Method, apparatus, and storage medium for monitoring crop health

Abstract

The embodiment of the application provides a kind of method for monitoring crop health condition, device and storage medium. Method includes: obtaining the multi-modal image data of crop to be measured, multi-modal image data includes visible light image data, near-infrared image data, thermal infrared image data and laser radar point cloud data;Multi-modal image data is input to improved YOLOV10 model, to obtain the current growth stage of crop to be measured;According to multi-modal image data, the normalized vegetation index of crop to be measured, thermal infrared average temperature and average plant height are determined;Normalized vegetation index, thermal infrared average temperature and average plant height are weighted and fused to process, to obtain the growth dynamic index of crop to be measured in current growth stage;According to growth dynamic index, the health condition of crop to be measured is monitored, can improve the reliability of crop growth stage monitoring, so that the monitoring of crop health condition is more accurate.

Description

Technical Field

[0001] This application relates to the field of crop health monitoring technology, and more specifically to a method, apparatus and storage medium for monitoring crop health status. Background Technology

[0002] Currently, crop growth stage monitoring mainly relies on single-modal image data and traditional manual assessment methods. However, these methods are susceptible to environmental factors such as changes in light intensity and shading. Especially for large-scale planting areas, relying on single-modal image data makes it difficult to provide comprehensive monitoring of rice growth status, resulting in low reliability of crop growth stage monitoring and consequently, inaccurate monitoring of crop health status. Summary of the Invention

[0003] The purpose of this application is to provide a method, apparatus, and storage medium for monitoring crop health status, in order to solve the problem that the existing technology relies on single-modal image data, which makes the monitoring of crop growth stages less reliable and thus results in inaccurate monitoring of crop health status.

[0004] To achieve the above objectives, a first aspect of this application provides a method for monitoring crop health, comprising:

[0005] Acquire multimodal image data of the crop to be tested, including visible light image data, near-infrared image data, thermal infrared image data, and lidar point cloud data;

[0006] Multimodal image data is input into an improved YOLOV10 model to obtain the current growth stage of the crop under test. The improved YOLOV10 model includes a feature fusion layer, which fuses multimodal image data through the Transformer and Attention mechanisms.

[0007] The normalized vegetation index, mean thermal infrared temperature, and mean plant height of the crop under test were determined based on multimodal image data.

[0008] The normalized vegetation index, average thermal infrared temperature, and average plant height were weighted and fused to obtain the growth dynamic index of the crop under test at the current growth stage.

[0009] The health status of the crop under test is monitored based on the growth dynamic index.

[0010] In this embodiment, the weighted fusion processing of the normalized vegetation index, the average thermal infrared temperature, and the average plant height to obtain the growth dynamic index of the crop under test at the current growth stage includes: obtaining a first initial weight corresponding to the normalized vegetation index, a second initial weight corresponding to the average thermal infrared temperature, and a third initial weight corresponding to the average plant height; adjusting the first initial weight, the second initial weight, and the third initial weight based on a Bayesian algorithm and according to at least one of the current growth stage, the location of the planting area where the crop under test is located, and the type of the crop under test; and performing weighted fusion processing on the normalized vegetation index, the average thermal infrared temperature, and the average plant height according to the adjusted first preset weight, the adjusted second preset weight, and the adjusted third preset weight to obtain the growth dynamic index.

[0011] In this embodiment of the application, the normalized vegetation index, the average thermal infrared temperature, and the average plant height are weighted and fused according to the adjusted first preset weight, the adjusted second preset weight, and the adjusted third preset weight to obtain the growth dynamic index. This includes: weighting the normalized vegetation index according to the adjusted first preset weight to obtain a first weight; weighting the average thermal infrared temperature according to the adjusted second preset weight to obtain a second weight; weighting the average plant height according to the adjusted third preset weight to obtain a third weight; and determining the total weight among the first weight, the second weight, and the third weight, whereby the total weight is the growth dynamic index.

[0012] In this embodiment of the application, monitoring the health status of the crop under test based on the growth dynamic index includes: obtaining multiple preset ranges corresponding to the crop type of the crop under test, wherein each preset range corresponds to a preset level, the preset level corresponding to the preset range with a larger lower limit value is higher, and the health status corresponding to the higher the preset level is better; determining the target range in which the growth dynamic index is located from the multiple preset ranges; and determining the health status of the crop under test based on the preset level corresponding to the target range.

