CN116879297B - Soil moisture collaborative inversion method, device, equipment and medium - Google Patents
Soil moisture collaborative inversion method, device, equipment and medium Download PDFInfo
- Publication number
- CN116879297B CN116879297B CN202311147561.XA CN202311147561A CN116879297B CN 116879297 B CN116879297 B CN 116879297B CN 202311147561 A CN202311147561 A CN 202311147561A CN 116879297 B CN116879297 B CN 116879297B
- Authority
- CN
- China
- Prior art keywords
- soil moisture
- inversion
- data
- remote sensing
- result
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000002689 soil Substances 0.000 title claims abstract description 259
- 238000000034 method Methods 0.000 title claims abstract description 77
- 230000003287 optical effect Effects 0.000 claims abstract description 59
- 238000012544 monitoring process Methods 0.000 claims abstract description 36
- 230000004927 fusion Effects 0.000 claims description 42
- 238000002310 reflectometry Methods 0.000 claims description 40
- 238000011176 pooling Methods 0.000 claims description 39
- 230000015654 memory Effects 0.000 claims description 29
- 230000006870 function Effects 0.000 claims description 22
- 238000012549 training Methods 0.000 claims description 12
- 238000012937 correction Methods 0.000 claims description 8
- 238000005259 measurement Methods 0.000 claims description 8
- 230000004913 activation Effects 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 238000010801 machine learning Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 9
- 238000010586 diagram Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 5
- 238000013527 convolutional neural network Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 4
- 238000001228 spectrum Methods 0.000 description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- 208000027066 STING-associated vasculopathy with onset in infancy Diseases 0.000 description 3
- 239000000443 aerosol Substances 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000001956 neutron scattering Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 238000012876 topography Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 230000007787 long-term memory Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000007500 overflow downdraw method Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000000704 physical effect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 239000002028 Biomass Substances 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000003750 conditioning effect Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 125000004435 hydrogen atom Chemical group [H]* 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000005416 organic matter Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 150000003839 salts Chemical class 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N22/00—Investigating or analysing materials by the use of microwaves or radio waves, i.e. electromagnetic waves with a wavelength of one millimetre or more
- G01N22/04—Investigating moisture content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
- G06V10/809—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of classification results, e.g. where the classifiers operate on the same input data
- G06V10/811—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of classification results, e.g. where the classifiers operate on the same input data the classifiers operating on different input data, e.g. multi-modal recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pathology (AREA)
- Chemical & Material Sciences (AREA)
- Multimedia (AREA)
- Databases & Information Systems (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Immunology (AREA)
- Biophysics (AREA)
- Analytical Chemistry (AREA)
- Mathematical Physics (AREA)
- Biochemistry (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Electromagnetism (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The application provides a soil moisture collaborative inversion method, a device, equipment and a medium, and relates to the technical field of remote sensing monitoring, wherein the method comprises the following steps: acquiring optical satellite remote sensing data and long-time-sequence microwave satellite remote sensing data; extracting a remote sensing index corresponding to the optical satellite remote sensing data; inputting the remote sensing index into a pre-trained soil moisture inversion model of the optical satellite data to obtain a first soil moisture inversion result; inputting the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result; and fusing the two results, and correcting the fused result through ground monitoring data to obtain a target soil moisture inversion result. The application can obtain better inversion effect in different areas and different environments, improves the precision, stability and resolution of soil moisture inversion, and reduces soil moisture inversion errors.
Description
Technical Field
The application relates to the technical field of remote sensing monitoring, in particular to a soil moisture collaborative inversion method, a device, equipment and a medium.
Background
The soil moisture monitoring has important significance for agricultural management, crop growth monitoring and the like. At present, the soil moisture monitoring technology is mainly divided into two main types, namely direct and indirect, wherein a direct method is used for collecting soil samples in the field and analyzing and determining the moisture content of soil in a laboratory, but the method is only suitable for collecting soil moisture information in small areas, and is difficult to realize the monitoring of soil moisture in large areas; the indirect method comprises a resistance method, a neutron scattering method, a Time Domain Reflectometer (TDR) method, a Frequency Domain Reflectometer (FDR) method, a remote sensing method and the like, and in the related technology, the remote sensing method is generally influenced by various factors such as atmospheric conditions, terrains, surface temperatures and the like, so that the soil moisture inversion accuracy is lower.
Disclosure of Invention
The application aims to provide a soil moisture collaborative inversion method, device, equipment and medium, which can obtain good inversion effects in different areas and different environments, improve the accuracy, stability and resolution of soil moisture inversion and reduce soil moisture inversion errors.
In a first aspect, the application provides a method of soil moisture collaborative inversion, the method comprising: acquiring optical satellite remote sensing data and long-time-sequence microwave satellite remote sensing data of a target area; extracting a remote sensing index corresponding to the optical satellite remote sensing data; the remote sensing indexes comprise normalized vegetation indexes, soil adjustment vegetation indexes and normalized differential moisture indexes; inputting the normalized vegetation index, the soil adjustment vegetation index and the normalized difference moisture index into a pre-trained optical satellite data soil moisture inversion model to obtain a first soil moisture inversion result; the long-time sequence microwave satellite remote sensing data, the long-time sequence meteorological data and the normalized vegetation index are input into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result; the microwave satellite data soil moisture inversion model comprises an input layer, a channel attention module, a space attention module, a data fusion layer, a long-period memory network layer and an output layer, wherein the channel attention module is used for evaluating the importance of each channel, and the space attention module is used for representing the influence of space information on inversion results; and fusing the first soil moisture inversion result and the second soil moisture inversion result, and correcting the fusion result through ground monitoring data to obtain a target soil moisture inversion result.
In an alternative embodiment, extracting a remote sensing index corresponding to the optical satellite remote sensing data includes: extracting the reflectivity of a near infrared band, the reflectivity of a red light band and the reflectivity of a short wave infrared band corresponding to the optical satellite remote sensing data; calculating a normalized vegetation index based on the reflectivity of the near infrared band and the reflectivity of the red band; calculating a soil adjustment vegetation index based on the reflectivity of the near infrared band, the reflectivity of the red band and the soil adjustment factor; and calculating the normalized difference moisture index based on the reflectivity of the near infrared band and the reflectivity of the short wave infrared band.
In an alternative embodiment, the training step of the pre-trained soil moisture inversion model of the optical satellite data includes: acquiring a normalized vegetation index, a soil adjustment vegetation index and a normalized difference moisture index corresponding to optical satellite remote sensing data, and constructing sample data; marking sample data through ground measurement data, and training a preselected machine learning module through the marked sample data to obtain a trained optical satellite data soil moisture inversion model.
