Sky-ground net integrated high-standard farmland intelligent monitoring management system

Abstract

The invention discloses a sky-ground network integrated high-standard farmland intelligent monitoring management system, and belongs to the field of farmland monitoring management systems. The system is provided with a space monitoring data acquisition unit, an air monitoring data acquisition unit and a ground monitoring data acquisition unit; the system can be combined with satellite remote sensing equipment, unmanned aerial vehicle equipment and a plurality of sensor nodes in a multi-level manner to jointly form a sky-ground integrated data acquisition network for target farmland data acquisition, so that the comprehensiveness and accuracy of data acquisition are greatly improved; meanwhile, mapping layer three-dimensional modeling is carried out on data values acquired and converted by various sensors which are uniformly distributed at certain intervals, so that the display intuition, convenience and layering sense of different data on the target farmland are greatly improved; a user can more intuitively master the health states of crops at different positions of the target farmland and the advantages and disadvantages of the crop growth environment.

Description

Technical Field

[0001] The present invention relates to the field of farmland monitoring and management systems, and more specifically, to a high-standard farmland intelligent monitoring and management system integrating sky, ground and network. Background Art

[0002] With the advancement of agricultural modernization, precision agriculture and smart agriculture have gradually become important directions for improving farmland productivity and resource utilization efficiency. Especially in the construction and management of high-standard farmland, how to achieve comprehensive and real-time monitoring of crop growth environment and health status has become the key to modern agricultural management. Although existing users have made a lot of improvements in farmland monitoring and management technology, there are still the following shortcomings: 1. Incomplete data collection coverage. Existing farmland monitoring mainly relies on manual inspections or single equipment collection, which cannot achieve comprehensive data collection at multiple levels in space, air and ground, resulting in incomplete collection of crop growth environment data and pest and disease information, affecting the accuracy and comprehensiveness of data analysis; 2. Data visualization is not intuitive. Existing farmland monitoring systems usually only display information through simple data reports or two-dimensional charts, which makes it difficult for users to intuitively understand the health status of crops at different locations in the farmland and the advantages and disadvantages of the crop growth environment. It is impossible to quickly and accurately grasp the situation at different locations in the farmland, which is not conducive to accurate monitoring, management and decision-making of farmland; 3. The remote linkage control capability is limited. The existing farmland monitoring and management technology can only provide one-way data collection and simple monitoring, and lacks linkage and control functions with management equipment such as irrigation, fertilization, and pesticide spraying. It is not convenient for users to directly conduct effective real-time management and emergency response of farmland through the remote management system, which affects management efficiency; Therefore, we proposed an integrated sky-ground network high-standard farmland intelligent monitoring and management system to solve the above problems. Summary of the invention

[0003] 1. Technical issues to be solved In view of the problems existing in the prior art, the purpose of the present invention is to provide a high-standard farmland intelligent monitoring and management system integrating sky, ground and network. The system firstly, through the set space monitoring data acquisition unit, aerial monitoring data acquisition unit and ground monitoring data acquisition unit, can be respectively combined with satellite remote sensing equipment, unmanned aerial vehicle equipment and a number of sensor nodes in a multi-level manner to form a sky-ground integrated data acquisition network for target farmland data collection, which greatly improves the comprehensiveness and accuracy of data collection; at the same time, the system evenly distributes a number of sensor nodes used to collect ground data information on the target farmland at a certain interval, so that the various sensors included in the sensor nodes can be at different positions on the target farmland Simultaneous monitoring at all locations can improve the comprehensiveness of data collection at different locations on the target farmland, and on this basis, continue to use the data values ​​converted from multiple sensors to map the layers for three-dimensional modeling, which greatly improves the intuitiveness, convenience and layering of the display of different data on the target farmland, allowing users to more intuitively grasp the health status of crops at different locations on the target farmland and the advantages and disadvantages of the crop growth environment; at the same time, the system is also equipped with a number of equipment unit interfaces, which can directly communicate and connect equipment that can irrigate the target farmland, fertilize equipment, and disinfest and spray equipment, so as to achieve the effect of user remote control and improve the efficiency of effective real-time management of farmland directly through the remote management system and the ability to handle emergencies.

[0004] 2. Technical solution

[0005] To solve the above problems, the present invention adopts the following technical solutions.

