A method and device for determining the thickness of a black soil layer and a medium

Abstract

This application discloses a method, apparatus, and medium for determining the thickness of a black soil layer, relating to the field of soil layer thickness determination technology. The method includes: performing absorbance conversion and smoothing on the reflectance of any pixel in a preprocessed target image at various wavelengths within the effective band, obtaining the processed absorbance of the current pixel at various wavelengths within the effective band, and inputting this processed absorbance into a soil organic matter content prediction model to obtain a predicted value of the soil organic matter content of the current pixel; determining a pixel-level vertical distribution map of soil organic matter content based on the predicted values ​​and spatial coordinates of the soil organic matter content of each pixel in the preprocessed target image; determining the black soil layer identification result based on the pixel-level vertical distribution map of soil organic matter content; and determining the black soil layer thickness based on the black soil layer identification result and the pixel-level vertical distribution map of soil organic matter content. This application improves the accuracy of black soil layer thickness determination.

Description

Technical Field

[0001] This application relates to the field of soil layer thickness determination technology, and in particular to a method, apparatus and medium for determining the thickness of black soil layers. Background Technology

[0002] Black soil, as an important topsoil layer, is rich in organic matter and possesses favorable physicochemical properties. The thickness of the black soil layer not only reflects the soil's maturity but also plays a crucial role in rapidly identifying soil fertility outdoors. Furthermore, changes in the black soil layer are of significant guiding value for assessing the degree of black soil degradation. Accurate monitoring of black soil layer thickness helps to comprehensively understand and grasp the distribution and characteristics of regional land resources, providing a scientific basis and technical support for agricultural production and land use planning.

[0003] Traditional methods for identifying black soil layers mainly rely on visual inspection and colorimetric charts. This approach is heavily influenced by subjective judgment, making it difficult to achieve high-precision and standardized identification, especially in areas with blurred soil layer transitions or significant human disturbance, where misjudgments are prone to occur. Furthermore, traditional methods cannot quantify the continuous changes of black soil layers across soil profiles, failing to meet the demands of high-precision research. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, and medium for determining the thickness of black soil layers, so as to solve the problem of low accuracy in determining the thickness of black soil layers.

[0005] To achieve the above objectives, this application provides the following solution.

[0006] Firstly, this application provides a method for determining the thickness of a black soil layer, including:

[0007] Acquire a target image; the target image is a hyperspectral image containing the original profile of the target soil column; the spectral value of each pixel in the hyperspectral image includes the reflectance and spatial coordinates of the pixel at multiple wavelengths in a preset band;

[0008] The target image is preprocessed to obtain a preprocessed target image;

[0009] Any pixel in the preprocessed target image is identified as the current pixel, and the reflectance of the current pixel at each wavelength in the effective band is determined.

[0010] The reflectance of the current pixel at each wavelength in the effective band is sequentially converted and smoothed to obtain the processed absorbance of the current pixel at each wavelength in the effective band.

[0011] The processed absorbance of the current pixel at each wavelength in the effective band is input into the soil organic matter content prediction model to obtain the predicted value of the soil organic matter content of the current pixel; the soil organic matter content prediction model is determined based on the spectral learning machine model.

[0012] Based on the predicted values ​​and spatial coordinates of soil organic matter content of each pixel in the preprocessed target image, a pixel-level vertical distribution map of soil organic matter content in the original profile of the target soil column is determined.

[0013] Based on the pixel-level vertical distribution map of soil organic matter content in the undisturbed profile of the target soil column, the pixel-level black soil distribution map of the undisturbed profile of the target soil column is determined.

[0014] Based on the pixel-level black soil distribution map of the original profile of the target soil column, the black soil layer identification result of the original profile of the target soil column is determined.

[0015] Based on the black soil layer identification results and pixel-level black soil distribution map of the original profile of the target soil column, the thickness of the black soil layer in the original profile of the target soil column is determined.