In an alternative embodiment, the long time sequence microwave satellite remote sensing data, the long time sequence meteorological data and the normalized vegetation index are input into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result, which comprises the following steps: the method comprises the steps of inputting long-time-sequence microwave satellite remote sensing data, long-time-sequence meteorological data and normalized vegetation indexes into a microwave satellite data soil moisture inversion model for normalization input into an input layer; carrying out maximum pooling treatment and average pooling treatment on input data through a channel attention module, inputting a maximum pooling result and an average pooling result into a multi-layer perceptron to learn channel dimension characteristics and channel importance, and obtaining a channel attention value through Sigmoid function mapping treatment; carrying out maximum pooling treatment and average pooling treatment on the channel attention value at each time point through the spatial attention module, splicing pooling results according to the channels, carrying out convolution operation on the spliced results, and calculating through an activation function to obtain the spatial attention value; adding the spatial attention values corresponding to each time point of each channel through a data fusion layer to obtain fused single-channel time sequence data; and introducing inversion influence degree of time information on the single-channel time sequence data through the long-short-period memory network layer to obtain a second soil moisture inversion result.
In an alternative embodiment, fusing the first soil moisture inversion result and the second soil moisture inversion result includes: performing data downscaling treatment on the second soil moisture inversion result according to the high-resolution vegetation coverage type to obtain a downscaled second soil moisture inversion result; fusion is performed by the following formula:
wherein,for the fused inversion result, +.>Inversion result for first soil moisture>Value element set,/->Inversion result for second soil moisture of downscaling +.>Wherein A and B elements satisfy +.>；/>Inversion results for soil moisture->Value element set,/->Inversion of the result for the second soil moisture>Is a set of value elements of (a).
In an alternative embodiment, correcting the fusion result by ground monitoring data to obtain a target soil moisture inversion result includes: calculating error values between ground monitoring data and remote sensing inversion results of the same point location, and calculating error values corresponding to pixels of other point locations in the remote sensing data through inverse distance interpolation; and correcting the fusion result based on the error value to obtain a target soil moisture inversion result.
In an alternative embodiment, calculating an error value between ground monitoring data and a remote sensing inversion result of the same point location, and calculating error values corresponding to pixels of other point locations in the remote sensing data through inverse distance interpolation, including: calculating a first error value between the ground monitoring data and the remote sensing inversion result of the same point location; determining an error weighting function based on the pixel distance from the ground monitoring point to the same point location; and determining second error values corresponding to pixels of other points in the remote sensing data based on the first error values and the error weighting function.
In a second aspect, the present invention provides an apparatus for soil moisture co-inversion, the apparatus comprising: the data acquisition module is used for acquiring optical satellite remote sensing data and long-time-sequence microwave satellite remote sensing data of a target area; the index extraction module is used for extracting a remote sensing index corresponding to the optical satellite remote sensing data; the remote sensing indexes comprise normalized vegetation indexes, soil adjustment vegetation indexes and normalized differential moisture indexes; the first inversion module is used for inputting the normalized vegetation index, the soil adjustment vegetation index and the normalized difference moisture index into a pre-trained optical satellite data soil moisture inversion model to obtain a first soil moisture inversion result; the second inversion module is used for inputting the long-time sequence microwave satellite remote sensing data, the long-time sequence meteorological data and the normalized vegetation index into the microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result; the microwave satellite data soil moisture inversion model comprises an input layer, a channel attention module, a space attention module, a data fusion layer, a long-period memory network layer and an output layer, wherein the channel attention module is used for evaluating the importance of each channel, and the space attention module is used for representing the influence of space information on inversion results; and the fusion correction module is used for fusing the first soil moisture inversion result and the second soil moisture inversion result, correcting the fusion result through ground monitoring data and obtaining a target soil moisture inversion result.
In a third aspect, the application provides an electronic device comprising a processor and a memory, the memory storing computer executable instructions executable by the processor to perform the method of soil moisture co-inversion of any of the preceding embodiments.
In a fourth aspect, the application provides a computer readable storage medium storing computer executable instructions that, when invoked and executed by a processor, cause the processor to implement a method of soil moisture co-inversion according to any of the preceding embodiments.
According to the soil moisture collaborative inversion method, device, equipment and medium provided by the application, through the microwave satellite data soil moisture inversion model comprising the input layer, the channel attention module, the spatial attention module, the data fusion layer, the long-term memory network layer and the output layer, inversion accuracy is improved, model generalization capability and adaptability are improved, and good inversion effects can be obtained in different areas and different environments; the soil moisture is inverted by the optical remote sensing data and the microwave remote sensing data respectively, and the optical and microwave remote sensing inversion results are fused, so that the advantages are complemented, and the accuracy and stability of soil moisture inversion are improved; and correcting the fusion result through ground test data, so that the spatial resolution of the inversion result is improved, and the soil water inversion error is reduced.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for soil moisture collaborative inversion provided by an embodiment of the application;
FIG. 2 is a schematic diagram of a soil moisture inversion model of microwave satellite data according to an embodiment of the present application;
FIG. 3 is a schematic diagram of downscaling of a pure pixel according to an embodiment of the present application;
FIG. 4 is a schematic diagram of downscaling of a hybrid pixel according to an embodiment of the present application;
FIG. 5 is a process flow diagram of a specific soil moisture inversion with optical and microwave remote sensing cooperation provided by an embodiment of the present application;
FIG. 6 is a block diagram of an apparatus for soil moisture collaborative inversion according to an embodiment of the present application;
fig. 7 is a block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments of the present application. The components of the embodiments of the present application generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the application, as presented in the figures, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures.
Soil moisture is a key parameter in researches of hydrologic cycle, ecological environment, climate change and the like, and the acquisition of high-resolution long-time sequence soil moisture information has important significance for agricultural management, crop growth monitoring and the like. At present, soil moisture monitoring technologies are mainly divided into two main categories, namely direct and indirect.
(1) The direct method is to collect soil samples in the field and analyze and determine the moisture content of the soil in a laboratory, and has the advantages of higher measurement accuracy, difficult influence of atmosphere, vegetation and soil texture and type, but is only suitable for collecting soil moisture information in small areas, and difficult to realize monitoring of soil moisture in large areas.