[0006] The high-standard farmland intelligent monitoring and management system integrating sky, ground and network includes farmland data collection module, data transmission module, data processing and analysis module, monitoring and early warning module, decision support module, 3D modeling module and user interaction module; The farmland data collection module is internally provided with a space monitoring data collection unit, an aerial monitoring data collection unit and a ground monitoring data collection unit, and the space monitoring data collection unit, the aerial monitoring data collection unit and the ground monitoring data collection unit together constitute a sky-ground integrated data collection network for target farmland data collection, the space monitoring data collection unit is data-communication connected with a satellite remote sensing device, and the satellite remote sensing device is arranged above the target farmland, and is used to collect the location information, area data information and crop category information of the target farmland using satellite remote sensing technology above the target farmland, the aerial monitoring data collection unit is data-communication connected with an unmanned aerial vehicle device, and the unmanned aerial vehicle device is equipped with a multi-spectral high-definition camera and a thermal imaging camera, which are used to take pictures of the growth status and pest and disease conditions of crops in the target farmland, the ground monitoring data collection unit is data-communication connected with a plurality of sensor nodes, and the plurality of sensor nodes are evenly distributed on the target farmland at a certain interval, and the sensor nodes include a soil temperature and humidity sensor, a soil conductivity sensor, a soil pH sensor, an air temperature and humidity sensor, a carbon dioxide sensor, a wind speed sensor and a light intensity sensor; The data transmission module uses the Internet of Things (IoT) network technology to transmit and upload the data collected by the farmland data collection module to the data processing and analysis module in real time; The data processing and analysis module is based on a cloud computing platform and uses artificial intelligence and big data analysis and processing technology to clean and classify the data transmitted and uploaded by the data transmission module. At the same time, the data processing and analysis module also uses machine learning and deep learning algorithms to train the target farmland in combination with historical data to achieve automatic abnormal data recognition, pest and disease recognition, and crop growth status assessment, so as to facilitate the provision of personalized management suggestions to users in the later stage; The monitoring and early warning module is connected to the data processing and analysis module, and is used to monitor the results output by the data processing and analysis module in real time, identify potential risks of the target farmland, and issue corresponding early warnings; The decision support module provides decision suggestions on planting, fertilization, irrigation and pest control based on the potential risk results identified by the monitoring and early warning module; The user interaction module is communicatively connected to a mobile control terminal and a computer control terminal, and the user interaction module displays in real time through the mobile control terminal and the computer control terminal the data collected by the farmland data collection module, the potential risk results identified by the monitoring and early warning module, and the decision suggestions output by the decision support module, allowing the user to perform remote control; The three-dimensional modeling module is internally provided with a primary modeling unit, which establishes a proportionally reduced target farmland model and a crop model growing on the target farmland model by using three-dimensional modeling technology including but not limited to Solidworks according to the location information, area data information and crop category information of the target farmland collected by the spatial monitoring data collection unit, so as to facilitate user viewing and management.

[0007] Furthermore, the user interaction module is provided with a plurality of equipment unit interfaces for communicating and connecting equipment for irrigating the target farmland, equipment for fertilizing, and equipment for pest control and spraying pesticides, etc., so as to achieve the effect of user remote control.

[0008] Furthermore, the data processing and analysis module also combines the photos of crop growth status, extracts the plant height and the number of fruits of the crops in the photos, and generates corresponding crop growth status data values; The data processing and analysis module also combines the photos of crop pests and diseases, extracts the number of pests in the photos, and generates corresponding crop pest data values; The data processing and analysis module also combines the data collected by the soil temperature and humidity sensor, the soil conductivity sensor, the soil pH value sensor, the air temperature and humidity sensor, the carbon dioxide sensor, the wind speed sensor and the light intensity sensor to generate corresponding soil temperature and humidity data values, soil conductivity data values, soil pH value data values, air temperature and humidity data values, carbon dioxide data values, wind speed data values, and light intensity data values.

[0009] Furthermore, a secondary modeling unit is also provided inside the three-dimensional modeling module, and the secondary modeling unit generates a color layer suspended on the target farmland model according to various data values, for example, the crop growth status data values ​​are modeled in the form of a red layer, the crop pest data values ​​are modeled in the form of an orange layer, the soil temperature and humidity data values ​​are modeled in the form of a yellow layer, the soil conductivity data values ​​are modeled in the form of a green layer, the soil pH value data values ​​are modeled in the form of a cyan layer, the air temperature and humidity data values ​​are modeled in the form of a blue layer, the carbon dioxide data values ​​are modeled in the form of a purple layer, the wind speed data values ​​are modeled in the form of a black layer, and the illumination data values ​​are modeled in the form of a gray layer.

[0010] Furthermore, a layer mapping unit is also provided inside the three-dimensional modeling module. The layer mapping unit first presets standard values ​​for various data values, and presets standard colors for each color layer according to the preset standard values, and then maps the various data values ​​actually generated to the corresponding color layers. The greater the various data actually generated is than the preset standard value, the darker the color on the corresponding color layer. Conversely, the smaller the various data actually generated is than the preset standard value, the lighter the color on the corresponding color layer. Therefore, the depth of the color layer can intuitively reflect the quality of the crop growth environment.

[0011] Furthermore, the data processing and analysis module combines photos of crop growth status and uses computer vision technology, a deep learning model based on a convolutional neural network (CNN), and an image processing algorithm to extract and calculate the plant height and the number of fruits of the crop from the photos. The data processing and analysis module combines photos of crop disease and pest conditions and uses computer vision technology, a deep learning model based on a convolutional neural network (CNN), and an image processing algorithm to extract and calculate the number of pests from the photos.

[0012] Furthermore, the data processing and analysis module combines analog signals, such as voltage and current, collected by the soil temperature and humidity sensor, soil conductivity sensor, soil pH value sensor, air temperature and humidity sensor, carbon dioxide sensor, wind speed sensor and light intensity sensor, and then converts the analog signals into digital signals through an analog-to-digital converter (ADC), thereby generating corresponding soil temperature and humidity data values, soil conductivity data values, soil pH value data values, air temperature and humidity data values, carbon dioxide data values, wind speed data values, and light intensity data values.

