Spraying parameter acquisition method based on plant canopy density and spraying machine

Abstract

The invention provides a spraying parameter acquisition method based on plant canopy density and a spraying machine, and relates to the technical field of agricultural plant protection, and the method comprises the steps: determining the point cloud number of a to-be-sprayed plant in a plurality of canopy regions according to the canopy point cloud data of the to-be-sprayed plant; determining a leaf area index of each canopy region of the to-be-sprayed plant according to the number of point clouds of the to-be-sprayed plant in the plurality of canopy regions; determining the canopy density of each canopy region of the to-be-sprayed plant according to the leaf area index of each canopy region of the to-be-sprayed plant; inputting the canopy density of each canopy area of the to-be-sprayed plant into the constructed correlation model to obtain a spraying parameter corresponding to each canopy area of the to-be-sprayed plant; the association model is used for representing an association relationship between the canopy density and the spraying parameters. According to the method and the device, the spraying parameters corresponding to each canopy region are accurately obtained, so that the spraying accuracy and efficiency are improved.

Description

Technical Field

[0001] The invention relates to the technical field of agricultural plant protection, and in particular to a method for obtaining spray parameters based on plant canopy density and a sprayer. Background Art

[0002] Fruit tree planting is an important part of modern agricultural production. With the use and popularization of pesticides, how to improve the accuracy and efficiency of spraying, reduce waste and environmental pollution and other issues still need to be solved.

[0003] The traditional method of spraying fruit trees usually relies on manual experience to adjust the spray volume and then spray the whole tree. This method has many shortcomings. First, the spray volume is usually large, which can easily lead to excessive use of pesticides, increase costs, and cause unnecessary pollution to the environment; second, due to the differences in the growth status of different fruit trees, a uniform spray volume is difficult to meet the actual needs of each tree, which may lead to insufficient spraying in some areas or excessive spraying in some areas, affecting the effect of pesticides and even causing waste of resources.

[0004] In recent years, variable rate spraying (VRS) technology based on crop growth characteristics has gradually become an effective means to solve the problems of traditional spraying. Variable rate spraying technology uses real-time sensing data, combined with parameters such as crop growth status and vegetation coverage, to dynamically adjust the spray volume to achieve accurate and efficient pesticide application.

[0005] However, existing variable spray technologies mostly rely on static image recognition, sensor measurement or simple algorithm models, resulting in the accuracy and efficiency of spraying still need to be improved. Summary of the invention

[0006] The present invention provides a method for obtaining spray parameters based on plant canopy density and a spray machine, so as to solve the defect that the accuracy and efficiency of spraying in the prior art need to be improved, and achieve accurate acquisition of spray parameters, thereby improving the accuracy and efficiency of spraying.

[0007] The present invention provides a method for obtaining spray parameters based on plant canopy density, comprising: Determining the number of point clouds of the plants to be sprayed in multiple canopy areas according to the canopy point cloud data of the plants to be sprayed; Determining the leaf area index of each canopy area of ​​the plant to be sprayed according to the number of point clouds of the plant to be sprayed in multiple canopy areas; Determining the canopy density of each canopy area of ​​the plant to be sprayed according to the leaf area index of each canopy area of ​​the plant to be sprayed; The canopy density of each canopy area of ​​the plant to be sprayed is input into the constructed correlation model to obtain the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; the correlation model is used to characterize the correlation between the canopy density and the spray parameters.

[0008] In some embodiments, determining the number of point clouds of the plants to be sprayed in multiple canopy areas according to the canopy point cloud data of the plants to be sprayed includes: Dividing the canopy of the plant to be sprayed into a plurality of canopy areas; Acquiring point cloud data of the plurality of canopy regions according to the canopy point cloud data of the plants to be sprayed; Divide the point cloud data of each canopy area into multiple voxels and calculate the number of point clouds in each voxel; The number of point clouds of the plants to be sprayed in the multiple canopy areas is determined according to the number of point clouds in each voxel and the multiple voxels corresponding to each canopy area.

[0009] In some embodiments, before acquiring the point cloud data of the plurality of canopy regions according to the canopy point cloud data of the plant to be sprayed, the method further includes: Based on the Euclidean distance clustering algorithm, the canopy point cloud data of the plant group is clustered to obtain the cluster center point of each plant; A cylindrical space model is established with the cluster center point of each plant as the center; The cylindrical space model is used to segment the canopy point cloud data of the plant group to obtain the canopy point cloud data of each individual plant.

[0010] In some embodiments, before clustering the canopy point cloud data of the plant group based on the Euclidean distance clustering algorithm to obtain the cluster center point of each plant, the method further includes: Obtain initial point cloud data of vegetation canopy from multiple different measuring stations; Based on the target sphere fitting, the initial point cloud data is registered to obtain the first point cloud data; According to random sampling and curvature sampling, the first point cloud data is layered and simplified to obtain second point cloud data; Based on the HSI threshold and through filtering, the second point cloud data is filtered to obtain the canopy point cloud data of the plant population.

[0011] In some embodiments, the registering the initial point cloud data based on the target sphere fitting to obtain the first point cloud data includes: Performing target sphere detection on the initial point cloud data and extracting the coordinates of the sphere center; Based on the spherical center coordinates and the ID of the target sphere, matching the same target sphere is performed in the initial point cloud data of different measuring stations; According to each matched target sphere, the corresponding relationship between the initial point cloud data of different measuring stations is established; According to the corresponding relationship, the transformation matrix between the initial point cloud data of different measuring stations is calculated; The transformation matrix is ​​registered with the initial point cloud data based on the ICP algorithm to obtain first point cloud data.

[0012] In some embodiments, the step of performing hierarchical simplification on the first point cloud data according to random sampling and curvature sampling to obtain second point cloud data includes: Constructing a rectangular bounding box surrounding the first point cloud data; Dividing the spatial structure of the rectangular bounding box into multiple levels; The first point cloud data in each layer is sampled by using random sampling and curvature sampling to obtain the second point cloud data.