[0013] In this embodiment of the application, the method further includes: generating a crop health distribution map of the planting area where the crop to be tested is located based on the growth dynamic index.

[0014] In this embodiment of the application, the method further includes: marking crop areas in the crop health distribution map where the health status is lower than expected; and sending the marked crop health distribution map to a display device.

[0015] In this embodiment of the application, the method further includes: generating a field management report based on the health status of the crop to be tested; and sending the field management report to the user terminal.

[0016] In the embodiments of this application, the improved YOLOV10 model is optimized for lightweighting through a pruning algorithm.

[0017] A second aspect of this application provides an apparatus for monitoring crop health, comprising:

[0018] The memory is configured to store instructions;

[0019] The processor is configured to retrieve instructions from memory and, when executing the instructions, to implement the aforementioned method for monitoring crop health.

[0020] A third aspect of this application provides a machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the aforementioned method for monitoring crop health.

[0021] The above technical solution acquires multimodal image data of the crop under test, including visible light image data, near-infrared image data, thermal infrared image data, and lidar point cloud data. This multimodal image data is then input into an improved YOLOv10 model to determine the current growth stage of the crop. The improved YOLOv10 model includes a feature fusion layer that fuses the multimodal image data using Transformer and Attention mechanisms. Based on the multimodal image data, the normalized vegetation index (NVI), average thermal infrared temperature, and average plant height of the crop are determined. The NVI, average thermal infrared temperature, and average plant height are then weighted and fused to obtain the growth dynamic index of the crop at its current growth stage. Monitoring the health status of the crop based on this growth dynamic index improves the reliability of crop growth stage monitoring and makes crop health status monitoring more accurate.

[0022] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0023] The accompanying drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the following detailed description to explain the embodiments of this application, but do not constitute a limitation on the embodiments of this application. In the drawings:

[0024] Figure 1 The schematic diagram illustrates a process flow diagram of a method for monitoring crop health according to an embodiment of this application;

[0025] Figure 2 The illustration shows a flowchart of determining the current growth stage of a crop to be tested according to an embodiment of this application;

[0026] Figure 3 The schematic diagram illustrates a process for determining weight factors based on a Bayesian algorithm according to an embodiment of this application;

[0027] Figure 4 A schematic flowchart of a method for monitoring crop health according to another embodiment of this application is shown.

[0028] Figure 5 The diagram illustrates the internal structure of a computer device according to an embodiment of this application. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0030] It should be noted that if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0031] Figure 1 A schematic flowchart illustrating a method for monitoring crop health according to an embodiment of this application is shown. Figure 1 As shown in one embodiment of this application, a method for monitoring crop health status is provided, comprising the following steps:

[0032] Step 101: Acquire multimodal image data of the crop to be tested. The multimodal image data includes visible light image data, near-infrared image data, thermal infrared image data, and lidar point cloud data.

[0033] Multimodal image data can be acquired using aerial equipment or remote sensing devices. Aerial equipment may include drones or other devices capable of aerial inspection. For example, if data is acquired using aerial equipment, the processor can control the equipment to cruise the planting area where the crop to be measured is located at a preset altitude. The preset altitude can be set according to actual conditions. The overlap of the cruise can also be set when the aerial equipment cruises the crop area; for example, the overlap can be set to 75%.

[0034] The flight equipment is equipped with image acquisition devices. These devices can be cameras, camcorders, recorders, or other equipment with image acquisition capabilities. The processor can acquire multimodal image data of the crop under test collected by the flight equipment. This multimodal image data includes visible light image data, near-infrared image data, thermal infrared image data, and lidar point cloud data.

[0035] Visible light image data can be used to extract basic morphological features of the crop under test; near-infrared image data can be used to identify the leaf health status of the crop; and thermal infrared image data can be used to detect the temperature distribution in the planting area of ​​the crop, for example, it can help identify whether the crop is growing abnormally. LiDAR point cloud data can be used to extract three-dimensional structural features of the crop, for example, it can help identify features such as plant height.

[0036] Step 102: Input the multimodal image data into the improved YOLOV10 model to obtain the current growth stage of the crop under test. The improved YOLOV10 model includes a feature fusion layer, which fuses the multimodal image data through the Transformer mechanism and the Attention mechanism.

