A seedling water and fertilizer management method, system, electronic device and storage medium

By using multi-source information fusion technology and artificial intelligence algorithms, the growth status of seedlings can be monitored in real time, which solves the problem of relying on experience-based decision-making for water and fertilizer management in tidal seedling cultivation, realizes refined water and fertilizer management, and improves seedling growth efficiency.

CN116076211BActive Publication Date: 2024-11-12INTELLIGENT EQUIPMENT RESEARCH CENTER BEIJING ACADEMY OF AGRICULTURE AND FORESTRY SCIENCES
View PDF 11 Cites 0 Cited by

Patent Information

Application Number
CN202211567249.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-12-07
Publication Date
2024-11-12
Estimated Expiration
2042-12-07

Smart Images

  • Figure CN116076211B_ABST
    Figure CN116076211B_ABST
Patent Text Reader

Abstract

The present invention provides a seedling water and fertilizer management method, system, electronic device and storage medium, which belongs to the field of agricultural information technology, including: obtaining leaf association data before a preset growth period, and determining seedling growth prediction information based on the leaf association information; obtaining a leaf image set before the preset growth period, and determining the seedling growth period using the leaf image set and the seedling growth prediction information; obtaining seedling growth environment data, and determining the amount of seedling irrigation and fertilization in combination with the seedling growth environment data and the seedling growth period. The present invention realizes real-time monitoring and feedback of seedling growth by utilizing multi-source information fusion technology and artificial intelligence algorithms, and further formulates a fertilization and irrigation management system based on the current number of leaves and the actual growth environment of the crop, which is conducive to managers further analyzing the crop water consumption law and the factors affecting the emergence rate, and provides the possibility for exploring a better and more efficient control strategy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of agricultural information technology, and in particular to a seedling water and fertilizer management method, system, electronic equipment and storage medium. Background Art

[0002] Tidal seedling cultivation is a process in which plants are grown in containers filled with soilless cultivation substrates. A timer is used to control a water pump to periodically pump fertilizer solution from a storage tank into multiple cultivation troughs or bed boxes. The plants and substrate absorb the remaining fertilizer solution and then naturally flow back into the storage tank for recycling.

[0003] Tidal seedling raising, a new type of water-saving irrigation seedling raising method, solves the problem that water and fertilizer are applied to the leaves in the traditional top spray seedling raising method, but cannot penetrate into the roots, resulting in low water and fertilizer utilization efficiency. It has been widely promoted in recent years due to its fine water and fertilizer control, high degree of standardization of system construction, and efficient management. However, since seedlings are more sensitive to water and fertilizer than mature seedlings, and have higher requirements for water and fertilizer supply, traditional experimental methods cannot form a systematic water and fertilizer irrigation model. Therefore, empirical methods are still used to manage all crops, and the methods are generally based on "empirical decision-making" and "closed-eye decision-making". Summary of the invention

[0004] The present invention provides a seedling water and fertilizer management method, system, electronic device and storage medium, which are used to solve the defects in the prior art that the seedling water and fertilizer management for tidal seedbeds only relies on experience and manual work, resulting in insufficient management and insufficient timely information acquisition.

[0005] In a first aspect, the present invention provides a seedling water and fertilizer management method, comprising:

[0006] Acquire leaf-related data before a preset growth period, and determine seedling growth prediction information based on the leaf-related information;

[0007] Acquire a leaf image set before the preset growth period, and determine the seedling growth period using the leaf image set and the seedling growth prediction information;

[0008] The seedling growth environment data is obtained, and the amount of irrigation and fertilization for the seedlings is determined by combining the seedling growth environment data with the seedling growth period.

[0009] According to a seedling water and fertilizer management method provided by the present invention, leaf-related data before a preset growth period is obtained, and seedling growth prediction information is determined based on the leaf-related information, including:

[0010] Using a preset neural network model to train a seedling image dataset before the preset growth period to obtain the center point of the seedling stem;

[0011] Identify a single leaf image based on the center point of the seedling stem, and determine the center point of the leaf in the single leaf image;

[0012] Determine a leaf star angle in the single leaf image, and determine the leaf associated data according to the leaf star angle;

[0013] The leaf-related data is used to derive the probability of new leaves growing out of leaves in the preset growth period.

