A method and system for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs)

Abstract

This invention relates to the field of unmanned aerial vehicle (UAV) application technology, specifically disclosing a method and system for acquiring soybean seedling characteristics based on UAVs. The method includes sending a satellite image acquisition request to a map service to obtain satellite images of the soybean planting area; identifying the satellite images to determine regional demarcation lines; establishing a detection path based on the regional demarcation lines; sending the detection path to a UAV assembly; acquiring the motion parameters of the UAV assembly in real time; and determining soybean seedling characteristics based on the motion parameters. Specifically, while the UAV assembly moves along the detection path, it acquires images of the planting area in real time, identifies the images, and adjusts the motion parameters in real time based on the identification results. This invention integrates the data identification process into the UAV, allowing for rapid determination of seedling characteristics by acquiring the UAV's motion parameters through a locator. Furthermore, it can locate potentially defective points through parameter differences. The source data processed is only location information, resulting in high identification efficiency.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) application technology, specifically a method and system for acquiring soybean seedling characteristics based on UAVs. Background Technology

[0002] Soybean growing areas are mostly flat and relatively low-lying, with strong water and fertilizer retention capacity, moderate soil moisture content, deep topsoil, and moderate soil compaction, which promotes root growth and development and is most suitable for soybean growth.

[0003] The flat terrain of soybean planting areas is very conducive to intelligent identification processes. Therefore, many drone-based soybean seedling characteristic identification schemes have emerged in the existing technology.

[0004] However, most existing drone-based soybean seedling characteristic recognition schemes involve collecting images of the planting area by drones, and then having a data processing center identify the images to determine the characteristics of the soybean seedlings. In this case, the data processing center's data processing process involves identifying the images of the planting area. When the soybean planting area is large, the data processing process is extremely demanding. How to reduce the data processing pressure is the technical problem that this invention aims to solve. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs) to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs), the method comprising:

[0008] The area of ​​the soybean planting region is obtained, and a combination of drones is selected based on the area and a preset ratio; the combination of drones includes a coarse inspection drone and a fine inspection drone; the detection accuracy of the coarse inspection drone is less than that of the fine inspection drone.

[0009] Send a satellite image request to the map service to obtain satellite images of the soybean growing area;

[0010] The satellite image is identified to determine the regional dividing lines, and a detection path is established based on the regional dividing lines.

[0011] The detection path is sent to the drone assembly, the motion parameters of the drone assembly are acquired in real time, and the characteristics of soybean seedling stage are determined based on the motion parameters.

[0012] The drone assembly acquires images of the planting area in real time while moving along the detection path, identifies the images of the planting area, and adjusts the motion parameters in real time based on the identification results.

[0013] As a further aspect of the present invention: the step of obtaining the area of ​​the soybean planting area and selecting a combination of drones based on the area and a preset ratio includes:

[0014] Receive the planting area image and boundary coordinates with scale input from the management;

[0015] Contour recognition is performed on the planting area image to determine the planting area boundary;

[0016] The planting area boundary is adjusted according to the boundary coordinates and scale.

[0017] Calculate the area area based on the adjusted planting area boundary, and then look up the minimum detection quantity in a preset quantity table based on the area area.

[0018] The drone combination is determined based on the minimum number of detections and the preset ratio.

[0019] As a further aspect of the present invention: the step of identifying the satellite image, determining the region demarcation line, and establishing a detection path based on the region demarcation line includes:

[0020] Obtain the time information of the satellite image, and query the environmental information in the weather service based on the time information;

[0021] Select target satellite images based on environmental information;

[0022] The target satellite image is subjected to color value conversion to determine a single-value image;

[0023] The pixels in the single-value image are traversed according to the preset Gaussian operator, and the region dividing line is determined based on the traversal result.

[0024] Establish a detection path based on the area separation line.

[0025] As a further aspect of the present invention: the step of traversing the pixels in a single-value image according to a preset Gaussian operator and determining the region dividing line based on the traversal result includes:

[0026] The Gaussian mapping value of each pixel is calculated by traversing the single-value image according to the preset Gaussian operator.

[0027] The Gaussian mapping value is compared with the original value, and the deviation ratio is calculated;

[0028] When the deviation ratio reaches a preset deviation threshold, the corresponding pixel is marked;

[0029] Count the marked pixels to determine the region dividing lines;

[0030] The Gaussian operator is:

[0031] ;

[0032] When there is no pixel at the corresponding position, a virtual pixel is created based on the center pixel.

