Regional soil moisture monitoring method based on airborne spectrum reconstruction optical satellite remote sensing

Abstract

The invention discloses a regional soil moisture monitoring method based on airborne spectrum reconstruction optical satellite remote sensing, and relates to the technical field of soil moisture monitoring. The method comprises the following steps: acquiring satellite remote sensing data of a to-be-monitored area and acquiring unmanned aerial vehicle remote sensing data of a local area in the to-be-monitored area; performing spectrum reconstruction on the satellite remote sensing data of the to-be-monitored area according to the unmanned aerial vehicle remote sensing data of the local area to obtain reconstructed satellite spectrum data of the satellite remote sensing data; and inputting the reconstructed satellite spectrum data into a pre-trained soil moisture inversion model to obtain the soil moisture content of the to-be-monitored area. The method can accurately and rapidly monitor the soil moisture.

Description

Technical Field

[0001] The invention relates to the technical field of soil moisture monitoring, and in particular to a regional soil moisture monitoring method using airborne spectral reconstruction optical satellite remote sensing. Background Art

[0002] Soil moisture is usually quantified as soil moisture content, which has important research value in crop growth monitoring, regional drought monitoring and early warning, efficient use of water resources, precision irrigation and smart agriculture.

[0003] At present, optical satellite remote sensing can be used to monitor soil moisture in large-scale irrigation areas. However, due to the limitation of spatial resolution and the dynamic changes of vegetation coverage (surface spatial heterogeneity), the accuracy of soil moisture monitoring is low. Low-altitude UAV remote sensing has the characteristics of ultra-high spatial resolution and fast response speed. It can capture rich surface information and has higher accuracy in soil moisture monitoring than satellite remote sensing. However, for large-scale irrigation areas, it is time-consuming and laborious to collect UAV remote sensing data covering the entire irrigation area, and the timeliness of soil moisture monitoring products cannot be guaranteed.

[0004] Therefore, there is an urgent need for a method that can accurately and quickly monitor soil moisture. Summary of the invention

[0005] Based on this, it is necessary to provide a regional soil moisture monitoring method based on airborne spectral reconstruction optical satellite remote sensing to address the above technical problems. This method can accurately and quickly monitor soil moisture.

[0006] The present invention adopts the following technical solutions:

[0007] The present invention provides a method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing, comprising:

[0008] Acquire satellite remote sensing data of the area to be monitored and collect drone remote sensing data of local areas in the area to be monitored;

[0009] According to the UAV remote sensing data of the local area, the satellite remote sensing data of the monitored area is spectrally reconstructed to obtain the reconstructed satellite spectral data of the satellite remote sensing data;

[0010] The reconstructed satellite spectral data are input into the pre-trained soil moisture inversion model to obtain the soil moisture content of the monitored area.

[0011] Preferably, the satellite remote sensing data of the monitored area is spectrally reconstructed based on the UAV remote sensing data of the local area to obtain reconstructed satellite spectral data of the satellite remote sensing data, including:

[0012] Upscale the drone remote sensing data of the local area to determine the drone spectral data of the area to be monitored;

[0013] Determine satellite spectral data of the area to be monitored based on satellite remote sensing data of the area to be monitored;

[0014] The satellite spectral data is reconstructed through the UAV spectral data to obtain reconstructed satellite spectral data.

[0015] Preferably, the spectral data includes band reflectance and spectral index; the satellite spectral data is reconstructed through the UAV spectral data to obtain reconstructed satellite spectral data, including:

[0016] The satellite band reflectivity is reconstructed through the UAV spectral data to obtain the reconstructed satellite band reflectivity;

[0017] The reconstructed satellite spectral index is determined according to the reconstructed satellite band reflectivity and the preset conversion function.

[0018] Preferably, the calculation method of reconstructing the satellite band reflectivity is:

[0019] B r =B o {1+k·[FVC / (B u ·B o )]}

[0020] Among them, B r is the reconstructed satellite band reflectivity; B0 is the satellite band reflectivity; FVC is the fractional vegetation coverage corresponding to the spatial resolution scale of the optical satellite; k is the adjustment factor, B u It is the upscaled UAV spectral data, including the upscaled UAV band reflectance and spectral index.

