Rice maturity sensing system and method based on multi-source heterogeneous agricultural condition data fusion

Abstract

The invention provides a rice maturity sensing system and method based on multi-source heterogeneous agricultural condition data fusion, and realizes rapid and accurate sensing of rice maturity through multi-time-sequence and multi-source heterogeneous agricultural condition data fusion and an intelligent sensing algorithm. The system comprises a data acquisition module, a data preprocessing module, a data fusion module, an intelligent sensing module and a decision support module. Through multi-source heterogeneous data fusion and an intelligent sensing algorithm, the rice maturity sensing precision and efficiency are improved, and more accurate and real-time equipment and data support are provided for agricultural production.

Description

Technical Field

[0001] The present invention relates to the field of agricultural information technology, and particularly relates to a rice maturity perception system and method based on the fusion of multi-source heterogeneous agricultural situation data. Background Art

[0002] With the development of precision agriculture, the acquisition and analysis of agricultural situation data have become particularly important. Traditional methods for acquiring agricultural situation data mainly rely on ground sampling. This method is not only time-consuming and laborious, but also difficult to achieve large-scale real-time monitoring. In recent years, the development of satellite remote sensing and unmanned aerial vehicle (UAV) technologies has provided new ways for the acquisition of agricultural situation data. However, the data obtained by these technologies have the characteristics of multi-source heterogeneity. How to effectively fuse these data and extract useful information has become a hot and difficult issue in current research.

[0003] Therefore, how to provide a rice maturity perception system and method based on the fusion of multi-source heterogeneous agricultural situation data to solve the problems existing in the prior art is of great significance for its application. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a rice maturity perception system and method based on the fusion of multi-source heterogeneous agricultural situation data to solve the problem that ground sampling is not only time-consuming and laborious, but also difficult to achieve large-scale real-time monitoring of rice maturity.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A rice maturity perception system based on the fusion of multi-source heterogeneous agricultural situation data, comprising:

[0007] A data acquisition module, including a satellite remote sensing unit: using the MSI sensor of the Sentinel-2 satellite to obtain NDVI (Normalized Difference Vegetation Index) data with a resolution of 10 meters, and the coverage period is 5 days;

[0008] A UAV remote sensing unit: using the DJI Matrice 300RTK equipped with a Parrot Sequoia+ multispectral camera, flying at an altitude of 100 meters to obtain the canopy coverage rate (CC) and leaf area index (LAI) with a resolution of 5 centimeters;

[0009] A ground sampling unit: measuring the rice plant height (PH) and leaf nitrogen content (LNC) through a handheld CropSense sensor, and the sampling density is 10 points per hectare;

[0010] The data preprocessing module includes:

[0011] Data alignment unit: Using GIS spatial registration technology, satellite, drone, and ground data are unified to the UTM coordinate system (Zone 50N), and the time is aligned to UTC standard time;

[0012] Data standardization unit: Resample satellite data to 1-meter resolution, normalize drone data to the [0, 1] interval, and interpolate ground data to generate continuous rasters;

[0013] It is used to align and standardize the collected multi-source heterogeneous data;

[0014] Data fusion module, using a spatio-temporal convolutional neural network (ST-CNN), with multi-temporal NDVI, CC, LAI, and PH data as inputs, and the output is a fused rice growth feature map; The network structure includes 3 3D convolutional layers (kernel size = 3×3×3) and 2 LSTM layers (hidden size = 128);

[0015] Rice maturity perception module, based on the fused data, uses an SVM classifier to divide the rice maturity stages (vegetative growth stage, heading stage, milk ripening stage, dough ripening stage, full ripening stage);

[0016] Output module, outputs the maturity prediction results to the farm management system in GeoJSON format, and provides an API interface for real-time query on the mobile side.

[0017] As a preferred technical solution of the present invention, the data preprocessing module includes:

[0018] Data alignment unit, used to synchronize time and align space for data from different times and sources;

[0019] Data standardization unit, used to convert data from different sources into a unified format and standard.

[0020] As a preferred technical solution of the present invention, the data fusion module uses a deep learning model, including but not limited to convolutional neural network (CNN), recurrent neural network (RNN), or graph neural network (GNN), for processing the spatio-temporal features of multi-source heterogeneous data.

[0021] As a preferred technical solution of the present invention, the decision support module, based on the perception results, combines historical data and meteorological data, uses a prediction model to generate crop growth trend predictions and pest warnings, and provides accurate fertilization, irrigation, and pest control suggestions.

[0022] As a preferred technical solution of the present invention, the intelligent perception module uses low-altitude remote sensing technology of unmanned aerial vehicles, combines the data obtained by multi-spectral sensors, and uses machine learning algorithms to calculate the growth parameters of crops in real time and generate a high-precision agricultural situation map.

