Fertilization management and control system and method based on unmanned aerial vehicle

Abstract

The invention discloses a fertilization management and control system and method based on an unmanned aerial vehicle, relates to the technical field of unmanned aerial vehicle fertilization management and control, and solves the technical problems in the prior art that the management and control of the unmanned aerial vehicle fertilization process are inaccurate, effective absorption of fertilizers by crops cannot be ensured, and the fertilization precision of the unmanned aerial vehicle is difficult to ensure. When the fertilization time is determined, whether an abnormal rainfall interval exists in a rainfall change curve of a target area or not is recognized, if yes, it is judged that fertilization work can be completed at the time interval of the adjacent abnormal rainfall interval, and if yes, the starting time of the time interval serves as the fertilization time; or when the precipitation amount of the target area is not high and relatively gentle, and when the total precipitation amount reaches the set threshold value, taking the corresponding moment as the fertilization moment; when the fertilization moment is determined, the influence of rainwater on fertilizer absorption is comprehensively considered, the unmanned aerial vehicle is controlled to start fertilization at the fertilization moment, and it can be guaranteed that the fertilizer cannot be washed away and can be fully absorbed by crops.

Description

Technical Field

[0001] This application belongs to the field of agriculture and relates to the technology of drone fertilization control. Specifically, it is a fertilization control system and method based on drones. Background Art

[0002] Fertilization plays an important role in agricultural production and can increase the yield of crops. However, over-fertilization may lead to fertilizer loss and leakage, causing pollution to the soil and water bodies and posing potential hazards to the environment. Therefore, scientific and reasonable fertilization strategies are crucial for reducing agricultural non-point source pollution.

[0003] In the traditional fertilization process, the fertilization process mainly relies on the experience of fertilization personnel. With the development of drone technology, drones are playing an increasingly important role in agricultural production, especially in the field of crop fertilization. Currently, the main method of drone fertilization is to carry a specific amount or concentration of fertilizer by the drone at a set time and complete the fertilization process according to the pre-set fertilization route. Before fertilization, the fertilization time is still mainly determined manually, and a pre-set fertilization route is required. The fertilization time will affect the absorption of fertilizer by crops, and the drone will be affected by environmental factors during the fertilization process, which may lead to uneven fertilization. Therefore, the existing solutions are not only time-consuming and laborious but also result in poor overall fertilization effects.

[0004] This application provides a fertilization control system and method based on drones to solve the above technical problems. Summary of the Invention

[0005] This application aims to solve at least one of the technical problems existing in the prior art; for this purpose, this application proposes a fertilization control system and method based on drones, which are used to solve the technical problems that the control of the drone fertilization process in the prior art is inaccurate, it is impossible to ensure the effective absorption of fertilizer by crops, and it is difficult to ensure the fertilization accuracy of drones.

[0006] To achieve the above object, the first aspect of this application provides a fertilization control system based on drones, including a fertilization control module, and a data acquisition module and a drone connected thereto;

[0007] Data acquisition module: used to collect soil information and image data of the target area, and collect meteorological data through the meteorological platform; where the target area is the area that needs to be fertilized, the soil information is used to identify the element content in the soil, and the image data is used to identify the growth status of the crops;

[0008] Fertilization control module: used to determine the fertilizer deficiency information map of the target area according to the soil information and image data, and formulate a fertilization prescription according to the fertilizer deficiency information map; where the fertilization prescription is the fertilizer requirement information of each position in the target area; and,

[0009] It is used to determine the fertilization time and fertilization route according to meteorological data, and carry out fertilization according to the fertilization route at the fertilization time; during the fertilization process, the drone is controlled and adjusted according to the fertilization prescription.

[0010] Preferably, determining the fertilization time according to meteorological data includes:

[0011] Extract the precipitation at each moment from the meteorological data; use the moment as the independent variable and the precipitation as the dependent variable to fit and establish a precipitation change curve; among them, the interval between each moment is 10 minutes, 15 minutes or 30 minutes;

[0012] Determine whether there is an abnormal rainfall interval according to the precipitation change curve; if so, determine the fertilization time based on the abnormal rainfall interval; if not, determine the fertilization time based on the precipitation change curve.

[0013] Preferably, determining the fertilization time based on the abnormal rainfall interval includes:

[0014] Identify the corresponding abnormal rainfall interval in the precipitation change curve and calculate the time interval between adjacent abnormal rainfall intervals; among them, the precipitation in the abnormal rainfall interval will affect the fertilization effect;

[0015] When the meteorological data of the time interval meet the flight requirements of the drone and the time interval is greater than the time threshold, take the start time corresponding to the time interval as the fertilization time; among them, the time threshold is set according to the absorption efficiency of the crops to the fertilizer under the corresponding meteorological data.