[0014] According to a seedling water and fertilizer management method provided by the present invention, determining a leaf star angle in the single leaf image, and determining the leaf associated data by the leaf star angle includes:

[0015] The angle between the preset upper left corner in the single leaf image, the center point of the seedling stem and the center point of the leaf is taken as the leaf star angle;

[0016] A leaf-related data coordinate system is established with the total radiation of the seedling planting environment as the horizontal coordinate and the leaf star angle as the vertical coordinate, and the leaf-related data is obtained from the leaf-related data coordinate system.

[0017] According to a seedling water and fertilizer management method provided by the present invention, a leaf image set before the preset growth period is obtained, and the seedling growth period is determined by using the leaf image set and the seedling growth prediction information, including:

[0018] Constructing a convolutional neural network model based on the leaf image set to obtain seedling growth status information;

[0019] Based on the seedling growth status information, reflection spectrum imaging is acquired at a preset time, and based on the reflection spectrum imaging and the seedling growth prediction information, the seedling growth period is output.

[0020] According to a seedling water and fertilizer management method provided by the present invention, a convolutional neural network model is constructed based on the leaf image set to obtain seedling growth status information, including:

[0021] Performing data enhancement processing on the leaf image set to obtain an enhanced data set;

[0022] Annotating the enhanced data set based on the preset growth period, and dividing the annotated enhanced data set into a training set, a validation set, and a test set according to a preset ratio;

[0023] Using a preset deep learning framework, construct the convolutional neural network model based on the training set, the validation set, and the test set;

[0024] Initializing the parameters of the convolutional neural network model and training the convolutional neural network model using a stochastic gradient descent method to obtain a detection model;

[0025] The leaf image set is input into the detection model to obtain the seedling growth status information.

[0026] According to a seedling water and fertilizer management method provided by the present invention, based on the seedling growth status information, a reflection spectrum imaging is obtained at a preset time, and based on the reflection spectrum imaging and the seedling growth prediction information, the seedling growth period is output, including:

[0027] Based on the seedling growth status information, extracting a leaf image set at the third leaf stage;

[0028] Extracting a leaf bottom spectrum curve, a leaf middle spectrum curve and a leaf tip spectrum curve from the leaf image set at the third leaf stage respectively;

[0029] Based on the leaf bottom spectrum curve, the leaf mid-spectrum curve, the leaf tip spectrum curve and the seedling growth prediction information, the number of overlapping leaves is obtained;

[0030] The fourth leaf stage growth period and the fifth leaf stage growth period are outputted according to the number of overlapping leaves.

[0031] According to a seedling water and fertilizer management method provided by the present invention, the seedling growth environment data is obtained, and the amount of irrigation and fertilization for the seedling is determined by combining the seedling growth environment data and the seedling growth period, including:

[0032] Obtaining the current radiation value and the current temperature value, and accumulating the current temperature value to obtain the hourly accumulated temperature;

[0033] If it is determined that the hourly accumulated temperature reaches the critical hourly accumulated temperature and the current radiation value is greater than the critical radiation value, irrigation and fertilization are started with the seedling irrigation and fertilization amount, otherwise irrigation waiting is performed.

[0034] According to a seedling water and fertilizer management method provided by the present invention, the critical hourly accumulated temperature, the seedling irrigation and fertilization amount, and the concentration of the seedling irrigation and fertilization amount are all linearly related to the number of seedling leaves.

[0035] In a second aspect, the present invention further provides a seedling water and fertilizer management system, comprising:

[0036] A first processing module, configured to obtain leaf-related data before a preset growth period, and determine seedling growth prediction information based on the leaf-related information;

[0037] A second processing module is used to obtain a leaf image set before the preset growth period, and determine the seedling growth period by using the leaf image set and the seedling growth prediction information;

[0038] The third processing module is used to obtain the seedling growth environment data, and determine the amount of irrigation and fertilization for the seedlings in combination with the seedling growth environment data and the seedling growth period.

[0039] In a third aspect, the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a seedling water and fertilizer management method as described above is implemented.

[0040] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described seedling water and fertilizer management methods.

[0041] In a fifth aspect, the present invention further provides a computer program product, comprising a computer program, which, when executed by a processor, implements any of the above-mentioned seedling water and fertilizer management methods.