[0013] 3. Beneficial effects

[0014] Compared with the prior art, the advantages of the present invention are: (1) In this scheme, the system firstly combines the space monitoring data collection unit, the aerial monitoring data collection unit and the ground monitoring data collection unit with satellite remote sensing equipment, UAV equipment and several sensor nodes at multiple levels to form a sky-ground integrated data collection network for target farmland data collection, which greatly improves the comprehensiveness and accuracy of data collection; (2) In this scheme, the system evenly distributes several sensor nodes used to collect ground data information on the target farmland at a certain distance, so that the multiple sensors included in the sensor nodes can monitor at different locations on the target farmland at the same time, improving the comprehensiveness of data collection at different locations on the target farmland, and on this basis, continue to use the data values ​​converted from the multiple sensors to map the layers for three-dimensional modeling, which greatly improves the intuitiveness, convenience and layering of the display of different data on the target farmland, allowing users to more intuitively grasp the health status of crops and the advantages and disadvantages of the crop growth environment at different locations on the target farmland; (3) In this solution, the system is also provided with a number of equipment unit interfaces, which can directly communicate and connect the equipment for irrigating the target farmland, the equipment for fertilizing, and the equipment for pest control and spraying, so as to achieve the effect of user remote control and improve the efficiency of effective real-time management of farmland directly through the remote management system and the ability of emergency handling. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic diagram of the sky-ground integrated data acquisition network of the present invention; Figure 2 It is a schematic diagram of the overall system architecture of the present invention; Figure 3 The system part of the present invention is shown in FIG. Figure 1 ; Figure 4 The system part of the present invention is shown in FIG. Figure 2 .

[0016] Description of the numbers in the figure: 1. Farmland data collection module; 101. Space monitoring data collection unit; 102. Aerial monitoring data collection unit; 103. Ground monitoring data collection unit; 2. Data transmission module; 3. Data processing and analysis module; 4. Monitoring and early warning module; 5. Decision support module; 6. 3D modeling module; 601. primary modeling unit; 602. secondary modeling unit; 603. layer mapping unit; 7. User interaction module; 701. Mobile control terminal; 702. Computer control terminal; 703. Equipment unit interface. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the specification of the present invention; it is obvious that the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments, and all other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making creative work are within the scope of protection of the present invention.

[0018] Example 1

[0019] See also Figure 1-Figure 4 The working principle of the integrated high-standard farmland intelligent monitoring and management system of sky-ground network is as follows: 1. The principle of data collection is: First, the satellite remote sensing equipment above the target farmland is remotely connected to the space monitoring data acquisition unit 101 of the farmland data acquisition module 1 in the system, so that the space monitoring data acquisition unit 101 can collect the location information, area data information and crop type information of the target farmland by using satellite remote sensing technology. At the same time, the drone equipment equipped with a multi-spectral high-definition camera and a thermal imaging camera is remotely connected to the air monitoring data acquisition unit 102 of the farmland data acquisition module 1 in the system, so that the air monitoring data acquisition unit 102 can take pictures of the growth status and pest and disease conditions of the crops in the target farmland and obtain the corresponding pictures. A plurality of sensor nodes consisting of conductivity sensors, soil pH value sensors, air temperature and humidity sensors, carbon dioxide sensors, wind speed sensors and light intensity sensors are evenly distributed on the target farmland at a certain interval, and then remotely communicated with the ground monitoring data acquisition unit 103 of the farmland data acquisition module 1 in the system, so that the ground monitoring data acquisition unit 103 can collect soil temperature and humidity simulation signals, soil conductivity simulation signals, soil pH value simulation signals, air temperature and humidity simulation signals, carbon dioxide simulation signals, wind speed simulation signals and light intensity simulation signals at different positions of the target farmland, thereby forming a sky-ground integrated data acquisition network for target farmland data acquisition; (II) The principles of data processing and analysis are as follows: First, the Internet of Things (IoT) network technology is used to transmit and upload the collected photo information and various analog signals to the data processing and analysis module 3 in real time through the data transmission module 2. Then, based on the cloud computing platform, the photo information and various analog signal data are cleaned and classified using artificial intelligence and big data analysis and processing technology within the data processing and analysis module 3. At the same time, the data processing and analysis module 3 also uses machine learning and deep learning algorithms to train the target farmland with historical data to achieve automatic abnormal data recognition, pest and disease recognition, and crop growth status assessment, so as to facilitate the provision of personalized management suggestions to users in the later stage. Then, based on the data processing and analysis module 3, in combination with the photos of the crop growth status and using computer vision technology, a deep learning model based on a convolutional neural network (CNN), and an image processing algorithm, the plant height and the number of fruits of the crop are extracted and calculated from the photos to generate corresponding crop growth status data values. The data processing and analysis module 3 combines the photos of the crop disease and insect pest conditions and uses computer vision technology, a deep learning model based on a convolutional neural network (CNN), and an image processing algorithm to extract and calculate the number of pests from the photos to generate corresponding crop pest data values. The data processing and analysis module 3 combines the analog signals collected by the soil temperature and humidity sensor, the soil conductivity sensor, the soil pH value sensor, the air temperature and humidity sensor, the carbon dioxide sensor, the wind speed sensor, and the light intensity sensor, such as voltage and current, and then converts the analog signals into digital signals through an analog-to-digital converter (ADC), thereby generating corresponding soil temperature and humidity data values, soil conductivity data values, soil pH value data values, air temperature and humidity data values, carbon dioxide data values, wind speed data values, and light intensity data values; (III) The principle of 3D modeling is: First, the primary modeling unit 601 in the three-dimensional modeling module 6 establishes a proportionally reduced target farmland model and a crop model growing on the target farmland model based on the location information, area data information and crop category information of the target farmland collected by the space monitoring data collection unit 101, using three-dimensional modeling technology including but not limited to Solidworks, so as to facilitate user viewing and management. Then the secondary modeling unit 602 generates a color layer suspended on the target farmland model based on the above-mentioned various data values. For example, the crop growth status data value is modeled in the form of a red layer, the crop pest data value is modeled in the form of an orange layer, the soil temperature and humidity data value is modeled in the form of a yellow layer, the soil conductivity data value is modeled in the form of a green layer, the soil pH value data value is modeled in the form of a cyan layer, the air temperature and humidity data value is modeled in the form of a blue layer, the carbon dioxide data value is modeled in the form of a purple layer, and the wind speed data value is modeled in the form of a blue layer. The data values ​​are modeled in the form of a black layer, and the illuminance data values ​​are modeled in the form of a gray layer. Then, the layer mapping unit 603 presets standard values ​​for various data values ​​in advance, and presets standard colors for each color layer according to the preset standard values. Then, the various data values ​​actually generated are mapped to the corresponding color layers. The larger the various data actually generated is than the preset standard value, the darker the color on the corresponding color layer. On the contrary, the smaller the various data actually generated is than the preset standard value, the lighter the color on the corresponding color layer. Thus, the depth of the color layer can intuitively reflect the quality of the crop growth environment. At this time, the model diagram of the color layer corresponding to various sensors or devices suspended on the target farmland model can be presented. In addition, each layer of the color layer model not only has the corresponding sensor node position adapted to the target farmland position, but also has the layer color change of the corresponding sensor node position, so that the established three-dimensional model is displayed more clearly and intuitively. (IV) The principle of user remote control is: Based on the above, the monitoring and early warning module 4 monitors the various data values ​​of the data processing and analysis module 3 in real time. If the various real-time data values ​​processed by the data processing and analysis module 3 are greater than or less than the preset standard value, it means that the monitoring project corresponding to the data value has a risk. At this time, the corresponding early warning information can be issued, and notified and displayed through the mobile control terminal and the computer control terminal. Then, the decision support module 5 provides decision suggestions on planting, fertilization, irrigation and pest control based on the potential risk results identified by the monitoring and early warning module 4, so as to facilitate the user to make active interference in a timely and correct manner in the future; When the user actively interferes, the equipment that can irrigate the target farmland, the equipment that applies fertilizers, and the equipment that removes insects and sprays pesticides are first connected to the corresponding equipment unit interface 703 for communication. At this time, the user can control these devices to operate through the mobile control terminal and the computer control terminal, so as to deal with problems that arise in the crop planting process in a timely manner.