[0016] Preferably, determining the fertilization time based on the precipitation change curve includes:

[0017] Set a rainfall threshold; among them, the rainfall threshold is used to improve the penetration and absorption efficiency of the fertilizer;

[0018] Mark the function of the precipitation change curve as F(t) and mark the rainfall threshold as YY; solve the formula Get t2 and take the moment corresponding to t2 as the fertilization time; among them, t1 is the moment when it starts to rain in the precipitation change curve.

[0019] Preferably, controlling and adjusting the drone according to the fertilization prescription during the fertilization process includes:

[0020] At the fertilization time, control the drone to start fertilization work according to the fertilization route; among them, the fertilization route includes the flight altitude, flight attitude and flight path of the drone;

[0021] When the drone flies a set distance according to the fertilization route, extract the fertilization images during the fertilization process within the set distance; analyze the fertilization images to determine whether it is necessary to adjust the fertilization route.

[0022] Preferably, determining whether to adjust the fertilization route includes:

[0023] Analyzing the fertilization image through a machine learning algorithm to extract the fertilization area in the fertilization image; wherein, the fertilization area is the effective ground area covered by the fertilizer during the drone fertilization process;

[0024] Comparing the fertilization area with the preset coverage area corresponding to the fertilization route to determine the area deviation information; adjusting the fertilization route according to the fertilization area and the area deviation information.

[0025] Preferably, extracting the fertilization area in the fertilization image includes:

[0026] Identifying the fertilizer coverage area in the fertilization image through a machine learning algorithm; wherein, the fertilizer coverage area is the ground area covered by the fertilizer during the drone fertilization process;

[0027] Identifying the invalid areas in the fertilizer coverage area according to the fertilization prescription, and removing the invalid areas from the fertilizer coverage area to obtain the fertilization area.

[0028] Preferably, determining the area deviation information includes:

[0029] Comparing the fertilization areas corresponding to all fertilization images with the preset coverage area, and calculating the deviation distance, deviation direction and drone height between the two; wherein, the deviation distance is the mean value or the maximum value;

[0030] Judging the data correlation between the deviation distance and the deviation direction and the meteorological data; when the data correlation is greater than the correlation threshold, integrating the deviation distance, the deviation direction and the drone height into the area deviation information.

[0031] Preferably, adjusting the fertilization route according to the fertilization area and the area deviation information includes:

[0032] Calculating the area boundary of the fertilization area;

[0033] Adjusting the fertilization route according to the area boundary and the area deviation information to obtain a new fertilization route; wherein, the fertilization area corresponding to the new fertilization route is greater than or equal to the target area.

[0034] The second aspect of the present application provides a fertilization control method based on a drone, including:

[0035] Collecting the soil information and image data of the target area, and collecting meteorological data through a meteorological platform; wherein, the target area is the area that needs to be fertilized, the soil information is used to identify the element content in the soil, and the image data is used to identify the growth state of the crops;

[0036] Determine the fertilizer deficiency information map of the target area based on soil information and image data, and formulate a fertilization prescription according to the fertilizer deficiency information map; wherein, the fertilization prescription is the fertilizer requirement information for each position in the target area.

[0037] Determine the fertilization time and fertilization route according to meteorological data, and perform fertilization according to the fertilization route at the fertilization time; during the fertilization process, control and adjust the drone according to the fertilization prescription.

[0038] Compared with the prior art, the beneficial effects of the present application are:

[0039] 1. The present application determines the fertilization time and fertilization route according to meteorological data, and performs fertilization according to the fertilization route at the fertilization time; when determining the fertilization time, by identifying whether there is an abnormal rainfall interval in the precipitation change curve of the target area, if there is, it is judged whether the time interval between adjacent abnormal rainfall intervals can complete the fertilization work, and if so, the start time of the time interval is used as the fertilization time; or when the precipitation in the target area is not high and relatively flat, when the total precipitation reaches the set threshold, the corresponding time is used as the fertilization time; when determining the fertilization time, the present application comprehensively considers the influence of rainwater on fertilizer absorption, and controls the drone to start fertilization work at the fertilization time, which can ensure that the fertilizer will not be washed away and can be fully absorbed by crops.