[0042] The seedling water and fertilizer management method, system, electronic device and storage medium provided by the present invention realize real-time monitoring and feedback of seedling growth by utilizing multi-source information fusion technology and artificial intelligence algorithm, and further formulate a fertilization and irrigation management system according to the current number of leaves and the actual growth environment of the crop, which is conducive to managers to further analyze the water consumption patterns of crops and the factors affecting the emergence rate, and provides the possibility for discovering better and more efficient control strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 It is a schematic diagram of the process of the seedling water and fertilizer management method provided by the present invention;

[0045] Figure 2 It is a schematic diagram of a seedling recognition convolutional neural network model provided by the present invention;

[0046] Figure 3 It is a structural schematic diagram of the seedling water and fertilizer management system provided by the present invention;

[0047] Figure 4 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] Figure 1 Schematic diagram of the process of the seedling water and fertilizer management method provided by the present invention, such as Figure 1 As shown, including:

[0050] Step 100: Acquire leaf-related data before a preset growth period, and determine seedling growth prediction information based on the leaf-related information;

[0051] Step 200: Acquire a leaf image set before the preset growth period, and determine the seedling growth period using the leaf image set and the seedling growth prediction information;

[0052] Step 300: Acquire seedling growth environment data, and determine the amount of irrigation and fertilization for the seedlings in combination with the seedling growth environment data and the seedling growth period.

[0053] In view of the current status of seedling management in tidal seedbeds, the present invention uses information technology and computer technology to provide a new crop growth feedback system, and adjusts the supply of water, fertilizer and gas in real time according to the feedback information, providing a new idea for the visual intelligent management and control of seedlings.

[0054] The system includes six parts: tidal seedbed assembly, three-degree-of-freedom linear unit gantry slide assembly, sensor monitoring assembly, water and fertilizer supply assembly and human-computer interaction input and output assembly. Among them, the tidal seedbed assembly includes a tidal seedbed and a seedling tray; the three-degree-of-freedom linear unit gantry slide assembly includes a servo motor, a photoelectric sensor, a guide rail, a slider, a belt, a supporting mechanism, etc.; the sensor monitoring assembly includes a high-definition camera, an online spectrometer, an air temperature and humidity sensor, a total light radiation sensor, a liquid level meter, a flow meter, etc.; the water and fertilizer supply assembly includes a liquid storage unit, a mother liquid unit, a liquid return unit, a water and fertilizer integrated head, a liquid supply and return pipeline, a liquid supply and return valve, etc. Preferably, a diaphragm solenoid valve is selected for liquid supply, and a three-position two-way pressureless electric valve is selected for liquid return; the human-computer interaction input and output assembly includes a communication unit, an upper display control unit and a remote service center.

[0055] Based on the above system, the present invention also provides a seedling leaf tracking method, which utilizes multi-source information fusion technology and artificial intelligence algorithm to realize real-time monitoring and feedback of seedling growth based on multi-parameter coupling, and then gives a specific number of leaves, and further formulates a fertilization and irrigation management system based on the current number of leaves and the actual growth environment of the crop.

[0056] In addition, the system stores all parameters related to seedling cultivation, including environmental parameters, irrigation parameters, setting parameters and control parameters, and provides a data download interface, which can further analyze the water consumption patterns of crops and factors affecting the emergence rate, providing the possibility for discovering better and more efficient control strategies.

[0057] Specifically, firstly, the leaf growth, leaf quantity and radiation amount before the seedlings reach the three-leaf stage are obtained, leaf-related data are established, and the growth information of the seedlings after the three-leaf stage is predicted based on the leaf-related data;

[0058] Then, based on the convolutional neural network model and reflectance spectrum, the growth period of the seedlings was determined according to the leaf image set before the three-leaf stage and the seedling growth prediction information;

[0059] Finally, according to the seedling growth period and the current growth environment data of the seedlings, corresponding water and fertilizer irrigation measures are implemented.

[0060] The present invention realizes real-time monitoring and feedback of seedling growth by utilizing multi-source information fusion technology and artificial intelligence algorithm, and further formulates a fertilization and irrigation management system according to the current number of leaves and the actual growth environment of the crop, which is conducive to managers to further analyze the water consumption patterns of crops and the factors affecting the emergence rate, and provides the possibility for discovering better and more efficient control strategies.

[0061] Based on the above embodiment, step 100 includes:

[0062] Using a preset neural network model to train a seedling image dataset before the preset growth period to obtain the center point of the seedling stem;

[0063] Identify a single leaf image based on the center point of the seedling stem, and determine the center point of the leaf in the single leaf image;

[0064] Determine a leaf star angle in the single leaf image, and determine the leaf associated data according to the leaf star angle;

[0065] The leaf-related data is used to derive the probability of new leaves growing out of leaves in the preset growth period.

