Solute region spectrum extraction method, system and storage medium

Abstract

The present application discloses a solute region spectrum extraction method, system and storage medium, belonging to the field of liquid detection technology. The method is used for liquid-solid conversion LIBS liquid detection scenarios. The method comprehensively considers the information of different elements and different spectral lines in the same droplet, and performs a morphological recognition before and after the weighted fusion of multiple element information, so as to accurately extract the solute region spectrum in the coverage acquisition raw data for quantitative analysis, and eliminate the interference of invalid spectra without solute on LIBS analysis. Compared with the traditional rectangular dot matrix coverage scanning method, the spectral intensity after correction of the present invention has higher stability, and can effectively improve the sensitivity of quantitative analysis, which is conducive to the further development and application of LIBS liquid detection technology.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of liquid detection technology, and in particular to a solute region spectrum extraction method, system and storage medium. Background Art

[0002] Laser-Induced Breakdown Spectroscopy (LIBS) technology uses a focusing lens to focus a high-energy pulsed laser on a tiny area on the sample surface, forming a high-temperature, high-pressure plasma. By collecting and analyzing the plasma radiation light, the wavelength and relative spectral intensity of the emission characteristic spectrum are obtained, and then the element type and element content of the sample to be tested are deduced. This technology has the advantages of simultaneous multi-element detection, simple or no sample preparation, real-time rapid analysis and in-situ detection, and is widely used in many fields such as metal analysis, environmental testing, atomic energy and nuclear engineering, life sciences, and cultural relics protection.

[0003] LIBS technology has obvious advantages in solid detection and has also been widely studied and applied in liquid detection. In the existing LIBS liquid detection methods, although direct water body detection does not require any sample preparation, it is easy to cause problems such as liquid splashing and plasma quenching. Liquid flow detection converts static liquid into flowing liquid, effectively reducing problems such as liquid surface fluctuation and droplet splashing. However, liquid flow detection requires additional auxiliary equipment for liquid flow generation, and the laser object is still liquid. In order to overcome these problems, researchers have proposed liquid-solid conversion auxiliary detection, which completely avoids problems such as liquid splashing and plasma quenching by converting liquid into solid solute deposits. The substrate liquid-solid conversion method drops a small amount of liquid sample onto a non-absorbent substrate, forms a solute deposition layer by evaporation and drying, and then performs LIBS analysis. Among the LIBS liquid-solid conversion detection methods, compared with the three methods of freezing, extraction and electrodeposition, the sample preparation of the substrate liquid-solid conversion method is simpler and more efficient, does not require a complex system, and uses less reagents, providing a low-cost and efficient sample processing method for liquid detection. The substrate liquid-solid conversion method has great potential in the field of liquid detection.

[0004] In the LIBS liquid detection of substrate liquid-solid conversion, the solute deposits after liquid-solid conversion are usually distributed in a nearly circular or elliptical shape. In order to obtain the LIBS spectral data of the entire solute deposition layer, the rectangular dot matrix scanning method is currently commonly used to collect the deposition layer in full coverage. However, the use of rectangular dot matrix to scan the circular or elliptical solute deposition layer in full coverage will inevitably introduce invalid LIBS spectra without solutes, resulting in reduced spectral stability and detection sensitivity, affecting the LIBS liquid-solid conversion quantitative detection performance. Summary of the invention

[0005] The present application provides a solute region spectrum extraction method, system and storage medium to solve the problem in the related art that a rectangular dot matrix is ​​used to fully cover a circular or elliptical solute deposition layer, thereby introducing an invalid LIBS spectrum without solute, thereby reducing spectral stability and detection sensitivity. The technical solution is as follows:

[0006] On the one hand, a solute region spectrum extraction method is provided, the method is used in a liquid-solid conversion LIBS liquid detection scenario, the method comprising:

[0007] Obtaining a substrate sample, wherein the substrate sample comprises a solute deposition layer of an element to be detected, wherein the solute deposition layer of the element to be detected is formed by dripping liquid;

[0008] Performing laser scanning on the solute deposition layer of the element to be measured and acquiring corresponding LIBS raw spectrum data;

[0009] Determine available spectral lines of each element to be measured according to the LIBS original spectral data, wherein the available spectral lines meet the signal-to-noise ratio requirement;

[0010] Using the available spectral lines of the respective elements to be measured, an element distribution image of the respective elements to be measured in the LIBS raw spectrum data is formed;

[0011] Perform the first morphological recognition on all element distribution images to obtain the masks corresponding to different spectral lines of each element to be tested;

[0012] The masks corresponding to different spectral lines of each element to be measured are weighted and fused to obtain an accurate droplet distribution image;

[0013] Performing a second morphological recognition on the accurate droplet distribution image to obtain a mask representing the distribution of the solute deposition layer of the element to be measured;

[0014] The solute region spectrum of the element to be measured is extracted from the LIBS original spectrum data through the mask of the solute deposition layer distribution of the element to be measured.