[0040] 2. The present application controls and adjusts the drone according to the fertilization prescription during the fertilization process. Specifically, after starting the drone fertilization at the fertilization time, the video stream within the set distance is collected by the drone, and the video stream is analyzed to determine whether the fertilization area conforms to the preset coverage area. If not, the fertilization route is adjusted according to the regional deviation information; the present application can automatically identify the influence of the climate environment on the fertilization effect through the video stream within the set distance, can automatically calibrate the fertilization route, and can effectively improve the fertilization accuracy. Description of the Drawings

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0042] Figure 1 It is a schematic diagram of the method steps of the fertilization control method in the embodiment of the present application;

[0043] Figure 2 It is a schematic diagram of the system principle of the fertilization control system in the embodiment of the present application;

[0044] Figure 3 It is a schematic diagram of the method steps of determining the fertilization time in the embodiment of the present application. Detailed implementation manners

[0045] The technical solutions of the present application will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0046] With the rapid development of UAV technology, UAVs have been applied in various fields. Due to the characteristics of rapid takeoff and landing and strong mobility of UAVs, they are also widely used in the agricultural field, mainly for fertilization. UAVs can carry various sensors and intelligent control systems, fertilize crops according to preset routes and fertilization amounts, and are not restricted by terrain during the fertilization process, nor will they cause damage to crops and soil.

[0047] Although UAVs have great advantages, currently when using UAVs for fertilization, the UAVs are only responsible for the specific fertilization process, and other processes are still determined by the staff. The staff needs to determine when to fertilize and what route to take for fertilization in combination with empirical data. This method is mainly based on experience and it is difficult to accurately quantify the fertilization effect. Moreover, UAVs are easily affected by various environmental factors during the fertilization process. Once the UAVs are affected and the fertilization effect is different from the expected one, the existing solutions are also difficult to adjust in a timely manner.

[0048] To solve the above problems, the present application provides a fertilization control system and method based on UAVs.

[0049] Please refer to Figure 1 - Figure 2 , an embodiment of the first aspect of the present application provides a fertilization control system based on UAVs, including a fertilization control module, as well as a data acquisition module and a UAV connected thereto;

[0050] Data acquisition module: used to collect soil information and image data of the target area, and collect meteorological data through a meteorological platform;

[0051] Fertilization control module: used to determine the fertilizer deficiency information map of the target area according to the soil information and image data, formulate a fertilization prescription according to the fertilizer deficiency information map; and, used to determine the fertilization time and fertilization route according to the meteorological data, and perform fertilization according to the fertilization route at the fertilization time; control and adjust the UAV according to the fertilization prescription during the fertilization process.

[0052] The fertilization control system of this embodiment mainly consists of two major modules: a fertilization control module and a data acquisition module. The data acquisition module is mainly responsible for collecting various data to assist the fertilization control module in determining the fertilization prescription and timely adjusting the fertilization process of the drone. The fertilization control module determines when the fertilization effect is the best, when there is a deviation in the drone fertilization, and how to make adjustments based on the received various data. At the same time, the fertilization control module also conducts data interaction with the drone responsible for fertilization to monitor and adjust the fertilization process of the drone. It should be noted that the drone in this application is an existing fertilization drone, and only the fertilization time and fertilization route of the existing fertilization drone are automatically adjusted through remote control.

[0053] The target area in this embodiment refers to the area that needs to be fertilized. If the target area is large, it can be divided, and then each divided area can be analyzed. The soil information is used to identify the element content in the soil, which can be measured by a soil nutrient analyzer. During the measurement process, soil samples at different positions in the target area are collected for analysis, and the analysis results are used for the subsequent formulation of the fertilization prescription. The image data is used to identify the status of the crops, which can be collected by remote sensing satellites or drones. The vegetation index of each position in the target area can be determined by analyzing the image data, and the growth status of the crops can be reflected by using the vegetation index of different bands. It should be emphasized that although this application takes crops as an example when explaining the technical solution, its ultimate goal is to improve the fertilization effect of the drone. Therefore, this technical solution is also applicable to other fields where drones can be used for fertilization.

[0054] Before fertilizing with the drone, it is necessary to collect the soil information and image data of the target area, determine the element content in the soil at each position using the soil information, determine the growth status of the crops using the image data, and combine the element content and the growth status of the crops in the target area to determine the fertilizer requirement information of the crops, so as to construct a fertilizer requirement information map of the target area. This fertilizer requirement information map represents the type and quantity of fertilizers required at each position in the target area, providing a data basis for drone fertilization.

[0055] Next, considering that the fertilization effect has a great relationship with meteorological data, if a heavy rainstorm occurs after or during fertilization, it may cause the fertilizer to be washed away, thus affecting the fertilization effect. Therefore, this embodiment first determines whether there will be extreme weather that affects the fertilization effect based on the collected meteorological data (prediction data obtained through the meteorological platform), and then determines when it is best to start fertilizing. After starting fertilization, due to environmental interference, the actual fertilization effect of the drone may be inconsistent with the expected fertilization effect, which will cause a deviation in the fertilization area. If fertilization continues along the original fertilization route, on the one hand, it will waste fertilizer, and on the other hand, it will cause the target area not to be fully fertilized, affecting the overall fertilization effect.