[0020] Example 2

[0021] In view of the above embodiment 1, for further description, please refer to Figure 1-Figure 4 The technical means to realize the high-standard farmland intelligent monitoring and management system with integrated sky-ground network are as follows: First, the data is cleaned based on the artificial intelligence and big data analysis and processing technology of the cloud computing platform. The cleaning process is to first use the moving average algorithm technology of the data smoothing algorithm to smooth the continuous time series data by calculating the average value of the data in a certain time window, remove short-term fluctuation noise, and make the data trend clearer. Then, the algorithm technology based on statistical principles of outlier detection is used to calculate the mean and standard deviation of the data. Data that deviates from the mean by more than 3 times the standard deviation is regarded as an outlier to ensure the accuracy and reliability of the data. Then, the KMeans clustering algorithm technology of the clustering analysis algorithm is used to divide the data into different categories according to the similarity of the data. For example, soil-related data (soil temperature and humidity, conductivity, pH value, etc.) are clustered into one category, air environment-related data (air temperature and humidity, carbon dioxide concentration, wind speed, etc.) are clustered into one category, and crop growth-related data (crop height, fruit quantity, pest and disease conditions, etc.) are clustered into one category. Such classification helps to conduct specialized analysis and processing for different types of data, improving analysis efficiency and accuracy; For photos of crop growth status, target detection algorithms (such as Faster RCNN) are used to identify crop plants and fruits in the photos. Then, based on the deep learning model of the convolutional neural network (CNN) and the image processing algorithm, the crop plant height (by converting the ratio of the pixel height of the plant in the photo to the actual shooting height) and the number of fruits (by counting the identified fruit targets) are calculated, thereby generating the corresponding crop growth status data value. For photos of pests and diseases, target detection algorithms are also used to identify pests, and then the number of pests is counted to generate crop pest data values.

[0022] The analog signals (such as voltage and current) collected by the soil temperature and humidity sensors are converted into digital signals through the analog-to-digital converter (ADC) to obtain the soil temperature and humidity data values. At the same time, the data collected by the soil conductivity sensor, soil pH sensor, air temperature and humidity sensor, carbon dioxide sensor, wind speed sensor and light intensity sensor are converted and calibrated accordingly to generate accurate data values, ensuring that the data is within the appropriate numerical range and has actual physical meaning.

[0023] The source of training data is to use the target farmland combined with historical data for training. Years of crop growth data, environmental data, and corresponding pest and disease occurrence and crop yield data are used as training sets to provide comprehensive samples for model training. Then, the support vector machine (SVM) in machine learning and the long short-term memory network (LSTM) algorithm in deep learning are used to train the model. In pest and disease identification and crop growth status assessment, classification and judgment can be made according to data characteristics. At the same time, the long short-term memory network is suitable for processing time series data, which has advantages in analyzing dynamic changes in crop growth and predicting pest and disease development trends. In actual monitoring, the currently collected soil moisture, temperature, crop growth height and other data are input into the trained model. The model judges whether the crop growth is normal and whether there is a risk of pests and diseases based on the pre-learned pattern. For example, by learning the trend of soil fertility changes, the model can identify abnormal conditions such as a sudden drop in soil fertility; by learning the crop growth rate pattern, it can judge whether the crop growth rate is abnormal.

