Boom sprayer drift potential testing device and method based on computer vision

Abstract

The present invention relates to a device and method for testing the drift potential of a boom sprayer based on computer vision. The testing device includes a droplet receiving platform, a remote control sliding cover, a droplet receiving device bracket, a camera, and a tripod; a droplet receiving device, receiving materials, and marking points are arranged on the droplet receiving platform; the droplet receiving device is a hard flat plate, receiving materials are attached to the center of the droplet receiving device, and multiple marking points are attached to the corners of the droplet receiving device; a droplet receiving device bracket is provided at the bottom of the droplet receiving platform, and the droplet receiving device bracket has leveling legs to ensure that the droplet receiving platform is level; the camera is mounted on the tripod, and the camera lens is aimed obliquely downward at the droplet receiving device. The present invention detects the change in deposition amount over time and calculates the drift potential by using computer vision methods, without the need to add expensive fluorescent indicators or special instruments to detect the deposition amount, thus saving manpower and material resources.

Description

Technical Field

[0001] The present invention belongs to the technical field of plant protection machinery application in precision agriculture, and relates to a device and method for testing the drift potential of a boom sprayer based on computer vision. Background Art

[0002] Pesticide drift refers to the portion of pesticide that is not deposited on the target crop during application and is instead carried away by the wind. In recent years, with the increasing emphasis on environmental protection, spray drift control has become increasingly important.

[0003] The drift potential of a sprayer is of great reference value for guiding pesticide application operations and determining the spray drift buffer zone. The current drift potential of a sprayer is based on a dedicated test bench (GB / T 40346-2021, ISO 22401:2015), which has a complex structure and is relatively expensive. Each measurement requires researchers to arrange a large number of samples, add fluorescent indicators, and use fluorescence emission spectroscopy to measure the deposition amount. This process is extremely cumbersome, requires a lot of manpower and material resources, and has very low work efficiency. In order to improve the accuracy of the evaluation results, multiple repetitions are required, which further increases the consumption of materials for this method, as well as the time and energy investment of researchers. Summary of the Invention

[0004] In response to the above problems, the present invention proposes a device and method for testing the drift potential of a boom sprayer based on computer vision to solve the problems of complex operation and high cost of special equipment in current technology.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A device for testing the drift potential of a boom sprayer based on computer vision comprises a droplet receiving platform 1, a remote control sliding cover 2, a droplet receiving device bracket 3, a camera 4 and a tripod 5; wherein,

[0007] The droplet receiving platform 1 is provided with a droplet receiving device 6, a receiving material 7 and marking points 8; the droplet receiving device 6 is a hard flat plate, the center of the droplet receiving device 6 is pasted with the receiving material 7, and the corners of the droplet receiving device 6 are pasted with multiple marking points 8;

[0008] A droplet receiving device bracket 3 is provided at the lower portion of the droplet receiving platform 1. The droplet receiving device bracket 3 has leveling legs to ensure that the droplet receiving platform 1 is level. A camera 4 is mounted on a tripod 5, with the lens of the camera 4 pointing obliquely downward toward the droplet receiving device 6.

[0009] The remote control sliding cover 2 can slide on the droplet receiving device bracket 3; the remote control sliding cover 2 includes an electric push rod 10 at the bottom, one end of the electric push rod 10 is fixed on the remote control sliding cover 2 itself, and the other end of the electric push rod 10 is fixedly connected to the edge of the droplet receiving device bracket 3; when the electric push rod 10 is in the shortest position, the remote control sliding cover 2 is located at a fixed position close to the edge of the droplet receiving device bracket 3, at this time, the remote control sliding cover 2 is in an open state, the receiving material 7 is completely exposed, and the droplets can be deposited on the receiving material 7 and captured by the camera 4; when the electric push rod 10 is extended, the remote control sliding cover 2 slides under the action of the electric push rod 10, completely covering the droplet receiving platform 1, at this time, the remote control sliding cover 2 is in a closed state.

