Methods, apparatus, equipment and media for synergistic inversion of multiple vegetation growth parameters

Abstract

This disclosure relates to a method, apparatus, device, and medium for collaborative inversion of multiple vegetation growth parameters, belonging to the technical field of image processing. The method includes: acquiring spectral remote sensing data of a target vegetation area output by a UAV and measured data of the target vegetation area output by a ground sampling device; aligning the spectral remote sensing data and the measured data with the location parameters of the target vegetation area to obtain an inversion feature set; inputting reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain inversion data corresponding to each vegetation sensitive feature; updating the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature; for the updated inversion feature set, performing the step of inputting reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain inversion data corresponding to each vegetation sensitive feature, until the updated inversion feature set meets the set conditional expression.

Description

Technical Field

[0001] This disclosure relates to the field of image processing technology, and more specifically, to a method, apparatus, device, and medium for the collaborative inversion of multiple growth parameters of vegetation. Background Technology

[0002] Food security is a crucial foundation of national security. Rice, as one of my country's most important food crops, holds an irreplaceable strategic position in ensuring the supply of staple food for residents and maintaining social stability. The formation of rice yield is closely related to its growth status. Key growth parameters such as leaf area index, leaf water content, and chlorophyll content directly reflect the rice canopy structure, physiological activity, and photosynthetic capacity, serving as important bases for assessing rice growth, predicting yield, and guiding field management. Especially during the critical growth stages of rice, timely and accurate monitoring of these growth parameters is essential for achieving precision rice cultivation, disaster early warning, and yield assurance.

[0003] Traditional methods for obtaining rice growth parameters mainly rely on manual field measurements or small-scale experimental observations. Although these methods are highly accurate, they suffer from low operational efficiency, high labor intensity, and insufficient spatial representativeness, making it difficult to meet the needs of large-scale, continuous rice growth monitoring. Summary of the Invention

[0004] One objective of this disclosure is to provide a new technical solution for the collaborative inversion of multiple vegetation growth parameters.

[0005] According to a first aspect of this disclosure, a method for synergistic inversion of multiple vegetation growth parameters is provided, the method comprising: Acquire spectral remote sensing data of the target vegetation area output by the UAV and measured data of the target vegetation area output by the ground sampling equipment; The spectral remote sensing data and the measured data are aligned with the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation sensitive feature; The reference data corresponding to each vegetation sensitive feature is input into the preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature; Update the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature; For the updated inversion feature set, the step of inputting the reference data corresponding to each vegetation sensitive feature into the preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature is performed until the updated inversion feature set meets the set condition formula.

[0006] Optionally, the spectral remote sensing data and the measured data are aligned using the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation-sensitive feature, including: Based on the latitude and longitude information reflected in the measured data, spatially aligned data is obtained by mapping it to the spatial coordinate system associated with the spectral remote sensing data; Based on the spatial alignment data, determine the index characteristics of the target vegetation area; Using the measured data as a reference, the correlation between the spectral remote sensing data and the spectral band, the exponential feature, and the corresponding growth parameter is determined, and each vegetation sensitive feature in the spatially aligned data and the reference data corresponding to each vegetation sensitive feature are obtained.

[0007] Optionally, determining the index characteristics of the target vegetation area based on the spectral remote sensing data includes: Based on the spectral remote sensing data, the following indexes were determined as index characteristics of the target vegetation area: Normalized Difference Vegetation Index, Core Normalized Vegetation Index, Green Normalized Vegetation Index, Normalized Red Edge Index, Enhanced Vegetation Index, Improved Enhanced Vegetation Index, Soil-Adjusted Vegetation Index, Optimized Soil-Adjusted Vegetation Index, Chlorophyll Index, Green Band Chlorophyll Index, Red Edge Band Chlorophyll Index, Ratio Vegetation Index, Improved Chlorophyll Reflectance Absorption Index, and Structure-Insensitive Pigment Index.

[0008] Optionally, the inversion model is a regression model constructed based on at least one of the following algorithms: linear regression, random forest, support vector regression, KNN, GBDT, and XGBoost.