[0024] For crop growth status assessment, the principal component analysis (PCA) algorithm is used to reduce the dimension of multiple growth-related data (such as plant height, fruit number, chlorophyll content, etc.), extract the main characteristic components, and then predict the crop yield trend through the constructed regression model (such as multivariate linear regression), so as to comprehensively evaluate the crop growth status. In terms of pest and disease identification, the convolutional neural network (CNN) is used to extract features from pest and disease photos, build a classification model, identify the types of pests and diseases (such as aphids, leaf spot, etc.), and combine the time series analysis algorithm (such as the ARIMA model) to analyze the time pattern of pest and disease occurrence, predict the development trend of pests and diseases, so as to take prevention and control measures in advance.

[0025] For photos of crop growth status, target detection algorithms in computer vision technology (such as FasterRCNN) are used. This algorithm can accurately identify the location of crop plants and fruits in photos. For example, in a photo containing a large area of ​​corn crops, the target detection algorithm can clearly outline the outline of each corn plant and the location of corn cobs (fruits), and then extract features with the help of a deep learning model based on a convolutional neural network (CNN). The CNN model can learn various characteristic patterns of crop plants and fruits by training a large amount of image data labeled with crop features. For example, it can identify the leaf shape and texture of corn plants and the size and color of corn cobs, and then use image processing algorithms to calculate the height of crop plants. This is achieved by converting the proportional relationship between the pixel height of the plant in the photo and the actual shooting height. For example, if the parameters of the shooting camera (such as focal length, shooting distance, etc.) are known, the actual plant height can be calculated from the pixel height of the plant in the photo according to the principle of similar triangles. At the same time, the identified fruit targets are counted to obtain the number of fruits. These characteristic information such as plant height and fruit number constitute part of the crop growth status data value. The plant height, fruit quantity and other data calculated above are integrated to generate data values ​​that can fully reflect the growth status of crops. For example, for an orchard, the average plant height and fruit quantity distribution of each fruit tree can be obtained. These data can be used as a basis for evaluating the overall growth status of the orchard, such as judging whether the fruit trees are growing vigorously and whether management measures such as pruning or fertilization are needed.

[0026] For photos of crop pests and diseases, the target detection algorithm in computer vision technology is also used to identify pests. After being trained with a large amount of pest and disease image data, the algorithm can accurately distinguish pests from crop backgrounds and other irrelevant objects. For example, in a photo of vegetable leaves infested by aphids, the target detection algorithm can quickly locate and mark the location of aphids, and then further analyze the pest characteristics based on the deep learning model and image processing algorithm of the convolutional neural network (CNN) to ensure accurate identification of pest species (although species identification is not mentioned in the document, this is a common further processing method) and accurate counting of the number of pests. For example, the morphology, color and other features of aphids are identified by the CNN model, and the target segmentation and counting methods in the image processing algorithm are combined to obtain the accurate number of aphids in the photo. The number of pests counted is used as the crop pest data value. This data value can intuitively reflect the degree of crop damage by pests and diseases. For example, in a farmland, if photos from different locations show that the number of pests is large and widely distributed, it means that the farmland faces a serious risk of pests and diseases, and prevention and control measures need to be taken in time.

[0027] Soil temperature and humidity sensors, soil conductivity sensors, soil pH sensors, air temperature and humidity sensors, carbon dioxide sensors, wind speed sensors, and light intensity sensors are distributed on the target farmland to collect environmental data in real time. These sensors collect analog signals, such as voltage and current. For example, the soil temperature and humidity sensor outputs corresponding voltage signals according to the changes in soil moisture content and temperature; the light intensity sensor generates current signals of different sizes according to the different light intensity. The analog-to-digital converter (ADC) converts these analog signals into digital signals. The ADC converts continuously changing analog signals into discrete digital values ​​by sampling, quantizing, and encoding analog signals. For example, for the voltage signal output by the soil moisture sensor, the ADC collects the voltage value at a certain sampling frequency and converts it into the corresponding digital code, which represents the size of the soil moisture. After ADC conversion, the soil temperature and humidity data values, soil conductivity data values, soil pH data values, air temperature and humidity data values, carbon dioxide data values, wind speed data values, and light intensity data values ​​are obtained. Then, the corresponding range conversion and calibration processing are performed on these data. Range conversion is the conversion of the digital value output by the sensor into the numerical range of the actual physical quantity, such as converting the digital output of the soil moisture sensor into a percentage moisture value (such as 0 - 100%). Calibration processing is to ensure the accuracy of the data by comparing and adjusting with standard measuring equipment or known standard values ​​to eliminate the influence of factors such as sensor errors and environmental interference. For example, soil moisture sensors are calibrated regularly using standard soil samples to make the measurement data more accurate and reliable. These processed data values ​​can accurately reflect the actual conditions of farmland soil and environment, providing a basis for subsequent data analysis and decision-making.

[0028] Example 3

[0029] In view of the above-mentioned embodiment 1 and embodiment 2, for further description, please refer to Figure 1-Figure 4 , the sky-ground-network integrated high-standard farmland intelligent monitoring and management system of the present invention is described in detail in conjunction with specific embodiments: Data collection and transmission Space monitoring data collection: The space monitoring data acquisition unit 101 is connected to the satellite remote sensing equipment arranged above the target farmland. The satellite remote sensing equipment uses its advanced remote sensing technology to accurately collect the location information, area data information and crop type information of the target farmland. For example, in a large high-standard farmland, the satellite remote sensing equipment scans the farmland at a specific time interval (such as once a week) to obtain the boundary coordinates, total area and type distribution of the crop planted in the farmland. This information is transmitted to the data processing and analysis module 3 in real time through the data transmission module 2 using the Internet of Things (IoT) network technology.

