Pesticide fog drop deposition online detection device based on optical camera

Abstract

The invention discloses a pesticide fog drop deposition on-line detection device based on an optical camera, and relates to the technical field of plant protection fog drop deposition detection research, the pesticide fog drop deposition on-line detection device comprises a light source, an upper shell, a lower shell, a transparent acrylic plate, a projection module, a high definition camera, a Raspberry Pi and a power supply; a first steering engine is arranged on the upper shell, before the transparent acrylic plate obtains fog drops, the first steering engine works to drive the rotating shaft to rotate so as to drive the fog shielding cover to rotate, and the fog shielding cover is used for shielding the upper shell to prevent the fog drops from entering the upper shell; a first steering engine is arranged on the upper shell, a second steering engine is arranged on the lower shell, before the high-definition camera starts to work, the first steering engine rotates the fog shielding cover by 180 degrees so as to cancel shielding of the upper shell, and the second steering engine rotates the upper shell to the position over the lower shell. An information processing terminal and a Raspberry Pi are created by using a python language, the information processing terminal is mainly used for control and display, the Raspberry Pi is mainly used for picture analysis, a controllable closed structure design and a developing system are combined for fog drop detection, and the influence of external illumination is effectively avoided.

Description

Technical Field

[0001] The invention relates to the technical field of plant protection droplet deposition detection research, and in particular to an optical camera-based online detection device for pesticide droplet deposition. Background Art

[0002] Visual analysis is being applied in more and more fields due to its strong adaptability. Optical cameras are widely used in the field of plant protection detection for the current plant protection industry's demand for precision detection and excellent anti-interference capabilities. They have high precision and efficiency in detection, can be used with various peripherals, have strong automation capabilities, are easy to operate and maintain. Compared with the commonly used water-sensitive paper for detecting droplet deposition, as well as resistance sensors and capacitance sensors, optical cameras can be reused and have high detection accuracy. By performing a series of processing and analysis on the images obtained by the camera, more levels of droplet detection data can be obtained, which can greatly improve the detection of droplet deposition quality.

[0003] With the development of the manufacturing industry, 3D printing technology has become more and more mature. Various complex structures can be produced through 3D printing technology. Compared with traditional manufacturing methods, it reduces the cost and cycle of production and improves the accuracy of products. At the same time, it can also customize products according to customer needs.

[0004] Control has always been the core part of a device. With the continuous development of electronic technology, more and more affordable and easy-to-use tools have been developed. Raspberry Pi is a card-type computer the size of a credit card. It has all the basic functions of a computer and has begun to be widely used in all walks of life. As an open source, low-cost, small single-board computer, it has many advantages, including affordable and cost-effective costs; the open source of its hardware and software can greatly promote the creativity and personalized needs of users. Users can add various peripherals and modules as needed to meet their functional needs. In addition, Raspberry Pi supports multiple programming languages ​​and development environments, including Python, C++, Java, etc., making it widely used in various industries. Summary of the invention

[0005] In view of the shortcomings in the prior art, the present invention provides an online detection device for pesticide droplet deposition based on an optical camera. An information processing terminal and a Raspberry Pi are created using the Python language. The information processing terminal is mainly used for control and display, and the Raspberry Pi is mainly used for image analysis. A controllable closed structure design combined with a development system is used for droplet detection to effectively avoid the influence of external light.

[0006] The present invention achieves the above technical objectives through the following technical means.

[0007] An optical camera-based online detection device for pesticide droplet deposition comprises a light source, an upper shell, a lower shell, a transparent acrylic plate, a projection module, a high-definition camera, a Raspberry Pi and a power supply; a first servo is arranged on the upper shell, and before the transparent acrylic plate acquires droplets, the first servo works to drive a rotating shaft to rotate, thereby driving a fog cover to rotate, and the fog cover is used to shield the upper shell from droplets entering the upper shell; a second servo is arranged on the lower shell, and before the high-definition camera starts working, the first servo rotates the fog cover 180 degrees to cancel the shielding of the upper shell, and the second servo rotates the upper shell to be directly above the lower shell.

