Plant protection vehicle and method for laser pest killing thereof

Abstract

The application provides a plant protection vehicle and a method for laser pest killing. The plant image in a target crop area is collected by an image acquisition device. The target pest is identified from the collected plant image by a host computer. If the target pest is identified, the target pest is positioned to obtain the positioning information of the target pest, and a pest killing control instruction is sent to a rudder based on the positioning information. The rudder adjusts the rudder freedom platform according to the received pest killing control instruction, controls the laser head to emit laser to the target pest, and realizes laser pest killing. The plant protection vehicle and the laser pest killing method provided by the application integrate pest identification and pest positioning, adjust the laser head on the rudder freedom platform by the rudder, realize accurate positioning of the laser on the target pest, and realize efficient pest control.

Description

Technical Field

[0001] This invention relates to the field of agricultural pest control technology, and more particularly to a plant protection vehicle and a method for achieving laser pest control. Background Technology

[0002] Currently, agricultural pest control generally relies on pesticide spraying, but this method has several problems. For example, pesticide spraying causes environmental pollution, long-term use of the same pesticide can lead to pesticide resistance in pests, and manual spraying results in low operational efficiency. Furthermore, existing laser pest control devices are mostly fixed in form, meaning their position, shooting angle, and laser emission angle are all fixed and cannot be moved or adjusted, thus lacking mobility and precise positioning capabilities.

[0003] Therefore, the existing technology needs further improvement. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this invention is to provide a plant protection vehicle and a method for laser pest control, thereby improving the accuracy and reliability of pest removal.

[0005] In a first aspect, this application discloses a plant protection vehicle for laser pest control, comprising: a support frame, at least one servo motor mounted on the support frame, at least one image acquisition device, and a host computer connected to both the servo motor and the image acquisition device; a laser head is mounted on the servo motor's degree-of-freedom platform.

[0006] The image acquisition device is used to capture crop images of the target crop area;

[0007] The host computer is used to acquire the crop image, identify the crop image, determine whether it contains target pests, and if target pests are identified, send pest control commands to the servo motor.

[0008] The servo motor is used to adjust the servo motor's degree of freedom platform according to the insect extermination control command, and control the laser head to emit laser to the target pest, so as to achieve laser insect extermination.

[0009] Optionally, the image acquisition device is a camera, which is mounted on the top of the bracket and has a narrow-band filter.

[0010] Optionally, the bracket has a long strip-shaped structure, the camera is located in the middle of the bracket, and there are two laser heads, symmetrically arranged on both sides of the camera.

[0011] Optionally, the host computer stores a target object recognition model; the target object recognition model is used to receive the crop image and identify whether the crop image contains a target pest; the target object recognition model is trained based on the YOLOv5 network model.

[0012] Secondly, this application provides a method for laser pest control using an agricultural vehicle, comprising:

[0013] Use an image acquisition device to acquire crop images within the target crop area;

[0014] The host computer is used to identify target pests in the collected crop images. If a target pest is identified, it is located to obtain the location information of the target pest, and then the pest control command is sent to the servo motor based on the location information.

[0015] The servo motor adjusts its degree of freedom platform according to the received insect-killing control command, and controls the laser head to emit laser to the target pest, so as to achieve laser insect killing.

[0016] Optionally, the step of identifying target pests in the acquired images includes:

[0017] The crop image is input into a trained target recognition model, and the output is the recognition result of the target pest; wherein, the target recognition model is trained based on the YOLOv5 network model.

[0018] Optionally, the step of locating the target pest and obtaining its location information includes:

[0019] Obtain the normalized position coordinates of the top left and top right corners of the target bounding box output by the target object recognition model;

[0020] The pixel coordinates of the top left and top right corners of the target bounding box in the crop image are obtained by converting the normalized position coordinates of the top left and top right corners of the target bounding box.

[0021] When the laser head and the camera are on the same plane, the pixel coordinates of the upper left and upper right corners of the target frame are used as the positioning reference to calculate the planar coordinates of the upper left and upper right corners of the target pest in the Cartesian coordinate system.

[0022] Alternatively, when the laser head and the camera are in two different planes, the pixel coordinates of the upper left and upper right corners of the target frame are used as the positioning reference. The homography matrix is ​​used to map the coordinates of the upper left and upper right corners of the target pest in the image coordinate system to the Cartesian coordinate system, so as to obtain the planar coordinates of the upper left and upper right corners of the target pest in the Cartesian coordinate system.