[0033] As a further aspect of the present invention: the step of establishing a detection path based on the region dividing line includes:

[0034] Determine the area to be inspected and its center point based on the area dividing lines;

[0035] The path between the center points is established based on the Dijkstra algorithm, serving as the detection direction;

[0036] The detection path is determined in direction based on the area to be inspected and the detection area of ​​the drone;

[0037] Wherein, the union of the detection areas of the UAV along the detection path is greater than the area of ​​the region to be inspected.

[0038] As a further aspect of the present invention: the step of sending the detection path to the UAV assembly, acquiring the motion parameters of the UAV assembly in real time, and determining the characteristics of soybean seedlings based on the motion parameters includes:

[0039] Send the detection path to the drone assembly;

[0040] Real-time acquisition of motion parameters of the drone combination, and simultaneous calculation of parameter differences between the coarse inspection drone and the fine inspection drone;

[0041] Soybean seedling characteristics are determined based on the motion parameters and the parameter differences.

[0042] The present invention also provides a soybean seedling stage characteristic acquisition system based on unmanned aerial vehicles (UAVs), the system comprising:

[0043] The drone selection module is used to acquire the area of ​​the soybean planting area and select a drone combination based on the area and a preset ratio; the drone combination includes a coarse inspection drone and a fine inspection drone; the detection accuracy of the coarse inspection drone is lower than that of the fine inspection drone.

[0044] The satellite image acquisition module is used to send a satellite image acquisition request to the map service to obtain satellite images of the soybean planting area;

[0045] The detection path establishment module is used to identify the satellite image, determine the region dividing line, and establish a detection path based on the region dividing line;

[0046] The motion analysis module is used to send the detection path to the UAV assembly, acquire the motion parameters of the UAV assembly in real time, and determine the characteristics of soybean seedlings based on the motion parameters.

[0047] The drone assembly acquires images of the planting area in real time while moving along the detection path, identifies the images of the planting area, and adjusts the motion parameters in real time based on the identification results.

[0048] As a further aspect of the present invention: the UAV selection module includes:

[0049] A boundary determination unit is used to perform contour recognition on the planting area image to determine the planting area boundary;

[0050] A boundary adjustment unit is used to adjust the boundary of the planting area according to the boundary coordinates and scale.

[0051] The data query unit is used to calculate the area based on the adjusted planting area boundary and to query the minimum detection quantity in a preset quantity table based on the area.

[0052] The combination determination unit is used to determine the combination of drones based on the minimum number of detections and a preset ratio.

[0053] As a further aspect of the present invention: the detection path establishment module includes:

[0054] An environmental information query unit is used to obtain time information from satellite images and query environmental information in weather services based on the time information.

[0055] The selection unit is used to select target satellite images based on environmental information.

[0056] A color value conversion unit is used to perform color value conversion on the target satellite image to determine a single-value image;

[0057] The separator line determination unit is used to traverse the pixels in the single-value image according to a preset Gaussian operator and determine the region separator line based on the traversal result.

[0058] An execution unit is established to create a detection path based on the region dividing line.

[0059] As a further aspect of the present invention: the motion analysis module includes:

[0060] The path transmission unit is used to transmit the detected path to the UAV assembly;

[0061] The parameter difference calculation unit is used to acquire the motion parameters of the UAV combination in real time and simultaneously calculate the parameter difference between the coarse inspection UAV and the fine inspection UAV.

[0062] The feature determination unit is used to determine soybean seedling characteristics based on the motion parameters and the parameter differences.

[0063] Compared with the prior art, the beneficial effects of the present invention are: the present invention integrates the data recognition process into the drone, and the characteristics of seedlings can be quickly determined by obtaining the motion parameters of the drone through the locator. In addition, points that may have defects can be located by the parameter difference. The source data processed is only location information, which is extremely simple and has high recognition efficiency. Attached Figure Description

[0064] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention.

[0065] Figure 1 This is a flowchart of a method for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs).

[0066] Figure 2 This is the first sub-flowchart of the method for acquiring soybean seedling characteristics based on drones.

[0067] Figure 3 This is the second sub-flow flowchart of the method for obtaining soybean seedling characteristics based on drones.

