WO2023231408A1 - Automatic fruit harvesting apparatus mounted on unmanned aerial vehicle, and control method therefor - Google Patents

Abstract

An automatic fruit harvesting apparatus mounted on an unmanned aerial vehicle, the apparatus comprising: a fruit positioning assembly (01), a control and information transmission device (02) and a picking mechanism (03), wherein the fruit positioning assembly (01) comprises an RGB-D camera (0101), a microprocessor (0102) and a photoelectric sensor (0103); the control and information transmission device (02) comprises an Arduino development board (0201), a relay (0202), an electric motor control board (0203) and a USB-TypeB data cable (0204); and the picking mechanism (03) comprises a battery (0301), an electric motor protective housing (0302), an electric motor (0303), a lead screw (0304), a sleeve (0305), a support rod (0306), a steel bar (0307), and scissors (0309) having a clamping mechanism (0308). When in use, the photoelectric sensor (0103) transmits a signal to the microprocessor (0102) upon detecting that a fruiting mother branch has entered the middle of the scissors (0309); and after receiving the signal from the photoelectric sensor (0103), the microprocessor (0102) runs a fruit positioning program to position fruits and determine whether an unmanned aerial vehicle has reached a destination position. The apparatus is mounted on an unmanned aerial vehicle platform, can adapt to the complex terrain conditions of hilly orchards, and thus effectively solves the problem whereby a picking robot fails to operate in dealing with fruits on tall fruit trees. The present invention further relates to a control method for an automatic fruit harvesting apparatus mounted on an unmanned aerial vehicle.

Description

An automatic fruit harvesting device mounted on a drone and its control method Technical field

The invention belongs to the technical field of agricultural machinery, and in particular relates to an automatic fruit harvesting device mounted on a drone and a control method thereof.

Background technique

Litchi, longan and other bunch-shaped fruits are specialty fruits in tropical and subtropical regions. They have important economic value and are widely planted in the hilly areas of southern China. Since lychees and longans grow in clusters and the fruits are scattered, picking such fruits requires a lot of labor and is costly. Currently, picking bunch-shaped fruits such as lychees and longans is mainly done manually, which is not only laborious and time-consuming, but also as rural labor shifts to non-agricultural industries, the surplus rural labor force is gradually decreasing. During the ripening season, if you encounter hot weather, it is easy for the fruit quality to deteriorate due to delayed picking. Therefore, in order to reduce the cost of picking bunch-shaped fruits such as lychees and longans, it is of economic value to develop agricultural robots that can automatically pick such fruits. Relevant researchers have developed fruit harvesting robots. On this basis, based on the terrain conditions of mountain orchards and the fruit cluster growth characteristics of tall lychee and longan trees, more suitable harvesting robots need to be developed.

In recent years, drones have been widely used in agricultural production, including plant protection, crop monitoring, and crop yield assessment. Compared with ground walking mechanical equipment, UAVs have the advantages of good terrain adaptability and high efficiency. Therefore, UAVs can be used to perform picking tasks in unstructured orchard environments. Research on installing lightweight harvesting devices and their control methods on UAVs is of great significance to the development of automatic fruit picking UAVs.

technical problem

The main purpose of the present invention is to overcome the shortcomings and deficiencies of the existing technology. The present invention provides an automatic fruit harvesting device mounted on a drone and a control method thereof to improve the automation level of bunch-type fruit picking such as lychees and longans.

Technical solutions

In order to achieve the above objects, the present invention adopts the following technical solutions:

On the one hand, the present invention provides an automatic fruit harvesting device mounted on a drone, including a fruit positioning component, control and information transmission equipment and a picking mechanism;

The fruit positioning component includes an RGB-D camera, a microprocessor and a photoelectric sensor; the RGB-D camera is installed in the middle of the support rod of the picking mechanism, and the RGB-D camera simultaneously acquires the color and outline of the target. and location characteristics; the microprocessor is installed at the top of the drone, and the photoelectric sensor includes a diffuse reflection infrared transmitter and an infrared receiver. The infrared rays emitted by the infrared transmitter have the result that the mother branch enters the middle of the scissors. When the infrared receiver receives the diffusely reflected infrared ray, this information is fed back to the microprocessor in the form of a photoelectric signal;

