Spectral curve similarity evaluation method based on bionic vision

Abstract

The invention discloses a spectral curve similarity evaluation method based on bionic vision, and solves the technical problems that only the size difference of spectrums can be reflected, the shape difference of the spectrums cannot be well reflected, and the absorption peaks of the spectrums cannot be well reflected. The method is similar to a human eye color distinguishing method, windowing processing is carried out on the spectrum curve obtained by the spectrograph, interested spectrum band information is reserved, spectrum band information except the interested spectrum band is compressed, and similarity evaluation of spectrum curve key information is facilitated; the method has the advantages of being small in calculation amount and easy to realize real-time processing of the algorithm.

Description

Technical Field

[0001] The invention relates to a method for evaluating spectral curve similarity, and in particular to a method for evaluating spectral curve similarity based on bionic vision. Background Art

[0002] A spectrum curve refers to the relationship curve between the wavelength of the spectrum and other variables. For example, if the intensity of the incident light remains unchanged, the energy corresponding to light of different frequencies is measured, and a curve of the relationship between the two is drawn.

[0003] Spectra are called the optical "fingerprints" of substances. Spectral curves are widely used in the classification and identification of substances, which requires the evaluation of spectral curve similarity, and the similarity of two or more spectral curves is judged based on the evaluation results. The existing spectral curve similarity evaluation methods mainly include the following:

[0004] 1. Euclidean distance method: The similarity between spectra is measured by the Euclidean distance between two spectral curves. This method can only reflect the size difference of the spectrum, but cannot reflect the shape difference of the spectrum well;

[0005] 2. Spectral angle matching method: This method evaluates the similarity by calculating the angle between two spectral curves. This method cannot well reflect the characteristics of the spectrum such as absorption peaks. Summary of the invention

[0006] The purpose of the present invention is to solve the technical problems that the existing spectral curve similarity evaluation method can only reflect the size difference of the spectrum, but cannot well reflect the shape difference of the spectrum, and cannot well reflect the absorption peak of the spectrum, and to provide a spectral curve similarity evaluation method based on bionic vision.

[0007] In order to achieve the above object, the present invention adopts the following technical solution:

[0008] A spectral curve similarity evaluation method based on bionic vision is special in that it includes the following steps:

[0009] S1. Obtaining original spectral data using a dispersive spectrometer;

[0010] S2, determining the number of spectrum segments N of the original spectrum data and the wavelength range of each spectrum segment, N ≥ 3, and dividing the original spectrum data into N spectrum segments according to the wavelength range, and drawing the original spectrum curves of the N spectrum segments;

[0011] S3, normalizing the original spectrum curves of the N spectrum segments to obtain normalized spectrum curves of the N spectrum segments;

[0012] S4, determining a Gaussian window function of a normalized spectrum curve of each spectrum segment;

[0013] S5, performing windowing processing on the normalized spectrum curve of each spectrum segment using a corresponding window function to obtain windowed curves of N spectrum segments;

[0014] S6, calculating the windowed curve integral value of each spectrum segment;

[0015] S7, select reference spectrum data in the spectrum curve database, and divide the reference spectrum data into N reference spectrum segments according to the number of spectrum segments N in S2 and the wavelength range of each spectrum segment, draw N reference spectrum curves, and calculate the reference windowed curve integral value of each spectrum segment according to the method in S3-S6;

[0016] S8. Evaluate whether the original spectral curve is similar to the reference spectral curve according to the integral value of the windowed curve of each spectral segment in S6 and the integral value of the reference windowed curve of each spectral segment in S7, and complete the spectral curve similarity evaluation based on bionic vision.

[0017] In S1, the original spectrum data includes wavelengths and energies corresponding to the corresponding wavelengths.

[0018] In S4, the expression of the Gaussian window function is:

[0019]

[0020] Among them, ω i (n) is the window function within a certain band, n i0 is the wavelength of the center point of the i-th spectrum segment, α i is the Gaussian window width parameter of the i-th spectral segment, n is the wavelength in the i-th spectral segment, where n is a positive integer.

[0021] In S5, the expression for windowing the normalized spectrum curve of each spectrum segment using a corresponding window function is:

[0022] f(n i )′=f(n i )·ω(n i );

[0023] Among them, f i (n)′ is the windowing curve of the i-th spectrum segment, f i (n) is the original spectral curve in the i-th spectral segment.