[0013] In some embodiments, the spray parameters include: wind speed, driving speed and spray flow rate.

[0014] The present invention also provides a sprayer, comprising: a moving wheel, a bearing plate, a fan, a liquid storage tank, a spray head, a data acquisition module, a laser radar and an electronic control system; The moving wheel is installed under the bearing plate, and the bearing plate carries the fan, the liquid storage tank, the laser radar and the electronic control system; The liquid storage tank is connected to the nozzle, and the electronic control system is respectively connected to the laser radar, the nozzle, the fan, the moving wheel and the data acquisition module; The moving wheel is used to drive the sprayer to move; The bearing plate is used to provide a bearing platform; The fan is used to provide wind speed during spraying; The liquid storage tank is used to store the liquid to be sprayed; The spray head is used to spray the liquid; The data acquisition module is used to collect the current rotation speed of the fan, the current driving speed of the moving wheel and the current spray flow rate of the nozzle; The laser radar is used to collect canopy point cloud data of the plants to be sprayed in the forward direction of the sprayer; The electronic control system is used to execute the spray parameter acquisition method based on the plant canopy density as described in any one of claims 1 to 7 according to the canopy point cloud data of the plant to be sprayed, and output the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; and automatically adjust the spray flow rate of the nozzle, the rotation speed of the fan and the travel speed of the moving wheel according to the spray parameters corresponding to each canopy area of ​​the plant to be sprayed.

[0015] In some embodiments, the data acquisition module includes: an encoder, a unit sensor, and a flow sensor; The encoder is nested in the connecting shaft between the fan and the motor, the unit sensor is arranged beside the electronic control system, the flow sensor is nested in the infusion tube of the nozzle, and the electronic control system is respectively connected to the encoder, the unit sensor and the flow sensor; The motor is used to provide power to the fan; The encoder is used to collect the current rotation speed of the fan; The unit sensor is used to collect the current travel speed of the moving wheel; The flow sensor is used to collect the current spray flow of the nozzle.

[0016] In some embodiments, the electronic control system includes: a computer, a flow controller, a wind speed controller, and a speed controller; The flow controller is connected to the flow sensor, the nozzle and the computer respectively; the wind speed controller is connected to the encoder, the fan and the computer respectively; the speed controller is connected to the unit sensor, the moving wheel and the computer respectively; The computer is used to execute the method for obtaining spray parameters based on plant canopy density according to any one of claims 1 to 7 according to the canopy point cloud data of the plant to be sprayed, and output the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; The flow controller is used to automatically adjust the spray flow of the nozzle according to the current spray flow of the nozzle and the spray flow in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed; The wind speed controller is used to automatically adjust the speed of the fan according to the current speed of the fan and the wind speed in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed; The speed controller is used to automatically adjust the travel speed of the moving wheel according to the current travel speed of the moving wheel and the travel speed in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed.

[0017] The spray parameter acquisition method and sprayer based on plant canopy density provided by the present invention divide the canopy of the plant to be sprayed into multiple canopy areas, determine the leaf area index of each canopy area of ​​the plant to be sprayed according to the number of point clouds of the plant to be sprayed in multiple canopy areas, determine the canopy density of each canopy area of ​​the plant to be sprayed according to the leaf area index of each canopy area of ​​the plant to be sprayed, and then obtain the spray parameters corresponding to each canopy area of ​​the plant to be sprayed by constructing an association model for characterizing the association relationship between canopy density and spray parameters. The present invention realizes accurate acquisition of the spray parameters corresponding to each canopy area by real-time identification of the canopy density of each canopy area of ​​the plant, and adjusts the spray parameters according to the canopy density of each canopy area, thereby improving the accuracy and efficiency of spraying, reducing pesticide waste, maximizing the effect of pesticide use, and reducing the environmental burden. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] 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.

[0019] Figure 1 It is one of the flow diagrams of the method for obtaining spray parameters based on plant canopy density provided by the present invention.

[0020] Figure 2 This is the second flow chart of the method for obtaining spray parameters based on plant canopy density provided by the present invention.

[0021] Figure 3 This is the third flow chart of the method for obtaining spray parameters based on plant canopy density provided by the present invention.

[0022] Figure 4 This is one of the structural schematic diagrams of the sprayer provided by the present invention.

[0023] Figure 5 This is the second structural schematic diagram of the sprayer provided by the present invention.

[0024] Figure 6 It is a schematic diagram of the structure of the electric control system provided by the present invention.

[0025] Reference numerals: DETAILED DESCRIPTION

[0026] 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.

[0027] Figure 1 is one of the flow diagrams of the method for obtaining spray parameters based on plant canopy density provided by the present invention, such as Figure 1 As shown, the present invention provides a method for obtaining spray parameters based on plant canopy density, comprising: Step 110, determining the number of point clouds of the plants to be sprayed in multiple canopy areas based on the canopy point cloud data of the plants to be sprayed.

[0028] Specifically, a laser radar (LiDAR) is used to obtain the number of canopy point clouds of the plants to be sprayed, and the canopies of the plants to be sprayed are divided into regions to obtain the number of point clouds of the plants to be sprayed in multiple canopy regions.

[0029] Step 120, determining the leaf area index of each canopy area of ​​the plant to be sprayed according to the number of point clouds of the plant to be sprayed in multiple canopy areas.

[0030] Specifically, for example, the canopy of the plant to be sprayed is divided into an upper region, a middle region, a lower region and a region close to the ground.

[0031] Upper area: Located at the top of the canopy, usually exposed to strong sunlight. The leaf density here is relatively small, but due to sufficient sunlight, it may be a high-incidence area for pests and diseases, especially during the crop maturity period, when the reproduction of pests and the spread of diseases may first affect the upper part of the canopy.

