Crop irrigation regulation control method and system based on phenotypic characteristics

Abstract

The application discloses a crop irrigation regulation control method and system based on phenotypic characteristics, and relates to the field of crop irrigation regulation control. The method comprises the following steps: acquiring information data of different crop partitions in a target farmland; for any crop partition, the following steps are performed: based on a crop phenotypic model, determining phenotypic characteristic information according to crop image data; the crop phenotypic model is obtained by adjusting and training based on an improved YOLOv8 network, according to historical crop image data of known phenotypic characteristic information, with the minimum loss function as the target; the improved YOLOv8 network comprises a C2fCA network module with an added attention mechanism; based on a phenotypic-water requirement mapping model, determining water requirement information according to the phenotypic characteristic information and field environment information; and performing corresponding irrigation control according to the water requirement information, so as to realize irrigation regulation of crops in the target farmland. The application aims to automatically realize individualized crop irrigation regulation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of crop irrigation regulation control, in particular to a crop irrigation regulation control method and system based on phenotypic characteristics. BACKGROUND

[0002] Water resource shortage and efficient water use are very serious problems that need to be considered in the irrigation of field crops. At the same time, the aging and shortage of labor force is increasingly prominent, and it is urgent to transform the traditional crop irrigation method through new generation information technology to empower traditional agriculture with digitalization, to ensure food security and water resource security, and to improve production efficiency.

[0003] At present, the irrigation modes of field corn include flooding irrigation, border irrigation, surface drip irrigation and film drip irrigation, and the control method is mainly manual control. Some demonstration areas have realized automatic control, but the irrigation head function is limited to simple functions such as timing, and the irrigation pipe control method mainly adopts the method of main pipe plus parallel branch, by installing a manual ball valve at the connection between each branch and the main pipe to realize zoned management.

[0004] The above irrigation methods do not consider the growth state of corn itself, ignoring the coupling relationship between phenotypic characteristics and water requirement state. In recent years, plant phenotyping methods based on image recognition have been gradually applied to agricultural production, but there are still the following problems:

[0005] Existing phenotypic detection is mainly based on monocular vision, lacks depth information, and has limited precision; existing models do not integrate phenotypic recognition and intelligent control, and cannot form a closed-loop autonomous decision-making system; irrigation control mainly relies on artificial judgment or static models, and it is difficult to realize individualized crop irrigation regulation. SUMMARY

[0006] The purpose of the present application is to provide a crop irrigation regulation control method and system based on phenotypic characteristics, which can automatically realize individualized crop irrigation regulation.

[0007] To achieve the above purpose, the present application provides the following solutions:

[0008] In a first aspect, the present application provides a crop irrigation regulation control method based on phenotypic characteristics, comprising:

[0009] obtaining information data of different crop zones in a target farmland; the information data includes crop image data and field environment information;

[0010] For any crop zone, the following are performed:

[0011] determine phenotype feature information according to the crop image data based on a crop phenotype model; the crop phenotype model is obtained by adjusting and training a YOLOv8 network based on historical crop image data with known phenotype feature information, with the objective of minimizing a loss function; the improved YOLOv8 network includes a C2fCA network module with an added attention mechanism;

[0012] determine water requirement information according to the phenotype feature information and the field environment information based on a phenotype-water requirement mapping model; the water requirement information includes target humidity;

[0013] perform corresponding irrigation control according to the water requirement information to achieve irrigation regulation of crops in the target farmland.

[0014] In an embodiment, after obtaining information data of different crop partitions in the target farmland, the method further includes:

[0015] perform horizontal flipping processing on the crop image data to obtain flipped crop image data;

[0016] perform random adjustment of saturation, brightness, and contrast on the flipped crop image data to obtain adjusted crop image data;

[0017] perform sharpening and enhancement processing on the adjusted crop image data.

[0018] In an embodiment, a mathematical expression of the phenotype-water requirement mapping model is:

[0019] ;

[0020] wherein, is the target humidity; is leaf length; is leaf angle; is plant height; is air temperature; is air humidity; 、 、 、 and are fitting coefficients.