[0056] Before fertilization, it is necessary to determine which fertilizers are lacking in the soil and which fertilizers are lacking in the crops. The fact that the crops lack fertilizers does not necessarily mean that the soil lacks fertilizers. It may be that the crops cannot absorb or have not had time to absorb them. Therefore, it is necessary to combine soil information and image data for joint judgment. The judgment method can be referred to as follows:

[0057] 1. Determine the target area, then collect soil samples from the target area, analyze the soil samples through a soil nutrient analyzer to obtain the content of each element in each soil sample, and thus establish a distribution map of element content in the target area. The more soil samples, the more accurate the element content distribution map, but the collection of soil samples needs to be comprehensively considered in terms of work efficiency. Of course, soil nutrient analyzers or other soil analysis equipment can also be set at fixed positions, and data transmission can be completed through the Internet of Things to reduce the cost of soil nutrient analysis.

[0058] 2. Collect image data of the target area. This image data can be a vegetation index image. Collect the image data of the target area before fertilization, and input the image data of the target area and the element content distribution map into the trained artificial intelligence model after preprocessing to obtain a fertilizer requirement information map of the target area.

[0059] The training of the artificial intelligence model can be referred to the following steps:

[0060] Step 1, in the data preparation stage, measure the content of various elements at each of multiple positions in the farmland through a soil nutrient rapid test device; use the remote sensing sensor carried by the drone to obtain multiple vegetation index images of the crops; Step 2, in the data processing stage, obtain the element distribution map of each element in the farmland according to the content of various elements at each of multiple positions, and preprocess and merge the element distribution map and the vegetation index image into a multi-channel image; Step 3, in the model training stage, construct a decision-making model, and set a set of trainable parameters for each of the preset elements, and use the multi-channel image to train the decision-making model and multiple sets of trainable parameters. The decision-making model includes an encoder and a decoder.

[0061] The training and use of the artificial intelligence model can refer to the Chinese invention patent with the application number 2023113031863 and the name of a fertilization decision-making method and system based on multi-vegetation index images. This patent discloses using image data and a soil nutrient analyzer to determine the fertilizer requirement information map of plants, especially the model construction and data processing processes are disclosed in detail and will not be elaborated here.

[0062] The fertilization information map can be understood as the fertilizer information required by crops at each location in the target area, which may vary due to different locations. Since the fertilizer information required by each crop in the same season is significantly different, the target area can be divided secondarily according to the crop type to improve the fertilization efficiency of the drone and reduce the fertilization difficulty. It should be noted that the fertilizer requirement information map calculates the fertilizer information required at each location in the target area through soil information and image data, but the fertilizer requirement information map cannot be used as the fertilization prescription for the drone. Therefore, there is a certain loss in the conversion process between the two. For example, if 100 units are required, considering factors such as fertilizer absorption, the actual fertilization amount has to exceed 100 units. This conversion relationship can be established by analyzing historical data, such as testing and obtaining the relationship between the actual fertilization amount and the actual absorption amount of the same type of land and crops. Of course, when the difference between the two is not significant, the fertilizer requirement information map can be directly converted into a fertilization prescription. The fertilization prescription is the fertilizer requirement information at each location in the target area, that is, the type of fertilizer and the fertilization amount required at each location determined according to the fertilizer requirement information map.

[0063] After determining the fertilization prescription, the fertilization amount and fertilization type that the drone needs to apply to each location in the target area can be clarified. During the fertilization process of crops, the absorption of fertilizers also needs to be focused on. The absorption of fertilizers is mainly related to the soil humidity. When the soil has a certain humidity, it can accelerate the absorption of fertilizers by the soil, thereby improving the absorption rate of fertilizers by crops and realizing the efficient conversion of fertilizers. The main method to increase soil humidity is natural humidification, that is, to increase the soil moisture content through rainfall. However, if fertilizers are applied before rainfall, the fertilizers that have not been infiltrated in time may be washed away by the upcoming rain, resulting in a significant reduction in the fertilizer absorption efficiency; if fertilizers are applied during rainfall, the fertilizers will also be washed away if the rainfall is large; and if fertilizers are applied after rainfall, if the rainfall lasts for a long time, fertilization needs to wait. After the rainfall ends, when the staff issue a fertilization instruction, continuous intervention by the staff is still required. Therefore, determining the fertilization time is also very important for drone fertilization.