[0032] Middle area: Located in the middle area of ​​the crown, it usually has a more even distribution of leaves. There are more leaves here with moderate density. This part of the leaves usually occupies the main part of the plant to absorb sunlight and photosynthesize, and is also a potential spread area for diseases and pests.

[0033] Lower area: Located at the bottom of the crown, the density of leaves is usually higher, but because it is closer to the ground, it is less affected by light and has poor air circulation, so moisture and pathogens are easily accumulated here.

[0034] Area close to the ground: The area at the bottom of the crown is very close to the ground and has high air humidity. The leaves in the area close to the ground have low resistance to diseases and pests and are easily affected by diseases and pests in the bottom layer, so special attention needs to be paid.

[0035] According to the number of point clouds of each canopy area of ​​the plant to be sprayed, the leaf area index of each canopy area of ​​the plant to be sprayed is determined. The specific formula is as follows: In the formula, represents the leaf area index of the upper canopy area, Represents the total number of point clouds in the entire canopy, represents the number of point clouds in the upper canopy area, represents the leaf area index of the middle area of ​​the canopy, represents the number of point clouds in the middle area of ​​the canopy, Leaf area index representing the lower canopy area represents the number of point clouds in the lower canopy area, Represents the leaf area index close to the ground, Represents the number of point clouds close to the ground area.

[0036] A data pair includes the number of point clouds and the leaf area index of each canopy area. The number of point clouds of multiple canopy areas can be obtained by using lidar or sensors carried by drones, and the corresponding leaf area index can be obtained at the same time and location by direct measurement or using remote sensing technology (e.g., ground spectral sensors, drones, etc.).

[0037] Step 130, determining the canopy density of each canopy area of ​​the plant to be sprayed according to the leaf area index of each canopy area of ​​the plant to be sprayed.

[0038] Specifically, the leaf area index is an important parameter to characterize the density of the plant canopy. The canopy density (CanopyDensity) usually describes the density of the leaves in the plant canopy. According to the leaf area index of each canopy area of ​​the plant to be sprayed, the canopy density of each canopy area of ​​the plant to be sprayed can be determined.

[0039] The relationship between leaf area index and canopy density is shown below: In the formula, C represents the canopy density, and K is an empirical coefficient that depends on the plant species, canopy type, and growth stage. Represents leaf area index.

[0040] The canopy structures of different plant species vary greatly, so it is necessary to determine an appropriate coefficient K based on the canopy structure of the specific plant. For example, denser plants (such as forest trees) usually have higher k values, while sparse herbaceous plants (such as some crops) have lower K values.

[0041] Step 140, input the canopy density of each canopy area of ​​the plant to be sprayed into the constructed association model to obtain the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; the association model is used to characterize the association relationship between the canopy density and the spray parameters.

[0042] Specifically, the correlation model can output the best spray parameters according to different canopy densities, and the correlation model can be obtained through a liquid medicine deposition test. In some embodiments, the spray parameters may include: wind speed, driving speed, and spray flow rate.

[0043] For example, taking cotton as the spraying object, three experimental factors, namely spray flow rate, wind speed and forward speed of the sprayer, are selected, and multiple levels are set to conduct spray deposition experiments on cotton with different canopy densities, and the pesticide utilization rate and deposition uniformity (coefficient of variation) of each group of experiments are obtained. According to the different emphasis on the pursuit of efficient utilization of pesticides or uniform deposition of liquid medicine in the spraying requirements, the pesticide utilization rate and deposition uniformity are given corresponding weights to construct the spray objective function. A Gaussian response surface is established based on the parameters and experimental results of the spray deposition experiment, and then the optimal solution of each spray objective function is solved by a genetic algorithm, thereby obtaining the optimal association model corresponding to different canopy densities.

[0044] After the constructed correlation model is obtained, the canopy density of each canopy area of ​​the plant to be sprayed is input into the constructed correlation model to obtain the spray parameters corresponding to each canopy area of ​​the plant to be sprayed.

[0045] The spray parameter acquisition method based on the plant canopy density provided by the embodiment of the present invention divides the canopy of the plant to be sprayed into multiple canopy areas, determines the leaf area index of each canopy area of ​​the plant to be sprayed according to the number of point clouds of the plant to be sprayed in the multiple canopy areas, determines the canopy density of each canopy area of ​​the plant to be sprayed according to the leaf area index of each canopy area of ​​the plant to be sprayed, and then obtains the spray parameters corresponding to each canopy area of ​​the plant to be sprayed by constructing an association model for characterizing the association between canopy density and spray parameters. The present invention realizes accurate acquisition of the spray parameters corresponding to each canopy area by real-time identification of the canopy density of each canopy area of ​​the plant, and adjusts the spray parameters according to the canopy density of each canopy area, thereby improving the accuracy and efficiency of spraying, reducing pesticide waste, maximizing the effect of pesticide use, and reducing the environmental burden.

[0046] In some embodiments, determining the number of point clouds of the plants to be sprayed in multiple canopy areas based on the canopy point cloud data of the plants to be sprayed includes: Divide the canopy of the plant to be sprayed into a plurality of canopy areas; According to the canopy point cloud data of the plants to be sprayed, point cloud data of multiple canopy areas are obtained; Divide the point cloud data of each canopy area into multiple voxels and calculate the number of point clouds in each voxel; According to the number of point clouds in each voxel and the multiple voxels corresponding to each canopy area, the number of point clouds of the plants to be sprayed in multiple canopy areas is determined.

[0047] Specifically, the canopy of the plant to be sprayed is divided into a plurality of canopy regions, and then the point cloud data of the canopy of the plant to be sprayed is divided into the plurality of canopy regions to obtain the point cloud data of the plurality of canopy regions.

[0048] The point cloud data of each canopy area is divided into multiple voxels, each voxel represents a spatial unit, and the number of point clouds in each voxel is calculated.

