Crop soil moisture content remote sensing inversion method, device and system, and storage medium

Abstract

The invention discloses a crop soil moisture content remote sensing inversion method, device and system, and a storage medium. The method comprises the following steps: S1, obtaining an unmanned aerial vehicle multispectral remote sensing image of crops in a predetermined area; s2, removing a soil background and an image shadow from the multi-spectral remote sensing image of the unmanned aerial vehicle; s3, training a back propagation neural network model and a partial least square regression model according to the unmanned aerial vehicle multispectral remote sensing image after the soil background and the image shadow are removed to obtain an inversion model; and S4, inputting the unmanned aerial vehicle multispectral remote sensing image of the crops in the target area into the inversion model, and performing inversion of the soil moisture content of the crops in each growth period. By adopting the technical scheme provided by the invention, the crop soil water content can be accurately inverted.

Description

Technical Field

[0001] The invention belongs to the technical field of remote sensing inversion, and in particular relates to a remote sensing inversion method and device, system and storage medium for crop soil moisture content. Background Art

[0002] As one of the important indicators of agricultural monitoring, soil moisture content plays a guiding role in making decisions on precise irrigation, evaluating crop growth conditions and estimating crop yields. Traditional soil moisture content monitoring methods include drying method, gamma ray method, neutron instrument measurement method, etc. Although such methods have high single-point monitoring accuracy, they are time-consuming and labor-intensive, and cannot provide continuous soil moisture content monitoring data in real time, non-contact, and non-destructively. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a remote sensing inversion method and device, system and storage medium for crop soil moisture content, which can accurately invert the crop soil moisture content.

[0004] To achieve the above object, the present invention adopts the following technical solution:

[0005] A remote sensing inversion method for crop soil moisture content, comprising:

[0006] Step S1, obtaining a multispectral remote sensing image of crops in a predetermined area using an unmanned aerial vehicle;

[0007] Step S2, removing soil background and image shadows from the multispectral remote sensing image of the UAV;

[0008] Step S3, training a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing image of the UAV after removing the soil background and image shadow, to obtain an inversion model;

[0009] Step S4: input the multispectral remote sensing image of the crops in the target area from the UAV into the inversion model to invert the soil moisture content of the crops at each growth stage.

[0010] Preferably, the crop is silage corn.

[0011] Preferably, in step S2, soil background and image shadows are removed from the multispectral remote sensing image of the UAV by using the maximum likelihood method in supervised classification.

[0012] Preferably, step S3 comprises:

[0013] Extract the canopy reflectance of silage corn based on the multispectral remote sensing images of UAVs after removing the soil background and image shadows;

[0014] According to the canopy reflectance of silage corn, the spectral index sensitive to soil moisture content at different growth stages of silage corn was obtained;

[0015] Based on the UAV multispectral remote sensing images after removing the soil background and image shadows, the canopy reflectance and spectral index of silage corn, the back propagation neural network model and the partial least squares regression model were trained respectively, and the first inversion model for the silage corn tasseling period and the second inversion model for the silage corn seedling and jointing stages were obtained.

[0016] The present invention also provides a remote sensing inversion device for crop soil moisture content, comprising:

[0017] An acquisition module is used to acquire UAV multispectral remote sensing images of crops in a predetermined area;

[0018] The preprocessing module is used to remove soil background and image shadows from multispectral remote sensing images of UAVs;

[0019] A training module is used to train a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing images of the UAV after removing the soil background and image shadows, so as to obtain an inversion model;

[0020] The inversion module is used to input the multispectral remote sensing images of crops in the target area from UAVs into the inversion model to invert the soil moisture content at each growth stage of the crops.

[0021] Preferably, the crop is silage corn.

[0022] Preferably, the preprocessing module removes soil background and image shadows from the UAV multispectral remote sensing image through the maximum likelihood method in supervised classification.

[0023] Preferably, the training module includes:

[0024] An extraction unit, used for extracting the canopy reflectance of silage corn based on the multispectral remote sensing image of the UAV after removing the soil background and image shadow;

[0025] A processing unit, for obtaining a spectral index sensitive to soil moisture content at different growth stages of silage corn according to the canopy reflectance of silage corn;

[0026] The training unit is used to train a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing images of the unmanned aerial vehicle after removing the soil background and image shadows, the reflectance of the silage corn canopy and the spectral index, respectively, to obtain the first inversion model for the silage corn tasseling period and the second inversion model for the silage corn seedling and jointing periods.

[0027] The present invention also provides a crop soil moisture remote sensing inversion system, comprising: a memory and a processor, wherein the memory stores a computer program run by the processor, and the computer program executes the crop soil moisture remote sensing inversion method when run by the processor.

[0028] The present invention also provides a storage medium, on which a computer program is stored, and when the computer program is run, a remote sensing inversion method for crop soil moisture content is executed.