The control and information transmission equipment includes an Arduino development board, a relay, a motor control board and a USB-TypeB data line; the relay is connected to the Arduino development board, and the motor control boards are a power control board and a drive control board respectively. ; The power control board is composed of an electromagnetic switch and a circuit board and is used to control the start and stop of the motor; the drive control board is used to control the input voltage and current of the motor; the USB-TypeB data line is used to Connecting the Arduino development board and the microprocessor enables program downloading and data communication;

The picking mechanism includes a battery, a motor protective shell, a motor, a screw rod, a sleeve, a support rod, a steel bar, and scissors with a clamping mechanism; the battery provides electrical energy for the automatic fruit harvesting device; the motor protection The shell and the support rod are fixedly connected with screws; the motor is connected to one end of the screw rod; the other end of the screw rod is connected to the sleeve; the other end of the sleeve is connected to the steel bar with screws; The other end of the steel bar is connected to one blade of the scissors through screws; the other blade of the scissors is fixedly connected to the support rod through screws; the two clamping pieces of the clamping mechanism on the scissors are respectively connected with the scissors. The two blades are secured with screws.

As a preferred technical solution, the RGB-D camera includes a color camera and an infrared camera; the color camera provides information on three channels of red, green and blue for collecting RGB images; the infrared camera provides A depth information channel is used to collect depth images.

As the preferred technical solution, the microprocessor has built-in memory, NVIDIA Jetson TX2 GPU and 8GB RAM; the memory is a computer-readable storage device, which stores the relevant information of the fruit positioning algorithm written in the Ubuntu18.04 operating system and Python programming language. Program; NVIDIA Jetson TX2 GPU and 8GB RAM are used to execute programs related to the Python programming language.

As a preferred technical solution, the Arduino development board includes an AVR microcontroller, a crystal oscillator or oscillator and a DC power supply, and contains digital input/output pins, analog inputs, crystal oscillator clock, power jack, ICSP connector and reset button.

As a preferred technical solution, the processing core of the AVR microcontroller is ATMEGA328P.

The relay is connected to the Arduino development board through three DuPont lines; the three DuPont lines are respectively connected to the positive and negative poles and signal output ports on the Arduino development board to provide power and input signals to the relay.

As a preferred technical solution, the motor protective shell is made of 3D printed plastic parts; the motor is a DC motor, used to control the movement of a blade on the scissors; the support rod is made of carbon fiber material

On the other hand, the present invention provides a control method for an automatic fruit harvesting device mounted on a drone, which includes the following steps:

The drone drives the automatic fruit harvesting device to fly to the first fruit tree;

Use an RGB-D camera to collect RGB images and depth images of the fruit and input them to the microprocessor;

The blurred image judgment program runs on the microprocessor and only clear images are retained;

The fruit positioning program runs on the microprocessor, and the positioning information is processed and converted into the flight path information of the UAV, which is transmitted to the flight controller to control the UAV to fly to the destination location;

When the photoelectric sensor detects that the fruit-bearing mother branch enters the middle of the scissors, it emits a signal to the microprocessor;

After receiving the signal from the photoelectric sensor, the microprocessor runs the fruit positioning program, locates the fruit again and determines whether the drone has reached the destination location;

The microprocessor sends a running program signal to the Arduino development board. After the Arduino development board runs the program, the start relay switch is closed;

The DC motor starts and links the sleeve through the screw rod. The sleeve drives the steel bar, and the steel bar drives a blade of the scissors to move and cut off the fruit-bearing mother branches;

The clamping device on the scissors clamps the fruit-bearing mother branch and sends a signal to the flight controller;

The drone flies above the fruit collection basket, opens the scissors, and the fruit falls into the fruit collection basket to complete fruit picking.

As a preferred technical solution, a fuzzy image judgment program is run on the microprocessor and only clear images are retained, specifically as follows:

The RGB image collected in real time by the RGB-D camera is processed by the LoG operator and the gray variance of the RGB image is calculated. If the gray variance value is lower than the preset threshold k, it can be automatically determined to be a blurred image and needs to be collected again. The image is then judged until the grayscale variance value of the collected RGB image is higher than the threshold, and then this clear RGB image and depth image can be used as the input image.