[0010] A method for testing the drift potential of a boom sprayer based on computer vision using the apparatus for testing the drift potential of a boom sprayer based on computer vision comprises the following steps:

[0011] S1. Place the droplet receiving platform 1 and the camera 4 in the forward direction of the boom sprayer 9 to be tested; stick new receiving material 7 on the droplet receiving device 6, and then close the remote control sliding cover 2;

[0012] S2. Prepare a colorable spray solution using carmine or lemon yellow, and adjust the spray pressure to the pressure required for the test;

[0013] S3, start all nozzles on the sprayer boom of the sprayer boom sprayer 9 and start moving at the test speed;

[0014] S4. When the spray boom of the boom sprayer 9 passes the test device by 2 m ± 0.1 m, the remote control sliding cover 2 is manually triggered to open, and the camera 4 is simultaneously turned on to start recording video;

[0015] S5, the remote control sliding cover 2 is in the open state and receives the droplets for at least 5 minutes, after which the remote control sliding cover 2 is manually triggered to close, and the video recording of the camera 4 is simultaneously stopped;

[0016] S6. Analyze the video of camera 4 to obtain the change of deposition amount over time D(t);

[0017] S6.1. Analyze each frame in the video separately, and take the frame in which the remote control sliding cover 2 is fully opened as frame 0;

[0018] S6.2. First, convert the color image captured by camera 4 into a grayscale image and extract the marker point 8 in the field of view of camera 4. Perform an affine transformation based on the coordinates of marker point 8 to obtain an orthographic image of receiving material 7.

[0019] S6.3. Binarize the orthophoto grayscale image obtained in step S6.2 using the OTSU binarization method to obtain a binary image of the fog droplets;

[0020] S6.4. Segment the area of ​​each droplet from the binary image of the droplets obtained in step S6.3 and calculate the total deposition amount according to formula 1;

[0021] d i =1.12A 0.468 Formula 1

[0022] In formula 1, d i is the diameter of the droplet corresponding to the spot, in μm; A is the area of ​​the droplet spot on the receiving material 7, in μm 2 ;

[0023] The corresponding droplet volume is:

[0024]

[0025] In formula 2, V i is the volume of the droplet corresponding to the spot, in μm 3 ;d i is the diameter of the corresponding droplet, in μm;

[0026] Then the deposition amount D of the current frame is:

[0027]

[0028] In formula 3, D is the deposition amount of the current frame, in μm 3 / cm 2 ; V i The volume of the droplet corresponding to each spot, in μm 3 ; S is the area of ​​the receiving material 7, in cm 2 ;

[0029] S6.5. Calculate the deposition amount for each frame of image to obtain the relationship D(t) between the deposition amount and time.

[0030] S7, calculating the drift potential index according to the relationship D(t) between the amount of sedimentation and time obtained in step S6;

[0031]

[0032] In formula 4, DPV is the drift potential index, dimensionless; D i The deposition amount obtained for each frame, in μm 3 / cm 2 ;RSD is the reference spray deposition value below the spray boom, in μm 3 / cm2 , calculated from the measured forward speed and nozzle flow rate.

[0033] In step S3, to ensure that the boom sprayer 9 can achieve a stable spraying state when passing through the test device, the boom sprayer 9 starts spraying at least 20 meters in front of the droplet receiving platform 1 and stops spraying at least 20 meters away from the droplet receiving platform 1.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] Compared with traditional methods, the computer vision-based boom sprayer drift potential testing device and method provided by the present invention detects the change of deposition amount over time and calculates the drift potential through computer vision methods. It does not require the addition of expensive fluorescent indicators or special instruments to detect the deposition amount, thus saving manpower and material resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of a device for testing the drift potential of a boom sprayer based on computer vision according to the present invention;

[0037] Figure 2 Schematic diagram of the droplet receiving device of the present invention;

[0038] Figure 3 A droplet receiving device 6 in the camera's field of view;

[0039] Figure 4 is a grayscale image of the receiving material 7 restored to an orthographic perspective;

[0040] Figure 5 is the grayscale image after binarization;

[0041] Figure 6 Top view of the equipment layout for field testing.