[0016] In one embodiment, image preprocessing is performed on the target image to obtain a preprocessed target image, including:

[0017] The target image is then subjected to whiteboard correction to obtain the whiteboard-corrected target image.

[0018] The original profile of the non-target soil column in the target image after whiteboard correction is cropped to obtain the cropped target image;

[0019] The reflectance of each pixel in the cropped target image at multiple wavelengths in the preset band is input into the soil differentiation model to obtain the soil differentiation result of each pixel in the cropped target image; the soil differentiation result is either soil or non-soil.

[0020] In the cropped target image, the reflectance values ​​of all pixels that are not soil in the soil differentiation result are set to null, resulting in the preprocessed target image.

[0021] In one embodiment, the reflectance of the current pixel at each wavelength in the effective band is sequentially converted and smoothed to obtain the processed absorbance of the current pixel at each wavelength in the effective band, including:

[0022] The absorbance of the current pixel at each wavelength in the effective band is converted to absorbance at each wavelength in the effective band.

[0023] By using SG smoothing filter and first derivative transformation, the absorbance of the current pixel at each wavelength in the effective band is smoothed to obtain the processed absorbance of the current pixel at each wavelength in the effective band.

[0024] In one embodiment, the process of determining the soil organic matter content prediction model includes:

[0025] Obtain the reflectance of each pixel in the hyperspectral image of the undisturbed profile of each soil column in the soil imaging hyperspectral library at each wavelength in the effective band.

[0026] Any layer in the hyperspectral image of any original column profile is selected as the layer to be screened, and any layer in the preprocessed target image is selected as the target layer.

[0027] Based on the reflectance of all pixels in the layer to be screened at each wavelength in the effective band, the average reflectance of the layer to be screened at each wavelength in the effective band is determined. Based on the reflectance of all pixels in the target layer at each wavelength in the effective band, the average reflectance of the target layer at each wavelength in the effective band is determined.

[0028] Based on the average reflectance of the layer to be screened at each wavelength in the effective band and the average reflectance of the target layer at each wavelength in the effective band, the spectral similarity coefficient between the layer to be screened and the target layer is determined, and the spectral similarity coefficient between the layer to be screened and all layers in the preprocessed target image is determined as the judgment data of the layer to be screened.

[0029] Each layer to be screened whose data exceeds a preset threshold is determined as a training sample layer.

[0030] Based on the processed absorbance of all pixels in each training sample layer, the mean absorbance of the corresponding training sample layer is determined.

[0031] The actual values ​​of soil organic matter content in each training sample layer were obtained from the soil imaging hyperspectral library;

[0032] Initialize the spectral learning machine model;

[0033] The average absorbance of each training sample layer is used as input, and the actual soil organic matter content of the corresponding training sample layer is used as output to train the spectral learning machine model, thereby obtaining the soil organic matter content prediction model.

[0034] In one embodiment, based on the predicted values ​​and spatial coordinates of soil organic matter content at each pixel in the preprocessed target image, a pixel-level vertical distribution map of soil organic matter content in the undisturbed profile of the target soil column is determined, including:

[0035] The predicted values ​​of soil organic matter content of each pixel in each preprocessed target image are mapped to the corresponding spatial coordinates to generate a two-dimensional rasterized image, thus obtaining a pixel-level vertical distribution map of soil organic matter content in the original profile of the target soil column.

[0036] In one embodiment, based on the pixel-level vertical distribution map of soil organic matter content in the undisturbed profile of the target soil column, a pixel-level black soil distribution map of the undisturbed profile of the target soil column is determined, including:

[0037] In the vertical distribution map of soil organic matter content at the pixel level of the original profile of the target soil column, pixels with predicted soil organic matter content of 0 or missing value are identified as non-soil pixels, and other pixels are identified as soil pixels.