[0008] In the above scheme, when the high-definition camera is working, the upper shell and the lower shell are separated by a transparent acrylic plate, a projection film is arranged on the lower side of the transparent acrylic plate, the projection film is facing the high-definition camera, the high-definition camera is arranged in the lower shell, and the high-definition camera is connected to a power supply and a Raspberry Pi; a light source is arranged above the upper shell, and the light emitted by the light source is irradiated on the transparent acrylic plate, presenting an image of fog droplets on the transparent acrylic plate on the projection film.

[0009] In the above scheme, the power supply is powered by connecting to the GPIO power pin of the Raspberry Pi, and then the light source, the first servo, the second servo and the high-definition camera are powered and controlled through the GPIO pin of the Raspberry Pi.

[0010] In the above scheme, the high-definition camera is connected to the Raspberry Pi via a USB interface, and the opencv installed on the Raspberry Pi is used to control the high-definition camera. The boundary between the image droplet outline and the background is enhanced by adjusting the backlight contrast of the high-definition camera to obtain a clear droplet image.

[0011] The above scheme also includes an information processing terminal, which includes parameter input, command sending, data receiving, data sorting, data interface display and data storage. The data interface display includes the number of droplets, droplet coverage, droplet deposition amount, running time, concentration parameter input and particle size distribution diagram.

[0012] In the above scheme, the information processing terminal inputs parameters and transmits them to the Raspberry Pi, which starts to start two servos to convert the upper shell and the lower shell from the droplet receiving state to the droplet detection state. Specifically, the upper shell and the lower shell are closed by the two servos, and the image of the droplets on the projection film is acquired by a high-definition camera. The acquired images are processed, analyzed, calculated and integrated, and the result data including the number of deposited droplets, coverage, deposition amount, operating time and particle size distribution are sent to the Raspberry Pi to complete the detection of pesticide droplet deposition.

[0013] In the above scheme, the process and result acquisition of the Raspberry Pi processing the fog droplet image obtained by the high-definition camera include the following steps:

[0014] (a1) graying the acquired image, denoising the processed image, and finally performing adaptive binarization on the acquired image;

[0015] (a2) Fitting the droplet contours of the processed binary image, placing the fitted image in image 0 for subsequent image processing, using the circularity to preliminarily screen the closed droplet contours in image 0, and screening out the complete single droplet contours and placing them in an empty image 1;

[0016] (a3) The droplet contours in image 0 are filtered out using the circularity to filter out the broken droplet contours (n), and the image centroid (Xc, Yc) is used to calculate and draw a circle at the broken droplet contours to complete them;

[0017] The centroid calculation of the above image is:

[0018]

[0019] For each point (Xi, Yi) on the above contour, use the Euclidean distance formula to calculate the distance Di between it and the centroid (Xc, Yc), calculate the average value Rc of Di, and use (Xc, Yc) as the center and Rc as the radius to fill and repair the broken droplet contour with a solid circle, and put the completed droplet contour in the second image;

[0020]

[0021]

[0022] (a4) Subtract image 1 and image 2 from image 0, and put the resulting droplet contour in image 3. The droplets in image 3 are basically sticky droplets. Use the image segmentation algorithm to process the sticky droplets in image 3, perform polygon fitting, and adjust the distance between points on the polygon. Select three adjacent points P on the polygon one by one. n-1 , P n , P n+1 , point P in the middle of the three points n With the points P on both sides respectively n-1 and P n+1 Draw a perpendicular bisector between the two points and find the point P where the two lines intersect k Coordinates, determine P k Whether it is within the outline of the adhesion droplet. If not, it is a concave point;

[0023] Above Coordinate calculation formula:

[0024]

[0025]

[0026]

[0027]

[0028]

[0029]

[0030] (a5) Calculate the concave points of each droplet contour respectively, and match and segment the concave points to obtain a single complete droplet contour, filter it using circularity, segment the remaining adhering droplets again, iterate until all adhering droplets are segmented, and place the segmented droplets in image 4;

[0031] (a6) Merge images 1 to 4 to obtain the final image. The actual area S of each droplet outline is obtained by converting the pixel points of the final image to the actual area ratio. n , calculate the number of droplet deposition D in the final image n , droplet deposition D d , droplet coverage D k , droplet size distribution D r and the running time of the device for calculation and integration;

[0032] Among them, the droplet deposition amount D d , the actual area S of each droplet profile is known n , droplet contact angle α and solution concentration ρ, calculate the spherical cap cross-sectional radius of each droplet Then calculate the droplet volume The droplet deposition amount Droplet coverage Droplet size The final result is obtained and sent to the information processing terminal for storage and display.