[0015] Optionally, obtaining a substrate sample includes:

[0016] Grinding the substrate to obtain a polished substrate with a uniform surface roughness;

[0017] basically cleaning after said grinding;

[0018] Adding droplets of a solution containing the element to be tested onto the surface of the polished substrate by a pipette;

[0019] Then the polished substrate is placed flat on a constant temperature heating table for heating;

[0020] After the heating is completed, a substrate sample containing a solute deposition layer of the element to be measured is obtained.

[0021] Optionally, the laser scanning of the solute deposition layer of the element to be measured and obtaining corresponding LIBS raw spectrum data includes:

[0022] Optimize laser energy and delay;

[0023] The solute deposition layer of the element to be measured at the substrate sample is laser scanned using a covering acquisition method to obtain LIBS original spectrum data, wherein the size of the laser sampling point array used is M×N, and it is necessary to ensure that the laser sampling point array covers the droplet deposition layer.

[0024] Optionally, determining the available spectral lines of each element to be measured according to the LIBS original spectral data, wherein the available spectral lines meet the signal-to-noise ratio requirement, includes:

[0025] Select the spectral line whose relative intensity of the target element in the NIST spectral line library is greater than the spectral line intensity threshold K;

[0026] The average of the spectral line intensity values ​​corresponding to the laser sampling point array is taken to obtain an average original spectrum;

[0027] The persistent lines are screened according to the average original spectrum to obtain the available spectral lines, wherein the screening condition is the spectral lines whose signal-to-noise ratio is greater than a signal-to-noise ratio threshold S.

[0028] Optionally, the content of the element distribution imaging includes: element distribution images of different elements and element distribution images corresponding to different spectral lines of the same element, and the resolution of the element distribution image is the size of the laser sampling point array.

[0029] Optionally, the first morphological identification is operated by acquiring the relative position of the droplet in the laser sampling M×N dot matrix and then extracting the spectrum of the solute region of the element to be measured from the LIBS original spectrum data;

[0030] The data processing of the first morphological recognition includes image binarization, image connected region acquisition, maximum connected region search, closing operation and opening operation.

[0031] Optionally, weighted fusion of masks corresponding to different spectral lines of each element to be measured to obtain an accurate droplet distribution image includes:

[0032] The average spectral intensity of the LIBS original spectral data corresponding to the selected spectral line of each element to be measured is used as the weight of the corresponding mask;

[0033] Perform weighted fusion on the masks corresponding to different spectral lines of each element to be measured;

[0034] An accurate droplet distribution image is obtained by integrating the profile information provided by multiple spectral lines of each element to be measured.

[0035] Optionally, the data processing of the second morphological recognition includes image binarization, image connected region acquisition, small area deletion, closing operation, maximum connected region search and opening operation.

[0036] On the other hand, a solute region spectrum extraction system is provided, the system is used for liquid-solid conversion LIBS liquid detection scenario, the system comprising:

[0037] A sample acquisition module, used for acquiring a substrate sample, wherein the substrate sample comprises a solute deposition layer of an element to be detected, wherein the solute deposition layer of the element to be detected is formed by dripping liquid;

[0038] A spectrum acquisition module, used for laser scanning the solute deposition layer of the element to be measured and obtaining the corresponding LIBS original spectrum data;

[0039] A spectral line selection module, used to determine the available spectral lines of each element to be measured according to the LIBS original spectral data, wherein the available spectral lines meet the signal-to-noise ratio requirement;

[0040] An element distribution module, used to form an element distribution image of each element to be measured in the LIBS original spectrum data using the available spectral lines of each element to be measured;

[0041] The primary recognition module is used to perform the first morphological recognition on all element distribution images to obtain the masks corresponding to different spectral lines of each element to be tested;

[0042] The information fusion module is used to perform weighted fusion of the masks corresponding to different spectral lines of each element to be measured to obtain an accurate droplet distribution image;

[0043] A secondary recognition module, used for performing a second morphological recognition on the accurate droplet distribution image to obtain a mask representing the distribution of the solute deposition layer of the element to be measured;

[0044] The solute region module is used to extract the solute region spectrum of the element to be measured from the LIBS original spectrum data through the mask of the solute deposition layer distribution of the element to be measured.