[0038] Soil pixels with predicted organic matter content greater than the soil organic matter content threshold are identified as black soil pixels, while soil pixels with predicted organic matter content less than the soil organic matter content threshold are identified as non-black soil pixels, thus obtaining a pixel-level black soil distribution map of the original profile of the target soil column.

[0039] In one embodiment, based on the pixel-level black soil distribution map of the undisturbed profile of the target soil column, the black soil layer identification result of the undisturbed profile of the target soil column is determined, including:

[0040] Determine the relative ratio of the number of black soil pixels to the number of soil pixels in each layer of the pixel-level black soil distribution map; each layer corresponds to a row of pixels in the pixel-level black soil distribution map.

[0041] Layers with a relative proportion greater than or equal to a preset proportion are identified as black soil layers, while layers with a relative proportion less than a preset proportion are identified as non-black soil layers, thus obtaining the black soil layer identification results of the original profile of the target soil column.

[0042] In one embodiment, the thickness of the black soil layer in the undisturbed profile of the target soil column is determined based on the black soil layer identification results and pixel-level black soil distribution map, including:

[0043] Following the order from the surface downwards, starting from the surface, the first layer among the first consecutive non-black soil layers of a predetermined number is determined as the termination layer;

[0044] The number of layers contained in the layer above the surface layer to the terminal layer is determined as the number of black soil layers in the original profile of the target soil column;

[0045] The ratio of black soil layers in the original profile of the target soil column to the total number of layers in the pixel-level black soil distribution map is calculated to obtain the proportion of black soil layers in the original profile of the target soil column.

[0046] The thickness of the black soil layer in the original profile of the target soil column is obtained by multiplying the proportion of the black soil layer in the original profile of the target soil column by the actual thickness of the original profile of the target soil column.

[0047] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for determining the thickness of the black soil layer.

[0048] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for determining the thickness of the black soil layer described above.

[0049] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0050] This application discloses a method, apparatus, and medium for determining the thickness of black soil layers, significantly improving the accuracy and efficiency of black soil layer thickness identification in black soil regions. First, imaging hyperspectral technology is used to rapidly and continuously scan soil column profiles to obtain sub-millimeter pixel-level hyperspectral images. After removing non-soil portions of the images, a high-precision vertical mapping of the organic matter content of the undisturbed soil profile is achieved based on a soil imaging hyperspectral library combined with spectral similarity analysis and a spectral learning machine algorithm. This not only significantly improves the efficiency and accuracy of soil information acquisition but also provides high-quality data support for in-depth understanding of the spatial variability and evolution of soil properties in the profile, making it possible to determine the thickness of black soil layers with high precision. Secondly, soil pixels with predicted organic matter content greater than the threshold (1%) are identified as black soil pixels, while those with content less than the threshold are identified as non-black soil pixels. This yields a pixel-level black soil distribution map of the original soil profile, enabling quantitative identification of black soil regions in the soil profile. This overcomes the subjectivity and reliance on expert experience inherent in traditional manual colorimetric card interpretation, and avoids the long cycle and high cost of traditional laboratory soil organic matter determination methods. Thirdly, by statistically analyzing whether the proportion of black soil pixels in the vertical profile exceeds a set threshold, the black soil layer interface is scientifically determined, thereby accurately calculating the black soil layer thickness. This significantly improves the accuracy, objectivity, and repeatability of black soil layer thickness measurement results, providing an efficient and reliable technical means for black soil resource investigation and monitoring. This application not only avoids the problems of low efficiency and strong subjectivity in black soil layer thickness measurement using traditional colorimetric card methods, but also achieves rapid and accurate measurement of black soil layer thickness, improving the efficiency and accuracy of black soil layer thickness measurement. This provides high-quality data support for in-depth understanding of the spatial variability and evolution of soil properties in the profile. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 A flowchart illustrating a method for determining the thickness of a black soil layer according to an embodiment of this application.

[0053] Figure 2 This is a pixel-level vertical distribution map of soil organic matter content.

