An adaptive paddy field weeding robot based on machine vision and its working method
The machine vision-based adaptive paddy field weeding robot achieves accurate identification and adaptive adjustment of weeds and crops, solving the problems of low identification accuracy, weak adaptive ability, easy damage to seedlings, and easy entanglement of weeds in existing technologies, and achieving efficient and pollution-free paddy field weeding effect.
Patent Information
- Application Number
- CN202611470629.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-21
- Publication Date
- 2026-11-17
AI Technical Summary
Existing methods for weeding paddy fields suffer from problems such as low recognition accuracy, weak adaptability, easy damage to seedlings, easy entanglement with weeds, poor operational stability, and environmental pollution caused by chemical weeding.
An adaptive paddy field weeding robot based on machine vision is adopted. It combines a machine vision recognition module, an adaptive weeding actuator and a main control system to achieve accurate identification and adaptive adjustment of weeds and crops. It is equipped with flexible weeding teeth, depth limit plates and anti-entanglement covers to avoid damaging seedlings and getting tangled in weeds.
It achieves high-precision, pollution-free weeding in paddy fields, protects the ecological environment, reduces labor costs and intensity, and improves the continuity and reliability of operations.
Smart Images

Figure CN123100387A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of agricultural robot technology, and in particular to an adaptive paddy field weeding robot based on machine vision and its working method. Background Technology
[0002] Currently, weed control in paddy fields mainly relies on three methods: manual weeding, mechanical weeding, and chemical weeding. With the development of smart agriculture, although intelligent weeding equipment based on machine vision has emerged, existing technologies still have many shortcomings.
[0003] Traditional manual weeding methods are extremely labor-intensive, inefficient, and costly, making them unsuitable for large-scale planting. While chemical herbicides are more efficient, long-term overuse can lead to herbicide resistance in weeds, cause serious pesticide residues, soil and water pollution, damage the ecological environment, and ultimately affect the quality and safety of agricultural products such as rice. Existing mechanical weeding equipment, such as some rotary or shovel weeders, suffers from the following problems in practical applications: First, their fixed mechanical structures cannot adaptively adjust to the actual plant and row spacing of crops in the field, resulting in poor row accuracy and easy damage to crop seedlings growing close to weeds; second, in the muddy and complex environment of paddy fields, traditional mechanical weeding components are easily entangled by weeds and straw, leading to downtime or damage, and poor operational continuity and stability; third, most traditional mechanical weeding equipment cannot accurately distinguish between weeds and crops, resulting in indiscriminate and indiscriminate operation, poor adaptability, and difficulty in guaranteeing the quality of operation.
[0004] Therefore, there is an urgent need in this field to develop an intelligent paddy field weeding robot that can achieve high-precision recognition, adaptive adjustment of operating parameters, anti-weed entanglement function, and pollution-free operation, in order to solve the comprehensive problems of low recognition accuracy, weak adaptive ability, easy damage to seedlings, easy entanglement with weeds, poor operation stability, and environmental pollution caused by chemical weeding in the existing technology. Summary of the Invention
[0005] The purpose of this invention is to provide an adaptive paddy field weeding robot based on machine vision and its working method, so as to solve the problems existing in the prior art.
[0006] To achieve the above objectives, the present invention provides the following solution: This invention provides an adaptive paddy field weeding robot based on machine vision, comprising: frame; A machine vision recognition module is installed at the front of the frame and is used to acquire field images in real time and identify weeds and crops; it includes a vision recognition module, an image processing unit, and a weed location unit, which are electrically connected in sequence. A walking drive device, which is connected to the bottom of the frame, is used to drive the weeding robot to move in the paddy field; it includes a drive motor, a transmission belt, and walking wheels; An adaptive weeding actuator is connected to the middle of the frame and is used to perform weeding actions. It includes a weeding wheel, a position adjustment mechanism, flexible weeding teeth, and a depth limiting plate. The drive motor is connected to the weeding wheel via the transmission belt. The position adjustment mechanism is connected to the weeding wheel. The flexible weeding teeth are disposed on the surface of the weeding wheel, and the depth limiting plate is disposed below the weeding wheel. The main control system is fixed on the frame and electrically connected to the machine vision recognition module, the walking drive device and the adaptive weeding actuator.