[0021] In an embodiment, performing corresponding irrigation control according to the water requirement information to achieve irrigation regulation of crops in the target farmland specifically includes:

[0022] determining irrigation water volume according to the water requirement information;

[0023] determining irrigation control decisions for corresponding crop partitions according to the irrigation water volume to achieve irrigation regulation of crops in the target farmland; the irrigation control decisions include irrigation duration.

[0024] In an embodiment, the expression of the irrigation water amount is:

[0025] ;

[0026] wherein, is the irrigation water amount; is the root coverage area per plant; is the effective root layer depth; is the target humidity; is the current humidity.

[0027] In an embodiment, the method for determining the crop phenotype model specifically comprises:

[0028] obtaining a training data set; the training data set includes historical crop image data with known phenotype feature information;

[0029] dividing the training data set into a training set, a validation set, and a test set according to a set proportion;

[0030] constructing an improved YOLOv8 network;

[0031] inputting the training set into the improved YOLOv8 network, adjusting and training the network parameters to minimize the loss function, and obtaining a trained network; the network parameters include: confidence;

[0032] inputting the validation set into the trained network for validation, and obtaining a validated network;

[0033] inputting the test set into the validated network for testing, and obtaining a tested network;

[0034] determining the tested network as the crop phenotype model.

[0035] In an embodiment, the crop irrigation adjustment control method based on phenotype characteristics further comprises:

[0036] obtaining information data corresponding to the crop irrigation adjustment in the target farmland, and feeding back to the upper computer for crop irrigation coordination.

[0037] In a second aspect, the present application provides a crop irrigation adjustment control system based on phenotype characteristics, which is realized by using a crop irrigation adjustment control method based on phenotype characteristics; the crop irrigation adjustment control system based on phenotype characteristics comprises an image acquisition module, an image recognition and processing module, an environment information acquisition module, an intelligent irrigation decision module, and an electromagnetic valve execution module.

[0038] The image recognition and processing module is connected with the image acquisition module; the image recognition and processing module and the environment information acquisition module are both connected with the intelligent irrigation decision module; the electromagnetic valve execution module is connected with the intelligent irrigation decision module;

[0039] The image acquisition module is configured to acquire crop image data of different crop sub-zones in the target farmland.

[0040] The environment information acquisition module is configured to acquire field environment information of different crop sub-zones in the target farmland.

[0041] The image recognition and processing module is configured to determine, for any crop sub-zone, phenotypic feature information based on a crop phenotypic model and according to the crop image data.

[0042] The intelligent irrigation decision module is configured to determine water requirement information based on a phenotypic-water requirement mapping model and according to the phenotypic feature information and the field environment information.

[0043] The electromagnetic valve execution module is configured to perform corresponding irrigation control according to the water requirement information, so as to realize irrigation regulation of crops in the target farmland.

[0044] In an embodiment, the image acquisition module adopts a binocular camera.

[0045] In an embodiment, the crop irrigation regulation control system based on phenotypic features further comprises a local storage module and a screen display module.

[0046] The local storage module and the screen display module are both connected with the binocular camera.

[0047] The local storage module is configured to store crop image data acquired by the binocular camera.

[0048] The screen display module is configured to display and view acquisition situation information of the binocular camera.

[0049] According to the specific embodiments provided in the present application, the following technical effects are disclosed:

[0050] The application provides a crop irrigation adjustment control method and system based on phenotypic characteristics, information data of different crop partitions in a target farmland is acquired; for any crop partition, the following are performed: based on a crop phenotypic model, phenotypic characteristic information is determined according to crop image data; the crop phenotypic model is obtained by adjusting and training based on historical crop image data of known phenotypic characteristic information, with the objective of minimizing a loss function; the improved YOLOv8 network includes a C2fCA network module with an added attention mechanism; based on a phenotypic-water requirement mapping model, water requirement information is determined according to the phenotypic characteristic information and field environment information; the application combines the crop phenotypic model and the phenotypic-water requirement mapping model, performs corresponding irrigation control according to the water requirement information, so as to realize irrigation adjustment of crops in the target farmland, thereby automatically realizing individualized crop irrigation adjustment. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0052] Figure 1 Flowchart of the crop irrigation adjustment control method based on phenotypic characteristics;