[0054] Figure 3 This is a schematic diagram of the first example of the black soil layer thickness identification results.

[0055] Figure 4 This is a schematic diagram of the second example of the black soil layer thickness identification results.

[0056] Figure 5 This is a schematic diagram of the third example of the black soil layer thickness identification results.

[0057] Figure 6 This is a schematic diagram of the fourth example of the black soil layer thickness identification results.

[0058] Figure 7 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0059] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0060] The purpose of this application is to provide a method, apparatus, and medium for determining the thickness of black soil layers, aiming to improve the accuracy of black soil layer thickness determination.

[0061] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] In one exemplary embodiment, such as Figure 1 As shown, a method for determining the thickness of a black soil layer is provided, including the following steps.

[0063] Step 1: Acquire the target image.

[0064] The target image is a hyperspectral image containing the original profile of the target soil column; the spectral values ​​of each pixel in the hyperspectral image include the reflectance and spatial coordinates of the pixel at multiple wavelengths in the preset band.

[0065] Specifically, soil columns of 0-100 cm were collected at designated sampling points in the study area using a soil auger, resulting in target soil columns. The collected target soil columns were placed in polyvinyl chloride (PVC) tubes and sealed with plastic film to prevent moisture evaporation and external interference, thus preserving the original state of the soil. The collected target soil columns were vertically bisected from the center of the top end to the bottom end. In a darkroom environment, a hyperspectral push-broom polarization analyzer (Hyperimager Plant Phenomics Analysis, HPPA) was used to scan the original profile of the bisected target soil columns, obtaining hyperspectral image data (i.e., target images) within the 375.5 nm–1017.3 nm band (i.e., the preset band).

[0066] Step 2: Perform image preprocessing on the target image to obtain the preprocessed target image.

[0067] As an optional implementation, step 2 includes steps 21-24.

[0068] Step 21: Perform whiteboard correction on the target image to obtain the whiteboard-corrected target image.

[0069] Step 22: Crop the non-target soil column original profile portion in the target image after whiteboard correction to obtain the cropped target image.

[0070] Specifically, when acquiring the target image, it may include not only the original profile of the target soil column, but also non-original profile parts such as the light absorption test platform, the calibration whiteboard, and the PVC pipe. Therefore, the non-original profile parts of the target image after whiteboard calibration are cropped to obtain the cropped target image, so that the cropped target image only contains the original profile of the target soil column.

[0071] Step 23: Input the reflectance of each pixel in the cropped target image at multiple wavelengths in the preset band into the soil differentiation model to obtain the soil differentiation result of each pixel in the cropped target image; the soil differentiation result is soil or non-soil.

[0072] Specifically, the soil differentiation model is obtained by training a random forest model. The process of determining the soil differentiation model includes: 1) acquiring multiple hyperspectral images containing soil; 2) selecting multiple regions of interest from each hyperspectral image containing soil; 3) using the actual soil partitioning results of each region of interest as training samples to train the random forest model to obtain the soil differentiation model.

[0073] Step 24: Set the reflectance of all pixels in the cropped target image that are not soil to null to obtain the preprocessed target image.

[0074] Step 3: Determine any pixel in the preprocessed target image as the current pixel, and determine the reflectance of the current pixel at each wavelength in the effective band.

[0075] Step 4: Perform absorbance conversion and smoothing on the reflectance of the current pixel at each wavelength in the effective band to obtain the processed absorbance of the current pixel at each wavelength in the effective band.

[0076] As an optional implementation, the effective band is selected by filtering the preset bands.

[0077] Specifically, the effective wavelength range is 434nm–927nm.

[0078] As an optional implementation, step 4 includes steps 41-42.

[0079] Step 41: Convert the reflectance of the current pixel at each wavelength in the effective band to obtain the absorbance of the current pixel at each wavelength in the effective band.