[0037] The improved YOLOv10 model in this embodiment is a model in which a feature fusion layer is added to the backbone network of the YOLOv10 model. The feature fusion layer fuses multimodal image data through the Transformer and Attention mechanisms. In order to significantly improve the accuracy and speed of subsequent prediction of the growth stage of the crop under test, the improved YOLOv10 model refers to the tenth generation version of the improved YOLO model.

[0038] Before processing multimodal image data, an improved YOLOv10 model needs to be trained. Specifically, multimodal data of rice at different growth stages (including RGB images, near-infrared images, thermal infrared images, and LiDAR point clouds) can be collected and input into the improved YOLOv10 model. The model achieves multimodal feature fusion and key region identification through the Transformer mechanism and spatial and channel attention mechanisms. The Transformer mechanism establishes global relationships between RGB, NIR, thermal infrared, and LiDAR data, such as fusing spectral features and three-dimensional structural features. The spatial attention mechanism focuses on crop distribution density during the seedling stage, detects the number of tillers during the tillering stage, strengthens plant height and stem strength features during the jointing stage, locks the panicle region during the heading stage, analyzes grain fullness during the grain filling stage, and evaluates harvesting standards (such as panicle maturity) during the maturity stage. The channel attention mechanism dynamically weights the importance of different modalities of data, for example, increasing the weight of thermal infrared data in water shortage monitoring. The conditions for training completion may include, but are not limited to, the prediction accuracy reaching a preset accuracy, the preset error being within a preset range, and the number of iterations reaching the predicted number. For example, the conditions for training completion include: the model's average accuracy (mAP) for classifying rice growth stages reaching 86.5%, or the loss function converging to a preset threshold, or the number of training iterations reaching 10,000.

[0039] To meet the computing demands of edge devices such as drones, in this embodiment, the improved YOLOv10 model can be lightweighted using a channel pruning algorithm to reduce redundant parameters and computational load, ensuring the real-time performance of the improved YOLOv10 model. The processor can input the acquired multimodal image data into the improved YOLOv10 model to obtain the current growth stage of the crop under test. The current growth stage may include seedling stage, tillering stage, jointing stage, booting stage, heading stage, and grain-filling stage.

[0040] In one embodiment, before inputting multimodal image data into the improved YOLOv10 model, the multimodal image data can be preprocessed. The preprocessing methods may include, but are not limited to, data alignment processing (such as synchronizing the acquisition time of multimodal data based on GPS timestamps), data registration processing (unifying the spatial coordinate system of RGB, NIR, thermal infrared images and LiDAR point clouds through affine transformation), and normalization processing (standardizing the pixel values ​​of different modal data to the [0,1] interval to eliminate dimensional differences) to ensure the consistency of different data sources in space and time.

[0041] like Figure 2 As shown, using the improved YOLOV10 model as the improved YOLO-V10 model, a flowchart is provided to determine the current growth stage of the crop to be tested.

[0042] After acquiring multimodal data, it can be input into the improved YOLO-V10 model. This multimodal data can include RGB (visible light image data), NIR (near-infrared image data), Thermal (thermal infrared image data), and LiDAR (lidar radar point cloud data). The backbone network of the improved YOLO-V10 model extracts the basic features of the multimodal data and outputs a basic feature map.

[0043] The multimodal feature fusion layer establishes global relationships between multimodalities through the Transformer mechanism and uses the Attention mechanism to enable the improved YOLOV10 model to highlight key regions or modalities, so as to output a fused comprehensive feature map. After lightweight optimization, it performs target detection and classification as well as growth stage identification to output the classification results of rice growth stages, such as tillering stage, jointing stage and heading stage.

[0044] Step 103: Determine the normalized vegetation index, average thermal infrared temperature, and average plant height of the crop to be tested based on the multimodal image data.

[0045] The processor can determine the normalized difference vegetation index, average thermal infrared temperature, and average plant height of the crop under test based on multimodal image data. Specifically, the processor can determine the normalized difference vegetation index of the crop under test based on near-infrared image data, determine the average thermal infrared temperature of the crop under test based on thermal infrared image data, and determine the average plant height of the crop under test based on lidar point cloud data.