[0068] Figure 4 This is the third sub-flow flowchart of the method for acquiring soybean seedling characteristics based on drones.

[0069] Figure 5 This is a block diagram of the composition structure of a UAV-based soybean seedling characteristic acquisition system. Detailed Implementation

[0070] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0071] Figure 1 The flowchart below illustrates a method for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs). In this embodiment of the invention, a method for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs) includes:

[0072] Step S100: Obtain the area of ​​the soybean planting area, and select a drone combination according to the area and a preset ratio; the drone combination includes a coarse inspection drone and a fine inspection drone; the detection accuracy of the coarse inspection drone is less than that of the fine inspection drone;

[0073] The soybean planting area is pre-demarcated by staff. Once the soybean planting area is demarcated, the area is known data. Based on the area, a combination of drones is selected to obtain characteristics of soybean seedlings. The combination of drones generally includes two types: one is a high-precision precision inspection drone, and the other is a low-precision coarse inspection drone.

[0074] Step S200: Send a satellite image acquisition request to the map service to obtain satellite images of the soybean planting area;

[0075] Existing map services generally include satellite image services, which can display satellite images of various regions. By sending a satellite image retrieval request to the map service and inputting the coordinates of the soybean planting area, a satellite image of the soybean planting area can be obtained. The accuracy of the satellite image is related to the permissions of the subject executing this method; the higher the permissions, the higher the accuracy of the satellite image.

[0076] Step S300: Identify the satellite image, determine the region dividing lines, and establish a detection path based on the region dividing lines;

[0077] Satellite images reflect the overall characteristics of the soybean planting process. By identifying satellite images, differences between different regions can be determined, and then the dividing lines between regions can be identified, which are called regional dividing lines. Based on the regional dividing lines, the detection path of the UAV combination can be determined.

[0078] Step S400: Send the detection path to the UAV assembly, acquire the motion parameters of the UAV assembly in real time, and determine the characteristics of soybean seedlings based on the motion parameters;

[0079] Once the detection path is determined, it is sent to the drone assembly. As the drone assembly moves along the detection path, it acquires images of the planting area in real time, identifies the images, and adjusts the motion parameters in real time based on the identification results. On this basis, the coarse inspection drone and the fine inspection drone may experience positional or velocity shifts. At this time, by analyzing the motion parameters, the state of each area in the soybean planting area can be determined.

[0080] Since the technical solution of this invention provides two drones with different levels of precision, the detection process itself can also be identified based on the parameter difference during the identification of seedling characteristics.

[0081] Figure 2 This is a flowchart of the first sub-process of the method for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs). The steps of acquiring the area of ​​the soybean planting area and selecting a combination of UAVs based on the area and a preset ratio include:

[0082] Step S101: Receive the planting area image with scale and boundary coordinates input by the management;

[0083] Step S102: Perform contour recognition on the planting area image to determine the planting area boundary;

[0084] Step S103: Adjust the boundary of the planting area according to the boundary coordinates and scale;

[0085] The process of determining soybean planting areas and calculating their area is mainly accomplished through boundary coordinates. In the technical solution of this invention, a new method is also provided, which is to determine a basic outline through the planting area image uploaded by the management. At this time, the boundary of the planting area can be determined through a limited number of boundary coordinates. When using the planting area image, the number of boundary coordinates required for the boundary determination process is less.

[0086] Step S104: Calculate the area based on the adjusted planting area boundary, and look up the minimum detection quantity in the preset quantity table based on the area;

[0087] Once the planting area boundary is determined, the area can be calculated based on the planting area boundary. The number of drones detected can be queried based on the area. The number of drones detected can be the number of drones used for coarse inspection, the number of drones used for fine inspection, or the total number.

[0088] Step S105: Determine the drone combination based on the minimum number of detections and the preset ratio;

[0089] There is a ratio between the number of rough inspection drones and the number of fine inspection drones. The drone combination can be determined based on this ratio and the minimum number of inspections determined above.

[0090] Figure 3 The second sub-flowchart of the method for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs) includes the following steps: identifying the satellite image, determining the region demarcation line, and establishing a detection path based on the region demarcation line.