[0053] Figure 2 Structure diagram of the crop irrigation adjustment control system based on phenotypic characteristics;

[0054] Figure 3 Working flowchart of the crop irrigation adjustment control system. DETAILED DESCRIPTION

[0055] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0056] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0057] In one exemplary embodiment, as shown in Figure 1 A crop irrigation adjustment control method based on phenotypic characteristics is provided, comprising:

[0058] Step 100: Obtain information data of different crop partitions in the target farmland. The information data includes crop image data and field environment information.

[0059] For any crop partition, the following is performed:

[0060] Step 200: Determine the phenotypic feature information according to the crop image data based on a crop phenotypic model. The crop phenotypic model is obtained by adjusting and training a known historical crop image data with known phenotypic feature information based on an improved YOLOv8 network with the objective of minimizing the loss function; the improved YOLOv8 network includes a C2fCA network module with an added attention mechanism.

[0061] The method for determining the crop phenotypic model specifically includes:

[0062] Obtain a training data set; the training data set includes historical crop image data with known phenotypic feature information.

[0063] Divide the training data set into a training set, a validation set, and a test set according to a set proportion.

[0064] Construct an improved YOLOv8 network; input the training set into the improved YOLOv8 network, and adjust and train the network parameters with the objective of minimizing the loss function to obtain a trained network; the network parameters include a confidence level.

[0065] Input the validation set into the trained network for validation to obtain a validated network; input the test set into the validated network for testing to obtain a tested network; determine the tested network as the crop phenotypic model.

[0066] Step 300: Determine the water requirement information based on the phenotypic feature information and the field environment information based on a phenotypic-water requirement mapping model. The water requirement information includes a target humidity.

[0067] The mathematical expression corresponding to the phenotypic-water requirement mapping model is:

[0068] .

[0069] wherein, is the target humidity; is the leaf length; is the leaf angle; is the plant height; is the air temperature; is the air humidity; , , , and are fitting coefficients.

[0070] Step 400: according to the water demand information, corresponding irrigation control is performed to realize irrigation regulation of crops in the target farmland.

[0071] As an optional implementation, according to the water demand information, corresponding irrigation control is performed to realize irrigation regulation of crops in the target farmland, specifically including:

[0072] According to the water demand information, the irrigation water quantity is determined; according to the irrigation water quantity, the irrigation control decision for the corresponding crop subarea is determined to realize irrigation regulation of crops in the target farmland; the irrigation control decision includes the irrigation duration.

[0073] The expression of the irrigation water quantity is:

[0074] .

[0075] Wherein, is the irrigation water quantity; is the root coverage area per plant; is the effective root layer depth; is the target humidity; is the current humidity.

[0076] In an embodiment, after obtaining the information data of different crop subareas in the target farmland, it further includes:

[0077] The crop image data is subjected to horizontal flip processing to obtain flipped crop image data; the flipped crop image data is subjected to random adjustment of saturation, brightness and contrast to obtain adjusted crop image data; the adjusted crop image data is subjected to sharpening and enhancement processing.

[0078] As an optional implementation, the crop irrigation regulation control method based on phenotypic characteristics further includes:

[0079] Obtaining corresponding information data of crops after irrigation regulation in the target farmland, and feeding back to the upper computer to perform crop irrigation coordination.

[0080] In an exemplary embodiment, a crop irrigation regulation control system based on phenotypic characteristics is provided, which is implemented by using a crop irrigation regulation control method based on phenotypic characteristics. As shown in Figure 2 The crop irrigation regulation control system based on phenotypic characteristics includes an image acquisition module, an image recognition and processing module, an environmental information acquisition module, an intelligent irrigation decision module and an electromagnetic valve execution module.

