Mineral identification method and system based on infrared spectrum

Abstract

The invention discloses a mineral identification method and system based on an infrared spectrum. The method comprises the following steps: acquiring a mineral spectrum of a to-be-identified sample; processing the mineral spectrum to obtain wavelengths, absorption depths and prominences of a plurality of absorption peaks of the sample to be identified; aiming at each absorption peak P of the sample to be identified, using wavelength as key input, querying in the standard mineral spectrum query tree to obtain a plurality of standard mineral absorption peaks in a preset value x range, and screening to obtain a standard mineral absorption peak RP closest to the query wavelength; according to the screened standard mineral absorption peak RP, obtaining all standard minerals possibly corresponding to the to-be-identified sample; if all the characteristic absorption peaks of a certain standard mineral correspond to each other in the to-be-identified sample, adding the standard mineral into a list of candidate minerals of the to-be-identified sample; and identifying the main mineral of the sample to be identified from the candidate mineral list. According to the invention, the expandability and reliability of the mineral identification method are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of infrared spectroscopy in mineral exploration, and more particularly to a mineral identification method and system based on infrared spectroscopy. Background Technology

[0002] Mineral infrared spectroscopy primarily utilizes different infrared bands to identify various minerals and alteration information. It is generally divided into three types: near-infrared (400-1000nm), short-wave infrared (1000-2500nm), and thermal infrared (6000-14500nm). Currently, short-wave infrared technology is the most widely used, employed to identify the spectral information of altered minerals containing hydroxyl groups, sulfates, and carbonates. Its advantages of low cost and high efficiency have led to its widespread application in mineral mapping and exploration of porphyry-epidermal hydrothermal deposits, VMS, skarn, and IOCG deposits. By extracting characteristic wavelengths and other parameters of different altered minerals, such as muscovite, chlorite, and tourmaline, short-wave infrared spectroscopy can be used to establish mineral spectral exploration markers for specific deposit types, delineating mineralized hydrothermal centers and key exploration target areas.

[0003] For example, in studies identifying hydrothermal centers in porphyry copper deposits, the distance to the hydrothermal center is usually related to the crystallinity of muscovite, which can be calculated by the ratio of the absorption depth near 2200 nm to the absorption depth near 1900 nm in short-wave infrared spectroscopy. Therefore, it is necessary to remove non-muscovite data from all samples, and the basis for this removal is mineral identification based on infrared spectroscopy.

[0004] There are generally two approaches to mineral infrared spectroscopy identification: one is to examine the characteristic absorption peaks of the mineral, and the other is to examine the fit between the standard spectrum of the mineral and the spectrum of the sample to be identified.

[0005] The first approach is the commonly used method of manual identification. Various minerals typically have representative characteristic absorption peaks. For example, muscovite usually has significant absorption peaks at 1400nm, 1900nm, 2200nm, 2350nm, and 2440nm. If the spectrum to be identified has these characteristics, it can be identified as muscovite.

[0006] The second approach uses methods commonly used in automated identification programs, including the traditional spectral angle method and the TSA algorithm in TSG software. Although the specific implementations of these algorithms differ, their core principle is to find a curve that best matches the spectrum of the mineral to be identified, fitted by one or more standard mineral spectra.

[0007] Currently, most mineral infrared spectroscopy software primarily uses the second approach, with the TSA algorithm being the most mature and representative. For both standard mineral samples and the sample to be identified, the TSA algorithm first takes the logarithm of the original reflectance curve, then uses cubic splines to remove the envelope and eliminate water effects, resulting in what is called the "scaled residual." Next, it uses a matrix approach to solve for one to three standard minerals that can fit the scaled residual of the sample to be identified according to a certain mixing ratio. The normalized sum of squared residuals (SRSS) between the fitted curve and the curve to be identified represents the error of the identification result. Other similar identification methods, although significantly different from TSA, all follow a similar least squares approach. The evaluation of their identification results is based on the sum of squared residuals or its variations.

[0008] Compared to the first approach, the second approach, while mathematically more elegant and rigorous, has significant drawbacks. Firstly, mineral identification using infrared spectroscopy focuses on the goodness of fit of the characteristic absorption peak positions of a mineral; the goodness of fit of non-absorption peak positions does not affect the identification result. Secondly, because non-absorption peak positions also participate in error calculation, the sum of squared residuals of the overall curve cannot represent the goodness of fit. A good fit might have poor absorption peak positions but good non-absorption peak positions, ultimately leading to a very good sum of squared residuals, misleading users into believing it is a very good identification result. This flaw has been demonstrated in the practical application of the TSG software. For example, when identifying the opal standard sample in its built-in tsgreflib example project, the TSG software incorrectly identified it as a mixture of montmorillonite and magnesite, with an SRSS of only 90.855 (according to the software manual, an SRSS greater than 1000 is considered unreliable). Therefore, SRSS cannot effectively indicate the reliability of the result. In other words, the identification result lacks quantifiable reliability.

