Intelligent lentinus edodes harvesting robot
By using a pneumatically driven gripper structure and intelligent image recognition technology, the shiitake mushroom harvesting robot achieves efficient and stable harvesting and mushroom protection, solving the problems of unstable gripping and mushroom damage in high-humidity environments of traditional robots, and improving harvesting quality and efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUAZHONG AGRI UNIV
- Filing Date
- 2025-09-16
- Publication Date
- 2026-07-31
AI Technical Summary
Traditional shiitake mushroom harvesting robots are prone to slipping or damaging the mushrooms in high humidity environments, resulting in unstable harvesting efficiency. Furthermore, the wounds on the mushrooms are susceptible to mold growth after harvesting, affecting their commercial value.
It adopts a pneumatically driven gripper structure, combining pneumatic clamping and gas delivery functions. During clamping, it blows air to dehumidify and after picking, it dries the wounds on the mushroom body. The gripper is equipped with ventilation holes and air chambers. The pneumatic system is connected to the air tank and piston assembly through air pipes. The gripper is equipped with anti-slip pads and pressure sensors. The servo motor drives the mechanical gripper to rotate, and it combines image recognition and deep learning models for precise picking.
It improves harvesting efficiency and mushroom protection, reduces mushroom damage and mold risk, enhances harvesting quality and shelf life, and reduces energy consumption and labor costs.
Smart Images

Figure CN120982354B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of agricultural automation technology, specifically to an intelligent mushroom harvesting robot. Background Technology
[0002] Shiitake mushroom harvesting is characterized by strong seasonality, high labor intensity, and high labor costs. With an aging population and a shortage of agricultural labor, the traditional manual harvesting method is no longer sufficient to meet the demands of large-scale, standardized production. Furthermore, manual harvesting is inefficient and prone to damaging the mushrooms due to improper handling, affecting their commercial value. Therefore, developing automated and intelligent harvesting equipment has become an urgent need for the transformation and upgrading of the edible mushroom industry.
[0003] Currently, some automated agricultural harvesting equipment has emerged both domestically and internationally. For example, the Jiangsu Provincial Agricultural Science and Technology Innovation Project developed an intelligent mushroom harvesting robot, which addresses the problems of high labor intensity, low harvesting efficiency, and high labor costs in mushroom cultivation factories. This is of great significance for promoting the industrialization and large-scale development of the mushroom industry. Although this robot has made breakthroughs in image recognition, its gripping mechanism uses a simple motor-driven gear rack or worm gear structure for opening and closing, resulting in limited functionality. The gripper can only perform basic gripping actions and cannot handle dew or culture medium deposits on the mushroom stem surface caused by high humidity, easily leading to slippage or injury. Based on these problems, an intelligent mushroom harvesting robot that ensures harvesting efficiency while maintaining stable harvesting is proposed. Summary of the Invention
[0004] To address the aforementioned issues, this invention provides an intelligent mushroom harvesting robot. Through a pneumatically driven gripping structure, it achieves an integrated function of briefly dehumidifying the surrounding environment during gripping and rapidly drying the wounds on the cut surface of the mushroom after successful harvesting.
[0005] To achieve the above objectives, the technical solution of the present invention is as follows: a shiitake mushroom intelligent harvesting robot, including a robotic arm, with a robotic claw at the end of the robotic arm, the robotic arm being signal-connected to a harvesting system, the robotic claw including a base, a pneumatic assembly and symmetrically arranged grippers, the pneumatic assembly being used to drive the grippers to clamp the stem of the shiitake mushroom, the pneumatic assembly including a piston assembly fixedly connected to one end of the base and air pipes symmetrically arranged on both sides of the piston assembly, a wedge block fixedly connected to the output end of the piston assembly, and each gripper having a sliding groove that slides and engages with both sides of the wedge block; a spring is sleeved on the piston assembly, with both ends of the spring fixedly connected to the side wall of the base and the wedge block respectively;
[0006] Each gripper has an air chamber, and each gripper has several ventilation holes located at the bottom of the air chamber. The end of the air pipe away from the piston assembly is connected to the air chamber.
[0007] The technical principles of the above solution are as follows:
[0008] When a robotic arm drives a robotic gripper to harvest shiitake mushrooms, air pressure is input into the piston assembly through an air pipe. The output end of the piston assembly drives the wedge block forward synchronously. Due to the sliding engagement between the two sides of the wedge block and the grooves inside the gripper, the lateral movement of the wedge block compresses the two gripper claws on both sides through the inclined surface, causing them to close towards the center, thereby clamping the stem of the shiitake mushroom. After the air pressure is released, the spring fitted on the piston assembly returns to its original position, pushing the wedge block to move in the opposite direction, and the gripper claws release under the action of the spring force.
[0009] Before the grippers close, gas is blown out through the air tube from the ventilation hole, briefly blowing away the moisture in and around the shiitake mushroom stem, reducing the impact of the humid environment during harvesting on the shiitake mushroom; after harvesting, when the air supply is continuous, the airflow from the ventilation hole can quickly dry the wounds on the cut surface of the shiitake mushroom.
[0010] The above approach has the following beneficial effects:
[0011] 1. This solution utilizes a pneumatic system to simultaneously achieve gripper clamping and gas delivery. When the pneumatically driven wedge closes the gripper, gas is simultaneously blown out through the ventilation holes, instantly dehumidifying the surface and surrounding area of the shiitake mushroom. This prevents mold or contamination caused by moisture during harvesting, improving harvest quality. It also reduces the need for additional components, simplifies the mechanical structure, and enhances the robot's operational efficiency.
[0012] 2. After traditional mechanical harvesting, the cut surfaces of shiitake mushrooms are prone to bacterial growth due to moisture, affecting their shelf life. This solution uses continuous airflow through ventilation holes after harvesting to act on the wounds, accelerating moisture evaporation and creating a dry surface. This is equivalent to providing immediate care for the mushroom bed (log) and the wounds of the shiitake mushrooms after harvesting, ensuring the growth of the next batch of shiitake mushrooms and preserving the freshness of the harvested mushrooms.
[0013] 3. In this solution, the pneumatic drive of the gripper and the airflow control of the ventilation holes form a linkage mechanism: during gripping, the airflow briefly dehumidifies to ensure stable gripping and prevent slippage caused by wetness; after harvesting, the airflow continues to dry the fruit, optimizing post-harvest quality. This design, which automatically switches airflow functions according to the stage of operation, reflects targeted optimization of the physiological characteristics of shiitake mushrooms. Compared with traditional mechanical grippers, it better meets the actual needs of agricultural harvesting scenarios and demonstrates the intelligent feature of "functions changing as needed".
[0014] Furthermore, the piston assembly includes a gas tank and a piston rod. The gas tank is fixedly connected to one side of the base, and a piston plate is slidably connected inside the gas tank. The piston plate divides the interior of the gas tank into a first chamber and a second chamber. The side of the first chamber away from the piston plate is connected to a gas delivery assembly for conveying gas, and all gas pipes are connected to the first chamber. The piston plate is located on one side of the second chamber and is fixedly connected to the piston rod. The end of the piston rod away from the piston plate is fixedly connected to a wedge block, and a spring is sleeved on the piston rod.
[0015] Beneficial effects: When the first chamber receives gas, it can both push the piston plate to move the piston rod to drive the gripper to close, and simultaneously supply gas to the gripper's air chamber through the air pipe; the spring is sleeved on the piston rod, and when the air pressure is removed, the spring force acts directly on the piston rod, making the wedge block reset more stable, avoiding the jamming problem that may occur in traditional single pneumatic reset, and ensuring that the opening and closing action of the gripper is precise and controllable.
[0016] Furthermore, the gas delivery assembly includes a gas pump that is signal-connected to the harvesting system.
[0017] Beneficial effects: The harvesting system can adjust the air pressure output of the air pump in real time according to the variety and maturity of the shiitake mushroom. For example, the air pressure can be reduced for thin and weak stems to reduce the clamping force, while the air pressure can be increased for thick stems to ensure a stable clamping.
[0018] Furthermore, a drive assembly for driving the mechanical gripper to rotate left and right is provided near the mechanical gripper on the robotic arm. The drive assembly includes a drive cavity and a servo motor. The drive cavity is located inside the end of the robotic arm. The servo motor is fixedly connected to the top wall inside the drive cavity. A transmission rod is coaxially fixedly connected to the output shaft of the servo motor. A first connecting rod is coaxially sleeved on the transmission rod. An arc-shaped groove is provided around the bottom of the drive cavity. The end of the first connecting rod away from the servo motor extends into the arc-shaped groove and is slidably connected to the arc-shaped groove. A second connecting rod is fixedly connected to the end of the first connecting rod located in the arc-shaped groove. A fixing ring is fixedly connected to the end of the second connecting rod away from the first connecting rod. The fixing ring is sleeved at the connection between the base and the gas tank.
[0019] Beneficial effects: The servo motor drives the fixed ring to rotate along the arc groove through the first and second connecting rods, which can realize the left and right rotation of the mechanical claw. The torque of the servo motor can adapt to the different growth angles of the shiitake mushrooms on the substrate, avoiding the failure of gripping due to angle deviation.
[0020] Furthermore, anti-slip pads are provided on the inner sidewalls of the grippers, and pressure sensors connected to the harvesting system are installed inside the anti-slip pads.
[0021] Beneficial effects: The anti-slip pad increases the friction between the gripper and the mushroom stem, especially when the stem surface is slightly damp, which can effectively prevent the mushroom from slipping during the gripping process and reduce the harvesting failure rate; the pressure sensor monitors the gripping force in real time and feeds it back to the harvesting system. The system can adjust the air pump pressure according to the preset threshold to ensure that the gripping force is within the optimal range that is both stable and does not damage the mushroom, avoiding the stem breakage or epidermal damage caused by the uncontrolled force of traditional mechanical grippers; the continuous signal from the pressure sensor can also serve as an auxiliary basis for judging whether the harvesting is successful (e.g., the sudden disappearance of the gripping force may mean that the mushroom has fallen off), providing data support for subsequent processing.
[0022] Furthermore, each gripper is equipped with a capacitive humidity sensor that is connected to the harvesting system signal, and each anti-slip pad has an opening to allow the capacitive humidity sensor to contact the mushroom stem.
[0023] Beneficial effects: The capacitive humidity sensor directly contacts the mushroom stem through the opening of the anti-slip pad, allowing for real-time acquisition of surface humidity data, avoiding interference from ambient humidity and ensuring higher data accuracy. Combined with preset humidity thresholds for different varieties in the harvesting system, it can determine in advance whether harvesting is suitable: if the humidity is too high, the system can first control the ventilation holes to blow air and dehumidify before harvesting, reducing the risk of clamping failure due to slipperiness or post-harvest mold; humidity data can also help determine the freshness of shiitake mushrooms (e.g., excessive moisture may indicate impending rot), further optimizing the harvesting and selection logic and improving harvest quality.
[0024] Furthermore, the harvesting system has built-in training datasets for several shiitake mushroom varieties. The harvesting system includes an image receiving module, a recognition module, a judgment module, and an execution module; among which:
[0025] The image receiving module is connected to camera components mounted on the side wall and base of the robotic arm. The image receiving module receives and processes image information of shiitake mushrooms from several angles synchronously acquired by the camera components using a preprocessing algorithm. The preprocessed key feature points are then transmitted to the recognition module. The preprocessing algorithm includes noise reduction of image information using Gaussian filtering, enhancement of image information contrast using histogram equalization, and extraction of key feature points from the image information using a scale-invariant feature transform algorithm. The key feature points include the surface texture and color of the shiitake mushroom cap, the opening and closing degree of the gills, and the relative size ratio between the stem and the cap.
[0026] Beneficial effects: The camera components on the sidewall and base of the robotic arm can simultaneously acquire multi-angle images, which can completely restore the three-dimensional morphology of the mushroom and avoid the loss of information from a single perspective; Gaussian filtering removes environmental interference, such as impurities or light reflection on the surface of the mushroom log; histogram equalization enhances contrast and highlights the boundary between the cap and the log; SIFT algorithm extracts key features, which significantly improves the accuracy and stability of feature points; the preprocessed key feature points are more consistent with the biological characteristics of shiitake mushroom growth (such as the gill opening degree being directly related to maturity), laying a reliable foundation for subsequent maturity recognition and three-dimensional localization and reducing the misjudgment rate.
[0027] Furthermore, the identification module receives key feature points. Based on the improved YOLOv7 deep learning model, which is trained and optimized by accessing the training dataset, it calls the corresponding training dataset of shiitake mushrooms according to the variety selected by the user. It identifies the maturity of the shiitake mushrooms based on the key feature points and performs three-dimensional spatial positioning of the mushroom body, including the center of the cap and the base of the stem, to obtain three-dimensional coordinate information. It also classifies the maturity of the identified shiitake mushrooms into three levels: "immature", "mature" and "overripe", and packages the three-dimensional coordinate information, variety information and level determination results into an identification result and transmits it to the judgment module.
[0028] Beneficial effects: The improved YOLOv7 model is optimized for the morphological characteristics of shiitake mushrooms. Combined with variety-specific training datasets, it can accurately distinguish between different shiitake mushroom varieties, avoiding misharvesting across varieties. Maturity is divided into three levels: "immature," "mature," and "overripe," accurately matching harvesting needs (harvesting only mature mushrooms), reducing resource waste (avoiding harvesting immature mushrooms) and quality loss (avoiding harvesting overripe mushrooms). The three-dimensional spatial positioning of the cap center and stipe base coordinates provides precise motion parameters for the robotic arm and drive components, ensuring that the robotic claw can accurately reach the gripping position "point-to-point," reducing ineffective movements and improving work efficiency.
[0029] Furthermore, the judgment module consists of three judgment logics, specifically:
[0030] One judgment is made to receive the recognition result. If the recognition result is determined to be "mature", then it can be picked and the three-dimensional coordinate information is transmitted to the execution module.
[0031] The second judgment is used to receive humidity data and variety information transmitted by the capacitive humidity sensor and the corresponding humidity threshold to assess whether the moisture content of the mushroom body affects harvesting. If it does not affect harvesting, the harvesting information is transmitted to the execution module to control the mechanical claw to harvest. If it does affect harvesting, the adjustment information is transmitted to the third judgment.
[0032] The system performs three checks to determine whether the shiitake mushroom has successfully detached from the substrate by comparing the measured torque with the standard torque range corresponding to the variety and maturity, and by combining the continuous pressure signal from the pressure sensor. If the harvest is successful, the subsequent processing information is transmitted to the execution module. If the harvest fails or adjustment information is received, the current processing information is output and transmitted to the execution module.
[0033] Beneficial effects: The hierarchical judgment logic constructs a progressive harvesting decision-making system, significantly improving the system's reliability.
[0034] A single assessment (maturity screening) ensures that subsequent operations are only performed on mature mushrooms from the source, avoiding ineffective work.
[0035] Secondary assessment (humidity assessment) addresses the susceptibility of shiitake mushrooms to humidity by intervening in cases of excessive humidity (e.g., dehumidifying before harvesting) to reduce harvesting failures caused by environmental factors.
[0036] Three judgments (torque and pressure signals to assess whether the harvesting was successful) can identify harvesting abnormalities (such as the mushroom not detaching from the stick or the gripper slipping) and trigger corresponding processing mechanisms (such as retrying or marking for manual handling), avoiding system lag or erroneous execution and improving the degree of automation.
[0037] Furthermore, the execution module is used to receive three-dimensional coordinate information, picking information, subsequent processing information, and current processing information, and transmit the corresponding data to the robotic arm or gripper to perform the operation, wherein:
[0038] The three-dimensional coordinate information is used to send motion commands to the robotic arm and servo motor, control the robotic gripper to move precisely to the location of the mushroom stem, and at the same time send clamping commands to the air pump.
[0039] The harvesting information is used to send a quantitative inflation command to the air pump, which drives the gripper to close and hold the stipe. At the same time, it receives real-time pressure data from the pressure sensor and adjusts the inflation amount to maintain the clamping force within the threshold range corresponding to the variety.
[0040] The subsequent processing information is used to send a retraction command to the robotic arm, controlling the robotic arm to move the robotic gripper to the temporary storage area;
[0041] The current processing information is divided into two execution logics: If it is adjustment information, a low-pressure continuous air supply command is sent to the air pump, and a dry airflow is blown onto the surface of the mushroom stem through the ventilation hole to reduce the surface humidity and then trigger the secondary judgment and harvesting process again; if it is harvesting failure information, an air release command is sent to the air pump, which resets the spring and causes the gripper to release. At the same time, a reset command is sent to the robotic arm to control the robotic gripper to return to a safe position, and the coordinates of the mushroom are marked as "awaiting manual processing" and stored in the system log.
[0042] Beneficial effects: The execution module controls the robotic arm and servo motor based on three-dimensional coordinate motion commands to achieve precise positioning and ensure that the gripper is accurately aligned with the mushroom stem; combined with pressure sensor feedback to adjust the inflation volume, it achieves adaptive gripping force control and protects the integrity of the mushroom body;
[0043] When abnormal handling logic occurs, such as dehumidifying first when the humidity is too high or marking manual handling when harvesting fails, the harvesting system is fault-tolerant, reducing the overall downtime caused by a single fault and improving the stability of continuous operation.
[0044] The phased instructions enable all components to work together, resulting in a clear and efficient process that improves efficiency compared to traditional manual harvesting.
[0045] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description
[0046] Figure 1 This is an isometric schematic diagram of an embodiment of the intelligent mushroom harvesting robot of the present invention;
[0047] Figure 2 This is an isometric schematic diagram of the mechanical claw portion of an embodiment of the intelligent mushroom harvesting robot of the present invention;
[0048] Figure 3 This is an isometric sectional view of the drive component of an embodiment of the intelligent mushroom harvesting robot of the present invention;
[0049] Figure 4 This is a schematic diagram of the system framework of an embodiment of the intelligent mushroom harvesting robot of the present invention.
[0050] The reference numerals in the accompanying drawings include: 1. robotic arm; 2. drive chamber; 3. air pump; 4. base; 5. gripper; 6. air tank; 7. first chamber; 8. piston plate; 9. air pipe; 10. piston rod; 11. second chamber; 12. slide groove; 13. spring; 14. wedge; 15. anti-slip pad; 16. servo motor; 17. transmission rod; 18. first connecting rod; 19. second connecting rod; 20. arc groove. Detailed Implementation
[0051] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. 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.
[0052] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They 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. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.
[0053] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0054] The following detailed description illustrates the specific implementation method:
[0055] Example:
[0056] As attached Figure 1As shown: A smart mushroom harvesting robot includes a robotic arm 1, with a mechanical claw at the end of the robotic arm 1. The robotic arm 1 is connected to a harvesting system via a signal. The mechanical claw includes a base 4, a pneumatic assembly, and symmetrically arranged grippers 5. The pneumatic assembly is used to drive the grippers 5 to clamp the stem of the mushroom. Figure 2 As shown, the pneumatic assembly includes a piston assembly fixedly connected to one end of the base 4 and air pipes 9 symmetrically arranged on both sides of the piston assembly. The piston assembly includes an air tank 6 and a piston rod 10. The air tank 6 is fixedly connected to one side of the base 4. A piston plate 8 is slidably connected inside the air tank 6. The piston plate 8 divides the interior of the air tank 6 into a first chamber 7 and a second chamber 11. The side of the first chamber 7 away from the piston plate 8 is connected to a gas delivery assembly for delivering gas. The gas delivery assembly includes an air pump 3 connected to the harvesting system signal, and all air pipes 9 are connected to the first chamber 7. The piston plate 8 is located on one side of the second chamber 11 and is fixedly connected to the piston rod 10. A wedge block 14 is fixedly connected to the end of the piston rod 10 away from the piston plate 8. A spring 13 is sleeved on the piston rod 10. The two ends of the spring 13 are fixedly connected to the side wall of the base 4 and the wedge block 14, respectively. Each of the grippers 5 has a sliding groove 12 that slides and engages with both sides of the wedge block 14.
[0057] Each of the grippers 5 has an air chamber, and each of the grippers 5 has several ventilation holes located at the bottom of the air chamber. The end of the air pipe 9 that is away from the air tank 6 is connected to the air chamber.
[0058] Each gripper 5 has an anti-slip pad 15 on its inner sidewall, and each anti-slip pad 15 contains a pressure sensor that is connected to the harvesting system. Each gripper 5 also contains a capacitive humidity sensor that is connected to the harvesting system, and each anti-slip pad 15 has an opening that allows the capacitive humidity sensor to contact the stem of the shiitake mushroom.
[0059] The robotic arm 1 is equipped with a drive assembly near the robotic gripper for driving the gripper to rotate left and right, such as... Figure 3 As shown, the drive assembly includes a drive cavity 2 and a servo motor 16. The drive cavity 2 is located inside the end of the robotic arm 1. The servo motor 16 is fixedly connected to the top wall inside the drive cavity 2. A transmission rod 17 is coaxially fixedly connected to the output shaft of the servo motor 16. A first connecting rod 18 is coaxially sleeved on the transmission rod 17. An arc-shaped groove 20 is provided around the bottom of the drive cavity 2. The end of the first connecting rod 18 away from the servo motor 16 extends into the arc-shaped groove 20 and is slidably connected to the arc-shaped groove 20. A second connecting rod 19 is fixedly connected to the end of the first connecting rod 18 located in the arc-shaped groove 20. A fixing ring is fixedly connected to the end of the second connecting rod 19 away from the first connecting rod 18. The fixing ring is sleeved at the connection between the base 4 and the gas tank 6.
[0060] The harvesting system has built-in training datasets for several shiitake mushroom varieties, such as... Figure 4 As shown, the harvesting system includes an image receiving module, a recognition module, a judgment module, and an execution module; wherein:
[0061] The image receiving module is connected to camera components mounted on the side wall of the robotic arm 1 and the base 4 respectively. The image receiving module is used to receive and process image information of shiitake mushrooms from several angles synchronously acquired by the camera components through a preprocessing algorithm, and transmit the preprocessed key feature points to the recognition module. The preprocessing algorithm includes noise reduction of image information by using Gaussian filtering, enhancement of image information contrast by using histogram equalization, and extraction of key feature points in image information by using scale-invariant feature transformation algorithm. The key feature points include the surface texture and color of the shiitake mushroom cap, the opening and closing degree of the gills, and the relative size ratio of the stem to the cap.
[0062] The identification module receives key feature points and is based on an improved YOLOv7 deep learning model. The YOLOv7 deep learning model is trained and optimized by accessing the training dataset. Based on the shiitake mushroom variety selected by the user, the corresponding shiitake mushroom training dataset is accessed. The maturity of the shiitake mushroom is identified based on the key feature points, and the mushroom body, including the center of the cap and the base of the stem, is located in three-dimensional space to obtain three-dimensional coordinate information. The module also classifies the maturity of the identified shiitake mushrooms into three levels: "immature", "mature" and "overripe". The three-dimensional coordinate information, variety information and level determination results are packaged into an identification result and transmitted to the judgment module.
[0063] The judgment module consists of three judgment logic steps, specifically:
[0064] One judgment is made to receive the recognition result. If the recognition result is determined to be "mature", then it can be picked and the three-dimensional coordinate information is transmitted to the execution module.
[0065] The second judgment is used to receive humidity data and variety information transmitted by the capacitive humidity sensor and the corresponding humidity threshold to assess whether the moisture content of the mushroom body affects harvesting. If it does not affect harvesting, the harvesting information is transmitted to the execution module to control the mechanical claw to harvest. If it does affect harvesting, the adjustment information is transmitted to the third judgment.
[0066] The system performs three checks to determine whether the shiitake mushroom has successfully detached from the substrate by comparing the measured torque with the standard torque range corresponding to the variety and maturity, and by combining the continuous pressure signal from the pressure sensor. If the harvest is successful, the subsequent processing information is transmitted to the execution module. If the harvest fails or adjustment information is received, the current processing information is output and transmitted to the execution module.
[0067] The execution module is used to receive three-dimensional coordinate information, picking information, subsequent processing information, and current processing information, and transmit the corresponding data to the robotic arm 1 or robotic gripper to perform operations.
[0068] The three-dimensional coordinate information is used to send motion commands to the robotic arm 1 and servo motor 16 to control the robotic gripper to move precisely to the location of the mushroom stem, and at the same time send clamping commands to the air pump 3.
[0069] The picking information is used to send a quantitative inflation command to the air pump 3, which drives the gripper 5 to close and hold the mushroom stem. At the same time, it receives real-time pressure data from the pressure sensor and adjusts the inflation amount to maintain the clamping force within the threshold range corresponding to the variety.
[0070] The subsequent processing information is used to send a retraction command to robotic arm 1, controlling robotic arm 1 to move the robotic gripper to the temporary storage area;
[0071] The current processing information is divided into two execution logics: if it is adjustment information, a low-pressure continuous air supply command is sent to the air pump 3, and a dry airflow is blown onto the surface of the mushroom stem through the ventilation hole to reduce the surface humidity and then trigger the secondary judgment and picking process again; if it is picking failure information, an air release command is sent to the air pump 3, which resets the spring 13 and drives the gripper 5 to release. At the same time, a reset command is sent to the robotic arm 1 to control the robotic gripper to return to a safe position, and the coordinates of the mushroom are marked as "awaiting manual processing" and stored in the system log.
[0072] The specific implementation process is as follows: Taking a real-world harvesting scenario at a shiitake mushroom cultivation base as an example, in a greenhouse environment, if harvesting is carried out after the surface humidity of the mushrooms has increased due to spraying operations, the mechanical claws are prone to slipping during traditional mechanical harvesting. This robot achieves reliable harvesting through the following collaborative mechanism: First, the harvesting system acquires images of the shiitake mushrooms using multi-angle cameras arranged on the robotic arm 1 and base 4. The image receiving module performs noise reduction, contrast enhancement, and feature extraction on the images. The recognition module analyzes the cap texture, color, gill state, and stem proportion based on a pre-trained YOLOv7 model to determine the maturity of the shiitake mushroom and performs three-dimensional localization. If the mushroom is identified as "mature," its three-dimensional coordinates are sent to the execution module.
[0073] The execution module then drives the robotic arm 1 to move the robotic claw to the target mushroom stem area, and simultaneously starts the air pump 3 to inflate the first chamber 7. The gas pushes the piston plate 8 to move, which in turn drives the wedge 14 forward through the piston rod 10, causing the two grippers 5 to close along the slide groove 12, thus clamping the mushroom stem. At this time, the spring 13 is in a compressed state, providing a restoring force for the wedge 14 to reset when the air pump 3 stops inflating. On the other hand, some gas flows into the air chamber of the gripper 5 through the air pipe 9 and is finally discharged through the ventilation hole, blowing air around the mushroom stem in the initial stage of clamping, reducing surface humidity, and thus improving clamping stability.
[0074] During this process, the pressure sensor monitors the clamping force in real time and feeds it back to the judgment module. If the pressure is abnormal, the system dynamically adjusts the air volume of air pump 3 to achieve closed-loop control of the clamping force. At the same time, the capacitive humidity sensor detects the humidity of the mushroom stem. If the humidity exceeds the set threshold, the judgment module will notify the execution module to start a low-pressure continuous blowing mode to dry the surface of the mushroom stem through the ventilation holes. The system will check again after a few seconds until the humidity is suitable before continuing the harvesting process.
[0075] When both the clamping force and humidity are within the normal range, the robotic arm 1 applies upward force to harvest the shiitake mushrooms. Pressure and torque sensors work together to determine whether the mushrooms have successfully detached from the substrate. If harvesting is successful, the execution module controls the robotic arm 1 to transfer the mushrooms to a temporary storage area and maintains air supply from the air pump 3 for a period of time to air-dry the harvested section and inhibit bacterial growth. If harvesting fails or an abnormality occurs, the air pump 3 stops supplying gas, the spring 13 drives the wedge 14 to reset, the gripper 5 releases, and the robotic arm 1 returns to a safe position. This position is recorded as the point requiring manual intervention for subsequent manual harvesting and identification.
[0076] To verify the harvesting efficiency and advantages of the intelligent mushroom harvesting robot proposed in this solution, the following experiment was designed:
[0077] I. Experimental Design
[0078] 1. Group settings
[0079] Experimental group: The intelligent mushroom harvesting robot of the present invention (equipped with pneumatic gripping, dehumidification and drying functions, and the harvesting system includes image recognition and three-stage judgment logic).
[0080] Control group: Using existing intelligent mushroom harvesting robot (motor-driven gear and rack gripper, no dehumidification and drying function, basic image recognition).
[0081] Blank group: Hand-picked (experienced pickers using traditional hand-picking methods).
[0082] 2. Experimental conditions
[0083] Shiitake mushroom variety: The "939" shiitake mushroom variety is used uniformly, and the mushroom logs are in a consistent growth state, with the shiitake mushrooms selected at the "mature" level.
[0084] Environmental conditions: Greenhouse temperature 22±2℃, humidity 85±5%, light intensity 500±100 lux.
[0085] Experimental scale: 100 shiitake mushrooms were treated in each group, and the experiment was repeated 3 times. The average value was taken.
[0086] II. Experimental Procedure
[0087] Preparation stage
[0088] Mark the shiitake mushrooms on the substrate that meet the maturity requirements, and record their locations by number.
[0089] Calibrate the camera components, pressure sensors, and humidity sensors of the test group robot; debug the robotic arm and grippers of the control group equipment.
[0090] Harvesting stage
[0091] Experimental group:
[0092] The robot uses its camera to capture multi-angle images of shiitake mushrooms, and its recognition module determines the maturity and locates the three-dimensional coordinates.
[0093] The robotic arm moves the robotic gripper to the stipe of the mushroom, the air pump inflates and drives the gripper to close, and at the same time the ventilation hole blows air to dehumidify.
[0094] After harvesting, the section is continuously air-dried, and the execution module controls the robotic arm to move it to the temporary storage area.
[0095] Control group:
[0096] The equipment uses basic visual recognition to locate shiitake mushrooms, and a motor-driven gripper holds the stem.
[0097] The robotic arm applies force to harvest the fruit, and without dehumidification or drying, the fruit is directly transferred to the temporary storage area.
[0098] Blank group:
[0099] Workers manually observe the maturity of shiitake mushrooms, pick the stems by hand, and place them in collection baskets.
[0100] Data collection phase
[0101] Record the total time, number of failed attempts to pick 100 shiitake mushrooms, and number of damaged mushrooms for each group.
[0102] The humidity value of the cross-section of shiitake mushrooms was measured using a humidity sensor after air drying for 5 minutes.
[0103] The energy consumption (kWh) of the air pump in the experimental group and the motor in the control group was statistically analyzed.
[0104] III. Detection Target Data
[0105] 1. Harvesting efficiency: The number of harvested items per unit time (pieces / hour).
[0106] 2. Clamping success rate: (Total number of harvested berries - Number of failed clamping attempts) / Total number of harvested berries × 100%.
[0107] 3. Mushroom damage rate: (Number of damaged mushrooms / Total number of harvested mushrooms) × 100% (Damage criteria: broken stem, damaged epidermis).
[0108] 4. Cross-sectional moisture: The moisture content (%) of the stipe cross-section measured 5 minutes after harvesting.
[0109] 5. Operational energy consumption: Total energy consumption (kWh) for each group to complete 100 harvests.
[0110] IV. Experimental Results Table
[0111] Harvesting efficiency (pieces / hour) 82.5±3.2 65.8±2.7 41.2±1.8 Clamping success rate (%) 96.7±1.5 83.3±2.3 92.1±1.9 Mushroom damage rate (%) 3.4±0.8 12.5±1.6 5.7±1.2 Cross-sectional humidity (%) 28.6±2.1 45.3±3.5 52.4±4.2 Operating energy consumption (kWh) 0.45±0.03 0.68±0.05 -
[0112] V. Discussion of Results
[0113] Harvesting efficiency and automation advantages: The experimental group's efficiency was 25.4% higher than the control group and 100.2% higher than the blank group, thanks to the coordinated control of the robotic arm and the harvesting system, especially the three-dimensional positioning and angle adjustment function of the drive components, which can quickly adapt to shiitake mushrooms at different growth angles. In contrast, the control group had a single gripper function, which took longer to position itself in the complex mushroom log environment; manual harvesting was limited by physical strength and concentration, resulting in the lowest efficiency.
[0114] Clamping success rate and damage control: The experimental group, using pneumatic clamping combined with dehumidification, achieved a significantly higher success rate than the control group (the motor-driven grippers were prone to slipping under high humidity), and the damage rate was reduced by 72.8%. The flexible clamping of the pneumatic system (with pressure sensors adjusting the inflation volume in real time) prevented stem breakage caused by uncontrolled mechanical gripping force, while the rigid clamping in the control group easily caused epidermal damage. Although the control group had a higher success rate, uneven manual force still led to some damage.
[0115] Cross-sectional humidity and preservation effect: The air-drying function in the experimental group reduced the cross-sectional humidity by 36.8% (compared to the control group) and 45.4% (compared to the blank group), which effectively inhibited bacterial growth and extended the shelf life. The control group had no air-drying measures, and the blank group had the highest humidity after manual harvesting, which was not conducive to subsequent storage.
[0116] Energy consumption and economy: The experimental group consumed 33.8% less energy than the control group. This is because the pneumatic system only supplies air on demand during clamping and drying, while the motor in the control group needs to run continuously to drive the gripper, demonstrating the energy-saving advantage of pneumatic drive. In large-scale production scenarios, the experimental group can further reduce labor costs and energy expenditures.
[0117] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.
Claims
1. A mushroom intelligent harvesting robot, comprising a mechanical arm (1), characterized in that, The robotic arm (1) is equipped with a mechanical claw at its end. The robotic arm (1) is connected to a harvesting system. The mechanical claw includes a base (4), a pneumatic assembly, and symmetrically arranged grippers (5). The pneumatic assembly is used to drive the grippers (5) to grip the stem of the shiitake mushroom. The pneumatic assembly includes a piston assembly fixedly connected to one end of the base (4) and air pipes (9) symmetrically arranged on both sides of the piston assembly. A wedge (14) is fixedly connected to the output end of the piston assembly. Each gripper (5) has a sliding groove (12) that slides and engages with both sides of the wedge (14). A spring (13) is sleeved on the piston assembly. The two ends of the spring (13) are fixedly connected to the side wall of the base (4) and the wedge (14) respectively. Each of the grippers (5) has an air chamber, and each of the grippers (5) has several ventilation holes at the bottom of the air chamber. The end of the air pipe (9) away from the piston assembly is connected to the air chamber. Among them, anti-slip pads (15) are provided on the inner side wall of the gripper (5), and pressure sensors connected to the harvesting system are provided in the anti-slip pads (15); Each of the grippers (5) is equipped with a capacitive humidity sensor that is connected to the harvesting system signal, and each of the anti-slip pads (15) has an opening that allows the capacitive humidity sensor to contact the stem of the shiitake mushroom. The harvesting system has built-in training datasets for several shiitake mushroom varieties. The system includes an image receiving module, a recognition module, a judgment module, and an execution module; among which: The image receiving module is connected to a camera component installed on the side wall of the robotic arm (1) and the base (4). The image receiving module is used to receive and process the image information of the shiitake mushroom from several angles synchronously acquired by the camera component through a preprocessing algorithm, and transmit the preprocessed key feature points to the recognition module. The preprocessing algorithm includes denoising the image information by using Gaussian filtering, enhancing the contrast of the image information by using histogram equalization, and extracting key feature points in the image information by using scale-invariant feature transformation algorithm. The key feature points include the surface texture and color of the shiitake mushroom cap, the opening and closing degree of the gills, and the relative size ratio of the stem and the cap. The identification module receives key feature points and is based on an improved YOLOv7 deep learning model. The YOLOv7 deep learning model is trained and optimized by accessing a training dataset. Based on the shiitake mushroom variety selected by the user, the corresponding shiitake mushroom training dataset is accessed. The maturity of the shiitake mushroom is identified based on key feature points, and the mushroom body, including the center of the cap and the base of the stem, is located in three-dimensional space to obtain three-dimensional coordinate information. The module also classifies the maturity of the identified shiitake mushrooms into three levels: "immature", "mature" and "overripe". The three-dimensional coordinate information, variety information and level determination results are packaged into an identification result and transmitted to the judgment module. The judgment module consists of three judgment logic steps, specifically: One judgment is made to receive the recognition result. If the recognition result is determined to be "ripe", then it can be picked and the three-dimensional coordinate information is transmitted to the execution module. The second judgment is used to receive humidity data and variety information transmitted by the capacitive humidity sensor and the corresponding humidity threshold to assess whether the moisture content of the mushroom body affects harvesting. If it does not affect harvesting, the harvesting information is transmitted to the execution module to control the mechanical claw to harvest. If it does affect harvesting, the adjustment information is transmitted to the third judgment. The system performs three checks to determine whether the shiitake mushroom has successfully detached from the substrate by comparing the measured torque with the standard torque range corresponding to the variety and maturity, and by combining the continuous pressure signal from the pressure sensor. If the harvest is successful, the subsequent processing information is transmitted to the execution module. If the harvest fails or adjustment information is received, the current processing information is output and transmitted to the execution module.
2. The intelligent mushroom harvesting robot according to claim 1, characterized in that, The piston assembly includes a gas tank (6) and a piston rod (10). The gas tank (6) is fixedly connected to one side of the base (4). A piston plate (8) is slidably connected inside the gas tank (6). The piston plate (8) divides the inside of the gas tank (6) into a first chamber (7) and a second chamber (11). The side of the first chamber (7) away from the piston plate (8) is connected to a gas delivery assembly for delivering gas, and the gas pipes (9) are all connected to the first chamber (7). The piston plate (8) is located on one side of the second chamber (11) and is fixedly connected to the piston rod (10). The end of the piston rod (10) away from the piston plate (8) is fixedly connected to the wedge (14). A spring (13) is sleeved on the piston rod (10).
3. The intelligent mushroom harvesting robot according to claim 2, characterized in that, The gas delivery assembly includes a gas pump (3) that is connected to the harvesting system via a signal.
4. The intelligent mushroom harvesting robot according to claim 3, characterized in that, The robotic arm (1) is provided with a drive assembly for driving the robotic claw to rotate left and right near the robotic claw. The drive assembly includes a drive cavity (2) and a servo motor (16). The drive cavity (2) is located inside the end of the robotic arm (1). The servo motor (16) is fixedly connected to the top wall inside the drive cavity (2). A transmission rod (17) is coaxially fixedly connected to the output shaft of the servo motor (16). A first connecting rod (18) is coaxially sleeved on the transmission rod (17). An arc groove (20) is provided around the bottom of the drive cavity (2). The end of the first connecting rod (18) away from the servo motor (16) extends into the arc groove (20) and slides in connection with the arc groove (20). A second connecting rod (19) is fixedly connected to the end of the first connecting rod (18) located in the arc groove (20). A fixing ring is fixedly connected to the end of the second connecting rod (19) away from the first connecting rod (18). The fixing ring is sleeved at the connection between the base (4) and the gas tank (6).
5. The intelligent mushroom harvesting robot according to claim 4, characterized in that, The execution module is used to receive three-dimensional coordinate information, picking information, subsequent processing information and current processing information and transmit the corresponding data to the robotic arm (1) or robotic gripper to perform operations, wherein: The three-dimensional coordinate information is used to send motion commands to the robotic arm (1) and servo motor (16), control the robotic claw to move precisely to the location of the stalk, and send clamping commands to the air pump (3); The picking information is used to send a quantitative inflation command to the air pump (3), which drives the gripper (5) to close and hold the stalk. At the same time, it receives real-time pressure data from the pressure sensor and adjusts the inflation amount to maintain the clamping force within the threshold range corresponding to the variety. The subsequent processing information is used to send a retraction command to the robotic arm (1) to control the robotic arm (1) to move the robotic gripper to the temporary storage area; The current processing information is divided into two execution logics: if it is adjustment information, a low-pressure continuous air supply command is sent to the air pump (3), and a dry airflow is blown to the surface of the mushroom stem through the ventilation hole to reduce the surface humidity and then trigger the secondary judgment and picking process again; if it is picking failure information, an air release command is sent to the air pump (3), so that the spring (13) is reset and the gripper (5) is released. At the same time, a reset command is sent to the robotic arm (1) to control the robotic gripper to return to the safe position and mark the coordinates of the mushroom as "awaiting manual processing" and store it in the system log.