A method and system for detecting heavy metals in soil

Abstract

The present invention discloses a soil heavy metal detection method and system. The detection method combines near-infrared spectral analysis with hyperspectral imaging technology. First, a soil photo and spectral image are obtained. Then, spectral feature data is extracted from the spectral image, and the spectral angle is calculated with the spectral feature data in the spectral library. The similarity between the two is determined by the size of the spectral angle, so as to qualitatively analyze the types of soil heavy metals. Then, the concentration of soil heavy metals is quantitatively calculated by the preset linear regression equation calculation program of the spectral feature data and the absolute concentration. The detection system includes a flight device, an analysis system, and an optical camera, a multispectral sensor and a GPS positioning system carried on the flight device; the spectral library is set in the analysis system. The detection system and method have the advantages of fast, non-destructive, large-area detection and monitoring, and have broad application prospects in soil pollution detection and monitoring.

Description

Technical Field

[0001] The present invention relates to the technical field of soil heavy metal detection, and in particular to a soil heavy metal detection method and system. Background Art

[0002] At present, the standard method for soil heavy metal detection is still mainly based on laboratory confirmatory analysis, such as atomic absorption spectrometry (AAS), atomic fluorescence spectrometry (AFS), inductively coupled plasma optical emission spectrometry (ICP-OES), inductively coupled plasma mass spectrometry (ICP-MS), wavelength dispersion X-ray fluorescence spectrometry (WDXRF), etc. However, the above methods require complex and time-consuming sample preparation and digestion or extraction, or large instrument size and complex and precise hardware configuration, which cannot be used for on-site and rapid analysis of soil, making it difficult to monitor and prevent soil heavy metal pollution in a timely and effective manner from the source. Summary of the invention

[0003] In order to overcome the shortcomings of the prior art, one of the purposes of the present invention is to provide a soil heavy metal detection method, which can detect the types of heavy metals in the soil through multi-spectral imaging and sensing technology, and then detect the concentration of heavy metals in the soil through a set calculation concentration relationship; the second purpose of the present invention is to provide a soil heavy metal detection system, which carries a multi-spectral sensor and an optical camera on a flying device to realize on-site and rapid analysis of the types and concentrations of heavy metals in the soil.

[0004] One of the purposes of the present invention is achieved by the following technical solution:

[0005] A method for detecting heavy metals in soil comprises the following steps:

[0006] 1) Establish a spectral library that stores spectral characteristic data of soils contaminated with known types and concentrations of heavy metals;

[0007] 2) Obtain soil photos and spectral images of the soil area to be tested;

[0008] 3) Extracting spectral characteristic data from the spectral image obtained in step 2), and calculating the spectral angle with the reference spectral characteristic data in the spectral library, comparing the similarity between the two, thereby obtaining the types of heavy metals contained in the soil of the test area, and then calculating the absolute concentration of heavy metals in the soil through the set linear regression equation.

[0009] Furthermore, step 2) also includes the following steps: after collecting the spectral image, the analysis system uses an image segmentation method to remove image content that is irrelevant to the soil.

[0010] Furthermore, step 2) also includes the following steps: using the standard normal transformation method to correct each spectral feature data, the specific steps are: subtracting the average value of all spectra from a single original spectrum, and then dividing it by the standard deviation of the single spectral feature data.

[0011] Furthermore, step 3) also includes the following step: using spectral angle mapping to compare the similarity between the spectral feature data obtained in step 2) and the spectral feature data in the spectral library to obtain a correlation index.

[0012] Furthermore, step 3) also includes the following steps: comparing the calculated correlation index with a predetermined value, selecting data greater than or equal to the predetermined value as data for linear regression analysis, obtaining a linear regression equation, and calculating the absolute concentration of the heavy metal.

[0013] Furthermore, in step 3), the predetermined value may be any value within the range of 0.9 to 1.0.

[0014] Furthermore, the step 4) is: comparing the absolute concentration obtained in step 3) with the known concentration value sampled at a specific point in the same test area. If the error is less than 0.1, the absolute concentration obtained in step 3) is the heavy metal concentration in the test area.