[0009] Since the second approach is a statistically-based "black box" method, improving its recognition accuracy typically involves training with more samples. However, acquiring large amounts of standard mineral infrared spectral data is a labor-intensive process, making it difficult to improve model accuracy through large datasets in practice. Therefore, this approach easily reaches its bottleneck.

[0010] The first approach typically uses expert systems. While more traditional, this "white-box" nature makes it easier to interpret, fix defects, and has a higher functional ceiling. However, there are currently no stable and reliable software systems implemented using this approach. Summary of the Invention

[0011] The main objective of this invention is to provide a scalable and highly reliable mineral identification method and system based on infrared spectroscopy.

[0012] The technical solution adopted in this invention is: A method for constructing a standard mineral spectral lookup tree is provided, including the following steps: The infrared spectrum is processed to obtain descriptive information of each absorption peak of each standard mineral, including the wavelength, absorption depth and prominence of the absorption peak; and the characteristic absorption peaks of the standard minerals that meet the preset conditions are marked. Create a one-dimensional KD tree structure for a standard mineral spectral lookup tree, denoted as KDT. The keys of this KD tree structure are the wavelengths of the absorption peaks of the standard minerals, and the values ​​are descriptive information about the absorption peaks of the standard minerals; using preset values... x represents the interval for enumerating all wavelengths between the minimum value XMin and the maximum value XMax of the absorption peaks for each standard mineral, including XMin and XMin+. x, XMin+2 x, ..., XMax, and add them to the KDT.

[0013] According to the above technical solution, the condition for judging the characteristic absorption peak is: the minimum prominence of the absorption peak, PMin, is greater than a preset value.

[0014] Following the above technical solution, the preset value The value range of x is [15nm, 25nm].

[0015] Following the above technical solution, the data processing for infrared spectra specifically involves removing the envelope and normalizing the reflectance data in each infrared spectrum.

[0016] Following the above technical solution, the Hull-Quotient algorithm is specifically used to remove the envelope.

[0017] The present invention also provides a standard mineral spectral query tree, which is a query tree constructed by the construction method of the standard mineral spectral query tree described in the above technical solution.

[0018] This invention also provides a mineral identification method based on infrared spectroscopy, which is based on the standard mineral spectral lookup tree described in the above technical solution. This mineral identification method includes the following steps: Obtain the mineral spectrum of the sample to be identified; The mineral spectrum is processed to obtain the wavelength, absorption depth, and prominence of multiple absorption peaks of the sample to be identified; For each absorption peak P of the sample to be identified, the wavelength is used as the key input to search for the preset value in the standard mineral spectrum lookup tree. Multiple standard mineral absorption peaks within the x range are selected, and the standard mineral absorption peak RP closest to the query wavelength is obtained. Based on the selected standard mineral absorption peaks RP, all possible standard minerals corresponding to the sample to be identified are obtained; if all characteristic absorption peaks of a certain standard mineral are corresponding in the sample to be identified, then the standard mineral is added to the list of candidate minerals of the sample to be identified. Identify the principal mineral of the sample to be identified from the list of candidate minerals.

[0019] Following the above technical solution, the steps also include: The absorption peaks in the sample to be identified whose absorption depth and prominence are both greater than the threshold are marked as the main absorption peaks. Calculate the distance D between each absorption peak P of the sample to be identified and the corresponding standard mineral absorption peak RP: If the wavelength of P is between the minimum value XMin and the maximum value XMax of the wavelength of RP, then D is 0; if the wavelength of P is less than the minimum value XMin or greater than the maximum value XMax of the wavelength of RP, then D is the distance between the wavelength of P and the minimum value XMin or the maximum value XMax of the wavelength of RP multiplied by the absorption depth of P. For each candidate mineral in the list, calculate the number of all absorption peaks N, the number of corresponding main absorption peaks NP, and the sum of all absorption peak distances DS for the sample to be identified. The candidate mineral with the largest corresponding NP value is the main mineral of the sample to be identified. When the corresponding NP values ​​are equal, the candidate mineral with the largest corresponding N value is the main mineral of the sample to be identified. When the corresponding N values ​​are equal, the candidate mineral with the smallest corresponding DS is the main mineral of the sample to be identified. The DS of the main mineral is the error of the main mineral result.

[0020] Following the above technical solution, if there is a main mineral in the sample to be identified without a corresponding absorption peak, it indicates the presence of a secondary mineral. In this case, among the remaining candidate minerals, the mineral that corresponds to the absorption peak and has the smallest DS is identified as the secondary mineral. The DS of the secondary mineral is the result error of the secondary mineral.

