Multi-source spectrum library-based banksoldier alignment and integration method and system

Abstract

The invention provides a multi-source spectrum library-based beacon alignment and integration method and system, and the method comprises the steps: providing a multi-source spectrum library, firstly, selecting a reference spectrum library and a to-be-aligned spectrum library from the multi-source spectrum library; then, sequentially through a homonymous spectrum cluster-based sablet selection strategy, a waveform shaping-based background extraction strategy and a background component-based sablet alignment strategy, carrying out sablet alignment on the to-be-aligned spectrum library and the reference spectrum library to obtain a sablet alignment spectrum library; secondly, integrating every two of the beacon alignment spectrum library and the reference spectrum library to obtain a new integrated spectrum library; and finally, taking the integrated spectrum library as a reference spectrum library, selecting another spectrum library to be aligned from the multi-source spectrum libraries, and circularly traversing until the multi-source spectrum libraries are integrated into a single spectrum library. According to the invention, the stable and effective integration of the multi-source spectrum library is realized through a pawn alignment strategy.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and in particular to a benchmark alignment and integration method and system for a multi-source spectral library. Background Art

[0002] At present, in response to different application needs, existing technologies have built spectral libraries targeting many different application fields, and have widely applied these spectral libraries to different fields such as spectral unmixing, spectral super-resolution, spectral fitting, spectral analysis, and optical remote sensing image classification.

[0003] Compared with a single spectral library, spectral libraries from different sources contain more categories and diverse spectral information. Gathering and integrating spectral information from spectral libraries from different sources can help further improve the accuracy of spectral library information processing and the breadth of its application. However, the differences in acquisition conditions and acquisition equipment involved in spectral libraries from different sources will lead to differences in spectral information, which in turn will affect the processing and application of spectral library information.

[0004] Therefore, how to comprehensively utilize the spectral information provided by spectral libraries of different institutions and provide a multi-source spectral library integration method and system to consistently collect and integrate spectral information from multiple spectral libraries from different sources has important practical significance and actual application value. Summary of the invention

[0005] The present application aims to solve one of the technical problems in the related art at least to some extent.

[0006] To this end, the first purpose of the present application is to propose a benchmark alignment and integration method and system for a multi-source spectral library, which can consistently collect and integrate spectral information from multiple spectral libraries from different sources, thereby achieving stable and effective integration of the multi-source spectral library.

[0007] To achieve the above-mentioned purpose, the first embodiment of the present application proposes a benchmark alignment and integration method for a multi-source spectral library, comprising:

[0008] S1, provides a multi-source spectral library {Z n}, n=1,…,N,N≥2;

[0009] S2, in the multi-source spectral library {Z n Select reference spectral library Z in i and the spectral library Z to be aligned j , i≠j, i=1,2,…,N, j=1,2,…,N;

[0010] S3, cluster-based model selection strategy, respectively, in the reference spectral library Z i and the spectral library Z to be aligned jSelect a reference spectral cluster that satisfies the spectral consistency between libraries from the spectral clusters with the same name and the spectral clusters to be aligned

[0011] S4, background subtraction strategy based on waveform shaping, respectively, in the reference spectrum cluster and the spectral clusters to be aligned Extract the reference background component that satisfies the spectral consistency within the cluster and the background component to be aligned

[0012] S5, the benchmark alignment strategy based on the background component and the reference spectral library Z i Solve the spectral library Z to be aligned j , obtain the benchmark aligned spectral library Z j ′, and integrated spectral library

[0013] S6, will integrate spectral libraries As a reference spectral library, and in the multi-source spectral library {Z n}Select the next spectral library Z to be aligned k , k≠j, k=1,2,…,N; repeat steps S3 to S5 until the multi-source spectral library {Z n}Integrated into a single spectral library

[0014] Optionally, the reference spectral library Z i and the spectral library Z to be aligned j have the same spectral resolution.