(2) Indirect measurement methods include a resistance method, a neutron scattering method, a Time Domain Reflectometer (TDR) method, a Frequency Domain Reflectometer (FDR) method, a remote sensing method and the like. The resistance method estimates the moisture content by measuring the resistance change of the soil, is simple and quick, but is sensitive to the soil type and the salt content, so that the precision is easily affected; neutron scattering estimates moisture content by measuring the number of neutrons scattered by hydrogen atoms in the soil, which is highly accurate, but complex and expensive in equipment; time Domain Reflectometry (TDR) and Frequency Domain Reflectometry (FDR) methods estimate moisture content by measuring the propagation velocity or frequency of electromagnetic waves in the soil, which is highly accurate but expensive in equipment; the remote sensing method measures and estimates the soil moisture content in a large range by analyzing data collected from sensors on satellites or unmanned aerial vehicles, and can cover a large range, but is influenced by factors such as topography, vegetation, atmospheric conditions and the like, so that the soil moisture inversion accuracy is not high.
The remote sensing method comprises a spectrum inversion method, a microwave remote sensing inversion method, a thermal infrared remote sensing inversion method, a soil humidity-temperature model inversion method and the like. The spectrum inversion method utilizes the spectrum characteristics of soil to invert soil moisture, and spectrum remote sensing can provide high-resolution data, but is greatly influenced by atmospheric conditions; the microwave remote sensing inversion method is an important means for inverting soil moisture, passive microwaves can acquire soil moisture information in a large range, but the resolution is lower, usually between tens and hundreds of kilometers, and the resolution of active microwaves is higher, which can reach several meters to tens of meters, but the signals can be influenced by terrains; the thermal infrared remote sensing inversion method is based on the relationship between the surface thermal infrared radiation and the physical properties of the soil, and inversion of the soil moisture is performed by measuring the thermal infrared radiation of the surface, so that the thermal infrared remote sensing can acquire the soil moisture information under the cloud-free condition, but is influenced by the surface temperature; the soil humidity-temperature model inversion method is based on physical properties of soil, soil moisture is inverted by utilizing a soil humidity-temperature model, the accuracy of the method is influenced by multiple factors, the thermal conductivity of the soil is not only related to the moisture content, but also influenced by factors such as the soil type, the organic matter content, the temperature and the like, and the inversion accuracy can be influenced.
Based on the above, the embodiment of the application provides a soil moisture collaborative inversion method, a device, equipment and a medium, which can obtain better inversion effects in different areas and different environments, improve the accuracy, stability and resolution of soil moisture inversion and reduce soil moisture inversion errors.
The embodiment of the application provides a soil moisture collaborative inversion method, which is shown in fig. 1 and mainly comprises the following steps:
step S110, optical satellite remote sensing data and long-time sequence microwave satellite remote sensing data of a target area are obtained.
The target area is an area to be monitored for soil moisture, and the size and area of the area can be selected according to the configuration of the actual remote sensing equipment, and the area is not particularly limited herein.
Step S120, extracting a remote sensing index corresponding to the optical satellite remote sensing data.
The remote sensing indexes comprise normalized vegetation indexes, soil adjustment vegetation indexes and normalized differential moisture indexes.
The normalized vegetation index (NDVI) is used to characterize the surface vegetation coverage and biomass, and can be used indirectly for soil moisture inversion because of the correlation between vegetation coverage and soil moisture.
Soil conditioning vegetation index (SAVI) is used to characterize the effect of soil background reflection on vegetation index.
The normalized differential moisture index (NDWI) is an index for monitoring surface moisture and it uses mainly data in both the near infrared and short wave infrared bands.
Step S130, inputting the normalized vegetation index, the soil adjustment vegetation index and the normalized difference moisture index into a pre-trained optical satellite data soil moisture inversion model to obtain a first soil moisture inversion result.
And step S140, inputting the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result.
The microwave satellite data soil moisture inversion model comprises an input layer, a channel attention module, a space attention module, a data fusion layer, a long-period memory network layer and an output layer, wherein the channel attention module is used for evaluating the importance of each channel, and the space attention module is used for representing the influence of space information on inversion results;
and step S150, fusing the first soil moisture inversion result and the second soil moisture inversion result, and correcting the fusion result through ground monitoring data to obtain a target soil moisture inversion result.
The method will be specifically described below.
In one embodiment, the extracting the remote sensing index corresponding to the optical satellite remote sensing data may include the following steps 1-1 to 1-4:
and step 1-1, extracting the reflectivity of a near infrared band, the reflectivity of a red light band and the reflectivity of a short wave infrared band corresponding to the optical satellite remote sensing data.
And step 1-2, calculating a normalized vegetation index based on the reflectivity of the near infrared band and the reflectivity of the red band. Ndvi= (NIR-R)/(nir+r), where NIR is the reflectivity in the near infrared band and R is the reflectivity in the red band.
And 1-3, calculating a soil adjustment vegetation index based on the reflectivity of the near infrared band, the reflectivity of the red band and the soil adjustment factor. SAVI= (1+L) × (NIR-R)/(NIR+R+L). Where L is a constant, typically taking a value of 0.5.
And step 1-4, calculating a normalized difference moisture index based on the reflectivity of the near infrared band and the reflectivity of the short wave infrared band. Ndwi= (NIR-SWIR)/(nir+swir), where SWIR is the reflectance in the short wave infrared band.
Further, the training step of the pre-trained soil moisture inversion model of the optical satellite data may include the following steps 2-1 and 2-2:
Step 2-1, obtaining a normalized vegetation index, a soil adjustment vegetation index and a normalized difference moisture index corresponding to optical satellite remote sensing data, and constructing sample data;
and 2-2, marking sample data through ground measurement data, and training a preselected machine learning module through the marked sample data to obtain a trained soil moisture inversion model of the optical satellite data.
The optical satellite data soil moisture inversion model may be a convolutional neural network (Convolutional Neural Networks, CNN) or a support vector machine model (Support Vector Machine, SVM).
In one embodiment, the model relationship of features to soil moisture may be established by convolutional neural network CNN. The input data are NDVI, SAVI and NDWI data G, the metafile G is an mxn×3 matrix, and the sample flag data is ground measurement data V (mxn matrix).
When the support vector machine is selected, 60% of the data set is used as a training set, 20% is used as a verification set, 20% is used as a test set, and model training is carried out to obtain the optical satellite data soil water inversion model M1.