[0033] Beneficial effects:

[0034] 1. The present invention uses Python language to create an information processing terminal and a Raspberry Pi to realize online detection of pesticide droplet deposition. The information processing terminal includes functions of parameter input, command sending, command receiving, data integration, data display and data storage. The information processing terminal is designed as software and can be installed on any display device that supports Python language. After the Raspberry Pi receives the command from the information processing terminal, the servo is started to close the upper shell of the device and turn on the light source, and the deposited droplets on the acrylic plate are clearly projected on the back of the projection film. At the same time, a high-definition camera obtains a picture and transmits the picture to the Raspberry Pi. The Raspberry Pi processes, intercepts, analyzes, calculates and integrates the picture, and sends the number, coverage, deposition amount and particle size distribution of the deposited droplets to the information processing terminal, thereby realizing quality detection of pesticide spray.

[0035] 2. The purpose of the mist shield in the present invention is to prevent mist droplets from entering the upper shell of the device when the device is operating in the field, thereby affecting the accuracy of the light source and mist droplet detection operation.

[0036] 3. In the present invention, after receiving the command from the information processing terminal, the first servo operates to open the fog cover, and then the second servo operates to move the upper shell in a circle until it is closed with the lower shell. The device turns on the light source to project the image of the mist droplets deposited on the acrylic plate, and the outline of the mist droplets is clearly presented on the back of the projection film so that the high-definition camera can capture the picture. The projection film fits tightly with the acrylic plate to ensure the accuracy of the mist droplet image. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the overall design structure of the droplet deposition device provided by the present invention;

[0038] Figure 2 This is a schematic diagram of the droplet deposition device provided by the present invention in an open state;

[0039] Figure 3 The upper shell of the droplet deposition device provided by the present invention;

[0040] Figure 4 The mist shielding cover of the mist droplet deposition device provided by the present invention;

[0041] Figure 5 It is the light source of the droplet deposition device provided by the present invention;

[0042] Figure 6 The acrylic plate of the droplet deposition device provided by the present invention;

[0043] Figure 7 It is a projection film of a droplet deposition device provided by the present invention;

[0044] Figure 8 It is a high-definition camera for a droplet deposition device provided by the present invention;

[0045] Fig. 9 It is the droplet deposition device Raspberry Pi provided by the present invention;

[0046] Fig.10 It is the power supply of the droplet deposition device provided by the present invention;

[0047] Fig.11 It is a steering gear of a droplet deposition device provided by the present invention;

[0048] Fig.12 It is the information processing terminal interface of the droplet deposition device provided by the present invention;

[0049] Fig.13 The invention discloses a method for measuring a droplet deposition device.

[0050] 1-light source; 2-upper shell; 3-first servo; 4-transparent acrylic plate; 5-fog cover; 6-projection film; 7-HD camera; 8-rotating shaft; 9-lower shell; 10-power supply; 11-Raspberry Pi. DETAILED DESCRIPTION