[0015] The second object of the present invention is achieved by adopting the following technical solution:

[0016] A soil heavy metal detection system, the detection method of the detection system is the above-mentioned soil heavy metal detection method, comprising a flying device, an analysis system, and an optical camera, a multi-spectral sensor and a GPS positioning system carried on the flying device; the optical camera, the GPS positioning system and the multi-spectral sensor are all connected to the analysis system by electrical signals, and the analysis system is provided with a spectral library storing spectral characteristic data of soil contaminated with known types and concentrations of heavy metals;

[0017] A spectrum library storing spectrum characteristic data of soil contaminated by known types and concentrations of heavy metals is established in the analysis system; the flying device receives instructions from the GPS positioning system and flies to the sky above a specific soil area to be tested, an optical camera carried on the flying device takes photos of the soil and a multi-spectral sensor obtains a spectrum image of the soil, and then transmits the photos and spectrum images of the soil to the analysis system in the form of electrical signals; the analysis system extracts spectrum characteristic data from the obtained spectrum image, calculates the spectrum angle with the reference spectrum characteristic data in the spectrum library, compares the similarity between the two, thereby obtaining the types of heavy metals contained in the soil in the test area, and then calculates the absolute concentration of the heavy metals in the soil through a set linear regression equation.

[0018] Furthermore, the analysis system is used to carry a spectral library that stores spectral characteristic data of soil contaminated with known types and concentrations of heavy metals, receive spectral images transmitted by a multi-spectral sensor, extract spectral characteristic data therein, and then correct the extracted spectral characteristic data. Subsequently, the spectral angle is calculated based on the corrected spectral characteristic data and the reference spectral characteristic data of the spectral library, and the similarity between the two is compared to obtain the types of heavy metals contained in the soil in the test area, and then the absolute concentration of the heavy metals in the soil is calculated through a set linear regression equation.

[0019] Furthermore, the soil heavy metal detection system also includes an image storage device, which is electrically connected to the optical camera and the analysis system respectively; the image storage device is used to store the soil photos of the optical camera and transmit the stored soil photo data to the analysis system.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] (1) The detection method of the present invention combines near-infrared spectral analysis with hyperspectral imaging technology. First, a photo and spectral image of the soil in the test area are obtained. Then, spectral feature data are extracted from the spectral image and the spectral angle is calculated with the spectral feature data in the spectral library. The smaller the spectral angle, the greater the similarity. Since the spectral feature data in the spectral library corresponds to a specific type of heavy metal, the type of heavy metal in the measured soil can be inferred. Then, the concentration of heavy metals in the soil can be calculated by a preset absolute concentration and spectral feature data linear regression equation calculation program.

[0022] (2) In the detection method of the present invention, before comparing the acquired spectral feature data with the spectral library, the data is preprocessed using the standard normalized variation (SNV). This method is used to eliminate the effects of surface scattering and optical path changes on the spectrum. SNV assumes that the absorbance values ​​of each wavelength point in each spectrum meet a certain distribution. The process is to subtract the average value of the spectrum from the original spectrum and divide it by the standard deviation of the spectral data. Since SNV corrects the spectrum of a single sample, even if the spectral data varies greatly between samples, it can ensure a strong correction ability.

[0023] (3) The detection system of the present invention carries an optical camera and a multi-spectral sensor on a flying device to capture a wide range of soil images and soil spectral characteristic data, and transmits the soil photos and soil spectral images to the analysis system. The analysis system is provided with a spectral library that stores spectral characteristic data of soil contaminated with known types and concentrations of heavy metals. The spectral characteristics of the soil obtained from the spectral image are calculated with the known spectral characteristic data in the spectral library to obtain a spectral angle, and the similarity between the two is determined by the size of the spectral angle, so as to qualitatively analyze the types of heavy metals in the soil. The analysis system can then quantitatively analyze the concentration of heavy metals in the soil through a calculation program of a linear regression equation between the preset spectral characteristic data and the absolute concentration. The detection system of the present invention has the advantages of rapid, non-destructive, and large-area detection and monitoring, and has broad application prospects in soil pollution detection and monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a flow chart of a soil heavy metal detection method according to an embodiment of the present invention. ;

[0025] Figure 2 This is a system framework diagram for soil heavy metal detection according to an embodiment of the present invention.

[0026] Figure 3 for Figure 2 The flowchart of the soil heavy metal detection system is shown.

[0027] In the figure: 110, flying device; 120, analysis system; 130, image storage device; 140, optical camera; 150, multi-spectral sensor; 160, GPS positioning system. DETAILED DESCRIPTION

[0028] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods. It should be noted that, under the premise of no conflict, the various embodiments or technical features described below can be arbitrarily combined to form a new embodiment.