[0024] In S6, the formula for calculating the windowed curve integral value of each spectrum segment is:

[0025]

[0026] Among them, t iis the integral value of the windowed curve in the i-th spectral segment, m is the minimum wavelength of the i-th spectral segment, and M is the maximum wavelength of the i-th spectral segment.

[0027] The specific process of S8 is as follows:

[0028] S8.1. For the windowed curve integral value of each spectral segment in S6 and the reference windowed curve integral value of each spectral segment in S7, calculate the error value △t in the i-th spectral segment. i , the formula is:

[0029] △t i =(t i -t i ') 2 ;

[0030] Among them, t′ i is the integral value of the reference windowing curve in the i-th spectrum segment;

[0031] S8.2. Set the error threshold P in the i-th spectrum segment i ;

[0032] S8.3, judging the error value in the i-th spectrum segment and the error threshold P i If the error value in all spectrum segments is not greater than the corresponding error threshold P i , then the original spectral curve is similar to the reference spectral curve, otherwise they are not similar.

[0033] Beneficial effects of the present invention:

[0034] 1. The present invention provides a spectral curve similarity evaluation method based on bionic vision, which is similar to the method of human eyes distinguishing colors. The spectral curve obtained by the spectrometer is subjected to windowing processing, so that the spectral band information of interest is retained and the spectral band information other than the spectral band of interest is compressed, which is conducive to the similarity evaluation of the key information of the spectral curve;

[0035] 2. The present invention provides a spectral curve similarity evaluation method based on bionic vision, which has small calculation amount and is easy to realize real-time processing of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a color perception curve of the human eye;

[0037] Figure 2 It is a color perception curve of the human eye simulated by the window function;

[0038] Figure 3 It is a schematic diagram of the original spectrum in Example S1 of a spectral curve similarity evaluation method based on bionic vision of the present invention;

[0039] Figure 4Schematic diagram of a Gaussian window function used in Example S3 of a spectral curve similarity evaluation method based on bionic vision of the present invention;

[0040] Figure 5 It is a curve graph obtained after windowing the original spectral curve of each spectral segment using a corresponding window function in Example S4 of a spectral curve similarity evaluation method based on bionic vision of the present invention;

[0041] Figure 6 It is a schematic diagram of visualization of similarity of three spectrum segments when the original spectrum curve is similar to the reference spectrum curve in Example S7 of a spectrum curve similarity evaluation method based on bionic vision of the present invention;

[0042] Figure 7 It is a schematic diagram of visualization of the similarity of three spectral segments when the original spectral curve is not similar to the reference spectral curve in Example S7 of the spectral curve similarity evaluation method based on bionic vision of the present invention.

[0043] Specific implementation party

[0044] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] This embodiment is based on the fact that the human eye is a precise and efficient imaging system. The human retina is composed of cone cells that recognize the three colors of red (R), green (G), and blue (B) and a type of rod cell. Under good light conditions, the three types of cone cells perceive the three colors of red, green, and blue. The color perception curve is as follows: Figure 1 As shown in the figure, after computer synthesis, humans can perceive the colorful world. According to some literature, the human eye can perceive more than 160,000 colors. This discrimination ability is very powerful. From the spectral response curve of cone cells, it can be seen that the human eye's perception of red (R), green (G), and blue (B) is not a "pure color". Each color has a range similar to a mountain peak, and there is overlap between red, green, and blue.

[0046] The color curve of human eye perception simulated by window function is as follows Figure 2 As shown, it can be seen that in the human eye perception color curve simulated by the window function, the spectral information of interest is retained, and the spectral information outside the spectral band of interest is compressed, which is conducive to the similarity evaluation of the key information of the spectral curve.

[0047] Some people have the ability of "tetrachromatic vision", which means that their eyes can perceive a fourth color besides red, green and blue, so they can distinguish more kinds of colors.

[0048] Based on the above theoretical background, the present invention proposes a method for evaluating the similarity of spectral curves based on bionic vision, which performs windowing processing on the dispersion spectrum, uses the processed data to obtain a new spectral curve, and then performs a similarity evaluation method.