[0080] Step 42: Using SG smoothing filter and first derivative transformation, smooth the absorbance of the current pixel at each wavelength in the effective band to obtain the processed absorbance of the current pixel at each wavelength in the effective band.

[0081] Specifically, the absorbance is smoothed, which effectively eliminates baseline drift and background noise, and enhances the stability and recognizability of the absorbance.

[0082] Step 5: Input the processed absorbance of the current pixel at each wavelength in the effective band into the soil organic matter content prediction model to obtain the predicted value of the soil organic matter content of the current pixel.

[0083] The soil organic matter content prediction model is determined based on a spectral learning machine model.

[0084] As an optional implementation, step 5, the process of determining the soil organic matter content prediction model, includes steps 51-59.

[0085] Step 51: Obtain the reflectance of each pixel in the hyperspectral image of the undisturbed profile of each soil column in the soil imaging hyperspectral library at each wavelength in the effective band.

[0086] Step 52: Determine any layer in the hyperspectral image of any original column profile as the layer to be screened, and determine any layer in the preprocessed target image as the target layer.

[0087] Step 53: Based on the reflectance of all pixels in the layer to be screened at each wavelength in the effective band, determine the average reflectance of the layer to be screened at each wavelength in the effective band. Based on the reflectance of all pixels in the target layer at each wavelength in the effective band, determine the average reflectance of the target layer at each wavelength in the effective band.

[0088] Step 54: Based on the average reflectance of the layer to be screened at each wavelength in the effective band and the average reflectance of the target layer at each wavelength in the effective band, determine the spectral similarity coefficient between the layer to be screened and the target layer, and use the spectral similarity coefficient between the layer to be screened and all layers in the preprocessed target image as the judgment data for the layer to be screened.

[0089] Step 55: Determine each layer to be screened whose data is greater than the preset threshold as a training sample layer.

[0090] The preset threshold is 0.8.

[0091] Step 56: Determine the average absorbance of the corresponding training sample layer based on the processed absorbance of all pixels in each training sample layer.

[0092] Step 57: Obtain the actual values ​​of soil organic matter content for each training sample layer from the soil imaging hyperspectral library.

[0093] Step 58: Initialize the spectral learning machine model.

[0094] Step 59: Using the average absorbance of each training sample layer as input and the actual soil organic matter content of the corresponding training sample layer as output, train the spectral learning machine model to obtain the soil organic matter content prediction model.

[0095] Calculate the coefficients of determination for the modeling set and the validation set respectively. Root mean square error and relative percentage difference The specific formula for evaluating model accuracy is as follows.

[0096] .

[0097] .

[0098] .

[0099] in, This represents the actual value of soil organic matter content in the i-th training sample layer of the validation set. This represents the predicted value of soil organic matter content in the i-th training sample layer of the validation set. This represents the average of the actual soil organic matter content values ​​across all training sample layers in the validation set. This indicates the number of training sample layers in the validation set.

[0100] Step 6: Based on the predicted values ​​and spatial coordinates of soil organic matter content of each pixel in the preprocessed target image, determine the pixel-level vertical distribution map of soil organic matter content in the original profile of the target soil column.

[0101] As an optional implementation, step 6 includes:

[0102] The predicted values ​​of soil organic matter content of each pixel in each preprocessed target image are mapped to the corresponding spatial coordinates to generate a two-dimensional rasterized image, thus obtaining a pixel-level vertical distribution map of soil organic matter content in the original profile of the target soil column.

[0103] Specifically, the pixel-level vertical distribution map of soil organic matter content is as follows: Figure 2 As shown. Figure 2 In this context, SOM represents soil organic matter content.

[0104] Step 7: Based on the pixel-level vertical distribution map of soil organic matter content in the undisturbed profile of the target soil column, determine the pixel-level black soil distribution map of the undisturbed profile of the target soil column.

[0105] As an optional implementation, step 7 includes steps 71-72.