Referring to fig. 2, the microwave satellite data soil moisture inversion model comprises an input layer, a channel attention module, a space attention module, a data fusion layer, a Long Short-Term Memory (LSTM) layer and an output layer, and is shown in fig. 2.
In one embodiment, the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index are input into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result, and the method can comprise the following steps 3-1 to 3-5:
and 3-1, inputting long-time sequence microwave satellite remote sensing data, long-time sequence meteorological data and normalized vegetation indexes into a microwave satellite data soil moisture inversion model for normalization and input into an input layer. Specifically, active microwaves (VV, HH, VH, HV), NDVI, vegetation cover type, slope direction, accumulated rainfall, average temperature, average humidity, average wind speed are included.
Alternatively, the input data may be normalized data. Specific:
1) The target area long time sequence active microwave data is preprocessed, a backward scattering coefficient is extracted, and image shooting parameters are extracted;
2) The target area long time sequence accumulated rainfall data, average temperature data, average humidity, average wind speed data and NDVI data are processed into data with the same spatial resolution of active microwaves through an upscaling method;
3) The vegetation coverage type data, gradient data and slope data of the target area are expanded into long time sequence data.
And 3-2, carrying out maximum pooling treatment and average pooling treatment on the input data through a channel attention module, inputting the maximum pooling result and the average pooling result into a multi-layer perceptron to carry out channel dimension feature and channel importance learning, and obtaining a channel attention value through Sigmoid function mapping treatment. Specifically, the multidimensional data of each time point of the module is subjected to maximum pooling and average pooling, the pooled result is sent to a multi-layer perceptron (MLP) to learn channel dimension characteristics and importance of each channel, the MLP output result is subjected to addition operation, and the channel attention value is obtained through Sigmoid function mapping processing.
And 3-3, carrying out maximum pooling treatment and average pooling treatment on the channel attention value at each time point through the spatial attention module, splicing the pooling results according to the channels, carrying out convolution operation on the splicing results, and calculating through an activation function to obtain the spatial attention value. Specifically, the result of channel attention model excitation at each time point of the module is subjected to maximum pooling and average pooling, the pooled result is spliced according to the channel, convolution operation is performed on the spliced result, and a spatial attention value is obtained through an activation function.
And 3-4, adding the spatial attention values corresponding to each time point of each channel through a data fusion layer to obtain fused single-channel time sequence data. Specifically, the channels of the results of the excitation of the spatial attention model at each time point are added to obtain the fused time sequence input characteristics, and each time point is taken as single-channel data.
And 3-5, introducing inversion influence of time information on the single-channel time sequence data through the long-term and short-term memory network layer to obtain a second soil moisture inversion result.
The long-short-term memory network Layer (LSTM) is an improved cyclic neural network, can solve the problem that RNNs cannot handle long-distance dependence, can also solve the problems of gradient explosion or gradient disappearance and the like common in the neural network, and is very effective in processing sequence data.
After the data set is constructed, 60% of the data set is used as a training set, 20% is used as a verification set, 20% is used as a test set, and model training is carried out to obtain the microwave satellite data soil moisture inversion model M2.
Obtaining a first soil moisture inversion result through a pre-trained soil moisture inversion model of the optical satellite data, and marking the first soil moisture inversion result as the first soil moisture inversion result . Obtaining a second soil moisture inversion result through a microwave satellite data soil moisture inversion model, and marking the second soil moisture inversion result as the second soil moisture inversion result。
In one embodiment, in obtaining the first soil moisture inversion result and the second soil moisture inversion result, fusing the first soil moisture inversion result and the second soil moisture inversion result may include the following steps 4-1 and 4-2:
and 4-1, performing data downscaling treatment on the second soil moisture inversion result according to the high-resolution vegetation coverage type to obtain a downscaled second soil moisture inversion result.
In order to realize the fusion of optical and microwave inversion results, a large-scale microwave result is firstly obtainedDownscaling according to the high resolution vegetation coverage type data. For the pure pixel, the soil moisture keeps the original scale pixel value, as shown in fig. 3, the left side is the microwave pure pixel, the soil moisture inversion value is a, and the pixel soil moisture after scale reduction is a. For mixed pixels, the soil moisture value of the downscaled pixels adopts the inversion value of the similar pure pixels as the soil moisture value because the inversion result has poor reliability, and is shown in fig. 4.
The method for calculating the soil moisture value of each small pixel after the scale is reduced comprises the following steps:
1) According to the high-resolution vegetation coverage type classification obtained in the first step, decomposing an active microwave soil inversion resultPicture element, acquisition of a set of picture elements->Wherein->Is of the pixel type (will be maximumThe duty ratio ground object type is used as a small pixel type, mixed pixels are not distinguished any more), and the floor object type is +.>For the parent element type->And->For the center coordinates of the picture elements>For the soil moisture value to be determined, < > a->Is the soil moisture value of the parent pixel.
2) Taking i=1;
3) If it isThen->Then step 6) is performed, otherwise step 4) is performed;
4) RetrievalMiddle distance>Less than d and of the type +.>Obtaining a set R;
wherein,is->Pixel type->Is->Type of picture element.
5）Taking distance +.>The nearest value is taken as soil moisture value, if R is empty,/->。
6) i=i+1, and ends if i > k, otherwise step 3) is performed.
Inversion of results for active microwavesDownscaling pixel by pixel to finally obtain a downscaling active microwave soil moisture inversion result>。
Step 4-2, fusing by the following formula:
wherein,for the fused inversion result, +.>Inversion result for first soil moisture>Value element set,/->Inversion result for second soil moisture of downscaling +.>Wherein A and B elements satisfy +. >；/>Inversion results for soil moisture->Value element set,/->Inversion of the result for the second soil moisture>Is a set of value elements of (a).
The optical image is easily influenced by various factors to cause an optical image inversion error, so that the influence of the optical image inversion error can be reduced by introducing an optical image influence factor by adopting a variable weight coefficient in data fusion.
A and B were determined by influencing factors (atmospheric humidity, aerosol, cloud layer):
where a is a weight reference, typically taking a=0.5;
is the atmospheric humidity influencing factor, +.>，/>Is air humidity;
is aerosol influencing factor, +.>，/>Is aerosol thickness;
is a cloud affecting factor, ++>，/>Is cloud optical thickness.