As a preferred technical solution, a fruit positioning program is run on the microprocessor to process the positioning information and convert it into UAV flight path information, specifically as follows:

After running the fruit positioning program on the microprocessor, the maximum circumscribed rectangular boxes of multiple candidate fruits are output and the corresponding confidence values are given, and the areas marked by all rectangular boxes with confidence levels greater than 0.75 are saved to obtain the RGB location of the fruit. Pixel coordinates in the image; map this coordinate to the optimized depth image to extract the spatial information of the fruit. By fusing the pixel coordinates of the fruit in the RGB image and the depth information extracted on the depth image, the RGB-D shape of the fruit is obtained. The spatial coordinates in the camera coordinate system; through the positional relationship between the RGB-D camera and the center of the drone, the coordinates in the RGB-D camera coordinate system are converted into coordinates in the drone coordinate system, and the drone can be calculated The flight destination coordinates of the aircraft.

As a preferred technical solution, the DC motor is started, and the sleeve is linked through the screw rod. The sleeve drives the steel bar, and the steel bar drives the blade on one side of the scissors to move to cut off the fruit-bearing mother branches, specifically:

The DC motor starts to rotate after starting, driving the screw rod to move in a circle. The screw rod and the sleeve are connected through threads. The sleeve moves horizontally on the screw rod. The sleeve and the steel bar are fixed together. The steel bar and the scissors are connected. The blade on one side is fixed. At this time, the blade on one side of the scissors makes an arc-like motion and closes with the fixed blade on the other side of the scissors to cut off the fruit-bearing mother branch.

beneficial effects

Compared with the existing technology, the present invention has the following advantages and beneficial effects:

(1) The present invention is mounted on a drone platform and can adapt to the complex terrain conditions of hilly orchards, effectively solving the problem that picking robots cannot work when facing fruits on tall fruit trees.

(2) The present invention combines an RGB-D camera, a microprocessor, an Arduino development board and a picking mechanism to realize automatic positioning and automatic control of picking of fruits, providing a foundation for the development of automatic picking drones.

(3) The picking mechanism proposed by the present invention is equipped with a clamping device, which can cut off the fruit branches while preventing the fruits from falling, and can effectively avoid damage caused by the fruits falling after being picked.

Description of the drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without exerting creative efforts.

Figure 1 is an overall schematic diagram of the automatic fruit harvesting device according to Embodiment 1 of the present invention;

Figure 2 is a schematic structural perspective view of the automatic fruit harvesting device according to Embodiment 1 of the present invention;

Figure 3 is an overall schematic diagram of the automatic fruit harvesting device mounted on a drone according to Embodiment 1 of the present invention;

Figure 4 is a flow chart of the control method of the automatic fruit harvesting device in Embodiment 2 of the present invention.

Explanation of reference numbers:

01. Fruit positioning component; 02. Control and information transmission equipment; 03. Picking mechanism; 0101. RGB-D camera; 0102. Microprocessor; 0103. Photoelectric sensor; 0201. Arduino development board; 0202. Relay; 0203. Motor Control board; 0204, USB-TypeB data cable; 0301, battery; 0302, motor protective case; 0303, motor; 0304, screw rod; 0305, sleeve; 0306, support rod; 0307, steel bar; 0308, belt clamp Holding mechanism; 0309, scissors.

Embodiments of the invention

In order to enable those in the technical field to better understand the solution of the present application, the technical solution in the embodiment of the present application will be clearly and completely described below in conjunction with the drawings in the embodiment of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the scope of protection of this application.

In this application, unless otherwise expressly stated or limited, the terms "mounted", "connected", "connected", "fixed" and other terms should be interpreted broadly. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, or it can be an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be an internal connection between two components. Connectivity can also be surface contact only. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

Example 1 :

As shown in Figure 1, this embodiment provides an automatic fruit harvesting device mounted on a drone, including a fruit positioning component 01, a control and information transmission device 02, and a picking mechanism 03.

Further, as shown in Figure 2, the fruit positioning component 01 includes an RGB-D camera 0101, a microprocessor (host computer) 0102 and a photoelectric sensor 0103; the RGB-D camera 0101 is installed on the support of the picking mechanism. The middle position of pole 0306 is composed of a color camera and an infrared camera; the color camera provides information on three channels of red, green and blue, used to collect RGB images; the infrared camera provides a depth information channel, which is used to collect RGB images. Used to collect depth images; the RGB-D camera 0101 can simultaneously obtain the color, contour and position characteristics of the target; the microprocessor 0102 is installed at the top of the drone, with built-in memory, NVIDIA Jetson TX2 GPU and 8GB RAM; the memory is a computer-readable storage device, which stores the relevant programs of the fruit positioning algorithm written in the Ubuntu18.04 operating system and Python programming language; the NVIDIA Jetson TX2 GPU and 8GB RAM are used to execute related programs in Python programming language; the photoelectric sensor 0103 contains a diffuse reflection infrared transmitter and an infrared receiver. The infrared rays emitted by the infrared transmitter have results when the mother branch enters the middle of the scissors 0309 , the infrared receiver will receive the diffusely reflected infrared rays and feedback this information to the microprocessor 0102 in the form of photoelectric signals.