[0021] Preferably, the process of constructing the soil moisture inversion model includes:

[0022] Obtain original satellite training sample data and drone training sample data, as well as the corresponding soil moisture content;

[0023] Determine original satellite sample spectral data according to original satellite training sample data;

[0024] The original satellite sample spectral data is spectrally reconstructed according to the UAV training sample data to obtain the reconstructed satellite sample spectral data;

[0025] The extreme learning machine model is trained by reconstructing the satellite sample spectral data to obtain the soil moisture inversion model.

[0026] Preferably, the method further comprises:

[0027] Before training the soil moisture inversion model, the original satellite sample spectral data and the reconstructed satellite sample spectral data were correlated with the soil moisture content.

[0028] Preferably, the method further comprises:

[0029] The extreme learning machine model is trained through the original satellite sample spectral data to obtain the soil moisture inversion model under the original satellite sample spectral data;

[0030] The extreme learning machine model is trained by reconstructing the satellite sample spectral data to obtain the soil moisture inversion model under the reconstructed satellite sample spectral data;

[0031] The performance tests were carried out on the soil moisture inversion model under the original satellite sample spectral data and the soil moisture inversion model under the reconstructed satellite sample spectral data, and the original performance test results of the soil moisture inversion model under the original satellite sample spectral data and the performance test results of the soil moisture inversion model under the reconstructed satellite sample spectral data were obtained.

[0032] The present invention provides a regional soil moisture monitoring device using airborne spectral reconstruction optical satellite remote sensing, comprising:

[0033] An acquisition module is used to acquire satellite remote sensing data of the area to be monitored and collect drone remote sensing data of a local area in the area to be monitored;

[0034] A reconstruction module is used to perform spectral reconstruction of the satellite remote sensing data of the monitored area based on the UAV remote sensing data of the local area, so as to obtain reconstructed satellite spectral data of the satellite remote sensing data;

[0035] The monitoring module is used to input the reconstructed satellite spectral data into a pre-trained soil moisture inversion model to obtain the soil moisture content of the area to be monitored.

[0036] The present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing is implemented.

[0037] The present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing is implemented.

[0038] At least one of the above technical solutions adopted by the present invention can achieve the following beneficial effects:

[0039] The satellite remote sensing data of the monitored area is spectrally reconstructed through the UAV remote sensing data of the local area in the monitored area, so that the reconstructed satellite spectral data is integrated into the UAV remote sensing data of the local area. In this way, the soil moisture content of the monitored area determined by the reconstructed satellite spectral data will be more accurate and timeliness is guaranteed, thereby realizing accurate and rapid soil moisture monitoring in the monitored area. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0041] Figure 1 A schematic flow chart of a method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing provided by the present invention;

[0042] Figure 2 A schematic flow chart of another method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing provided by the present invention;

[0043] Figure 3 A schematic flow chart of another method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing provided by the present invention;

[0044] Figure 4 A matrix schematic diagram of the correlation analysis between the original and reconstructed band reflectance and spectral index and soil moisture content provided by the present invention;

[0045] Figure 5 A scatter point accuracy comparison diagram of a soil moisture inversion model under original and reconstructed satellite spectral data provided by the present invention;

[0046] Figure 6 A schematic diagram of a regional soil moisture monitoring device using airborne spectral reconstruction optical satellite remote sensing provided by the present invention;

[0047] Figure 7 A schematic diagram of a computer device for implementing a method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing provided by the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the specific embodiments of the present invention and the corresponding drawings. 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 making creative work are within the scope of protection of the present invention.

[0049] The multi-scale fusion of satellite and UAV optical remote sensing data can greatly improve the accuracy of satellite remote sensing monitoring of soil moisture. However, for large-scale irrigation areas, it is time-consuming and laborious to collect UAV remote sensing data covering the entire irrigation area, and the timeliness of soil moisture monitoring products cannot be guaranteed.

[0050] Therefore, the present invention provides a regional soil moisture monitoring method based on airborne spectral reconstruction optical satellite remote sensing. The method integrates the optical remote sensing data of unmanned aerial vehicles in the local area to reconstruct the reflectivity and spectral index of the satellite remote sensing band, so as to improve the accuracy of satellite remote sensing monitoring of soil moisture and realize high temporal and spatial resolution and high precision monitoring of soil moisture in a large range of irrigation areas.

[0051] The technical solutions provided by various embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0052] Figure 1 This is a flow chart of a method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing in the present invention, which specifically includes the following steps:

[0053] S101, acquiring satellite remote sensing data of the area to be monitored and collecting drone remote sensing data of a local area in the area to be monitored.