[0106] Step 71: In the vertical distribution map of soil organic matter content in the undisturbed profile of the target soil column, the pixels with a predicted value of 0 or a missing value of soil organic matter content are identified as non-soil pixels, and the other pixels are identified as soil pixels.

[0107] Specifically, in step 24, all reflectance values ​​corresponding to pixels in the cropped target image that are not soil are set to null (only the spatial coordinates of each pixel are retained). When using the soil organic matter content prediction model for prediction, the predicted soil organic matter content of each pixel that is not soil will be assigned 0 or a missing value.

[0108] Step 72: Soil pixels with predicted soil organic matter content greater than the soil organic matter content threshold are identified as black soil pixels, and soil pixels with predicted soil organic matter content less than the soil organic matter content threshold are identified as non-black soil pixels, thereby obtaining a pixel-level black soil distribution map of the original profile of the target soil column.

[0109] Specifically, the threshold for soil organic matter content is 1%.

[0110] Step 8: Based on the pixel-level black soil distribution map of the original profile of the target soil column, determine the black soil layer identification result of the original profile of the target soil column.

[0111] As an optional implementation, step 8 includes steps 81-82.

[0112] Step 81: Determine the relative ratio of the number of black soil pixels to the number of soil pixels in each layer of the pixel-level black soil distribution map; each layer corresponds to a row of pixels in the pixel-level black soil distribution map.

[0113] Step 82: Identify the layers with a relative proportion greater than or equal to the preset proportion as black soil layers, and identify the layers with a relative proportion less than the preset proportion as non-black soil layers, thereby obtaining the black soil layer identification results of the original profile of the target soil column.

[0114] Specifically, the preset ratio is 90%. Determining the black soil layer identification result through a relative ratio can effectively reduce false positives caused by a single abnormal pixel, and improve the stability and spatial continuity of profile layer identification. Black soil layer identification was performed on soil column profiles of multiple examples, and the black soil layer identification results are as follows: Figures 3-6 As shown. Figures 3-6 The dashed line indicates the location of the fault tail.

[0115] Step 9: Based on the black soil layer identification results and pixel-level black soil distribution map of the original profile of the target soil column, determine the thickness of the black soil layer in the original profile of the target soil column.

[0116] As an optional implementation, step 9 includes steps 91-94.

[0117] Step 91: Following the order from the surface downwards, starting from the surface, determine the first layer among the first consecutive non-black soil layers of a predetermined number as the termination layer.

[0118] Step 92: Determine the number of black soil layers in the original profile of the target soil column, including the layer above the surface layer to the termination layer.

[0119] Specifically, for example, when, starting from the surface, there are 30 consecutive layers from layer 41 to layer 70 that are all "non-black soil layers", then layer 41 is considered the termination layer, and the number of black soil layers is 40.

[0120] Step 93: Calculate the ratio between the number of black soil layers in the undisturbed profile of the target soil column and the total number of layers in the pixel-level black soil distribution map to obtain the proportion of black soil layers in the undisturbed profile of the target soil column.

[0121] Step 94: Multiply the proportion of the black soil layer in the original profile of the target soil column by the actual thickness of the original profile of the target soil column to obtain the thickness of the black soil layer in the original profile of the target soil column.

[0122] Furthermore, to verify the accuracy of the black soil layer thickness determination method of this application, the black soil layer thickness of each pixel is determined based on the reflectance of multiple pixels using the method of this application. The black soil layer thickness of the pixels identified by the colorimetric card is used as a reference, and the determination coefficient between the black soil layer thickness obtained by the method of this application and the black soil layer thickness identified by the colorimetric card is calculated respectively. With root mean square error The results are shown in Table 1.

[0123] Table 1. Coefficient of Determination and Root Mean Square Error

[0124]

[0125] The results show that the method in this application has high accuracy in determining the thickness of the black soil layer, providing important technical support for the digital characterization of profile attributes and the formulation of protection strategies in the black soil region.