Further, as shown in Figure 2, the control and information transmission device 02 includes an Arduino development board 0201, a relay 0202, a motor control board 0203, a USB-TypeB data line 0204 and related circuit lines; the Arduino development board 0201 It consists of an AVR microcontroller, a crystal oscillator or oscillator and a DC power supply. It contains digital input/output pins (6 of which can be used as PWM output), analog input, crystal oscillator clock, power jack, ICSP connector and reset button; The processing core of the AVR microcontroller is ATMEGA328P; the relay 0202 is an electrical control device, an automatic switch that uses small current to control large current operation. It is connected to the Arduino development board 0201 through three DuPont lines; Three DuPont lines are connected to the positive and negative poles and signal output ports on the Arduino development board 0201 respectively, and are used to provide power and input signals to the relay 0202; the motor control board 0203 is the power control board and the drive control board respectively; The power control board is composed of an electromagnetic switch and a circuit board and is used to control the start and stop of the motor 0303; the drive control board is used to control the input voltage and current of the motor 0303; the USB-TypeB data line 0204 Used to connect the Arduino development board 0201 and the microprocessor 0102 to enable program downloading and data communication; the related circuit lines provide power and data transmission for the entire control device.

Further, as shown in Figure 2, the picking mechanism 03 includes a battery 0301, a motor protective shell 0302, a motor 0303, a screw rod 0304, a sleeve 0305, a support rod 0306, a steel bar 0307, and scissors with a clamping mechanism 0308. 0309 consists of; the battery 0301 provides electric energy for the automatic fruit harvesting device; the motor protective shell 0302 and the support rod 0306 are fixedly connected by screws; the motor 0303 is connected to one end of the screw 0304; the wire The other end of the rod 0304 is connected to the sleeve 0305; the other end of the sleeve 0305 is connected to the steel bar 0307 with screws; the other end of the steel bar 0307 is connected to a blade of the scissors 0309 through screws; The other blade of the scissors 0309 is fixedly connected to the support rod 0306 through screws; the two clamping pieces of the clamping mechanism 0308 on the scissors 0309 are respectively fixed with the two blades of the scissors 0309 with screws; the motor protective shell 0302 is made of 3D printed plastic parts; the motor 0303 is a DC motor, used to control the movement of the blade on the scissors 0309; the support rod 0306 is made of carbon fiber material.

The overall schematic diagram of the automatic fruit harvesting device mounted on a drone is shown in Figure 3.

Example 2:

This embodiment provides a control method for an automatic fruit harvesting device mounted on a drone, as shown in Figure 4, including the following steps:

(1) The drone drives the automatic fruit harvesting device to fly to the first fruit tree, uses the RGB-D camera 0101 to collect the RGB image and depth image of the fruit, and inputs them to the microprocessor 0102;

(2) Run the blurred image judgment program on the microprocessor 0102 to retain only clear images. The specific method is:

The RGB image collected in real time by the RGB-D camera 0101 is processed by the LoG operator and the grayscale variance of the RGB image is calculated. If the grayscale variance value is lower than the preset threshold k, it can be automatically determined as a blurred image and needs to be re- Collect the image and then make a judgment. Until the grayscale variance value of the collected RGB image is higher than the threshold, the clear RGB image and depth image can be used as the input image.

(3) The fruit positioning program runs on the microprocessor 0102, which processes the positioning information and converts it into the flight path information of the UAV, and transmits it to the flight controller to control the UAV to fly to the destination location. The specific method is:

After running the fruit positioning program on the microprocessor 0102, the maximum circumscribed rectangular boxes of multiple candidate fruits are output and the corresponding confidence values are given, and the areas marked by all rectangular boxes with confidence levels greater than 0.75 are saved to obtain the location of the fruit. Pixel coordinates in the RGB image; map this coordinate to the optimized depth image to extract the spatial information of the fruit. By fusing the pixel coordinates of the fruit in the RGB image and the depth information extracted on the depth image, the RGB- The spatial coordinates in the coordinate system of the D camera 0101; through the positional relationship between the RGB-D camera 0101 and the center of the drone, the coordinates in the coordinate system of the RGB-D camera 0101 are converted into coordinates in the coordinate system of the drone, which can be calculated Get the flight destination coordinates of the drone.