[0064] Please refer to Figure 3 , the method for determining the fertilization time provided in this example can be referred to as follows: Extract the precipitation amount at each moment from the meteorological data; Fit and establish a precipitation change curve with the moment as the independent variable and the precipitation amount as the dependent variable;

[0065] Determine whether there is an abnormal rainfall interval according to the precipitation change curve; If yes, determine the fertilization time based on the abnormal rainfall interval; If no, determine the fertilization time based on the precipitation change curve.

[0066] The meteorological data is obtained through a third-party meteorological platform, and the interval between each moment is 10 minutes, 15 minutes or 30 minutes. The precipitation change curve is actually a curve of the precipitation amount changing with time, which can be used to predict the total precipitation in a certain future period.

[0067] Based on the precipitation change curve, it can be judged whether there is an abnormal rainfall interval in the future period. If there is no abnormal rainfall interval, it indicates that the precipitation is small in the future period, and fertilization can be carried out on the target area during the precipitation process. If there is an abnormal rainfall interval, it is necessary to avoid fertilization during the abnormal rainfall interval to prevent the fertilizer from being washed away.

[0068] Determining the fertilization time based on the abnormal rainfall interval includes:

[0069] Identifying the corresponding abnormal rainfall interval in the precipitation change curve and calculating the time interval between adjacent abnormal rainfall intervals; when the meteorological data of the time interval meet the requirements for the drone to fly and the time interval is greater than the time threshold, the start time corresponding to the time interval is used as the fertilization time.

[0070] The precipitation in the abnormal rainfall interval will affect the fertilization effect. The main impact is that it will wash away the fertilizer (including solid and liquid) from the fertilization target position, which may lead to a large deviation between the fertilization amount at each position in the target area and the preset fertilization amount, and may even cause the fertilizer to be washed away from the target area. The time threshold is set according to the absorption efficiency of the crops for the fertilizer under the corresponding meteorological data, that is, after the drone fertilizes, it can be effectively absorbed after the time corresponding to the time threshold, and the increase in rainfall does not affect the current fertilization effect.

[0071] The precipitation change curve is a continuous curve, which may include the rainfall conditions for several days. Rainfall, especially heavy rain, is very likely to occur intermittently. If the fertilizer can be effectively absorbed after one precipitation ends and before another precipitation, fertilization between these two precipitations does not affect the absorption effect of the fertilizer.

[0072] The determination of the abnormal rainfall interval can start from the peak value of the precipitation change curve. Counting from the peak value of the precipitation to both sides, the total precipitation of a precipitation curve is obtained. If the total precipitation of this interval exceeds the set precipitation threshold, it is determined that the fertilizer will be washed away (or mostly washed away) in this interval, and this interval can be defined as the abnormal rainfall interval. It should be noted that the precipitation threshold is mainly used to judge whether the fertilizer will be washed away, so it is related to the fertilizer type, crop type and soil type, and can be set after experiments.

[0073] The absorption of fertilizers requires a process. If fertilization is carried out after the end of the previous abnormal rainfall interval, after the fertilizer falls on the land, it needs an absorption time. For example, after 24 hours or 48 hours, if the next abnormal rainfall interval arrives within this absorption time, the fertilizer will still be washed away, affecting the absorption of the fertilizer. Based on this, a time threshold can be set. Therefore, whether the gap between two abnormal rainfall intervals is suitable for fertilization also needs to determine whether the corresponding time interval is greater than the set time threshold. If the time interval is greater than the time threshold, it means that fertilization is carried out in time within this time period, and the fertilizer can be well absorbed before the next abnormal rainfall interval arrives.

[0074] Of course, if there will be no other abnormal rainfall intervals after the end of the last abnormal rainfall interval in the precipitation change curve, then the end moment of the last abnormal rainfall interval is taken as the fertilization moment.

[0075] For some crops, if fertilization is carried out immediately after the precipitation ends, the accumulated rainwater may affect the root respiration of the crops. For this situation, a certain buffer time can be set, and fertilization is carried out after this buffer time. In some other preferred embodiments, when judging whether the time interval of the abnormal rainfall interval can be fertilized, the following formula can be used for judgment: buffer time + time interval ≥ time threshold, which can better ensure the absorption of fertilizers.

[0076] Determining the fertilization moment based on the precipitation change curve includes:

[0077] Setting a rainfall threshold; marking the function of the precipitation change curve as F(t), and marking the rainfall threshold as YY; solving the formula to obtain t2, and taking the moment corresponding to t2 as the fertilization moment.