[0023] Optionally, the step of controlling the emission of a laser to the positioning information to eliminate the target pest includes:

[0024] Based on the coordinates of the upper left and upper right corners of the target box, the horizontal swing angle of the laser head is calculated, so that the emitted laser moves between the upper left and upper right corners of the target box within the horizontal swing angle range, forming a linear coverage of the area within the target box;

[0025] The formula for calculating the horizontal swing angle of the laser head is:

[0026]

[0027]

[0028] Where θ is the horizontal swing angle of the laser head when it swings to the upper left corner of the target box, α is the horizontal swing angle of the laser head when it swings to the upper right corner of the target box, and x... left Hy is the x-coordinate of the top-left corner of the target bounding box. top The x-coordinate of the top-left corner of the target bounding box is given by [reference]. right Hy is the x-coordinate of the top right corner of the target bounding box. top The y-coordinate of the top right corner of the target box.

[0029] Optionally, the step of adjusting the servo motor's degree of freedom platform according to the received insect-killing control command, and controlling the laser head to emit laser light towards the target pest, includes:

[0030] The servo motor adjusts the servo motor's degree of freedom platform via PWM pulse width adjustment so that the laser head emits laser light that swings horizontally between angles θ and α.

[0031] The calculation method for the PWM pulse width of the laser at the upper left corner of the target box is as follows:

[0032]

[0033] The method for calculating the PWM pulse width of the laser at the upper right corner of the target box is as follows:

[0034]

[0035] Optionally, before the step of locating the target pest and obtaining its location information, the method further includes:

[0036] Calculate the correction term for the horizontal swing angle of the laser head based on the physical offset of the camera and laser head during installation;

[0037] The horizontal swing angle of the laser head is corrected based on the calculated correction term.

[0038] Beneficial effects:

[0039] This invention provides a plant protection vehicle and a method for laser pest control. The vehicle utilizes an image acquisition device to capture crop images within a target crop area. A host computer identifies target pests from the captured crop images. If a target pest is identified, it is located to obtain its location information. Based on this location information, a pest control command is sent to a servo motor. The servo motor adjusts its degree of freedom platform according to the received pest control command, controlling the laser head to emit laser light towards the target pest, thus achieving laser pest control. The plant protection vehicle and laser pest control method provided by this invention achieve efficient pest control by integrating pest identification and pest location. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the three-dimensional structure of the plant protection vehicle provided by the present invention;

[0041] Figure 2 This is a front view of the plant protection vehicle provided in an embodiment of the present invention;

[0042] Figure 3 This is a front view of the plant protection vehicle provided in an embodiment of the present invention;

[0043] Figure 4 This is a flowchart of the laser insecticidal steps of the method in this embodiment of the invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0045] Current agricultural pest control methods are generally categorized into agricultural control, biological control, physical control, chemical control, and integrated pest management. Agricultural control involves adjusting the internal structure of the agricultural ecosystem, improving the crop's growing environment, and enhancing the crop's own resistance to pests. Biological control utilizes natural enemies and pathogenic microorganisms to control pest populations, offering advantages such as no environmental pollution and no disruption to the ecological balance. Physical control uses physical agents or mechanical means such as light, heat, electricity, and radiation to control pests, offering advantages such as simple operation and no pollution. Chemical control uses chemical pesticides to control pests, offering advantages such as rapid effectiveness and ease of operation; however, long-term use can lead to pesticide resistance in pests and environmental pollution. Integrated pest management combines the above methods to construct a comprehensive pest control system.

[0046] Of the methods mentioned above, agricultural and biological control are environmentally friendly but slow to take effect, while chemical control is fast-acting but pollutes the environment and the pesticides sprayed may damage crops. Therefore, none of the above control methods can simultaneously achieve the effects of fast effect, environmental friendliness, high efficiency and energy saving.

[0047] Existing technologies also disclose methods to attract pests by trapping them, thereby achieving the effect of pest control. However, most pest control devices are fixed and cannot be moved or located, thus lacking the function of real-time pest identification and elimination.

[0048] To achieve rapid and environmentally friendly control of crop pests, this application provides a plant protection vehicle for laser pest control and a method for using the vehicle. The vehicle captures images of crops within a target area, identifies the presence of target pests, and locates their positions. Based on the located pest position, the laser head is adjusted to aim at the target pest, thus eliminating it. This embodiment combines image recognition with target localization for highly efficient and environmentally friendly pest control.