[0066] Wherein, determining the leaf star angle in the single leaf image, and determining the leaf associated data by the leaf star angle, comprises:

[0067] The angle between the preset upper left corner in the single leaf image, the center point of the seedling stem and the center point of the leaf is taken as the leaf star angle;

[0068] A leaf-related data coordinate system is established with the total radiation of the seedling planting environment as the horizontal coordinate and the leaf star angle as the vertical coordinate, and the leaf-related data is obtained from the leaf-related data coordinate system.

[0069] Specifically, firstly, commonly used neural network models, such as commonly used deep neural networks (DNN), long short-term memory networks (LSTM), and residual networks (ResNet), are used to train seedling image datasets before a preset growth period to obtain the center point of the seedling stem, denoted as M;

[0070] Then, a single leaf is identified based on the center point M, the center point N of the leaf is determined, and a unique ID is assigned to the leaf based on the center point N of the leaf;

[0071] Define the angle between the upper left corner of a single leaf image, the center point M of the seedling stem and the center point N of the leaf as the leaf star angle, take the total radiation received by the planting environment as the horizontal coordinate, take the leaf star angle of each ID as the vertical coordinate, establish the leaf associated data coordinate system, and obtain the leaf associated data from the leaf associated data coordinate system;

[0072] The leaf association data before the three-leaf stage were used to infer the growth of seedlings after the three-leaf stage, and the probability of new leaves emerging was obtained, expressed as a percentage.

[0073] It should be noted that the data set collected here must be shot at fixed points to ensure the consistency of the input image.

[0074] The present invention adopts the method of locating the center point to track and predict the leaves in real time, and integrates the reflected spectrum information at the same time, which can solve the bottleneck problem of the application of machine vision in agriculture and the inability to identify overlapping leaves.

[0075] Based on the above embodiment, step 200 includes:

[0076] Constructing a convolutional neural network model based on the leaf image set to obtain seedling growth status information;

[0077] Based on the seedling growth status information, reflection spectrum imaging is acquired at a preset time, and based on the reflection spectrum imaging and the seedling growth prediction information, the seedling growth period is output.

[0078] Wherein, constructing a convolutional neural network model based on the leaf image set to obtain seedling growth status information includes:

[0079] Performing data enhancement processing on the leaf image set to obtain an enhanced data set;

[0080] Annotating the enhanced data set based on the preset growth period, and dividing the annotated enhanced data set into a training set, a validation set, and a test set according to a preset ratio;

[0081] Using a preset deep learning framework, construct the convolutional neural network model based on the training set, the validation set, and the test set;

[0082] Initializing the parameters of the convolutional neural network model and training the convolutional neural network model using a stochastic gradient descent method to obtain a detection model;

[0083] The leaf image set is input into the detection model to obtain the seedling growth status information.

[0084] Wherein, based on the seedling growth status information, obtaining reflection spectrum imaging at a preset time, and outputting the seedling growth period based on the reflection spectrum imaging and the seedling growth prediction information, including:

[0085] Based on the seedling growth status information, extracting a leaf image set at the third leaf stage;

[0086] Extracting a leaf bottom spectrum curve, a leaf middle spectrum curve and a leaf tip spectrum curve from the leaf image set at the third leaf stage respectively;

[0087] Based on the leaf bottom spectrum curve, the leaf mid-spectrum curve, the leaf tip spectrum curve and the seedling growth prediction information, the number of overlapping leaves is obtained;

[0088] The fourth leaf stage growth period and the fifth leaf stage growth period are outputted according to the number of overlapping leaves.

[0089] Specifically, the present invention also adopts a real-time monitoring feedback method for seedling growth based on multi-parameter coupling, uses a convolutional neural network model and reflectance spectrum, and couples decisions on different growth stages of seedlings to achieve adaptive control of seedling water and fertilizer.

[0090] First, if Figure 2 As shown in the figure, a large number of image data sets of seedlings at different growth stages were selected to build a convolutional neural network model, which was trained to recognize the real-time growth status of crops, represented by the number of leaves L, which was divided into one-leaf stage, two-leaf stage and three-leaf stage.