In order to further improve inversion accuracy, correcting the fusion result through ground monitoring data to obtain a target soil moisture inversion result, the method can comprise the following steps of 5-1 and 5-2:
step 5-1, calculating error values between ground monitoring data and remote sensing inversion results of the same point location, and calculating error values corresponding to pixels of other point locations in the remote sensing data through inverse distance interpolation;
in particular implementations, the following steps 5-1.1 to 5-1.3 may be included:
step 5-1.1, calculating a first error value between ground monitoring data and a remote sensing inversion result of the same point location;
Step 5-1.2, determining an error weighting function based on pixel distances from the ground monitoring point to the same point;
and 5-1.3, determining second error values corresponding to pixels of other points in the remote sensing data based on the first error values and the error weighting function.
And 5-2, correcting the fusion result based on the error value to obtain a target soil moisture inversion result.
In practical application, compared with a remote sensing inversion result, the ground detection data is more accurate. The method comprises the steps of monitoring through ground observation points or sampling points to obtain ground soil moisture detection data, and correcting remote sensing inversion data V by adopting the ground data.
Firstly, calculating the difference value between ground measurement data and the remote sensing inversion result of the same point location, calculating pixel error values of other point locations in the remote sensing data by using inverse distance interpolation, and further adjusting the inversion result of each pixel to achieve the effect of correcting the inversion result.
The error weighting function is:
where p is any positive real number, typically p=2;the ground measurement point-to-pixel distance is calculated according to the formula:
(,/>) Is the j-th pixel coordinate, (-)>，/>) The point coordinates are measured for the ground.
Calculating pixel j error correction value：
Wherein,for remote sensing inversion error at the ith ground detection point,/- >For ground detection value, ++>And n is the number of ground observation data points for the inversion value of the corresponding position.
The pixel j inversion correction formula is as follows:
executing the formula by all pixel inversion results to finish inversion result correction and finally generating a soil moisture inversion remote sensing inversion product。
Fig. 5 shows a specific process flow of performing optical and microwave remote sensing cooperative soil moisture inversion, and the optical and microwave remote sensing inversion results are fused by adopting an optical and active microwave cooperative inversion method, so that advantages are complemented, and the accuracy and stability of soil moisture inversion are improved. The application constructs the neural network model comprising the channel attention mechanism module, the space attention mechanism module and the LSTM module, introduces channel importance information, time information and space information for active microwave inversion of soil moisture, reflects the associated influence of multi-type data (comprising data such as topography factors, rainfall, temperature and the like), history information and pixel surrounding pixel information on the soil moisture in the model, improves inversion accuracy, improves model generalization capability and adaptability, and can obtain better inversion effect in different areas and different environments. The scheme adopts a comprehensive fusion method to finish inversion data fusion, and improves data reliability and accuracy. Firstly, the scale of an active microwave inversion result is reduced based on the composition of mixed pixels, the optical and active microwave inversion result is fused on the basis, and finally, the fusion result is corrected through ground test data, so that the inversion result is spatially resolved, and the soil water inversion error is reduced.
In summary, the application introduces channel importance information, time information and space information for active microwave inversion of soil moisture by constructing the neural network model comprising the channel attention mechanism module, the space attention mechanism module and the LSTM module, reflects the associated influence of multi-type data (comprising data such as topography factors, rainfall, temperature and the like), history information and pixel surrounding pixel information on the soil moisture in the model, improves inversion accuracy, improves model generalization capability and adaptability, and can obtain better inversion effect in different areas and different environments. And the soil moisture is inverted by adopting optical and active microwave remote sensing data, and the optical and microwave remote sensing inversion results are fused, so that the advantages are complementary, and the accuracy and stability of soil moisture inversion are improved. And the inversion data fusion is completed by adopting a comprehensive fusion method, so that the reliability and accuracy of the data are improved. Firstly, the scale of an active microwave inversion result is reduced based on the composition of mixed pixels, the optical and active microwave inversion result is fused on the basis, and finally, the fusion result is corrected through ground test data, so that the spatial resolution of the inversion result is improved, and the soil water inversion error is reduced.
Based on the above method embodiment, the present application further provides a second aspect, and the present application provides an apparatus for soil moisture collaborative inversion, as shown in fig. 6, where the apparatus includes the following parts:
the data acquisition module 610 is configured to acquire optical satellite remote sensing data and long-time-sequence microwave satellite remote sensing data of a target area;
the index extraction module 620 is configured to extract a remote sensing index corresponding to the remote sensing data of the optical satellite; the remote sensing indexes comprise normalized vegetation indexes, soil adjustment vegetation indexes and normalized differential moisture indexes;
the first inversion module 630 is configured to input the normalized vegetation index, the soil adjustment vegetation index, and the normalized difference moisture index to a pre-trained soil moisture inversion model of optical satellite data, to obtain a first soil moisture inversion result;
the second inversion module 640 is configured to input the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result; the microwave satellite data soil moisture inversion model comprises an input layer, a channel attention module, a space attention module, a data fusion layer, a long-period memory network layer and an output layer, wherein the channel attention module is used for evaluating the importance of each channel, and the space attention module is used for representing the influence of space information on inversion results;
The fusion correction module 650 is configured to fuse the first soil moisture inversion result and the second soil moisture inversion result, and correct the fusion result through ground monitoring data to obtain a target soil moisture inversion result.
In an alternative embodiment, the index extraction module 620 is further configured to:
extracting the reflectivity of a near infrared band, the reflectivity of a red light band and the reflectivity of a short wave infrared band corresponding to the optical satellite remote sensing data;
calculating a normalized vegetation index based on the reflectivity of the near infrared band and the reflectivity of the red band;
calculating a soil adjustment vegetation index based on the reflectivity of the near infrared band, the reflectivity of the red band and the soil adjustment factor;
and calculating the normalized difference moisture index based on the reflectivity of the near infrared band and the reflectivity of the short wave infrared band.
In an alternative embodiment, the apparatus further includes a model training module configured to:
acquiring a normalized vegetation index, a soil adjustment vegetation index and a normalized difference moisture index corresponding to optical satellite remote sensing data, and constructing sample data;
marking sample data through ground measurement data, and training a preselected machine learning module through the marked sample data to obtain a trained optical satellite data soil moisture inversion model.