[0054] The area to be monitored is an area where soil moisture is to be monitored, and the local area may be a representative area of ​​soil moisture that is preset in the area to be monitored.

[0055] The remote sensing data of the area to be monitored can be directly collected by satellite, and the remote sensing data of the local area in the area to be monitored can be collected by drone. After the satellite collects the remote sensing data of the area to be monitored, it sends it to the server to obtain the satellite remote sensing data of the area to be monitored; after the drone collects the remote sensing data of the local area in the area to be monitored, it can send it to the server to obtain the drone remote sensing data of the local area in the area to be monitored.

[0056] The server mentioned in the present invention may be a server set on a business platform, or a device such as a desktop computer, a notebook computer, etc. that can execute the solution of the present invention. For the convenience of description, the following description is only based on the server as the execution subject.

[0057] S102, performing spectral reconstruction on the satellite remote sensing data of the monitored area according to the UAV remote sensing data of the local area, to obtain reconstructed satellite spectral data of the satellite remote sensing data.

[0058] Among them, Figure 2 As shown, according to the UAV remote sensing data of the local area, the satellite remote sensing data of the monitored area is spectrally reconstructed to obtain the reconstructed satellite spectral data of the satellite remote sensing data, including the following steps:

[0059] S201, upscaling the UAV remote sensing data of the local area to determine the UAV spectral data of the area to be monitored.

[0060] Among them, the UAV remote sensing data of the local area can be converted into the UAV spectral data of the area to be monitored by upscaling conversion. The upscaling factor can be determined according to the needs, for example, from 5 cm resolution to 10 m resolution.

[0061] Optionally, the drone remote sensing data of the local area can also be input into the first learning model, and the drone spectral data of the area to be monitored can be obtained by upscaling through the first learning model. The drone spectral data includes the spectral data of all areas in the area to be monitored; specifically, in the first learning model, the drone spectral data of the local area can be determined based on the drone remote sensing data of the local area, and then the drone spectral data of other areas of the area to be monitored can be learned from the drone spectral data of the local area to obtain the drone spectral data of other areas of the area to be monitored, thereby obtaining the drone spectral data of all areas in the area to be monitored.

[0062] Alternatively, the drone remote sensing data of the local area can be input into the second learning model, and the drone spectral data of the area to be monitored can be obtained by upscaling the data through the second learning model. The drone spectral data includes the spectral data of all areas in the area to be monitored; specifically, in the second learning model, other areas of the area to be monitored are first learned from the drone remote sensing data of the local area to obtain the drone remote sensing data of other areas of the area to be monitored, thereby obtaining the drone remote sensing data of all areas in the area to be monitored, and the drone remote sensing data of all areas in the area to be monitored are analyzed to obtain the drone spectral data of the area to be monitored.

[0063] S202, determining satellite spectral data of the area to be monitored based on the satellite remote sensing data of the area to be monitored.

[0064] The satellite remote sensing data of the area to be monitored may include satellite spectral data.

[0065] S203, reconstructing the satellite spectral data through the UAV spectral data to obtain reconstructed satellite spectral data.

[0066] Among them, the spectral data includes band reflectivity and spectral index; the satellite spectral data is reconstructed through the UAV spectral data to obtain reconstructed satellite spectral data, including: reconstructing the satellite band reflectivity through the UAV spectral data to obtain the reconstructed satellite band reflectivity; determining the reconstructed satellite spectral index according to the reconstructed satellite band reflectivity and a preset conversion function.

[0067] Among them, the calculation method of reconstructing the satellite band reflectivity is:

[0068] B r =B o {1+k·[FVC / (B u ·B o )]} (1)

[0069] Among them, B r is the reconstructed satellite band reflectivity; B0 is the satellite band reflectivity; FVC is the fractional vegetation coverage corresponding to the spatial resolution scale of the optical satellite; k is the adjustment factor, B u It is the drone band reflectivity or drone spectral index.

[0070] The reconstructed spectral index is a function of the reconstructed band reflectance, that is, the conversion function is shown in formula (2).

[0071] S rec =f(B r ) (2)

[0072] Among them, S rec is the reconstructed satellite spectral index, f(·) is B r To S rec As a function of , the reconstructed satellite spectral index can include vegetation index, salt index and brightness index.

[0073] For example, as shown in Table 1, Table 1 is a conversion function from band reflectance to spectral index.