[0049] The following, in conjunction with the accompanying drawings, provides a more detailed description of a laser pest control method and a plant protection vehicle used for laser pest control disclosed in this application.

[0050] Firstly, this application discloses a plant protection vehicle for laser pest control, such as... Figure 1 As shown, the system includes: a support 3, at least one servo motor 1 mounted on the support 3, at least one image acquisition device 4, and a host computer connected to both the servo motor and the image acquisition device; a laser head 2 is mounted on the servo motor's degree-of-freedom platform. At least one servo motor and one image acquisition device are mounted on the support, and the laser head is mounted on the servo motor. The laser head emits a laser beam. When a target pest is irradiated by the laser, the outer surface of the pest rapidly carbonizes, thus achieving the goal of eliminating the pest.

[0051] The image acquisition device 4 is used to capture crop images of the target crop area. The image acquisition device can be a high-definition camera, a smartphone, an action camera, or other device capable of capturing images of the target crop area.

[0052] To acquire detailed information about objects within crop images, one specific implementation uses a 2-megapixel high-definition camera mounted on the top of the agricultural vehicle. This camera can capture real-time images of the target crop area at 30fps. Furthermore, to suppress visible light interference, an 850nm narrowband filter is added to the front of the camera to enhance the contrast of the infrared reflection image, thereby more accurately identifying target objects from the captured crop images.

[0053] The host computer is used to acquire the crop image, identify the crop image, determine whether it contains target pests, and if target pests are identified, send pest control commands to the servo motor.

[0054] The image acquisition device is connected to a host computer. The host computer acquires crop images captured by the image acquisition device and identifies whether target pests are present in the crop images. The communication connection between the image acquisition device and the host computer in this step can be wireless or wired. Wireless communication can be achieved through Wi-Fi or Bluetooth, while wired communication can be achieved through a wired connection.

[0055] The host computer identifies target pests, including aphids and other types of pests, by analyzing captured crop images. Specifically, the host computer is equipped with a pre-trained target identification module for recognizing target pests in images, and this module is used to identify the aforementioned target pests in the crop images.

[0056] The servo motor 1 is used to adjust the servo motor degree of freedom platform according to the insect extermination control command, and control the laser head 2 to emit laser to the target pest, so as to achieve laser insect extermination.

[0057] A servo motor is a high-precision actuator that can quickly and accurately rotate to a specified angle and maintain stability based on input signals, such as PWM (Pulse Width Modulation). It combines a motor, a reduction gear set, a position sensor, and a control circuit, achieving angle positioning through a feedback mechanism. In this embodiment, a servo motor is used to adjust the angle of the laser head, ensuring that the adjusted laser head is aimed at the target pest, thereby achieving rapid elimination of the pest.

[0058] Specifically, the servo motor is equipped with a miniature motor, a reduction gear set, a position sensor, and a control circuit. The servo motor receives PWM pulses from an external controller. The control circuit converts the pulse width into a target angle value. The position sensor detects the current angle and compares it with the target value to generate an error signal. The control circuit adjusts the direction and magnitude of the motor current based on the error, driving the gear set to rotate until the error approaches zero. Once the target angle is reached, the motor stops driving, and the gear set locks the position.

[0059] The value-preserving vehicle disclosed in this embodiment is set up within a target crop area. An image acquisition device (e.g., a camera) on the support frame captures real-time images of the crops within the target crop area. A host computer connected to the image acquisition device acquires the crop images captured by the image acquisition device and uses its internal target object recognition module to identify whether the crop images contain target pests. If so, the host computer locates the position of the target pest in the crop image and sends a pest control command to the servo motor based on the position coordinates of the target pest. The servo motor adjusts the current angle of the laser head according to the target rotation angle value contained in the pest control command, so that the laser head is aligned with the location of the target pest.

[0060] Combination Figure 2 and Figure 3 As shown, to facilitate precise location of the target pest, this embodiment features a long, narrow support structure with the camera positioned in the center. Two servo motors and two laser heads are included, with the laser heads mounted on the servo motors, which are symmetrically positioned on either side of the camera. To allow the laser emitted by the laser heads to rotate and position over a wider area, the servo motors are installed near the edges at both ends of the support structure, thus expanding the scanning area of ​​the two laser heads.

[0061] In order to ensure that the laser emitted by the laser head can be more accurately aligned with the target pest, the laser head and the camera are positioned on the same plane, and the emission direction and the shooting direction are the same.