[0042] The accompanying drawings are as follows:

[0043] 1. Droplet receiving platform 2. Remote control sliding cover

[0044] 3. Droplet receiving device bracket 4. Camera

[0045] 5. Tripod 6. Mist droplet receiving device

[0046] 7. Receive materials 8. Marking points

[0047] 9. Sprayer boom DETAILED DESCRIPTION

[0048] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0049] like Figure 1 As shown in FIG, the computer vision-based sprayer drift potential test device of the present invention comprises a droplet receiving platform 1, a remote control sliding cover 2, a droplet receiving device bracket 3, a camera 4 and a tripod 5. Figure 2 As shown, the droplet receiving platform 1 is provided with a droplet receiving device 6, a receiving material 7 and marking points 8. The droplet receiving device 6 is a hard flat plate, the receiving material 7 is attached to the center of the droplet receiving device 6, and a plurality of marking points 8 are attached to the corners of the droplet receiving device 6.

[0050] A droplet receiving device bracket 3 is provided at the lower portion of the droplet receiving platform 1. The droplet receiving device bracket 3 has leveling legs to ensure that the droplet receiving platform 1 is level. A camera 4 is mounted on a tripod 5, with the lens of the camera 4 directed obliquely downward toward the droplet receiving device 6.

[0051] The remote-controlled sliding cover 2 is capable of sliding on the mist receiving device support 3. An electric push rod 10 is located beneath the remote-controlled sliding cover 2. One end of the push rod 10 is fixed to the remote-controlled sliding cover 2 itself, and the other end is fixedly connected to the edge of the mist receiving device support 3. When the push rod 10 is at its shortest position, the remote-controlled sliding cover 2 is located near the edge of the mist receiving device support 3, and the remote-controlled sliding cover 2 is in an open state. When the push rod 10 is extended, the remote-controlled sliding cover 2 slides under the force of the push rod 10, completely covering the mist receiving platform 1, and the remote-controlled sliding cover 2 is in a closed state.

[0052] When the remote control sliding cover 2 is in the closed state, the remote control sliding cover 2 slides to the top of the droplet receiving platform 1, and the remote control sliding cover 2 completely covers the droplet receiving platform 1. Figure 1 As shown. Because receiving material 7 is placed on droplet receiving platform 1, droplets cannot land on receiving material 7. When the remote-controlled sliding cover 2 is open, it slides to a fixed position near its edge with the droplet receiving device bracket 3, completely exposing receiving material 7. Droplets can deposit on receiving material 7 and be captured by camera 4. The remote-controlled sliding cover 2 requires manual control by the experimenter. When the slide is triggered, a signal is automatically synchronized to camera 4, causing camera 4 to start or stop recording as the remote-controlled sliding cover 2 moves.

[0053] When the remote control sliding cover 2 is in the open state, the receiving material 7 and the marking point 8 appear completely in the field of view of the camera 4 , and the camera 4 will not block the mist droplets falling on the receiving material 7 .

[0054] After pesticide droplets are atomized by the nozzle, they produce droplets of varying sizes. These droplets have different settling times. Large droplets settle onto the target surface more quickly, while small droplets settle more slowly. This settling process is easily affected by meteorological factors, leading to drift. Therefore, measuring the change in deposition over time can reveal the percentage of smaller droplets deposited. This percentage is called the drift potential index. In the experiment, droplets deposited on the target after the sprayer boom has traveled 2 meters are designated as small droplets. A larger percentage indicates a greater number of suspended small droplets generated after the sprayer has traveled, posing a greater risk of drift.

[0055] A method for testing the drift potential of a boom sprayer based on computer vision using the apparatus for testing the drift potential of a boom sprayer based on computer vision comprises the following steps:

[0056] S1, such as Figure 6 As shown, arrange the test site, placing the droplet receiving platform 1 and camera 4 in the direction of travel of the boom sprayer 9 to be tested. Apply new receiving material 7 to the droplet receiving device 6, then close the remote control sliding cover 2. Subsequent experiments should be conducted in calm wind conditions. If the ambient wind speed exceeds 0.5 m / s during the experiment, the data will be invalid.