[0045] On the other hand, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer-readable storage medium, and the computer program is loaded and executed by a processor to implement the solute region spectrum extraction method as described above.

[0046] In the embodiment of the present application, a solute region spectrum extraction method, system and storage medium are provided, which comprehensively considers the information of different elements and different spectral lines in the same droplet, and performs a morphological recognition before and after the weighted fusion of multiple element information, so as to accurately extract the solute region spectrum in the original data of the coverage collection for quantitative analysis, and eliminate the interference of invalid spectra without solute on LIBS analysis. Compared with the traditional rectangular dot matrix coverage scanning method, the spectral intensity after correction of the present invention has higher stability, and can effectively improve the sensitivity of quantitative analysis, which is conducive to the further development and application of LIBS liquid detection technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flow chart of a solute region spectrum extraction method provided by an exemplary embodiment of the present application is shown;

[0048] Figure 2 A flow chart of a solute region spectrum extraction method based on morphological spectrum selection provided by an exemplary embodiment of the present application is shown;

[0049] Figure 3 A schematic diagram of a liquid-solid conversion LIBS liquid detection experimental system provided by an exemplary embodiment of the present application is shown;

[0050] Figure 4 A real picture of a zinc substrate after full laser coverage acquisition of liquid-solid conversion LIBS liquid detection provided by an exemplary embodiment of the present application is shown;

[0051] Figure 5 An element distribution imaging diagram of CrII283.560 nm in a liquid-solid conversion LIBS coverage acquisition mode provided by an exemplary embodiment of the present application is shown;

[0052] Figure 6 The mask of the solute deposition layer in the laser sampling point array obtained by the solute regional spectrum extraction method provided by an exemplary embodiment of the present application is shown;

[0053] Figure 7 A comparison chart of the Cd element coverage collection in a Cd, Mn, Cr mixed solution provided by an exemplary embodiment of the present application and the quantitative results of the present invention is shown;

[0054] Figure 8 A comparison chart of Mn element coverage collection in a Cd, Mn, Cr mixed solution provided by an exemplary embodiment of the present application and the quantitative results of the present invention is shown;

[0055] Fig. 9 A comparison chart of the Cr element coverage collection in a Cd, Mn, Cr mixed solution provided by an exemplary embodiment of the present application and the quantitative results of the present invention is shown;

[0056] Fig.10 A structural block diagram of a solute region spectrum extraction system provided by an exemplary embodiment of the present application is shown. DETAILED DESCRIPTION

[0057] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0058] The term "multiple" as used herein refers to two or more than two. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the related objects are in an "or" relationship.

[0059] Please refer to Figure 1 , which shows a flow chart of a solute region spectrum extraction method provided by an exemplary embodiment of the present application. The method is used in a liquid-solid conversion LIBS liquid detection scenario, and the method comprises:

[0060] Step 101 , obtaining a substrate sample, wherein the substrate sample comprises a solute deposition layer of an element to be detected, and the solute deposition layer of the element to be detected is formed by dripping liquid.

[0061] In a possible implementation, step 101 includes the following content.

[0062] Content 1. Grind the substrate to obtain a polished substrate with uniform surface roughness.

[0063] Content 2: Basic cleaning after grinding.

[0064] Content 3: Use a pipette to drop a solution containing the element to be tested onto the surface of the polished substrate.

[0065] Content 4. Then place the polished substrate flat on a constant temperature heating table for heating.

[0066] Content 5. After the heating is completed, a substrate sample containing a solute deposition layer of the element to be measured is obtained.

[0067] In one example, the flattened zinc metal substrate surface was polished with 1200-grit sandpaper to ensure uniform surface roughness, and then washed three times with alcohol and deionized water to ensure that there were no residual impurities on the zinc substrate surface that would affect LIBS detection. The substrate was then placed flat on a constant temperature heating table set at 70°C for heating, and then 40 μL of a solution containing the elements to be tested, Cd, Mn, and Cr (the concentration ratio of the three elements was 1:1:1) was dripped onto the substrate surface using a pipette gun, and finally a substrate sample containing a solute deposition layer of the elements to be tested was obtained.