[0062] Furthermore, the host computer stores a target object recognition model; the target object recognition model is used to receive the crop image and identify whether the crop image contains a target pest; the target object recognition model is trained based on the YOLOv5 network model.

[0063] This embodiment employs a pre-trained target recognition model to identify whether target pests are present in crop images. The network structure of this target recognition model is built upon the YOLOv5 network model and trained using a large number of target pest image samples. In this embodiment, a YOLOv5 classification model is used, trained on a self-collected dataset of over 10,000 images, achieving a recognition accuracy exceeding 96%. The output target bounding box coordinates are referenced to their size within the image.

[0064] Once the target object recognition model outputs the target bounding box locating the target pest, the coordinates corresponding to the target bounding box are transformed to the Cartesian coordinate system. Based on the planar coordinates of the target pest in the Cartesian coordinate system, the servo motor platform is controlled to adjust the rotation angle of the laser head, positioning the laser head on the straight line where the target pest is located, thereby eliminating the target pest.

[0065] Multiple plant protection vehicles provided by this invention can be set up within the target crop area, or only one plant protection vehicle can be set up. For ease of use, a set of wheels can be installed on the frame of the plant protection vehicle to enable it to move within the target crop area, facilitating the elimination of pests throughout the target crop area.

[0066] The plant protection vehicle disclosed in this embodiment has a simple structure and is easy to manufacture. It uses deep learning to identify target pests and utilizes servo motors to position the laser head to target the pests, thus achieving intelligent laser pest control. This not only overcomes the shortcomings of existing pest control devices, such as immobility and high energy consumption, but also demonstrates that the plant protection vehicle provided in this embodiment, by using lasers for pest control, is not only highly efficient but also environmentally friendly and capable of real-time operation, thus possessing high potential for widespread application.

[0067] Secondly, this application provides a method for laser pest control using an agricultural vehicle, such as... Figure 4 As shown, it includes:

[0068] Step S1: Use an image acquisition device to acquire crop images within the target crop area.

[0069] In this step, a high-definition camera is used to take pictures of the target crop area in real time or at fixed intervals to obtain crop images within the target crop area.

[0070] Step S2: Use the host computer to identify target pests in the collected crop images. If a target pest is identified, locate the target pest, obtain the target pest's location information, and send pest control commands to the servo motor based on the location information.

[0071] The host computer identifies target pests in crop images and determines whether there are target pests in the crop images. If any type of target pest is identified, the position of the target pest in the image is located, and then the pest control command is sent to the servo motor based on the location information of the target pest.

[0072] Specifically, the step of identifying target pests from the acquired images includes:

[0073] The crop image is input into a trained target recognition model, and the output is the recognition result of the target pest; wherein, the target recognition model is trained based on the YOLOv5 network model.

[0074] The target pest recognition model processes crop images and outputs the recognition result of the target pest, which is a bounding box containing the location information of the target pest and its confidence score. The bounding box is a rectangular bounding box that marks the location and range of the target pest.

[0075] When training the object recognition model, the input image size was 640×640. The training parameters were: initial learning rate 0.01, SGD optimizer, 100 training cycles, and a final test set accuracy of 96.3%. The target bounding boxes output by the trained object recognition model were in the format of normalized center coordinates and width and height (x_center, y_center, w, h), with a confidence threshold set to 0.7.

[0076] Furthermore, the step of locating the target pest and obtaining its location information includes:

[0077] Step S21: Obtain the normalized position coordinates of the upper left and upper right corners of the target box output by the target object recognition model.

[0078] Step S22: Based on the normalized position coordinates of the upper left and upper right corners of the target bounding box, convert to obtain the pixel coordinates of the upper left and upper right corners of the target bounding box in the crop image.

[0079] Since the target recognition model outputs normalized coordinates (relative to the feature map or input image size), this step requires converting the target box coordinates to pixel coordinates. The conversion method is to multiply the normalized position coordinates by the scaling factor from the feature map to the original image to obtain the actual pixel coordinates. For example, if the feature map size is (13, 13) and the input image size is (416, 416), then the normalized coordinates need to be multiplied by 416 / 13 = 32 to convert them to pixel coordinates.