[0091] Here, an image dataset of seedlings at different growth stages is obtained, and data enhancement is performed on the image dataset, specifically including:

[0092] The camera was placed at a 90° angle to the horizontal direction of the crop row and about 1m away from the seedbed. Images of seedlings at different stages were collected to construct an image dataset. The image dataset was enhanced by methods such as 90-degree random rotation, random horizontal flipping, random vertical flipping, and random pseudo-reflection transformation.

[0093] After obtaining the image data sets of seedlings at different growth stages, the image data needs to be normalized and then the image data sets are enhanced; the enhanced data sets are annotated using annotation software, and the annotated data sets are randomly divided into training set, validation set, and test set in a ratio of 6:2:2; on the TensorFlow deep learning framework, a convolutional neural network model is built to initialize the model parameters, and the training set is input into the detection model for training based on the stochastic gradient descent method; after the training is completed, the image to be identified is input into the detection model, and the growth stage of the seedling is output.

[0094] When building a convolutional neural network model on the TensorFlow deep learning framework, the model is mainly composed of two main blocks, namely the feature extraction block and the classification block. In the model, the feature extraction block extracts features from the input image through the convolution layer, and the classification block determines the category.

[0095] In addition, the convolutional neural network model parameters are initialized, and the training set is input into the detection model for training based on the stochastic gradient descent method, specifically including: using Xavier to initialize the weights and using zero to initialize the bias; using the stochastic gradient descent method to train the parameters, and when the error value is equal to or less than the expected value, the training is terminated and the weights are updated.

[0096] After the training is completed, the image to be identified is input into the detection model, and the growth stage of the seedling is output. The image data obtained by the camera is input into the detection model to obtain the output prediction value to determine the growth stage of the seedling.

[0097] The present invention obtains the judgment result of the seedling growth period through convolutional neural network model training, which is beneficial to subsequent irrigation and fertilization according to the seedling growth period.

[0098] Based on the above embodiment, step 300 includes:

[0099] Obtaining the current radiation value and the current temperature value, and accumulating the current temperature value to obtain the hourly accumulated temperature;

[0100] If it is determined that the hourly accumulated temperature reaches the critical hourly accumulated temperature and the current radiation value is greater than the critical radiation value, irrigation and fertilization are started with the seedling irrigation and fertilization amount, otherwise irrigation waiting is performed.

[0101] Among them, the critical hourly accumulated temperature, the seedling irrigation and fertilization amount and the concentration of the seedling irrigation and fertilization amount are all linearly related to the number of seedling leaves.

[0102] Specifically, based on the aforementioned embodiments, the present invention also provides an irrigation and fertilization management method based on multi-parameter coupling, that is, integrating the comprehensive data of a high-definition camera, an online spectrometer, an air temperature and humidity sensor, and a total light radiation sensor to perform irrigation management.

[0103] The current environment of the crop is collected every 1 minute, and the current radiation value is recorded as Wi and the current temperature value is Ti. The temperature is accumulated to obtain the hourly accumulated temperature When the hourly accumulated temperature reaches the critical hourly accumulated temperature Ttl, the current radiation value is judged. If the current radiation value is greater than the critical radiation, irrigation is started; otherwise, it is waited; further, the critical hourly accumulated temperature Ttl is a linear function related to the number of leaves L, which is represented by Ttl=f1(L). This solves the problem of excessive growth caused by the inability to avoid cloudy days during scheduled irrigation.

[0104] Furthermore, the irrigation amount ET is also a positively correlated function of the leaf number L, which is represented by ET=f2(L); further, the irrigation solution concentration EC increases with the increase of the leaf number L to ensure growth, which is represented by EC=1200+100L.

[0105] The present invention completely links the three elements of irrigation and fertilization: irrigation start time, irrigation amount and irrigation liquid concentration with the number of leaves, which can not only fully save water and fertilizer, but also ensure that each irrigation is carried out at a suitable temperature, which is beneficial to crop absorption and growth, is not prone to excessive growth, and effectively improves the quality of seedlings.

[0106] The seedling water and fertilizer management system provided by the present invention is described below. The seedling water and fertilizer management system described below and the seedling water and fertilizer management method described above can be referenced to each other.