[0007] Preferably, the frame is also equipped with a seedling protection gantry, an anti-tangling cover, and an emergency stop switch. The seedling protection gantry is used to keep crop seedlings within the gantry space to avoid being crushed. The anti-tangling cover is installed on the outside of the adaptive weeding actuator to prevent weeds from tangling. The emergency stop switch is electrically connected to the main control system.
[0008] Preferably, the position adjustment mechanism drives the weeding wheel to move laterally and / or longitudinally, for adjusting the working distance and soil entry position of the weeding wheel.
[0009] Preferably, the adaptive weeding actuator is provided in three groups, and each group of the adaptive weeding actuator is provided with a weeding wheel, a drive motor, a transmission belt, a position adjustment mechanism and a depth limiting plate.
[0010] Preferably, the flexible weeding teeth are made of a wear-resistant flexible material.
[0011] Preferably, the visual recognition module is positioned facing forward, and the image processing unit and the weed positioning unit are integrated inside the frame.
[0012] Preferably, the walking drive device further includes a pulley, and the transmission belt cooperates with the pulley to provide stable walking drive force in muddy terrain.
[0013] Preferably, the weeding robot further includes a solar power panel and a battery. The solar power panel is fixed above the frame and electrically connected to the battery. The battery is electrically connected to the main control system to provide power.
[0014] Preferably, the rack is further provided with a power supply interface and a data transmission interface. The power supply interface is used to connect to an external power source, and the data transmission interface is used to communicate with a remote control terminal.
[0015] This invention also provides a working method for an adaptive paddy field weeding robot based on machine vision, comprising: The walking drive device drives the weeding robot to move in the paddy field. The machine vision recognition module collects field images in real time and distinguishes between crop plants and weeds through the image processing unit. The weed positioning unit completes the positioning and automatic row alignment of weeds. Based on the identification and positioning results, the main control system controls the position adjustment mechanism to automatically adjust the spacing, rotation speed, and soil penetration depth of the weeding wheels; The weeding wheel drives the flexible weeding teeth to rotate to remove weeds, while the depth limiting plate controls the penetration depth of the flexible weeding teeth into the soil, and the anti-tangling cover prevents weeds and straw from getting tangled in the mechanism, thus completing the adaptive and precise weeding operation.
[0016] The present invention achieves the following beneficial technical effects compared to the prior art: This invention provides an adaptive paddy field weeding robot based on machine vision and its working method. Through the collaborative work of the machine vision recognition module, the main control system, and the adaptive weeding execution mechanism, it can identify weeds and crops in real time and automatically align them. The main control system drives the position adjustment mechanism to automatically adjust the spacing, speed, and penetration depth of the weeding wheels according to the recognition results. This achieves fully automatic, high-precision, and targeted physical weeding for paddy fields with different varieties and planting patterns. It not only completely avoids the environmental pollution and pesticide residue problems caused by chemical herbicides, protecting the ecological environment and improving crop quality, but also effectively prevents damage to seedlings and roots through the seedling protection gantry and depth limit plate. Combined with the anti-entanglement cover, it solves the stubborn problem of weed entanglement, greatly improves the continuity and reliability of the operation, significantly reduces labor costs and labor intensity, and effectively solves the problems of poor adaptability, easy damage to seedlings, easy entanglement of weeds, and unstable operation in the prior art. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a schematic diagram of the structure of the present invention; Figure 2 This is a schematic diagram of the frame and overall layout structure of the present invention; Figure 3 This is a schematic diagram of the adaptive weeding actuator of the present invention; In the diagram: 1. Frame; 2. Machine vision recognition module; 3. Weeding wheel; 4. Drive motor; 5. Conveyor belt; 6. Vision recognition module; 7. Image processing unit; 8. Weed positioning unit; 9. Battery; 10. Power supply interface; 11. Emergency stop switch; 12. Pulley; 13. Flexible weeding teeth; 14. Position adjustment mechanism; 15. Depth limit plate; 16. Anti-tangling cover; 17. Weeding outlet; 18. Solar power panel; 19. Seedling protection gantry. Detailed Implementation
[0019] Unless otherwise specified, the terms "connection" and "linkage" used in this application include both direct and indirect connections (linkages). In the description of this invention, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention.