[0081] The image recognition and processing module is connected with the image acquisition module; the image recognition and processing module and the environmental information acquisition module are both connected with the intelligent irrigation decision module; the electromagnetic valve execution module is connected with the intelligent irrigation decision module.

[0082] The image acquisition module is configured to acquire crop image data of different crop sub-zones in the target farmland. The image acquisition module employs a binocular camera.

[0083] The environmental information acquisition module is configured to acquire field environmental information of different crop sub-zones in the target farmland. The image recognition and processing module is configured to determine, for any crop sub-zone, phenotypic characteristic information based on a crop phenotypic model and according to the crop image data.

[0084] The intelligent irrigation decision module is configured to determine water requirement information based on a phenotypic-water requirement mapping model and according to the phenotypic characteristic information and the field environmental information.

[0085] The electromagnetic valve execution module is configured to perform corresponding irrigation control according to the water requirement information, so as to achieve irrigation regulation of crops in the target farmland.

[0086] In an embodiment, the crop irrigation regulation control system based on phenotypic characteristics further comprises a local storage module and a screen display module.

[0087] The local storage module and the screen display module are both connected to the binocular camera. The local storage module is configured to store crop image data acquired by the binocular camera. The screen display module is configured to display and view acquisition situation information of the binocular camera.

[0088] The present application identifies phenotypic characteristics such as corn stalk width, leaf area, and inclination angle based on a binocular camera combined with an improved YOLOv8 model (network), and constructs a precise phenotypic model (crop phenotypic model).

[0089] The present application uses an improved YOLOv8 network structure optimized by a CA attention mechanism + EIoU loss, to improve target recognition accuracy and training convergence speed.

[0090] The present application constructs a phenotypic-water requirement mapping model, fuses environmental parameters (temperature and humidity, illumination) and crop phenotypes, and realizes precise water requirement identification of crops.

[0091] Based on the identification result, an electromagnetic valve control signal is automatically generated to realize intelligent irrigation regulation of sub-zones and crops. In addition, the system mentioned in the present application has edge computing capability, and can realize real-time identification and decision-making on the spot in the farmland, thereby improving irrigation timeliness and reliability.

[0092] The image acquisition module employs a binocular camera to acquire crop image data. The image recognition and processing module is deployed on an edge computing terminal and is configured to identify phenotypic characteristics of crops. The environmental information acquisition module acquires environmental parameters such as soil humidity and air temperature and humidity. The intelligent irrigation decision module generates an irrigation strategy by comprehensively considering phenotypic characteristics and environmental parameters. The electromagnetic valve execution module controls electromagnetic valves to accurately perform irrigation according to decision-making instructions.

[0093] The method mentioned in the present application, taking corn intelligent irrigation control as an example, combines Figure 3 the working flow diagram of the crop irrigation regulation and control system in the present application, and the method comprises the following steps:

[0094] Step 1: image acquisition (performed by an image acquisition module).

[0095] A binocular camera acquisition module capable of automatically acquiring images of corn plants is built, which comprises a binocular camera module, the binocular images acquired by the binocular camera can provide depth images for corn stalk width calculation and provide more accurate data information for corn stalk width calculation; a camera shooting module, which can realize the start and stop of the camera and the interval setting of the timing shooting time; a local storage module, which can store the acquired binocular images in the module for easy downloading and use; a screen display module, which can operate and control the camera on the screen module, making it easy to check the working condition of the binocular camera acquisition module on site and troubleshoot problems at any time; an encapsulated housing module, which makes it easy to place the binocular camera image module acquisition module outdoors.

[0096] Step 2: image preprocessing and target recognition (performed by an image recognition and processing module).

[0097] The corn plant dataset is horizontally flipped to simulate images taken at various angles.

[0098] The image is randomly adjusted in terms of saturation, brightness and contrast to simulate images taken under different weather conditions.

[0099] Different sharpening processing is performed on the image to simulate pictures taken under harsh conditions, thereby better improving the robustness of the model.