[0080] The bounding box is labeled with normalized center coordinates and width / height (x_center, y_center, w, h), which needs to be converted to the actual pixel coordinates in the entire image. The following are the conversion methods for the top-left and top-right corner coordinates:

[0081] Assuming the image size is (width, height) and the YOLO annotation value is (x_center, y_center, w, h), the value range of this annotation value is [0, 1].

[0082] Coordinates of the top left corner:

[0083]

[0084] Coordinates of the top right corner:

[0085]

[0086] For example, if the image size is 800×600 and YOLO outputs (0.5,0.5,0.3,0.4), then the top left corner is (280,180) and the top right corner is (520,180).

[0087] Step S23: When the laser head and the camera are on the same plane, the pixel coordinates of the upper left and upper right corners of the target frame are used as the positioning reference to calculate the planar coordinates of the upper left and upper right corners of the target pest in the Cartesian coordinate system.

[0088] Alternatively, when the laser head and the camera are in two different planes, the pixel coordinates of the upper left and upper right corners of the target frame are used as the positioning reference. The homography matrix is ​​used to map the coordinates of the upper left and upper right corners of the target pest in the image coordinate system to the Cartesian coordinate system, so as to obtain the planar coordinates of the upper left and upper right corners of the target pest in the Cartesian coordinate system.

[0089] If the laser head and the camera are on the same plane, then the top left and top right corners of the target box in the Cartesian coordinate system can be obtained directly without coordinate system transformation.

[0090] If the laser head and the camera are not on the same plane, the pixel coordinates in the image coordinate system need to be transformed to the planar coordinates in the Cartesian coordinate system through the homography matrix, so that the laser head can locate the coordinate position.

[0091] In detail, the laser head is located at the top left corner of the image (origin (0,0)), and the target bounding box is located on the same horizontal line as the laser head. The specific steps are as follows:

[0092] Coordinate system transformation:

[0093] Image coordinate system: the origin (0,0) is at the top left corner, the x-axis is to the right, and the y-axis is downward.

[0094] Cartesian coordinate system: the y-axis needs to be reversed (upward is positive), and the coordinates of the target pest are taken from the upper right corner of the target box (x). left ,y top ), converted to (x left Hy top ), where H is the image height.

[0095] Step S3: The servo motor adjusts its degree of freedom platform according to the received insect-killing control command, and controls the laser head to emit laser to the target pest, so as to achieve laser insect killing.

[0096] In this step, the servo receives the insect-killing control command from the host computer, parses the insect-killing control command to extract the target rotation angle of the servo's degree of freedom platform, and controls the rotation of the servo's degree of freedom platform to achieve the horizontal swing angle of the laser head.

[0097] The pest control command is generated based on the planar coordinates of the target pest, which includes the target rotation angle of the servo motor platform. The pest control command also includes: when the servo motor platform rotates to the target rotation angle, the laser head is triggered to emit a laser to kill the target pest.

[0098] Furthermore, the step of controlling the emission of laser light to the positioning information to eliminate the target pests includes:

[0099] Based on the coordinates of the upper left and upper right corners of the target box, the horizontal swing angle of the laser head is calculated, so that the emitted laser moves between the upper left and upper right corners of the target box within the horizontal swing angle range, forming a linear coverage of the area within the target box;

[0100] The formula for calculating the horizontal swing angle of the laser head is:

[0101]

[0102]

[0103] Where θ is the horizontal swing angle of the laser head when it swings to the upper left corner of the target box, α is the horizontal swing angle of the laser head when it swings to the upper right corner of the target box, and x... left Hy is the x-coordinate of the top-left corner of the target bounding box. top The x-coordinate of the top-left corner of the target bounding box is given by [reference]. right Hy is the x-coordinate of the top right corner of the target bounding box. top The y-coordinate of the top right corner of the target box.

[0104] In practical implementation, for the upper left corner of the target frame, the horizontal swing angle θ of the laser head is:

[0105]

[0106] x left The x-coordinate (in pixels) of the target point.

[0107] Hy top : Transformed y-coordinate (in pixels)

[0108] The upper right corner of the target bounding box is the coordinate point of the target pest (x... right ,y top The transformed coordinates are (x) right Hy top ), where H is the image height.

[0109] For the upper right corner of the target bounding box, the horizontal swing angle α of the laser head is:

[0110]

[0111] x right : The x-coordinate (in pixels) of the target point.

[0112] Hy top : Transformed y-coordinate (in pixels).