[0015] Optionally, the step S3 specifically includes:

[0016] S31, obtaining the reference spectrum library Z i and the spectral library Z to be aligned j The cluster number sequence of the spectral clusters with the same name {C ij};

[0017] S32, in the reference spectral library Z i and the spectral library Z to be aligned j The sequence of spectral clusters of the same name {C ij}Select the cth spectral cluster in and spectral clusters c=1,2,…,C ij ;

[0018] S33, calculation of spectral clusters and spectral clusters Spectral correlation of Spectral correlation The mean minimum and spectral correlation of The maximum variance and spectral correlation of The ratio of the minimum mean value to the maximum variance value S c ;

[0019] S34, in the reference spectral library Z i and the spectral library Z to be aligned j The spectral cluster sequence C of the same name ij Select the e-th spectral cluster and spectral clusters c≠e,e=1,2,…,C ij , and repeat step S33 until the maximum ratio S is obtained. p The corresponding p-th spectral cluster and spectral clusters

[0020] Optionally, the reference spectral library Z i and the spectral library Z to be aligned j The spectral cluster sequence C of the same name ij The c-th spectral cluster and spectral clusters Spectral correlation of The Pearson correlation coefficient and the average spectrum in the corresponding spectrum cluster are solved in sequence and satisfy:

[0021]

[0022] in, Spectral cluster The average spectrum within the cluster; Spectral cluster The average spectrum within the cluster.

[0023] Optionally, the step S4 specifically includes:

[0024] S41, respectively cluster the reference spectrum and the spectral cluster to be aligned The spectral curve is decomposed into signal component, background component and noise component respectively;

[0025] S42, respectively obtaining the reference spectrum clusters and the spectral cluster to be aligned The reference spectral clusters are respectively and the spectral cluster to be aligned Perform waveform shaping to obtain the reference spectrum clusters after waveform shaping and the spectral cluster to be aligned The shaping background component of

[0026] S43, respectively performing waveform shaping on the reference spectrum clusters and the spectral cluster to be aligned The shaping background component is solved, and the reference background component is obtained based on the solution value and the background component to be aligned

[0027] Optionally, the reference spectrum cluster in step S42 and the spectral cluster to be aligned The maximum amplitude and the reference spectrum cluster after waveform shaping and the spectral cluster to be aligned Respectively meet:

[0028]

[0029] Among them, α is the weight factor; The reference spectrum cluster The maximum amplitude of is the spectral cluster to be aligned The maximum amplitude of The reference spectrum cluster The signal component of is the spectral cluster to be aligned The signal component of The reference spectrum cluster The noise content; is the spectral cluster to be aligned The noise content; The reference spectral cluster after shaping The background weight of The spectral clusters to be aligned after shaping background component.

[0030] Optionally, the benchmark alignment strategy based on the background component and the reference spectral library Z i Solve the spectral library Z to be aligned j Before the steps, also include:

[0031] Calculate and obtain the reference background component respectively The average and the background component to be aligned The average

[0032] Optionally, in step S5, the benchmark alignment spectral library Z j 'satisfy:

[0033]

[0034] in, is the reference background component The average value of The background component to be aligned The average value of , J is an all-one matrix.

[0035] To achieve the above-mentioned purpose, the second embodiment of the present application proposes a benchmark alignment and integration system for a multi-source spectral library, comprising:

[0036] A storage module, used for storing multi-source spectral library data;

[0037] An acquisition module, used for selectively extracting the multi-source spectral library data stored in the storage module to obtain a reference spectral library and a spectral library to be aligned;

[0038] A first processing module is used for performing calculation processing on the spectrum clusters with the same name in the reference spectrum library and the spectrum library to be aligned according to the cluster standard selection strategy, so as to screen and obtain the reference spectrum cluster and the spectrum cluster to be aligned that meet the spectrum consistency from the spectrum clusters with the same name in the reference spectrum library and the spectrum library to be aligned;

[0039] A second processing module is used to obtain a reference background component and a background component to be aligned that satisfy spectral consistency from the reference spectrum cluster and the spectrum cluster to be aligned according to a background extraction strategy of waveform shaping;

[0040] a third processing module, configured to perform computational processing on the reference background component, the background component to be aligned and the spectral library to be aligned according to the benchmark alignment strategy of the background component, so as to obtain a benchmark aligned spectral library that is benchmark aligned with the reference spectral library, and to integrate the reference spectral library and the benchmark aligned spectral library to obtain a pairwise integrated spectral library;

[0041] The control output module is used to control the switching of the data extraction path of the reference spectrum library by the acquisition module, so that the third processing module can output the integrated spectrum library to the acquisition module until the multi-source spectrum libraries are integrated into a single integrated spectrum library.