[0049] According to the multiple voxels corresponding to each canopy area, the number of point clouds in the multiple voxels is accumulated to obtain the number of point clouds of the plants to be sprayed in each canopy area.

[0050] The spray parameter acquisition method based on the plant canopy density provided in an embodiment of the present invention first divides the canopy of the plant to be sprayed into regions to reduce the point cloud space, then divides the point cloud data of each canopy region into multiple voxels, further reducing the point cloud space, and finally accurately obtains the number of point clouds in each canopy region of the plant to be sprayed through the number of point clouds in each voxel.

[0051] In some embodiments, Figure 2 FIG. 2 is a flow chart of the method for obtaining spray parameters based on plant canopy density provided by the present invention. Figure 2 As shown, before acquiring point cloud data of multiple canopy areas according to the canopy point cloud data of the plant to be sprayed, the method further includes: Step 210: clustering the canopy point cloud data of the plant group based on the Euclidean distance clustering algorithm to obtain the cluster center point of each plant.

[0052] Specifically, the target point cloud data is clustered using the Euclidean distance clustering segmentation algorithm to extract the cluster center point of each plant. Set the cluster tolerance (for example, 0.02 meters, or 2 centimeters), which determines which points in the point cloud are considered to be part of the same cluster. Set the minimum and maximum cluster sizes to avoid too small or too large clusters affecting the results.

[0053] Step 220, establishing a cylindrical space model with the cluster center point of each plant as the center.

[0054] Specifically, a cylindrical space model is established with the cluster center point of each plant as the center, that is, the radius and height of the cylindrical space model are set.

[0055] The radius of the cylinder space model is usually set according to the row spacing or plant size of the crop. For example, if the row spacing of the crop is fixed, the radius of the cylinder can be set to half the row spacing to ensure that each cylinder space roughly corresponds to one plant in a row.

[0056] The height of the column space model is usually determined according to the growth height of the crop. For different growth stages, the height of the plant will be different, so the height of the column space model needs to be set according to the specific growth and development stage.

[0057] Step 230, using the cylindrical space model to segment the canopy point cloud data of the plant group to obtain the canopy point cloud data of each individual plant.

[0058] Specifically, since each cylindrical space roughly corresponds to a plant in a row, the canopy point cloud data of the plant group is segmented using the cylindrical space model to obtain the canopy point cloud data of each individual plant.

[0059] In some embodiments, the segmentation result may be verified to check the accuracy of the segmentation. If the accuracy of the segmentation does not meet the requirement, re-clustering and segmentation may be performed.

[0060] The variable spraying method based on plant canopy density provided by the embodiment of the present invention clusters the canopy point cloud data of the plant group based on the Euclidean distance clustering algorithm to obtain the cluster center point of each plant, establishes a cylindrical space model with the cluster center point of each plant as the center, and uses the cylindrical space model to segment the canopy point cloud data of the plant group to obtain the canopy point cloud data of each individual plant, thereby accurately acquiring the canopy point cloud data of each individual plant, which is useful for subsequent processing of the canopy point cloud data of the plant to be sprayed.

[0061] In some embodiments, Figure 3 FIG. 3 is a flow chart of the method for obtaining spray parameters based on plant canopy density provided by the present invention, such as Figure 3 As shown, before clustering the canopy point cloud data of the plant group based on the Euclidean distance clustering algorithm to obtain the cluster center point of each plant, it also includes: Step 310 , obtaining initial point cloud data of the vegetation canopy from a plurality of different measuring stations.

[0062] Specifically, a 3D laser scanner (such as a ground-based LiDAR scanning system) is used to obtain initial point cloud data of the vegetation canopy. These data contain the phenotypic structure and spatial information of the vegetation, providing a basis for subsequent density analysis.

[0063] Step 320: Based on the target sphere fitting, the initial point cloud data is registered to obtain the first point cloud data.

[0064] Specifically, the initial point cloud data collected from multiple different measuring stations are registered to ensure the consistency and accuracy of the data. A point cloud registration method based on target sphere fitting can be used to achieve point cloud data registration of different measuring stations by fitting the sphere to solve the coordinates of the sphere center.

[0065] Step 330 , hierarchically simplify the first point cloud data according to random sampling and curvature sampling to obtain second point cloud data.

[0066] Specifically, the point cloud data with tens of millions of points is simplified to improve the efficiency of subsequent processing and reduce the requirements for computer hardware and software. The layered simplification method can be used to combine random sampling and curvature sampling algorithms to achieve point cloud data simplification.

[0067] Step 340 , filtering the second point cloud data based on the HSI threshold and through filtering to obtain canopy point cloud data of the plant population.

[0068] Specifically, the collected RGB point cloud data is converted to the HSI color space, and the appropriate HSI threshold is set according to the characteristics of the crop plant canopy. Then, the second point cloud data is filtered using the straight-through filter combined with the HSI color model to automatically extract the canopy point cloud data of the plant group. This method improves work efficiency and provides data support for the segmentation of the plant group canopy point cloud data.

[0069] The variable spray method based on plant canopy density provided in the embodiment of the present invention realizes automatic and efficient extraction of canopy point cloud data of plant groups by fitting, simplifying and filtering the initial point cloud data of the vegetation canopy.

[0070] In some embodiments, the initial point cloud data is registered based on the target sphere fitting to obtain the first point cloud data, including: Perform target ball detection on the initial point cloud data and extract the coordinates of the ball center; Based on the similarity of the sphere center coordinates and the ID of the target sphere, the same target sphere is matched in the initial point cloud data of different measuring stations; According to each matched target sphere, the corresponding relationship between the initial point cloud data of different measuring stations is established; According to the corresponding relationship, the transformation matrix between the initial point cloud data of different measuring stations is calculated; Based on the ICP algorithm, the transformation matrix is ​​registered on the initial point cloud data to obtain the first point cloud data.