In an alternative embodiment, the second inversion module 640 is further configured to:
the method comprises the steps of inputting long-time-sequence microwave satellite remote sensing data, long-time-sequence meteorological data and normalized vegetation indexes into a microwave satellite data soil moisture inversion model for normalization input into an input layer;
carrying out maximum pooling treatment and average pooling treatment on input data through a channel attention module, inputting a maximum pooling result and an average pooling result into a multi-layer perceptron to learn channel dimension characteristics and channel importance, and obtaining a channel attention value through Sigmoid function mapping treatment;
carrying out maximum pooling treatment and average pooling treatment on the channel attention value at each time point through the spatial attention module, splicing pooling results according to the channels, carrying out convolution operation on the spliced results, and calculating through an activation function to obtain the spatial attention value;
adding the spatial attention values corresponding to each time point of each channel through a data fusion layer to obtain fused single-channel time sequence data;
and introducing inversion influence degree of time information on the single-channel time sequence data through the long-short-period memory network layer to obtain a second soil moisture inversion result.
In an alternative embodiment, the fusion correction module 650 is further configured to:
Performing data downscaling treatment on the second soil moisture inversion result according to the high-resolution vegetation coverage type to obtain a downscaled second soil moisture inversion result;
fusion is performed by the following formula:
wherein,for the fused inversion result, +.>Inversion result for first soil moisture>Value element set,/->Inversion result for second soil moisture of downscaling +.>Wherein A and B elements satisfy +.>；/>Inversion results for soil moisture->Value element set,/->Inversion of the result for the second soil moisture>Is a set of value elements of (a).
In an alternative embodiment, correcting the fusion result by ground monitoring data to obtain a target soil moisture inversion result includes:
calculating error values between ground monitoring data and remote sensing inversion results of the same point location, and calculating error values corresponding to pixels of other point locations in the remote sensing data through inverse distance interpolation;
and correcting the fusion result based on the error value to obtain a target soil moisture inversion result.
In an alternative embodiment, the fusion correction module 650 is further configured to:
calculating a first error value between the ground monitoring data and the remote sensing inversion result of the same point location;
Determining an error weighting function based on the pixel distance from the ground monitoring point to the same point location;
and determining second error values corresponding to pixels of other points in the remote sensing data based on the first error values and the error weighting function.
The implementation principle and the produced technical effects of the soil moisture collaborative inversion device provided by the embodiment of the application are the same as those of the method embodiment, and for the sake of brief description, reference is made to corresponding contents in the method embodiment of the soil moisture collaborative inversion to the places where the embodiment of the soil moisture collaborative inversion device is not mentioned.
An embodiment of the present application further provides an electronic device, as shown in fig. 7, which is a schematic structural diagram of the electronic device, where the electronic device 100 includes a processor 71 and a memory 70, where the memory 70 stores computer executable instructions that can be executed by the processor 71, and the processor 71 executes the computer executable instructions to implement a method of the above-mentioned soil moisture collaborative inversion.
In the embodiment shown in fig. 7, the electronic device further comprises a bus 72 and a communication interface 73, wherein the processor 71, the communication interface 73 and the memory 70 are connected by the bus 72.
The memory 70 may include a high-speed random access memory (RAM, random Access Memory), and may further include a non-volatile memory (non-volatile memory), such as at least one magnetic disk memory. The communication connection between the system network element and the at least one other network element is achieved via at least one communication interface 73 (which may be wired or wireless), which may use the internet, a wide area network, a local network, a metropolitan area network, etc. Bus 72 may be an ISA (Industry Standard Architecture ) bus, PCI (Peripheral Component Interconnect, peripheral component interconnect standard) bus, or EISA (Extended Industry Standard Architecture ) bus, among others. The bus 72 may be classified as an address bus, a data bus, a control bus, or the like. For ease of illustration, only one bi-directional arrow is shown in FIG. 7, but not only one bus or type of bus.
The processor 71 may be an integrated circuit chip with signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuitry in hardware or instructions in software in the processor 71. The processor 71 may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but also digital signal processors (Digital Signal Processor, DSP for short), application specific integrated circuits (Application Specific Integrated Circuit, ASIC for short), field-programmable gate arrays (Field-Programmable Gate Array, FPGA for short) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present application may be embodied directly in the execution of a hardware decoding processor, or in the execution of a combination of hardware and software modules in a decoding processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in a memory and the processor 71 reads the information in the memory and, in combination with its hardware, performs the steps of the method of soil moisture co-inversion of the previous embodiments.
The embodiment of the application also provides a computer readable storage medium, which stores computer executable instructions that, when being called and executed by a processor, cause the processor to implement the above-mentioned soil moisture collaborative inversion method, and the specific implementation can be found in the foregoing method embodiments, which are not repeated herein.
The method, apparatus, device and medium for soil moisture collaborative inversion provided by the embodiments of the present application include a computer readable storage medium storing program codes, where the instructions included in the program codes may be used to execute the method described in the foregoing method embodiment, and specific implementation may refer to the method embodiment and will not be described herein.
The relative steps, numerical expressions and numerical values of the components and steps set forth in these embodiments do not limit the scope of the present application unless it is specifically stated otherwise.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer readable storage medium executable by a processor. Based on this understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In the description of the present application, it should be noted that the terms "first," "second," and the like are used merely to distinguish between descriptions and should not be construed as indicating or implying relative importance.
In the description of the present application, it should also be noted that, unless explicitly specified and limited otherwise, the terms "connected," "connected," and "connected" are to be construed broadly, and may be either fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present application will be understood in specific cases by those of ordinary skill in the art.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the application.
Claims (9)
1. A method of soil moisture collaborative inversion, the method comprising:
acquiring optical satellite remote sensing data and long-time-sequence microwave satellite remote sensing data of a target area;
extracting a remote sensing index corresponding to the optical satellite remote sensing data; the remote sensing indexes comprise normalized vegetation indexes, soil adjustment vegetation indexes and normalized differential moisture indexes;
inputting the normalized vegetation index, the soil adjustment vegetation index and the normalized difference moisture index into a pre-trained optical satellite data soil moisture inversion model to obtain a first soil moisture inversion result;
inputting the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result; the microwave satellite data soil moisture inversion model comprises an input layer, a channel attention module, a space attention module, a data fusion layer, a long-period memory network layer and an output layer, wherein the channel attention module is used for evaluating the importance of each channel, and the space attention module is used for representing the influence of space information on inversion results;
Fusing the first soil moisture inversion result and the second soil moisture inversion result, and correcting the fusion result through ground monitoring data to obtain a target soil moisture inversion result;
inputting the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result, wherein the method comprises the following steps of:
inputting the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index into a microwave satellite data soil moisture inversion model for normalization input into an input layer;
the channel attention module is used for carrying out maximum pooling treatment and average pooling treatment on input data, inputting the maximum pooling result and the average pooling result into a multi-layer perceptron to carry out channel dimension characteristic and channel importance learning, and obtaining a channel attention value through Sigmoid function mapping treatment;
carrying out maximum pooling treatment and average pooling treatment on the channel attention value through each time point of the spatial attention module, splicing pooling results according to channels, carrying out convolution operation on the spliced results, and calculating through an activation function to obtain the spatial attention value;
Adding the spatial attention values corresponding to each time point of each channel through the data fusion layer to obtain fused single-channel time sequence data;
and introducing inversion influence of time information on the single-channel time sequence data through the long-short-period memory network layer to obtain a second soil moisture inversion result.