[0046] Step 104: Perform weighted fusion processing on the normalized vegetation index, average thermal infrared temperature, and average plant height to obtain the growth dynamic index of the crop under test at the current growth stage.

[0047] The processor can perform weighted fusion processing on the normalized vegetation index, average thermal infrared temperature, and average plant height to obtain the growth dynamic index of the crop under test at the current growth stage.

[0048] In this embodiment, the weighted fusion processing of the normalized vegetation index, the average thermal infrared temperature, and the average plant height to obtain the growth dynamic index of the crop under test at the current growth stage includes: obtaining a first initial weight corresponding to the normalized vegetation index, a second initial weight corresponding to the average thermal infrared temperature, and a third initial weight corresponding to the average plant height; adjusting the first initial weight, the second initial weight, and the third initial weight based on a Bayesian algorithm and according to at least one of the current growth stage, the location of the planting area where the crop under test is located, and the type of the crop under test; and performing weighted fusion processing on the normalized vegetation index, the average thermal infrared temperature, and the average plant height according to the adjusted first preset weight, the adjusted second preset weight, and the adjusted third preset weight to obtain the growth dynamic index.

[0049] The processor can acquire a first initial weight corresponding to the normalized vegetation index, a second initial weight corresponding to the average thermal infrared temperature, and a third initial weight corresponding to the average plant height. The processor can adjust the first, second, and third initial weights based on a Bayesian algorithm, according to at least one of the following: the current growth stage, the location of the planting area of ​​the crop under test, and the type of the crop under test.

[0050] like Figure 3 The diagram illustrates a flowchart for determining weight factors based on a Bayesian algorithm, including prior setting, data input, posterior update, and weight output. During prior setting, the prior distribution of the weight factors can be defined, using historical data or expert experience as input. During data input, multimodal data, current growth stage, geographical location of the planting area, and crop type can be input. During posterior update, the posterior distribution can be calculated based on Bayesian algorithm, and the weight factors can be sampled using the MCM method. During weight output, the optimized weight factors are output, and CRGDI can be further calculated. The specific steps are as follows:

[0051] 1. Bayesian model construction

[0052] Definition of the objective function: Let the mapping relationship between the Comprehensive Growth Dynamics Index (CRGDI) and the actual growth status (such as yield and health level) be:

[0053] y = α·NDVI + β·Thermal avg +γ·H mean +∈

[0054] Where y is the true growth state label (e.g., the standardized yield value), and ∈ represents the observation noise (assuming it follows a normal distribution ∈ ~ N(0,σ). 2 )).

[0055] Prior distribution setting: The initial values ​​of the weighting factors α, β, γ are set based on historical data or expert experience.

[0056]

[0057] Where, μ α μ β μ γ The initial weight mean, such as μ α =0.4, μ β =0.3, μ γ =0.3, variance σ 2 This reflects the initial uncertainty.

[0058] Assuming the observed data follows a Gaussian distribution, the likelihood function is:

[0059] p(y|α,β,γ)=N(y|α·NDVI+β·Thermal avg +γ·H mean ,σ 2 )

[0060] 2. Dynamic weight adjustment mechanism

[0061] Input feature encoding: growth stage is divided into categorical variables, such as seedling stage = 0, tillering stage = 1, etc.; planting area location is encoded as latitude and longitude coordinates or climate characteristics (temperature, rainfall, etc.); crop type is encoded as rice = 1, others = 0.

[0062] Posterior distribution update: Based on Bayes' theorem, the posterior distribution of the weight factors is as follows:

[0063] p(α,β,γ|y)∝p(y|α,β,γ)·p(α)·p(β)·p(γ)

[0064] The posterior distribution is approximated using the Markov Chain Monte Carlo (MCMC) method. The optimal values ​​of the weighting factors are obtained through Hamiltonian Monte Carlo (HMC) sampling.

[0065] Phased optimization: The prior distribution is dynamically adjusted based on the current growth stage. During the heading stage, which is sensitive to spectrum and temperature, the prior means of α (NDVI) and β (temperature) can be increased. During the grain-filling stage, since plant height is strongly correlated with grain fullness, the prior weight of γ (plant height) can be increased.