[0071] Specifically, the target spheres are detected in the initial point cloud data by finding the point sets with the local maximum curvature, which correspond to the surface of the sphere. For each detected target sphere, an algorithm (such as the Random Sample Consensus (RANSAC) algorithm) is used to fit the sphere and extract the coordinates of the sphere center.

[0072] Usually, target spheres are matched in the initial point cloud data of different measuring stations based on the similarity of the sphere center coordinates and the same ID of the target spheres. For each matched target sphere, a mapping relationship between the initial point cloud data of different measuring stations is established.

[0073] For example, suppose there are two stations A and B, and a target ball is detected in the point cloud data of station A. And determine the coordinates of the center of the sphere as , the matching target sphere is also detected in the point cloud data of station B The coordinates of the center of the sphere are Then, this mapping relationship can be expressed as , that is, the mapping from the coordinates of the center of the target ball in station A to the coordinates of the center of the target ball in station B.

[0074] Using this mapping relationship, calculate the transformation matrix between the initial point cloud data of different stations, that is, the transformation matrix (including translation and rotation) from the initial point cloud data of one station to the initial point cloud data of another station. Use the Iterative Closest Point (ICP) algorithm or other point cloud registration algorithms to continuously optimize the transformation matrix through an iterative process, and further optimize the registration result of the initial point cloud data according to the optimized transformation matrix until the predetermined registration accuracy is reached to obtain the first point cloud data.

[0075] Verify the accuracy of the registration by comparing the initial point cloud data after registration with the known target sphere position. Analyze the registration error to ensure that the registration result meets the engineering or research requirements.

[0076] In some embodiments, before detecting the target sphere, it is necessary to select and arrange the target sphere, collect data, and pre-process the point cloud. The specific process is as follows: Target sphere selection and placement: Select target spheres with high reflectivity to ensure they can be clearly identified in laser scanning. Arrange target spheres within the field of view of each station to ensure they can be detected in the data of multiple stations.

[0077] Data collection: Use a 3D laser scanner to scan the target area, while ensuring that the target sphere is included in the scan data.

[0078] Point cloud preprocessing: De-noise the collected initial point cloud data to remove the noise generated during the scanning process. Apply filtering algorithms, such as radius filtering or statistical filtering, to further clean up the point cloud data.

[0079] The spray parameter acquisition method based on plant canopy density provided by an embodiment of the present invention performs target sphere detection on initial point cloud data and extracts the coordinates of the sphere center; based on the similarity of the sphere center coordinates and the ID of the target sphere, the same target sphere is matched in the initial point cloud data of different measuring stations; according to each matched target sphere, a corresponding relationship between the initial point cloud data of different measuring stations is established; according to the corresponding relationship, the transformation matrix between the initial point cloud data of different measuring stations is calculated; based on the ICP algorithm, the transformation matrix is ​​aligned with the initial point cloud data to obtain the first point cloud data, thereby improving the alignment accuracy.

[0080] In some embodiments, the first point cloud data is hierarchically simplified according to random sampling and curvature sampling to obtain the second point cloud data, including: Constructing a rectangular bounding box surrounding the first point cloud data; Divide the spatial structure of the rectangular bounding box into multiple levels; Random sampling and curvature sampling are used to sample the first point cloud data in each layer to obtain the second point cloud data.

[0081] Specifically, before constructing the rectangular bounding box, a statistical filtering method is used to eliminate outliers and noise points to provide a clean data basis for the subsequent simplification algorithm.

[0082] A rectangular bounding box (such as a cuboid bounding box or a cube bounding box) is constructed to enclose the first point cloud data, where the rectangular bounding box can enclose all the first point cloud data.

[0083] The spatial structure of the rectangular bounding box is divided into multiple levels. In some embodiments, at each level, the point cloud space is further divided into smaller units, such as voxel units, for subsequent processing.

[0084] Random sampling is a simple and direct point cloud simplification method that reduces the amount of data by randomly selecting a certain number of point clouds. The curvature of each point in the point cloud is calculated, and the curvature can provide important information about the local shape and characteristics of the underlying surface. Sampling is performed based on the local curvature value of the point cloud, and points with larger curvature are preferentially retained, which usually correspond to feature points or edge points in the point cloud.

[0085] Apply random sampling and curvature sampling at different levels to sample the first point cloud data to obtain the second point cloud data to achieve a better simplification effect. In different levels, the ratio of random sampling and curvature sampling can be different. For example, the ratio of random sampling can be increased in flat areas, and the ratio of curvature sampling can be increased in complex or curvature areas.

[0086] The spray parameter acquisition method based on plant canopy density provided by an embodiment of the present invention first constructs a rectangular bounding box enclosing the first point cloud data, then divides the spatial structure of the rectangular bounding box into multiple levels, and finally applies random sampling and curvature sampling at different levels to achieve flexible sampling, adapt to different types of sampling areas, simplify the number of point clouds, so as to improve the work efficiency of subsequent processing and reduce the requirements for computer hardware and software.

[0087] Figure 4 This is one of the structural schematic diagrams of the sprayer provided by the present invention. Figure 5 This is the second structural diagram of the spray machine provided by the present invention, such as Figure 4 and 5 As shown, the present invention provides a sprayer, comprising: a moving wheel 1, a carrying plate 2, a fan 3, a liquid storage tank 4, a spray head 5, a data acquisition module 6, a laser radar 7 and an electronic control system 8.