[0100] The enhanced dataset is uniformly named as IMGxxx format, and is divided into training set, validation set and test set according to the ratio of 8:1:1.

[0101] The corn stalk dataset is labeled using the open source tool LabelImg, and the Annotations and JPEGImages folders are used to store the label file and dataset image file. A predefined-class text document is created, the label category stem is input, the LabelImg software is opened, the minimum recognition rectangle box is added to the target, and the LabelImg software will generate the corresponding txt label file to adapt to the YOLOv8 detection model.

[0102] The algorithm of YOLOv8 is improved, including: fusing the CA attention mechanism module with the C2f module in the original Backbone, replacing the Bottleneck in the original C2f module with the CA attention module, redesigning it as a C2fCA network module, and repeatedly adding attention mechanisms. The CA attention mechanism can make the attention block capture the long-distance relationship in one direction while preserving the spatial information in the other direction, so that the position information can be saved in the generated attention map to focus on the area of interest and help the network better and more accurately locate the target. At the same time, the default loss function CIoU in the original box-IoU is replaced by EIoU, which splits the aspect ratio loss term into the difference between the predicted width and height and the minimum bounding box width and height, accelerating the convergence of the prediction box and improving the regression accuracy of the prediction box.

[0103] After training the improved YOLOv8 network based on the pretreated corn stalk data set, the improved YOLOv8 network is also evaluated, and the one with the best evaluation effect is selected as the improved YOLOv8 network. The corn plant image is input into the trained improved YOLOv8 network, including the Backbone backbone network redesigned C2fCA network module for convolution processing of the corn plant image, and then the CA attention mechanism is used to better and more accurately locate the target to obtain the first feature map; the first feature map is input into the Neck network to obtain the second feature map; the second feature map is predicted, the loss function is adjusted, the regression accuracy of the prediction box is improved, and the confidence of the output corn stem diameter, leaf length and canopy width is improved.

[0104] Step 3: Environmental information collection (performed by the environmental information acquisition module).

[0105] The environmental information closely related to the growth of corn includes: soil moisture content (current humidity) (unit: %); air temperature (unit: °C); air humidity (unit: %). The above data is obtained by deploying wireless sensor nodes in the field and synchronously transmitted to the edge device.

[0106] Step 4: Water requirement estimation and irrigation decision generation (performed by the intelligent irrigation decision module).

[0107] The system uses the identified corn plant phenotypic characteristics (such as leaf length , plant height , stem diameter , leaf angle ) and field environmental information: air temperature , air humidity , soil moisture content (current humidity) We constructed a phenotypic-water requirement mapping model to achieve dynamic water requirement identification and precise irrigation control for individual crops.

[0108] .

[0109] All variables are collected data. Irrigation volume is determined by comparing target humidity with actual soil moisture. :

[0110] .

[0111] in, This refers to the amount of irrigation water. This refers to the root system coverage area of ​​a single plant. Effective root depth; Target humidity; This represents the current humidity.

[0112] Based on the calculation results, the specific solenoid valve opening time is generated. It is used for precise control of water volume.

[0113] Irrigation plans can be adaptively adjusted based on crop conditions and environmental parameters to achieve water conservation and increased yield.

[0114] Step 5: Irrigation control command execution (executed by the solenoid valve control module).

[0115] Irrigation commands are sent to the smart solenoid valve via IoT communication protocols (such as RS485 or LoRa). The solenoid valve then adjusts its operation based on the opening time. It automatically completes the irrigation process.

[0116] Step 6: Feedback and iteration (executed in a coordinated manner by the entire system).

[0117] After irrigation is completed, image and sensor data are reacquired for model correction and control strategy optimization. An online model update mechanism or adaptive parameter tuning mechanism can be introduced. This forms a closed-loop control system of "identification-decision-execution-feedback," improving system intelligence and robustness.

[0118] The advantages of this application compared to existing technologies are:

[0119] (1) High recognition accuracy: Based on the binocular depth map and the improved YOLOv8 structure, the improved YOLOv8 network improves the accuracy of corn phenotypic recognition; the precision P, recall R, and mean accuracy mAP0.5 and mAP0.5:0.95 reached 96.80%, 94.10%, 96.60%, and 77.00%, respectively.