[0021] The present invention also provides a mineral identification system based on infrared spectroscopy, comprising: A standard mineral spectral lookup tree is a lookup tree constructed using the method described above. The spectral acquisition module is used to acquire the mineral spectrum of the sample to be identified; The spectral processing module is used to process the mineral spectrum to obtain the wavelength, absorption depth, and prominence of multiple absorption peaks of the sample to be identified. The absorption peak screening module is used to search for each absorption peak P of the sample to be identified, using the wavelength as the key input, and retrieve the peaks within the preset range from the standard mineral spectrum lookup tree. Multiple standard mineral absorption peaks within the x range are selected, and the standard mineral absorption peak RP closest to the query wavelength is obtained. The candidate mineral list generation module is used to obtain all the standard minerals that may correspond to the sample to be identified based on the absorption peak RP of the screened standard minerals; if all the characteristic absorption peaks of a certain standard mineral are corresponding in the sample to be identified, then the standard mineral is added to the list of candidate minerals of the sample to be identified. The identification module is used to identify the main mineral of the sample to be identified from the list of candidate minerals.

[0022] Following the above technical solution, the identification module is specifically used for: marking absorption peaks in the sample to be identified whose absorption depth and prominence are both greater than a threshold as main absorption peaks; calculating the distance D between each absorption peak P of the sample to be identified and the corresponding standard mineral absorption peak RP: if the wavelength of P is between the minimum value XMin and the maximum value XMax of the wavelength of RP, then D is 0; if the wavelength of P is less than the minimum value XMin or greater than the maximum value XMax of the wavelength of RP, then D is the distance between the wavelength of P and the minimum value XMin or the maximum value XMax of the wavelength of RP multiplied by the absorption depth of P; for each candidate mineral in the list, calculating the number N of all absorption peaks corresponding to the sample to be identified, the number NP of the corresponding main absorption peaks, and the sum of all absorption peak distances DS, where the candidate mineral with the largest corresponding NP value is the main mineral of the sample to be identified; when the corresponding NP values ​​are equal, the candidate mineral with the largest corresponding N value is the main mineral of the sample to be identified; when the corresponding N values ​​are equal, the candidate mineral with the smallest corresponding DS is the main mineral of the sample to be identified; where the DS of the main mineral is the error of the main mineral result.

[0023] The present invention also provides a computer storage medium storing a computer program executable by a processor, the computer program performing the mineral identification method based on infrared spectroscopy described in the above technical solution.

[0024] The beneficial effects of this invention are as follows: By constructing a standard mineral spectral lookup tree, this invention correlates the wavelength, absorption intensity, and salience of each absorption peak of a standard mineral. Furthermore, by enumerating wavelengths at equal intervals within the lookup tree, it indirectly achieves interval-based lookup of absorption peaks. Based on this standard mineral spectral lookup tree, the absorption peaks of the standard minerals closest to each absorption peak of the sample to be identified are found through absorption peak wavelength lookup. This allows for the rapid generation of a list of candidate minerals, from which the principal mineral of the sample to be identified can be determined, thereby improving identification efficiency. Moreover, the number of identifiable minerals can be expanded by extending the standard mineral spectral lookup tree. The scalability of the identification method is greatly improved by creating spectral lookup trees for various standard minerals.

[0025] Furthermore, by using the method of matching absorption peaks to qualitatively identify minerals, the error of the identification result can be output, which is the sum of the distances between the identification result and the absorption peaks of the standard mineral spectrum, allowing users to quantitatively evaluate the reliability of the identification result.

[0026] Furthermore, if the sample to be identified contains a main mineral without a corresponding absorption peak, it indicates the presence of a secondary mineral, and the corresponding secondary mineral can be further searched among the remaining candidate minerals.

[0027] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a flowchart of the method for constructing a quasi-mineral spectral query tree according to an embodiment of the present invention; Figure 2 This is a flowchart of a mineral identification method based on infrared spectroscopy according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating how the main mineral of a sample to be identified is derived from a list of candidate minerals according to an embodiment of the present invention. Figure 4 This is a schematic diagram of the structure of a mineral identification system based on infrared spectroscopy according to an embodiment of the present invention; Figure 5 This is the infrared spectrum of the sample to be identified in an embodiment of the present invention; Figure 6 This is the short-wave infrared spectrum of a standard sample of muscovite in an embodiment of the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0031] It should be noted that the illustrations provided in the embodiments of the present invention are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0032] In this invention, it should also be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. Furthermore, the terms "first" and "second" are used only for descriptive and distinguishing purposes and should not be construed as indicating or implying relative importance.

[0033] Furthermore, it should be noted that the features of the various embodiments of the present invention can be combined or integrated in whole or in part, and as those skilled in the art will understand, they can interact and operate in different ways. Each embodiment can be implemented independently of each other or in association with one another.

