CN110553989A - method for removing spectrum baseline - Google Patents

Abstract

the invention discloses a method for removing a spectrum baseline, and belongs to the technical field of spectrum analysis. Firstly, reading original data; then finding out inflection points of the trend of the spectrum baseline; then fitting the spectral baselines within each region; and finally, subtracting the fitted spectrum base line of each region from the original data to realize spectrum base line correction. The spectrum baseline is divided into a plurality of areas, and polynomial fitting is carried out on the spectrum baseline in each area, so that the algorithm for deducting the spectrum baseline is realized. No parameters need to be set in advance and the spectral baselines are applicable for all different baseline trends.

Description

Method for removing spectrum baseline

Technical Field

The invention relates to the technical field of spectral analysis, in particular to a method for removing a spectral baseline.

background

there are many academic papers and patents that have published the results of algorithmic research on removing the spectral baseline, among which the following are more common: polynomial iterative fitting method, wavelet transform method, first derivative method, etc. However, in practice, these methods have respective limitations.

The polynomial iterative fitting method needs to set the polynomial order in advance to determine the baseline trend. For multi-channel systems with different wavelength ranges, baseline trends generated by noise and stray light interference are different, and a polynomial of a certain order cannot be uniformly and perfectly fitted;

The wavelet transform rule requires setting of wavelet base type, wavelet decomposition level number and high frequency filtering threshold. However, which kind of wavelet base or how many decomposition layers is suitable for a specific spectral line in practice has no literature guidance, and needs to be adjusted and tried manually continuously, which brings trouble to the pretreatment work of the spectral line. Although the literature compares the influence of different wavelet bases and different decomposition layer numbers on a certain spectral line, the result shows that the baseline removing effect is the best under a certain parameter; however, the conclusion is not popularized to other spectral lines, so that the method has no any guiding significance in practice;

For the first derivative method, the invention patent with application number 201410006439.5 discloses a spectral baseline correction method based on first derivative peak finding and spline fitting, and in the description, the original spectrum is really too simple, and the used spectral line can be regarded as a baseline expressed by a quadratic polynomial plus several standard Gaussian peaks. However, in the actually generated spectral line, besides the gaussian peak shape, there may be shoulders and overlapping peaks, which cannot be identified by the first derivative method, resulting in distortion of the spectral line.

Disclosure of Invention

the invention aims to provide a method for removing a spectrum baseline, which aims to solve the problems that the existing method for removing the spectrum baseline can only correct the baseline trend of a certain type, or needs to manually set parameters and has low processing efficiency.

In order to solve the above technical problem, the present invention provides a method for removing a spectral baseline, comprising:

Reading original data;

step two, finding out an inflection point of the baseline trend;

Step three, fitting a base line in each area;

and step four, subtracting the fitted base line of each area from the original data to realize base line correction.

Optionally, the raw data includes a plurality of CCD collected data.

Optionally, in the second step, the formula I (I) < I (I-1) < I (I-2) and I (I) < I (I +1) < I (I +2) are determined according to the spectrum intensity, and an inflection point of the baseline trend is found; wherein I (i) is the light intensity value of the ith sampling point in the spectral line; i (I-1) is the light intensity value of the I-1 th sampling point; i (I-2) is the light intensity value of the I-2 sampling point; i (I +1) is the light intensity value of the (I +1) th sampling point; i (I +2) is the light intensity value of the (I +2) th sampling point.

optionally, before the second step, the method for removing the spectral baseline further includes:

through the comparison of the intensity values between adjacent acquisition points, the acquisition points where all the minimum light intensity values on the spectral line are located are found;

and removing the acquisition points where the minimum light intensity values on the shoulder peaks and the overlapped peaks are located through a threshold judgment formula.

alternatively, the minimum value on the spectral line is found according to the spectral intensity judgment formulas I (I) < I (I-1) < I (I-2) and I (I) < I (I +1) < I (I + 2).