[0088] The moving wheel 1 is installed under the bearing plate 2, and the bearing plate 2 carries the fan 3, the liquid storage tank 4, the laser radar 7 and the electronic control system 8; The liquid storage tank 4 is connected to the nozzle 5, and the electronic control system 8 is respectively connected to the laser radar 7, the nozzle 5, the fan 3, the moving wheel 1 and the data acquisition module 6; The moving wheel 1 is used to drive the sprayer to move; The carrying plate 2 is used to provide a carrying platform; The fan 3 is used to provide wind speed during spraying; The liquid storage tank 4 is used to store the liquid to be sprayed; The nozzle 5 is used for spraying liquid; The data acquisition module 6 is used to collect the current rotation speed of the fan 3, the current travel speed of the moving wheel 1 and the current spray flow rate of the nozzle 5; The laser radar 7 is used to collect the canopy point cloud data of the plants to be sprayed in the forward direction of the sprayer; The electronic control system 8 is used to execute any of the above-mentioned spray parameter acquisition methods based on the plant canopy density according to the canopy point cloud data of the plant to be sprayed, and output the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; and automatically adjust the spray flow rate of the nozzle 5, the rotation speed of the fan 3 and the driving speed of the moving wheel 1 according to the spray parameters corresponding to each canopy area of ​​the plant to be sprayed.

[0089] Specifically, the sprayer is moved by the moving wheels 1. The sprayer comprises at least one moving wheel 1. For example, the number of the moving wheels 1 may be 1, 2, 3 or 4.

[0090] The moving wheel 1 is installed under the bearing plate 2, which provides a bearing platform. The bearing plate 2 carries a fan 3, a liquid storage tank 4, a laser radar 7 and an electronic control system 8. The fan 3 is used to provide wind speed during spraying.

[0091] The liquid storage tank 4 is connected to a spray head 5 . The liquid storage tank 4 stores liquid to be sprayed, and the spray head 5 sprays the liquid stored in the liquid storage tank 4 .

[0092] The electronic control system 8 is respectively connected to the laser radar 7, the nozzle 5, the fan 3, the moving wheel 1 and the data acquisition module 6. The laser radar 7 collects the canopy point cloud data of the plants to be sprayed in the forward direction of the sprayer, and transmits the collected canopy point cloud data to the electronic control system 8. The electronic control system 8 executes any of the above-mentioned spray parameter acquisition methods based on the density of the plant canopy according to the canopy point cloud data of the plants to be sprayed, and outputs the spray parameters corresponding to each canopy area of ​​the plants to be sprayed.

[0093] The data acquisition module 6 collects the current rotation speed of the fan 3, the current driving speed of the moving wheel 1, and the current spray flow rate of the nozzle 5. The electronic control system 8 obtains the current rotation speed of the fan 3, the current driving speed of the moving wheel 1, and the current spray flow rate of the nozzle 5 from the data acquisition module 6.

[0094] The electronic control system 8 automatically adjusts the spray flow rate of the nozzle 5, the rotation speed of the fan 3 and the driving speed of the moving wheel 1 according to the current rotation speed of the fan 3, the current driving speed of the moving wheel 1, the current spray flow rate of the nozzle 5 and the spray parameters corresponding to each canopy area of ​​the plant to be sprayed, so that the adjusted wind speed, driving speed and spray flow rate meet the requirements of the spray parameters corresponding to each canopy area of ​​the plant to be sprayed.

[0095] The sprayer provided in the embodiment of the present invention collects the canopy point cloud data of the plants to be sprayed in the forward direction of the sprayer through the laser radar 7, and the data acquisition module 6 collects the current rotation speed of the fan 3, the current driving speed of the moving wheel 1 and the current spray flow rate of the nozzle 5. The electronic control system 8 executes the spray parameter acquisition method based on the density of the plant canopy according to the canopy point cloud data of the plants to be sprayed, and outputs the spray parameters corresponding to each canopy area of ​​the plants to be sprayed. Then, according to the current rotation speed of the fan 3, the current driving speed of the moving wheel 1, the current spray flow rate of the nozzle 5 and the spray parameters corresponding to each canopy area of ​​the plants to be sprayed, the spray flow rate of the nozzle 5, the rotation speed of the fan 3 and the driving speed of the moving wheel 1 are automatically adjusted. Variable spraying is achieved by adjusting the spray parameters, thereby improving the spraying accuracy and efficiency.

[0096] In some embodiments, the data acquisition module 6 includes: an encoder 61, a unit sensor 62 and a flow sensor 63; The encoder 61 is embedded in the connecting shaft between the fan 3 and the motor 9, the unit sensor 62 is arranged on the electronic control system 8, the flow sensor 63 is embedded in the infusion tube of the nozzle 5, and the electronic control system 8 is connected to the encoder 61, the unit sensor 62 and the flow sensor 63 respectively; The motor 9 is used to provide power to the fan 3; The encoder 61 is used to collect the current rotation speed of the fan 3; The unit sensor 62 is used to collect the current travel speed of the moving wheel 1; The flow sensor 63 is used to collect the current spray flow of the nozzle 5 .

[0097] Specifically, Figure 5 As shown, the data acquisition module 6 includes an encoder 61 , a unit sensor 62 and a flow sensor 63 .

[0098] The encoder 61 is embedded in the connecting shaft between the fan 3 and the motor 9 . The motor 9 provides power to the fan 3 . The encoder 61 collects the current rotation speed of the fan 3 and transmits the collected current rotation speed of the fan 3 to the electronic control system 8 .

[0099] The unit sensor 62 is arranged beside the electronic control system 8, and the unit sensor 62 collects the current travel speed of the moving wheel 1, and transmits the collected current travel speed of the moving wheel 1 to the electronic control system 8. For example, the unit sensor 62 is a GPS locator, and the current travel distance and current travel time of the sprayer are obtained through the GPS locator, thereby obtaining the current travel speed of the moving wheel 1.

[0100] The flow sensor 63 is nested in the infusion tube of the nozzle 5 . The flow sensor 63 collects the current spray flow of the nozzle 5 and transmits the collected current spray flow of the nozzle 5 to the electronic control system 8 .