[0126] In one exemplary embodiment, a computer device is provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for determining the thickness of a black soil layer.

[0127] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements a method for determining the thickness of a black soil layer.

[0128] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 7 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network. When the computer program is executed by the processor, it implements a method for determining the thickness of a black soil layer.

[0129] Those skilled in the art will understand that Figure 7The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0130] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0131] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0132] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0133] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0134] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for determining the thickness of a black soil layer, characterized in that, The method for determining the thickness of the black soil layer includes: Acquire a target image; the target image is a hyperspectral image containing the original profile of the target soil column; the spectral value of each pixel in the hyperspectral image includes the reflectance and spatial coordinates of the pixel at multiple wavelengths in a preset band; The target image is preprocessed to obtain a preprocessed target image; Any pixel in the preprocessed target image is identified as the current pixel, and the reflectance of the current pixel at each wavelength in the effective band is determined. The reflectance of the current pixel at each wavelength in the effective band is sequentially converted and smoothed to obtain the processed absorbance of the current pixel at each wavelength in the effective band. The processed absorbance of the current pixel at each wavelength in the effective band is input into the soil organic matter content prediction model to obtain the predicted value of the soil organic matter content of the current pixel; the soil organic matter content prediction model is determined based on the spectral learning machine model. Based on the predicted values ​​and spatial coordinates of soil organic matter content of each pixel in the preprocessed target image, a pixel-level vertical distribution map of soil organic matter content in the original profile of the target soil column is determined. Based on the pixel-level vertical distribution map of soil organic matter content in the undisturbed profile of the target soil column, the pixel-level black soil distribution map of the undisturbed profile of the target soil column is determined. Based on the pixel-level black soil distribution map of the original profile of the target soil column, the black soil layer identification result of the original profile of the target soil column is determined. Based on the black soil layer identification results and pixel-level black soil distribution map of the original profile of the target soil column, the thickness of the black soil layer in the original profile of the target soil column is determined.

2. The method for determining the thickness of the black soil layer according to claim 1, characterized in that, The target image is preprocessed to obtain a preprocessed target image, including: The target image is then subjected to whiteboard correction to obtain the whiteboard-corrected target image. The original profile of the non-target soil column in the target image after whiteboard correction is cropped to obtain the cropped target image; The reflectance of each pixel in the cropped target image at multiple wavelengths in the preset band is input into the soil differentiation model to obtain the soil differentiation result of each pixel in the cropped target image; the soil differentiation result is either soil or non-soil. In the cropped target image, the reflectance values ​​of all pixels that are not soil in the soil differentiation result are set to null, resulting in the preprocessed target image.

3. The method for determining the thickness of the black soil layer according to claim 1, characterized in that, The reflectance of the current pixel at each wavelength in the effective band is sequentially converted and smoothed to obtain the processed absorbance of the current pixel at each wavelength in the effective band, including: The absorbance of the current pixel at each wavelength in the effective band is converted to absorbance at each wavelength in the effective band. By using SG smoothing filter and first derivative transformation, the absorbance of the current pixel at each wavelength in the effective band is smoothed to obtain the processed absorbance of the current pixel at each wavelength in the effective band.