[0051] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0052] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "axial", "radial", "vertical", "horizontal", "inner", "outer" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0053] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0054] The wireless connection indicator of the information processing terminal lights up. Enter the concentration of the measured solution in the concentration input column of the information processing terminal. Click the "Start" button of the information processing terminal. The information processing terminal sends the concentration parameter to the Raspberry Pi. After receiving the command, the Raspberry Pi drives the servo to turn the droplet deposition device from the initial open state to the Figure 2 Transformed into a closed state as shown Figure 1 As shown, turn on the light source 1 and the high-definition camera 7, project the outline of the deposited droplets on the transparent acrylic plate 4 onto the projection film 6, and the high-definition camera 7 performs automatic focus, aperture adjustment, and contrast adjustment, and saves the parameters as environmental variables. After the adjustment, the high-definition camera 7 will obtain a clear and complete image of the droplet outline on the projection film 6, and send it to the Raspberry Pi. After receiving the image, the Raspberry Pi 11 will process, analyze, calculate and integrate it, and send the result data including the number of deposited droplets, coverage, deposition amount, running time and particle size distribution to the information processing terminal. The information processing terminal will integrate and display the acquired data, store the data number, and finally complete the detection of pesticide droplet deposition.

[0055] If an error is encountered during the detection process in the present invention, the "Stop" button of the information processing terminal can be clicked to stop the detection process. Clicking "Start" will continue the detection. Clicking the "Reset" button will reset the device to the initial connection state and clear the information processing terminal data and interface.

[0056] Combined with Figure 1 The figure shows a schematic diagram of the overall design structure of the contoured droplet deposition device, which includes an upper part and a lower part; the upper part includes an upper shell 2, a light source 1, a servo 3 and a fog cover 5; the lower part includes a high-definition camera 7, a power supply 10, a Raspberry Pi 11, a lower shell 9, a transparent acrylic plate 4, a projection film 6 and a rotating shaft 8; the entire device outputs power through the servo.

[0057] Combined with Figure 2 The figure shows the schematic diagram of the profile droplet deposition device in the open state, which is in the working state and waiting for the spray droplets to be deposited on the acrylic plate. After the deposition is completed, the information processing terminal sends a command to the droplet deposition device, and the droplet deposition device runs the program to rotate the steering gear, so that the upper shell is closed. Figure 1state, start the high-definition camera 7 to obtain a picture of the target area, then identify and calculate the contour of the deposited droplets in the picture, and send the calculation result back to the information processing terminal for display and storage.

[0058] The information processing terminal and Raspberry Pi 11 are programmed using Python, and multiple servers are integrated into the information processing terminal through the Socket protocol. The information processing terminal is programmed with a visual operation interface, so that the information processing terminal can send commands and receive data to all HD cameras and Raspberry Pis that are turned on.

[0059] Combined with Figure 3 The upper shell 2 in the contoured droplet deposition device shown is designed with a thin-wall structure to reduce the operating burden of the steering gear. The material is black ABS plastic. In order to meet the light-shielding requirements of the device, the upper shell of the device is produced and processed by 3D printing technology to ensure the accuracy of the processing. The upper shell performs circular motion through the coordinated operation of the steering gear and the rotating shaft.

[0060] Combined with Figure 3 The mist cover 5 in the contoured droplet deposition device shown is connected to the upper shell 2 through a rotating shaft 8, and the mist cover is driven to rotate by a steering gear. It is made of hard PVC plastic material, which is light in weight and has good waterproof and corrosion resistance. Its main function is to effectively prevent droplets from entering the upper shell during field spraying operations, thereby ensuring the accuracy of droplet detection.

[0061] Combined with Figure 5 The light source 1 in the contoured droplet deposition device shown is an LED low-voltage, low-power light source, which is loaded in the middle of the upper shell and adopts a countersunk hole design. A lamp cover is designed for it to prevent dust and water. The LED light source has a long service life, fast response speed, and high safety, and can clearly image the droplets deposited on the acrylic plate on the projection film.

[0062] Combined with Figure 6 The acrylic plate 4 in the contoured droplet deposition device shown has excellent light transmittance, with a light transmittance of more than 92%. It is light in weight and has excellent corrosion resistance. It is mainly used for droplet deposition during pesticide spraying and has excellent droplet adhesion characteristics. The adhered droplets will not flow freely due to the tilt of the acrylic plate, and the morphological characteristics of the droplets after deposition are well preserved, ensuring the accuracy and stability of the detection of deposited droplets.