[0113] With an image height H = 600 and the coordinates of the top left corner point being (280, 180), after coordinate transformation, it becomes (280, 420). Therefore, the horizontal swing angle θ of the laser head is:

[0114]

[0115] Once the horizontal swing angle range of the laser head is calculated, the servo motor platform swings back and forth at a frequency of 10Hz between θ and α according to the PWM value, and the laser focus forms a linear coverage within the target frame area. The laser power is set to no less than 2W (wavelength 980nm) and the pulse frequency is 1kHz to ensure instantaneous carbonization of the insect's exoskeleton.

[0116] Specifically, the step of adjusting the servo motor's degree of freedom platform according to the received insect-killing control command, and controlling the laser head to emit laser light towards the target pest, includes:

[0117] The servo motor adjusts the servo motor's degree of freedom platform via PWM pulse width adjustment so that the laser head emits laser light that swings horizontally between angles θ and α.

[0118] The calculation method for the PWM pulse width of the laser at the upper left corner of the target box is as follows:

[0119]

[0120] The method for calculating the PWM pulse width of the laser at the upper right corner of the target box is as follows:

[0121]

[0122] In this step, angle control commands for the servo motor degree-of-freedom platform are generated based on the located coordinate information. The servo motor controls the adjustment angle of the servo motor degree-of-freedom platform through PWM pulse width. After the adjustment is completed, the host computer controls the laser head to emit laser to eliminate the target pest. In a specific embodiment, the control parameters for adjusting the servo motor degree-of-freedom platform include: PWM period parameter, PWM pulse width corresponding to the upper left corner of the laser positioning target frame, and PWM pulse width corresponding to the upper right corner of the laser positioning target frame.

[0123] In specific implementations, the period of the PWM pulse width is fixed. For example, the PWM period is a fixed 20ms (50Hz), and the pulse width ranges from 0.5ms (0°) to 2.5ms (180°), with a linear mapping.

[0124] The formula for calculating the PWM pulse width corresponding to positioning the laser at the upper left corner of the target box is:

[0125]

[0126] The formula for calculating the PWM pulse width corresponding to the laser positioning at the upper right corner of the target box is:

[0127]

[0128] The formula for converting laser positioning at the top left or top right corner of the target box into a PWM value (assuming a PWM resolution of 12 bits and a range of 0 to 4095) is as follows:

[0129]

[0130] Specifically, the host computer outputs the PWM value to control the servo motor.

[0131] Furthermore, in order to achieve more accurate location of the target pest, before the step of locating the target pest and obtaining its location information, the method further includes:

[0132] Based on the physical offset of the camera and laser head during installation, a correction term for the horizontal swing angle of the laser head is calculated; the horizontal swing angle of the laser head is then corrected based on the calculated correction term.

[0133] Since the camera and laser head will have a physical offset during installation, in order to overcome the positioning deviation caused by the physical offset, this step calculates a correction term to correct the physical offset, thereby improving the positioning accuracy.

[0134] In practical implementation, the calculation principle of coordinate offset compensation is as follows:

[0135] There is a physical offset between the camera and the laser head (Δx = 5cm, Δy = 10cm). After actual measurement using a calibration plate, a correction term is added to the angle calculation.

[0136]

[0137] Where H is the height of the image, y is the y-coordinate of the location, and x is the x-coordinate of the location.

[0138] This invention provides a plant protection vehicle and a method for laser pest control. The vehicle utilizes an image acquisition device to capture crop images within a target crop area. A host computer identifies target pests from the captured crop images. If a target pest is identified, it is located to obtain its location information. Based on this location information, a pest control command is sent to a servo motor. The servo motor adjusts its degree of freedom platform according to the received pest control command, controlling the laser head to emit laser light towards the target pest, thus achieving laser pest control. The plant protection vehicle and laser pest control method provided by this invention achieve efficient pest control by integrating pest identification and pest location.

[0139] This invention achieves precise extermination of agricultural pests through high-precision target detection, rapid coordinate transformation, and dynamic laser control. It is characterized by high efficiency, environmental friendliness, and low power consumption, and is suitable for large-scale smart agriculture scenarios.

[0140] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0141] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0142] It is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present application.