[0074] Table 1

[0075]

[0076] Among them, B, R, G, and NIR are the reflectances of the bands corresponding to blue, red, green, and near infrared (NIR), respectively.

[0077] S103, inputting the reconstructed satellite spectral data into a pre-trained soil moisture inversion model to obtain the soil moisture content of the area to be monitored.

[0078] The reconstructed satellite band reflectivity and the reconstructed satellite spectral index are input into the pre-trained soil moisture inversion model to obtain the soil moisture content of the monitored area output by the soil moisture inversion model.

[0079] Preferably, the construction process of the soil moisture inversion model includes: obtaining original satellite training sample data and unmanned aerial vehicle training sample data, and the corresponding soil moisture content; determining the original satellite sample spectral data according to the original satellite training sample data, and spectrally reconstructing the original satellite sample spectral data according to the unmanned aerial vehicle training sample data to obtain reconstructed satellite sample spectral data; training the extreme learning machine model by reconstructing the satellite sample spectral data to obtain the soil moisture inversion model.

[0080] Specifically, local representative (different vegetation coverage conditions) UAV multispectral remote sensing data, optical satellite remote sensing images and soil moisture data can be obtained first and preprocessed. Among them, the measured soil moisture content and UAV multispectral remote sensing data are obtained by conducting field experiments, and the satellite remote sensing data is obtained by downloading from the official website of satellite remote sensing data. The number of pixels upscaled from the UAV to the satellite remote sensing scale is not less than 1,000, and the number of ground soil moisture sampling is not less than 120. The UAV multispectral remote sensing data is used as the UAV training sample data, and the optical satellite remote sensing image is used as the satellite training sample data.

[0081] First, the satellite training sample data is spectrally reconstructed according to the UAV training sample data to obtain reconstructed satellite sample spectral data; it should be noted that the method of obtaining the reconstructed satellite sample spectral data is the same as the method of obtaining the reconstructed satellite spectral data of the satellite remote sensing data in the above embodiment, and this embodiment will not be repeated here. The reconstructed satellite sample spectral data includes the reconstructed satellite sample band reflectance and the reconstructed satellite sample spectral index.

[0082] The k value in the reconstructed satellite sample band reflectance, that is, the k value in formula (1), is determined by constructing a linear correlation function r(k). The linear correlation function is constructed by formula (3)-formula (6). Formula (3) is the calculation of the linear correlation coefficient between the reconstructed satellite sample band reflectance and the corresponding soil moisture content.

[0083]

[0084] Among them, S m is the soil mass moisture content.

[0085]

[0086] According to formula (4), formula (3) can be converted into formula (5).

[0087]

[0088] When the sampling number n is given, M0, M1, M2, M3, M4, and M5 are also determined. At this time, they all become constants, and equation (5) is expressed as equation (6):

[0089]

[0090] Where r(k) is the linear correlation function between the reconstructed band reflectance and soil moisture content, r(k)∈[-1,1].

[0091] Differentiating both sides of equation (6) with respect to k, we can obtain equation (7):

[0092]

[0093] Equation (7) has an extreme value (maximum value) of the correlation r(k). Let dr / dk=0 in equation (7), and we have equation (8):

[0094]

[0095] For the linear correlation function r(k), there are four special cases: (i), (ii), (iii) and (iv).

[0096] (i) When k = (M1M4-M2M5) / (M2M4-M1M3) in equation (6), we have equation (9):

[0097]

[0098] Among them, (k e , r e ) is the only extreme point of the function r(k) in the real number range, r e is the unique minimum value of the function r(k).

[0099] (ii) When k=0 in equation (6), we have equation (10):

[0100]

[0101] Among them, r O Indicates B o With S m The linear correlation coefficient of the original satellite band reflectivity and soil moisture.

[0102] (iii) When k = -M1 / M2 in equation (6), r(k) = 0, and there is no linear correlation between the reconstructed satellite band reflectivity and soil moisture.

[0103] (iv) When k→∞ in equation (6), we have equation (11):

[0104]

[0105] Among them, r ∞Indicates B r With S m Linear correlation coefficient at positive and negative infinity.