[0009] Optionally, the updated inversion feature set includes a first inversion feature mapped from the inversion data corresponding to the first vegetation sensitive feature and an inherent second inversion feature.

[0010] Optionally, the set condition includes the following: the determination coefficient of the previous inversion feature set is less than the determination coefficient of the subsequent inversion feature set; the root mean square error of the previous inversion feature set is greater than the root mean square error of the subsequent inversion feature set; or the number of iterations corresponding to the subsequent inversion feature set exceeds a set number.

[0011] Optionally, the determination coefficient and root mean square error of any inversion feature set are obtained by using the model prediction value of the corresponding inversion model and the corresponding measured data.

[0012] According to a second aspect of this disclosure, a vegetation multi-growth parameter synergistic inversion device is also provided, the device comprising: The acquisition module is used to acquire spectral remote sensing data of the target vegetation area output by the UAV and measured data of the target vegetation area output by the ground sampling equipment; The first obtaining module is used to align the spectral remote sensing data and the measured data with the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation sensitive feature; The second obtaining module is used to input the reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature; The update module is used to update the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature; The third module is used to perform the step of inputting the reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature for the updated inversion feature set, until the updated inversion feature set meets the set condition formula.

[0013] According to a third aspect of this disclosure, a computer system is also provided, the computer system including a processor, which, when executing program instructions or code, implements the vegetation multi-growth parameter collaborative inversion method of the first aspect.

[0014] For example, the computer system also includes a memory for storing program instructions or code.

[0015] According to a fourth aspect of this disclosure, a computer-readable storage medium is also provided, wherein a computer program is stored therein, wherein the computer program is configured to execute the above-described vegetation multi-growth parameter collaborative inversion method at runtime.

[0016] According to a fifth aspect of this disclosure, a computer program product is also provided, comprising a computer program that, when executed, causes a computer to perform the steps of the above-described vegetation multi-growth parameter collaborative inversion method.

[0017] According to a sixth aspect of this disclosure, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to execute the above-described vegetation multi-growth parameter collaborative inversion method through the computer program.

[0018] One beneficial effect of this disclosure is that the vegetation multi-growth parameter collaborative inversion method provided by the present invention can acquire spectral remote sensing data of the target vegetation area output by a UAV and measured data of the target vegetation area output by a ground sampling device; align the spectral remote sensing data and the measured data with the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation sensitive feature; input the reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain inversion data corresponding to each vegetation sensitive feature; update the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature; for the updated inversion feature set, execute the step of inputting the reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain inversion data corresponding to each vegetation sensitive feature, until the updated inversion feature set meets the set conditional expression. By acquiring spectral remote sensing data output by a UAV and measured data output by a ground sampling device, and constructing an inversion feature set by aligning the data with location parameters, it is possible to effectively integrate multi-source, multi-scale observation information and improve the problem of insufficient information from a single data source.

[0019] Other features and advantages of the embodiments of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0020] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments of the present disclosure and, together with their description, serve to explain the principles of the embodiments of the present disclosure.

[0021] Figure 1 A flowchart illustrating a method for synergistic inversion of multiple vegetation growth parameters according to some embodiments is shown. Figure 2 A flowchart of a vegetation multi-growth parameter collaborative inversion method according to some embodiments is shown; Figure 3 A schematic diagram of a vegetation multi-growth parameter collaborative inversion device according to some embodiments is shown; Figure 4 A schematic diagram of the hardware structure of an electronic device according to some embodiments is shown. Detailed Implementation

[0022] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention.

[0023] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use.

[0024] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0025] In all the examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0026] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0027] <Method Implementation> Figure 1 This is a flowchart illustrating a method for the coordinated inversion of multiple vegetation growth parameters according to one embodiment. The implementing entity is an electronic device, such as a personal computer, server, or dedicated image processing equipment.

[0028] like Figure 1 As shown, the vegetation multi-growth parameter synergistic inversion method of this embodiment may include the following steps S110 to S150: Step S110: Acquire spectral remote sensing data of the target vegetation area output by the UAV and measured data of the target vegetation area output by the ground sampling equipment.