Claims

1. A method for laser pest control using an agricultural vehicle, characterized in that, An application is made to a plant protection vehicle for laser pest control, the plant protection vehicle comprising: a bracket, at least one servo motor mounted on the bracket, at least one image acquisition device, and a host computer connected to both the servo motor and the image acquisition device; a laser head is mounted on the servo motor's degree-of-freedom platform; The method includes: Use an image acquisition device to acquire crop images within the target crop area; The host computer is used to identify target pests in the collected crop images. If a target pest is identified, it is located to obtain the location information of the target pest, and then the pest control command is sent to the servo motor based on the location information. The servo motor adjusts its degree of freedom platform according to the received insect-killing control command, and controls the laser head to emit laser to the target pest, so as to achieve laser insect killing; The steps for identifying target pests from the acquired images include: The crop image is input into a trained target recognition model, and the output is the recognition result of the target pest; wherein, the target recognition model is trained based on the YOLOv5 network model; The steps for locating the target pest and obtaining its location information include: Obtain the normalized position coordinates of the top left and top right corners of the target bounding box output by the target object recognition model; The pixel coordinates of the top left and top right corners of the target bounding box in the crop image are obtained by converting the normalized position coordinates of the top left and top right corners of the target bounding box. When the laser head and the camera are on the same plane, the pixel coordinates of the upper left and upper right corners of the target frame are used as the positioning reference to calculate the planar coordinates of the upper left and upper right corners of the target pest in the Cartesian coordinate system. Alternatively, when the laser head and the camera are in two different planes, the pixel coordinates of the upper left and upper right corners of the target frame are used as the positioning reference. The homography matrix is ​​used to map the coordinates of the upper left and upper right corners of the target pest in the image coordinate system to the Cartesian coordinate system, so as to obtain the planar coordinates of the upper left and upper right corners of the target pest in the Cartesian coordinate system. The step of controlling the emission of laser light to the positioning information to eliminate the target pests includes: Based on the coordinates of the upper left and upper right corners of the target box, the horizontal swing angle of the laser head is calculated, so that the emitted laser moves between the upper left and upper right corners of the target box within the horizontal swing angle range, forming a linear coverage of the area within the target box; The formula for calculating the horizontal swing angle of the laser head is: ； ； in, This represents the horizontal swing angle of the laser head when it swings to the coordinates of the upper left corner of the target frame. This represents the horizontal swing angle of the laser head when it swings to the coordinates of the upper right corner of the target frame. The x-coordinate of the top-left corner of the target box. The y-coordinate of the top-left corner of the target box. The x-coordinate of the top right corner of the target box. The y-coordinate of the top right corner of the target box; The steps of adjusting the servo motor's degree of freedom platform according to the received insect-killing control command, and controlling the laser head to emit laser light to the target pest, include: The servo motor adjusts its degree of freedom platform via PWM pulse width modulation, so that the laser head emits laser light at an angle... and oscillates horizontally between them; The calculation method for the PWM pulse width of the laser at the upper left corner of the target box is as follows: ； The method for calculating the PWM pulse width of the laser at the upper right corner of the target box is as follows: 。 2. The laser insecticidal method according to claim 1, characterized in that, Before the step of locating the target pest and obtaining its location information, the method further includes: Calculate the correction term for the horizontal swing angle of the laser head based on the physical offset of the camera and laser head during installation; The horizontal swing angle of the laser head is corrected based on the calculated correction term.

3. A plant protection vehicle that uses the laser pest control method as described in any one of claims 1-2 for laser pest control, characterized in that, include: A bracket, at least one servo motor mounted on the bracket, at least one image acquisition device, and a host computer connected to both the servo motor and the image acquisition device; A laser head is installed on the servo motor's degree-of-freedom platform; The image acquisition device is used to capture crop images of the target crop area; The host computer is used to acquire the crop image, identify the crop image, determine whether it contains target pests, and if target pests are identified, send pest control commands to the servo motor. The servo motor is used to adjust the servo motor degree of freedom platform according to the insect extermination control command, and control the laser head to emit laser to the target pest, so as to achieve laser insect extermination.

4. The plant protection vehicle for laser pest control according to claim 3, characterized in that, The image acquisition device is a camera, which is mounted on the top of the bracket and has a narrow-band filter.

5. The plant protection vehicle for laser pest control according to claim 3, characterized in that, The bracket has a long strip-shaped structure, the camera is located in the middle of the bracket, and there are two laser heads, which are symmetrically arranged on both sides of the camera.

6. The plant protection vehicle for laser pest control according to claim 3, characterized in that, The host computer stores a target object recognition model; the target object recognition model is used to receive the crop image and identify whether the crop image contains a target pest; the target object recognition model is trained based on the YOLOv5 network model.

Images