[0120] (2) Fast response speed: edge computing + local deployment of recognition model, real-time response. Image / point cloud collection to recognition completion time ≤ 2s / plant; environmental data collection + fusion decision output ≤ 1s; irrigation instruction delay ≤ 500ms; overall response cycle (from recognition to execution) ≤ 3s, realizing closed-loop control.

[0121] (3) Decision intelligence: based on crop individual phenotype (such as plant height, leaf length, stem diameter, leaf angle, etc.) and current environmental data to build water demand estimation model, realizing individual water calculation. Support ≥ 5 kinds of crop phenotype parameter fusion modeling (such as 、 、 、 , LAI), which can adaptively adjust model parameters (such as irrigation coefficient dynamic adjustment), accurately estimate single plant target water content error ≤ ± 5%, realize single plant level and regional level double scale irrigation decision and control.

[0122] (4) Control precision: using electromagnetic valve + flow sensor combination to realize quantitative control and closed-loop verification. Can realize: irrigation error ≤ ± 3% (target irrigation water quantity and actual irrigation water quantity ratio); system water saving rate improvement ≥ 30% (compared with fixed time irrigation method); crop unit water efficiency improvement ≥ 25% (unit irrigation water output ratio improvement); energy consumption reduction ≥ 15% (reducing invalid pumping and overflow).

[0123] (5) Strong adaptability: applicable to different corn varieties and different environmental conditions.

[0124] From the perspective of technical integration: build a unique combination. This system integrates crop phenotype recognition, environmental parameter sensing and edge intelligent decision-making into one, forming an intelligent irrigation platform with "sensing-identification-decision-execution" closed loop. Compared with existing irrigation systems mainly based on fixed time control or single parameter feedback, this system can realize individual-level precision irrigation control for crops, with high integration and dynamic adaptability.

[0125] From the perspective of application matching degree: emphasize customization and complexity response. For the problems of large crops such as corn that have dramatic phenotype changes in different periods and complex water demand rules, the system drives irrigation decision based on crop phenotype parameters, without relying on manual intervention or being limited to environmental threshold trigger control. Existing solutions cannot realize "phenotype-water-irrigation" continuous mapping, and cannot meet the needs of individualized precision irrigation under complex spatio-temporal distribution.

[0126] From the edge real-time control perspective: emphasize the "sense-decision-control" closed loop. The system deploys recognition models and decision logic based on edge computing architecture, without relying on remote servers to achieve second-level recognition and response, complete local autonomous control closed loop. Most existing systems still rely on central server processing and manual intervention, cannot run continuously in the field environment such as lack of network and unstable data, so it is difficult to replace the high adaptability and autonomous decision-making ability of the system in complex field scenarios.

[0127] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present disclosure.

[0128] The principles and implementation modes of the present application are described by applying specific examples herein, and the above descriptions of the embodiments are only used to help understand the method of the present application and its core idea; meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation modes and application ranges will have changes. In conclusion, the content of the present description should not be understood as a limitation of the present application.

Claims

1. A crop irrigation regulation and control method based on phenotypic characteristics, characterized in that, include: Obtain information data on different crop zones within the target farmland; The information data includes: crop image data and field environment information; For any crop zone, the following is performed: Based on the crop phenotypic model, phenotypic feature information is determined according to the crop image data; the crop phenotypic model is obtained by adjusting and training based on the improved YOLOv8 network and the historical crop image data with known phenotypic feature information, with the goal of minimizing the loss function; the improved YOLOv8 network includes a C2fCA network module with added attention mechanism. Based on the phenotypic-water requirement mapping model, water requirement information is determined according to the phenotypic feature information and the field environment information; the water requirement information includes: target humidity; Based on the water demand information, corresponding irrigation control is carried out to achieve irrigation regulation of crops in the target farmland; The mathematical expression corresponding to the phenotype-water requirement mapping model is: ； in, Target humidity; For leaf length; The blade tilt angle; Plant height; For temperature; Air humidity; , , , as well as All are fitting coefficients.