[0030] Aerial monitoring data collection: The aerial monitoring data acquisition unit 102 works in conjunction with the drone equipment equipped with a multispectral high-definition camera and a thermal imaging camera. When it is necessary to monitor the growth status of crops and the situation of pests and diseases, the operator controls the drone to take off through the mobile control terminal 701 or the computer control terminal 702, and shoots the target farmland according to the preset route. For example, during the critical period of crop growth (such as flowering period and fruiting period), the drone performs flight shooting twice a week. The multispectral high-definition camera can obtain the reflection information of crops in different spectral bands, so as to analyze the growth indicators such as chlorophyll content and nitrogen content of crops; the thermal imaging camera can detect the surface temperature distribution of crops and promptly discover the abnormal heating area of ​​crops caused by pests and diseases. The photographed photo data is transmitted to the data processing and analysis module 3 in real time through wireless communication technology.

[0031] Ground monitoring data collection: The ground monitoring data acquisition unit 103 is connected to a number of sensor nodes evenly distributed on the target farmland at a certain interval (such as one every 10 meters). These sensor nodes include soil temperature and humidity sensors, soil conductivity sensors, soil pH sensors, air temperature and humidity sensors, carbon dioxide sensors, wind speed sensors, and light intensity sensors. The sensor nodes collect data in real time. For example, the soil temperature and humidity sensor collects data every 15 minutes, converts the collected analog signals (such as voltage and current) into digital signals through an analog-to-digital converter (ADC), and then transmits the data to the data processing and analysis module 3 through the Internet of Things technology.

[0032] Data processing and analysis Data cleaning and classification: The data processing and analysis module 3 is based on the cloud computing platform. After receiving various data from the data transmission module 2, it uses artificial intelligence and big data analysis and processing technology to clean the data. First, a data smoothing algorithm, such as the moving average method, is used to smooth the continuous time series data (such as the data of soil moisture changes over time) to remove short-term fluctuation noise and make the data trend clearer. For outlier detection, a method based on statistical principles is used, such as calculating the mean and standard deviation of the data, and the data that deviates from the mean by more than 3 times the standard deviation is regarded as an outlier. Then, a cluster analysis algorithm (such as KMeans clustering) is used to divide the data into different categories according to the similarity of the data, for example, soil-related data (soil temperature and humidity, conductivity, pH value, etc.) are clustered into one category, air environment-related data (air temperature and humidity, carbon dioxide concentration, wind speed, etc.) are clustered into one category, and crop growth-related data (crop height, fruit quantity, pest and disease conditions, etc.) are clustered into one category, which is convenient for subsequent targeted analysis.

[0033] Feature extraction and data value generation: Combined with the photos of crop growth status, the target detection algorithm in computer vision technology (such as FasterRCNN) is used to first identify the crop plants and fruits in the photos, and then the deep learning model based on convolutional neural network (CNN) and image processing algorithm are used to calculate the plant height of the crop (by converting the ratio of the pixel height of the plant in the photo to the actual shooting height), the number of fruits (by counting the identified fruit targets) and other features to generate the corresponding crop growth status data value. For photos of pests and diseases, the target detection algorithm is also used to identify pests, and then the number of pests is counted to generate crop pest data values. Combined with the data collected by various sensors, the analog signal collected by the soil temperature and humidity sensor is converted into a digital signal through an analog-to-digital converter (ADC) to obtain the soil temperature and humidity data value; the data collected by the soil conductivity sensor, soil pH sensor, air temperature and humidity sensor, carbon dioxide sensor, wind speed sensor and light intensity sensor are converted and calibrated accordingly to generate accurate soil conductivity data values, soil pH data values, air temperature and humidity data values, carbon dioxide data values, wind speed data values, and light intensity data values.

[0034] Abnormal data identification, pest and disease identification and crop growth status assessment: The support vector machine (SVM) in machine learning and the long short-term memory network (LSTM) algorithm in deep learning are used to train the target farmland combined with historical data. Years of crop growth data, environmental data, and corresponding pest and disease occurrence and crop yield data are used as training sets to train models that can identify abnormal data (such as a sudden drop in soil fertility, abnormal crop growth rate, etc.), pests and diseases (such as identifying the type and severity of pests and diseases through the number of pests and crop surface damage characteristics) and assess crop growth status (such as judging whether the crop is in a state of water shortage, fertilizer shortage, vigorous growth, etc.). In actual monitoring, the trained model is used to analyze and process real-time data. For example, the currently collected soil moisture, temperature, crop growth height and other data are input into the model, and the model judges whether the crop growth is normal and whether there is a risk of pests and diseases based on the pre-learned pattern.

[0035] For crop growth status assessment, the principal component analysis (PCA) algorithm is used to reduce the dimension of multiple growth-related data (such as plant height, number of fruits, chlorophyll content, etc.), extract the main characteristic components, and then predict the crop yield trend through the constructed regression model (such as multivariate linear regression), so as to evaluate the crop growth status. In terms of pest and disease identification, the convolutional neural network (CNN) is used to extract features from pest and disease photos, build a classification model, and identify the types of pests and diseases (such as aphids, leaf spot, etc.). At the same time, combined with the time series analysis algorithm (such as the ARIMA model), the time pattern of pest and disease occurrence is analyzed, the development trend of pests and diseases is predicted, and prevention and control measures are taken in advance.