[0078] The rainfall threshold is used to improve the penetration and absorption efficiency of fertilizers. If the precipitation reaches the rainfall threshold, the fertilizer can penetrate well into the soil so that the crops can absorb it better. In the above formula, t1 is the moment when it starts to rain in the precipitation change curve and is a determined value; if the total precipitation from moment t1 to moment t2 is greater than the rainfall threshold, it means that fertilization at this time can ensure the rapid absorption of fertilizers. Therefore, moment t2 can be taken as the fertilization moment. Of course, the buffer time mentioned above can also be taken into account and the fertilization moment can be appropriately postponed.

[0079] It should be noted that when determining the fertilization moment based on the precipitation change curve, if there are no abnormal rainfall moments in the precipitation change curve, it can be understood as continuous gentle rain. Therefore, after determining the fertilization moment, the drone fertilizes at the fertilization moment, and subsequent precipitation will not have a great impact on the absorption of fertilizers, but some losses are understandable.

[0080] After determining the fertilization prescription and fertilization time, the drone can be controlled to carry out fertilization work according to the preset fertilization route. The fertilization route can be planned comprehensively based on the size of the target area, the type of fertilizer, and the type of crops. To ensure the effective absorption of fertilizers by crops, the fertilization time is determined based on precipitation. However, precipitation is accompanied by variable weather conditions, such as changes in wind force, air pressure, etc. If the preset fertilization route is still followed in different meteorological environments, the expected fertilization effect may not be achieved. Therefore, it is necessary to control and adjust the drone in a timely manner during the fertilization process.

[0081] During the fertilization process, the drone is controlled and adjusted according to the fertilization prescription, including:

[0082] At the fertilization time, the drone is controlled to start fertilization work according to the fertilization route; when the drone flies a set distance along the fertilization route, the fertilization images during the fertilization process within the set distance are extracted; the fertilization images are analyzed to determine whether the fertilization route needs to be adjusted.

[0083] The fertilization route includes the flight altitude, flight attitude, and flight path of the drone. The fertilization route is preferably planned before fertilization. If it is planned during the fertilization process, the difficulty will increase, and the influence of the meteorological environment on the fertilization of the drone can be predicted. Therefore, in this embodiment, the fertilization process of the drone is regulated by adjusting the fertilization route later.

[0084] When the fertilization route, fertilization time, and fertilization prescription are clear, the drone is controlled to fertilize according to the fertilization route at the fertilization time. After flying a set distance along the fertilization route, the video stream is collected through the high-definition camera carried by the drone, and then the image data in the video stream is analyzed to determine whether the meteorological environment has an impact on the fertilization process. It should be noted that the set distance can be set to a fixed value, such as 10 meters, 20 meters, etc. The main purpose is to provide enough distance to collect the video stream.

[0085] After the video stream is collected within the set distance, the frame images can be extracted from the video stream, and it is judged whether the fertilizer is spread in the preset area after fertilization by the drone according to the frame images. If there is a deviation, it is very likely that it is affected by the meteorological environment. The specific reason for "very likely" will be described in detail later. To determine whether the fertilization route needs to be adjusted, the following steps can be referred to:

[0086] The fertilization images are analyzed by machine learning algorithms to extract the fertilization areas in the fertilization images; the fertilization areas are compared with the preset coverage areas corresponding to the fertilization route to determine the area deviation information; the fertilization route is adjusted according to the fertilization areas and the area deviation information.

[0087] The machine learning algorithm analyzes the fertilization image, which actually extracts the fertilizer coverage area from the fertilization image. Existing image processing algorithms can achieve feature extraction, and the construction, training, and types of machine learning algorithms will not be elaborated here. The fertilization area in this embodiment is the effective ground area covered by the fertilizer during the drone fertilization process. First, the fertilizer coverage area is extracted from the frame images of the video stream, that is, the ground area covered by the fertilizer after the drone fertilization. However, due to the influence of soil moisture, fertilizer type, crop variety, etc., if there is less fertilizer on the ground, it may cause the crops to be unable to absorb sufficient fertilizer, resulting in a reduction in the fertilization effect in this area. For this, the fertilization amount at each position can be obtained according to the fertilization prescription, and this fertilization amount is compared with the fertilization amount at each position in the fertilizer coverage area. The area with insufficient fertilization amount is regarded as an invalid area. The invalid area is excluded from the fertilizer coverage area to obtain the fertilization area, that is, the effective ground area covered by the fertilizer. The amount of fertilization can be realized by identifying the concentration or quantity of the fertilizer in the image, and it can be identified through the image data of the fertilizer falling above the crops.

[0088] In some other preferred embodiments, if the fertilization amount on the ground cannot be identified through image data, it can be detected manually and the result is fed back to the fertilization control module. Of course, the fertilizer coverage area can also be directly used as the fertilization area without considering the influence of the fertilization amount.