[0057] S2. Prepare a colorable spray solution using carmine or lemon yellow, and adjust the spray pressure to the pressure required for the test;

[0058] S3. Start all nozzles on the boom of the boom sprayer 9 and begin moving at the test speed. To ensure that the boom sprayer 9 can achieve a stable spraying state when passing through the test device, the boom sprayer 9 starts spraying at least 20 meters in front of the droplet receiving platform 1 and stops spraying at least 20 meters away from the droplet receiving platform 1;

[0059] S4. When the spray boom of the boom sprayer 9 passes the test device by 2 m ± 0.1 m, the remote control sliding cover 2 is manually triggered to open, and the camera 4 is simultaneously turned on to start recording video;

[0060] S5, the remote control sliding cover 2 is in the open state and receives the droplets for at least 5 minutes, after which the remote control sliding cover 2 is manually triggered to close, and the video recording of the camera 4 is simultaneously stopped;

[0061] S6. Analyze the video of camera 4 to obtain the change of deposition amount over time D(t);

[0062] S6.1. Analyze each frame in the video separately, and take the frame in which the remote control sliding cover 2 is fully opened as the 0th frame.

[0063] S6.2, such as Figure 3As shown in FIG, the receiving material 7 and the marking point 8 in the field of view of the camera 4 are shown. First, the color image captured by the camera 4 is converted into a grayscale image and the marking point 8 in the field of view of the camera 4 is extracted. An affine transformation is performed based on the coordinates of the marking point 8 to obtain an orthographic image of the receiving material 7, as shown in FIG. Figure 4 shown.

[0064] S6.3. Use the OTSU binarization method to binarize the orthophoto grayscale image obtained in step S6.2 to obtain a binary image of the fog droplets, such as Figure 5 shown.

[0065] S6.4. Segment the area of ​​each droplet from the binary image of the droplets obtained in step S6.3 and calculate the total deposition amount according to formula 1.

[0066] d i =1.12A 0.468 Formula 1

[0067] In formula 1, d i is the diameter of the droplet corresponding to the spot, in μm; A is the area of ​​the droplet spot on the receiving material 7, in μm 2 .

[0068] The corresponding droplet volume is:

[0069]

[0070] In formula 2, V i is the volume of the droplet corresponding to the spot, in μm 3 ;d i is the diameter of the corresponding droplet, in μm.

[0071] Then the deposition amount D of the current frame is:

[0072]

[0073] In formula 3, D is the deposition amount of the current frame, in μm 3 / cm 2 ; V i The volume of the droplet corresponding to each spot, in μm 3 ; S is the area of ​​the receiving material 7, in cm 2 .

[0074] S6.5. Calculate the deposition amount for each frame of image to obtain the relationship D(t) between the deposition amount and time.

[0075] S7. Calculate the drift potential index based on the relationship D(t) between the amount of sedimentation and time obtained in step S6.

[0076]

[0077] In formula 4, DPV is the drift potential index, dimensionless; D i The deposition amount obtained for each frame, in μm 3 / cm 2 ;RSD is the reference spray deposition value below the spray boom, in μm 3 / cm 2 , calculated from the measured forward speed and nozzle flow rate.

