Growth-control drug spraying method, device, electronic equipment and storage medium

Abstract

The present application provides a method, device, electronic device and storage medium for spraying growth-controlling drugs, wherein the method for spraying growth-controlling drugs includes: determining the under-growing area of ​​crops in the target plot; converting the target plot into a target array according to the spraying width of the drone and the under-growing area of ​​crops; the target array includes multiple array points, the distance between each array point is equal to the spraying width of the drone, and the multiple array points include normal array points and abnormal array points, and the abnormal array points are array points in the under-growing area of ​​crops; based on the reinforcement learning model, a spraying path is determined for the drone in the target array, and the spraying path must pass through each normal array point, each normal array point is passed only once, and cannot pass through each abnormal array point; control the drone to spray the growth-controlling drugs on the crops in the target plot according to the spraying path. The above-mentioned spraying method can improve the accuracy of the drone spraying growth-controlling drugs and avoid excessive spraying.

Description

Technical Field

[0001] The present application relates to the field of smart agricultural technology, and more specifically, to a method, device, electronic device and storage medium for spraying growth-control drugs. Background Art

[0002] In modern agriculture, drones have become an important tool for improving production efficiency. Using drones to spray pesticides in farmland can effectively cover a wide area of ​​land, ensuring that every crop receives the necessary plant protection treatment. However, despite the convenience and efficiency brought by this technology, it also faces the challenge of precise control, especially in the application of spraying growth control drugs.

[0003] Growth control drugs are special chemicals used to regulate crop growth. Appropriate use can promote healthy growth of crops and increase yield and quality. However, the use of this drug requires very precise control, because excessive spraying may excessively inhibit crop growth and lead to crop yield reduction. Although drones can quickly cover large areas of farmland, the overlap of their spraying paths and insufficient control often lead to uneven or excessive spraying of drugs, especially under improper operation. Summary of the invention

[0004] The present application provides a method, device, electronic device and storage medium for spraying growth-controlling drugs, the purpose of which is to improve the accuracy of spraying growth-controlling drugs by unmanned aerial vehicles and avoid excessive spraying.

[0005] In a first aspect, an embodiment of the present application provides a method for spraying growth-control drugs, the method comprising:

[0006] Using the preset sensors and / or image analysis modules of the drone, the undergrown crop areas in the target plot are identified;

[0007] According to the spraying width of the drone and the crop undergrowth area, the target plot is converted into a target array; the target array includes a plurality of array points, the distance between each array point is equal to the spraying width of the drone, and the plurality of array points include normal array points and abnormal array points, and the abnormal array points are array points in the crop undergrowth area;

[0008] Based on the reinforcement learning model, a spraying path is determined for the drone in the target array, and the spraying path must pass through each normal array point, each normal array point is passed only once, and cannot pass through each abnormal array point;

[0009] The drone is controlled to spray growth control drugs on the crops in the target plot according to the spraying path.

[0010] Optionally, only one crop is planted in the target plot, and the method of using a preset sensor and / or an image analysis module of the drone to determine the undergrown area of ​​the crop in the target plot includes:

[0011] Controlling the drone to traverse each position of the target plot, and using a preset sensor and / or an image analysis module of the drone to identify a first distance at each position, where the first distance is the distance between the top of the crop at each position and the drone;

[0012] According to the first distance at each position, an undergrown crop area in the target plot is determined.

[0013] Optionally, determining the undergrown crop area in the target plot according to the first distance at each position includes:

[0014] Calculate the mean and standard deviation of multiple first distances;

[0015] determining a first threshold value according to the standard deviation;

[0016] For the first distance at each position, if the first distance is smaller than the average value, the difference between the first distance and the average value is calculated; if the difference is larger than the first threshold, it is determined that the position belongs to an undergrown crop area.

[0017] Optionally, at least two crops are planted in the target plot, and the method of using a preset sensor and / or an image analysis module of the drone to determine the undergrown area of ​​the crops in the target plot includes:

[0018] Controlling the drone to traverse each position of the target plot, and using the image analysis module of the drone to identify the type of crops at each position, and using a preset sensor and / or image analysis module of the drone to identify a first distance at each position, wherein the first distance is the distance between the top of the crops at each position and the drone;

[0019] According to the crop type and the first distance at each position, an undergrown crop area in the target plot is determined.