[0042] To achieve the above objectives, the third aspect of the present application proposes a computer-readable storage medium for storing non-temporary computer-readable instructions. When the non-temporary computer-readable instructions are executed by a computer, the computer executes the benchmark alignment and integration method of the multi-source spectral library described in the above embodiment.

[0043] The method and system for aligning and integrating the benchmarks of the multi-source spectral library provided in the present application at least have the following beneficial effects:

[0044] The present application provides a benchmark alignment and integration method and system for a multi-source spectral library, including providing a multi-source spectral library; first, selecting a reference spectral library and a spectral library to be aligned in the multi-source spectral library; then, aligning the spectral library to be aligned with the benchmark of the reference spectral library through a benchmark selection strategy based on spectral clusters of the same name, a background extraction strategy based on waveform shaping, and a benchmark alignment strategy based on background components, to obtain a benchmark alignment spectral library; then, integrating the benchmark alignment spectral library with the reference spectral library in pairs to obtain a new integrated spectral library; finally, using the integrated spectral library as a reference spectral library, selecting another spectral library to be aligned from the multi-source spectral library, thereby traversing in a loop until the multi-source spectral library is integrated into a single spectral library. The present application achieves stable and effective integration of multi-source spectral libraries through a benchmark alignment strategy.

[0045] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0047] Figure 1 The figure is a flow chart of a benchmark alignment and integration method of a multi-source spectral library according to an embodiment of the present application.

[0048] Figure 2 The figure is a flow chart of another benchmark alignment and integration method of a multi-source spectral library according to an embodiment of the present application.

[0049] Figure 3 A reference background component according to an embodiment of the present application is shown The average The background component to be aligned The average relationship curve.

[0050] Figure 4 A pairwise integrated spectral library according to an embodiment of the present application is shown Typical spectral curve. DETAILED DESCRIPTION

[0051] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0052] At present, the research on the integration methods of spectral libraries from multiple different sources is relatively limited. The usual integration methods are spectral library integration methods based on data models, or spectral library integration methods that simply stack all spectra in different spectral libraries in sequence. However, the spectral library integration method based on data models can only improve the interoperability of spectral libraries obtained from different institutions. The spectral library integration method based on simple stacking cannot effectively improve the accuracy of subsequent applications and processing, especially the accuracy of application processing such as spectral super-resolution and spectral fitting. In addition, the above methods do not fully consider the different effects of acquisition conditions and acquisition equipment differences on spectral signal components and background components, which makes it difficult for the existing multi-source spectral library integration process to ensure the overall consistency of the statistical characteristics of the spectrum after integration, which also leads to the inability of the existing multi-source spectral library integration method to obtain satisfactory multi-source spectral library integration effects.

[0053] Based on the above problems, an embodiment of the present application provides a benchmark alignment and integration method and system for a multi-source spectral library, which selects a reference spectral library and a spectral library to be aligned from the multi-source spectral library, and aligns the spectral library to be aligned with the benchmark of the reference spectral library through a benchmark selection strategy based on spectral clusters of the same name, a background extraction strategy based on waveform shaping, and a benchmark alignment strategy based on background components, so that the spectral library to be aligned after the benchmark alignment can be directly integrated with the reference spectral library into a new integrated spectral library, and finally the integrated spectral library is used as the reference spectral library, and another spectral library to be aligned is selected from the multi-source spectral library, thereby cyclically traversing until the multi-source spectral libraries are integrated into a single spectral library, thereby achieving stable and effective integration of the multi-source spectral libraries.

[0054] According to a first aspect of the present application, a method for aligning and integrating a multi-source spectral library is provided, such as Figure 1 and Figure 2 As shown, the method comprises the following steps:

[0055] S1, provides a multi-source spectral library {Z n}, n=1,…,N,N≥2;

[0056] S2, in the multi-source spectral library {Z n Select reference spectral library Z in i and the spectral library Z to be aligned j , i≠j, i=1,2,…,N, j=1,2,…,N;

[0057] S3, cluster-based model selection strategy, respectively, in the reference spectral library Z i and the spectral library Z to be aligned j Select a reference spectral cluster that satisfies the spectral consistency between libraries from the spectral clusters with the same name and the spectral clusters to be aligned

[0058] S4, background subtraction strategy based on waveform shaping, respectively, in the reference spectrum cluster and the spectral clusters to be aligned Extract the reference background component that satisfies the spectral consistency within the cluster and the background component to be aligned