[0029] The present invention removes soil background and image shadows from multispectral remote sensing images of unmanned aerial vehicles; trains a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing images of unmanned aerial vehicles after removing the soil background and image shadows to obtain an inversion model; inputs the multispectral remote sensing images of unmanned aerial vehicles of crops in the target area into the inversion model to invert the soil moisture content of crops in various growth stages. The technical solution of the present invention can accurately invert the soil moisture content of crops. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0031] Figure 1 The present invention is a flowchart of a method for remote sensing inversion of soil moisture content of crops according to an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0034] Embodiment 1:

[0035] like Figure 1 As shown, an embodiment of the present invention provides a remote sensing inversion method for crop soil moisture content, comprising:

[0036] Step S1, obtaining a multispectral remote sensing image of crops in a predetermined area using an unmanned aerial vehicle;

[0037] Step S2, removing soil background and image shadows from the multispectral remote sensing image of the UAV;

[0038] Step S3, training a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing image of the UAV after removing the soil background and image shadow, to obtain an inversion model;

[0039] Step S4: input the multispectral remote sensing image of the crops in the target area from the UAV into the inversion model to invert the soil moisture content of the crops at each growth stage.

[0040] As an implementation manner of the embodiment of the present invention, the crop is silage corn.

[0041] As an implementation method of an embodiment of the present invention, in step S2, soil background and image shadows are removed from the multispectral remote sensing image of the UAV by using the maximum likelihood method in supervised classification. The maximum likelihood method in supervised classification is a nonlinear classification method with the minimum probability of classification error based on the Bayesian criterion. Specifically, it includes:

[0042] Step S21, dividing the soil, shadow, and corn regions of interest by synthesizing the false color image of the multispectral remote sensing image of the UAV;

[0043] Step S22, determining the degree of separation by calculating the type of the divided region of interest: the value range is 0-2, if it is less than 1.8, the degree of separation is not enough, and the region of interest needs to be reselected;

[0044] Step S23, if the separation degree is greater than 1.8, then input the false color image and use the maximum likelihood method for classification;

[0045] Step S24, using Majority analysis to eliminate small patches in the classification result, so that the classification result image becomes smooth;

[0046] Step S25: Generate a confusion matrix using the classification results after eliminating small patches, and perform accuracy verification using the Kappa coefficient calculated using the confusion matrix.

[0047] As an implementation of the embodiment of the present invention, step S3 includes:

[0048] Step S31, extracting the canopy reflectance of silage corn based on the multispectral remote sensing image of the UAV after removing the soil background and image shadow;

[0049] Step S32, obtaining spectral indices sensitive to soil moisture content at different growth stages of silage corn according to the silage corn canopy reflectance;

[0050] Step S33, based on the UAV multispectral remote sensing image after removing the soil background and image shadow, the canopy reflectance and spectral index of the silage corn, respectively train the back propagation neural network model and the partial least squares regression model to obtain the first inversion model for the silage corn tasseling period and the second inversion model for the silage corn seedling and jointing stages.

[0051] Further, step S31 includes:

[0052] Step S311, performing raster-to-surface conversion on the crop band of the multispectral remote sensing image of the UAV after removing the soil background and image shadow, to produce a mask vector file;

[0053] Step S312, extracting the canopy images of each band after removing the soil background and image shadows through the vector file mask;

[0054] Step S313: adding field sampling points to the canopy images of each band, performing multi-value extraction to point processing, and obtaining the canopy reflectance of each band at each sampling point.

[0055] Furthermore, in step 32, the UAV multispectral image data has the characteristics of complex combinations, and redundant and interfering spectral variables will affect the precision and accuracy of the soil moisture inversion model, and thus it is necessary to screen the sensitive spectral index combination. The Pearson correlation analysis method is used to evaluate the strength of the linear relationship between the spectral index and the soil moisture content, and the calculation formula is as follows:

[0056]

[0057] Among them, X i and Y i They represent the sample values ​​of the multispectral remote sensing image of the UAV and the sample values ​​of the soil moisture content in the root zone of silage corn, respectively. and They represent the sample means of the two, respectively, and n is the number of samples. The value range of r is [-1, 1]. A value of 0.8-1.0 indicates a very strong correlation, 0.6-0.8 indicates a strong correlation, 0.4-0.6 indicates a moderate correlation, 0.2-0.4 indicates a weak correlation, and 0.0-0.2 indicates a very weak correlation or no correlation.

[0058] Embodiment 2:

[0059] The embodiment of the present invention further provides a remote sensing inversion device for crop soil moisture content, comprising:

[0060] An acquisition module is used to acquire UAV multispectral remote sensing images of crops in a predetermined area;

[0061] The preprocessing module is used to remove soil background and image shadows from multispectral remote sensing images of UAVs;

[0062] A training module is used to train a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing images of the UAV after removing the soil background and image shadows, so as to obtain an inversion model;

[0063] The inversion module is used to input the multispectral remote sensing images of crops in the target area from UAVs into the inversion model to invert the soil moisture content at each growth stage of the crops.