[0106] The reconstructed satellite sample spectral data includes the reconstructed satellite sample band reflectance and the reconstructed satellite sample spectral index; k can be calculated according to formula (8): m , the original satellite band reflectance B0 corresponding to the satellite training sample data, the vegetation coverage FVC and the drone multispectral band reflectance or spectral index B corresponding to the upscaled drone training sample data u The reconstructed satellite sample band reflectance is calculated according to formula (1), and the reconstructed satellite sample spectral index is calculated based on the reconstructed satellite sample band reflectance.

[0107] After obtaining the reconstructed satellite sample spectral data, the extreme learning machine model is trained by reconstructing the satellite sample spectral data to obtain a soil moisture inversion model under the reconstructed satellite sample spectral data. Specifically, a machine learning algorithm such as an extreme learning machine is used to construct a soil moisture inversion regression model for reconstructed satellite sample spectral data to obtain a soil moisture inversion model.

[0108] In an exemplary embodiment, the embodiment includes: before training the soil moisture inversion model, performing correlation analysis on the original satellite sample spectral data and the reconstructed satellite sample spectral data and the soil moisture content, respectively, wherein the correlation analysis adopts Pearson correlation.

[0109] In addition, the extreme learning machine model is trained by the original satellite sample spectral data to obtain the soil moisture inversion model under the original satellite sample spectral data; the extreme learning machine model is trained by the reconstructed satellite sample spectral data to obtain the soil moisture inversion model under the reconstructed satellite sample spectral data; the soil moisture inversion model under the original satellite sample spectral data and the soil moisture inversion model under the reconstructed satellite sample spectral data are performance tested to obtain the original performance test results of the soil moisture inversion model under the original satellite sample spectral data and the performance test results of the soil moisture inversion model under the reconstructed satellite sample spectral data.

[0110] Among them, the soil moisture inversion model under the original satellite sample spectral data and the soil moisture inversion model under the reconstructed satellite sample spectral data are an inversion regression model.

[0111] Optionally, the coefficient of determination (R 2 ) and root mean square error (RMSE) were used to evaluate the accuracy of the soil moisture inversion model constructed using the original satellite sample spectral data and the soil moisture inversion model constructed using the reconstructed satellite sample spectral data, respectively, and R 2The increase or decrease of RMSE is used to evaluate the improvement effect of the soil moisture inversion model under the reconstructed satellite sample spectral data compared with the soil moisture inversion model under the original satellite sample spectral data.

[0112] In an exemplary embodiment, Figure 3 As shown, the present invention also provides a regional soil moisture monitoring method of airborne spectral reconstruction optical satellite remote sensing, specifically, obtaining satellite images of the sample area collected by the satellite, and then preprocessing the satellite images to obtain original spectral data, and obtaining drone images of the sample area collected by a drone, and preprocessing the drone images to obtain preprocessed drone image data, and upscaling the drone image processing to obtain upscaled drone multispectral data.

[0113] Then, the original satellite spectral data was reconstructed through the upscaled UAV multispectral data to obtain the reconstructed satellite spectral data. The original satellite spectral data and the reconstructed spectral data were respectively correlated with the measured soil moisture content. The original satellite spectral data and the reconstructed satellite spectral data were used as independent variables, and the measured soil moisture content was used as the dependent variable. The soil moisture inversion model under the original satellite spectral data and the soil moisture inversion model under the reconstructed satellite spectral data were trained. Then, the evaluation index (R 2 The soil moisture inversion model under the original satellite spectral data and the soil moisture inversion model under the reconstructed satellite spectral data were evaluated by using the mean square error (RMSE) and RMSE to determine that the soil moisture inversion model under the reconstructed satellite spectral data has better performance.

[0114] The regional soil moisture monitoring method using airborne spectral reconstruction optical satellite remote sensing provided by the present invention is further described below through a specific embodiment.

[0115] Firstly, the local (four plots with different vegetation coverage were selected) unmanned aerial vehicle multispectral remote sensing data obtained from the Jiefangzha Shahaoqu irrigation area in the Hetao irrigation area of ​​site A, the measured soil moisture data on four typical plots collected simultaneously, and the downloaded cloud-free Landsat 8 satellite multispectral remote sensing data corresponding to the study area were used to construct an irrigation area soil moisture inversion model based on the original satellite spectral data and an irrigation area soil moisture inversion model based on the satellite spectral data reconstructed by airborne spectrum, respectively. The inversion accuracy and improvement effect of each model were compared through evaluation indicators.