[0020] Optionally, determining the undergrown crop area in the target plot according to the crop type and the first distance at each position includes:

[0021] For each crop type, calculating an average value and a standard deviation of a plurality of first distances corresponding to the crop type;

[0022] determining a first threshold value according to the standard deviation;

[0023] For each first distance corresponding to the crop type, if the first distance is less than the average value, the difference between the first distance and the average value is calculated; if the difference is greater than the first threshold, it is determined that the corresponding position of the first distance belongs to the crop undergrowth area.

[0024] Optionally, converting the target plot into a target array according to the spraying width of the drone and the crop undergrowth area includes:

[0025] According to the spraying width of the drone and the length and width of the target plot, a plurality of array points are determined, wherein the distance between every two adjacent array points is equal to the spraying width;

[0026] For each position belonging to the crop undergrowth area, an array point closest to the position is determined as an abnormal array point, and the remaining array points except the abnormal array point are determined as normal array points, so as to obtain a target array.

[0027] Optionally, before determining a spraying path for the drone in the target array based on the reinforcement learning model, the method further includes:

[0028] A bottleneck sub-region is identified from the target array; the bottleneck sub-region refers to an area that the drone can only enter but cannot exit when spraying according to the requirement that the drone must pass through each normal array point, each normal array point only once, and cannot pass through each abnormal array point;

[0029] If there are multiple identified bottleneck sub-regions, a target bottleneck sub-region closest to the boundary of the target plot is determined from the multiple bottleneck sub-regions, and the remaining bottleneck sub-regions are adjusted to non-bottleneck sub-regions, thereby obtaining an adjusted target array.

[0030] A second aspect of the present application provides a control growth drug spraying device, the device comprising:

[0031] An area determination module is used to determine the undergrown area of ​​crops in the target plot by using the preset sensor and / or image analysis module of the drone;

[0032] An array conversion module is used to convert the target plot into a target array according to the spraying width of the drone and the crop undergrowth area; the target array includes a plurality of array points, the distance between each array point is equal to the spraying width of the drone, and the plurality of array points include normal array points and abnormal array points, and the abnormal array points are array points in the crop undergrowth area;

[0033] A path determination module, for determining a spraying path for the UAV in the target array based on a reinforcement learning model, wherein the spraying path must pass through each normal array point, each normal array point is passed only once, and cannot pass through each abnormal array point;

[0034] The spraying control module is used to control the UAV to spray the growth control drugs on the crops in the target plot according to the spraying path.

[0035] A third aspect of an embodiment of the present application provides an electronic device, the electronic device comprising a memory and a processor;

[0036] The memory is used to store application programs;

[0037] The processor is used to run the application program stored in the memory to implement any of the above-mentioned methods for spraying growth-control drugs.

[0038] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores an application program, and when the application program is executed by a processor, it is used to implement any of the above-mentioned methods for spraying growth-control drugs.

[0039] The method for spraying control drugs provided by the present application is adopted. First, the undergrown area of ​​crops in the target plot is determined. Then, according to the spraying width of the drone and the undergrown area of ​​crops, the target plot is converted into a target array. The target array includes multiple array points. The distance between each array point is equal to the spraying width of the drone. The multiple array points include normal array points and abnormal array points. The abnormal array points are array points in the undergrown area of ​​crops. Then, based on the reinforcement learning model, a spraying path is determined for the drone in the target array. The spraying path must pass through each normal array point, each normal array point only passes through once, and cannot pass through each abnormal array point. Finally, the drone is controlled to spray the control drugs on the crops in the target plot according to the spraying path. In this way, on the one hand, it can avoid the drone spraying the control drugs on the undergrown area, causing the crops to further undergrowth. On the other hand, it can avoid the drone repeatedly spraying the control drugs on the crops that need to be controlled, resulting in excessive spraying of the control drugs. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0041] Figure 1It is a schematic diagram of the flow chart of the method for spraying growth-control drugs provided in one embodiment of the present application;

[0042] Figure 2 is a schematic diagram of generating a target array provided by an embodiment of the present application;

[0043] Figure 3 is a schematic diagram of a target array provided in one embodiment of the present application;

[0044] Figure 4 is a schematic diagram of the existence of multiple bottleneck sub-regions in the target array in this application;

[0045] Figure 5 is a schematic diagram of multiple positive samples provided by an embodiment of the present application;

[0046] Figure 6 is a schematic diagram of multiple negative samples provided by an embodiment of the present application;

[0047] Figure 7 It is a structural schematic diagram of a growth control drug spraying device provided in one embodiment of the present application;

[0048] Figure 8 It is a structural block diagram of an electronic device proposed in one embodiment of the present disclosure. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0050] In modern agriculture, drones have become an important tool for improving production efficiency. Using drones to spray pesticides in farmland can effectively cover a wide area of ​​land, ensuring that every crop receives the necessary plant protection treatment. However, despite the convenience and efficiency brought by this technology, it also faces the challenge of precise control, especially in the application of spraying growth control drugs.