[0068] Optionally, the zinc metal substrate in the above embodiment is zinc metal, and the material of the substrate may also be Al, Cu, or Si, without limitation.

[0069] The constant temperature heating stage was kept horizontal during the entire sample preparation process. The zinc substrate was cut into 2 cm × 2 cm blocks to prevent solute splashing caused by thermal effects and close distances between droplets during the evaporation of multiple droplets, or splashing of solute deposits caused by laser shock waves during LIBS spectrum acquisition, which would affect qualitative or quantitative analysis.

[0070] In a possible implementation, step 102 includes the following content.

[0071] Step 102 : laser scanning is performed on the solute deposition layer of the element to be measured and corresponding LIBS raw spectrum data is obtained.

[0072] Content 1: Optimize laser energy and delay.

[0073] The spectral intensity will increase first and then stabilize with the laser energy. The optimization means that in the embodiment of the present application, a laser energy that can make the spectral intensity reach a saturated state is selected to obtain a sufficiently large spectral intensity.

[0074] Content 2: Use the covering acquisition method to laser scan the solute deposition layer of the element to be measured at the substrate sample and obtain the LIBS original spectrum data. The size of the laser sampling point array used is M×N, and it is necessary to ensure that the laser sampling point array covers the droplet deposition layer.

[0075] There is no specific limitation on the range of M×N, and those skilled in the art can set it according to the size of the droplets. For example, the laser scanning dot matrix is ​​scanned with a step size of 0.5 mm to fully cover the entire droplet deposition layer after drying. Assuming that the droplet diameter is 8 mm, a 17*17 dot matrix is ​​sufficient to cover it.

[0076] In one example, the laser energy and delay were optimized before the LIBS experiment, where the laser energy was optimized to 100 mJ and the delay was optimized to 2 μs. Figure 3 As shown in the figure, the paraxial spectrum acquisition method is used, and the system integrates an autofocus module to reduce the negative impact on the stability of the LIBS spectrum caused by the change in the distance between the sample surface and the focusing lens due to the uneven surface of the zinc substrate or the substrate is not completely horizontal during the LIBS sampling process. The above system and experimental parameters are used to complete the coverage LIBS spectrum acquisition of the solution to be tested after liquid-solid conversion.

[0077] Furthermore, the dry droplet deposition layer on the substrate surface was scanned using a coverage acquisition method to obtain the original spectral data obtained by LIBS coverage acquisition. The laser sampling point array size used was 17×17, and a three-dimensional electric displacement platform was used to ensure that the center distance between any two adjacent laser ablation points was 0.5 mm.

[0078] In addition, during the LIBS spectrum acquisition process, it is necessary to ensure that the laser sampling point array completely covers the solute deposition layer. The distribution of the zinc substrate and the solution deposit after laser sampling in liquid-solid conversion LIBS liquid detection is as follows Figure 4 As shown in the figure, the laser sampling dot matrix completely covers the solute deposition layer, the white dot matrix is ​​the laser ablation pit caused by LIBS collection, and the nearly circular white substance is the solute deposition after the liquid-solid conversion of the solution to be tested. Due to the existence of the coffee ring effect, the white deposition is not evenly distributed on the zinc substrate, which is also an important reason for the randomness of the quantitative accuracy of the LIBS internal sampling method.

[0079] Step 103 , laser scanning is performed on the solute deposition layer of the element to be measured and corresponding LIBS raw spectrum data is obtained.

[0080] In a possible implementation, step 103 includes the following content.

[0081] Content 1: Select the spectral line whose relative intensity of the target element in the NIST spectral line library is greater than the spectral line intensity threshold K;

[0082] Content 2: averaging the spectral line intensity values ​​corresponding to the laser sampling point array to obtain an average original spectrum;

[0083] Content three: The persistent lines are screened according to the average original spectrum to obtain the available spectral lines, wherein the screening condition is the spectral lines whose signal-to-noise ratio is greater than a signal-to-noise ratio threshold S.

[0084] In general, step 103 realizes automatic spectral line selection. In one example, a persistent line with a relative intensity of the target element greater than 100 in the NIST spectral line library is selected, and the original spectral data obtained by the overlay acquisition is averaged to obtain an average original spectrum, and the spectral line with a signal-to-noise ratio greater than 1.5 in the selected persistent line is selected from the average original spectrum as an available spectral line. Specifically, the average original spectrum is the average of 289 LIBS spectra obtained by the overlay dot matrix acquisition, which includes the spectrum of the solute deposition layer and the spectrum of the zinc substrate without solute.