[0029] like Figure 1As shown, a method for detecting heavy metals in soil comprises the following steps:

[0030] 1) Establish a spectral library that stores spectral characteristic data of soils contaminated with known types and concentrations of heavy metals;

[0031] 2) Obtain soil photos and spectral images of the soil area to be tested;

[0032] 3) Extracting spectral characteristic data from the spectral image obtained in step 2), and calculating the spectral angle with the reference spectral characteristic data in the spectral library, comparing the similarity between the two, thereby obtaining the types of heavy metals contained in the soil of the test area, and then calculating the absolute concentration of heavy metals in the soil through the set linear regression equation.

[0033] Step 2) also includes the following steps 2.1 and 2.2. In step 2.1, after collecting the spectral image, the image segmentation method is used to remove the image content irrelevant to the soil. In step 2.2, the standard normal transformation method is used to correct each spectral feature data, which is specifically implemented as follows: subtract the average value of all spectra from a single original spectrum, and then divide it by the standard deviation of the single spectral feature data.

[0034] Step 3) also includes the following steps 3.1 and 3.2. Wherein, step 3.1 is to use spectral angle mapping to compare the similarity between the spectral feature data obtained in step 2) and the spectral feature data of the spectral library to obtain a correlation index. Wherein, step 3.2 is to compare the calculated correlation index with a predetermined value, select data greater than or equal to the predetermined value as data for linear regression analysis, obtain a linear regression equation, and calculate the absolute concentration of heavy metals. The predetermined value can be any value in the range of 0.9 to 1.0.

[0035] In this embodiment, the soil heavy metal detection method further includes step 4). Step 4) is: comparing the absolute concentration obtained in step 3) with the known concentration value sampled at a specific point in the same test area. If the error is less than 0.1, the absolute concentration obtained in step 3) is the heavy metal concentration of the test area.

[0036] like Figure 2 As shown, a soil heavy metal detection system includes a flying device 110, an analysis system 120, an image storage device 130, and an optical camera 140, a multi-spectral sensor 150 and a GPS positioning system 160 carried on the flying device 110; the optical camera 140, the GPS positioning system 160 and the multi-spectral sensor are all electrically connected to the analysis system 120; the image storage device 130 is electrically connected to the optical camera 140 and the analysis system 120, respectively, and the image storage device 130 is used to store photos of the soil and spectral images of the soil.

[0037] The flying device 110 receives the instruction of the GPS positioning system 160 and flies to the sky above the specific soil area to be tested. The optical camera 140 carried on the flying device 110 takes a photo of the soil and the multi-spectral sensor obtains the spectral image of the soil, and then transmits the photo and spectral image of the soil to the analysis system 120 in the form of electrical signals; a spectral library storing spectral characteristic data of soil contaminated with known types and concentrations of heavy metals is established in the analysis system 120. The analysis system 120 is used to receive the spectral image transmitted by the multi-spectral sensor, extract the spectral characteristic data therein, and then correct the extracted spectral characteristic data. Then, the spectral angle is calculated based on the corrected spectral characteristic data and the reference spectral characteristic data of the spectral library, and the similarity between the two is compared, so as to obtain the types of heavy metals contained in the soil of the test area, and then the absolute concentration of the heavy metals in the soil is calculated through a set linear regression equation.

[0038] The analysis system 120 is loaded with a VideometerLab multi-spectral measurement system, which processes the spectral image obtained by the multi-spectral sensor 150 and extracts characteristic data of the spectral image. The multi-spectral sensor 150 is a sensor that simultaneously records electromagnetic energy in multiple bands, mainly a scanner concept, and involves a device for collecting electromagnetic waves reflected from a refractive index into a lens and a reflector, separating the collected electromagnetic waves through a filter, separating the separated wavelengths according to the separated wavelengths, and recording the image on a magnetic tape.

[0039] like Figure 3 As shown, the following Figure 2 The detection process of the soil heavy metal detection system is shown in the figure. The specific detection process is described as follows.

[0040] 1) Establishing a spectral library storing spectral characteristic data of soil contaminated with known types and concentrations of heavy metals in the analysis device;

[0041] 2) The flying device 110 flies to the sky above the soil area to be tested through the GPS positioning system 160, and the optical camera 140 carried on the flying device 110 takes a photo of the soil, and then the multi-spectral sensor obtains the spectral image of the soil, and then transmits the photo and spectral image of the soil to the analysis system 120 in the form of electrical signals; the analysis system 120 uses the image segmentation method to remove the image content irrelevant to the soil,

[0042] 3) The analysis system 120 extracts spectral feature data from the spectral image obtained in step 1). Specifically, based on the differences in spectral features of different components, the spectral feature extraction first uses Canonical Discriminant Analysis (CDA) and threshold setting to complete background removal and region of interest (ROI) segmentation. The threshold segmentation algorithm defines the regional affiliation of different targets in the image by thresholds. The basic principle of threshold segmentation is: first select a grayscale threshold in the grayscale value range of the image, and then apply the threshold to all pixels in the image. Then, a statistical method is used to obtain the average spectral reflectance value of the soil in the test area (region of interest); the image feature extraction also first performs CDA and threshold segmentation, and then extracts the characteristic values ​​of heavy metal types through Blob analysis.