[0066] 3. Training and Validation

[0067] Data structure: Multimodal field data acquired by collection equipment covers different growth stages, geographical locations (such as rice paddies in Anhui and Hunan), and environmental conditions. Labels are then defined, and agronomists annotate the actual growth status of the rice, such as health level 1-5, and quantify yield ranges.

[0068] Optimization process: Load prior distribution parameters, generate candidate values ​​for weight factors through MCMC, calculate acceptance probability, and terminate when the chain trajectory is stable or the maximum number of iterations is reached.

[0069] Performance verification: The correlation coefficient (R²) between the CRGDI predicted value and the actual yield was used to verify the results. 2 The CRGDI prediction R-value after Bayesian weight optimization is validated on the test set using two metrics: mean squared error (MSE) and weighted average error (SSE). 2 Reaching 0.82, compared to a fixed-weight R 2 =0.68, a significant improvement.

[0070] 4. Implementation Examples

[0071] Taking the rice heading stage as an example, input the current growth stage (heading stage, coded as 3), the location is a farm in Wuhu, Anhui (latitude and longitude coding), and the crop type is rice (coded as 1). First, prior adjustment is performed, increasing the prior means of α and β to 0.5 and 0.4 respectively, reflecting the high sensitivity of this stage to spectral and temperature conditions. Then, posterior sampling is performed. After MCMC sampling, the optimal weights are obtained: α = 0.52, β = 0.38, and γ = 0.10. Finally, CRGDI is calculated: CRGDI = 0.52 × NDVI + 0.38 × Thermal avg +0.10×H mean Among them, RGDI refers to the growth dynamics index, NDVI refers to the normalized difference vegetation index, and Thermal... avg This refers to the average thermal infrared temperature, H. mean This refers to the average plant height.

[0072] The processor can perform weighted fusion processing on the normalized vegetation index, average thermal infrared temperature, and average plant height according to the adjusted first preset weight, the adjusted second preset weight, and the adjusted third preset weight to obtain the growth dynamic index. The growth dynamic index is used to measure the growth status of the crop under test.

[0073] In this embodiment of the application, the normalized vegetation index, the average thermal infrared temperature, and the average plant height are weighted and fused according to the adjusted first preset weight, the adjusted second preset weight, and the adjusted third preset weight to obtain the growth dynamic index. This includes: weighting the normalized vegetation index according to the adjusted first preset weight to obtain a first weight; weighting the average thermal infrared temperature according to the adjusted second preset weight to obtain a second weight; weighting the average plant height according to the adjusted third preset weight to obtain a third weight; and determining the total weight among the first weight, the second weight, and the third weight, whereby the total weight is the growth dynamic index.

[0074] The processor can weight the normalized vegetation index according to the adjusted first preset weight to obtain a first weight value. The processor can weight the average thermal infrared temperature according to the adjusted second preset weight to obtain a second weight value. The processor can weight the average plant height according to the adjusted third preset weight to obtain a third weight value. The processor can determine the total weight value among the first, second, and third weight values, and the total weight value is the growth dynamic index.

[0075] In one embodiment, the growth dynamics index can be determined by the following formula:

[0076] CRGDI = α × NDVI + β × Thermal avg +γ×H mean

[0077] Wherein, CRGDI refers to the growth dynamics index, α refers to the adjusted first preset weight, NDVI refers to the normalized vegetation index, β refers to the adjusted second preset weight, and Thermal... avg This refers to the average thermal infrared temperature, γ refers to the adjusted third preset weight, and H... mean This refers to the average plant height.

[0078] Step 105: Monitor the health status of the crop under test based on the growth dynamic index.

[0079] The processor can monitor the health status of the crop under test based on the growth dynamic index. For example, the health status of the crop under test can be determined based on the range of the growth dynamic index, or based on the magnitude between the growth dynamic index and a preset threshold.

[0080] In this embodiment of the application, monitoring the health status of the crop under test based on the growth dynamic index includes: obtaining multiple preset ranges corresponding to the crop type of the crop under test, wherein each preset range corresponds to a preset level, the preset level corresponding to the preset range with a larger lower limit value is higher, and the health status corresponding to the higher the preset level is better; determining the target range in which the growth dynamic index is located from the multiple preset ranges; and determining the health status of the crop under test based on the preset level corresponding to the target range.