[0064] As an implementation manner of the embodiment of the present invention, the crop is silage corn.

[0065] As an implementation method of the embodiment of the present invention, the preprocessing module removes soil background and image shadows from the multi-spectral remote sensing image of the unmanned aerial vehicle through the maximum likelihood method in supervised classification.

[0066] As an implementation of an embodiment of the present invention, the training module includes:

[0067] An extraction unit, used for extracting the canopy reflectance of silage corn based on the multispectral remote sensing image of the UAV after removing the soil background and image shadow;

[0068] A processing unit, for obtaining a spectral index sensitive to soil moisture content at different growth stages of silage corn according to the canopy reflectance of silage corn;

[0069] The training unit is used to train a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing images of the unmanned aerial vehicle after removing the soil background and image shadows, the reflectance of the silage corn canopy and the spectral index, respectively, to obtain the first inversion model for the silage corn tasseling period and the second inversion model for the silage corn seedling and jointing periods.

[0070] Embodiment 3:

[0071] An embodiment of the present invention also provides a crop soil moisture remote sensing inversion system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a crop soil moisture remote sensing inversion method when executed by the processor.

[0072] Embodiment 4:

[0073] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and when the computer program is run, a remote sensing inversion method for crop soil moisture content is executed.

[0074] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A remote sensing inversion method for crop soil moisture content, It is characterized in that include: Step S1, obtaining a multispectral remote sensing image of crops in a predetermined area using an unmanned aerial vehicle; Step S2, removing soil background and image shadows from the multispectral remote sensing image of the UAV; Step S3, training a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing image of the UAV after removing the soil background and image shadow, to obtain an inversion model; Step S4: input the multispectral remote sensing image of the crops in the target area from the UAV into the inversion model to invert the soil moisture content of the crops at each growth stage.

2. The remote sensing inversion method for crop soil moisture content according to claim 1, It is characterized in that The crop is silage corn.

3. The remote sensing inversion method for crop soil moisture content according to claim 2, It is characterized in that In step S2, the soil background and image shadows are removed from the multispectral remote sensing image of the UAV by using the maximum likelihood method in supervised classification.

4. The remote sensing inversion method for crop soil moisture content according to claim 3, It is characterized in that Step S3 includes: Extract the canopy reflectance of silage corn based on the multispectral remote sensing images of UAVs after removing the soil background and image shadows; According to the canopy reflectance of silage corn, the spectral index sensitive to soil moisture content at different growth stages of silage corn was obtained; Based on the UAV multispectral remote sensing images after removing the soil background and image shadows, the canopy reflectance and spectral index of silage corn, the back propagation neural network model and the partial least squares regression model were trained respectively, and the first inversion model for the silage corn tasseling period and the second inversion model for the silage corn seedling and jointing stages were obtained.

5. A remote sensing inversion device for crop soil moisture content, It is characterized in that include: An acquisition module is used to acquire UAV multispectral remote sensing images of crops in a predetermined area; The preprocessing module is used to remove soil background and image shadows from multispectral remote sensing images of UAVs; A training module is used to train a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing images of the UAV after removing the soil background and image shadows, so as to obtain an inversion model; The inversion module is used to input the multispectral remote sensing images of crops in the target area from UAVs into the inversion model to invert the soil moisture content at each growth stage of the crops.

6. The crop soil moisture remote sensing inversion device according to claim 5, It is characterized in that The crop is silage corn.

7. The crop soil moisture remote sensing inversion device according to claim 6, It is characterized in that The preprocessing module removes soil background and image shadows from UAV multispectral remote sensing images through the maximum likelihood method in supervised classification.

8. The crop soil moisture remote sensing inversion device according to claim 7, It is characterized in that The training modules include: An extraction unit, used for extracting the canopy reflectance of silage corn based on the multispectral remote sensing image of the UAV after removing the soil background and image shadow; A processing unit, for obtaining a spectral index sensitive to soil moisture content at different growth stages of silage corn according to the canopy reflectance of silage corn; The training unit is used to train a back propagation neural network model and a partial least squares regression model based on the multispectral remote sensing images of the unmanned aerial vehicle after removing the soil background and image shadows, the reflectance of the silage corn canopy and the spectral index, respectively, to obtain the first inversion model for the silage corn tasseling period and the second inversion model for the silage corn seedling and jointing periods.

9. A remote sensing inversion system for crop soil moisture content, It is characterized in that include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the remote sensing inversion method for crop soil moisture content as described in any one of claims 1 to 4 is executed.

10. A storage medium, It is characterized in that The storage medium stores a computer program, which, when running, executes the remote sensing inversion method for crop soil moisture content according to any one of claims 1 to 4.

Images