[0034] like Figure 1 As shown, the method for constructing a standard mineral spectral lookup tree according to an embodiment of the present invention is characterized by comprising the following steps: S101. Obtain the infrared spectra of multiple samples used as standard minerals; S102. Perform data processing on the infrared spectrum to obtain descriptive information of each absorption peak of each standard mineral, including the wavelength, absorption depth and prominence of the absorption peak; and mark the characteristic absorption peaks of the standard minerals that meet the preset conditions. S103. Create a one-dimensional KD tree structure for a standard mineral spectral lookup tree, denoted as KDT. The keys of this KD tree structure are the wavelengths of the absorption peaks of the standard minerals, and the values ​​are descriptive information about the absorption peaks of the standard minerals; using preset values... x represents the interval for enumerating all wavelengths between the minimum value XMin and the maximum value XMax of the absorption peaks for each standard mineral, including XMin and XMin+. x, XMin+2 x, ..., XMax, and add them to the KDT.

[0035] It can create standard mineral spectral query trees for multiple minerals. Among them, preset values... The value range of x is [15nm, 25nm].

[0036] Furthermore, the data processing for the infrared spectrum specifically involves removing the envelope and normalizing the reflectance data in each infrared spectrum.

[0037] Specifically, the standard mineral spectral lookup tree for each mineral contains descriptive information for each absorption peak after the mineral's spectral reflectance (e.g., after removing the envelope and normalizing using the Hull-Quotient algorithm). Each absorption peak description should include the following: the minimum and maximum possible wavelengths of the absorption peak (XMin and XMax); the minimum and maximum absorption depths (YMin and YMax); and the minimum and maximum saliences (PMin and PMax).

[0038] The system can determine whether an absorption peak is a characteristic absorption peak based on preset conditions. That is, the absorption peak must appear in the mineral, and PMin must be greater than a preset value (e.g., 0.02). This information can be obtained through manual analysis of standard mineral infrared spectral data or automatically based on preset conditions.

[0039] The ability to identify minerals can be extended by expanding the standard mineral spectral model library.

[0040] In step S103, a standard mineral peak index is also established while constructing the query tree. For each absorption peak description in the spectral query tree for each mineral, all wavelengths between XMin and XMax can be enumerated in 20nm intervals, i.e., XMin, XMin+20nm, XMin+40nm, ... up to XMax. Each wavelength is used as the key, and the standard mineral absorption peak as the value, and added to the KDT. Ultimately, by inputting a wavelength into the KDT, the nearest standard mineral absorption peak can be retrieved.

[0041] The reason for using The x-interval enumeration of values ​​from XMin to XMax is used as the key because subsequent steps will use them as keys. x represents the nearest standard mineral absorption peak to the maximum distance query. However, KD trees cannot directly perform interval queries, therefore, by using... By enumerating values ​​between XMin and XMax at intervals of x, a range query is indirectly achieved. This way, if a wavelength to be queried is located within the range of XMin of a certain absorption peak... x to XMax+ The absorption peak can be found within the range of x.

[0042] like Figure 2 As shown, the mineral identification method based on infrared spectroscopy in this embodiment of the invention is based on the standard mineral spectral query tree constructed in the above embodiment. The mineral identification method includes the following steps: S201. Obtain the mineral spectrum of the sample to be identified; S202. Process the mineral spectrum to obtain the wavelength, absorption depth and prominence of multiple absorption peaks of the sample to be identified; S203. For each absorption peak P of the sample to be identified, use the wavelength as the key input to search for the preset value in the standard mineral spectrum lookup tree. Multiple standard mineral absorption peaks within the x range are selected, and the standard mineral absorption peak RP closest to the query wavelength is obtained. S204. Based on the selected standard mineral absorption peaks RP, obtain all the standard minerals that may correspond to the sample to be identified; if all the characteristic absorption peaks of a certain standard mineral are corresponding in the sample to be identified, then add the standard mineral to the list of candidate minerals of the sample to be identified. S205. Identify the main mineral of the sample to be identified from the list of candidate minerals.

[0043] Among them, such as Figure 3 As shown, step S205 specifically includes: S2051. Mark the absorption peaks in the sample to be identified that have both absorption depth and prominence greater than the threshold as the main absorption peaks. S2052. Calculate the distance D between each absorption peak P of the sample to be identified and the corresponding standard mineral absorption peak RP: If the wavelength of P is between the minimum value XMin and the maximum value XMax of the wavelength of RP, then D is 0; if the wavelength of P is less than the minimum value XMin or greater than the maximum value XMax of the wavelength of RP, then D is the distance between the wavelength of P and the minimum value XMin or the maximum value XMax of the wavelength of RP multiplied by the absorption depth of P. S2053. For each candidate mineral in the list, calculate the total number of absorption peaks N, the total number of main absorption peaks NP, and the sum of all absorption peak distances DS for the sample to be identified. The candidate mineral with the largest NP value is the main mineral of the sample to be identified. When the corresponding NP values ​​are equal, the candidate mineral with the largest corresponding N value is the main mineral of the sample to be identified. When the corresponding N values ​​are equal, the candidate mineral with the smallest corresponding DS is the main mineral of the sample to be identified. The DS of the main mineral is the error of the main mineral result. For example, for the mineral spectrum x of the sample to be identified, a list of candidate minerals including four candidate minerals (abcd) is obtained from multiple KDTs. The calculation results are as follows: N=5 absorption peaks corresponding to x and a, with NP=4 main absorption peaks; N=4 absorption peaks corresponding to x and b, with NP=4 main absorption peaks; and so on. Then, the main mineral of the sample to be identified is determined based on the calculation results.