[0085] After the above automatic spectral line selection, the selected spectral lines of each target element are Cr element: CrII283.560nm, CrI357.870nm, CrI359.350n, CrI425.430nm, CrI428.970nm, CrI520.840nm, etc.; Mn element: MnII257.610nm, MnII259.370nm, MnII260.570nm, MnI279.480nm, MnII293.930nm, MnII294.920nm, MnI403.080nm, MnI403.310nm, etc.; Cd element: CdII226.500nm, CdI228.800nm, etc.

[0086] Step 104: Form an element distribution image of each element to be measured in the LIBS original spectrum data using the available spectral lines of each element to be measured.

[0087] Optionally, the content of element distribution imaging includes the following: element distribution images of different elements and element distribution images corresponding to different spectral lines of the same element, and the resolution of the element distribution image is the size of the laser sampling point array.

[0088] In one example, after selecting available spectral lines of multiple target elements, multiple spectral lines of the selected multiple elements are used to perform element distribution imaging on the target elements in the raw data obtained by overlay acquisition, so as to obtain a preliminary image of the solute deposition layer in the raw data obtained by overlay acquisition, such as Figure 5 The figure shows the distribution image of CrII283.560nm under the overlay acquisition mode. The element distribution image includes: element distribution images of different elements, element distribution images corresponding to different spectral lines of the same element, and the image resolution is the size of the laser sampling point array, that is, 17×17.

[0089] Step 105: Perform a first morphological recognition on all element distribution images to obtain masks corresponding to different spectral lines of each element to be tested.

[0090] Optionally, the first morphological recognition operation mode is to obtain the relative position of the droplet in the laser sampling M×N dot matrix and then extract the spectrum of the solute area of ​​the element to be measured in the LIBS raw spectrum data. Here, the droplet refers to the droplet of the solution containing the element to be measured added to the surface of the polished substrate by a pipette gun.

[0091] The data processing of the first morphological recognition includes image binarization, image connected region acquisition, maximum connected region search, closing operation and opening operation.

[0092] In a possible implementation, image binarization is used to convert the element distribution grayscale image obtained by the characteristic spectral intensity into a binary image, wherein the global threshold for binarization is calculated based on the element distribution grayscale image using the Otsu method, and the Otsu method selects a suitable global threshold by minimizing the intra-class variance of the thresholded black and white pixels. The global threshold can use different thresholds according to the characteristics of different droplets to ensure that the element distribution information is retained as much as possible during the binarization process.

[0093] In a possible implementation, the image connected region acquisition is used to acquire the connected region of the binary image, and each connected region is labeled. Here, all connected regions are found in a four-connected manner.

[0094] In a possible implementation, the maximum connected region is found by calculating the area of ​​each connected region, thereby selecting the connected component with the largest area in the binary image, and retaining and only retaining the connected component with the largest area in the binary image.

[0095] In a possible implementation, the closing operation is performed by first dilating and then corroding to eliminate narrow discontinuities between connected regions and fill small voids in connected regions to make the connected regions fuller and more connected.

[0096] In one possible implementation, the opening operation is an erosion followed by a dilation operation to smooth the contours of the object, break up narrow discontinuities, and eliminate small protrusions.

[0097] Step 106 , weighted fusion of masks corresponding to different spectral lines of each element to be measured is performed to obtain an accurate droplet distribution image.

[0098] In a possible implementation, step 106 includes the following content.

[0099] Content 1: Use the average spectral intensity of the LIBS raw spectral data corresponding to the selected spectral lines of each element to be measured as the weight of the corresponding mask;

[0100] Content 2: Weighted fusion of masks corresponding to different spectral lines of each element to be tested;

[0101] Content 3: Obtain an accurate droplet distribution image that integrates the contour information provided by multiple spectral lines of each element to be measured.

[0102] In general, step 106 realizes information fusion. In order to eliminate the one-sidedness of the solute distribution mask corresponding to a single spectral line, which in turn leads to the problem of poor accuracy in selecting the solute region spectrum, the solute deposition layer distribution information provided by different spectral lines of three different target elements in the same droplet is integrated, and the multiple element distribution masks of multiple spectral lines are fused to obtain a comprehensive distribution grayscale map of the droplet.