[0029] In this embodiment, as Figure 2 As shown, spectral remote sensing data refers to imagery data collected by a multispectral sensor mounted on a UAV over a target vegetated area (such as a rice paddy). This spectral remote sensing data includes at least red, green, blue, near-infrared, and red-edge bands. Measured data refers to real vegetation growth parameters measured simultaneously in the same area by ground sampling equipment (such as a handheld photosynthesis meter or leaf area index meter), such as leaf area index (LAI), chlorophyll content (LCC), and leaf water content (LWC).

[0030] For example, in step S110, to obtain high-quality, multi-temporal data sources, multiple aerial observations can be conducted during the key growth stages of the target vegetation. For instance, during the early growth stage to early tillering stage of rice from July to August 2025, a drone platform equipped with a multispectral sensor can be used to conduct regular aerial observations of the experimental field, acquiring two sets of image data per month to form a multi-temporal dataset. Simultaneously, ground data collection is carried out at each drone aerial sampling point: leaf area index is measured in situ using an LAI-2200 vegetation canopy analyzer; chlorophyll content is measured using a handheld SPAD chlorophyll meter; and leaf water content is calculated by collecting leaf samples, drying them in the laboratory, and weighing them. Through these methods, high-quality paired data for model construction and validation can be obtained.

[0031] Step S120: Align the spectral remote sensing data and the measured data with the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation sensitive feature.

[0032] In this embodiment, the location parameter refers to the latitude and longitude coordinate information used for spatial registration. The purpose of data alignment is to establish a one-to-one correspondence between pixels in UAV imagery and measured values ​​of ground samples in space. The inversion feature set is a collection of various remote sensing features sensitive to vegetation growth status (such as reflectance in specific bands, vegetation indices, etc.) and their corresponding ground-measured reference data.

[0033] In this embodiment, the expression for the inverted feature set is as follows: in, This represents the spectral remote sensing data corresponding to the i-th sample point, including spectral bands and vegetation indices; The measured data represents the i-th sample point; .

[0034] For example, in order to reduce the impact of single-pixel noise and positioning error, during data alignment, all pixel values ​​within a preset window (such as 5×5 pixels) around the location of the ground sample point can be extracted and their average value can be calculated as the remote sensing feature value corresponding to the sample point, thereby constructing a more robust inversion feature set.

[0035] In some embodiments of this disclosure, step S120, which involves aligning the spectral remote sensing data and the measured data using the location parameters of the target vegetation area to obtain an inversion feature set, may specifically include the following steps S1210 to S1230: Step S1210: Map the latitude and longitude information reflected by the measured data to the spatial coordinate system associated with the spectral remote sensing data to obtain spatially aligned data.

[0036] For example, a handheld RTK positioning device is used to acquire high-precision latitude and longitude coordinates of ground sampling points, and these coordinates are uniformly converted to a spatial coordinate reference system consistent with the UAV remote sensing image. Through spatial mapping, it is ensured that each ground sampling point can find its precise corresponding position on the remote sensing image, thereby extracting the spectral information of that position and generating spatially aligned data.

[0037] Step S1220: Determine the index characteristics of the target vegetation area based on the spatial alignment data.

[0038] For example, the index features can be multiple vegetation indices. For instance, based on spectral remote sensing data, the following can be determined as index features for a target vegetation area: Normalized Difference Vegetation Index (NDVI), Kernel Normalized Vegetation Index (kNDVI), Green Normalized Vegetation Index (GNDVI), Normalized Red Edge Index (NDRE), Enhanced Vegetation Index (EVI), Modified Enhanced Vegetation Index (EVI2), Soil Adjusted Vegetation Index (SAVI), Optimized Soil Adjusted Vegetation Index (OSAVI), Chlorophyll Index (LCI), Green Band Chlorophyll Index (CIgreen), Red Edge Chlorophyll Index (CIrededge), Ratio Vegetation Index (RVI), Modified Chlorophyll Reflectance Absorption Index (MCARI), and Structure Insensitive Pigment Index (SIPI).