[0051] Growth control drugs are special chemicals used to regulate crop growth. Appropriate use can promote healthy growth of crops and increase yield and quality. However, the use of this drug requires very precise control, because excessive spraying may excessively inhibit crop growth and lead to crop yield reduction. Although drones can quickly cover large areas of farmland, the overlap of their spraying paths and insufficient control often lead to uneven or excessive spraying of drugs, especially under improper operation.

[0052] In order to improve the accuracy of spraying control drugs by drones and avoid excessive spraying, the present application provides a control drug spraying method through the following embodiments. Figure 1 , Figure 1 1 is a flow chart of a method for spraying a growth-control drug provided in an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:

[0053] S110: Using the preset sensor and / or image analysis module of the drone, determine the undergrown crop area in the target plot.

[0054] Among them, the under-growing crop area refers to: the area in the target plot where the crop growth is weak. For example, corn is planted in the target plot, and the normal growing corn plants in the target plot are usually taller than 1.2 meters, but there are some areas in the target plot where the corn plants are generally lower than 0.8 meters, and these areas are the under-growing crop areas. Or, corn is planted in the target plot, and the normal growing corn plants in the target plot are usually taller than 1.2 meters, and the number of leaves is usually more than 10, but there are some areas in the target plot where the corn plants are generally lower than 0.8 meters, and the number of leaves is usually less than 7, then these areas are the under-growing crop areas.

[0055] In some scenarios, only one crop is planted in the target plot, such as only corn or only soybean. In this scenario, the above step S110 may include the following sub-steps:

[0056] S110-1: Control the drone to traverse each position of the target plot, and use the preset sensor and / or image analysis module of the drone to identify the first distance at each position, where the first distance is the distance between the top of the crop at each position and the drone.

[0057] Specifically, a coordinate system can be set for the target plot in the drone, and the drone can be controlled to traverse the entire target plot in a zigzag path. When the drone traverses the plot, it records the coordinates of the current position at a preset time interval (for example, every 1 second), and detects the distance between the crops under the drone and the drone through a distance sensor or a 3D camera, that is, the first distance. In this way, multiple sets of data can be obtained, each set of data including a position coordinate and a first distance.

[0058] S110 - 2 : Determine the undergrown crop area in the target plot according to the first distance at each position.

[0059] Specifically, firstly, the average value and standard deviation of multiple first distances are calculated; then the first threshold is determined according to the standard deviation, for example, the first threshold is set to 2 times the standard deviation, or to 3 times the standard deviation; then for the first distance at each position, if the first distance is less than the average value, the difference between the first distance and the average value is calculated, and if the difference is greater than the first threshold, it is determined that the position belongs to the crop undergrowth area. It should be noted that the above is only an optional method for determining the crop undergrowth area, and the present disclosure can also use other methods to determine the crop undergrowth area, such as determining the crop undergrowth area by a pre-trained image recognition algorithm.

[0060] In other scenarios, at least two crops are planted in the target plot, for example, soybeans and corn are planted in the target plot at the same time, and the soybean planting belt and the corn planting belt are adjacent to and parallel to each other. In this scenario, the above step S110 may include the following sub-steps:

[0061] S110-A: Control the drone to traverse each location of the target plot, and use the image analysis module of the drone to identify the type of crops at each location, and use the preset sensor and / or image analysis module of the drone to identify the first distance at each location, where the first distance is the distance between the top of the crop at each location and the drone.

[0062] Specifically, the image analysis module of the drone includes a crop recognition model, which is a neural network that can be used to identify the type of crops. In addition, a coordinate system can be set for the target plot in the drone first, and the drone can be controlled to traverse the entire target plot in a zigzag path. When the drone traverses the plot, it records the coordinates of the current position at preset time intervals (for example, every 1 second), and on the other hand, collects crop images under the drone, and inputs the collected crop images into the crop recognition model to determine the type of crops here. On the other hand, the distance between the crops under the drone and the drone is detected by a distance sensor or a 3D camera, that is, the first distance. In this way, multiple sets of data can be obtained, each set of data includes a position coordinate, a crop type information, and a first distance.