[0116] Step 1: Soil sample collection and drone flight tests were carried out in four selected typical fields from August 12 to 15, 2019. The visible light bands of the drone multispectral data, namely blue, green, and red, were used for upscaling. The five-point sampling method was used to collect soil samples at a depth of 0-10 cm. The mass moisture content S of the soil was calculated by the drying method (105°C, 24h constant temperature treatment). m , the calculation formula is:

[0117]

[0118] Among them, m1 is the weight of wet soil plus empty aluminum box, m2 is the weight of dry soil plus empty aluminum box, and m3 is the weight of the empty aluminum box.

[0119] Download Landsat 8 multispectral satellite remote sensing data from USGS Earth Explorer. The acquired Landsat 8 satellite image is a Collection 2 Level-1 product that has been geometrically corrected. The imaging time of the image is August 15, 2019, which is close to the time of field sample collection. The Landsat 8 satellite image data is further processed by ENVI software, including radiation calibration, atmospheric correction, clipping, etc., and then the reflectance data of each band at the sampling point is extracted. In the present invention, the blue (Blue), green (Green), red (Red), near infrared (NIR), shortwave infrared-1 (SWIR1) and shortwave infrared-2 (SWIR2) of Landsat 8 optical satellite data are selected for reconstruction.

[0120] Step 2: FVC is obtained by extracting the proportion of vegetation-like elements in high-resolution UAV multispectral images at a given scale (30 m). The calculation formula is shown below.

[0121]

[0122] Where: N v and N no-v They are the number of vegetation and non-vegetation pixels corresponding to the spatial resolution scale of the Landsat 8 satellite.

[0123] B u The S3 salt index calculated by upscaling UAV remote sensing data is used in the present invention, that is, S3 u , the calculation formula is as follows:

[0124] S3 u =Green u ×Red u / Blue u (14)

[0125] Among them, Blue u 、Green u and Red u They are the reflectances of blue, green and red bands under UAV remote sensing data that were upscaled to 30m spatial resolution. The upscaling method used is the pixel aggregation method.

[0126] The k value formula at the extreme point of the linear r(k) function is used to calculate the k value of each reconstructed band reflectivity. m Value, i.e. k m-Blue , k m-Green , k m-Red , k m-NIR , k m-SWIR1 and k m-SWIR2 , and finally the reflectivity of each reconstructed band is obtained, as shown in formulas (15)-(20).

[0127] Blue re =Blue{1+k m-Blue ·[FVC / (S3 u ·Blue)]} (15)

[0128] Green re =Green{1+k m-Green ·[FVC / (S3 u Green)]} (16)

[0129] Red re =Red{1+k m-Red ·[FVC / (S3 u ·Red)]} (17)

[0130] NIR re =NIR{1+k m-NIR ·[FVC / (S3 u ·NIR)]} (18)

[0131] SWIR1 re =SWIR1{1+k m-SWIR1 ·[FVC / (S3 u ·SWIR1)]} (19)

[0132] SWIR2 re =SWIR2{1+k m-SWIR2 ·[FVC / (S3 u ·SWIR2)]} (20)

[0133] Step 3: Based on the band reflectance of the original Landsat 8 optical satellite data, 8 original spectral indices (including 4 salt indices, 3 vegetation indices and 1 brightness index) are constructed; based on the band reflectance of the reconstructed Landsat 8 optical satellite data, 8 reconstructed spectral indices (including 4 reconstructed salt indices, 3 reconstructed vegetation indices and 1 reconstructed brightness index) are constructed. The specific calculation formula is shown in Table 2:

[0134] Table 2

[0135]

[0136]

[0137] Among them, B, G, B and NIR represent the original blue, green, red and near-infrared band reflectances, that is, B = Blue, G = Green, R = Red; B re , G re , R re and NIR re Represent the reconstructed blue, green, red and near-infrared reflectance, that is, B re =Blue re , G re =Green re , R re =Red re .

[0138] Step 4: Correlation analysis is performed on the measured soil moisture content with the original and reconstructed satellite band reflectance and spectral index. The results are presented in the Pearson correlation matrix. Figure 4 , where the larger the circle, the stronger the correlation. Figure 4 It can be seen that compared with the original satellite band reflectivity and spectral index, the correlation between the reconstructed satellite band reflectivity and spectral index and soil moisture content is significantly enhanced.

