using a Gaussian line shape to obtain the peak wavelength; Step 4.3: Subtract the peak wavelength obtained by fitting from the wavelength of the standard spectrum to obtain the wavelength calibration coefficient; Step 4.4: Calibrate the spectral wavelength of the measured spectral data after baseline removal based on the wavelength calibration coefficient; thereby completing the baseline correction of the spectral data measured by the precision fiber optic spectrometer.

[0025] Through the above steps, the baseline correction of the precision fiber optic spectrometer is achieved, and precise spectral data with improved indicators such as dynamic range, spectral accuracy, and sensitivity are obtained.

[0026] The baseline correction method of the present invention only involves the processing of spectral measurement data. It does not require any modification to the hardware structure of the existing spectrometer and can be directly integrated into the software analysis process, with extremely high engineering adaptability and deployability. At the same time, the method supports automated processing procedures, which is conducive to system integration and batch application, and embodies the significant advantages of "simple, efficient and low-cost". It systematically solves the problems of limited dynamic range, reduced accuracy and insufficient sensitivity caused by baseline interference in traditional spectral measurements. Through high-precision baseline fitting and removal, the real signal distribution is effectively restored, and the performance upper limit of the precision fiber optic spectrometer is significantly improved. It is particularly suitable for scenarios such as high-speed optical communication, modulation analysis, and sensor detection that have extremely high requirements for signal-to-noise ratio and power accuracy. The method is compatible with a variety of precision fiber optic spectrometer platforms, including spectrometer systems based on the stimulated Brillouin scattering principle, heterodyne coherent detection principle, and Fourier transform principle. The method has strong versatility and a wide range of adaptability. It can be expanded to a variety of precision spectral measurement systems as a general data processing module to achieve accurate identification and stable removal of baseline components.

[0027] Example 2 Based on Example 1, this embodiment provides a spectrometer baseline correction system based on improved empirical mode decomposition, such as Figure 2 Shown, including: a sliding window smoothing module configured to preprocess the acquired measured spectral data using a sliding window smoothing method; An improved empirical mode decomposition module is configured to use an improved empirical mode decomposition algorithm to fit the preprocessed measured spectral data to obtain baseline data, wherein the improved empirical mode decomposition algorithm uses least squares fitting to construct upper and lower boundary envelopes of the spectral data; A baseline removal module is configured to subtract the fitted baseline data from the acquired measured spectrum data to obtain measured spectrum data after baseline removal; The correction module is configured to perform outlier detection and removal, spectral power calibration, and wavelength calibration on the measured spectral data after baseline removal to obtain spectral data after baseline correction.

[0028] It should be noted here that the various modules in this embodiment correspond one-to-one to the various steps in Example 1, and the specific implementation processes are the same, which will not be repeated here.

[0029] Example 3 This embodiment provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of the spectrometer baseline correction method based on improved empirical mode decomposition in Example 1 are completed.

[0030] Example 4 This embodiment provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps of the spectrometer baseline correction method based on improved empirical mode decomposition in Example 1 are completed.

[0031] The foregoing description is merely a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. Those skilled in the art will readily appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present disclosure shall be included within the scope of protection of the present disclosure.

[0032] Although the above describes the specific implementation methods of the present disclosure in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present disclosure. Those skilled in the art should understand that on the basis of the technical solution of the present disclosure, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present disclosure.

Claims

1. A spectrometer baseline correction method based on improved empirical mode decomposition, characterized in that: The steps include: The acquired measured spectral data are preprocessed using a sliding window smoothing method; Using an improved empirical mode decomposition algorithm to fit the preprocessed measured spectral data to obtain baseline data, the improved empirical mode decomposition algorithm uses least squares fitting to construct upper and lower boundary envelopes of the spectral data; Subtracting the fitted baseline data from the acquired measured spectrum data to obtain measured spectrum data after baseline removal, to obtain measured spectrum data after baseline removal; The measured spectral data after baseline removal are subjected to outlier detection and removal, spectral power calibration and wavelength calibration in sequence to obtain baseline-corrected spectral data.

2. The spectrometer baseline correction method based on improved empirical mode decomposition according to claim 1, wherein: The method for preprocessing the measured spectral data using a sliding window smoothing method comprises the following steps: The total number of sampling points is calculated based on the wavelength measurement range of the precision fiber optic spectrometer for baseline correction and the resolution of the spectrum, and the window width of the sliding window is set based on the total number of sampling points; The acquired spectral data is processed using the set sliding window. For each data point i, the data points half the size of the sliding window before and after the data point i are taken, and the average value of the data points in the sliding window is calculated as the value of the data point i. After all data points are processed, the spectral data after sliding window smoothing is obtained.