[0059] S5, the benchmark alignment strategy based on the background component and the reference spectral library Z i Solve the spectral library Z to be aligned j , obtain the benchmark aligned spectral library Z j ′, and integrated spectral library

[0060] S6, will integrate spectral libraries As a reference spectral library, and in the multi-source spectral library {Z n}Select the next spectral library Z to be aligned k , k≠j, k=1,2,…,N; repeat steps S3 to S5 until the multi-source spectral library {Z n}Integrated into a single spectral library

[0061] In some embodiments, the multi-source spectral library {Z n} can be N different source spectral libraries, N is a natural number greater than or equal to 2, and the multi-source spectral library {Z n Any spectral library in} should have the same spectral resolution, for example, the i-th spectral library Among them, λ x With M i are the number of bands and the number of pixels in each band of the i-th spectral library of the input multi-source spectral library respectively.

[0062] As an example, a multi-source spectral library {Z n} 480 typical mineral spectra from the United States Geological Survey (USGS) spectral library and 430 typical mineral spectra from the Jet Propulsion Laboratory (JPL) spectral library were used. The spectral coverage range of both is 400nm~2500nm. At the same time, the USGS spectral library Z 1 With JPL Spectral Library Z 2 The spectral resolution is consistent.

[0063] In some embodiments, step S3 specifically includes:

[0064] S31, obtain reference spectrum library Z iand the spectral library Z to be aligned j The sequence of spectral clusters of the same name {C ij}.

[0065] S32, based on the reference spectral library Z i and the spectral library Z to be aligned j The sequence of spectral clusters of the same name {C ij}Select the cth spectral cluster in and spectral clusters c=1,2,…,C ij ;

[0066] S33, calculation of spectral clusters and spectral clusters Spectral correlation of Spectral correlation The mean minimum and spectral correlation of The maximum variance and spectral correlation of The ratio of the minimum mean value to the maximum variance value S c ;

[0067] S34, in the reference spectral library Z i and the spectral library Z to be aligned j The spectral cluster sequence C of the same name ij Select the e-th spectral cluster and spectral clusters c≠e,e=1,2,…,C ij , and repeat step S33 until the maximum ratio S is obtained. p The corresponding p-th spectral cluster and spectral clusters

[0068] Specifically, since the spectral consistency between multi-source spectral libraries is the key to the integration of multi-source spectral libraries, and this spectral consistency exists in the spectral clusters with the same name corresponding to different spectral libraries, it is assumed that the reference spectral library Z i and the spectral library Z to be aligned j The spectral clusters of the same name contain C ij The clusters have the same name as the spectral clusters, and C ij The spectral clusters with the same name together constitute the spectral cluster sequence with the same name in step S31. ij}, a sequence of spectral clusters with the same name {C ij Any cluster (cth cluster) in the reference spectral library Z i The corresponding spectral cluster In the spectral library Z to be aligned j The corresponding spectral cluster

[0069] Therefore, based on the Pearson correlation coefficient, spectral cluster The average spectrum within the cluster and spectral clusters The average spectrum within the cluster You can calculate the spectral cluster Intra-cluster spectral correlation of and spectral clusters Intra-cluster spectral correlation in,

[0070]

[0071] Secondly, we solve and calculate the intra-cluster spectral correlation The minimum mean value of Varimax The ratio S c .

[0072] Considering that the spectral cluster with higher spectral consistency within the cluster is more stable and representative, the maximum value S of the ratio of the two is obtained. p The corresponding position p, and the spectral cluster with the same name as the pth cluster is taken as the model spectral cluster and The above process can be further described as:

[0073]

[0074] In some embodiments, step S4 specifically includes:

[0075] S41, respectively cluster the reference spectrum and the spectral cluster to be aligned The spectral curve is decomposed into signal component, background component and noise component respectively;

[0076] S42, respectively obtaining the reference spectrum clusters and the spectral cluster to be aligned The reference spectral clusters are respectively and the spectral cluster to be aligned Perform waveform shaping; respectively obtain the reference spectrum clusters after waveform shaping and the spectral cluster to be aligned The shaping background component of

[0077] S43, respectively performing waveform shaping on the reference spectrum clusters and the spectral cluster to be aligned The shaping background component is solved, and the reference background component is obtained based on the solution value and the background component to be aligned