[0139] Step 5, extreme learning machine is an artificial neural network model training algorithm, which usually consists of an input layer, a hidden layer and an output layer. During the model training process, the connection weights of the input layer and the hidden layer are random and do not need to be adjusted. Only the number of hidden layer nodes of the network needs to be set to generate the global optimal solution. In the present invention, the number of hidden layer nodes is uniformly set to 20, and the construction of the extreme learning machine model is completed by MATLAB software. The ratio of the number of samples in the modeling set and the verification set is set to 2:1, that is, the number of samples in the modeling set is 79, and the number of samples in the verification set is 40. Taking all the original band reflectances and spectral indices as independent variables and soil moisture content as the dependent variable, the extreme learning machine regression algorithm is used to construct the soil moisture inversion model under the original satellite data; taking all the reconstructed band reflectances and spectral indices as independent variables and soil moisture content as the dependent variable, the extreme learning machine regression algorithm is used to construct the soil moisture inversion model under the reconstructed satellite data. The results are as follows. Figure 5 shown.

[0140] Step 6: The coefficient of determination can reflect the fitting effect of the model. 2 The closer it is to 1, the smaller the RMSE is, indicating that the model prediction effect is better and the error between the predicted value and the measured value is smaller. 2 and RMSE, as shown in Table 3.

[0141]

[0142] Among them, y i represents the measured value, represents the predicted value, represents the average of the measured values, and n represents the number of samples.

[0143] Table 3

[0144]

[0145] It should be noted that the results in Table 3 show that the R of the extreme learning machine model constructed using the reconstructed satellite spectral data is lower than that of the extreme learning machine model constructed using the original satellite spectral data. 2 And the increase or decrease of RMSE; + means increase, - means decrease.

[0146] Combination Figure 5 As shown in Table 3, the soil moisture inversion model validation set R constructed using only the original satellite data is 2 is 0.441, and the RMSE is 0.034 cm 3 / cm 3 In contrast, the soil moisture inversion model validation set R based on the reconstructed satellite data constructed using the extreme learning machine regression algorithm 2 is 0.663, and the RMSE is 0.025 cm 3 / cm 3 . Validation set R 2 The increase was 0.222 (about 50.34%), and the RMSE was reduced by 0.009cm 3 / cm 3 (The decrease was about 26.47%). It can be seen that the integration of local area drone multispectral data can effectively improve the accuracy of the soil moisture model inverted by a single satellite optical remote sensing data, improve the accuracy of satellite remote sensing monitoring of soil moisture, and achieve high temporal and spatial resolution and high-precision monitoring of soil moisture in a large range of irrigation areas. The present invention can provide a theoretical basis and technical support for improving the accuracy of regional soil moisture monitoring based on satellite remote sensing spectral data.

[0147] When using the regional soil moisture monitoring method for airborne spectral reconstruction optical satellite remote sensing provided by the present invention, it is not necessary to Figure 1 The steps are executed in the order shown. The specific execution order of the steps can be determined according to needs, and the present invention does not limit this.

[0148] The above is a method for monitoring regional soil moisture using airborne spectral reconstruction optical satellite remote sensing provided by one or more embodiments of the present invention. Based on the same idea, the present invention also provides a corresponding device for monitoring regional soil moisture using airborne spectral reconstruction optical satellite remote sensing, such as Figure 6 shown.

[0149] Figure 6 A schematic diagram of a regional soil moisture monitoring device using airborne spectral reconstruction optical satellite remote sensing provided by the present invention is shown in FIG. 60 . The device 600 includes:

[0150] An acquisition module 601 is used to acquire satellite remote sensing data of the area to be monitored and collect drone remote sensing data of a local area in the area to be monitored;

[0151] The reconstruction module 602 is used to perform spectral reconstruction on the satellite remote sensing data of the monitored area according to the UAV remote sensing data of the local area, so as to obtain reconstructed satellite spectral data of the satellite remote sensing data;

[0152] The monitoring module 603 is used to input the reconstructed satellite spectral data into a pre-trained soil moisture inversion model to obtain the soil moisture content of the area to be monitored.

[0153] For the specific definition of the regional soil moisture monitoring device of airborne spectral reconstruction optical satellite remote sensing, please refer to the definition of the regional soil moisture monitoring method of airborne spectral reconstruction optical satellite remote sensing above, which will not be repeated here. Each module in the above-mentioned regional soil moisture monitoring device of airborne spectral reconstruction optical satellite remote sensing can be fully or partially implemented by software, hardware and their combination. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0154] The present invention also provides a computer-readable storage medium, which stores a computer program, which can be used to execute the above Figure 1 A regional soil moisture monitoring method based on airborne spectral reconstruction of optical satellite remote sensing is provided.