[0044] Furthermore, if there is a main mineral in the sample to be identified that does not have a corresponding absorption peak, then among the remaining candidate minerals, the mineral that corresponds to the absorption peak and has the smallest DS is selected as the secondary mineral, and the DS of the secondary mineral is the error of the secondary mineral result.

[0045] The above-described method uses absorption peak matching for qualitative mineral identification. Users can create standard mineral spectral query trees for various minerals, and multiple query trees can be used to construct a standard mineral spectral model library, thus infinitely expanding the standard mineral spectral model library that the algorithm can identify. This method also outputs the error of the identification result, i.e., the sum of the distances to the standard mineral spectral absorption peaks, allowing users to quantitatively evaluate the reliability of the identification results.

[0046] In a specific embodiment of the present invention, the mineral identification method based on infrared spectroscopy consists of three steps: establishing a standard mineral spectral model library (i.e., composed of a complete standard mineral spectral query tree), establishing a standard mineral peak position index, and mineral identification. Specifically, the method includes the following steps: 1) For the mineral spectrum of the sample to be identified, the spectral reflectance data is first normalized after envelope removal using the Hull-Quotient algorithm. Then, signal processing methods are used to find the wavelength, depth, and saturation of the absorption peaks. Absorption peaks with a depth greater than 0.1 and a saturation greater than 0.1 are marked as main absorption peaks.

[0047] 2) For each sample absorption peak, denoted as P, use the wavelength as input to search for the nearest standard mineral absorption peak in KDT, with a maximum search distance of 20 nm. The search results will be the nearest standard mineral absorption peak to P within a 20 nm range to the left and right. Perform the following operations on the search results to filter out unwanted results: a. Because in the previous steps when constructing the KDT, the wavelength range of an absorption peak was enumerated from small to large at 20nm intervals and then inserted into the KDT, the KDT query results may show the same standard mineral absorption peak appearing multiple times, and these duplicate results must be eliminated.

[0048] b. The query results may contain two different absorption peaks of the same standard mineral corresponding to the same wavelength. In this case, only the absorption peak of the standard mineral closest to the query wavelength should be retained.

[0049] After screening, the nearest standard mineral corresponding to P within ±20 nm is obtained, and each mineral corresponds to only one absorption peak. For each corresponding standard mineral absorption peak, denoted as RP, the distance D between P and RP is calculated as follows: a. If the wavelength of P is between XMin and XMax of RP, then D is 0. b. If the wavelength of P is less than XMin or greater than XMax, then D is the distance between the wavelength of P and XMin or XMax multiplied by the depth of P.

[0050] 3) Integrate the standard minerals corresponding to each peak obtained in the previous step to obtain all possible standard minerals corresponding to the sample. If all characteristic absorption peaks of a certain standard mineral have corresponding values ​​in the sample, then add the mineral to the candidate mineral list.

[0051] 4) For each candidate mineral, calculate the number of all absorption peaks N, the number of main absorption peaks NP, the distance between all absorption peaks and DS (i.e., sum of all D) of its corresponding sample.

[0052] 5) The mineral with the highest NP value is the main mineral; when NP values ​​are equal, the mineral with the highest N value is the main mineral; when N values ​​are equal, the mineral with the lowest DS value is the main mineral. The DS value of the main mineral is the error of the main mineral result.

[0053] 6) If a main mineral in the sample has no corresponding absorption peak, it indicates the presence of a secondary mineral. Among the remaining candidate minerals, find the one with the smallest DS (Dissipation Limit) that corresponds to the absorption peak; this is the secondary mineral. The DS of the secondary mineral represents the error in the secondary mineral result.

[0054] 7) The smaller the error of the result, the more reliable the result.

[0055] like Figure 4 As shown, the mineral identification system based on infrared spectroscopy in this embodiment of the invention includes: A standard mineral spectral query tree is a query tree constructed using the standard mineral spectral query tree construction method described in the above embodiments. Multiple standard mineral spectral query trees can form a standard mineral spectral model library. The spectral acquisition module is used to acquire the mineral spectrum of the sample to be identified; The spectral processing module is used to process the mineral spectrum to obtain the wavelength, absorption depth, and prominence of multiple absorption peaks of the sample to be identified. The absorption peak screening module is used to search for each absorption peak P of the sample to be identified, using the wavelength as the key input, and retrieve the peaks within the preset range from the standard mineral spectrum lookup tree. Multiple standard mineral absorption peaks within the x range are selected, and the standard mineral absorption peak RP closest to the query wavelength is obtained. The candidate mineral list generation module is used to obtain all the standard minerals that may correspond to the sample to be identified based on the absorption peak RP of the screened standard minerals; if all the characteristic absorption peaks of a certain standard mineral are corresponding in the sample to be identified, then the standard mineral is added to the list of candidate minerals of the sample to be identified. The identification module is used to identify the main mineral of the sample to be identified from the list of candidate minerals.