[0023] A method for perceiving the maturity of rice based on the fusion of multi-source heterogeneous agricultural situation data

[0024] S1: Obtain multi-temporal and multi-source heterogeneous agricultural situation data through multi-spectral sensors carried by satellites, unmanned aerial vehicles and ground sampling equipment;

[0025] S2: Preprocess the collected multi-source heterogeneous data, including denoising, spatio-temporal alignment and standardization processing, to generate a unified spatio-temporal data set;

[0026] S3: Based on a unified intelligent model framework, fuse the multi-source heterogeneous data to generate a comprehensive agricultural situation data set

[0027] S4: Use low-altitude remote sensing technology of unmanned aerial vehicles, based on the fused data, to perceive key growth parameters of rice in real time, such as height, canopy coverage rate, leaf area, vegetation index, etc.;

[0028] S5: Generate agricultural production decision-making suggestions based on the perception results and provide real-time data support.

[0029] As a preferred technical solution of the present invention, the data fusion in step 3 uses a deep learning model, including but not limited to a convolutional neural network (CNN), a recurrent neural network (RNN) or a graph neural network (GNN), for extracting spatio-temporal features of multi-source heterogeneous data.

[0030] As a preferred technical solution of the present invention, the low-altitude remote sensing technology of unmanned aerial vehicles in step 4 combines multi-spectral sensor data, uses machine learning algorithms to calculate the growth parameters of crops in real time, and generates a high-precision agricultural situation map.

[0031] As a preferred technical solution of the present invention, the decision support in step 5 is based on the perception results, combines historical data and meteorological data, uses a prediction model to generate crop growth trend predictions and pest warnings, and provides precise fertilization, irrigation and pest control suggestions.

[0032] Compared with the prior art, the beneficial effects of the present invention are:

[0033] The present invention realizes the rapid and accurate perception of the maturity of rice through the fusion of multi-source heterogeneous agricultural situation data and intelligent perception algorithms, and provides more precise and real-time equipment and data support for agricultural production. The specific beneficial effects include:

[0034] 1. The accuracy of rice maturity perception is improved through multi-source heterogeneous data fusion;

[0035] 2. The rapid perception of rice maturity is realized by using intelligent perception algorithms;

[0036] 3. Precise agricultural production decision-making support is provided according to the perception results.

[0037] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, so as to be implemented in accordance with the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the following takes the preferred embodiments of the present application and combines the drawings to describe in detail as follows.

[0038] Those skilled in the art will become more clear about the above and other purposes, advantages and features of the present application according to the following detailed description of the specific embodiments of the present application in conjunction with the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to actual scale.

[0040] Figure 1 It is the system block diagram of the present invention;

[0041] Figure 2 It is the method flow chart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. In the following description, specific details such as specific configurations and components are provided only to help a comprehensive understanding of the embodiments of the present application. Therefore, those skilled in the art should clearly understand that various changes and modifications can be made to the embodiments described here without departing from the scope and spirit of the present application. In addition, descriptions of known functions and structures are omitted for clarity and conciseness in the embodiments.

[0043] In addition, this application may repeat reference numerals and / or letters in different instances. Such repetition is for the purpose of simplicity and clarity and does not in itself indicate a relationship between the various embodiments and / or configurations discussed.

[0044] As used herein, the term "and / or" is merely a description of the relationship between associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, B exists alone, and both A and B exist simultaneously. The term " / and" as used herein describes another relationship between associated objects, indicating that two relationships may exist. For example, A / and B may represent: A exists alone, and both A and B exist. Additionally, the character " / " as used herein generally indicates that the associated objects before and after are in an "or" relationship.

[0045] It should also be noted that, as used herein, relational terms such as first and second are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including", or any other variation thereof is intended to cover non-exclusive inclusion.

[0046] Please refer to Figure 1 and 2 , the present invention provides a technical solution for a rice maturity perception system and method based on multi-source heterogeneous agricultural situation data fusion:

[0047] I. Embodiment 1 (System Architecture)

[0048] A rice maturity perception system based on multi-source heterogeneous agricultural situation data fusion includes the following components:

[0049] 1. Data acquisition module:

[0050] 1.1. Satellite remote sensing unit: The MSI sensor of the Sentinel-2 satellite is used to obtain NDVI (Normalized Difference Vegetation Index) data with a resolution of 10 meters, and the coverage period is 5 days;

[0051] 1.2. UAV remote sensing unit: The DJI Matrice 300 RTK is used to carry the Parrot Sequoia+ multispectral camera, with a flight altitude of 100 meters, to obtain the canopy coverage (CC) and leaf area index (LAI) with a resolution of 5 centimeters;

[0052] 1.3 Ground sampling unit: The rice plant height (PH) and leaf nitrogen content (LNC) are measured by a handheld CropSense sensor, and the sampling density is 10 points per hectare;