[0078] Specifically, in theory, the multi-source spectral library {Z n Any spectral library in} Any spectral library Z i A given spectral curve can be decomposed into signal component, background component and noise component. In other words, the reference spectral library Z i Spectral library Z j They satisfy the following relations respectively, namely:

[0079]

[0080] in, is the reference spectral cluster signal component, and is the spectral cluster to be aligned signal component, and is the reference spectral cluster The background component of Spectral clusters to be aligned The background component of is the reference spectral cluster The noise component of is the spectral cluster to be aligned The noise component of

[0081] Secondly, the reference spectral cluster and the spectral clusters to be aligned The maximum amplitude satisfies the following relationship, namely:

[0082]

[0083] Wherein, λ is the wavelength, α is the weight factor, and α>1.

[0084] Therefore, for the reference spectral cluster and the spectral clusters to be aligned The waveform shaping can be specifically expressed as the following relationship, namely:

[0085]

[0086] in, are the reference spectral clusters after shaping signal component and background component of is the spectral cluster to be aligned after shaping The signal component and background component of the reference spectrum cluster after shaping and the spectral clusters to be aligned The signal component Background Component Both are greater than zero.

[0087] Finally, the variational mode decomposition (VMD) is used to solve the reference spectrum cluster after waveform shaping. Background weight The low frequency mode u i The spectral cluster to be aligned Background weight The low frequency mode u j , and then obtain the reference spectrum cluster before shaping and the spectral clusters to be aligned The background component is:

[0088]

[0089] In some embodiments, the background component-based benchmark alignment strategy described in step S5 is based on the fact that slight differences in chemical composition and physical structure in spectral clusters with the same name generally only affect the signal component of the spectrum rather than the background component. The average The background component to be aligned The average Treat the aligned spectral library Z j Solve it and you can get the benchmark aligned spectral library Z j ',Right now:

[0090]

[0091] in, is the reference background component The average value of The background component to be aligned The average value of , J is an all-one matrix.

[0092] Therefore, the benchmark aligned spectral library Z j ′ is the reference spectral library Z that can be directly aligned with the benchmark i Perform pairwise integration to obtain an integrated spectral library Thus, the reference spectral library Z i and the spectral library Z to be aligned j integration.

[0093] Similarly, the integrated spectral library As a reference spectral library, the multi-source spectral library {Z n A spectral library Z in k As the spectral library to be aligned, k≠j, k=1,2,…,N, that is, the spectral library Zk Integrated spectral library Integrate, and repeat this cycle, and finally the multi-source spectral library {Z n}Integrated into a single spectral library The stable and effective integration of multi-source spectral libraries is achieved.

[0094] As an example, a multi-source spectral library {Z n}Specifically, it can include the preprocessed USGS spectral library Z 1 With JPL Spectral Library Z 2 , USGS spectral library Z 1 The size is 188×480, that is JPL Spectral Library Z 2 The size is 188×430, that is

[0095] When selectively extracting the spectral library Z 1 As a reference spectral library, spectral library Z 2 When the spectral library to be aligned is used, the reference spectral library Z 1 and the spectral library Z to be aligned 2 The spectral cluster sequence of the same name is "almandine", "albaite", "calcite", "diopside", "pyroxene" and "montmorillonite" (C ij =6), from which reference spectrum clusters that meet spectral consistency can be screened and obtained and the spectral clusters to be aligned

[0096] Correspondingly, the background extraction strategy based on waveform shaping can also be used according to the reference spectrum cluster. Calculate the reference background component And according to the spectral clusters to be aligned Calculate and obtain the background component to be aligned When the weight factor α is set to α=1.1, the reference background component The background component to be aligned The average The relationship curve is as follows Figure 2 shown.

[0097] Based on the background component standard alignment strategy, solve the aligned spectral library Z′ 2 , you can convert the USGS spectral library Z 1 With JPL Spectral Library Z 2 Integrate with each other to obtain a pairwise integrated spectral library Among them, the spectral library is integrated two by two Some typical spectra in Figure 3 shown.