[0020] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.
[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0022] The purpose of this invention is to provide an adaptive paddy field weeding robot based on machine vision and its working method, so as to solve the problems existing in the prior art.
[0023] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0024] Example 1: Please see Figures 1 to 3 This invention provides an adaptive paddy field weeding robot based on machine vision. The weeding robot includes a frame 1, a machine vision recognition module, a walking drive device, an adaptive weeding execution mechanism, and a main control system. The frame 1 serves as the foundation for the entire robot, with a compact overall layout and stable center of gravity, enabling it to adapt well to the complex environment of paddy fields. The machine vision recognition module is installed at the front of the frame 1 and is used to acquire real-time images of the paddy field during operation, quickly and accurately identifying weeds and crops such as rice. The walking drive device is connected to the bottom of the frame 1 and is used to drive the entire weeding robot to move stably in the muddy environment of the paddy field. The adaptive weeding execution mechanism is connected to the middle of the frame 1 and is the core component for performing weeding actions. The main control system is fixed to the frame 1 and is electrically connected to the machine vision recognition module, the walking drive device, and the adaptive weeding execution mechanism, respectively, for receiving and processing information and issuing control commands.
[0025] Furthermore, to achieve precise visual recognition, the machine vision recognition module specifically includes a visual recognition module 2, an image processing unit 7, and a weed positioning unit 8. The visual recognition module 2 is positioned facing forward to ensure optimal field coverage. The image processing unit 7 and the weed positioning unit 8 are integrated inside the frame 1, and are electrically connected sequentially. The visual recognition module 2 transmits the acquired image data to the image processing unit 7, which analyzes the images to distinguish between crop plants and weeds. The weed positioning unit 8 then accurately locates and automatically rows the weeds based on the analysis results. The main control system controls subsequent weeding actions based on the position information transmitted by the weed positioning unit 8.
[0026] To achieve stable movement in paddy field environments, the walking drive device includes a drive motor 4, a transmission belt 5, and walking wheels. In this embodiment, the walking wheels are also the weeding wheels 3 used for subsequent weeding, and simultaneously serve as walking wheels, achieving an integrated design for both walking and weeding functions. Please refer to [link / reference]. Figure 1 The drive motor 4 is fixed inside the frame 1, and the pulley 12 is fixedly connected to the shaft of the weeding wheel 3. The pulley 12 is connected to the output shaft of the drive motor 4 via a transmission belt 5, thereby driving the weeding wheel 3 to rotate. This belt drive method provides stable driving force in muddy and slippery paddy field terrain, reducing slippage. Furthermore, for ease of debugging and emergency handling, the frame 1 is also equipped with an emergency stop switch 11 electrically connected to the main control system.
[0027] Please refer to this carefully. Figure 2 and Figure 3The adaptive weeding actuator is the core improvement of this invention. This mechanism includes a weeding wheel 3, a position adjustment mechanism 14, flexible weeding teeth 13, and a depth limiting plate 15. As mentioned earlier, the drive motor 4 is connected to the weeding wheel 3 via a transmission belt 5, causing the weeding wheel 3 to rotate and weed while moving. The position adjustment mechanism 14 is connected to the weeding wheel 3 and electrically connected to the main control system. It can move the weeding wheel 3 laterally and longitudinally according to the instructions of the main control system, thereby automatically adjusting the working distance and soil entry position of the weeding wheel 3 to adapt to different plant and row spacings for different rice varieties. The flexible weeding teeth 13 are evenly distributed on the surface of the weeding wheel 3. These flexible teeth are made of wear-resistant flexible material, and their length and hardness are specially configured to remove weeds by the roots during rotation, while avoiding damage to rice seedlings and effectively preventing weed stalks from tangling on the weeding wheel 3. A depth limiting plate 15 is positioned below the weeding wheel 3 to limit the penetration depth of the flexible weeding teeth 13 during operation, preventing damage to the crop roots due to excessive penetration. Furthermore, an anti-tangling cover 16 is installed on the outside of the adaptive weeding actuator to further prevent detached weeds or straw from tangling around the mechanism's shafts and other parts, significantly improving the continuity and reliability of the operation.