Optionally, the minimum values on the shoulder peak and the overlapping peak are removed by the following threshold judgment formula:

I_{Acromion, overlapping peak}>3/2*I_min；

wherein, I_{Acromion, overlapping peak}shoulder, spectral intensity over overlapping peaks; i is_minthe minimum light intensity value in the spectrum composed of all the sampling points.

Optionally, fitting the spectral baseline in each region comprises:

dividing the spectral line into n-1 areas according to the n inflection points found in the step two;

Within each individual zone, the baseline trend is a primary or secondary curve.

The invention provides a method for removing a spectrum baseline, which comprises the steps of firstly reading original data; then finding out inflection points of the trend of the spectrum baseline; then fitting the spectral baselines within each region; and finally, subtracting the fitted spectrum base line of each region from the original data to realize spectrum base line correction. The spectrum baseline is divided into a plurality of areas, and polynomial fitting is carried out on the spectrum baseline in each area, so that the algorithm for deducting the spectrum baseline is realized. No parameters need to be set in advance and the spectral baselines are applicable for all different baseline trends.

The invention has the following beneficial effects:

(1) the parameters do not need to be manually set, the presetting is reduced, and the processing efficiency is improved;

(2) The method can uniformly process any type of baseline trend;

(3) But also to complex spectral lines containing shoulder peaks and overlapping peaks.

drawings

FIG. 1 is a schematic flow diagram of a method for removing a spectral baseline according to the present invention;

FIG. 2 is a schematic diagram of importing raw data;

FIG. 3 is a schematic diagram of a quad interpolation;

FIG. 4 is a schematic diagram of the acquisition of all minima points on the spectrum;

FIG. 5 is a schematic diagram of removing a shoulder peak and a minimum point on an overlapping peak;

FIG. 6 is a schematic diagram of obtaining a baseline inflection point;

FIG. 7 is a schematic of a fit for each region;

FIG. 8 is a schematic of a baseline correction of the raw spectrum.

Detailed Description

The following describes a method for removing a spectral baseline according to the present invention in further detail with reference to the accompanying drawings and specific examples. Advantages and features of the present invention will become apparent from the following description and from the claims. It is to be noted that the drawings are in a very simplified form and are not to precise scale, which is merely for the purpose of facilitating and distinctly claiming the embodiments of the present invention.

Example one

The invention provides a method for removing a spectrum baseline, the flow of which is shown in figure 1, and the method comprises the following steps:

Step S11, reading original data;

Step S13, finding out the inflection point of the baseline trend;

step S14, fitting a baseline within each region;

In step S15, the fitted baseline for each region is subtracted from the raw data to achieve baseline correction.

firstly, reading original data, wherein the original data comprises a plurality of CCD (charge coupled device) collected data, and the size and the number of CDDs (compact disc detectors) of each CCD collected data need to be known so as to process the collected data of each CDD one by one;

Then, according to the spectral intensity judgment formulas I (I) < I (I-1) < I (I-2) and I (I) < I (I +1) < I (I +2), finding the minimum value on the baseline; wherein I (i) is the light intensity value of the ith sampling point of the spectral line; i (I-1) is the light intensity value of the I-1 th sampling point; i (I-2) is the light intensity value of the I-2 sampling point; i (I +1) is the light intensity value of the (I +1) th sampling point; i (I +2) is the light intensity value of the (I +2) th sampling point;