[0098] According to a second aspect of the present application, a benchmark alignment and integration system for a multi-source spectral library is provided, the system comprising:

[0099] A storage module, used for storing input multi-source spectral library data;

[0100] An acquisition module, used for selectively extracting the multi-source spectral library data stored in the storage module to obtain a reference spectral library and a spectral library to be aligned;

[0101] A first processing module is used for performing calculation processing on the spectrum clusters with the same name in the reference spectrum library and the spectrum library to be aligned according to the cluster standard selection strategy, so as to screen and obtain the reference spectrum cluster and the spectrum cluster to be aligned that meet the spectrum consistency from the spectrum clusters with the same name in the reference spectrum library and the spectrum library to be aligned;

[0102] A second processing module is used to obtain a reference background component and a background component to be aligned that satisfy spectral consistency from the reference spectrum cluster and the spectrum cluster to be aligned according to a background extraction strategy of waveform shaping;

[0103] a third processing module, configured to perform computational processing on the reference background component, the background component to be aligned and the spectral library to be aligned according to the benchmark alignment strategy of the background component, so as to obtain a benchmark aligned spectral library that is benchmark aligned with the reference spectral library, and to integrate the reference spectral library and the benchmark aligned spectral library to obtain a pairwise integrated spectral library;

[0104] The control output module is used to control the switching of the data extraction path of the reference spectrum library by the acquisition module, so that the third processing module can output the integrated spectrum library to the acquisition module until the multi-source spectrum libraries are integrated into a single integrated spectrum library.

[0105] According to the third aspect of the present application, a computer-readable storage medium is also provided for storing non-temporary computer-readable instructions. When the non-temporary computer-readable instructions are executed by a computer, the computer executes the benchmark alignment and integration method of a multi-source spectral library described in any of the above embodiments.

[0106] In summary, the present application provides a benchmark alignment and integration method and system for a multi-source spectral library, including providing a multi-source spectral library, first, selecting a reference spectral library and a spectral library to be aligned in the multi-source spectral library; then, aligning the spectral library to be aligned with the benchmark of the reference spectral library through a benchmark selection strategy based on spectral clusters of the same name, a background extraction strategy based on waveform shaping, and a benchmark alignment strategy based on background components, to obtain a benchmark aligned spectral library; then, integrating the benchmark aligned spectral library with the reference spectral library in pairs to obtain a new integrated spectral library; finally, using the integrated spectral library as the reference spectral library, selecting another spectral library to be aligned from the multi-source spectral library, thereby traversing in a loop until the multi-source spectral library is integrated into a single spectral library. The present application achieves stable and effective integration of multi-source spectral libraries through a benchmark alignment strategy.

[0107] In the description of the foregoing embodiments, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the described specific features, structures, materials or characteristics may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they contradict each other.

[0108] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

Claims

1. A benchmark alignment and integration method for a multi-source spectral library, characterized in that: include: S1, provides a multi-source spectral library {Z n }, n=1,...,N, N≥2; S2, in the multi-source spectral library {Z n Select reference spectral library Z in i and the spectral library Z to be aligned j ,i≠j,i=1,2,...,N,j=1,2,...,N; S3, cluster-based model selection strategy, respectively, in the reference spectral library Z i and the spectral library Z to be aligned j Select a reference spectral cluster that satisfies the spectral consistency between libraries from the spectral clusters with the same name and the spectral clusters to be aligned S4, background subtraction strategy based on waveform shaping, respectively, in the reference spectrum cluster and the spectral clusters to be aligned Extract the reference background component that satisfies the spectral consistency within the cluster and the background component to be aligned S5, the benchmark alignment strategy based on the background component and the reference spectral library Z i Solve the spectral library Z to be aligned j , obtain the benchmark aligned spectral library Z′ j , and integrated spectral libraries S6, will integrate spectral libraries As a reference spectral library, and in the multi-source spectral library {Z n }Select the next spectral library Z to be aligned k , k≠j, k=1,2,...,N; repeat steps S3 to S5 until the multi-source spectral library {Z n }Integrated into a single spectral library 2. The method according to claim 1, characterized in that The reference spectral library Z i and the spectral library Z to be aligned j have the same spectral resolution.