[0089] The deviation considered in this example during the fertilization process is mainly caused by meteorological data (meteorological environment). Therefore, the deviation brought in a small range is consistent, that is, the offset distance, offset direction, etc. of the fertilizer from the drone to the ground are consistent. Therefore, the following steps can be referred to for determining the regional deviation information:

[0090] Compare the fertilization areas corresponding to all fertilization images with the preset coverage area, and calculate the deviation distance, deviation direction, and drone height between the two; judge the data correlation degree of the deviation distance and deviation direction with the meteorological data; when the data correlation degree is greater than the correlation threshold, integrate the deviation distance, deviation direction, and drone height into the regional deviation information.

[0091] During the drone fertilization process, if there is a deviation between the actual fertilizer coverage area and the preset coverage area, it may also be caused by improper setting of the fertilization equipment of the drone. Therefore, the influence of the fertilization equipment is excluded through the data correlation degree. The specific method is to judge whether the deviation direction and deviation distance have a high correlation with the meteorological data. If so, the influence of the fertilization equipment is excluded. If it is determined that it is affected by the meteorological data, the deviation distance, deviation direction between the fertilization area and the preset coverage area, and the corresponding drone height are integrated into the regional deviation information. It can be understood that the preset coverage area can be the fertilizer coverage area calculated according to the standard meteorological data (under the set wind force and set pressure) when planning the fertilization route.

[0092] After determining the fertilization area, calculate the area boundary of the fertilization area; adjust the fertilization route according to the area boundary and area deviation information to obtain a new fertilization route. There may be deviations not only in size but also in position between the preset coverage areas corresponding to the fertilization areas.

[0093] If there is a deviation in size, in order to ensure that the drone can cover the entire target area during fertilization, it is necessary to adjust the spacing between adjacent paths in the fertilization route to ensure that after fertilization along the new fertilization route, the effective coverage areas of the fertilizer can preferably overlap. If there is a deviation in position, then the entire fertilization course is translated according to the deviation distance and deviation direction in the area deviation information. The ultimate effect is also that the fertilization area corresponding to the adjusted fertilization route is responsible for the entire target area. When necessary, the flight altitude of the drone can also be adjusted to ensure that the effective coverage area of the fertilizer can cover the entire target area.

[0094] Exemplarily, if the fertilization area is the same size as the corresponding preset coverage area but is deviated 20 cm to the east in position, then translating the fertilization route 20 cm to the west can achieve correction.

[0095] If the fertilization area is different in size from the corresponding preset coverage area and the area boundary of the fertilization area deviates 20 cm from the preset coverage area along the deviation direction, then translating the fertilization route 20 cm in the opposite direction of the deviation direction can achieve correction.

[0096] An embodiment of the second aspect of the present application provides a method for fertilization control based on a drone, including:

[0097] Collect soil information and image data of the target area, and collect meteorological data through a meteorological platform; wherein, the target area is the area that needs to be fertilized, the soil information is used to identify the element content in the soil, and the image data is used to identify the growth state of the crops;

[0098] Determine the fertilizer deficiency information map of the target area according to the soil information and image data, and formulate a fertilization prescription according to the fertilizer deficiency information map; wherein, the fertilization prescription is the fertilizer requirement information for each position in the target area.

[0099] Determine the fertilization time and fertilization route according to the meteorological data, and perform fertilization according to the fertilization route at the fertilization time; control and adjust the drone according to the fertilization prescription during the fertilization process.

[0100] The above embodiments are only used to illustrate the technical method of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present application can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present application.

Claims

1. An unmanned aerial vehicle-based fertilization control system, characterized in that, It includes a fertilization control module, as well as a data acquisition module and a drone connected thereto; Data acquisition module: It is used to collect soil information and image data of the target area, and collect meteorological data through a meteorological platform; wherein, the target area is the area that needs to be fertilized, the soil information is used to identify the element content in the soil, and the image data is used to identify the growth status of crops; Fertilization control module: It is used to determine the fertilizer deficiency information map of the target area according to the soil information and image data, and formulate a fertilization prescription according to the fertilizer deficiency information map; wherein, the fertilization prescription is the fertilizer requirement information of each position in the target area; and, It is used to determine the fertilization time and fertilization route according to the meteorological data, and perform fertilization according to the fertilization route at the fertilization time; during the fertilization process, the drone is controlled and adjusted according to the fertilization prescription.