[0101] The sprayer provided in the embodiment of the present invention collects the current rotation speed of the fan 3 through the encoder 61, the unit sensor 62 collects the current driving speed of the moving wheel 1, and the flow sensor 63 collects the current spray flow of the nozzle 5, so as to realize real-time monitoring of the fan 3, the moving wheel 1 and the nozzle 5, and transmits the collected current rotation speed of the fan 3, the current driving speed of the moving wheel 1 and the current spray flow of the nozzle 5 to the electronic control system 8, so that the electronic control system 8 can regulate the fan 3, the moving wheel 1 and the nozzle 5.

[0102] In some embodiments, the electronic control system 8 includes: a computer 81, a flow controller 82, a wind speed controller 83 and a speed controller 84; The flow controller 82 is connected to the flow sensor 63, the nozzle 5 and the computer 8 respectively; the wind speed controller 83 is connected to the encoder 61, the fan 3 and the computer 81 respectively; the speed controller 84 is connected to the unit sensor 62, the moving wheel 1 and the computer 81 respectively; The computer 81 is used to execute any of the above-mentioned spray parameter acquisition methods based on plant canopy density according to the canopy point cloud data of the plant to be sprayed, and output the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; The flow controller 82 is used to automatically adjust the spray flow of the nozzle 5 according to the current spray flow of the nozzle 5 and the spray flow in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed; The wind speed controller 83 is used to automatically adjust the speed of the fan 3 according to the current speed of the fan 3 and the wind speed in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed; The speed controller 84 is used to automatically adjust the travel speed of the moving wheel 1 according to the current travel speed of the moving wheel 1 and the travel speed in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed.

[0103] Specifically, Figure 6 It is a schematic diagram of the structure of the electric control system provided by the present invention, such as Figure 6 As shown, the electronic control system 8 includes a computer 81 , a flow controller 82 , a wind speed controller 83 and a speed controller 84 .

[0104] The flow controller 82 is connected to the flow sensor 63, the nozzle 5 and the computer 81 respectively, and the flow controller 82 obtains the current spray flow of the nozzle 5 from the flow sensor 63, and obtains the spray parameters corresponding to each canopy area of ​​the plant to be sprayed from the computer 81. The flow controller 82 compares the current spray flow of the nozzle 5 with the spray flow in the spray parameters corresponding to each canopy area of ​​the plant to be sprayed, and automatically adjusts the spray flow of the nozzle 5 according to the difference between the two, for example, by automatically adjusting the size and number of the nozzle holes of the nozzle 5 to achieve automatic adjustment of the spray flow of the nozzle 5, so that the current spray flow of the nozzle 5 after adjustment is the spray flow in the spray parameters corresponding to each canopy area of ​​the plant to be sprayed, so as to achieve closed-loop feedback control of the spray flow and ensure the adjustment accuracy of the spray flow.

[0105] The wind speed controller 83 is connected to the encoder 61, the fan 3 and the computer 81 respectively, and the wind speed controller 83 obtains the current rotation speed of the fan 3 from the encoder 61, and obtains the spray parameters corresponding to each canopy area of ​​the plant to be sprayed from the computer 81. The wind speed controller 83 determines the current wind speed according to the current rotation speed of the fan 3, compares the current wind speed with the wind speed in the spray parameters corresponding to each canopy area of ​​the plant to be sprayed, and automatically adjusts the rotation speed of the fan 3 according to the difference between the two, for example, indirectly adjusts the rotation speed of the fan 3 by adjusting the power provided by the motor 9, so that the current wind speed provided by the adjusted fan 3 is the wind speed in the spray parameters corresponding to each canopy area of ​​the plant to be sprayed, thereby realizing closed-loop feedback control of the wind speed and ensuring the adjustment accuracy of the wind speed.

[0106] The speed controller 84 is connected to the unit sensor 62, the moving wheel 1 and the computer 81 respectively. The speed controller 84 obtains the current driving speed of the moving wheel 1 from the unit sensor 62, and obtains the spray parameters corresponding to each canopy area of ​​the plant to be sprayed from the computer 81. The speed controller 84 compares the current driving speed of the moving wheel 1 with the driving speed in the spray parameters corresponding to each canopy area of ​​the plant to be sprayed, and automatically adjusts the driving speed of the moving wheel 1 according to the difference between the two, so that the current driving speed of the moving wheel 1 after adjustment is the driving speed in the spray parameters corresponding to the plant to be sprayed, thereby realizing closed-loop feedback control of the driving speed and ensuring the adjustment accuracy of the driving speed.

[0107] The sprayer provided by the embodiment of the present invention automatically adjusts the spray flow of the nozzle 5 according to the current spray flow of the nozzle 5 and the spray flow in the spray parameters corresponding to each canopy area of ​​the plant to be sprayed through the flow controller 82, automatically adjusts the rotation speed of the fan 3 according to the current rotation speed of the fan 3 and the wind speed in the spray parameters corresponding to each canopy area of ​​the plant to be sprayed through the wind speed controller 83, and automatically adjusts the driving speed of the moving wheel 1 according to the current driving speed of the moving wheel 1 and the driving speed in the spray parameters corresponding to each canopy area of ​​the plant to be sprayed through the speed controller 84, thereby realizing closed-loop feedback control of the spray parameters, ensuring the adjustment accuracy of the spray parameters, and further improving the spraying accuracy and efficiency.

[0108] 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 method for obtaining spray parameters based on plant canopy density, characterized in that: include: Determining the number of point clouds of the plants to be sprayed in multiple canopy areas according to the canopy point cloud data of the plants to be sprayed; Determining the leaf area index of each canopy area of ​​the plant to be sprayed according to the number of point clouds of the plant to be sprayed in multiple canopy areas; Determining the canopy density of each canopy area of ​​the plant to be sprayed according to the leaf area index of each canopy area of ​​the plant to be sprayed; The canopy density of each canopy area of ​​the plant to be sprayed is input into the constructed correlation model to obtain the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; the correlation model is used to characterize the correlation between the canopy density and the spray parameters.