[0107] Figure 3 It is a structural schematic diagram of the seedling water and fertilizer management system provided by the present invention, such as Figure 3 As shown, it includes: a first processing module 31, a second processing module 32 and a third processing module 33, wherein:

[0108] The first processing module 31 is used to obtain leaf-related data before a preset growth period, and determine seedling growth prediction information based on the leaf-related information; the second processing module 32 is used to obtain a leaf image set before the preset growth period, and determine the seedling growth period using the leaf image set and the seedling growth prediction information; the third processing module 33 is used to obtain seedling growth environment data, and determine the amount of irrigation and fertilization for the seedlings in combination with the seedling growth environment data and the seedling growth period.

[0109] The present invention realizes real-time monitoring and feedback of seedling growth by utilizing multi-source information fusion technology and artificial intelligence algorithm, and further formulates a fertilization and irrigation management system according to the current number of leaves and the actual growth environment of the crop, which is conducive to managers to further analyze the water consumption patterns of crops and the factors affecting the emergence rate, and provides the possibility for discovering better and more efficient control strategies.

[0110] Figure 4 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 4 As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430 and a communication bus 440, wherein the processor 410, the communication interface 420 and the memory 430 complete mutual communication through the communication bus 440. The processor 410 may call the logic instructions in the memory 430 to execute the seedling water and fertilizer management method, which includes: obtaining leaf association data before a preset growth period, and determining seedling growth prediction information based on the leaf association information; obtaining a leaf image set before the preset growth period, and determining the seedling growth period using the leaf image set and the seedling growth prediction information; obtaining seedling growth environment data, and determining the amount of irrigation and fertilization for the seedlings in combination with the seedling growth environment data and the seedling growth period.

[0111] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0112] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the seedling water and fertilizer management method provided by the above methods, which includes: obtaining leaf association data before a preset growth period, and determining seedling growth prediction information based on the leaf association information; obtaining a leaf image set before the preset growth period, and determining the seedling growth period using the leaf image set and the seedling growth prediction information; obtaining seedling growth environment data, and determining the amount of irrigation and fertilization for the seedlings in combination with the seedling growth environment data and the seedling growth period.

[0113] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it is implemented to execute the seedling water and fertilizer management method provided by the above-mentioned methods. The method includes: obtaining leaf association data before a preset growth period, and determining seedling growth prediction information based on the leaf association information; obtaining a leaf image set before the preset growth period, and determining the seedling growth period using the leaf image set and the seedling growth prediction information; obtaining seedling growth environment data, and determining the amount of irrigation and fertilization for the seedlings in combination with the seedling growth environment data and the seedling growth period.

[0114] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0115] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A seedling water and fertilizer management method, characterized in that: include: Acquiring leaf-related data before a preset growth period, and determining seedling growth prediction information based on the leaf-related data; Acquire a leaf image set before the preset growth period, and determine the seedling growth period using the leaf image set and the seedling growth prediction information; Acquire seedling growth environment data, and determine the amount of irrigation and fertilization for the seedlings by combining the seedling growth environment data and the seedling growth period; Acquiring a leaf image set before the preset growth period, and determining the seedling growth period using the leaf image set and the seedling growth prediction information, comprising: Constructing a convolutional neural network model based on the leaf image set to obtain seedling growth status information; Based on the seedling growth status information, obtaining reflection spectrum imaging at a preset time, and outputting the seedling growth period based on the reflection spectrum imaging and the seedling growth prediction information; Acquiring leaf-related data before a preset growth period, and determining seedling growth prediction information based on the leaf-related data, including: Using a preset neural network model to train a seedling image dataset before the preset growth period to obtain the center point of the seedling stem; Identify a single leaf image based on the center point of the seedling stem, and determine the center point of the leaf in the single leaf image; Determine a leaf star angle in the single leaf image, and determine the leaf associated data according to the leaf star angle; The leaf-related data is used to derive the probability of new leaves growing out of leaves in the preset growth period; Determining a leaf star angle in the single leaf image, and determining the leaf associated data from the leaf star angle, comprising: The angle between the preset upper left corner in the single leaf image, the center point of the seedling stem and the center point of the leaf is taken as the leaf star angle; A leaf-related data coordinate system is established with the total radiation of the seedling planting environment as the horizontal coordinate and the leaf star angle as the vertical coordinate, and the leaf-related data is obtained from the leaf-related data coordinate system.