[0028] To enable simultaneous multi-line operations, in a preferred embodiment of the present invention, such as Figure 1 As shown, the adaptive weeding actuator has three sets, each set consisting of a weeding wheel 3, a drive motor 4, a transmission belt 5, a position adjustment mechanism 14, and a depth limiting plate 15. This configuration allows the weeding robot to simultaneously weed three rows of paddy field crops, significantly improving operational efficiency. During actual operation, the main control system controls each set of position adjustment mechanisms 14 based on visual recognition results, thereby achieving independent adaptive adjustment of each row's weeding actuator.
[0029] Please continue reading. Figure 1 and Figure 2To prevent damage to crop seedlings during movement and weeding, a seedling protection gate 19 is installed on the frame 1. The position of the seedling protection gate 19 corresponds to the crop row, ensuring that rice seedlings are kept within the gate's internal space, thus preventing the robot from crushing them during movement. In addition, the frame 1 is equipped with a power supply interface 10 and a data transmission interface. The power supply interface 10 is used to connect to an external power source to power the device or charge the battery 9, while the data transmission interface is used to communicate with a remote control terminal to remotely monitor and dynamically adjust operating parameters. Regarding energy supply, the weeding robot also includes a solar panel 18 and a battery 9. The solar panel 18 is fixed above the frame 1, making full use of solar energy. The solar panel 18 is electrically connected to the battery 9, which in turn is electrically connected to the main control system and other electrical components to provide the power required for operation. This helps extend the robot's continuous operating time in the field, achieving energy conservation and environmental protection.
[0030] The following details the working method of the adaptive paddy field weeding robot based on machine vision according to the present invention. At the start of operation, the main control system activates the walking drive device. The drive motor 4 drives the weeding wheels 3 to rotate via the transmission belt 5, thereby driving the robot to move stably in the paddy field. Simultaneously, the machine vision recognition module installed at the front of the frame 1 begins operation. The vision recognition module 2 acquires real-time images of the field ahead and transmits the image data to the image processing unit 7. The image processing unit 7 quickly distinguishes between rice plants and weeds in the image, while the weed positioning unit 8 accurately locates the weeds and automatically aligns the machine with the crop rows. Based on the recognition and positioning results transmitted by the weed positioning unit 8, the main control system calculates the optimal operating parameters in real time and drives the position adjustment mechanism 14 to automatically adjust the lateral spacing, rotation speed, and soil penetration depth of each set of weeding wheels 3. Subsequently, the rotating weeding wheels 3 drive the flexible weeding teeth 13 on their surface to rotate, accurately removing the target weeds, while the depth limiting plate 15 located below the weeding wheels 3 strictly controls the soil penetration depth to avoid damage to the rice roots. Throughout the operation, the anti-entanglement cover 16 continuously functions to prevent weeds and straw from getting tangled on the rotating mechanism, while the seedling protection gantry 19 ensures that rice seedlings remain within a safe passage and are not crushed by the robot. All image information and operation parameters can be transmitted back to the remote control terminal in real time via the data transmission interface, allowing operators to dynamically adjust the operation parameters as needed. Through the continuous cycle of the above process, the robot ultimately completes a fully adaptive, high-precision weeding operation in paddy fields.
[0031] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0032] It should be noted that the components mentioned in the above embodiments are all general standard parts or components known to those skilled in the art. Their structures and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.
[0033] This invention has illustrated its principles and implementation methods using specific examples. The descriptions of these embodiments are merely illustrative of the method and its core ideas; furthermore, those skilled in the art will recognize that modifications may be made to the specific implementation methods and application scope based on the principles of this invention. Therefore, the content of this specification should not be construed as limiting the invention.