2. The crop irrigation regulation and control method based on phenotypic characteristics according to claim 1, characterized in that, After obtaining information data on different crop zones within the target farmland, the process also includes: The crop image data is horizontally flipped to obtain the flipped crop image data; The saturation, brightness, and contrast of the flipped crop image data are randomly adjusted to obtain the adjusted crop image data. The adjusted crop image data is then sharpened and enhanced.

3. The crop irrigation regulation and control method based on phenotypic characteristics according to claim 1, characterized in that, Based on the water demand information, corresponding irrigation control is implemented to regulate irrigation for crops in the target farmland, specifically including: The irrigation water volume is determined based on the water demand information. Irrigation control decisions are made for the corresponding crop zones based on the irrigation water volume, so as to achieve irrigation regulation of crops in the target farmland; the irrigation control decisions include irrigation duration.

4. The crop irrigation regulation and control method based on phenotypic characteristics according to claim 3, characterized in that, The expression for the irrigation water volume is: ； in, This refers to the amount of irrigation water. This refers to the root system coverage area of ​​a single plant. Effective root depth; Target humidity; This represents the current humidity.

5. The crop irrigation regulation and control method based on phenotypic characteristics according to claim 1, characterized in that, The method for determining the crop phenotypic model specifically includes: Obtain a training dataset; the training dataset includes historical crop image data with known phenotypic features. The training dataset is divided into a training set, a validation set, and a test set according to a set ratio; Construct an improved YOLOv8 network; The training set is input into the improved YOLOv8 network, and the network parameters are adjusted and trained with the goal of minimizing the loss function to obtain the trained network; the network parameters include: confidence; The validation set is input into the trained network for validation, resulting in a validated network. The test set is input into the verified network for testing to obtain the tested network. The tested network was identified as a crop phenotypic model.

6. The crop irrigation regulation and control method based on phenotypic characteristics according to claim 1, characterized in that, The crop irrigation regulation and control method based on phenotypic characteristics further includes: The system acquires information data corresponding to the irrigation adjustment of crops in the target farmland and feeds it back to the host computer for crop irrigation coordination.

7. A crop irrigation regulation and control system based on phenotypic characteristics, characterized in that, The crop irrigation regulation and control system based on phenotypic features is implemented using the crop irrigation regulation and control method based on phenotypic features as described in any one of claims 1-6; the crop irrigation regulation and control system based on phenotypic features includes: an image acquisition module, an image recognition and processing module, an environmental information acquisition module, an intelligent irrigation decision module, and a solenoid valve execution module; The image recognition and processing module is connected to the image acquisition module; both the image recognition and processing module and the environmental information acquisition module are connected to the intelligent irrigation decision module; the solenoid valve execution module is connected to the intelligent irrigation decision module. The image acquisition module is used to acquire crop image data of different crop zones within the target farmland; The environmental information acquisition module is used to acquire field environmental information of different crop zones within the target farmland; The image recognition and processing module is used to partition any crop and determine phenotypic feature information based on the crop image data according to the crop phenotypic model. The intelligent irrigation decision module is used to determine water demand information based on the phenotypic feature information and the field environment information, according to the phenotypic-water demand mapping model. The solenoid valve actuator module is used to perform corresponding irrigation control based on the water demand information, so as to achieve irrigation regulation of crops in the target farmland.

8. The crop irrigation regulation and control system based on phenotypic characteristics according to claim 7, characterized in that, The image acquisition module uses a binocular camera.

9. The crop irrigation regulation and control system based on phenotypic characteristics according to claim 8, characterized in that, The crop irrigation regulation and control system based on phenotypic characteristics also includes: a local storage module and a screen display module; Both the local storage module and the screen display module are connected to the binocular camera. The local storage module is used to store crop image data acquired by the binocular camera; The screen display module is used to display and view the acquisition information of the binocular camera.

Images