[0043] After extracting the spectral feature data, it is corrected. The analysis system 120 uses the standard normalized variation (SNV) method to correct each spectral feature data. The specific steps are: subtract the average value of all spectra from a single original spectrum, and then divide it by the standard deviation of the single spectral feature data. This method is used to eliminate the effects of surface scattering and optical path changes on the spectrum. SNV assumes that the absorbance values ​​of each wavelength point in each spectrum meet a certain distribution. The process is to subtract the average value of the spectrum from the original spectrum and divide it by the standard deviation of the spectral data. Since SNV corrects the spectrum of a single sample, even if the spectral data varies greatly between samples, it can ensure a strong correction ability.

[0044] The spectral angle is calculated by comparing the corrected spectral feature data with the reference spectral feature data of the spectral library, and the similarity between the two is compared to obtain the types of heavy metals contained in the soil of the test area, and then the absolute concentration of heavy metals in the soil is calculated by the set linear regression equation. Specifically, the spectral angle is calculated by calculating the angular difference between the reference spectral vector of the reference spectral data and the multispectral image vector of the multispectral image data. Spectral Angle Mapper (SAM) is used, that is, the angle similarity coefficient angle cosine is used to describe the similarity between spectra. For two spectral vectors x and y, the mathematical definition of the spectral angle of the above SAM method is as follows: The smaller the spectral angle, the greater the similarity between spectra. When classifying, the average spectral vector of the general category is used as the reference spectrum, and the unknown sample is classified into the reference spectrum category with the smallest spectral angle. The specific algorithm flow of the spectral angle is as follows:

[0045] a) Calculate the angle θ between the pixel i at position (x, y) in the reference image and the actual image;

[0046] b) Repeat a) for each row of pixels, where each pixel has a corresponding position in a given window, and the total number of corners in the similar images is equal to the number of pixels in the reference image and the actual image;

[0047] c) applying K-means clustering algorithm to the corners in the similar images generated in b);

[0048] d) Mark the generated classes into classes with changes and classes without changes;

[0049] e) re-labeling the pixels whose neighboring pixels are concentrated in the focus area with the “most frequently occurring” label;

[0050] f) Evaluate the accuracy by comparing the mapped image generated by e) with the correlation image.

[0051] The analysis system 120 is preset with a linear regression equation for absolute concentration and spectral feature data, which can be used to calculate the concentration of heavy metals in the soil. The known formula is: relative concentration = absolute concentration / standard concentration, and the standard concentration is 1 mol / L, that is, the relative concentration is the absolute concentration. The linear regression equation for absolute concentration and spectral feature data is obtained by comparing the data of the spectral library with the heavy metal concentration as a reference. Among them, the linear regression equation is obtained by comparing the calculated correlation index with a predetermined value, and selecting data greater than or equal to the predetermined value as the data for linear regression analysis, and the predetermined value is 0.9 to 1.0.

[0052] 4) Concentration verification: Compare the absolute concentration obtained in step 3) with the known concentration value sampled at a specific point in the same test area. If the error is less than 0.1, the absolute concentration obtained in step 3) is the heavy metal concentration in the test area.

[0053] The above-mentioned embodiments are only preferred embodiments of the present invention and cannot be used to limit the scope of protection of the present invention. Any non-substantial changes and substitutions made by technicians in this field on the basis of the present invention shall fall within the scope of protection required by the present invention.