[0056] Specifically, the identification module is used to: mark absorption peaks in the sample to be identified whose absorption depth and prominence are both greater than a threshold as main absorption peaks; calculate the distance D between each absorption peak P of the sample to be identified and the corresponding standard mineral absorption peak RP: if the wavelength of P is between the minimum value XMin and the maximum value XMax of the wavelength of RP, then D is 0; if the wavelength of P is less than the minimum value XMin or greater than the maximum value XMax of the wavelength of RP, then D is the distance between the wavelength of P and the minimum value XMin or the maximum value XMax of the wavelength of RP multiplied by the absorption depth of P; for each candidate mineral in the list, calculate the number of all absorption peaks N, the number of corresponding main absorption peaks NP, and the sum of all absorption peak distances DS for the sample to be identified, where the candidate mineral with the largest corresponding NP value is the main mineral of the sample to be identified; when the corresponding NP values ​​are equal, the candidate mineral with the largest corresponding N value is the main mineral of the sample to be identified; when the corresponding N values ​​are equal, the candidate mineral with the smallest corresponding DS is the main mineral of the sample to be identified; where the DS of the main mineral is the error of the main mineral result.

[0057] In one embodiment of the present invention, the waveform of the infrared spectral sample to be identified is as follows: Figure 5 As shown, the absorption peaks of this sample are as follows: 1. Wavelength 1412, depth 0.47, salience 0.47, main absorption peak.

[0058] 2. Wavelength 1912, depth 1.00, salience 1.00, main absorption peak.

[0059] 3. Wavelength 2206, depth 0.59, salience 0.59, main absorption peak.

[0060] 4. Wavelength 2248, depth 0.22, salience 0.01.

[0061] 5. Wavelength 2345 nm, depth 0.40, salience 0.40, main absorption peak.

[0062] 6. Wavelength 2448 nm, depth 0.12, salience 0.12, main absorption peak.

[0063] First, establish a standard mineral spectral model library. For example... Figure 6 The image shows the short-wave infrared spectrum (Hull-Quotient, normalized) of a standard sample of muscovite: Through similar Figure 5 Analysis of shortwave infrared spectral data of multiple muscovite, biotite, montmorillonite, and dolomite yielded the following results.

[0064] Muscovite has 6 absorption peaks: 1. Wavelength range 1408-1416, depth range 0.63-0.87, salience range 0.63-0.87, characteristic peak.

[0065] 2. Wavelength range 1904-1920 nm, depth range 0.08-0.48 nm, salience range 0.08-0.48 nm, non-characteristic peak.

[0066] 3. Wavelength range 2112-2124, depth range 0.05-0.17, salience range 0.02-0.05, non-characteristic peak.

[0067] 4. Wavelength range 2200-2208, depth range 1.00-1.00, salience range 1.00-1.00, characteristic peak.

[0068] 5. Wavelength range 2344-2353, depth range 0.30-0.50, salience range 0.15-0.50, characteristic peak.

[0069] 6. Wavelength range 2428-2440 nm, depth range 0.32-0.43 nm, salience range 0.04-0.38 nm, characteristic peak.

[0070] Biotite has four absorption peaks: 1. Wavelength range 1396-1416, depth range 0.34-1.00, salience range 0.03-1.00, non-characteristic peak.

[0071] 2. Wavelength range 1912-1924, depth range 0.11-0.40, salience range 0.09-0.40, characteristic peak.

[0072] 3. Wavelength range 2248-2256, depth range 0.33-1.00, salience range 0.33-1.00, characteristic peak.

[0073] 4. Wavelength range 2336-2364, depth range 0.30-1.00, salience range 0.29-1.00, characteristic peak.

[0074] Montmorillonite has four absorption peaks: 1. Wavelength range 1412-1420 nm, depth range 0.48-0.99 nm, salience range 0.48-0.99 nm, characteristic peak.

[0075] 2. Wavelength range 1904-1912, depth range 1.00-1.00, salience range 1.00-1.00, characteristic peak.

[0076] 3. Wavelength range 2200-2216, depth range 0.28-0.62, salience range 0.19-0.55, characteristic peak.

[0077] 4. Wavelength range 2380-2452, depth range 0.07-0.27, salience range 0.07-0.27, characteristic peak.

[0078] Dolomite has 6 absorption peaks: 1. Wavelength range 1376-1428, depth range 0.11-0.42, salience range 0.11-0.42, non-characteristic peak.

[0079] 2. Wavelength range 1856-1868, depth range 0.09-0.23, salience range 0.04-0.19, non-characteristic peak.

[0080] 3. Wavelength range 1924-1984, depth range 0.14-1.00, salience range 0.14-1.00, characteristic peak.