[0103] In a possible implementation, the stronger the spectral intensity, the clearer and easier to distinguish the droplet distribution area in the element distribution map relative to the background (spectral line of the zinc substrate without solute), which means that the information carried by the element imaging map corresponding to the spectral line can more accurately and reliably reflect the distribution of the solute. Therefore, the average spectral intensity of the LIBS original spectral data covering the selected spectral lines of each element to be measured in the collected original spectral data is used as the weight of the corresponding mask, which is equivalent to assigning different weights according to the distribution information of the solute deposition layer carried by different spectral lines, thereby comprehensively utilizing the solute distribution information carried by all spectral lines. By weighted addition of the obtained multiple element distribution masks, an accurate grayscale image of the solute deposition layer that integrates the solute distribution information provided by multiple spectral lines of different elements is further obtained.

[0104] Step 107 , performing a second morphological recognition on the accurate droplet distribution image to obtain a mask representing the distribution of the solute deposition layer of the element to be measured.

[0105] Optionally, the data processing of the second morphological recognition includes image binarization, image connected region acquisition, small area deletion, closing operation, maximum connected region search and opening operation.

[0106] In one example, after weighted fusion of mask information, the position of the deposition layer in the grayscale image of the solute deposition layer distribution can represent the accurate distribution of the droplets relative to the coverage acquisition rectangular laser dot matrix. The mask of the deposition layer in the grayscale image can be used to extract the solute area spectrum from the coverage acquisition 17×17 original dot matrix data. Therefore, a second morphological recognition is performed on the fused droplet distribution image to obtain the final mask representing the distribution of the solute deposition of the substrate drying droplets.

[0107] In one example, if Figure 6 As shown in the figure, the evaporated droplet mask is obtained by the solute area spectrum extraction method based on morphological spectral selection. This mask consists of two parts: one is the black background representing the solute-free area on the zinc substrate, and the other is the white mask representing the solute area on the zinc substrate. Obviously, Figure 4 The distribution area of ​​white solute in the middle Figure 6 The distribution of the white mask relative to the black background and the shapes of the two are very similar. Figure 4 The white solute area in the middle is slightly expanded. This is because the ablation range of the sampling points outside the solute area overlaps with the solute deposition layer during the coverage collection process, resulting in a small amount of target elements being detected at the sampling points outside the solute area, which are then captured by the multi-element morphological selection method. Figure 4 Actual outline of the solute deposition layer in the elemental distribution, with the yellow area in the mask slightly expanded.

[0108] Step 108 , extracting the solute region spectrum of the element to be measured from the LIBS raw spectrum data through the mask of the solute deposition layer distribution of the element to be measured.

[0109] Among them, a mask of droplet distribution relative to the covering sampling point array has been obtained, through which the solute area spectrum can be extracted from the original spectral data obtained by covering sampling, eliminating the influence of invalid spectrum on qualitative and quantitative analysis.

[0110] Furthermore, based on the above-mentioned solute region spectrum extraction method, the content of the element to be analyzed in the substrate sample can be further predicted. The content of the method is as follows.

[0111] Step 1: Extract the solute region spectrum of the substrate sample by the solute region spectrum extraction method, remove the interference of invalid spectrum covering the original sampling data, and then obtain the corrected spectrum intensity I, and establish a calibration curve of the corrected spectrum intensity I and the element content Cs in the sample: Cs=kI+b, where k and b are coefficients.

[0112] Step 2: extract the solute region spectrum of the unknown concentration sample from the original data of the overlay sampling by the above-mentioned solute region spectrum extraction method, and predict the content of the element to be analyzed in the sample based on the established calibration curve. In other embodiments, the solute region spectrum can also be first extracted from the overlay sampling spectrum of the predicted sample, and then the substrate sample spectrum is corrected, and after the calibration curve of the training sample spectrum intensity I and the element content Cs in the sample is established, the content of the element to be analyzed in the sample is predicted.

[0113] Therefore, in summary, the above steps 101 to 108 realize the extraction of the solute region spectrum from the coverage sampling spectrum of the sample. On the basis of steps 101 to 108, steps 1 and 2 are added to realize the prediction of the content of the element to be analyzed in the substrate sample.

[0114] In this embodiment, after establishing the calibration curves of the normalized spectral intensities of CdII226.500nm, MnII257.610nm, and CrII283.560nm after correction and the Cd, Mn, and Cr contents Cs in the sample, the content of the elements in the corrected sample was predicted based on the calibration curves using a leave-one-out cross-validation method.