(4) When the photoelectric sensor 0103 detects that the fruit-bearing mother branch enters the middle of the scissors 0309, it emits a signal to the microprocessor 0102;

(5) After receiving the signal from the photoelectric sensor 0103, the microprocessor 0102 runs the fruit positioning program, locates the fruit again and determines whether the drone has reached the destination location;

(6) The microprocessor 0102 sends a running program signal to the Arduino development board 0201. After the Arduino development board 0201 runs the program, the start relay 0202 switch is closed;

(7) The DC motor 0303 starts and links the sleeve 0305 through the screw rod 0304. The sleeve 0305 drives the steel bar 0307, and the steel bar 0307 drives a blade of the scissors 0309 to move to cut off the fruit-bearing mother branches. The specific method is:

DC motor 0303 starts to rotate and drives the screw rod 0304 to move in a circle. The screw rod 0304 and the sleeve 0305 are connected through threads. The sleeve 0305 moves horizontally on the screw rod 0304. The sleeve 0305 is fixed to the steel bar 0307. Together, the steel bar 0307 and the blade on one side of the scissors 0309 are fixed. At this time, the blade on one side of the scissors 0309 makes an arc-like motion and closes with the fixed blade on the other side of the scissors 0309 to cut off the fruit-bearing mother branch.

(8) The clamping mechanism 0308 on the scissors 0309 clamps the fruit-bearing mother branch and sends a signal to the flight controller;

(9) The drone flies above the fruit collection basket, opens the scissors 0309, and the fruit falls into the fruit collection basket, completing fruit picking.

The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, etc. may be made without departing from the spirit and principles of the present invention. All simplifications should be equivalent substitutions, and are all included in the protection scope of the present invention.

Claims (10)

1. An automatic fruit harvesting device mounted on a drone, which is characterized by including a fruit positioning component, control and information transmission equipment and a picking mechanism;

   The fruit positioning component includes an RGB-D camera, a microprocessor and a photoelectric sensor; the RGB-D camera is installed in the middle of the support rod of the picking mechanism, and the RGB-D camera simultaneously acquires the color and outline of the target. and location characteristics; the microprocessor is installed at the top of the drone, and the photoelectric sensor includes a diffuse reflection infrared transmitter and an infrared receiver. The infrared rays emitted by the infrared transmitter have the result that the mother branch enters the middle of the scissors. When the infrared receiver receives the diffusely reflected infrared ray, this information is fed back to the microprocessor in the form of a photoelectric signal;

   The control and information transmission equipment includes an Arduino development board, a relay, a motor control board and a USB-TypeB data line; the relay is connected to the Arduino development board, and the motor control boards are a power control board and a drive control board respectively. ; The power control board is composed of an electromagnetic switch and a circuit board and is used to control the start and stop of the motor; the drive control board is used to control the input voltage and current of the motor; the USB-TypeB data line is used to Connecting the Arduino development board and the microprocessor enables program downloading and data communication;

   The picking mechanism includes a battery, a motor protective shell, a motor, a screw rod, a sleeve, a support rod, a steel bar, and scissors with a clamping mechanism; the battery provides electrical energy for the automatic fruit harvesting device; the motor protection The shell and the support rod are fixedly connected with screws; the motor is connected to one end of the screw rod; the other end of the screw rod is connected to the sleeve; the other end of the sleeve is connected to the steel bar with screws; The other end of the steel bar is connected to one blade of the scissors through screws; the other blade of the scissors is fixedly connected to the support rod through screws; the two clamping pieces of the clamping mechanism on the scissors are respectively connected with the scissors. The two blades are secured with screws.
2. An automatic fruit harvesting device mounted on a drone according to claim 1, characterized in that the RGB-D camera includes a color camera and an infrared camera; the color camera provides red, green, and blue Three channels of information are used to collect RGB images; the infrared camera provides a depth information channel for collecting depth images.
3. An automatic fruit harvesting device mounted on a drone according to claim 1, characterized in that the microprocessor has built-in memory, NVIDIA Jetson TX2 GPU and 8GB RAM; the memory is a computer-readable storage device, and The related programs of the fruit positioning algorithm written in the Ubuntu18.04 operating system and Python programming language are stored; the NVIDIA Jetson TX2 GPU and 8GB RAM are used to execute the related programs of the Python programming language.
4. An automatic fruit harvesting device mounted on a drone according to claim 1, characterized in that the Arduino development board includes an AVR microcontroller, a crystal oscillator or oscillator and a DC power supply, and contains digital input/ Output pins, analog inputs, crystal clock, power jack, ICSP header and reset button.
5. An automatic fruit harvesting device mounted on a drone according to claim 1, characterized in that the processing core of the AVR microcontroller is ATMEGA328P.