[0049] The present embodiment provides a method for evaluating spectral curve similarity based on bionic vision, comprising the following steps:

[0050] S1. Use a dispersive spectrometer to obtain original spectral data in 401 visible light bands, where the visible light band ranges from 350nm to 750nm; the original spectral data includes wavelength and energy corresponding to the corresponding wavelength; simulate the original spectral curve of the visible light band obtained by the dispersive spectrometer as shown in Figure 3 shown.

[0051] S2, determining the number of spectrum segments of the original spectrum data, 3, and the wavelength range of each spectrum segment, 350n-550nm, 400nm-700nm, 400nm-750nm, and dividing the original spectrum data into 3 spectrum segments according to the wavelength range, and drawing the original spectrum curve of each spectrum segment;

[0052] S3, normalizing the original spectral curves of the three spectral segments to obtain normalized spectral curves of the three spectral segments;

[0053] S4, determine the Gaussian window function of the normalized spectrum curve of each spectrum segment; the window function used in this embodiment is a Gaussian window function, such as Figure 4 As shown, the expression of the Gaussian window function is:

[0054]

[0055] Among them, ω i (n) is the window function within a certain band, n i0 is the wavelength of the center point of the i-th spectrum segment, α i is the Gaussian window width parameter of the i-th spectral segment, n is the wavelength in the i-th spectral segment, where n is a positive integer.

[0056] In this embodiment, the center point of the first window function is 450nm, and the width is 200nm (350nm-550nm); the center point of the second window function is 540nm, and the width is 300nm (400nm-700nm); the center point of the third window function is 570nm, and the width is 350nm (400nm-750nm). Thus, three window functions ω1(n), ω2(n), and ω3(n) are obtained;

[0057]

[0058] S5. Use the corresponding Gaussian window function to perform windowing processing on the original spectrum curve of each spectrum segment, such as Figure 5 As shown, the windowing curves of the three spectral segments are obtained; the expression is:

[0059] f1(n)′=f1(n)·ω1(n),350≤n<550,

[0060] f2(n)′=f2(n)·ω2(n),400≤n<700,

[0061] f3(n)′=f3(n)·ω3(n),400 <n≤750,

[0062] Among them, f1(n)′, f2(n)′, and f3(n)′ are the windowed curves of the 1st, 2nd, and 3rd spectral segments, and f1(n), f2(n), and f3(n) are the original spectral curves in the 1st, 2nd, and 3rd spectral segments.

[0063] S6. Calculate the integral value of the windowed curve in each spectrum segment; the formula is:

[0064]

[0065] Among them, t1, t2, and t3 are the integral values ​​of the windowed curves in the first, second, and third spectral segments.

[0066] S7, select reference spectrum data in the spectrum curve database, and according to the number of spectrum segments 3 in S2 and the wavelength range of each spectrum segment, the first wavelength range is 350nm-550nm, the center point is 450nm, and the width is 200nm; the second wavelength range is 400nm-700nm, the center point is 540nm, and the width is 300nm; the third wavelength range is 400nm-750nm, the center point is 570nm, and the width is 350nm, draw the reference spectrum curve of each spectrum segment, and the reference windowed curve integral value of each spectrum segment according to the method in S3-S6, t′1, t′2, t′3;

[0067] S8, evaluating whether the original spectral curve is similar to the reference spectral curve according to the integral value of the windowed curve of each spectral segment in S6 and the integral value of the reference windowed curve of each spectral segment in S7, and completing the spectral curve similarity evaluation based on bionic vision. The specific process is:

[0068] S8.1. For the integral value of the windowed curve in each spectral segment in S6 and the integral value of the reference windowed curve in the corresponding spectral segment in S7, calculate the error values ​​△t1, △t2, and △t3. The formula is:

[0069] △t1=(t1-t1') 2 ;

[0070] △t2=(t2-t'2) 2 ;

[0071] △t3=(t3-t'3) 2 .

[0072] S8.2, setting the error thresholds P1, P2, and P3 within the third spectrum segment;

[0073] S8.3, determine the error value in the third spectrum segment and the error thresholds P1, P2, and P3. If

[0074] △t1≤P1,△t2≤P2,△t3≤P3.