[0115] In one possible embodiment, a series of target elements with concentration gradients are prepared using deionized water, and the mixed solutions are Cd, Mn, and Cr (the concentration ratio of the three elements is 1:1:1), and the concentration gradients are 0.1 mg / L, 0.2 mg / L, 0.4 mg / L, 0.6 mg / L, 0.8 mg / L, and 1.0 mg / L. A standard solution with a volume of 40 μL of each concentration is added dropwise to the surface of the zinc substrate and combined with a constant temperature heating table to complete the liquid-solid conversion sample preparation. The LIBS spectrum of the deposit on the zinc substrate is collected using a covering collection method, and the solute region spectrum extraction is completed using the method proposed in the present invention.

[0116] Using the original spectrum intensity corresponding to the overlay acquisition method and the intensity corresponding to the solute region spectrum, the element calibration curve is established as follows Figure 7 (a) and (b) in Figure 8 (a) and (b) in Fig. 9 It can be seen that after the invalid spectra are eliminated by the solute region spectrum extraction method based on morphological spectrum selection of the present invention, the calibration curve R, RMSE and MAPE of the cross-validation prediction set established have no obvious changes, but the two performance indicators of LOD and RSD are significantly improved compared with the calibration curve established by the coverage acquisition method. The following index comparison is shown in Table 1 below.

[0117] Table 1 covers the simulation results of the collected spectra and the solute spectrum performance indicators extracted by the present invention

[0118]

[0119] In summary, the method provided in the embodiment of the present application applies morphological operations to the element distribution image covering the original spectrum data collected to extract the solute region spectrum, thereby eliminating the invalid spectrum collected in the substrate without solute distribution area, thereby improving the spectral stability and detection sensitivity. In addition, the method of the present invention does not require any hardware equipment to assist in collection based on the conventional LIBS liquid detection system, and the equipment complexity is relatively low.

[0120] Please refer to Fig.10 , shows a structural block diagram of a solute region spectrum extraction system provided by an exemplary embodiment of the present application, the system is used for liquid-solid conversion LIBS liquid detection scenario, the system includes:

[0121] The sample acquisition module 1001 is used to acquire a substrate sample, wherein the substrate sample includes a solute deposition layer of an element to be detected, and the solute deposition layer of the element to be detected is formed by dripping liquid;

[0122] The spectrum acquisition module 1002 is used to perform laser scanning on the solute deposition layer of the element to be measured and obtain the corresponding LIBS original spectrum data;

[0123] A spectral line selection module 1003 is used to determine the available spectral lines of each element to be measured according to the LIBS original spectral data, and the available spectral lines meet the signal-to-noise ratio requirement;

[0124] An element distribution module 1004 is used to form an element distribution image of each element to be measured in the LIBS original spectrum data using the available spectral lines of each element to be measured;

[0125] The primary recognition module 1005 is used to perform a first morphological recognition on all element distribution images to obtain masks corresponding to different spectral lines of each element to be tested;

[0126] The information fusion module 1006 is used to perform weighted fusion on the masks corresponding to different spectral lines of each element to be measured to obtain an accurate droplet distribution image;

[0127] A secondary recognition module 1007 is used to perform a second morphological recognition on the accurate droplet distribution image to obtain a mask representing the distribution of the solute deposition layer of the element to be measured;

[0128] The solute region module 1008 is used to extract the solute region spectrum of the element to be measured from the LIBS original spectrum data through the mask of the solute deposition layer distribution of the element to be measured.

[0129] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0130] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0131] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A solute region spectrum extraction method, characterized in that: The method is used in a liquid-solid conversion LIBS liquid detection scenario, and the method comprises: Obtaining a substrate sample, wherein the substrate sample comprises a solute deposition layer of an element to be detected, wherein the solute deposition layer of the element to be detected is formed by dripping liquid; Performing laser scanning on the solute deposition layer of the element to be measured and acquiring corresponding LIBS raw spectrum data; Determine available spectral lines of each element to be measured according to the LIBS original spectral data, wherein the available spectral lines meet the signal-to-noise ratio requirement; Using the available spectral lines of the respective elements to be measured, an element distribution image of the respective elements to be measured in the LIBS raw spectrum data is formed; Perform the first morphological recognition on all element distribution images to obtain the masks corresponding to different spectral lines of each element to be tested; The average spectral intensity of the LIBS original spectral data corresponding to the selected spectral line of each element to be measured is used as the weight of the corresponding mask; Perform weighted fusion on the masks corresponding to different spectral lines of each element to be measured; Obtain an accurate droplet distribution image that integrates the profile information provided by multiple spectral lines of each element to be measured; Performing a second morphological recognition on the accurate droplet distribution image to obtain a mask representing the distribution of the solute deposition layer of the element to be measured; The solute region spectrum of the element to be measured is extracted from the LIBS original spectrum data through the mask of the solute deposition layer distribution of the element to be measured.