[0081] 4. Wavelength range 2136-2148, depth range 0.05-0.14, salience range 0.03-0.13, characteristic peak.

[0082] 5. Wavelength range 2316-2328, depth range 0.98-1.00, salience range 0.98-1.00, characteristic peak.

[0083] 6. Wavelength range 2448-2464, depth range 0.19-0.52, salience range 0.19-0.53, characteristic peak.

[0084] The above four minerals will be used to construct a standard mineral spectral model library. This library should allow users to query the wavelength range, depth range, salience range, and whether a peak is a characteristic peak by mineral name and absorption peak number.

[0085] A KD tree class suitable for querying absorption peaks of nearby standard minerals can be implemented by wrapping the KDTree class in the Python module scipy.spatial. This class contains two member variables: a tree of type KDTree and mineralPeaks of type List[Tuple[str,int]]. The first element of each tuple in mineralPeaks is the name of the standard mineral, and the second element is the absorption peak number. Using the standard mineral name and absorption peak number, model information for a specific absorption peak of a given standard mineral can be retrieved from the standard mineral spectral model library.

[0086] During class initialization, each absorption peak of each mineral is inserted into a member variable. Taking the third absorption peak of dolomite as an example, wavelengths are first enumerated at 20nm intervals: 1924, 1944, 1964, and 1984. These four wavelengths are inserted into the `tree`, and the mineral name and absorption peak number are simultaneously inserted into `mineralPeaks`. The insertion order in the `tree` and `mineralPeaks` are kept consistent and of equal length. This is because the `tree` only returns the element number during querying, and the absorption peak model information is retrieved from `mineralPeaks` using this element number.

[0087] This class provides a Query interface that takes an input of wavelength as an integer and outputs a Dict[str, Tuple[int, int]], where the key is the mineral name, the first element of the value is the index, and the second element is the distance. Within this interface, the tree's query interface is called to retrieve the input wavelength, specifying distance_upper_bound as 20 in the parameter. It's important to note that the distance returned by this interface is not the distance returned by tree.query, but rather a distance calculated by comparing the input wavelength with the wavelength range of the absorption peak—that is, if the input wavelength is within the wavelength range of the absorption peak, the distance is 0; otherwise, it is the distance between the input wavelength and the minimum or maximum value of that range multiplied by the absorption peak depth of the sample.

[0088] The results of the search for the six absorption peaks of the sample are summarized and presented in Table 1 below: Table 1 Absorption peaks of each sample

[0089] These data show that characteristic peak 4 of dolomite has no corresponding value in this sample, while the other three minerals all have corresponding values ​​in this sample. Therefore, the candidate minerals are muscovite, biotite, and montmorillonite.

[0090] Muscovite has a N value (number of all absorption peaks) of 5 and an NP value (number of main absorption peaks) of 5; biotite has a N value of 4 and an NP value of 3; montmorillonite has a N value of 4 and an NP value of 4. Muscovite has the highest NP value, therefore it is the main mineral. However, absorption peak number 4 in the sample has no corresponding peak in muscovite, but only in biotite, therefore biotite is the accessory mineral.

[0091] The above calculations are easily implemented using the Python language, so the implementation methods will not be elaborated here.

[0092] This invention also provides a computer-readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, server, app store, etc., which stores a computer program, and the program performs corresponding functions when executed by a processor. In this embodiment, the computer-readable storage medium, when executed by a processor, implements the mineral identification method based on infrared spectroscopy of the method embodiment.

[0093] It should be noted that, depending on the implementation needs, the various steps / components described in this application can be broken down into more steps / components, or two or more steps / components or parts of the operation of steps / components can be combined into new steps / components to achieve the purpose of this invention.

[0094] The order of the steps in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0095] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A method of constructing a standard mineral spectral query tree, characterized by, The method comprises the following steps: Obtaining infrared spectra of a plurality of standard minerals; Processing the infrared spectra to obtain description information of each absorption peak of each standard mineral, including wavelength, absorption depth and prominence, and marking characteristic absorption peaks of standard minerals meeting preset conditions; A standard mineral spectrum query tree of one-dimensional K-D tree structure is created, denoted as KDT, a key of the K-D tree structure is a wavelength of an absorption peak of a standard mineral, and a value of the K-D tree structure is description information of the absorption peak of the standard mineral; a preset value x is an interval enumeration of all wavelengths between a minimum value XMin and a maximum value XMax of each absorption peak wavelength of a standard mineral, including XMin, XMin+ x, XMin+2 x, …, XMax, and added to the KDT. 2.The method of claim 1, wherein, The judgment condition of the characteristic absorption peak is that the minimum value PMin of the prominence of the absorption peak is greater than a preset value. 3.The method of claim 1, wherein, Pre-set value x is in the range [15 nm, 25 nm]. 4.The method of claim 1, wherein, The data processing of the infrared spectra specifically comprises: performing envelope removal and normalization processing on the reflectance data in each infrared spectrum.