Claims

1. A method for detecting heavy metals in soil, characterized in that: The following steps are involved: 1) Establish a spectral library that stores spectral characteristic data of soils contaminated with known types and concentrations of heavy metals; 2) Obtain soil photos and spectral images of the soil area to be tested; 3) extracting spectral characteristic data from the spectral image obtained in step 2), and calculating the spectral angle with the reference spectral characteristic data in the spectral library, comparing the similarity between the two, thereby obtaining the types of heavy metals contained in the soil of the test area, and then calculating the absolute concentration of heavy metals in the soil through a set linear regression equation; Specifically, step 2) also includes: The standard normal transformation method was used to correct each spectral feature data; calculating a spectral angle by calculating an angular difference between a reference spectral vector of the reference spectral data and a multispectral image vector of the multispectral image data; The algorithm flow of the specific spectral angle is as follows: a) Calculate the angle θ between the pixel i at position (x, y) in the reference image and the actual image; b) Repeat a) for each row of pixels, where each pixel has a corresponding position in a given window, and the total number of corners in the similar images is equal to the number of pixels in the reference image and the actual image; c) applying K-means clustering algorithm to the corners in the similar images generated in b); d) Mark the generated classes into classes with changes and classes without changes; e) re-labeling the pixels whose neighboring pixels are concentrated in the focus area with the "most frequently occurring" label; f) Evaluate the accuracy by comparing the mapped image generated by e) with the correlation image.

2. The soil heavy metal detection method according to claim 1, characterized in that: Step 2) also includes the following steps: after collecting the spectral image, using image segmentation method to remove image content irrelevant to the soil.

3. The soil heavy metal detection method according to claim 1, characterized in that: Step 2) also includes the following step of correcting each spectral feature data using a standard normal transformation method, which is specifically implemented as follows: subtracting the average value of all spectra from a single original spectrum, and then dividing it by the standard deviation of the single spectral feature data.

4. The soil heavy metal detection method according to claim 1, characterized in that: Step 3) also includes the following step: using spectral angle mapping to compare the similarity between the spectral feature data obtained in step 2) and the spectral feature data in the spectral library to obtain a correlation index.

5. The soil heavy metal detection method according to claim 4, characterized in that: Step 3) also includes the following steps: comparing the calculated correlation index with a predetermined value, selecting data greater than or equal to the predetermined value as data for linear regression analysis, obtaining a linear regression equation, and calculating the absolute concentration of the heavy metal.

6. The soil heavy metal detection method according to claim 5, characterized in that: In step 3), the predetermined value may be any value within the range of 0.9 to 1.

0.

7. The soil heavy metal detection method according to claim 1, characterized in that: The method further includes step 4), which is: comparing the absolute concentration obtained in step 3) with the known concentration value sampled at a specific point in the same test area. If the error is less than 0.1, the absolute concentration obtained in step 3) is the heavy metal concentration in the test area.

8. A soil heavy metal detection system, characterized in that: The detection method of the detection system is the soil heavy metal detection method as described in any one of claims 1 to 7, comprising a flying device, an analysis system, and an optical camera, a multi-spectral sensor and a GPS positioning system carried on the flying device; the optical camera, the GPS positioning system and the multi-spectral sensor are all connected to the analysis system by electrical signals, and the analysis system is provided with a spectral library storing spectral characteristic data of soil contaminated with known types and concentrations of heavy metals; A spectrum library storing spectrum characteristic data of soil contaminated by known types and concentrations of heavy metals is established in the analysis system; the flying device receives instructions from the GPS positioning system and flies to the sky above a specific soil area to be tested, an optical camera carried on the flying device takes photos of the soil and a multi-spectral sensor obtains a spectrum image of the soil, and then transmits the photos and spectrum images of the soil to the analysis system in the form of electrical signals; the analysis system extracts spectrum characteristic data from the obtained spectrum image, calculates the spectrum angle with the reference spectrum characteristic data in the spectrum library, compares the similarity between the two, thereby obtaining the types of heavy metals contained in the soil in the test area, and then calculates the absolute concentration of the heavy metals in the soil through a set linear regression equation.

9. The soil heavy metal detection system according to claim 8, characterized in that: The analysis system is used to carry a spectral library that stores spectral feature data of soil contaminated with known types and concentrations of heavy metals, receive spectral images transmitted by a multi-spectral sensor, extract spectral feature data therein, and then correct the extracted spectral feature data. Then, the spectral angle is calculated based on the corrected spectral feature data and the reference spectral feature data of the spectral library, and the similarity between the two is compared to obtain the types of heavy metals contained in the soil in the test area, and then the absolute concentration of the heavy metals in the soil is calculated through a set linear regression equation.

10. The soil heavy metal detection system according to claim 8, characterized in that: The soil heavy metal detection system also includes an image storage device, which is electrically connected to the optical camera and the analysis system respectively; the image storage device is used to store soil photos taken by the optical camera and transmit the stored soil photo data to the analysis system.

Images