2. The fertilization control system based on an unmanned aerial vehicle according to claim 1, wherein Determining the fertilization time according to the meteorological data includes: Extracting the precipitation at each moment from the meteorological data; using the moment as the independent variable and the precipitation as the dependent variable to fit and establish a precipitation change curve; wherein, the interval between each moment is 10 minutes, 15 minutes or 30 minutes; Determining whether there is an abnormal rainfall interval according to the precipitation change curve; if so, determining the fertilization time based on the abnormal rainfall interval; if not, determining the fertilization time based on the precipitation change curve.

3. The fertilization control system based on an unmanned aerial vehicle according to claim 2, wherein, Determining the fertilization time based on the abnormal rainfall interval includes: Identifying the corresponding abnormal rainfall interval in the precipitation change curve, and calculating the time interval between adjacent abnormal rainfall intervals; wherein, the precipitation in the abnormal rainfall interval will affect the fertilization effect; When the meteorological data of the time interval meets the drone flight requirements and the time interval is greater than the time threshold, taking the start time corresponding to the time interval as the fertilization time; wherein, the time threshold is set according to the fertilizer absorption efficiency of the crops under the corresponding meteorological data.

4. The fertilization control system based on an unmanned aerial vehicle according to claim 2, wherein Determining the fertilization time based on the precipitation change curve includes: Setting a rainfall threshold; wherein, the rainfall threshold is used to improve the penetration and absorption efficiency of fertilizers; Mark the function of the precipitation change curve as F(t), and mark the rainfall threshold as YY; solve the formula to obtain t2, and use the moment corresponding to t2 as the fertilization moment; where t1 is the moment when it starts to rain on the precipitation change curve.

5. The fertilization control system based on an unmanned aerial vehicle according to claim 1, wherein Controlling and adjusting the drone according to the fertilization prescription during the fertilization process includes: At the fertilization time, controlling the drone to start fertilization work according to the fertilization route; wherein, the fertilization route includes the flight altitude, flight attitude and flight path of the drone; When the drone flies a set distance according to the fertilization route, extracting the fertilization image during the fertilization process within the set distance; analyzing the fertilization image to determine whether the fertilization route needs to be adjusted.

6. The fertilization control system based on an unmanned aerial vehicle according to claim 5, wherein, Determining whether the fertilization route needs to be adjusted includes: Analyzing the fertilization image through a machine learning algorithm, and extracting the fertilization area in the fertilization image; wherein, the fertilization area is the effective ground area covered by the fertilizer during the drone fertilization process; Comparing the fertilization area with the preset coverage area corresponding to the fertilization route to determine the area deviation information; adjusting the fertilization route according to the fertilization area and the area deviation information.

7. The fertilization control system based on an unmanned aerial vehicle according to claim 6, characterized in that, Extracting the fertilization area in the fertilization image includes: Identifying the fertilizer coverage area in the fertilization image through a machine learning algorithm; wherein, the fertilizer coverage area is the ground area covered by the fertilizer during the drone fertilization process; Identify the invalid areas in the fertilizer coverage area according to the fertilization prescription, and exclude the invalid areas from the fertilizer coverage area to obtain the fertilization area.

8. The fertilization control system based on an unmanned aerial vehicle according to claim 6, characterized in that Determine the area deviation information, including: Compare the fertilization areas corresponding to all the fertilization images with a preset coverage area, and calculate the deviation distance, deviation direction, and UAV height between the two; where the deviation distance is the mean or the maximum value; Judge the data correlation between the deviation distance and deviation direction and the meteorological data; when the data correlation is greater than the correlation threshold, integrate the deviation distance, deviation direction, and UAV height into the area deviation information.

9. The fertilization control system based on an unmanned aerial vehicle according to claim 6, characterized in that, Adjust the fertilization route according to the fertilization area and the area deviation information, including: Calculate the area boundary of the fertilization area; Adjust the fertilization route according to the area boundary and the area deviation information to obtain a new fertilization route; where the fertilization area corresponding to the new fertilization route is greater than or equal to the target area.

10. A fertilization control method based on an unmanned aerial vehicle, which operates based on the fertilization control system based on an unmanned aerial vehicle according to any one of claims 1 to 9, characterized in that, Include: Collect the soil information and image data of the target area, and collect meteorological data through a meteorological platform; where the target area is the area that needs to be fertilized, the soil information is used to identify the element content in the soil, and the image data is used to identify the growth status of the crops; Determine the fertilizer deficiency information map of the target area according to the soil information and image data, and formulate a fertilization prescription according to the fertilizer deficiency information map; where the fertilization prescription is the fertilizer requirement information at each position in the target area; Determine the fertilization time and fertilization route according to the meteorological data, and perform fertilization according to the fertilization route at the fertilization time; during the fertilization process, control and adjust the UAV according to the fertilization prescription.

Images