[0091] Step S301: Obtain the time information of the satellite image, and query the environmental information in the weather service based on the time information;

[0092] Step S302: Select the target satellite image based on environmental information;

[0093] Step S303: Perform color value conversion on the target satellite image to determine a single-value image;

[0094] Step S304: Traverse the pixels in the single-value image according to the preset Gaussian operator, and determine the region dividing line according to the traversal result;

[0095] Step S305: Establish a detection path based on the area separation line.

[0096] The above content specifies the detection path for the UAV combination. First, steps S301 and S302 select satellite images and query environmental information (including lighting conditions and air conditions) through time information. Based on the environmental information, a target satellite image is selected from the satellite images. Then, the target satellite image is converted to color values ​​to obtain a single-value image. The single-value image can be a grayscale image. By analyzing the grayscale image, the region dividing line can be determined.

[0097] As a preferred embodiment of the technical solution of the present invention, the step of traversing the pixels in a single-value image according to a preset Gaussian operator and determining the region dividing line based on the traversal result includes:

[0098] The Gaussian mapping value of each pixel is calculated by traversing the single-value image according to the preset Gaussian operator.

[0099] The Gaussian mapping value is compared with the original value, and the deviation ratio is calculated;

[0100] When the deviation ratio reaches a preset deviation threshold, the corresponding pixel is marked;

[0101] Statistically count the marked pixels to determine the region dividing lines.

[0102] The above provides a specific scheme for determining region dividing lines. First, the Gaussian operator is used to traverse each pixel, and the Gaussian mapping value of the center point is calculated from the elements around the pixel. Then, the difference between the Gaussian mapping value and the original value is calculated, and the difference is represented by a ratio. If this difference is large, it means that the difference between the pixels around the center point and the center point is large. In this case, the corresponding pixels need to be marked, and the marked pixels are connected to obtain the region dividing line.

[0103] The Gaussian operator is:

[0104] ;

[0105] When there is no pixel at the corresponding position, a virtual pixel is created based on the center pixel.

[0106] The Gaussian operator mentioned above is a 5x5 matrix. At the boundary points of a single-valued image, there will be one side where no pixel exists. In this case, the current center point is copied, and a virtual point is created at the corresponding position.

[0107] As a preferred embodiment of the technical solution of the present invention, the step of establishing a detection path based on the region dividing line includes:

[0108] Determine the area to be inspected and its center point based on the area dividing lines;

[0109] The path between the center points is established based on the Dijkstra algorithm, serving as the detection direction;

[0110] The detection path is determined in direction based on the area to be inspected and the detection area of ​​the drone;

[0111] Wherein, the union of the detection areas of the UAV along the detection path is greater than the area of ​​the region to be inspected;

[0112] The process of determining the detection path mainly includes two aspects: the detection direction and the detection path. The detection direction is used to characterize the inspection sequence of each area to be inspected, and the detection path is used to characterize the inspection sequence of each area to be inspected. In simple terms, the detection direction determines which area to be inspected first, and the detection path moves within the area to be inspected to inspect the entire area.

[0113] Figure 4 The third sub-flow diagram of the method for acquiring soybean seedling characteristics based on UAVs includes the following steps: sending the detection path to the UAV assembly, acquiring the motion parameters of the UAV assembly in real time, and determining the soybean seedling characteristics based on the motion parameters.

[0114] Step S401: Send the detection path to the UAV assembly;

[0115] Step S402: Acquire the motion parameters of the drone combination in real time, and simultaneously calculate the parameter difference between the coarse inspection drone and the fine inspection drone;

[0116] Step S403: Determine the characteristics of soybean seedlings based on the motion parameters and the parameter differences.

[0117] The above content defines the process for determining the characteristics of soybean seedlings. During the detection process, the UAV identifies the acquired images of soybean seedlings to determine the growth status of the soybeans. Based on the growth status of the soybeans, motion parameters are determined, including motion position, motion speed, and motion acceleration.

[0118] Specifically, the location of the rough inspection drone is obtained in real time by the locator, and then the motion parameters are determined. Based on the motion parameters, the growth status of soybeans can be deduced, and the characteristics of soybean seedlings can be obtained.

[0119] Based on the above, since the recognition accuracy of coarse inspection drones and fine inspection drones is different, the recognition results may also be different. Compared with coarse inspection drones, fine inspection drones can detect some defect points. At this time, the movement speed will be reduced, resulting in a parameter difference between fine inspection drones and coarse inspection drones. By analyzing the parameter difference, some obvious abrupt changes can be identified (which can be reflected by the motion acceleration). The corresponding location is the location of the defect. Then, based on the location, the image collected by the drone is obtained for subsequent judgment, such as manual inspection.