[0036] 3D Modeling and Visualization Primary Modeling: The primary modeling unit 601 in the 3D modeling module 6 uses the Solidworks 3D modeling technology to establish a proportionally reduced target farmland model and a crop model growing thereon based on the location information, area data information, and crop category information of the target farmland collected by the spatial monitoring data collection unit 101. For example, for a rectangular farmland of 100 mu, a 3D model is established at a scale of 1:1000, and the boundaries of the farmland, ridges, and the distribution positions of different crops (such as wheat, corn, etc.) planted are clearly displayed in the model.

[0037] Secondary modeling and layer mapping: The secondary modeling unit 602 generates a color layer suspended on the target farmland model according to various data values. For example, the crop growth status data value is modeled in the form of a red layer. When the crop grows well, the red layer is lighter in color, and when the crop grows poorly, the color is darker; the crop pest data value is modeled in the form of an orange layer. The more pests there are, the darker the orange layer is. The layer mapping unit 603 presets standard values ​​for various data values ​​in advance. For example, the standard value of soil moisture is 40%-60%. When the actual soil moisture is 70%, the color of the yellow layer (soil temperature and humidity data value layer) will be darker than the standard humidity. In this way, users can intuitively understand the quality of the crop growth environment through the depth of the color layer.

[0038] Visual display: The user interaction module 7 displays the three-dimensional model of the target farmland model and various data layers in real time through the mobile control terminal 701 and the computer control terminal 702. The user can view the farmland model from different angles by touching the screen or operating the mouse, and click on different positions to view the detailed data values ​​of the position (such as soil moisture, crop growth height, etc.), which greatly improves the intuitiveness, convenience and layering of data display.

[0039] Monitoring, early warning and decision support Monitoring and early warning: The monitoring and early warning module 4 monitors the results output by the data processing and analysis module 3 in real time, and identifies potential risks in the target farmland. For example, when the crop pest data value exceeds a preset threshold (such as more than 10 pests per square meter), it is identified as a pest risk; when the soil moisture is lower than the standard value for many consecutive days and the crop growth status data shows that the crop grows slowly, it is identified as a drought risk. Once the risk is identified, a corresponding early warning is immediately issued, and the early warning information is notified to the user through the mobile control terminal 701 (such as push notification, SMS reminder) or the computer control terminal 702 (pop-up prompt).

[0040] Decision support: The decision support module 5 provides decision suggestions for planting, fertilization, irrigation and pest control based on the potential risk results identified by the monitoring and early warning module 4. For example, when drought risk is identified, it is recommended to start the irrigation equipment and provide reference values ​​for irrigation time and water volume; when pest risk is identified, it provides corresponding pesticide formulations and dosage recommendations based on the pest type, as well as the best time and method for application.

[0041] Remote Control Device connection and control: Several device unit interfaces 703 are provided on the user interaction module 7, and communication connections can be used for equipment for watering the target farmland, equipment for fertilizing, and equipment for pest control and spraying. For example, an intelligent irrigation system, an automatic fertilizer spreader, and drone plant protection equipment are connected through the device unit interface. The user can remotely control the start, stop, and operation parameter setting of these devices on the mobile control terminal 701 or the computer control terminal 702. When receiving a drought warning, the user can directly operate the smart irrigation system on the mobile phone and irrigate according to the irrigation parameters provided by the decision support module; when pests and diseases occur, the drone plant protection equipment can be remotely controlled to spray pesticides.

[0042] Real-time feedback and adjustments: During the operation of the equipment, data is continuously collected through sensor nodes, and the equipment operation effect is fed back in real time. For example, during the irrigation process, the soil moisture sensor monitors the changes in soil moisture in real time. When the preset suitable humidity is reached, the user can stop the irrigation equipment through remote control. At the same time, according to the actual situation, the user can adjust the equipment operation parameters according to the feedback data, such as adjusting the amount of fertilizer, the concentration of the pesticide, etc., to achieve precise management. It can be seen from the above embodiments that the sky-ground network integrated high-standard farmland intelligent monitoring and management system of the present invention can comprehensively and accurately collect farmland data, perform efficient data processing and analysis, realize intuitive and visual display through three-dimensional modeling, timely monitor and warn and provide decision-making support, and realize remote control at the same time, effectively improving the monitoring and management level and production efficiency of high-standard farmland.