2. The method of soil moisture collaborative inversion according to claim 1, wherein extracting a remote sensing index corresponding to the optical satellite remote sensing data comprises:
extracting the reflectivity of a near infrared band, the reflectivity of a red light band and the reflectivity of a short wave infrared band corresponding to the optical satellite remote sensing data;
calculating a normalized vegetation index based on the reflectivity of the near infrared band and the reflectivity of the red band;
calculating a soil adjustment vegetation index based on the reflectivity of the near infrared band, the reflectivity of the red band and the soil adjustment factor;
and calculating the normalized difference moisture index based on the reflectivity of the near infrared band and the reflectivity of the short wave infrared band.
3. The method of soil moisture collaborative inversion according to claim 1, wherein the training step of the pre-trained optical satellite data soil moisture inversion model comprises:
Acquiring a normalized vegetation index, a soil adjustment vegetation index and a normalized difference moisture index corresponding to optical satellite remote sensing data, and constructing sample data;
marking sample data through ground measurement data, and training a preselected machine learning module through the marked sample data to obtain a trained optical satellite data soil moisture inversion model.
4. The method of soil moisture collaborative inversion according to claim 1, wherein fusing the first soil moisture inversion result and the second soil moisture inversion result comprises:
performing data downscaling treatment on the second soil moisture inversion result according to the high-resolution vegetation coverage type to obtain a downscaled second soil moisture inversion result;
fusion is performed by the following formula:
wherein,for the fused inversion result, +.>Inversion result for first soil moisture>Value element set,/->Inversion result for second soil moisture of downscaling +.>Is a weight matrix of both A and B, wherein the A and B elements satisfy；/>Inversion results for soil moisture->Value element set,/->Inversion of the result for the second soil moisture >Is a set of value elements of (a).
5. The method of soil moisture collaborative inversion according to claim 1, wherein correcting the fusion result by ground monitoring data to obtain a target soil moisture inversion result comprises:
calculating error values between ground monitoring data and remote sensing inversion results of the same point location, and calculating error values corresponding to pixels of other point locations in the remote sensing data through inverse distance interpolation;
and correcting the fusion result based on the error value to obtain a target soil moisture inversion result.
6. The method of collaborative inversion of soil moisture according to claim 5, wherein calculating an error value between ground monitoring data and a remote sensing inversion result of the same point location and calculating error values corresponding to pixels of other points in the remote sensing data by inverse distance interpolation comprises:
calculating a first error value between the ground monitoring data and the remote sensing inversion result of the same point location;
determining an error weighting function based on the pixel distance from the ground monitoring point to the same point location;
and determining second error values corresponding to pixels of other points in the remote sensing data based on the first error values and the error weighting function.
7. An apparatus for soil moisture collaborative inversion, the apparatus comprising:
the data acquisition module is used for acquiring optical satellite remote sensing data and long-time-sequence microwave satellite remote sensing data of a target area;
the index extraction module is used for extracting a remote sensing index corresponding to the optical satellite remote sensing data; the remote sensing indexes comprise normalized vegetation indexes, soil adjustment vegetation indexes and normalized differential moisture indexes;
the first inversion module is used for inputting the normalized vegetation index, the soil adjustment vegetation index and the normalized difference moisture index into a pre-trained optical satellite data soil moisture inversion model to obtain a first soil moisture inversion result;
the second inversion module is used for inputting the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index into a microwave satellite data soil moisture inversion model to obtain a second soil moisture inversion result; the microwave satellite data soil moisture inversion model comprises an input layer, a channel attention module, a space attention module, a data fusion layer, a long-period memory network layer and an output layer, wherein the channel attention module is used for evaluating the importance of each channel, and the space attention module is used for representing the influence of space information on inversion results;
The fusion correction module is used for fusing the first soil moisture inversion result and the second soil moisture inversion result, correcting the fusion result through ground monitoring data and obtaining a target soil moisture inversion result;
the second inversion module is further used for inputting the long-time-sequence microwave satellite remote sensing data, the long-time-sequence meteorological data and the normalized vegetation index into a microwave satellite data soil moisture inversion model for normalization input into an input layer; the channel attention module is used for carrying out maximum pooling treatment and average pooling treatment on input data, inputting the maximum pooling result and the average pooling result into a multi-layer perceptron to carry out channel dimension characteristic and channel importance learning, and obtaining a channel attention value through Sigmoid function mapping treatment; carrying out maximum pooling treatment and average pooling treatment on the channel attention value through each time point of the spatial attention module, splicing pooling results according to channels, carrying out convolution operation on the spliced results, and calculating through an activation function to obtain the spatial attention value; adding the spatial attention values corresponding to each time point of each channel through the data fusion layer to obtain fused single-channel time sequence data; and introducing inversion influence of time information on the single-channel time sequence data through the long-short-period memory network layer to obtain a second soil moisture inversion result.
8. An electronic device comprising a processor and a memory, the memory storing computer executable instructions executable by the processor, the processor executing the computer executable instructions to implement the method of soil moisture co-inversion of any one of claims 1 to 6.