   The relay is connected to the Arduino development board through three DuPont lines; the three DuPont lines are respectively connected to the positive and negative poles and signal output ports on the Arduino development board to provide power and input signals to the relay.
6. An automatic fruit harvesting device mounted on a drone according to claim 1, characterized in that the motor protective shell is made of 3D printed plastic parts; the motor is a DC motor used to control the scissors. The movement of a blade; the support rod is made of carbon fiber material.
7. A control method for an automatic fruit harvesting device mounted on a drone according to any one of claims 1-6, characterized in that it includes the following steps:

   The drone drives the automatic fruit harvesting device to fly to the first fruit tree;

   Use an RGB-D camera to collect RGB images and depth images of the fruit and input them to the microprocessor;

   The blurred image judgment program runs on the microprocessor and only clear images are retained;

   The fruit positioning program runs on the microprocessor, and the positioning information is processed and converted into the flight path information of the UAV, which is transmitted to the flight controller to control the UAV to fly to the destination location;

   When the photoelectric sensor detects that the fruit-bearing mother branch enters the middle of the scissors, it emits a signal to the microprocessor;

   After receiving the signal from the photoelectric sensor, the microprocessor runs the fruit positioning program, locates the fruit again and determines whether the drone has reached the destination location;

   The microprocessor sends a running program signal to the Arduino development board. After the Arduino development board runs the program, the start relay switch is closed;

   The DC motor starts and links the sleeve through the screw rod. The sleeve drives the steel bar, and the steel bar drives a blade of the scissors to move and cut off the fruit-bearing mother branches;

   The clamping device on the scissors clamps the fruit-bearing mother branch and sends a signal to the flight controller;

   The drone flies above the fruit collection basket, opens the scissors, and the fruit falls into the fruit collection basket to complete fruit picking.
8. The control method according to claim 7, characterized in that a fuzzy image judgment program is run on the microprocessor to retain only clear images, specifically:

   The RGB image collected in real time by the RGB-D camera is processed by the LoG operator and the gray variance of the RGB image is calculated. If the gray variance value is lower than the preset threshold k, it can be automatically determined to be a blurred image and needs to be collected again. The image is then judged until the grayscale variance value of the collected RGB image is higher than the threshold, and then this clear RGB image and depth image can be used as the input image.
9. The control method according to claim 7, characterized in that a fruit positioning program is run on the microprocessor to process the positioning information and convert it into UAV flight path information, specifically:

   After running the fruit positioning program on the microprocessor, the maximum circumscribed rectangular boxes of multiple candidate fruits are output and the corresponding confidence values are given, and the areas marked by all rectangular boxes with confidence levels greater than 0.75 are saved to obtain the RGB location of the fruit. Pixel coordinates in the image; map this coordinate to the optimized depth image to extract the spatial information of the fruit. By fusing the pixel coordinates of the fruit in the RGB image and the depth information extracted on the depth image, the RGB-D shape of the fruit is obtained. The spatial coordinates in the camera coordinate system; through the positional relationship between the RGB-D camera and the center of the drone, the coordinates in the RGB-D camera coordinate system are converted into coordinates in the drone coordinate system, and the drone can be calculated The coordinates of the aircraft’s flight destination.
10. The control method according to claim 7, characterized in that the DC motor is started to link the sleeve through the screw rod, the sleeve drives the steel bar, and the steel bar drives one side blade of the scissors to move to cut off the fruit-bearing mother branch. Specifically:

    The DC motor starts to rotate after starting, driving the screw rod to move in a circle. The screw rod and the sleeve are connected through threads. The sleeve moves horizontally on the screw rod. The sleeve and the steel bar are fixed together. The steel bar and the scissors are connected. The blade on one side is fixed. At this time, the blade on one side of the scissors makes an arc-like motion and closes with the fixed blade on the other side of the scissors to cut off the fruit-bearing mother branch.