[0063] S110-B: Determine the undergrown crop area in the target plot according to the crop type and the first distance at each position.

[0064] Specifically, for each crop type, the average value and standard deviation of multiple first distances corresponding to the crop type are first calculated; then the first threshold is determined according to the standard deviation, for example, the first threshold is set to 2 times the standard deviation, or is set to 3 times the standard deviation; then for each first distance corresponding to the crop type, if the first distance is less than the average value, the difference between the first distance and the average value is calculated; if the difference is greater than the first threshold, it is determined that the corresponding position of the first distance belongs to the crop undergrowth area.

[0065] For ease of understanding, illustratively, it is assumed that soybeans and corn are planted in the target plot at the same time. In step S110-B, firstly, several groups of data with the crop type of corn are screened out from the multiple groups of data collected in step S110-A (for ease of explanation, each group of data screened out is referred to as the first data group below); for each first distance in each first data group, the average value and standard deviation of these first distances are calculated, and the first threshold is set to 3 times the standard deviation; then for the first distance in each first data group, if the first distance is less than the average value, the first distance is subtracted from the average value to obtain the difference between the two, and if the difference between the two is greater than the first threshold, it is determined that the position coordinates in the first data group belong to the crop undergrowth area, that is, the corn undergrowth area.

[0066] Then, several groups of data whose crop type is soybean are screened out from the multiple groups of data collected in step S110-B (for the sake of convenience, each group of screened data will be referred to as the second data group below); for each first distance in each second data group, the average value and standard deviation of these first distances are calculated, and the first threshold is set to 3 times the standard deviation; then for the first distance in each second data group, if the first distance is less than the average value, the first distance is subtracted from the average value to obtain the difference between the two; if the difference between the two is greater than the first threshold, it is determined that the position coordinates in the second data group belong to the crop undergrowth area, that is, the soybean undergrowth area.

[0067] S120: Convert the target plot into a target array according to the spraying width of the UAV and the crop undergrowth area; the target array includes a plurality of array points, the distance between each array point is equal to the spraying width of the UAV, and the plurality of array points include normal array points and abnormal array points, and the abnormal array points are array points in the crop undergrowth area.

[0068] In some specific embodiments, the target array can be obtained in the following manner: according to the spraying width of the UAV and the length and width of the target plot, a plurality of array points are determined, and the distance between every two adjacent array points is equal to the spraying width of the UAV; for each position belonging to the undergrown crop area, an array point closest to the position is determined as an abnormal array point, and the remaining array points except the abnormal array points are determined as normal array points, thereby obtaining the target array.

[0069] Specifically, first determine multiple array points based on the spraying width of the drone and the length and width of the target plot. For example, if the length of the target plot is 120 meters and the width is 75 meters, and the spraying width of the drone is 1.5 meters, then each row in the length direction has 81 array points, and the spacing between two adjacent array points is equal to 1.5 meters. Each column in the width direction has 51 array points, and the spacing between two adjacent array points is also equal to 1.5 meters. It should be noted that in this application, the distance between the array points is set to the spraying width of the drone. Its function is that when the drone sprays the growth control drug along the spraying path formed by the array points, the total spraying range can just cover the entire target plot (except for the undergrown area of ​​crops), and it will not cause repeated spraying, that is, it will not cause excessive spraying.

[0070] After multiple array points are determined, the specific attributes of each array point are determined according to the crop undergrowth area determined in step S110. Specifically, according to the crop undergrowth area determined in step S110, it is determined whether each array point is a normal array point or an abnormal array point.

[0071] For ease of understanding, illustratively, refer to Figure 2 and 3 , Figure 2 is a schematic diagram of generating a target array provided by an embodiment of the present application, Figure 3 is a schematic diagram of a target array provided in one embodiment of the present application. Figure 2 As shown, Figure 2 Each × in represents a location coordinate belonging to an undergrown crop area. Figure 2 Each ○ in represents an array point. For example, for each position coordinate ×, an array point ○ closest to the coordinate × can be determined as an abnormal array point. After multiple abnormal array points are determined, the remaining array points are regarded as normal array points. Finally, the following is obtained: Figure 3 As shown in the target array, each ○ in the target array represents a normal array point, and each × represents an abnormal array point.