[0081] The processor can acquire multiple preset ranges corresponding to the crop type of the crop being tested. Each preset range corresponds to a preset level; the higher the lower limit of the preset range, the higher the preset level, and the higher the preset level, the better the health condition. For example, the multiple preset ranges and the preset levels corresponding to each preset range can be set as follows: 0.8–1.0: Very healthy; 0.6–0.8: Healthy; 0.4–0.6: Mild stress or disease; 0.2–0.4: Moderate stress or disease; 0.0–0.2: Severe stress or disease.

[0082] The processor can determine the target range of the growth dynamic index from multiple preset ranges, and determine the health status of the crop under test according to the preset level corresponding to the target range. For example, if the preset level corresponding to the target range is "very healthy", the health status of the crop under test is "excellent"; if the preset level is "healthy", the health status is "good"; if the preset level is "mild stress or disease", the health status is "medium"; if the preset level is "moderate stress or disease", the health status is "poor"; and if the preset level is "severe stress or disease", the health status is "very poor".

[0083] In this embodiment of the application, the method further includes: generating a crop health distribution map of the planting area where the crop to be tested is located based on the growth dynamic index.

[0084] The processor can generate a crop health distribution map of the planting area where the crop under test is located, based on the growth dynamic index. The crop health distribution map can identify the growth dynamic index of the crop under test within the planting area. When the growth dynamic index is within different ranges or at different values, the area of ​​the crop under test on the crop health distribution map can be displayed in different colors, allowing users to understand the crop health status within the planting area.

[0085] In this embodiment of the application, the method further includes: marking crop areas in the crop health distribution map where the health status is lower than expected; and sending the marked crop health distribution map to a display device.

[0086] The processor can mark crop areas on the crop health distribution map where the health status is below expectations. For example, it can mark crop areas where the tested crop has a health status of medium or below. The processor can then send the marked crop health distribution map to a display device. This display device can be a display screen at a field workstation in the planting area or a display screen on a mobile terminal, or any other device with display capabilities.

[0087] In this embodiment of the application, the method further includes: generating a field management report based on the health status of the crop to be tested; and sending the field management report to the user terminal.

[0088] The processor can generate a field management report based on the health status of the crop under test. This report may include management recommendations based on the crop's health condition, such as increasing or decreasing irrigation or fertilization, spraying pesticides to control diseases, and adjusting planting density or crop rotation plans. The processor can send the field management report to the user terminal, prompting the user to make flexible adjustments to the crop's health status, thereby improving its survival rate and subsequent yield and quality.

[0089] In one embodiment, the crop to be tested may include, but is not limited to, rice and wheat.

[0090] like Figure 4 As shown, taking rice as an example, a flowchart of another method for monitoring crop health is provided.

[0091] When monitoring crop health, data can be collected, including RGB (visible light image data), NIR (near-infrared image data), Thermal (thermal infrared image data), and LiDAR (lidar radar point cloud data). The collected data can then be preprocessed, including but not limited to data alignment, registration, and normalization, to ensure spatial and temporal consistency across different data sources. Next, multimodal feature fusion can be performed. Specifically, the preprocessed data can be input into an improved YOLO-V10 model, which includes a feature fusion layer that fuses the preprocessed data using Transformer and Attention mechanisms. The improved YOLO-V10 model can then be used for target detection and classification to identify rice growth stages. Subsequently, the comprehensive index CRGDI can be calculated based on NDVI (Normalized Difference Vegetation Index), Thermal (mean thermal infrared temperature), and average plant height. The comprehensive index CRGDI is used to measure the growth status of the crop under test. Furthermore, a growth stage report and recommendations, including dynamic management measures, can be generated based on the comprehensive index CRGDI.

[0092] The above technical solution acquires multimodal image data of the crop under test, including visible light image data, near-infrared image data, thermal infrared image data, and lidar point cloud data. This multimodal image data is then input into an improved YOLOv10 model to determine the current growth stage of the crop. The improved YOLOv10 model includes a feature fusion layer that fuses the multimodal image data using Transformer and Attention mechanisms. Based on the multimodal image data, the normalized vegetation index (NVI), average thermal infrared temperature, and average plant height of the crop are determined. The NVI, average thermal infrared temperature, and average plant height are then weighted and fused to obtain the growth dynamic index of the crop at its current growth stage. Monitoring the health status of the crop based on this growth dynamic index improves the reliability of crop growth stage monitoring and makes crop health status monitoring more accurate.