4. The method for determining the thickness of the black soil layer according to claim 1, characterized in that, The process of determining the soil organic matter content prediction model includes: Obtain the reflectance of each pixel in the hyperspectral image of the undisturbed profile of each soil column in the soil imaging hyperspectral library at each wavelength in the effective band. Any layer in the hyperspectral image of any original column profile is selected as the layer to be screened, and any layer in the preprocessed target image is selected as the target layer. Based on the reflectance of all pixels in the layer to be screened at each wavelength in the effective band, the average reflectance of the layer to be screened at each wavelength in the effective band is determined. Based on the reflectance of all pixels in the target layer at each wavelength in the effective band, the average reflectance of the target layer at each wavelength in the effective band is determined. Based on the average reflectance of the layer to be screened at each wavelength in the effective band and the average reflectance of the target layer at each wavelength in the effective band, the spectral similarity coefficient between the layer to be screened and the target layer is determined, and the spectral similarity coefficient between the layer to be screened and all layers in the preprocessed target image is determined as the judgment data of the layer to be screened. Each layer to be screened whose data exceeds a preset threshold is determined as a training sample layer. Based on the processed absorbance of all pixels in each training sample layer, the mean absorbance of the corresponding training sample layer is determined. The actual values ​​of soil organic matter content in each training sample layer were obtained from the soil imaging hyperspectral library; Initialize the spectral learning machine model; The average absorbance of each training sample layer is used as input, and the actual soil organic matter content of the corresponding training sample layer is used as output to train the spectral learning machine model, thereby obtaining the soil organic matter content prediction model.

5. The method for determining the thickness of the black soil layer according to claim 1, characterized in that, Based on the predicted values ​​and spatial coordinates of soil organic matter content at each pixel in the preprocessed target image, a pixel-level vertical distribution map of soil organic matter content in the undisturbed profile of the target soil column is determined, including: The predicted values ​​of soil organic matter content of each pixel in each preprocessed target image are mapped to the corresponding spatial coordinates to generate a two-dimensional rasterized image, thus obtaining a pixel-level vertical distribution map of soil organic matter content in the original profile of the target soil column.

6. The method for determining the thickness of the black soil layer according to claim 1, characterized in that, Based on the pixel-level vertical distribution map of soil organic matter content in the undisturbed profile of the target soil column, a pixel-level black soil distribution map of the undisturbed profile of the target soil column is determined, including: In the vertical distribution map of soil organic matter content at the pixel level of the original profile of the target soil column, pixels with predicted soil organic matter content of 0 or missing value are identified as non-soil pixels, and other pixels are identified as soil pixels. Soil pixels with predicted organic matter content greater than the soil organic matter content threshold are identified as black soil pixels, while soil pixels with predicted organic matter content less than the soil organic matter content threshold are identified as non-black soil pixels, thus obtaining a pixel-level black soil distribution map of the original profile of the target soil column.

7. The method for determining the thickness of the black soil layer according to claim 1, characterized in that, Based on the pixel-level black soil distribution map of the undisturbed profile of the target soil column, the black soil layer identification results of the undisturbed profile of the target soil column are determined, including: Determine the relative ratio of the number of black soil pixels to the number of soil pixels in each layer of the pixel-level black soil distribution map; each layer corresponds to a row of pixels in the pixel-level black soil distribution map. Layers with a relative proportion greater than or equal to a preset proportion are identified as black soil layers, while layers with a relative proportion less than a preset proportion are identified as non-black soil layers, thus obtaining the black soil layer identification results of the original profile of the target soil column.

8. The method for determining the thickness of the black soil layer according to claim 1, characterized in that, Based on the black soil layer identification results and pixel-level black soil distribution map of the undisturbed profile of the target soil column, the thickness of the black soil layer in the undisturbed profile of the target soil column is determined, including: Following the order from the surface downwards, starting from the surface, the first layer among the first consecutive non-black soil layers of a predetermined number is determined as the termination layer; The number of layers contained in the layer above the surface layer to the terminal layer is determined as the number of black soil layers in the original profile of the target soil column; The ratio of black soil layers in the original profile of the target soil column to the total number of layers in the pixel-level black soil distribution map is calculated to obtain the proportion of black soil layers in the original profile of the target soil column. The thickness of the black soil layer in the original profile of the target soil column is obtained by multiplying the proportion of the black soil layer in the original profile of the target soil column by the actual thickness of the original profile of the target soil column.

9. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the method for determining the thickness of the black soil layer as described in any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the method for determining the thickness of the black soil layer as described in any one of claims 1-8.

Images