[0053] 2. Data preprocessing module

[0054] 2.1 Data Alignment Unit: Using GIS spatial registration technology, satellite, drone, and ground data are unified to the UTM coordinate system (Zone 50N), and the time is aligned to UTC standard time;

[0055] 2.2 Data Standardization Unit: Resample satellite data to 1-meter resolution, normalize drone data to the [0, 1] interval, and interpolate ground data to generate continuous rasters;

[0056] 3. Data Fusion Module:

[0057] 3.1 Adopt a spatio-temporal convolutional neural network (ST-CNN). The input is multi-temporal NDVI, CC, LAI, and PH data, and the output is a fused rice growth feature map;

[0058] 3.2 The network structure includes 3 3D convolutional layers (kernel size = 3×3×3) and 2 LSTM layers (hidden size = 128);

[0059] 4. Rice Maturity Perception Module:

[0060] Based on the fused data, use an SVM classifier to divide the rice maturity stages (vegetative growth stage, heading stage, milk ripening stage, dough ripening stage, full ripening stage)

[0061] 5. Output Module:

[0062] Output the maturity prediction results to the farm management system in GeoJSON format and provide an API interface for real-time query on the mobile side

[0063] II. Example 2 (Low-altitude Drone Remote Sensing)

[0064] 1. Flight Parameters: Flight altitude: 50 meters (ensuring a 5-centimeter resolution); Forward overlap rate: 80%, Side overlap rate: 70% (meeting the requirements for 3D reconstruction).

[0065] 2. Sensor Configuration

[0066] Multispectral bands: Blue (475nm), Green (560nm), Red (668nm), Red Edge (717nm), Near Infrared (842nm)

[0067] Spectral resolution: 10nm (FWHM)

[0068] 3. Data Processing Flow

[0069] 3.1 Generate orthophotos and DSM (Digital Surface Model) through Pix4D software

[0070] 3.2 Calculate OSAVI (Optimized Soil Adjusted Vegetation Index):

[0071]

[0072] 3.3 Combining the stem moisture content data from ground sampling, a random forest regression model is used to predict the maturity (R 2 > 0.92).

[0073] III. Example 3 (Deep learning fusion method)

[0074] 1. Input data:

[0075] 1.1 Satellite data: NDVI time series of Sentinel-2 (10 m / 5 days)

[0076] 1.2 UAV data: EVI2 index (Enhanced Vegetation Index)

[0077] 1.3 Ground data: Plant height (PH) and tiller number (TN)

[0078] 2. Model architecture

[0079] 2.1 A graph neural network (GNN) is adopted, where:

[0080] Node features: NDVI + EVI2 + PH + TN

[0081] Edge weights: Gaussian kernel function based on the Euclidean distance of the field

[0082] 2.2 Training strategy:

[0083] Loss function: Cross-entropy loss + Temporal consistency constraint

[0084] Optimizer: Adam (lr = 0.001)

[0085] 3. Output: Generate the field-level maturity probability distribution (Softmax output)

[0086] IV. Example 4 (Example of decision support system)

[0087] 1. Input data:

[0088] 1.1 Current maturity: 85% of the fields are in the dough stage (from the sensing module)

[0089] 1.2 Historical data: Yield in the same period in the past 3 years (database query)

[0090] 1.3 Meteorological data: Temperature and precipitation probability for the next 15 days (ECMWF forecast)

[0091] 2. Prediction model:

[0092] 2.1 Harvest Time Prediction: XGBoost Regression (MAE = 1.2 days)

[0093] 2.2 Pest and Disease Early Warning: LSTM Risk Score (Risk Level of Rice Blast: Level 3)

[0094] 3. Output Decisions:

[0095] 3.1 Suggested Harvest Time: September 15, 2025 ± 2 days

[0096] 3.2 Fertilization Suggestion: 10 kg / acre of urea (priority for the eastern block)

[0097] 3.3 Early Warning Notice: High humidity weather in the next 7 days. It is recommended to spray azoxystrobin.

[0098] V. Example 5 (Example of Edge Computing Deployment)

[0099] A lightweight edge computing implementation plan includes:

[0100] 1. Hardware Configuration:

[0101] 1.1 Edge Node: NVIDIA Jetson AGX Xavier (Computing Power: 32 TOPS)

[0102] 1.2 Communication Protocol: 5G Private Network (Uplink Bandwidth > 100 Mbps)

[0103] 2. Software Architecture:

[0104] 2.1 Data Preprocessing: TensorRT-Accelerated Standardized Pipeline (Latency < 50 ms)

[0105] 2.2 Inference Model: Quantized MobileNetV3 (INT8 Precision, Model Size: 3.5 MB)

[0106] 3. Federated Learning:

[0107] 3.1 Local Model Update Period: 24 hours

[0108] 3.2 Global Model Aggregation: Based on the FedAvg algorithm (Aggregation Server: AWS EC2)

[0109] The above are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited thereby. For those skilled in the art, the present invention can have various changes and modifications. Any changes, modifications, substitutions, integrations, and parameter changes made to these embodiments within the spirit and principle of the present invention by means of conventional substitutions or capable of achieving the same functions without departing from the principle and spirit of the present invention fall within the protection scope of the present invention.