[0039] In some examples, the expression for the Normalized Difference Vegetation Index (NDVI) is as follows: The expression for the nuclear normalized vegetation index (kNDVI) is as follows: The expression for the Green Normalized Difference Vegetation Index (GNDVI) is as follows: The expression for the Normalized Red Edge Index (NDRE) is as follows: The expression for the Enhanced Vegetation Index (EVI) is as follows: The improved Enhanced Vegetation Index (EVI2) is expressed as follows: The expression for the Soil Adjusted Vegetation Index (SAVI) is as follows: The optimized soil-adjusted vegetation index (OSAVI) is expressed as follows: The expression for the chlorophyll index (LCI) is as follows: The expression for the green band chlorophyll index (CIgreen) is as follows: The expression for the red-edge chlorophyll index (CIrededge) is as follows: The expression for the Ratio Vegetation Index (RVI) is as follows: The expression for the improved chlorophyll reflectance absorption index (MCARI) is as follows: The expression for the Structure Insensitive Pigment Index (SIPI) is as follows: in, , , , , express .

[0040] In these examples, by calculating multiple vegetation indices with different physiological meanings, information such as canopy structure, chlorophyll content, and water status of vegetation can be captured more comprehensively. For example, NDVI is sensitive to vegetation cover, while CIrededge is more sensitive to changes in chlorophyll content. The combination of multiple indices can provide a richer candidate set for subsequent feature selection.

[0041] Step S1230: Using the measured data as a reference, determine the correlation between the spectral remote sensing data and the spectral band, the exponential feature, and the corresponding growth parameter, and filter out each vegetation sensitive feature in the spatially aligned data and the reference data corresponding to each vegetation sensitive feature.

[0042] For example, Pearson correlation coefficient analysis can be used to calculate the correlation between each spectral band and each vegetation index and the measured leaf area index, chlorophyll content, and leaf water content on the ground. Features with high correlation to the target growth parameters are selected as candidate features. Based on this, the correlation between the candidate features is further analyzed, eliminating redundant features that are highly correlated with each other (e.g., correlation coefficient greater than 0.9). Finally, features that are representative of the target growth parameters and highly complementary are retained as the sensitive features for each vegetation species. This two-step feature optimization strategy can effectively avoid multicollinearity while reducing feature dimensionality, thereby improving the stability and generalization ability of the subsequent inversion model.

[0043] Step S130: Input the reference data corresponding to each vegetation sensitive feature into the preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature.

[0044] In this embodiment, the inversion model is a regression model built based on a machine learning algorithm. The reference data are spectral or exponential feature values ​​that are sensitive to specific growth parameters after screening, while the inversion data are vegetation growth parameter values ​​(such as LAI, LCC, etc.) predicted by the model.

[0045] For example, the inversion model can be a regression model built based on at least one of the following algorithms: linear regression, random forest, support vector regression, KNN, GBDT, and XGBoost. The expression of the regression model is as follows: in, The representative regression model can be linear regression, random forest, support vector regression, KNN, GBDT, XGBoost, etc. These are the model parameters for the corresponding inversion model.

[0046] In practical applications, a 4-fold cross-validation strategy can be used to divide the sample data, allowing for separate training and performance evaluation of different algorithms. For example, by comparing the performance of random forest regression and XGBoost in predicting leaf area index, the algorithm with higher R² and lower RMSE can be selected as the final initial inversion model for that growth parameter. In this way, the optimal machine learning model can be adaptively selected based on the response characteristics of different growth parameters.

[0047] Step S140: Update the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature.

[0048] In this embodiment, the first vegetation-sensitive feature can be a parameter selected from multiple growth parameters (such as LAI, LCC, LWC). Updating the inversion feature set refers to adding the inversion data of the first vegetation-sensitive feature predicted by the model as a new feature dimension to the original inversion feature set for inversion of other growth parameters. The updated inversion feature set includes a first inversion feature mapped from the inversion data corresponding to the first vegetation-sensitive feature and an inherent second inversion feature.

[0049] In this embodiment, the first reflecting feature The expression is as follows: For example, if the first vegetation-sensitive feature is the leaf area index, its initial inversion result is added as a new enhancement feature to the feature set used for inverting chlorophyll content. In this way, the feature set, which originally only contained spectra and vegetation indices, allows the model to make more accurate predictions by utilizing the physiological coupling relationship between different growth parameters (canopy structure affects photosynthetic pigment content).