Removing minimum values on the shoulder peaks and the overlapped peaks, wherein the minimum values can influence the judgment of the trend of the spectrum baseline, and the removing method is according to a threshold judgment formula: i is_{acromion, overlapping peak}>3/2*I_min(ii) a Wherein, I_{Acromion, overlapping peak}the light intensity values on the shoulder peaks are overlapped; i is_minthe minimum light intensity value in the spectrum formed by all the sampling points;

then judging the formula I (I) < I (I-1) < I (I-2) and I (I) < I (I +1) < I (I +2) according to the same spectrum intensity, and finding out the inflection point of the baseline trend; similarly, I (i) is the light intensity value of the ith sampling point of the spectral line; i (I-1) is the light intensity value of the I-1 th sampling point; i (I-2) is the light intensity value of the I-2 sampling point; i (I +1) is the light intensity value of the (I +1) th sampling point; i (I +2) is the light intensity value of the (I +2) th sampling point;

fitting a spectral baseline within each region; specifically, according to the acquisition points where the minimum values of all the non-shoulder peaks or the overlapping peaks found in the second step are located, n baseline trend inflection points are found by using the spectral intensity judgment formula again, and the spectral line is divided into n-1 areas; within each individual zone, the baseline trend is a primary or secondary curve;

and finally, subtracting the fitted spectrum base line of each region from the original data to realize spectrum base line correction.

the detailed steps are as follows:

The first step is as follows: importing raw data, as shown in FIG. 2;

the second step is that: performing a quadruple interpolation, as shown in fig. 3;

The third step: obtaining all minimum value points on the spectrum through a spectrum intensity judgment formula, as shown in fig. 4;

The fourth step: removing minimum value points on the shoulder peak and the overlapped peak through a threshold judgment formula, as shown in fig. 5;

The fifth step: obtaining a baseline inflection point, as shown in FIG. 6;

and a sixth step: fitting each region as shown in fig. 7; wherein the fitting baseline is a bold dashed line at the bottom, the original signal is a thin solid line, and the fitting baseline is below the original signal;

The seventh step: the raw spectral baseline was corrected as shown in fig. 8.

the above description is only for the purpose of describing the preferred embodiments of the present invention, and is not intended to limit the scope of the present invention, and any variations and modifications made by those skilled in the art based on the above disclosure are within the scope of the appended claims.

Claims (7)

1. A method of removing a spectral baseline, comprising:

reading original data;

Step two, finding out an inflection point of the baseline trend;

Step three, fitting a base line in each area;

And step four, subtracting the fitted base line of each area from the original data to realize base line correction.

2. The method for removing a spectral baseline of claim 1, wherein said raw data comprises a plurality of CCD collected data.

3. The method for removing spectral baseline according to claim 1, wherein in the second step, the inflection points of the baseline trend are found according to the spectral intensity judgment formulas I (I) < I (I-1) < I (I-2) and I (I) < I (I +1) < I (I + 2); wherein I (i) is the light intensity value of the ith sampling point in the spectral line; i (I-1) is the light intensity value of the I-1 th sampling point; i (I-2) is the light intensity value of the I-2 sampling point; i (I +1) is the light intensity value of the (I +1) th sampling point; i (I +2) is the light intensity value of the (I +2) th sampling point.

4. The method for removing a spectral baseline of claim 1, wherein prior to step two, the method for removing a spectral baseline further comprises:

through the comparison of the intensity values between adjacent acquisition points, the acquisition points where all the minimum light intensity values on the spectral line are located are found;

and removing the acquisition points where the minimum light intensity values on the shoulder peaks and the overlapped peaks are located through a threshold judgment formula.

5. the method for removing spectral baselines according to claim 4, wherein the minima on the spectral lines are found according to the spectral intensity judgment formulas I (I) < I (I-1) < I (I-2) and I (I) < I (I +1) < I (I + 2).

6. a method of removing a spectral baseline according to claim 4, wherein the minima on the shoulder and overlapping peaks are removed by the following threshold decision formula:

I_{Acromion, overlapping peak}>3/2*I_min；

Wherein, I_{acromion, overlapping peak}The light intensity values on the shoulder peaks are overlapped; i is_minthe minimum light intensity value in the spectrum composed of all the sampling points.

7. The method of removing a spectral baseline of claim 4, wherein fitting a spectral baseline in each region comprises:

Dividing the spectral line into n-1 areas according to the n inflection points found in the step two;

within each individual zone, the baseline trend is a primary or secondary curve.