[0155] The present invention also provides Figure 7 The structural diagram of the computer device shown in FIG. Figure 7 As shown in the figure, at the hardware level, the computer device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 1 A regional soil moisture monitoring method based on airborne spectral reconstruction of optical satellite remote sensing is provided.

[0156] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0157] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of the present invention.

Claims

1. A method for regional soil moisture monitoring using airborne spectral reconstruction optical satellite remote sensing, characterized in that: The method comprises: Acquiring satellite remote sensing data of the area to be monitored and collecting drone remote sensing data of a local area of ​​the area to be monitored; According to the UAV remote sensing data of the local area, spectrally reconstructing the satellite remote sensing data of the monitored area to obtain reconstructed satellite spectral data of the satellite remote sensing data; The reconstructed satellite spectral data is input into a pre-trained soil moisture inversion model to obtain the soil moisture content of the area to be monitored.

2. The method according to claim 1, characterized in that The step of reconstructing the spectrum of the satellite remote sensing data of the monitored area according to the UAV remote sensing data of the local area to obtain the reconstructed satellite spectrum data of the satellite remote sensing data includes: Upscaling the UAV remote sensing data of the local area to determine the UAV spectral data of the area to be monitored; Determining satellite spectral data of the area to be monitored based on the satellite remote sensing data of the area to be monitored; The satellite spectral data is reconstructed using the UAV spectral data to obtain the reconstructed satellite spectral data.

3. The method according to claim 2, characterized in that The spectral data includes band reflectance and spectral index; the satellite spectral data is reconstructed by using the unmanned aerial vehicle spectral data to obtain the reconstructed satellite spectral data, including: Reconstructing the satellite band reflectivity through the UAV spectral data to obtain a reconstructed satellite band reflectivity; The reconstructed satellite spectrum index is determined according to the reconstructed satellite band reflectivity and a preset conversion function.

4. The method according to claim 3, characterized in that The calculation method of reconstructing the satellite band reflectivity is: B r =B o {1+k·[FVC / (B u ·B o )]} Among them, B r is the reconstructed satellite band reflectivity; B0 is the satellite band reflectivity; FVC is the fractional vegetation coverage corresponding to the spatial resolution scale of the optical satellite; k is the adjustment factor, B u It is the drone band reflectivity or drone spectral index.

5. The method according to claim 1, characterized in that The construction process of the soil moisture inversion model includes: Obtain original satellite training sample data and drone training sample data, as well as the corresponding soil moisture content; Determine original satellite sample spectral data according to the original satellite training sample data; Performing spectral reconstruction on the original satellite sample spectral data according to the UAV training sample data to obtain reconstructed satellite sample spectral data; The soil moisture inversion model is obtained by training an extreme learning machine model through the reconstructed satellite sample spectral data.

6. The method according to claim 5, characterized in that The method further comprises: Before training the soil moisture inversion model, correlation analysis is performed on the original satellite sample spectral data and the reconstructed satellite sample spectral data with soil moisture content.

7. The method according to claim 5, characterized in that The method further comprises: The extreme learning machine model is trained by the original satellite sample spectral data to obtain a soil moisture inversion model under the original satellite sample spectral data; The extreme learning machine model is trained by the reconstructed satellite sample spectral data to obtain a soil moisture inversion model under the reconstructed satellite sample spectral data; The performance tests are performed on the soil moisture inversion model under the original satellite sample spectral data and the soil moisture inversion model under the reconstructed satellite sample spectral data to obtain the original performance test results of the soil moisture inversion model under the original satellite sample spectral data and the performance test results of the soil moisture inversion model under the reconstructed satellite sample spectral data.

8. A regional soil moisture monitoring device based on airborne spectral reconstruction optical satellite remote sensing, characterized in that: include: An acquisition module is used to acquire satellite remote sensing data of the area to be monitored and collect drone remote sensing data of a local area in the area to be monitored; A reconstruction module, used for spectrally reconstructing the satellite remote sensing data of the monitored area according to the UAV remote sensing data of the local area, so as to obtain reconstructed satellite spectral data of the satellite remote sensing data; The monitoring module is used to input the reconstructed satellite spectral data into a pre-trained soil moisture inversion model to obtain the soil moisture content of the area to be monitored.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

10. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

Images