[0050] Step S150: For the updated inversion feature set, re-execute the step of inputting the reference data corresponding to each vegetation sensitive feature into the preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature, until the updated inversion feature set meets the set condition formula.

[0051] In this embodiment, this is an iterative optimization process, and the iterative inversion model can be represented as follows: in, Indicates the number of iterations; For each iteration, using the inversion data corresponding to the first vegetation-sensitive feature, the updated feature set (containing the prediction results of other parameters from the previous iteration) is used to retrain and predict all growth parameters. The set conditional expression is a rule that controls the termination of the iteration; it defines when the accuracy of the inversion results no longer improves significantly.

[0052] For example, the set conditional expressions may include the following: the determination coefficient of the previous inversion feature set is less than the determination coefficient of the subsequent inversion feature set; the root mean square error of the previous inversion feature set is greater than the root mean square error of the subsequent inversion feature set; or the number of iterations corresponding to the subsequent inversion feature set exceeds a set number. Here, the determination coefficient and root mean square error of any inversion feature set are obtained from the model prediction values ​​of the corresponding inversion model and the corresponding measured data.

[0053] For example, the condition can be expressed as: After the t-th iteration, calculate the coefficient of determination between the model predictions and the ground-based measured values. and root mean square error When the (t+1)th iteration Compared to The increase is less than a preset minimum threshold (e.g., 0.01), or The rate of decrease tends to stabilize, or the number of iterations. The iteration terminates when the preset maximum number of iterations (e.g., 10) is reached. Through this multi-level progressive iterative optimization, complementary information between different growth parameters can be gradually integrated, making the prediction results of each parameter more physiologically consistent and improving accuracy in a synergistic way.

[0054] In this embodiment, the determination coefficient The expression is as follows: Root mean square error The expression is as follows: in, This represents the measured data of the i-th sample. This represents the predicted value output by the corresponding inversion model. This represents the average value of the measured data. This represents the number of samples.

[0055] As described in steps S110 to S150 above, this embodiment acquires multi-source remote sensing and ground-based measured data, performs spatial alignment and sensitive feature screening, and constructs an inversion feature set containing multiple vegetation sensitive features. Based on this, this embodiment abandons the traditional single-parameter independent inversion mode and innovatively introduces a multi-parameter collaborative iterative inversion mechanism: first, initial values ​​of each growth parameter are obtained using the initial feature set; then, these initial predicted values ​​are fed back to the feature set as new constraint features to invert other parameters again, and this process is repeated. Through this iterative strategy of "prediction-feedback-re-prediction," this embodiment can fully explore and utilize the inherent physiological coupling relationship between different vegetation growth parameters (such as leaf area and chlorophyll content), using the information of the inverted parameters as prior knowledge to constrain and optimize the inversion process of other parameters. This solves the problems of independent parameter inversion, insufficient information utilization, and bottlenecks in inversion accuracy in existing technologies, effectively improving the synergy, stability, and overall accuracy of multi-growth parameter inversion.

[0056] In some embodiments of this disclosure, another method for the coordinated inversion of multiple vegetation growth parameters is provided. This method, based on steps S110 to S150 of the aforementioned embodiments, preprocesses the spectral remote sensing data. The method includes the following steps S210 to S260: Step S210: Acquire spectral remote sensing data of the target vegetation area output by the UAV and measured data of the target vegetation area output by the ground sampling equipment.

[0057] Step S220: Perform grayscale processing and scaling on the multiple spectral remote sensing data to obtain the processed spectral remote sensing data.

[0058] In this embodiment, grayscale processing converts the multi-band spectral remote sensing data acquired by the UAV into single-channel brightness data, while scaling processing uniformly adjusts the spatial resolution of the image. These preprocessing operations aim to reduce the amount of data, eliminate the influence of sensor differences and changes in lighting conditions, and provide standardized input for subsequent feature extraction and analysis.