[0072] S130: Based on the reinforcement learning model, a spraying path is determined for the drone in the target array. The spraying path must pass through each normal array point, each normal array point is passed only once, and cannot pass through each abnormal array point.

[0073] Specifically, the Deep Q-Network (DQN) can be used as a reinforcement learning model, and the target array can be used as the environment of the reinforcement learning model. The environment needs to be able to respond to the actions of the drone, update the state according to the actions of the drone, and return the corresponding rewards. The state s of the reinforcement learning model includes the following elements: 1) Drone position: represented by coordinates (x, y), where x ∈ [0, 75], y ∈ [0, 120] (assuming that the length of the target plot is 120 meters and the width is 75 meters); 2) Array point state: a specific 4131-dimensional vector, each dimension represents the state of an array point (normal or abnormal); 3) Passage history: also a 4131-dimensional vector, each dimension indicates whether the corresponding array point has been passed. The action space A of the reinforcement learning model includes: moving one grid in the positive y direction (one grid is the distance between adjacent array points), moving one grid in the negative y direction, moving one grid in the positive x direction, and moving one grid in the negative x direction. The reward function of the reinforcement learning model is: successfully moving to a normal array point that has not been passed: +10 points; moving to an abnormal array point or crossing the boundary: -100 points; repeating a normal array point that has been visited: -50 points. The discount factor γ of the reinforcement learning model is set to 0.95. After the reinforcement learning model is set in the above manner, the reinforcement learning model can be trained for a period of time. After the reinforcement learning model has the ability to plan the spraying path, the reinforcement learning model can be used in step S130 to determine a spraying path for the drone. The spraying path will pass through each normal array point, and each normal array point will only be passed once, and will not pass through each abnormal array point.

[0074] S140: Control the drone to spray growth control drugs on the crops in the target plot according to the spraying path.

[0075] By executing the above steps S110 to S140, on the one hand, it can avoid that the drone sprays growth-control drugs on the areas that are already undergrowth, causing the crops to further undergrowth; on the other hand, it can avoid that the drone repeatedly sprays growth-control drugs on the crops that need growth control, causing excessive spraying of growth-control drugs.

[0076] In some cases, such as Figure 4 As shown, there may be multiple bottleneck sub-regions in a target array ( Figure 4 Each dotted box in the figure is a bottleneck sub-area). The so-called bottleneck sub-area refers to the area where the drone can only enter but not exit when the drone is spraying according to the requirements that it must pass through each normal array point, each normal array point can only pass once, and cannot pass through each abnormal array point. In this case, it is impossible to plan a spraying path for the drone according to the above spraying requirements.

[0077] In order to deal with the above situation, in some specific embodiments, before executing the above step S130, the bottleneck sub-area can be first identified from the target array; if the number of identified bottleneck sub-areas is multiple, the target bottleneck sub-area closest to the boundary of the target plot is determined from the multiple bottleneck sub-areas, and the remaining bottleneck sub-areas are adjusted to non-bottleneck sub-areas, thereby obtaining the adjusted target array. Then the above step S130 is performed based on the adjusted target array. It should be noted that by adjusting the target array in the above manner, the reinforcement learning model can eventually determine the target bottleneck sub-area as the end point of the spraying path. Since the target bottleneck sub-area is closest to the boundary of the target plot, the drone can fly to the end point of the spraying path (i.e., the target bottleneck sub-area) and then fly out of the target plot in a shorter distance, avoiding the drone from performing long-distance ineffective flights, improving the drone's operating efficiency, and saving drone power or energy.

[0078] Specifically, a pre-trained deep convolutional neural network (or a target detection neural network such as Faster R-CNN) can be used to identify bottleneck sub-regions from the target array. When training the neural network, the following can be generated in advance: Figure 5 As shown in the figure, multiple positive samples are generated as follows Figure 6 Multiple negative samples are shown, where the positive sample refers to an array pattern including a bottleneck sub-region, and the negative sample refers to an array pattern not including a bottleneck sub-region; then supervised learning is performed on the neural network based on these positive samples and negative samples, so that the neural network has the ability to identify the bottleneck sub-region.