2. The method for obtaining spray parameters based on plant canopy density according to claim 1, characterized in that: The step of determining the number of point clouds of the plants to be sprayed in multiple canopy areas according to the canopy point cloud data of the plants to be sprayed comprises: Dividing the canopy of the plant to be sprayed into a plurality of canopy areas; Acquiring point cloud data of the plurality of canopy regions according to the canopy point cloud data of the plants to be sprayed; Divide the point cloud data of each canopy area into multiple voxels and calculate the number of point clouds in each voxel; The number of point clouds of the plants to be sprayed in the multiple canopy areas is determined according to the number of point clouds in each voxel and the multiple voxels corresponding to each canopy area.

3. The method for obtaining spray parameters based on plant canopy density according to claim 2, characterized in that: Before acquiring the point cloud data of the plurality of canopy regions according to the canopy point cloud data of the plant to be sprayed, the method further includes: Based on the Euclidean distance clustering algorithm, the canopy point cloud data of the plant group is clustered to obtain the cluster center point of each plant; A cylindrical space model is established with the cluster center point of each plant as the center; The cylindrical space model is used to segment the canopy point cloud data of the plant group to obtain the canopy point cloud data of each individual plant.

4. The method for obtaining spray parameters based on plant canopy density according to claim 3, characterized in that: Before clustering the canopy point cloud data of the plant group based on the Euclidean distance clustering algorithm to obtain the cluster center point of each plant, the following steps are also included: Obtain initial point cloud data of vegetation canopy from multiple different measuring stations; Based on the target sphere fitting, the initial point cloud data is registered to obtain the first point cloud data; According to random sampling and curvature sampling, the first point cloud data is layered and simplified to obtain second point cloud data; Based on the HSI threshold and through filtering, the second point cloud data is filtered to obtain the canopy point cloud data of the plant population.

5. The method for obtaining spray parameters based on plant canopy density according to claim 4, characterized in that: The registering of the initial point cloud data based on the target sphere fitting to obtain the first point cloud data includes: Performing target sphere detection on the initial point cloud data and extracting the coordinates of the sphere center; Based on the spherical center coordinates and the ID of the target sphere, matching the same target sphere is performed in the initial point cloud data of different measuring stations; According to each matched target sphere, the corresponding relationship between the initial point cloud data of different measuring stations is established; According to the corresponding relationship, the transformation matrix between the initial point cloud data of different measuring stations is calculated; The transformation matrix is ​​registered with the initial point cloud data based on the ICP algorithm to obtain first point cloud data.

6. The method for obtaining spray parameters based on plant canopy density according to claim 4, characterized in that: The step of performing hierarchical simplification on the first point cloud data according to random sampling and curvature sampling to obtain second point cloud data includes: Constructing a rectangular bounding box surrounding the first point cloud data; Dividing the spatial structure of the rectangular bounding box into multiple levels; The first point cloud data in each layer is sampled by using random sampling and curvature sampling to obtain the second point cloud data.

7. The method for obtaining spray parameters based on plant canopy density according to claim 1, characterized in that: The spray parameters include: wind speed, driving speed and spray flow rate.

8. A spray machine, characterized in that: include: Moving wheels, carrying plates, fans, liquid storage tanks, nozzles, data acquisition modules, laser radars and electronic control systems; The moving wheel is installed under the bearing plate, and the bearing plate carries the fan, the liquid storage tank, the laser radar and the electronic control system; The liquid storage tank is connected to the nozzle, and the electronic control system is respectively connected to the laser radar, the nozzle, the fan, the moving wheel and the data acquisition module; The moving wheel is used to drive the sprayer to move; The bearing plate is used to provide a bearing platform; The fan is used to provide wind speed during spraying; The liquid storage tank is used to store the liquid to be sprayed; The spray head is used to spray the liquid; The data acquisition module is used to collect the current rotation speed of the fan, the current driving speed of the moving wheel and the current spray flow rate of the nozzle; The laser radar is used to collect canopy point cloud data of the plants to be sprayed in the forward direction of the sprayer; The electronic control system is used to execute the spray parameter acquisition method based on the plant canopy density as described in any one of claims 1 to 7 according to the canopy point cloud data of the plant to be sprayed, and output the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; and automatically adjust the spray flow rate of the nozzle, the rotation speed of the fan and the travel speed of the moving wheel according to the spray parameters corresponding to each canopy area of ​​the plant to be sprayed.

9. The spray machine according to claim 8, characterized in that The data acquisition module includes: an encoder, a unit sensor and a flow sensor; The encoder is nested in the connecting shaft between the fan and the motor, the unit sensor is arranged beside the electronic control system, the flow sensor is nested in the infusion tube of the nozzle, and the electronic control system is respectively connected to the encoder, the unit sensor and the flow sensor; The motor is used to provide power to the fan; The encoder is used to collect the current rotation speed of the fan; The unit sensor is used to collect the current travel speed of the moving wheel; The flow sensor is used to collect the current spray flow of the nozzle.

10. The spray machine according to claim 9, characterized in that The electronic control system includes: a computer, a flow controller, a wind speed controller and a speed controller; The flow controller is connected to the flow sensor, the nozzle and the computer respectively; the wind speed controller is connected to the encoder, the fan and the computer respectively; the speed controller is connected to the unit sensor, the moving wheel and the computer respectively; The computer is used to execute the method for obtaining spray parameters based on plant canopy density according to any one of claims 1 to 7 according to the canopy point cloud data of the plant to be sprayed, and output the spray parameters corresponding to each canopy area of ​​the plant to be sprayed; The flow controller is used to automatically adjust the spray flow of the nozzle according to the current spray flow of the nozzle and the spray flow in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed; The wind speed controller is used to automatically adjust the speed of the fan according to the current speed of the fan and the wind speed in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed; The speed controller is used to automatically adjust the travel speed of the moving wheel according to the current travel speed of the moving wheel and the travel speed in the spray parameter corresponding to each canopy area of ​​the plant to be sprayed.

Images