[0059] For example, the preprocessing steps may also include at least one of the following: outlier removal, radiometric correction, geometric correction, orthorectification, and image stitching. For instance, firstly, data distorted by flight attitude instability is removed; then, the original DN values ​​are converted to reflectance using radiometric calibration coefficients; next, geometric correction is performed on the image using positioning and attitude information (POS data) carried by the UAV; furthermore, orthorectification eliminates displacement caused by terrain and viewing angle, and the images acquired from multiple flight paths are seamlessly stitched into a complete orthorectified image covering the entire study area. Through these preprocessing steps, high-quality remote sensing images with multi-temporal and spatial consistency can be obtained, providing a reliable data foundation for subsequent inversion.

[0060] Step S230: Align the spectral remote sensing data and the measured data with the location parameters of the target vegetation area to obtain an inversion feature set.

[0061] Step S240: Input the reference data corresponding to each vegetation sensitive feature into the preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature.

[0062] Step S250: Update the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature.

[0063] Step S260: For the updated inversion feature set, re-execute the step of inputting the reference data corresponding to each vegetation sensitive feature into the preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature, until the updated inversion feature set meets the set condition formula.

[0064] In this embodiment, the specific implementation of steps S210, S230 to S260 is similar to steps S110 to S150 in the previous embodiment, and will not be repeated here.

[0065] As can be seen from steps S210 to S260 above, this embodiment performs a series of preprocessing operations on the spectral remote sensing data before data inversion, including grayscale conversion, scaling, radiometric correction, and geometric correction. This effectively eliminates noise, geometric distortion, and radiometric differences in the original image, generating a standardized, high-precision basic data product. This provides high-quality data assurance for subsequent feature extraction, alignment, and inversion model construction, further improving the accuracy and stability of collaborative inversion.

[0066] <Equipment Example 1> Figure 3 This is a schematic diagram of a vegetation multi-growth parameter collaborative inversion device according to one embodiment. Figure 3 As shown, the vegetation multi-growth parameter co-inversion device 300 may include: The acquisition module 310 is used to acquire spectral remote sensing data of the target vegetation area output by the UAV and measured data of the target vegetation area output by the ground sampling equipment; The first obtaining module 320 is used to align the spectral remote sensing data and the measured data with the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation sensitive feature; The second obtaining module 330 is used to input the reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature; The update module 340 is used to update the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature; The third module 350 is used to perform the step of inputting the reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature for the updated inversion feature set, until the updated inversion feature set meets the set condition formula.

[0067] Optionally, the first obtaining module 320 is used to map the latitude and longitude information reflected by the measured data to the spatial coordinate system associated with the spectral remote sensing data to obtain spatially aligned data; determine the exponential characteristics of the target vegetation area based on the spatially aligned data; and determine the correlation between the spectral remote sensing data and the spectral band, the exponential characteristics, and the corresponding growth parameters using the measured data as a reference, and filter to obtain each vegetation sensitive feature in the spatially aligned data and the reference data corresponding to each vegetation sensitive feature.

[0068] Optionally, the first obtaining module 320 is used to determine, based on the spectral remote sensing data, the normalized difference vegetation index, the kernel normalized vegetation index, the green normalized vegetation index, the normalized red edge index, the enhanced vegetation index, the improved enhanced vegetation index, the soil-adjusted vegetation index, the optimized soil-adjusted vegetation index, the chlorophyll index, the green band chlorophyll index, the red edge band chlorophyll index, the ratio vegetation index, the improved chlorophyll reflectance absorption index, and the structure-insensitive pigment index, as index features of the target vegetation area.

[0069] <Equipment Example 2> Figure 4 This is a schematic diagram of the hardware structure of an electronic device according to another embodiment.

[0070] like Figure 4 As shown, the electronic device 400 includes a processor 410 and a memory 420, the memory 420 for storing an executable computer program, and the processor 410 for executing methods as described in any of the above method embodiments under the control of the computer program.

[0071] Each module of the above-mentioned vegetation multi-growth parameter collaborative inversion device 300 can be implemented by the processor 410 in this embodiment executing the computer program stored in the memory 420, or it can be implemented by other structures, which are not limited here.

[0072] This invention can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of the invention.

[0073] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination thereof. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0074] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0075] The computer program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the computer-readable program instructions. This electronic circuitry can execute the computer-readable program instructions to implement various aspects of the invention.