[0063] Combined with Figure 7 The projection film 6 in the contour droplet deposition device shown can form an image on the back, and can clearly project the outline of the deposited droplets on the transparent acrylic plate onto the projection film, so as to facilitate subsequent image acquisition by a high-definition camera, thereby ensuring the accuracy of the deposited droplet detection.

[0064] Combined with Figure 8 The high-definition camera 7 in the contour droplet deposition device shown in the figure moves the upper shell of the device to the closed state of the device when the device receives the start signal, and then the device turns on the light source 1 and the high-definition camera 7 to take pictures of the droplet contour image on the projection film, and transmits the acquired image to the Raspberry Pi for processing, analysis and calculation, and finally obtains the required result data.

[0065] Combined with Fig. 9 The Raspberry Pi 11 in the contoured droplet deposition device shown in the figure has two Micro USB 2.0 interfaces, a Mini HDMI output interface, etc. In addition, it can connect various peripheral devices through the GPIO interface. After the Raspberry Pi receives the signal, the servo is started to put the device in a closed state, and then the light source 1 and the high-definition camera 7 are turned on. The high-definition camera 7 acquires the picture for processing, analysis, calculation and transmission, and the python language is used for compilation and processing. Finally, the final result of the picture processing is integrated and transmitted to the information processing terminal for display and storage.

[0066] Combined with Fig.10 The power supply 10 in the contoured droplet deposition device shown is the power source of the entire device. It is directly connected to the Raspberry Pi 11 through the GPIO interface for power supply. The power supply 10, two servos, and the high-definition camera 7 are all connected to the 3v and 5v power GPIO interfaces on the Raspberry Pi 11 through DuPont lines for power supply, and the on and off states of the peripheral devices are controlled by high and low levels.

[0067] Combined with Fig.11 The first servo 3 and the second servo in the contoured droplet deposition device shown cooperate with the rotating shaft 8, and their main function is to make the fog cover 5 and the upper shell 2 move in a circular motion after receiving instructions, and operate the device to the working open state and the detection closed state, so as to ensure the accuracy of the device's droplet deposition detection.

[0068] Combined with Fig.12 The information processing terminal in the contour droplet deposition device shown in the figure includes five node droplet quantity, droplet coverage, droplet deposition amount and running time display boxes, five node droplet particle size distribution summary charts, solution concentration input parameter box. And the trigger signal buttons "start", "reset" and "stop" correspond to different device control instructions.

[0069] Combined with Fig.13The measurement method of the droplet deposition device of the present invention is shown. The concentration parameters of the solution are input into the information processing terminal, and then the start button of the information processing terminal is clicked. The information processing terminal transmits the solution concentration parameters to the Raspberry Pi 11 through the wireless network. After receiving the parameters, the Raspberry Pi 11 starts to operate, and the steering gear drive device enters a closed state. The high-definition camera 7 obtains the outline image of the deposited droplets on the projection film and transmits it to the Raspberry Pi 11. The Raspberry Pi 11 performs a series of processing, analysis and calculation on the image, and summarizes the results. Finally, the obtained results are transmitted to the information processing terminal through the wireless network. The information processing terminal summarizes and organizes the received results, processes part of the data, and displays and stores the processed data on the interface for ready call at any time. The above-mentioned entire measurement method is built using the python language, and the transmission is realized through the wireless network.

[0070] The process and result acquisition of the droplet image obtained by the Raspberry Pi HD camera include the following steps:

[0071] (a1) graying the acquired image, denoising the processed image, and finally performing adaptive binarization on the acquired image;

[0072] (a2) Fitting the droplet contours of the processed binary image, placing the fitted image in image 0 for subsequent image processing, using the circularity to preliminarily screen the closed droplet contours in image 0, and screening out the complete single droplet contours and placing them in an empty image 1;

[0073] (a3) The droplet contours in image 0 are filtered out using the circularity to filter out the broken droplet contours (n), and the image centroid (Xc, Yc) is used to calculate and draw a circle at the broken droplet contours to complete them;

[0074] The centroid calculation of the above image is:

[0075]