9. A computer readable storage medium storing computer executable instructions which, when invoked and executed by a processor, cause the processor to implement the method of soil moisture co-inversion of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311147561.XA CN116879297B (en) | 2023-09-07 | 2023-09-07 | Soil moisture collaborative inversion method, device, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311147561.XA CN116879297B (en) | 2023-09-07 | 2023-09-07 | Soil moisture collaborative inversion method, device, equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116879297A CN116879297A (en) | 2023-10-13 |
| CN116879297B true CN116879297B (en) | 2023-12-12 |
Family
ID=88272121
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311147561.XA Active CN116879297B (en) | 2023-09-07 | 2023-09-07 | Soil moisture collaborative inversion method, device, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116879297B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117313563B (en) * | 2023-11-30 | 2024-02-27 | 江汉大学 | Configuration method of soil moisture reconstruction model under physical constraint and deep learning coupling |
| CN117421568B (en) * | 2023-12-19 | 2024-03-05 | 石家庄铁道大学 | Method, device, terminal and storage medium for positioning freezing and thawing monitoring points of side slope rock and soil mass |
| CN117787110B (en) * | 2024-02-26 | 2024-05-17 | 山东建筑大学 | Soil moisture inversion method and system based on deep learning model |
| CN117871472B (en) * | 2024-03-13 | 2024-06-11 | 江汉大学 | Method and device for reducing dimension of satellite soil moisture product |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106226260A (en) * | 2016-08-10 | 2016-12-14 | 武汉大学 | A kind of combination microwave and the Soil Moisture Inversion method of infrared remote sensing image |
| CN109344536A (en) * | 2018-10-30 | 2019-02-15 | 中国科学院、水利部成都山地灾害与环境研究所 | A kind of multi-level passive microwave soil moisture verification method coupling multi-source data |
| CN112488008A (en) * | 2020-12-04 | 2021-03-12 | 中国科学院地理科学与资源研究所 | Soil moisture measurement method and device based on multi-source remote sensing data fusion |
| CN113255874A (en) * | 2021-06-08 | 2021-08-13 | 中国农业科学院农业资源与农业区划研究所 | Optimized BP neural network-based soil moisture inversion method through microwave remote sensing |
| CN113836779A (en) * | 2020-06-23 | 2021-12-24 | 西北农林科技大学 | CNN-based farmland surface soil moisture inversion method for Sentinel multi-source data |
| CN116046692A (en) * | 2023-03-23 | 2023-05-02 | 航天宏图信息技术股份有限公司 | Soil heavy metal pollution monitoring method and device based on hyperspectrum |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230064454A1 (en) * | 2021-08-31 | 2023-03-02 | Satsure Analytics India Private Limited | System and method for generating soil moisture data from satellite imagery using deep learning model |
-
2023
- 2023-09-07 CN CN202311147561.XA patent/CN116879297B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106226260A (en) * | 2016-08-10 | 2016-12-14 | 武汉大学 | A kind of combination microwave and the Soil Moisture Inversion method of infrared remote sensing image |
| CN109344536A (en) * | 2018-10-30 | 2019-02-15 | 中国科学院、水利部成都山地灾害与环境研究所 | A kind of multi-level passive microwave soil moisture verification method coupling multi-source data |
| CN113836779A (en) * | 2020-06-23 | 2021-12-24 | 西北农林科技大学 | CNN-based farmland surface soil moisture inversion method for Sentinel multi-source data |
| CN112488008A (en) * | 2020-12-04 | 2021-03-12 | 中国科学院地理科学与资源研究所 | Soil moisture measurement method and device based on multi-source remote sensing data fusion |
| CN113255874A (en) * | 2021-06-08 | 2021-08-13 | 中国农业科学院农业资源与农业区划研究所 | Optimized BP neural network-based soil moisture inversion method through microwave remote sensing |
| CN116046692A (en) * | 2023-03-23 | 2023-05-02 | 航天宏图信息技术股份有限公司 | Soil heavy metal pollution monitoring method and device based on hyperspectrum |
Non-Patent Citations (1)
| Title |
|---|
| 土壤水分多源卫星遥感联合反演研究进展;蒋瑞瑞等;《自然资源遥感》;第1-13页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116879297A (en) | 2023-10-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116879297B (en) | Soil moisture collaborative inversion method, device, equipment and medium | |
| Chen et al. | Monitoring soil organic carbon in alpine soils using in situ vis‐NIR spectroscopy and a multilayer perceptron | |
| Zhou et al. | Mapping the concentrations of total suspended matter in Lake Taihu, China, using Landsat‐5 TM data | |
| Chou et al. | Assessment of the ASCAT wind error characteristics by global dropwindsonde observations | |
| CN113049500B (en) | Water quality detection model training and water quality detection method, electronic equipment and storage medium | |
| Zhang et al. | Detection of sea ice in sediment laden water using MODIS in the Bohai Sea: A CART decision tree method | |
| Wang et al. | Land cover change detection with a cross‐correlogram spectral matching algorithm | |
| CN113252583B (en) | Method for calculating alpine hay coverage based on hay vegetation index | |
| CN111879709B (en) | Lake water body spectral reflectivity inspection method and device | |
| Li et al. | A neural network-based approach for the detection of heavy precipitation using GNSS observations and surface meteorological data | |
| Aires et al. | Statistical approaches to assimilate ASCAT soil moisture information—I. Methodologies and first assessment | |
| He et al. | Bias correction for retrieval of atmospheric parameters from the Microwave Humidity and Temperature Sounder onboard the Fengyun-3C satellite | |
| CN111274911B (en) | Dense fog monitoring method based on wireless microwave attenuation characteristic transfer learning | |
| CN115494007A (en) | Random forest based high-precision rapid detection method and device for soil organic matters | |
| Rawat et al. | Parameterization of the modified water cloud model (MWCM) using normalized difference vegetation index (NDVI) for winter wheat crop: a case study from Punjab, India | |
| Campisi-Pinto et al. | Analysis of the Regularization Parameters of Primal–Dual Interior Method for Convex Objectives Applied to 1 H Low Field Nuclear Magnetic Resonance Data Processing | |
| Singh et al. | Improved generalized calibration of an impedance probe for soil moisture measurement at regional scale using Bayesian neural network and soil physical properties | |
| CN106525753A (en) | Convenient and simple remote-sensing soil moisture monitoring method | |
| Nassar et al. | To what extend does the Eddy Covariance footprint cutoff influence the estimation of surface energy fluxes using two source energy balance model and high-resolution imagery in commercial vineyards? | |
| Chiu et al. | Bayesian retrieval of complete posterior PDFs of oceanic rain rate from microwave observations | |
| Ettalbi et al. | Soil Moisture Retrieval in Bare Agricultural Areas Using Sentinel-1 Images | |
| Deb et al. | A modified tracer selection and tracking procedure to derive winds using water vapor imagers | |
| Meissner et al. | SMAP salinity retrievals near the sea-ice edge using multi-channel AMSR2 brightness temperatures | |
| Bocchiola | Use of scale recursive estimation for assimilation of precipitation data from TRMM (PR and TMI) and NEXRAD | |
| CN116124716A (en) | Water quality monitoring method and device based on proximity hyperspectral |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |