Tobacco leaf picking robot based on visual recognition

Abstract

The utility model relates to the field of tobacco production, concretely relates to a tobacco leaf picking robot based on visual identification, including chassis, mechanical arm and picking clamping jaw, the chassis includes caterpillar chassis and the connecting plate fixed in the top of caterpillar chassis, the side of connecting plate integrative support plate extending to the side below is formed, the support plate is wrapped caterpillar chassis and sets up with the front end exceeding caterpillar chassis front end, the chassis of this device adopts caterpillar chassis, and the mechanical arm picking movement is from below to above, greatly reduces the volume of picking robot, improves its mobility and walking stability, makes it more adapt to the interrow working environment and the picking of different plant different height, adopts the depth camera, can accurately identify the maturity of tobacco leaf, distinguishes the mature and immature state of tobacco leaf, improves the mechanization and automation of picking work, improves the operation efficiency, realizes the accurate picking of mature tobacco leaf.

Description

Technical Field

[0001] This utility model relates to the field of tobacco production, specifically to a tobacco leaf picking robot based on visual recognition. Background Technology

[0002] my country is one of the world's largest tobacco producers, and the tobacco economy plays a vital role in the national economy. Currently, tobacco harvesting in my country mainly relies on manual labor, which is characterized by harsh harvesting environments, high labor intensity, and low efficiency. There is an urgent need to improve the automation level of the harvesting process.

[0003] Existing harvesting robots with a high degree of mechanization and automation, such as the Chinese patent publication CN221127986U, disclose a tobacco leaf harvesting and collection device and method that uses a "gate"-shaped frame to span the upper part of the plant for harvesting, moving through an exposed track chassis. However, different types of tobacco leaves have different growth conditions and plant heights, making it difficult for this "gate"-shaped frame to meet the harvesting needs of tobacco plants of varying heights. If the plant height is too high, the frame needs to be raised accordingly. However, since the working environment is mostly between rows, an excessively high device would result in a high center of gravity, compromising stability and mobility. Furthermore, the exposed track chassis... The chassis is prone to damaging tobacco leaves hanging on the ground along the way, causing significant economic losses. In addition, Chinese Patent Publication No. CN119547641A discloses an intelligent tobacco leaf picking robot and its picking method, which uses anti-scratch ball-type and shearing-type ends to reduce damage during picking, but it cannot effectively reduce damage to the roots of tobacco leaves during picking. Another example is a broccoli picking robot disclosed in Chinese Patent Publication No. CN116114471A, which discloses a picking gripper that uses a ball screw to achieve the gripper closing action. It has a complex shape, is difficult to manufacture, and is large in size, making it unable to pick tobacco leaves flexibly and resulting in low actual picking efficiency. Utility Model Content

[0004] The purpose of this invention is to propose a visual recognition-based tobacco leaf harvesting robot to solve the technical problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: This utility model proposes a tobacco leaf picking robot based on visual recognition, including a chassis, a robotic arm and a picking gripper. The chassis includes a tracked chassis and a connecting plate fixed to the top of the tracked chassis. The connecting plate has an integrally formed support plate extending downward to the side. The support plate wraps around the tracked chassis and its front end extends forward beyond the front end of the tracked chassis. One end of the connecting plate is longitudinally connected to a lead screw device via bolts. A slider is slidably connected to the lead screw device, and a mechanical arm hanger is fixedly mounted on the slider. The robotic arm is fixedly connected to the robotic arm bracket by bolts. The end of the robotic arm is connected to a picking gripper, and the side of the picking gripper is also provided with a collection module for identifying tobacco leaves. The picking gripper includes a drive motor, a drive gear connected to the drive motor, and a drive rod that meshes with the drive gear. A flexible pad is installed at the end of the drive rod.

[0006] As a further improvement, a pressure sensor is provided inside the flexible pad.

[0007] As a further improvement, the lead screw device also includes a lead screw motor, a reducer, a screw, and a guide rail; One end of the reducer is connected to the lead screw motor, and the other end is connected to the screw. The screw is located inside the guide rail, and a slider is provided on the screw, wherein the slider cooperates with the guide rail.

[0008] As a further improvement, the reducer is mounted on a reducer bracket.

[0009] As a further improvement, the robotic arm includes a robotic arm chassis and multiple brackets connected by servo motors, and the end of the robotic arm is also provided with a servo disk for adjusting the angle of the picking gripper.

[0010] As a further improvement, a data acquisition box is also included, which is mounted on a connecting plate and has an inclined partition inside.

[0011] As a further improvement, the tracked chassis includes multiple track support wheels, guide wheels, tracks, a gearbox, and a motor; The track support wheel is mounted on the support frame, with the track passing over the guide wheel and placed on the track support wheel. The motor is connected to the gearbox, and the gearbox is connected to the guide wheel.

[0012] As a further improvement, a battery located inside the electrical cabinet housing is also included, which is used to power the motor.

[0013] As a further improvement, the acquisition module is a depth camera.

[0014] As a further improvement, the depth camera is a K210 monocular camera.

[0015] Compared with the prior art, this utility model achieves the following technical effects: The device uses a tracked chassis, with the robotic arm's harvesting movement proceeding from bottom to top. This significantly reduces the robot's size, improves its mobility and walking stability, and makes it more adaptable to working environments between rows and harvesting from different plant heights. Furthermore, the tracked chassis is encased in support plates, preventing damage to the bottom layer of tobacco leaves during movement. The support plates also lift the tobacco leaves, facilitating the depth camera's identification of harvesting points and preventing the identification points from being obscured by drooping leaves. This device uses a depth camera to accurately identify the maturity of tobacco leaves, distinguish between mature and immature tobacco leaves, improve the mechanization and automation of harvesting, increase work efficiency, and achieve precise harvesting of mature tobacco leaves.

[0016] The picking gripper of this device is a two-finger gripper with a flexible material pad at the end of the gripper to simulate the picking by hand, minimizing damage caused by picking with rigid materials, improving the economic benefits of tobacco leaves, and using gears and a four-bar linkage to realize the picking action, which simplifies the structure, reduces the size, reduces manufacturing costs, and improves the flexibility when picking deep into the bottom of the tobacco plant. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall structure of this utility model; Figure 2 This is a schematic diagram of the tracked chassis of this utility model; Figure 3 This is a schematic diagram of the lead screw structure of this utility model; Figure 4 This is a schematic diagram of the structure of the robotic arm of this utility model; Figure 5 This is a schematic diagram of the picking gripper of this utility model.

[0018] Reference numerals: 1. Chassis; 11. Tracked chassis; 112. Motor; 113. Guide wheel; 114. Battery; 115. Electrical cabinet shell; 116. Gearbox; 117. Track support wheel; 12. Connecting plate; 121. Support plate; 2. Depth camera; 3. Data acquisition box; 4. Screw assembly; 41. Screw motor; 42. Robotic arm bracket; 43. Screw; 44. Slider; 45. Guide rail; 46. Reducer; 47. Reducer bracket; 5. Robotic arm; 51. Robotic arm chassis; 52. Bearing; 53. Servo motor; 54. Bracket; 6. Harvesting gripper; 61. Drive motor; 62. Drive gear; 63. Drive rod; 64. Flexible gasket; 65. Fixing buckle; 66. Steering disc. Detailed Implementation

[0019] The embodiments of this application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application. Rather, the embodiments of this application include all variations, modifications, and equivalents falling within the spirit and scope of the appended claims.

[0020] This utility model proposes a tobacco leaf picking robot based on vision recognition, including a chassis 1, a robotic arm 5 and a picking gripper 6; The chassis 1 includes a tracked chassis 11 and a connecting plate 12 fixed to the top of the tracked chassis 11. The connecting plate 12 has an integrally formed support plate 121 extending downward to the side. The support plate 121 wraps around the tracked chassis 11 and its front end extends forward beyond the front end of the tracked chassis 11. One end of the connecting plate 12 is longitudinally connected to a lead screw device 4 by bolts. A slider 44 is slidably connected to the lead screw device 4, and a mechanical arm hanger 42 is fixed on the slider 44. The robotic arm 5 is fixedly connected to the robotic arm bracket 42 by bolts. The end of the robotic arm 5 is connected to a picking gripper 6, and the side of the picking gripper 6 is also provided with a collection module for identifying tobacco leaves. The picking gripper 6 includes a drive motor 61, a drive gear 62 connected to the drive motor 61, and a drive rod 63 that meshes with the drive gear 62 for transmission. A flexible pad 64 is installed at the end of the drive rod 63.

[0021] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0022] like Figure 1 As shown, the vision-based intelligent tobacco leaf picking robot of this embodiment mainly includes a chassis 1, a depth camera 2, a collection box 3, a lead screw device 4, a robotic arm 5, and a picking gripper 6.

[0023] Please see Figure 1 and Figure 2The chassis 1 includes a connecting plate 12 at the top connecting various components and a tracked chassis 11 at the bottom responsible for movement. The tracked chassis 11 has several track support wheels 117 fixed to the frame. The tracks pass over guide wheels 113 and rest on the track support wheels 117. The guide wheels 113 are driven by motors 112. A battery 114 housed in an electrical cabinet 115 supplies power to each motor 112. A reduction gearbox 116 is used to change the running speed of the tracks on both sides, enabling turning and other operations. Due to the smaller lateral spacing of the chassis 1, it is more suitable for working between furrows, eliminating the need to cross ridges. The lower chassis 1 also provides higher stability. Furthermore, the tracked design allows for the use of metal materials, providing good off-road capability for various complex terrains.

[0024] Please see Figure 1 and Figure 5 A depth camera 2 is mounted on the side of the picking gripper 6. It should be noted that the depth camera 2 can capture images of the tobacco leaves, directly and accurately obtaining their specific spatial location and distance. Using a neural network model, it performs deep learning tasks to send the information to the host computer, driving the robotic arm 5 and the picking gripper 6 to perform the picking action. In this embodiment, the depth camera 2 preferably uses a K210 monocular camera to provide instructions for the chassis 1, robotic arm 5, and picking gripper's actions. Using a deep learning network, algorithms for recognizing leaf color, shape, and drooping degree at different maturity levels are studied and constructed to achieve accurate identification of mature leaves.

[0025] Please see Figure 1 and Figure 3 A lead screw device 4 is longitudinally fixed to the connecting plate 12 by bolts. This device reacts to the spatial position information of the tobacco leaves, increasing the spatial height of the robotic arm 5. This significantly reduces the size of the robotic arm and the harvesting mechanism, greatly improving the overall flexibility of the machine. The lead screw device 4 includes a reducer 46 at its lower end and a reducer bracket 47 housing the reducer, and a lead screw motor 41 at its rear end, providing power to rotate the screw 43. The rotating screw 43 allows the slider 44 to slide longitudinally along the guide rail 45 to change its height.

[0026] It should be noted that the robotic arm bracket 42 is fixed on the slider 44 of the lead screw device 4. The robotic arm bracket 42 has bolt holes arranged on it, which correspond one-to-one with the holes on the robotic arm chassis 51. It is fixed by screws. The multi-screw connection can improve stability and bending strength, and can be adapted to different models of robotic arms 5 with the same base.

[0027] Please see Figure 4The robotic arm 5, also known as a multi-joint robot, contains four degrees of freedom. The robotic arm 5 includes a chassis 51 with screw holes corresponding to those on the robotic arm bracket 42. A bearing 52 is located at the upper end of the chassis 51. Driven by a servo motor 53, the angle of the robotic arm 5 along the z-axis can be changed, enabling harvesting at different angles. The main body of the robotic arm 5 consists of several supports 54. These "U"-shaped supports 54 use lightweight alloys such as aluminum alloy, effectively reducing the overall weight while providing sufficient bending and torsional strength. Each support segment 54 is mainly fixed with bolts, forming three joints. Torque is provided between each joint via the servo motor 53, and the harvesting action is achieved after receiving signals from the microcontroller. A drive motor 61 is placed in the end bracket of the robotic arm 5, connected to a servo disc 66 at the end of the harvesting gripper, changing the angle of the harvesting gripper 6, making the harvesting operation more flexible.

[0028] Please see Figure 5 The picking gripper 6 has a fixing buckle 65 at its end, which is fixed to the end of the robotic arm 5 by bolts. A depth camera 2 is installed on the side to identify the location of the salt field and the picking point. The picking action is completed by the drive gear 62 and the drive rod 63. The drive motor 61 drives the drive gear 62 to rotate, and through a four-bar linkage, the rotational motion of the drive gear 62 is converted into the horizontal movement of the flexible pad 64 at the end of the drive rod 63. The flexible pad 64 is preferably made of polyurethane plastic or a series of materials with elastomeric properties, flexibility and compatibility. A pressure sensor is embedded in the bottom of the flexible pad 64 to monitor the picking force applied when picking tobacco leaves in real time. The electrical signal generated by the pressure is transmitted to the microcontroller in real time to regulate the pressure and ensure that the pressure is always within the most suitable range. This picking gripper 6 uses a drive gear 62 and a drive rod 63 to complete the picking action, which greatly reduces the size and manufacturing cost of the picking gripper 6. At the same time, it combines the two-finger design of the picking gripper 6 with the index finger and thumb commonly used for picking tobacco leaves by hand, and the application of flexible pads 64 greatly simulates the picking of human hands. At the same time, it controls the pressure in real time, which minimizes the damage to tobacco leaves caused by the picking of metal parts and improves economic efficiency.

[0029] Please see Figure 1 The middle section of the connecting plate 12 houses the collection box 3, which is used for temporary storage of the harvested tobacco leaves. Inside the collection box 3 is a slanted partition. Although this reduces the space the collection box 3 can hold, placing the harvested tobacco leaves at an angle greatly reduces pressure on the lower tobacco leaves from the upper leaves, preventing them from breaking or bending. The collection box 3 is specially designed to hold enough tobacco leaves, with the root ends of the leaves placed high on the partition for easy manual collection and processing.

[0030] When this device is in use, the depth camera 2 collects the position information of the tobacco leaves and transmits the information to the control element. The control element drives the chassis 1 to move to a suitable position, passing through the bottom layer of tobacco leaves. The support plate 121 lifts the bottom layer of tobacco leaves. The depth camera 2, located at the front end of the robotic arm 5, penetrates into the bottom of the tobacco leaves to identify the picking position. The control screw device 4 is raised to a suitable height. The control element controls the robotic arm 5 to send the picking gripper 6 to a position close to the root of the tobacco leaves. The rudder disk 66 is rotated to the optimal picking angle, driving the picking gripper 6 to slowly close. The picking pressure is detected and negative feedback is sent to the control element to stabilize the picking pressure and realize the picking action.

[0031] It should be noted that in the description of this application, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0032] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the function involved, as will be understood by those skilled in the art to which embodiments of this application pertain.

[0033] 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.

[0034] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A tobacco leaf harvesting robot based on vision recognition, characterized in that, Includes chassis (1), robotic arm (5) and picking gripper (6); The chassis (1) includes a tracked chassis (11) and a connecting plate (12) fixed to the top of the tracked chassis (11). The connecting plate (12) has an integrally formed support plate (121) extending downward to the side. The support plate (121) wraps around the tracked chassis (11) and its front end extends forward beyond the front end of the tracked chassis (11). One end of the connecting plate (12) is longitudinally connected to a screw device (4) by bolts. A slider (44) is slidably connected to the screw device (4), and a mechanical arm hanger (42) is fixed on the slider (44). The robotic arm (5) is fixedly connected to the robotic arm bracket (42) by bolts. The end of the robotic arm (5) is connected to a picking gripper (6). The side of the picking gripper (6) is also provided with a collection module for identifying tobacco leaves. The picking gripper (6) includes a drive motor (61), a drive gear (62) connected to the drive motor (61), and a drive rod (63) meshing with the drive gear (62). A flexible pad (64) is installed at the end of the drive rod (63).

2. The tobacco leaf harvesting robot based on vision recognition according to claim 1, characterized in that, The flexible pad (64) is equipped with a pressure sensor.

3. The tobacco leaf harvesting robot based on vision recognition according to claim 1, characterized in that, The lead screw device (4) also includes a lead screw motor (41), a reducer (46), a screw (43), and a guide rail (45). One end of the reducer (46) is connected to the lead screw motor (41), and the other end is connected to the screw (43). The screw (43) is located inside the guide rail (45), and a slider (44) is provided on the screw (43), wherein the slider (44) cooperates with the guide rail (45).

4. The tobacco leaf harvesting robot based on vision recognition according to claim 3, characterized in that, The reducer (46) is mounted on the reducer bracket (47).

5. The tobacco leaf harvesting robot based on vision recognition according to claim 1, characterized in that, The robotic arm (5) includes a robotic arm chassis (51) and multiple brackets (54) connected by servo motors (53). The end of the robotic arm (5) is also provided with a servo disk (66) for adjusting the angle of the picking gripper (6).

6. The tobacco leaf harvesting robot based on vision recognition according to claim 1, characterized in that, It also includes a data collection box, which is mounted on a connecting plate and has an inclined partition inside.

7. The tobacco leaf harvesting robot based on vision recognition according to claim 1, characterized in that, The tracked chassis (11) includes multiple track support wheels (117), guide wheels (113), tracks, a gearbox (116), and a motor (112). The track support wheel (117) is mounted on the support frame, wherein the track passes around the guide wheel (113) and is placed on the track support wheel (117). The motor (112) is connected to the gearbox (116), and the gearbox (116) is connected to the guide wheel (113).

8. The tobacco leaf harvesting robot based on vision recognition according to claim 7, characterized in that, It also includes a battery (114) located in the electrical cabinet housing (115), which is used to supply power to the motor (112).

9. The tobacco leaf harvesting robot based on vision recognition according to claim 1, characterized in that, The acquisition module is a depth camera (2).

10. The tobacco leaf harvesting robot based on vision recognition according to claim 9, characterized in that, The depth camera (2) is a K210 monocular camera.

Images