3. The method according to claim 1, characterized in that The step S3 specifically includes: S31, obtaining the reference spectrum library Z i and the spectral library Z to be aligned j The cluster number sequence of the spectral clusters with the same name {C ij }; S32, in the reference spectral library Z i and the spectral library Z to be aligned j The sequence of spectral clusters of the same name {C ij }Select the cth spectral cluster in and spectral clusters S33, calculation of spectral clusters and spectral clusters Spectral correlation of and Spectral correlation and The mean minimum and spectral correlation of and The maximum variance and spectral correlation of and The ratio of the minimum mean value to the maximum variance value S c ; S34, in the reference spectral library Z i and the spectral library Z to be aligned j The spectral cluster sequence C of the same name ij Select the e-th spectral cluster and spectral clusters c≠e, e=1,2,...,C ij , and repeat step S33 until the maximum ratio S is obtained. p The corresponding p-th spectral cluster and spectral clusters 4. The method according to claim 3, characterized in that: The reference spectral library Z i and the spectral library Z to be aligned j The spectral cluster sequence C of the same name ij The c-th spectral cluster and spectral clusters Spectral correlation of and The Pearson correlation coefficient and the average spectrum in the corresponding spectrum cluster are solved in sequence and satisfy: in, Spectral cluster The average spectrum within the cluster; Spectral cluster The average spectrum within the cluster.

5. The method according to claim 1, characterized in that The step S4 specifically includes: S41, respectively cluster the reference spectrum and the spectral cluster to be aligned The spectral curve is decomposed into signal component, background component and noise component respectively; S42, respectively obtaining the reference spectrum clusters and the spectral cluster to be aligned The reference spectral clusters are respectively and the spectral cluster to be aligned Perform waveform shaping; respectively obtain the reference spectrum clusters after waveform shaping and the spectral cluster to be aligned The shaping background component of S43, respectively performing waveform shaping on the reference spectrum clusters and the spectral cluster to be aligned The shaping background component is solved, and the reference background component is obtained based on the solution value and the background component to be aligned 6. The method according to claim 5, characterized in that The reference spectrum cluster in step S42 and the spectral cluster to be aligned The maximum amplitude and the reference spectrum cluster after waveform shaping and the spectral cluster to be aligned Respectively meet: Among them, α(is the weight factor; The reference spectrum cluster The maximum amplitude of is the spectral cluster to be aligned The maximum amplitude of is the reference spectral cluster The signal component of is the spectral cluster to be aligned The signal component of The reference spectrum cluster The noise content; is the spectral cluster to be aligned The noise content; The reference spectral cluster after shaping The background weight of The spectral clusters to be aligned after shaping background component.

7. The method according to claim 1, characterized in that The benchmark alignment strategy based on the background component and the reference spectral library Z i Solve the spectral library Z to be aligned j Before the steps, also include: Calculate and obtain the reference background component respectively The average and the background component to be aligned The average 8. The method according to claim 6, characterized in that In step S5, the benchmark aligns the spectral library Z′ j satisfy: in, is the reference background component The average value of The background component to be aligned The average value of , J is an all-one matrix.

9. A benchmark alignment and integration system for multi-source spectral libraries, characterized in that: include: A storage module, used for storing multi-source spectral library data; An acquisition module, used for selectively extracting the multi-source spectral library data stored in the storage module to obtain a reference spectral library and a spectral library to be aligned; A first processing module is used for performing calculation processing on the spectrum clusters with the same name in the reference spectrum library and the spectrum library to be aligned according to the cluster standard selection strategy, so as to screen and obtain the reference spectrum cluster and the spectrum cluster to be aligned that meet the spectrum consistency from the spectrum clusters with the same name in the reference spectrum library and the spectrum library to be aligned; A second processing module is used to obtain a reference background component and a background component to be aligned that satisfy spectral consistency from the reference spectrum cluster and the spectrum cluster to be aligned according to a background extraction strategy of waveform shaping; a third processing module, configured to perform computational processing on the reference background component, the background component to be aligned and the spectral library to be aligned according to the benchmark alignment strategy of the background component, so as to obtain a benchmark aligned spectral library that is benchmark aligned with the reference spectral library, and to integrate the reference spectral library and the benchmark aligned spectral library to obtain a pairwise integrated spectral library; The control output module is used to control the switching of the data extraction path of the reference spectrum library by the acquisition module, so that the third processing module can output the integrated spectrum library to the acquisition module until the multi-source spectrum libraries are integrated into a single integrated spectrum library.

10. A computer-readable storage medium for storing non-transitory computer-readable instructions, characterized in that: When the non-transitory computer-readable instructions are executed by a computer, the computer is caused to execute the benchmark alignment and integration method for a multi-source spectral library as recited in any one of claims 1 to 8.

Images