2. The method according to claim 1, characterized in that The obtaining of the substrate sample comprises: Grinding the substrate to obtain a polished substrate with a uniform surface roughness; basically cleaning after said grinding; Adding droplets of a solution containing the element to be tested onto the surface of the polished substrate by a pipette; Then the polished substrate is placed flat on a constant temperature heating table for heating; After the heating is completed, a substrate sample containing a solute deposition layer of the element to be measured is obtained.

3. The method according to claim 1, characterized in that The laser scanning of the solute deposition layer of the element to be measured and obtaining the corresponding LIBS original spectrum data includes: Optimize laser energy and delay; The solute deposition layer of the element to be measured at the substrate sample is laser scanned using a covering acquisition method to obtain LIBS original spectrum data, wherein the size of the laser sampling point array used is M×N, and it is necessary to ensure that the laser sampling point array covers the droplet deposition layer.

4. The method according to claim 1, characterized in that: Determining available spectral lines of each element to be measured according to the LIBS original spectral data, wherein the available spectral lines meet the signal-to-noise ratio requirement, includes: Select the persistent line whose relative intensity of the target element in the NIST spectral line library is greater than the spectral line intensity threshold K; The average of the spectral line intensity values ​​corresponding to the laser sampling point array is taken to obtain an average original spectrum; The persistent lines are screened according to the average original spectrum to obtain the available spectral lines, wherein the screening condition is the spectral lines whose signal-to-noise ratio is greater than a signal-to-noise ratio threshold S.

5. The method according to claim 1, characterized in that The content of element distribution imaging includes: element distribution images of different elements and element distribution images corresponding to different spectral lines of the same element. The resolution of the element distribution image is the size of the laser sampling point array.

6. The method according to claim 1, characterized in that The operation mode of the first morphological recognition is to obtain the relative position of the droplet in the laser sampling M×N dot matrix and then extract the spectrum of the solute area of ​​the element to be measured in the LIBS original spectrum data; The data processing of the first morphological recognition includes image binarization, image connected region acquisition, maximum connected region search, closing operation and opening operation.

7. The method according to claim 1, characterized in that The data processing of the second morphological recognition includes image binarization, image connected region acquisition, small area deletion, closing operation, maximum connected region search and opening operation.

8. A solute region spectrum extraction system, characterized in that: The system is used for liquid-solid conversion LIBS liquid detection scenarios, and the system includes: A sample acquisition module, used for acquiring a substrate sample, wherein the substrate sample comprises a solute deposition layer of an element to be detected, wherein the solute deposition layer of the element to be detected is formed by dripping liquid; A spectrum acquisition module, used for laser scanning the solute deposition layer of the element to be measured and obtaining the corresponding LIBS original spectrum data; A spectral line selection module, used to determine the available spectral lines of each element to be measured according to the LIBS original spectral data, wherein the available spectral lines meet the signal-to-noise ratio requirement; An element distribution module, used to form an element distribution image of each element to be measured in the LIBS original spectrum data using the available spectral lines of each element to be measured; The primary recognition module is used to perform the first morphological recognition on all element distribution images to obtain the masks corresponding to different spectral lines of each element to be tested; An information fusion module is used to use the average spectral intensity of the LIBS original spectral data corresponding to the selected spectral lines of each element to be tested as the weight of the corresponding mask, and is also used to weightedly fuse the masks corresponding to different spectral lines of each element to be tested, and is also used to obtain an accurate droplet distribution image that fuses the contour information provided by multiple spectral lines of each element to be tested; A secondary recognition module, used for performing a second morphological recognition on the accurate droplet distribution image to obtain a mask representing the distribution of the solute deposition layer of the element to be measured; The solute region module is used to extract the solute region spectrum of the element to be measured from the LIBS original spectrum data through the mask of the solute deposition layer distribution of the element to be measured.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is loaded and executed by a processor to implement the solute region spectrum extraction method as described in any one of claims 1 to 7 above.

Images