5. A standard mineral spectral query tree, characterized by, The mineral identification method is based on the standard mineral spectrum query tree of claim 5, and the method comprises the following steps:

6. A method of mineral identification based on infrared spectroscopy, characterized in that, Obtaining a mineral spectrum of a sample to be identified; Processing the mineral spectrum to obtain the wavelength, absorption depth and prominence of a plurality of absorption peaks of the sample to be identified; According to the screened standard mineral absorption peaks RP, all standard minerals possibly corresponding to the sample to be identified are obtained; if all characteristic absorption peaks of a standard mineral have a corresponding in the sample to be identified, the standard mineral is added to a candidate mineral list of the sample to be identified; For each absorption peak P of the sample to be identified, using the wavelength as the key input, a plurality of standard mineral absorption peaks in the preset value x range are obtained in the standard mineral spectrum query tree, and the standard mineral absorption peak RP closest to the query wavelength is screened. Identifying a main mineral of the sample to be identified from the candidate mineral list. The process of identifying the main mineral of the sample to be identified from the candidate mineral list specifically comprises:

7. The infrared spectroscopy-based mineral identification method of claim 6, wherein, Marking absorption peaks with absorption depth and prominence greater than a threshold value in the sample to be identified as main absorption peaks; Calculating the distance D between each absorption peak P of the sample to be identified and the corresponding standard mineral absorption peak RP: if the wavelength of P is between the minimum value XMin and the maximum value XMax of the wavelength of RP, D is 0; if the wavelength of P is less than the minimum value XMin or greater than the maximum value XMax of the wavelength of RP, D is the distance between the wavelength of P and the minimum value XMin or the maximum value XMax of the wavelength of RP multiplied by the absorption depth of P; For each candidate mineral in the list, the number N of all absorption peaks corresponding to the sample to be identified, the number NP of corresponding main absorption peaks, and the sum DS of all absorption peak distances are calculated, wherein the candidate mineral with the maximum NP value is the main mineral of the sample to be identified; when the NP values are equal, the candidate mineral with the maximum N value is the main mineral of the sample to be identified; when the N values are equal, the candidate mineral with the minimum DS is the main mineral of the sample to be identified; wherein the DS of the main mineral is the result error of the main mineral. If there is no main mineral corresponding absorption peak in the sample to be identified, among the remaining candidate minerals, find a mineral that can correspond to the absorption peak and has the minimum DS, which is a subordinate mineral, and the DS of the subordinate mineral is the result error of the subordinate mineral.

8. The infrared spectroscopy-based mineral identification method of claim 7, wherein, The method comprises the following steps:

9. An infrared spectroscopy based mineral identification system, characterized by, The standard mineral spectrum query tree is constructed by the construction method of the standard mineral spectrum query tree according to any one of claims 1-4; A spectrum acquisition module is configured to obtain a mineral spectrum of a sample to be identified; A spectrum processing module is configured to process the mineral spectrum to obtain the wavelength, absorption depth and prominence of a plurality of absorption peaks of the sample to be identified; ​ The absorption peak screening module is configured to, for each absorption peak P of the sample to be identified, use the wavelength as a key input to query a standard mineral spectrum query tree to obtain a plurality of standard mineral absorption peaks within a preset value x of the query wavelength, and screen to obtain a standard mineral absorption peak RP closest to the query wavelength. The candidate mineral list generation module is configured to obtain all standard minerals possibly corresponding to the sample to be identified according to the screened standard mineral absorption peak RP; If all characteristic absorption peaks of a certain standard mineral have a corresponding in the sample to be identified, the standard mineral is added to the candidate mineral list of the sample to be identified; The identification module is configured to identify the main mineral of the sample to be identified from the candidate mineral list.

10. The infrared spectroscopy based mineral identification system of claim 9, wherein, The identification module is specifically configured to: mark an absorption peak in the sample to be identified as a main absorption peak if the absorption depth and the prominence of the absorption peak are both greater than a threshold value; calculate the distance D between each absorption peak P of the sample to be identified and the corresponding standard mineral absorption peak RP: if the wavelength of P is between the minimum value XMin and the maximum value XMax of the wavelength of RP, D is 0; if the wavelength of P is less than the minimum value XMin or greater than the maximum value XMax of the wavelength of RP, D is the distance between the wavelength of P and the minimum value XMin or the maximum value XMax of the wavelength of RP multiplied by the absorption depth of P; for each candidate mineral in the list, calculate the number N of all absorption peaks corresponding to the sample to be identified, the number NP of main absorption peaks corresponding to the sample to be identified, and the sum DS of the distances of all absorption peaks, wherein the candidate mineral corresponding to the maximum NP value is the main mineral of the sample to be identified; when the NP values are equal, the candidate mineral corresponding to the maximum N value is the main mineral of the sample to be identified; when the N values are equal, the candidate mineral corresponding to the minimum DS is the main mineral of the sample to be identified; The DS of the main mineral is the main mineral result error.

Images