[0076] For each point (Xi, Yi) on the above contour, use the Euclidean distance formula to calculate the distance Di between it and the centroid (Xc, Yc), calculate the average value Rc of Di, and use (Xc, Yc) as the center and Rc as the radius to fill and repair the broken droplet contour with a solid circle, and put the completed droplet contour in the second image;

[0077]

[0078]

[0079] (a4) Subtract image 1 and image 2 from image 0, and put the resulting droplet contour in image 3. The droplets in image 3 are basically sticky droplets. Use the image segmentation algorithm to process the sticky droplets in image 3, perform polygon fitting, and adjust the distance between points on the polygon. Select three adjacent points P on the polygon one by one. n-1 , P n , P n+1 , point P in the middle of the three points n With the points P on both sides respectively n-1 and P n+1 Draw a perpendicular bisector between the two points and find the point P where the two lines intersect k Coordinates, determine P k Whether it is within the outline of the adhesion droplet. If not, it is a concave point;

[0080] Above Coordinate calculation formula:

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087] (a5) Calculate the concave points of each droplet contour respectively, and match and segment the concave points to obtain a single complete droplet contour, filter it using circularity, segment the remaining adhering droplets again, iterate until all adhering droplets are segmented, and place the segmented droplets in image 4;

[0088] (a6) Merge images 1 to 4 to obtain the final image. The actual area S of each droplet outline is obtained by converting the pixel points of the final image to the actual area ratio. n , calculate the number of droplet deposition D in the final image n , droplet deposition D d , droplet coverage D k , droplet size distribution D r and the running time of the device for calculation and integration;

[0089] Among them, the droplet deposition amount D d , the actual area S of each droplet profile is known n , droplet contact angle α and solution concentration ρ, calculate the spherical cap cross-sectional radius of each droplet Then calculate the droplet volume The droplet deposition amount Droplet coverage Droplet size The final result is obtained and sent to the information processing terminal for storage and display.

[0090] In the present invention, the images of the Raspberry Pi high-definition camera in the contour drop deposition device at multiple locations can be processed and compared, thereby making the experimental results more reliable.

[0091] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0092] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention without departing from the principles and intent of the present invention.

Claims

1. An online detection device for pesticide droplet deposition based on an optical camera, characterized in that: The invention comprises a light source (1), an upper shell (2), a lower shell (9), a transparent acrylic plate (4), a projection module (6), a high-definition camera (7), a Raspberry Pi (11) and a power supply (10); a first steering gear (3) is arranged on the upper shell (2); before the transparent acrylic plate (4) obtains the mist droplets, the first steering gear (3) drives the rotating shaft (8) to rotate, thereby driving the mist cover (5) to rotate; the mist cover (5) is used to shield the upper shell (2) to prevent the mist droplets from entering the upper shell (2); a second steering gear is arranged on the lower shell (9); before the high-definition camera (7) starts to work, the first steering gear (3) rotates the mist cover (5) by 180 degrees to cancel the shielding of the upper shell (2); the second steering gear rotates the upper shell (2) to be directly above the lower shell (9).

2. The on-line detection device for pesticide droplet deposition based on an optical camera according to claim 1, characterized in that: When the high-definition camera (7) is working, the upper shell (2) and the lower shell (9) are separated by a transparent acrylic plate (4), a projection film (6) is arranged on the lower side of the transparent acrylic plate (4), the projection film (6) faces the high-definition camera (7), the high-definition camera (7) is arranged in the lower shell (9), and the high-definition camera (7) is connected to a power supply (10) and a Raspberry Pi (11); a light source (1) is arranged above the upper shell (2), and light emitted by the light source (1) is irradiated on the transparent acrylic plate (4), and a fog drop image on the transparent acrylic plate (4) is presented on the projection film (6).

3. The on-line detection device for pesticide droplet deposition based on an optical camera according to claim 1, characterized in that: The power supply (10) is connected to the GPIO power pin of the Raspberry Pi (11) to supply power, and then supplies power and controls the light source (1), the first servo (3), the second servo and the high-definition camera (8) through the GPIO pin of the Raspberry Pi (11).