[0120] In the technical solution of this invention, not only can the characteristics of seedlings be quickly determined through motion parameters, but also points that may have defects can be located through parameter differences. The source data processed is extremely simple, and the recognition efficiency is extremely high.

[0121] Figure 5 This is a block diagram of the composition structure of a UAV-based soybean seedling characteristic acquisition system. In this embodiment of the invention, a UAV-based soybean seedling characteristic acquisition system 10 includes:

[0122] The drone selection module 11 is used to obtain the area of ​​the soybean planting area and select a drone combination according to the area and a preset ratio; the drone combination includes a coarse inspection drone and a fine inspection drone; the detection accuracy of the coarse inspection drone is less than that of the fine inspection drone.

[0123] Satellite image acquisition module 12 is used to send a satellite image acquisition request to the map service to obtain satellite images of the soybean planting area;

[0124] The detection path establishment module 13 is used to identify the satellite image, determine the area separation line, and establish a detection path based on the area separation line;

[0125] Motion analysis module 14 is used to send the detection path to the UAV assembly, acquire the motion parameters of the UAV assembly in real time, and determine the characteristics of soybean seedlings based on the motion parameters.

[0126] The drone assembly acquires images of the planting area in real time while moving along the detection path, identifies the images of the planting area, and adjusts the motion parameters in real time based on the identification results.

[0127] In a preferred embodiment of the technical solution of the present invention, the UAV selection module 11 includes:

[0128] A boundary determination unit is used to perform contour recognition on the planting area image to determine the planting area boundary;

[0129] A boundary adjustment unit is used to adjust the boundary of the planting area according to the boundary coordinates and scale.

[0130] The data query unit is used to calculate the area based on the adjusted planting area boundary and to query the minimum detection quantity in a preset quantity table based on the area.

[0131] The combination determination unit is used to determine the combination of drones based on the minimum number of detections and a preset ratio.

[0132] Furthermore, the detection path establishment module 13 includes:

[0133] An environmental information query unit is used to obtain time information from satellite images and query environmental information in weather services based on the time information.

[0134] The selection unit is used to select target satellite images based on environmental information.

[0135] A color value conversion unit is used to perform color value conversion on the target satellite image to determine a single-value image;

[0136] The separator line determination unit is used to traverse the pixels in the single-value image according to a preset Gaussian operator and determine the region separator line based on the traversal result.

[0137] An execution unit is established to create a detection path based on the region dividing line.

[0138] Specifically, the motion analysis module 14 includes:

[0139] The path transmission unit is used to transmit the detected path to the UAV assembly;

[0140] The parameter difference calculation unit is used to acquire the motion parameters of the UAV combination in real time and simultaneously calculate the parameter difference between the coarse inspection UAV and the fine inspection UAV.

[0141] The feature determination unit is used to determine soybean seedling characteristics based on the motion parameters and the parameter differences.

[0142] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs), characterized in that, The method includes: The area of ​​the soybean planting region is obtained, and a combination of drones is selected based on the area and a preset ratio; the combination of drones includes a coarse inspection drone and a fine inspection drone; the detection accuracy of the coarse inspection drone is less than that of the fine inspection drone. Send a satellite image request to the map service to obtain satellite images of the soybean growing area; The satellite image is identified to determine the regional dividing lines, and a detection path is established based on the regional dividing lines. The detection path is sent to the drone array; the motion parameters of the drone array are acquired in real time, and the parameter difference between the coarse inspection drone and the fine inspection drone is calculated synchronously; the characteristics of soybean seedling stage are determined based on the motion parameters and the parameter difference. The drone assembly acquires images of the planting area in real time while moving along the detection path, identifies the images of the planting area, and adjusts the motion parameters in real time based on the identification results.

2. The method for acquiring soybean seedling characteristics based on unmanned aerial vehicles according to claim 1, characterized in that, The steps of obtaining the area of ​​the soybean planting region and selecting a drone combination based on the area and a preset ratio include: Receive the planting area image and boundary coordinates with scale input from the management; Contour recognition is performed on the planting area image to determine the planting area boundary; The planting area boundary is adjusted according to the boundary coordinates and scale. Calculate the area area based on the adjusted planting area boundary, and then look up the minimum detection quantity in a preset quantity table based on the area area. The drone combination is determined based on the minimum number of detections and the preset ratio.