[0093] Figure 1 and Figure 4 This is a flowchart illustrating a method for monitoring crop health in one embodiment. It should be understood that, although... Figure 1 and Figure 4 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise explicitly stated herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 and Figure 4At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0094] In one embodiment, a device for monitoring crop health is provided, comprising:

[0095] The memory is configured to store instructions;

[0096] The processor is configured to retrieve instructions from memory and, when executing the instructions, to implement the aforementioned method for monitoring crop health.

[0097] In one embodiment, a storage medium is provided on which a program is stored, which, when executed by a processor, implements the above-described method for monitoring crop health.

[0098] In one embodiment, a processor is provided for running a program, wherein the program executes the above-described method for monitoring crop health status.

[0099] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor A01, a network interface A02, memory (not shown), and a database (not shown) connected via a system bus. The processor A01 provides computing and control capabilities. The memory includes internal memory A03 and a non-volatile storage medium A04. The non-volatile storage medium A04 stores an operating system B01, a computer program B02, and a database (not shown). The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 stored in the non-volatile storage medium A04. The database stores data such as crop health status. The network interface A02 communicates with external terminals via a network connection. When executed by the processor A01, the computer program B02 implements a method for monitoring crop health status.

[0100] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0101] This application provides an apparatus including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs the following steps: acquiring multimodal image data of a crop to be tested, including visible light image data, near-infrared image data, thermal infrared image data, and lidar point cloud data; inputting the multimodal image data into an improved YOLOv10 model to obtain the current growth stage of the crop, wherein the improved YOLOv10 model includes a feature fusion layer that fuses the multimodal image data using a Transformer mechanism and an Attention mechanism; determining the normalized vegetation index, average thermal infrared temperature, and average plant height of the crop based on the multimodal image data; performing weighted fusion processing on the normalized vegetation index, average thermal infrared temperature, and average plant height to obtain a growth dynamic index of the crop at its current growth stage; and monitoring the health status of the crop based on the growth dynamic index.

[0102] In one embodiment, weighted fusion processing of the normalized vegetation index, average thermal infrared temperature, and average plant height to obtain the growth dynamic index of the crop under test at the current growth stage includes: obtaining a first initial weight corresponding to the normalized vegetation index, a second initial weight corresponding to the average thermal infrared temperature, and a third initial weight corresponding to the average plant height; adjusting the first initial weight, the second initial weight, and the third initial weight based on a Bayesian algorithm and according to at least one of the current growth stage, the location of the planting area where the crop under test is located, and the type of the crop under test; and performing weighted fusion processing of the normalized vegetation index, average thermal infrared temperature, and average plant height according to the adjusted first preset weight, the adjusted second preset weight, and the adjusted third preset weight to obtain the growth dynamic index.

[0103] In one embodiment, the normalized vegetation index, average thermal infrared temperature, and average plant height are weighted and fused according to the adjusted first preset weight, the adjusted second preset weight, and the adjusted third preset weight to obtain a growth dynamic index. This includes: weighting the normalized vegetation index according to the adjusted first preset weight to obtain a first weight; weighting the average thermal infrared temperature according to the adjusted second preset weight to obtain a second weight; weighting the average plant height according to the adjusted third preset weight to obtain a third weight; and determining the total weight among the first weight, the second weight, and the third weight, where the total weight is the growth dynamic index.

[0104] In one embodiment, monitoring the health status of a crop under test based on a growth dynamic index includes: obtaining multiple preset ranges corresponding to the crop type of the crop under test, wherein each preset range corresponds to a preset level, the preset level corresponding to a preset range with a larger lower limit value is higher, and the health status corresponding to a higher preset level is better; determining the target range in which the growth dynamic index is located from the multiple preset ranges; and determining the health status of the crop under test based on the preset level corresponding to the target range.

[0105] In one embodiment, the method further includes: generating a crop health distribution map of the planting area where the crop to be tested is located based on the growth dynamic index.

[0106] In one embodiment, the method further includes: marking crop areas in the crop health distribution map where the health status is lower than expected; and sending the marked crop health distribution map to a display device.