4. The on-line detection device for pesticide droplet deposition based on an optical camera according to claim 1, characterized in that: The high-definition camera (7) is connected to the Raspberry Pi (11) via a USB interface, and the opencv installed on the Raspberry Pi (11) is used to control the high-definition camera (7). The boundary between the image droplet outline and the background is strengthened by adjusting the backlight contrast of the high-definition camera (7), so as to obtain a clear droplet image.

5. The on-line detection device for pesticide droplet deposition based on an optical camera according to claim 1, characterized in that: It also includes an information processing terminal, which includes parameter input, command sending, data receiving, data sorting, data interface display and data storage, and the data interface display includes the number of droplets, droplet coverage, droplet deposition amount, running time, concentration parameter input and particle size distribution diagram.

6. The on-line detection device for pesticide droplet deposition based on an optical camera according to claim 5, characterized in that: The information processing terminal inputs parameters and transmits them to the Raspberry Pi (11). The Raspberry Pi (11) starts to activate two steering gears to convert the upper shell (2) and the lower shell (8) from a droplet receiving state to a droplet detection state. Specifically, the upper shell (2) and the lower shell (8) are sealed by the two steering gears, and the image of the droplets on the projection film (6) is captured by a high-definition camera (7). The captured image is processed, analyzed, calculated and integrated, and the result data including the number of deposited droplets, coverage, deposition amount, operating time and particle size distribution are sent to the Raspberry Pi (11), thereby completing the detection of pesticide droplet deposition.

7. The on-line detection device for pesticide droplet deposition based on an optical camera according to claim 6, characterized in that: The process of using a Raspberry Pi (11) to process the fog droplet image obtained by a high-definition camera (7) and the result acquisition includes the following steps: (a1) graying the acquired image, denoising the processed image, and finally performing adaptive binarization on the acquired image; (a2) Fitting the droplet contours of the processed binary image, placing the fitted image in image 0 for subsequent image processing, using the circularity to preliminarily screen the closed droplet contours in image 0, and screening out the complete single droplet contours and placing them in an empty image 1; (a3) The droplet contours in image 0 are filtered out using the circularity to filter out the broken droplet contours (n), and the image centroid (Xc, Yc) is used to calculate and draw a circle at the broken droplet contours to complete them; The centroid calculation of the above image is: For each point (Xi, Yi) on the above contour, use the Euclidean distance formula to calculate the distance Di between it and the centroid (Xc, Yc), calculate the average value Rc of Di, and use (Xc, Yc) as the center and Rc as the radius to fill and repair the broken droplet contour with a solid circle, and put the completed droplet contour in the second image; (a4) Subtract image 1 and image 2 from image 0, and put the resulting droplet contour in image 3. The droplets in image 3 are basically sticky droplets. Use the image segmentation algorithm to process the sticky droplets in image 3, perform polygon fitting, and adjust the distance between points on the polygon. Select three adjacent points P on the polygon one by one. n-1 , P n , P n+1 , point P in the middle of the three points n With the points P on both sides respectively n-1 and P n+1 Draw a perpendicular bisector between the two points and find the point P where the two lines intersect k Coordinates, determine P k Whether it is within the outline of the adhesion droplet. If not, it is a concave point; Above Coordinate calculation formula: (a5) Calculate the concave points of each droplet contour respectively, and match and segment the concave points to obtain a single complete droplet contour, filter it using circularity, segment the remaining adhering droplets again, iterate until all adhering droplets are segmented, and place the segmented droplets in image 4; (a6) Merge images 1 to 4 to obtain the final image. The actual area S of each droplet outline is obtained by converting the pixel points of the final image to the actual area ratio. n , calculate the number of droplet deposition D in the final image n , droplet deposition D d , droplet coverage D k , droplet size D r and the running time of the device for calculation and integration; Among them, the droplet deposition amount D d , the actual area S of each droplet profile is known n , droplet contact angle α and solution concentration ρ, calculate the spherical cap cross-sectional radius of each droplet Then calculate the droplet volume The droplet deposition amount Droplet coverage Droplet size The final result is obtained and sent to the information processing terminal for storage and display.

Images