Claims
1. An adaptive paddy field weeding robot based on machine vision, characterized in that, include: Rack (1); Machine vision recognition module (2), which is installed at the front of the frame (1) and is used to collect field images in real time and identify weeds and crops; It includes a visual recognition module (2), an image processing unit (7), and a weed positioning unit (8), which are electrically connected in sequence; The walking drive device is connected to the bottom of the frame (1) and is used to drive the weeding robot to move in the paddy field; it includes a drive motor (4), a transmission belt (5) and walking wheels; An adaptive weeding actuator is connected to the middle of the frame (1) and is used to perform weeding actions. It includes a weeding wheel (3), a position adjustment mechanism (14), flexible weeding teeth (13), and a depth limiting plate (15). The drive motor (4) is connected to the weeding wheel (3) via the transmission belt (5). The position adjustment mechanism (14) is connected to the weeding wheel (3). The flexible weeding teeth (13) are disposed on the surface of the weeding wheel (3). The depth limiting plate (15) is disposed below the weeding wheel (3). The main control system is fixed on the frame (1) and electrically connected to the machine vision recognition module (2), the walking drive device and the adaptive weeding actuator.
2. The machine vision-based adaptive paddy field weeding robot according to claim 1, characterized in that, The frame (1) is also equipped with a seedling protection gate (19), an anti-tangling cover (16), and an emergency stop switch (11). The seedling protection gate (19) is used to keep the crop seedlings in the gate space to avoid being crushed. The anti-tangling cover (16) is installed on the outside of the adaptive weeding actuator to prevent weeds from tangling. The emergency stop switch (11) is electrically connected to the main control system.
3. The machine vision-based adaptive paddy field weeding robot according to claim 1, characterized in that, The position adjustment mechanism (14) drives the weeding wheel (3) to move laterally and / or longitudinally to adjust the working distance and soil entry position of the weeding wheel (3).
4. The machine vision-based adaptive paddy field weeding robot according to claim 1, characterized in that, The adaptive weeding actuator is provided in three groups. Each group of the adaptive weeding actuator is provided with a weeding wheel (3), a drive motor (4), a transmission belt (5), a position adjustment mechanism (14), and a depth limiting plate (15).
5. The machine vision-based adaptive paddy field weeding robot according to claim 1, characterized in that, The flexible weeding teeth (13) are made of wear-resistant flexible material.
6. The adaptive paddy field weeding robot based on machine vision according to claim 1, characterized in that, The visual recognition module (2) is set to collect data in a forward direction, and the image processing unit (7) and the weed positioning unit (8) are integrated inside the frame (1).
7. The machine vision-based adaptive paddy field weeding robot according to claim 1, characterized in that, The walking drive device also includes a pulley (12), and the transmission belt (5) cooperates with the pulley (12) to provide stable walking drive force in muddy terrain.
8. The machine vision-based adaptive paddy field weeding robot according to claim 1, characterized in that, The weeding robot also includes a solar power panel (18) and a battery (9). The solar power panel (18) is fixed above the frame (1). The solar power panel (18) is electrically connected to the battery (9). The battery (9) is electrically connected to the main control system to provide power.
9. The machine vision-based adaptive paddy field weeding robot according to claim 1, characterized in that, The rack (1) is also provided with a power supply interface (10) and a data transmission interface. The power supply interface (10) is used to connect to an external power source, and the data transmission interface is used to communicate with a remote control terminal.
10. A working method for an adaptive paddy field weeding robot based on machine vision, as described in any one of claims 1 to 9, characterized in that, include: The walking drive device drives the weeding robot to move in the paddy field. The machine vision recognition module collects field images in real time and distinguishes between crop plants and weeds through the image processing unit (7). The weed positioning unit (8) completes the positioning and automatic row alignment of weeds. The main control system controls the position adjustment mechanism (14) to automatically adjust the spacing, rotation speed and soil penetration depth of the weeding wheel (3) based on the identification and positioning results; The weeding wheel (3) drives the flexible weeding teeth (13) to rotate to remove weeds. At the same time, the depth limiting plate (15) controls the soil penetration depth of the flexible weeding teeth (13). The anti-tangling cover (16) prevents weeds and straw from tangling with the mechanism, and finally completes the adaptive precision weeding operation.