[0107] In one embodiment, the method further includes: generating a field management report based on the health status of the crop to be tested; and sending the field management report to a user terminal.

[0108] In one embodiment, the improved YOLOV10 model is optimized for lightweighting through a pruning algorithm.

[0109] This application also provides a computer program product that, when executed on a data processing device, is adapted to execute a program that initializes method steps for monitoring crop health.

[0110] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0111] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will 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 program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0112] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0113] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0114] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0115] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0116] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0117] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0118] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for monitoring crop health status, characterized in that, The method includes: Acquire multimodal image data of the crop to be tested, including visible light image data, near-infrared image data, thermal infrared image data, and lidar point cloud data; The multimodal image data is input into an improved YOLOv10 model to obtain the current growth stage of the crop under test. The improved YOLOv10 model includes a feature fusion layer, which fuses the multimodal image data through a Transformer mechanism and an Attention mechanism. The normalized vegetation index, average thermal infrared temperature, and average plant height of the crop under test are determined based on the multimodal image data. The normalized vegetation index, average thermal infrared temperature, and average plant height are weighted and fused to obtain the growth dynamic index of the crop under test at the current growth stage. The health status of the crop under test is monitored based on the growth dynamic index.

2. The method for monitoring crop health status according to claim 1, characterized in that, The weighted fusion processing of the normalized vegetation index, average thermal infrared temperature, and average plant height to obtain the growth dynamic index of the crop under test at the current growth stage includes: Obtain a first initial weight corresponding to the normalized vegetation index, a second initial weight corresponding to the average thermal infrared temperature, and a third initial weight corresponding to the average plant height. Based on the Bayesian algorithm, the first initial weight, the second initial weight, and the third initial weight are adjusted according to at least one of the current growth stage, the location of the planting area where the crop to be tested is located, and the type of the crop to be tested. The normalized vegetation index, the average thermal infrared temperature, and the average plant height are weighted and fused according to the adjusted first preset weight, the adjusted second preset weight, and the adjusted third preset weight to obtain the growth dynamic index.

3. The method for monitoring crop health status according to claim 2, characterized in that, The weighted fusion process, which applies adjusted first preset weights, adjusted second preset weights, and adjusted third preset weights to the normalized vegetation index, the average thermal infrared temperature, and the average plant height, yields the growth dynamic index, including: The normalized vegetation index is weighted according to the adjusted first preset weight to obtain the first weight value; The average thermal infrared temperature is weighted according to the adjusted second preset weight to obtain a second weight value; The average plant height is weighted according to the adjusted third preset weight to obtain the third weight value; The total weight among the first weight, the second weight, and the third weight is determined, and the total weight is the growth dynamic index.

4. The method for monitoring crop health status according to claim 1, characterized in that, The monitoring of the health status of the crop under test based on the growth dynamic index includes: Multiple preset ranges corresponding to the crop type of the crop to be tested are obtained. Each preset range corresponds to a preset level. The higher the lower limit of the preset range, the higher the preset level. The higher the preset level, the better the health condition. The target range in which the growth dynamic index is located is determined from the plurality of preset ranges; The health status of the crop to be tested is determined according to a preset level corresponding to the target range.

5. The method for monitoring crop health status according to claim 1, characterized in that, The method further includes: A crop health distribution map of the planting area where the crop to be tested is located is generated based on the growth dynamic index.

6. The method for monitoring crop health status according to claim 5, characterized in that, The method further includes: Mark the crop areas in the crop health distribution map whose health status is lower than expected; Send the tagged crop health distribution map to the display device.

7. The method for monitoring crop health status according to claim 1, characterized in that, The method further includes: A field management report is generated based on the health status of the crop to be tested; The field management report is sent to the user terminal.

8. The method for monitoring crop health status according to claim 1, characterized in that, The improved YOLOv10 model is optimized for lightweighting through a pruning algorithm.

9. A device for monitoring crop health status, characterized in that, The device includes: The memory is configured to store instructions; The processor is configured to retrieve the instructions from the memory and, when executing the instructions, to implement the method for monitoring crop health according to any one of claims 1 to 8.

10. A machine-readable storage medium storing instructions thereon, characterized in that, When executed by a processor, the instruction causes the processor to be configured to perform the method for monitoring crop health according to any one of claims 1 to 8.

Images