2. The seedling water and fertilizer management method according to claim 1, characterized in that: A convolutional neural network model is constructed based on the leaf image set to obtain seedling growth status information, including: Performing data enhancement processing on the leaf image set to obtain an enhanced data set; Annotating the enhanced data set based on the preset growth period, and dividing the annotated enhanced data set into a training set, a validation set, and a test set according to a preset ratio; Using a preset deep learning framework, construct the convolutional neural network model based on the training set, the validation set, and the test set; Initializing the parameters of the convolutional neural network model and training the convolutional neural network model using a stochastic gradient descent method to obtain a detection model; The leaf image set is input into the detection model to obtain the seedling growth status information.

3. The seedling water and fertilizer management method according to claim 1, characterized in that: Based on the seedling growth state information, obtaining reflection spectrum imaging at a preset time, and outputting the seedling growth period based on the reflection spectrum imaging and the seedling growth prediction information, including: Based on the seedling growth status information, extracting a leaf image set at the third leaf stage; Extracting a leaf bottom spectrum curve, a leaf middle spectrum curve and a leaf tip spectrum curve from the leaf image set at the third leaf stage respectively; Based on the leaf bottom spectrum curve, the leaf mid-spectrum curve, the leaf tip spectrum curve and the seedling growth prediction information, the number of overlapping leaves is obtained; The fourth leaf stage growth period and the fifth leaf stage growth period are outputted according to the number of overlapping leaves.

4. The seedling water and fertilizer management method according to claim 1, characterized in that: Acquiring seedling growth environment data, combining the seedling growth environment data with the seedling growth period, and determining the amount of irrigation and fertilization for the seedlings, including: Obtaining the current radiation value and the current temperature value, and accumulating the current temperature value to obtain the hourly accumulated temperature; If it is determined that the hourly accumulated temperature reaches the critical hourly accumulated temperature and the current radiation value is greater than the critical radiation value, irrigation and fertilization are started with the seedling irrigation and fertilization amount, otherwise irrigation waiting is performed.

5. The seedling water and fertilizer management method according to claim 4, characterized in that: The critical hourly accumulated temperature, the seedling irrigation and fertilization amount, and the concentration of the seedling irrigation and fertilization amount are all linearly related to the number of seedling leaves.

6. A seedling water and fertilizer management system, characterized in that: include: A first processing module, configured to obtain leaf-related data before a preset growth period, and determine seedling growth prediction information based on the leaf-related data; A second processing module is used to obtain a leaf image set before the preset growth period, and determine the seedling growth period by using the leaf image set and the seedling growth prediction information; The third processing module is used to obtain the seedling growth environment data, and determine the amount of irrigation and fertilization for the seedlings in combination with the seedling growth environment data and the seedling growth period; Acquiring a leaf image set before the preset growth period, and determining the seedling growth period using the leaf image set and the seedling growth prediction information, comprising: Constructing a convolutional neural network model based on the leaf image set to obtain seedling growth status information; Based on the seedling growth status information, obtaining reflection spectrum imaging at a preset time, and outputting the seedling growth period based on the reflection spectrum imaging and the seedling growth prediction information; Acquiring leaf-related data before a preset growth period, and determining seedling growth prediction information based on the leaf-related data, including: Using a preset neural network model to train a seedling image dataset before the preset growth period to obtain the center point of the seedling stem; Identify a single leaf image based on the center point of the seedling stem, and determine the center point of the leaf in the single leaf image; Determine a leaf star angle in the single leaf image, and determine the leaf associated data according to the leaf star angle; The leaf-related data is used to derive the probability of new leaves growing out of leaves in the preset growth period; Determining a leaf star angle in the single leaf image, and determining the leaf associated data from the leaf star angle, comprising: The angle between the preset upper left corner in the single leaf image, the center point of the seedling stem and the center point of the leaf is taken as the leaf star angle; A leaf-related data coordinate system is established with the total radiation of the seedling planting environment as the horizontal coordinate and the leaf star angle as the vertical coordinate, and the leaf-related data is obtained from the leaf-related data coordinate system.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the seedling water and fertilizer management method as described in any one of claims 1 to 5 is implemented.

Citation Information

Patent Citations

  • Method and device for early detecting crop pests based on multisensor information fusion

    CN102084794A

  • Automatic detection method for time emergence of seedling of corns

    CN102592118A

  • Sparse representation-based semi-supervised hyperspectral image classification method

    CN106127225A

  • Device and method of intelligent light supplementation of flammulina velutipes in different growth periods based on convolutional neural network

    CN107223454A

  • Irrigation and fertilization device, system and method

    CN109673478A