Claims

1. A method for testing the drift potential of a boom sprayer based on computer vision, characterized by: The method uses a boom sprayer drift potential test device based on computer vision, wherein the test device comprises a droplet receiving platform (1), a remote control sliding cover (2), a droplet receiving device bracket (3), a camera (4) and a tripod (5); wherein, The droplet receiving platform (1) is provided with a droplet receiving device (6), a receiving material (7), and marking points (8); the droplet receiving device (6) is a hard flat plate, the receiving material (7) is attached to the center of the droplet receiving device (6), and a plurality of marking points (8) are attached to the corners of the droplet receiving device (6); A droplet receiving device bracket (3) is provided at the lower portion of the droplet receiving platform (1), wherein the droplet receiving device bracket (3) has leveling legs to ensure that the droplet receiving platform (1) is level; a camera (4) is mounted on a tripod (5), and a lens of the camera (4) is directed obliquely downward toward the droplet receiving device (6); The remote control sliding cover (2) can slide on the droplet receiving device bracket (3); the remote control sliding cover (2) includes an electric push rod (10) below, one end of the electric push rod (10) is fixed to the remote control sliding cover (2) itself, and the other end of the electric push rod (10) is fixedly connected to the edge of the droplet receiving device bracket (3); when the electric push rod (10) is in the shortest position, the remote control sliding cover (2) is located at a fixed position close to the edge of the droplet receiving device bracket (3), at this time, the remote control sliding cover (2) is in an open state, the receiving material (7) is completely exposed, and the droplets can be deposited on the receiving material (7) and captured by the camera (4); when the electric push rod (10) is extended, the remote control sliding cover (2) slides under the action of the electric push rod (10), completely covering the droplet receiving platform (1), at this time, the remote control sliding cover (2) is in a closed state; The method comprises the following steps: S1. Place the droplet receiving platform (1) and the camera (4) in the forward direction of the boom sprayer (9) to be tested; stick new receiving material (7) on the droplet receiving device (6), and then close the remote control sliding cover (2); S2. Prepare a colorable spray solution using carmine or lemon yellow, and adjust the spray pressure to the pressure required for the test; S3, starting all nozzles on the sprayer boom of the sprayer boom sprayer (9) and starting to move forward at the test speed; S4. When the spray boom of the spray boom sprayer (9) passes the test device by 2 m ± 0.1 m, the remote control sliding cover (2) is manually triggered to open, and the camera (4) is simultaneously turned on, so that the camera (4) starts recording video; S5, the remote control sliding cover (2) is in an open state and receives the mist droplets for not less than 5 minutes, and then the remote control sliding cover (2) is manually triggered to close, and the video recording of the camera (4) is stopped synchronously; S6. Analyze the video of camera (4) to obtain the change of deposition amount over time D(t); S6.

1. Analyze each frame in the video separately, and take the frame in which the remote control sliding cover (2) is fully opened as the 0th frame; S6.2, first converting the color image captured by the camera (4) into a grayscale image and extracting the identification point (8) in the field of view of the camera (4), and performing an affine transformation based on the coordinates of the identification point (8) to obtain an orthographic image of the receiving material (7); S6.

3. Binarize the orthophoto grayscale image obtained in step S6.2 using the OTSU binarization method to obtain a binary image of the fog droplets; S6.

4. Segment the area of ​​each droplet from the binary image of the droplets obtained in step S6.3 and calculate the total deposition amount according to formula 1; Formula 1 In formula 1, d i is the diameter of the droplet corresponding to the spot, in μm; A is the area of ​​the droplet spot on the receiving material (7), in μm 2 ; The corresponding droplet volume is: Formula 2 In formula 2, V i is the volume of the droplet corresponding to the spot, in μm 3 ;d i is the diameter of the corresponding droplet, in μm; Then the deposition amount D of the current frame is: Formula 3 In formula 3, D is the deposition amount of the current frame, in μm 3 / cm 2 ; V i The volume of the droplet corresponding to each spot, in μm 3 ; S is the area of ​​the receiving material (7), in cm 2 ; S6.

5. Calculate the deposition amount for each frame of image to obtain the relationship D(t) between the deposition amount and time. S7, calculating the drift potential index according to the relationship D(t) between the amount of sedimentation and time obtained in step S6; Formula 4 In formula 4, DPV is the drift potential index, dimensionless; D i The deposition amount obtained for each frame, in μm 3 / cm 2 ;RSD is the reference spray deposition value below the spray boom, in μm 3 / cm 2 , calculated from the measured forward speed and nozzle flow rate.

2. The method for testing the drift potential of a boom sprayer based on computer vision according to claim 1, wherein: In step S3, to ensure that the boom sprayer (9) can reach a stable spraying state when passing through the test device, the boom sprayer (9) starts spraying at least 20 m in front of the droplet receiving platform (1) and stops spraying after leaving the droplet receiving platform (1) by at least 20 m.

Images