[0076] Various aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0077] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0078] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0079] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions. It will be known to those skilled in the art that implementation in hardware, implementation in software, and implementation using a combination of software and hardware are equivalent.

[0080] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein. The scope of the invention is defined by the appended claims.

Claims

1. A method for synergistic inversion of multiple vegetation growth parameters, characterized in that, The method includes: Acquire spectral remote sensing data of the target vegetation area output by the UAV and measured data of the target vegetation area output by the ground sampling equipment; The spectral remote sensing data and the measured data are aligned with the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation sensitive feature; The reference data corresponding to each vegetation sensitive feature is input into the preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature; Update the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature; For the updated inversion feature set, the step of inputting the reference data corresponding to each vegetation sensitive feature into the preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature is performed until the updated inversion feature set meets the set condition formula.

2. The method according to claim 1, characterized in that, The process involves aligning the spectral remote sensing data and the measured data using the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation-sensitive feature, including: Based on the latitude and longitude information reflected in the measured data, spatially aligned data is obtained by mapping it to the spatial coordinate system associated with the spectral remote sensing data; Based on the spatial alignment data, determine the index characteristics of the target vegetation area; Using the measured data as a reference, the correlation between the spectral remote sensing data and the spectral band, the exponential feature, and the corresponding growth parameter is determined, and each vegetation sensitive feature in the spatially aligned data and the reference data corresponding to each vegetation sensitive feature are obtained.

3. The method according to claim 2, characterized in that, The step of determining the index characteristics of the target vegetation area based on the spectral remote sensing data includes: Based on the spectral remote sensing data, the following indexes were determined as index characteristics of the target vegetation area: Normalized Difference Vegetation Index, Core Normalized Vegetation Index, Green Normalized Vegetation Index, Normalized Red Edge Index, Enhanced Vegetation Index, Improved Enhanced Vegetation Index, Soil-Adjusted Vegetation Index, Optimized Soil-Adjusted Vegetation Index, Chlorophyll Index, Green Band Chlorophyll Index, Red Edge Band Chlorophyll Index, Ratio Vegetation Index, Improved Chlorophyll Reflectance Absorption Index, and Structure-Insensitive Pigment Index.

4. The method according to claim 1, characterized in that, The inversion model is a regression model constructed based on at least one of the following algorithms: linear regression, random forest, support vector regression, KNN, GBDT, and XGBoost.

5. The method according to claim 1, characterized in that, The updated inversion feature set includes a first inversion feature mapped from the inversion data corresponding to the first vegetation sensitive feature and an inherent second inversion feature.

6. The method according to claim 1, characterized in that, The set conditions include: the determination coefficient of the previous inversion feature set is less than the determination coefficient of the subsequent inversion feature set; the root mean square error of the previous inversion feature set is greater than the root mean square error of the subsequent inversion feature set; or the number of iterations corresponding to the subsequent inversion feature set exceeds a set number.

7. The method according to claim 6, characterized in that, The coefficient of determination and the root mean square error of any inversion feature set are obtained by using the model prediction value of the corresponding inversion model and the corresponding measured data.

8. A vegetation multi-growth parameter synergistic inversion device, characterized in that, The device includes: The acquisition module is used to acquire spectral remote sensing data of the target vegetation area output by the UAV and measured data of the target vegetation area output by the ground sampling equipment; The first obtaining module is used to align the spectral remote sensing data and the measured data with the location parameters of the target vegetation area to obtain an inversion feature set; wherein, the inversion feature set includes reference data corresponding to each vegetation sensitive feature; The second obtaining module is used to input the reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature; The update module is used to update the inversion feature set with the inversion data corresponding to the first vegetation sensitive feature; The third module is used to perform the step of inputting the reference data corresponding to each vegetation sensitive feature into a preset inversion model to obtain the inversion data corresponding to each vegetation sensitive feature for the updated inversion feature set, until the updated inversion feature set meets the set condition formula.

9. An electronic device, characterized in that, The system includes a memory and a processor, the memory being used to store a computer program; the processor being used to execute the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images