[0043] The above is only a preferred specific implementation of the present invention; however, the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solution and its improved conception within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. The high-standard farmland intelligent monitoring and management system with integrated sky and ground network is characterized by: It includes a farmland data collection module (1), a data transmission module (2), a data processing and analysis module (3), a monitoring and early warning module (4), a decision support module (5), a three-dimensional modeling module (6) and a user interaction module (7); The farmland data collection module (1) is internally provided with a space monitoring data collection unit (101), an aerial monitoring data collection unit (102) and a ground monitoring data collection unit (103); the space monitoring data collection unit (101), the aerial monitoring data collection unit (102) and the ground monitoring data collection unit (103) together constitute a sky-ground integrated data collection network for collecting target farmland data; the space monitoring data collection unit (101) is connected to a satellite remote sensing device for data communication; and the satellite remote sensing device is arranged above the target farmland and is used to collect the position information of the target farmland using satellite remote sensing technology above the target farmland. The aerial monitoring data collection unit (102) is connected to a drone device for data communication, and the drone device is equipped with a multi-spectral high-definition camera and a thermal imaging camera for taking photos of the growth status and pest and disease conditions of the crops in the target farmland. The ground monitoring data collection unit (103) is connected to a plurality of sensor nodes for data communication, and the plurality of sensor nodes are evenly distributed on the target farmland at a certain distance. The sensor nodes include a soil temperature and humidity sensor, a soil conductivity sensor, a soil pH sensor, an air temperature and humidity sensor, a carbon dioxide sensor, a wind speed sensor, and a light intensity sensor; The data transmission module (2) utilizes Internet of Things (IoT) network technology to transmit and upload the data collected by the farmland data collection module (1) to the data processing and analysis module (3) in real time; The data processing and analysis module (3) is based on a cloud computing platform and uses artificial intelligence and big data analysis and processing technology to clean and classify the data transmitted and uploaded by the data transmission module (2). At the same time, the data processing and analysis module (3) also uses machine learning and deep learning algorithms to train the target farmland in combination with historical data to achieve automated abnormal data recognition, pest and disease recognition, and crop growth status assessment; The monitoring and early warning module (4) is connected to the data processing and analysis module (3) and is used to monitor the results output by the data processing and analysis module (3) in real time, identify potential risks of the target farmland, and issue corresponding early warnings; The decision support module (5) provides decision suggestions on planting, fertilization, irrigation and pest control based on the potential risk results identified by the monitoring and early warning module (4); The user interaction module (7) is communicatively connected to a mobile control terminal (701) and a computer control terminal (702), and the user interaction module (7) displays in real time through the mobile control terminal (701) and the computer control terminal (702) the data collected by the farmland data collection module (1), the potential risk results identified by the monitoring and early warning module (4), and the decision suggestions output by the decision support module (5), allowing the user to perform remote control; The three-dimensional modeling module (6) is internally provided with a primary modeling unit (601). The primary modeling unit (601) uses three-dimensional modeling technology including but not limited to Solidworks to establish a proportionally reduced target farmland model and a crop model growing on the target farmland model based on the location information, area data information and crop category information of the target farmland collected by the spatial monitoring data collection unit (101), so as to facilitate user viewing and management.

2. The high-standard farmland intelligent monitoring and management system with integrated sky-ground network according to claim 1 is characterized by: The user interaction module (7) is provided with a plurality of equipment unit interfaces (703) for communicating with equipment for irrigating the target farmland, equipment for fertilizing, and equipment for removing insects and spraying pesticides, thereby achieving a remote control effect by the user.

3. The high-standard farmland intelligent monitoring and management system with integrated sky-ground network according to claim 1 is characterized by: The data processing and analysis module (3) also combines the photos of crop growth status, extracts the plant height and the number of fruits of the crops in the photos, and generates corresponding crop growth status data values; The data processing and analysis module (3) also combines the photos of crop pests and diseases, extracts the number of pests in the photos, and generates corresponding crop pest data values; The data processing and analysis module (3) also generates corresponding soil temperature and humidity data values, soil conductivity data values, soil pH value data values, air temperature and humidity data values, carbon dioxide data values, wind speed data values, and light intensity data values ​​in combination with data collected by the soil temperature and humidity sensor, the soil conductivity sensor, the soil pH value sensor, the air temperature and humidity data values, the carbon dioxide data values, the wind speed data values, and the light intensity data values.

4. The sky-ground-network integrated high-standard farmland intelligent monitoring and management system according to claim 3 is characterized by: The three-dimensional modeling module (6) is also provided with a secondary modeling unit (602) therein. The secondary modeling unit (602) generates a color layer suspended on the target farmland model according to various data values, for example, the crop growth status data value is modeled in the form of a red layer, the crop pest data value is modeled in the form of an orange layer, the soil temperature and humidity data value is modeled in the form of a yellow layer, the soil conductivity data value is modeled in the form of a green layer, the soil pH value data value is modeled in the form of a cyan layer, the air temperature and humidity data value is modeled in the form of a blue layer, the carbon dioxide data value is modeled in the form of a purple layer, the wind speed data value is modeled in the form of a black layer, and the illumination data value is modeled in the form of a gray layer.

5. The high-standard farmland intelligent monitoring and management system with integrated sky-ground network according to claim 4 is characterized by: The three-dimensional modeling module (6) is also provided with a layer mapping unit (603) therein. The layer mapping unit (603) first presets standard values ​​for various data values, and presets standard colors for various color layers according to the preset standard values, and then maps the various data values ​​actually generated to the corresponding color layers. The greater the various data actually generated is than the preset standard value, the darker the color on the corresponding color layer is. Conversely, the smaller the various data actually generated is than the preset standard value, the lighter the color on the corresponding color layer is. Thus, the depth of the color layer can intuitively reflect the quality of the crop growth environment.

6. The high-standard farmland intelligent monitoring and management system with integrated sky-ground network according to claim 3 is characterized by: The data processing and analysis module (3) combines photos of crop growth status and uses computer vision technology, a deep learning model based on a convolutional neural network (CNN), and an image processing algorithm to extract and calculate the plant height and the number of fruits of the crop from the photos. The data processing and analysis module (3) combines photos of crop disease and insect pest conditions and uses computer vision technology, a deep learning model based on a convolutional neural network (CNN), and an image processing algorithm to extract and calculate the number of pests from the photos.

7. The high-standard farmland intelligent monitoring and management system with integrated sky-ground network according to claim 3 is characterized by: The data processing and analysis module (3) combines analog signals, such as voltage and current, collected by the soil temperature and humidity sensor, the soil conductivity sensor, the soil pH value sensor, the air temperature and humidity sensor, the carbon dioxide sensor, the wind speed sensor, and the light intensity sensor, and then converts the analog signals into digital signals through an analog-to-digital converter (ADC), thereby generating corresponding soil temperature and humidity data values, soil conductivity data values, soil pH value data values, air temperature and humidity data values, carbon dioxide data values, wind speed data values, and light intensity data values.

Images