[0079] It should be noted that, in addition to using a pre-trained neural network to identify bottleneck sub-regions from a target array, other methods can also be used to identify bottleneck sub-regions. For example, after determining the target array, the drone or drone control terminal can display the target array to the user, and then receive the bottleneck sub-region position information manually input by the user, thereby obtaining the position and quantity information of the bottleneck sub-region in the target array.

[0080] When determining the distance between the bottleneck sub-region and the boundary of the target plot, the circumscribed convex polygon of the bottleneck sub-region may be determined first, and then the geometric center (or centroid) of the circumscribed convex polygon may be calculated, and then the distance between the geometric center (or centroid) and the four boundaries of the target plot may be calculated, and the minimum distance therebetween may be used as the distance between the bottleneck sub-region and the boundary of the target plot. In addition, when adjusting the remaining bottleneck sub-regions, for each of the remaining bottleneck sub-regions, each normal array point within the circumscribed convex polygon of the bottleneck sub-region may be adjusted to an abnormal array point, thereby adjusting the bottleneck sub-region to a non-bottleneck sub-region.

[0081] Based on the same inventive concept, the present application also provides a control drug spraying device, referring to Figure 7 , Figure 7 Schematic diagram of the structure of the control drug spraying device provided in one embodiment of the present application. Figure 7 As shown, the device comprises:

[0082] The area determination module 510 is used to determine the undergrown area of ​​crops in the target plot by using the preset sensor and / or image analysis module of the drone;

[0083] The array conversion module 520 is used to convert the target plot into a target array according to the spraying width of the drone and the crop undergrowth area; the target array includes a plurality of array points, the distance between each array point is equal to the spraying width of the drone, and the plurality of array points include normal array points and abnormal array points, and the abnormal array points are array points in the crop undergrowth area;

[0084] A path determination module 530 is used to determine a spraying path for the drone in the target array based on a reinforcement learning model, where the spraying path must pass through each normal array point, each normal array point is passed only once, and cannot pass through each abnormal array point;

[0085] The spraying control module 540 is used to control the UAV to spray growth control drugs on the crops in the target plot according to the spraying path.

[0086] In some specific embodiments, only one crop is planted in the target plot, and the area determination module 510 is specifically used to: control the drone to traverse each location of the target plot, and use the preset sensor and / or image analysis module of the drone to identify the first distance at each location, where the first distance is the distance between the top of the crop at each location and the drone; and determine the undergrown crop area in the target plot based on the first distance at each location.

[0087] In some specific embodiments, at least two crops are planted in the target plot, and the area determination module 510 is specifically used to: control the drone to traverse each location of the target plot, and use the drone's image analysis module to identify the type of crop at each location, and use the drone's preset sensor and / or image analysis module to identify the first distance at each location, where the first distance is the distance between the top of the crop at each location and the drone; determine the undergrown crop area in the target plot based on the crop type at each location and the first distance.

[0088] See also Figure 8 , Figure 86 is a structural block diagram of an electronic device proposed in an embodiment of the present disclosure, and the electronic device may be a drone, or may also be a smart phone, a computer, or other devices. The electronic device 600 includes a processor 610, a memory 620, and one or more application programs, wherein the one or more application programs are stored in the memory 620 and configured to be executed by the one or more processors 610, and the one or more programs are configured to execute the above-mentioned control drug spraying method.

[0089] The processor 610 may include one or more processing cores. The processor 610 uses various interfaces and lines to connect the various parts of the entire electronic device 600, and executes various functions and processes data of the electronic device 600 by running or executing instructions, programs, code sets or instruction sets stored in the memory 620, and calling data stored in the memory 620. Optionally, the processor 610 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 410 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 610, but may be implemented separately through a communication chip.

[0090] The memory 620 may include a random access memory (RAM) or a read-only memory (ROM). The memory 620 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 620 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc. The data storage area may also store data created by the electronic device 600 during use, etc.

[0091] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0092] 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 its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the pending claims of the invention.