[0075] If the original spectrum curve is similar to the reference spectrum curve, then they are not similar.

[0076] When the original spectrum curve is similar to the reference spectrum curve, the similarity visualization diagrams of the three spectrum segments are drawn respectively, such as Figure 6 As shown; when the original spectrum curve is not similar to the reference spectrum curve or the similarity is very small, a similarity visualization diagram of the three spectrum segments is drawn respectively, as shown Figure 7 As shown. Figure 6 and Figure 7 It can be seen that if the difference between the original spectral curve and the reference spectral curve in each band is less than the error threshold, the graphs of the two spectral curves drawn according to the integral values ​​of the windowed curves of each spectral band should be similar in shape.

[0077] The above description is only a specific embodiment of the present invention, and a comparison of the effects of the specific embodiments and the related comparative examples, but the protection scope of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A spectral curve similarity evaluation method based on bionic vision, characterized in that: The following steps are involved: S1. Obtaining original spectral data using a dispersive spectrometer; S2, determining the number of spectrum segments N of the original spectrum data and the wavelength range of each spectrum segment, N ≥ 3, and dividing the original spectrum data into N spectrum segments according to the wavelength range, and drawing the original spectrum curves of the N spectrum segments; S3, normalizing the original spectrum curves of the N spectrum segments to obtain normalized spectrum curves of the N spectrum segments; S4, determining a Gaussian window function of a normalized spectrum curve of each spectrum segment; S5, performing windowing processing on the normalized spectrum curve of each spectrum segment using a corresponding window function to obtain windowed curves of N spectrum segments; S6, calculating the windowed curve integral value of each spectrum segment; S7, select reference spectrum data in the spectrum curve database, and divide the reference spectrum data into N reference spectrum segments according to the number of spectrum segments N in S2 and the wavelength range of each spectrum segment, draw N reference spectrum curves, and calculate the reference windowed curve integral value of each spectrum segment according to the method in S3-S6; S8. Evaluate whether the original spectral curve is similar to the reference spectral curve according to the integral value of the windowed curve of each spectral segment in S6 and the integral value of the reference windowed curve of each spectral segment in S7, and complete the spectral curve similarity evaluation based on bionic vision.

2. The method for evaluating spectral curve similarity based on bionic vision according to claim 1, characterized in that: In S1, the original spectrum data includes wavelengths and energies corresponding to the corresponding wavelengths.

3. The method for evaluating spectral curve similarity based on bionic vision according to claim 2, characterized in that: In S4, the expression of the Gaussian window function is: Among them, ω i (n) is the window function within a certain band, n i0 is the wavelength of the center point of the i-th spectrum segment, α i is the Gaussian window width parameter of the i-th spectral segment, n is the wavelength in the i-th spectral segment, where n is a positive integer.

4. The method for evaluating spectral curve similarity based on bionic vision according to claim 3, characterized in that: In S5, the expression for windowing the normalized spectrum curve of each spectrum segment using a corresponding window function is: f(n i )′=f(n i )·ω(n i ); Among them, f i (n)′ is the windowing curve of the i-th spectrum segment, f i (n) is the original spectral curve in the i-th spectral segment.

5. The method for evaluating spectral curve similarity based on bionic vision according to claim 4, characterized in that: In S6, the formula for calculating the windowed curve integral value of each spectrum segment is: Among them, t i is the integral value of the windowed curve in the i-th spectral segment, m is the minimum wavelength of the i-th spectral segment, and M is the maximum wavelength of the i-th spectral segment.

6. The method for evaluating spectral curve similarity based on bionic vision according to claim 5, characterized in that: The specific process of S8 is as follows: S8.

1. For the windowed curve integral value of each spectral segment in S6 and the reference windowed curve integral value of each spectral segment in S7, calculate the error value △t in the i-th spectral segment. i , the formula is: △t i =(t i -t’ i ) 2 ; Among them, t' i is the integral value of the reference windowing curve in the i-th spectrum segment; S8.

2. Set the error threshold P in the i-th spectrum segment i ; S8.3, judging the error value in the i-th spectrum segment and the error threshold P i If the error value in all spectrum segments is not greater than the corresponding error threshold P i , then the original spectral curve is similar to the reference spectral curve, otherwise they are not similar.

Images