3. The method for acquiring soybean seedling characteristics based on unmanned aerial vehicles according to claim 1, characterized in that, The steps of identifying the satellite image, determining region demarcation lines, and establishing detection paths based on the region demarcation lines include: Obtain the time information of the satellite image, and query the environmental information in the weather service based on the time information; Select target satellite images based on environmental information; The target satellite image is subjected to color value conversion to determine a single-value image; The pixels in the single-value image are traversed according to the preset Gaussian operator, and the region dividing line is determined based on the traversal result. Establish a detection path based on the area separation line.

4. The method for acquiring soybean seedling characteristics based on unmanned aerial vehicles according to claim 3, characterized in that, The step of traversing the pixels in a single-value image according to a preset Gaussian operator and determining the region dividing line based on the traversal result includes: The Gaussian mapping value of each pixel is calculated by traversing the single-value image according to the preset Gaussian operator. The Gaussian mapping value is compared with the original value, and the deviation ratio is calculated; When the deviation ratio reaches a preset deviation threshold, the corresponding pixel is marked; Count the marked pixels to determine the region dividing lines; The Gaussian operator is: ； When there is no pixel at the corresponding position, a virtual pixel is created based on the center pixel.

5. The method for acquiring soybean seedling characteristics based on unmanned aerial vehicles according to claim 3, characterized in that, The step of establishing a detection path based on the region dividing line includes: Determine the area to be inspected and its center point based on the area dividing lines; The path between the center points is established based on the Dijkstra algorithm, serving as the detection direction; The detection path is determined in direction based on the area to be inspected and the detection area of ​​the drone; Wherein, the union of the detection areas of the UAV along the detection path is greater than the area of ​​the region to be inspected.

6. A system for acquiring soybean seedling characteristics based on unmanned aerial vehicles (UAVs), characterized in that, The system includes: The drone selection module is used to acquire the area of ​​the soybean planting area and select a drone combination based on the area and a preset ratio; the drone combination includes a coarse inspection drone and a fine inspection drone; the detection accuracy of the coarse inspection drone is lower than that of the fine inspection drone. The satellite image acquisition module is used to send a satellite image acquisition request to the map service to obtain satellite images of the soybean planting area; The detection path establishment module is used to identify the satellite image, determine the region dividing line, and establish a detection path based on the region dividing line; The motion analysis module is used to send the detection path to the UAV assembly, acquire the motion parameters of the UAV assembly in real time, and determine the characteristics of soybean seedlings based on the motion parameters. When the drone assembly moves along the detection path, it acquires images of the planting area in real time, identifies the images of the planting area, and adjusts the motion parameters in real time based on the identification results. The motion analysis module includes: The path transmission unit is used to transmit the detected path to the UAV assembly; The parameter difference calculation unit is used to acquire the motion parameters of the UAV combination in real time and simultaneously calculate the parameter difference between the coarse inspection UAV and the fine inspection UAV. The feature determination unit is used to determine soybean seedling characteristics based on the motion parameters and the parameter differences.

7. The soybean seedling stage characteristic acquisition system based on unmanned aerial vehicles according to claim 6, characterized in that, The drone selection module includes: A boundary determination unit is used to perform contour recognition on the planting area image to determine the planting area boundary; A boundary adjustment unit is used to adjust the boundary of the planting area according to the boundary coordinates and scale. The data query unit is used to calculate the area based on the adjusted planting area boundary and to query the minimum detection quantity in a preset quantity table based on the area. The combination determination unit is used to determine the combination of drones based on the minimum number of detections and a preset ratio.

8. The soybean seedling stage characteristic acquisition system based on unmanned aerial vehicles according to claim 6, characterized in that, The detection path establishment module includes: An environmental information query unit is used to obtain time information from satellite images and query environmental information in weather services based on the time information. The selection unit is used to select target satellite images based on environmental information. A color value conversion unit is used to perform color value conversion on the target satellite image to determine a single-value image; The separator line determination unit is used to traverse the pixels in the single-value image according to a preset Gaussian operator and determine the region separator line based on the traversal result. An execution unit is established to create a detection path based on the region dividing line.

Images