Claims

1. A method for spraying drugs to control growth, characterized in that: The method comprises: Using the preset sensors and / or image analysis modules of the drone, the undergrown crop areas in the target plot are identified; According to the spraying width of the drone and the crop undergrowth area, the target plot is converted into a target array; the target array includes a plurality of array points, the distance between each array point is equal to the spraying width of the drone, and the plurality of array points include normal array points and abnormal array points, and the abnormal array points are array points in the crop undergrowth area; Based on the reinforcement learning model, a spraying path is determined for the drone in the target array, and the spraying path must pass through each normal array point, each normal array point is passed only once, and cannot pass through each abnormal array point; Controlling the drone to spray growth control drugs on the crops in the target plot according to the spraying path; Before determining a spraying path for the drone in the target array based on the reinforcement learning model, the method further includes: A bottleneck sub-region is identified from the target array; the bottleneck sub-region refers to an area that the drone can only enter but cannot exit when spraying according to the requirement that the drone must pass through each normal array point, each normal array point only once, and cannot pass through each abnormal array point; If there are multiple identified bottleneck sub-regions, a target bottleneck sub-region closest to the boundary of the target plot is determined from the multiple bottleneck sub-regions, and the remaining bottleneck sub-regions are adjusted to non-bottleneck sub-regions, thereby obtaining an adjusted target array.

2. The method according to claim 1, characterized in that The target plot is planted with only one crop, and the use of a preset sensor and / or an image analysis module of the drone to determine the undergrown area of ​​the crop in the target plot includes: Controlling the drone to traverse each position of the target plot, and using a preset sensor and / or an image analysis module of the drone to identify a first distance at each position, where the first distance is the distance between the top of the crop at each position and the drone; According to the first distance at each position, an undergrown crop area in the target plot is determined.

3. The method according to claim 2, characterized in that The step of determining the undergrown crop area in the target plot according to the first distance at each position includes: Calculate the mean and standard deviation of multiple first distances; determining a first threshold value according to the standard deviation; For the first distance at each position, if the first distance is smaller than the average value, the difference between the first distance and the average value is calculated; if the difference is larger than the first threshold, it is determined that the position belongs to an undergrown crop area.

4. The method according to claim 1, characterized in that: At least two crops are planted in the target plot, and the method of using the preset sensor and / or image analysis module of the drone to determine the undergrown area of ​​the crops in the target plot includes: Controlling the drone to traverse each position of the target plot, and using the image analysis module of the drone to identify the type of crops at each position, and using a preset sensor and / or image analysis module of the drone to identify a first distance at each position, wherein the first distance is the distance between the top of the crops at each position and the drone; According to the crop type and the first distance at each position, an undergrown crop area in the target plot is determined.

5. The method according to claim 4, characterized in that Determining the undergrown crop area in the target plot according to the crop type and the first distance at each position includes: For each crop type, calculating an average value and a standard deviation of a plurality of first distances corresponding to the crop type; determining a first threshold value according to the standard deviation; For each first distance corresponding to the crop type, if the first distance is less than the average value, the difference between the first distance and the average value is calculated; if the difference is greater than the first threshold, it is determined that the corresponding position of the first distance belongs to the crop undergrowth area.

6. The method according to claim 3 or 5, characterized in that: The step of converting the target plot into a target array according to the spraying width of the drone and the crop undergrowth area comprises: According to the spraying width of the drone and the length and width of the target plot, a plurality of array points are determined, wherein the distance between every two adjacent array points is equal to the spraying width; For each position belonging to the crop undergrowth area, an array point closest to the position is determined as an abnormal array point, and the remaining array points except the abnormal array point are determined as normal array points, so as to obtain a target array.

7. A device for spraying drugs to control growth, characterized in that: The device comprises: An area determination module is used to determine the undergrown area of ​​crops in the target plot by using the preset sensor and / or image analysis module of the drone; An array conversion module is used to convert the target plot into a target array according to the spraying width of the drone and the crop undergrowth area; the target array includes a plurality of array points, the distance between each array point is equal to the spraying width of the drone, and the plurality of array points include normal array points and abnormal array points, and the abnormal array points are array points in the crop undergrowth area; A path determination module, for determining a spraying path for the UAV in the target array based on a reinforcement learning model, wherein the spraying path must pass through each normal array point, each normal array point is passed only once, and cannot pass through each abnormal array point; The spraying control module is used to control the UAV to spray the growth control drugs on the crops in the target plot according to the spraying path.

8. An electronic device, comprising a memory and a processor; The memory is used to store application programs; The processor is used to run the application stored in the memory to implement the growth control drug spraying method according to any one of claims 1 to 7.

9. A computer-readable storage medium, wherein an application is stored in the computer-readable storage medium, and when the application is executed by a processor, it is used to implement the growth control drug spraying method according to any one of claims 1 to 7.

Images