3. The spectrometer baseline correction method based on improved empirical mode decomposition according to claim 1, wherein: The improved empirical mode decomposition algorithm is used to fit the preprocessed measured spectral data to obtain baseline data, including the following steps: Step 21, identifying all extreme points of the pre-processed measured spectrum data Sp(n), including the maximum and minimum values ​​of the data points; Step 22: Using least squares fitting, obtain the upper boundary u0(n) and the lower boundary d0(n) as the upper and lower boundary envelopes of the construction; Step 23, calculating the average boundary based on the constructed upper and lower boundary envelopes; Step 24: Subtract the average boundary from the pre-processed measured spectrum data Sp(n) to obtain the preliminary oscillation component ; Step 25: Determine whether the obtained preliminary oscillation component meets the requirements of the eigenmode function. If so, use the preliminary oscillation component as the eigenmode function component. Otherwise, use the preliminary oscillation component instead of the preprocessed measured spectrum data to execute steps 21 to 24 until the eigenmode function component is obtained. Step 26: Minimize the residual generated by subtracting the intrinsic mode function component IMF1 obtained in the previous step from Sp(n), and perform multiple iterations to extract all IMF components until the remaining residual meets the set stopping condition. The remaining residual is the spectral baseline.

4. The spectrometer baseline correction method based on improved empirical mode decomposition according to claim 3, wherein: Determine the initial oscillation component The judgment conditions for meeting the requirements of the eigenmode function include: The total number of poles and zero crossings does not differ by more than; At any point, the sum of the upper and lower bounds of the local extreme envelope is zero; Alternatively, the remaining residual satisfies the set stopping condition that the remaining residual satisfies the function of becoming a constant, a slope, or only one extreme point.

5. The spectrometer baseline correction method based on improved empirical mode decomposition according to claim 1, wherein: The spectrum power calibration method is performed on the measured spectrum data after the baseline is removed, including the following steps: Select a standard light source to provide standard spectral power; The measured spectrum data after baseline removal is integrated according to the set spectrum region size, and the spectrum peak area is used as the measurement power; Subtract the integrated measured power from the standard spectrum power to obtain the power calibration coefficient; The spectral power of the measured spectral data after baseline removal is calibrated based on the power calibration coefficient.

6. The spectrometer baseline correction method based on improved empirical mode decomposition according to claim 1, wherein: The measured spectral data after baseline removal are sequentially detected and removed for outliers: For the measured spectral data after baseline removal, starting from the first data point, outlier detection is performed point by point. If a point is less than or equal to zero, it is replaced by a set small constant value to ensure that all sampling points of the measured spectral data after baseline removal are values ​​greater than zero.

7. The spectrometer baseline correction method based on improved empirical mode decomposition according to claim 1, wherein: The wavelength calibration method is performed on the measured spectral data after the baseline is removed, including the following steps: Select a standard light source to provide standard spectral wavelengths; The measured spectrum data after baseline removal is fitted with a Gaussian line shape to obtain the peak wavelength; Subtract the peak wavelength obtained by fitting from the standard spectrum wavelength to obtain the wavelength calibration coefficient; The spectrum wavelength of the measured spectrum data after baseline removal is calibrated based on the wavelength calibration coefficient.

8. A spectrometer baseline correction system based on improved empirical mode decomposition, characterized in that: include: a sliding window smoothing module configured to preprocess the acquired measured spectral data using a sliding window smoothing method; An improved empirical mode decomposition module is configured to use an improved empirical mode decomposition algorithm to fit the preprocessed measured spectral data to obtain baseline data, wherein the improved empirical mode decomposition algorithm uses least squares fitting to construct upper and lower boundary envelopes of the spectral data; A baseline removal module is configured to subtract the fitted baseline data from the acquired measured spectrum data to obtain measured spectrum data after baseline removal; The correction module is configured to perform outlier detection and removal, spectral power calibration, and wavelength calibration on the measured spectral data after baseline removal to obtain baseline-corrected spectral data.

9. An electronic device, characterized in that: The invention comprises a memory and a processor and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps of the spectrometer baseline correction method based on improved empirical mode decomposition according to any one of claims 1 to 7 are completed.

10. A computer-readable storage medium, characterized in that Used to store computer instructions, which, when executed by a processor, complete the steps of the spectrometer baseline correction method based on improved empirical mode decomposition according to any one of claims 1 to 7.

Images