Claims

1. A rice maturity perception system based on the fusion of multi-source heterogeneous agricultural situation data, characterized in that: Including: A data acquisition module, which is used to obtain multi-temporal and multi-source heterogeneous agricultural situation data from satellites, multi-spectral sensors carried by unmanned aerial vehicles (UAVs), and ground sampling. A data preprocessing module, which is used to align and standardize the collected multi-source heterogeneous data. A data fusion module, which is used to fuse the preprocessed multi-source heterogeneous data through a unified intelligent model framework to generate fused agricultural situation data. A rice maturity perception module, which is used to perceive and predict the maturity of rice in real time based on the fused agricultural situation data by using intelligent perception algorithms. An output module, which is used to output the rice maturity perception results to an agricultural production decision-making system.

2. The rice maturity perception system based on multi-source heterogeneous agricultural situation data fusion according to claim 1, characterized in that: The data acquisition module includes: A satellite remote sensing data acquisition unit, which is used to obtain the growth situation of rice on a large scale. A UAV low-altitude remote sensing data acquisition unit, which is used to obtain parameters such as high-resolution rice canopy coverage rate and leaf area index. A ground sampling data acquisition unit, which is used to obtain ground measured data such as rice height and vegetation index.

3. The rice maturity perception system based on multi-source heterogeneous agricultural situation data fusion according to claim 1, wherein: The data preprocessing module includes: A data alignment unit, which is used to synchronize the time and align the space of data from different times and different sources. A data standardization unit, which is used to convert data from different sources into a unified format and standard.

4. A rice maturity perception system based on the fusion of multi-source heterogeneous agricultural situation data as claimed in claim 1, characterized in that: The data fusion module adopts a deep learning model, including but not limited to a convolutional neural network (CNN), a recurrent neural network (RNN), or a graph neural network (GNN), which is used to process the spatio-temporal features of multi-source heterogeneous data.

5. The rice maturity perception system based on multi-source heterogeneous agricultural situation data fusion according to claim 1, characterized in that: The decision support module, based on the perception results, combines historical data and meteorological data, and uses a prediction model to generate crop growth trend predictions and pest warnings, and provides precise fertilization, irrigation, and pest control suggestions.

6. The rice maturity perception system based on multi-source heterogeneous agricultural situation data fusion according to claim 1, characterized in that: The intelligent perception module, through UAV low-altitude remote sensing technology, combines the data obtained by multi-spectral sensors, and uses machine learning algorithms to calculate the growth parameters of crops in real time and generate a high-precision agricultural situation map.

7. A method for perceiving rice maturity based on the fusion of multi-source heterogeneous agricultural situation data, characterized in that: S1: Obtain multi-temporal and multi-source heterogeneous agricultural situation data through satellites, multi-spectral sensors carried by UAVs, and ground sampling devices; S2: Preprocess the collected multi-source heterogeneous data, including denoising, spatio-temporal alignment, and standardization processing, to generate a unified spatio-temporal data set; S3: Based on a unified intelligent model framework, fuse the multi-source heterogeneous data to generate a comprehensive agricultural situation data set; S4: Use UAV low-altitude remote sensing technology to perceive key growth parameters such as the height, canopy coverage rate, leaf area, and vegetation index of rice in real time based on the fused data; S5: Generate agricultural production decision-making suggestions based on the perception results and provide real-time data support.

8. The rice maturity perception method based on multi-source heterogeneous agricultural situation data fusion according to claim 1, wherein: The data fusion in step 3 adopts a deep learning model, including but not limited to a convolutional neural network (CNN), a recurrent neural network (RNN), or a graph neural network (GNN), which is used to extract the spatio-temporal features of multi-source heterogeneous data.

9. The rice maturity perception method based on multi-source heterogeneous agricultural situation data fusion according to claim 1, characterized in that: The UAV low-altitude remote sensing technology in step 4 combines multi-spectral sensor data, and uses machine learning algorithms to calculate the growth parameters of crops in real time and generate a high-precision agricultural situation map.

10. A rice maturity perception method based on multi-source heterogeneous agricultural situation data fusion according to claim 1, characterized in that: The decision-making support in step 5 is based on the perception results, combined with historical data and meteorological data, and uses a prediction model to generate predictions of crop growth trends and pest and disease warnings, and provides precise fertilization, irrigation, and pest and disease control recommendations.

Images