Tobacco impurity picking gripper and system based on artificial intelligence and flexible grabbing technology
By integrating machine vision, deep learning, and flexible gripping technologies, the tobacco impurity removal equipment solves the integration, adaptability, and cost problems of existing equipment, achieving efficient and accurate impurity removal, reducing tobacco waste and mouthpiece clogging, and improving the level of automation in tobacco processing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- 南京焦耳科技有限责任公司
- Filing Date
- 2025-05-30
- Publication Date
- 2026-05-01
AI Technical Summary
Existing automated tobacco processing equipment suffers from poor system integration, limited adaptability, high cost, difficult maintenance, and problems such as tobacco waste and mouthpiece blockage caused by excessive suction force of flexible grippers.
The tobacco sorting gripper and system, which adopts artificial intelligence and flexible gripping technology, integrates machine vision, deep learning algorithms and flexible gripping technology. It accurately grasps debris through flexible finger modules and combines with a pneumatic conveying system to achieve full-process automation and efficient removal of impurities.
It achieves efficient and accurate impurity removal, reduces tobacco waste, prevents mouthpiece clogging, lowers production costs, and improves the working environment for workers.
Smart Images

Figure CN224179136U_ABST
Abstract
Description
A tobacco sorting gripper and system based on artificial intelligence and flexible gripping technology Technical Field
[0001] This utility model relates to the field of tobacco processing automation technology, and in particular to a tobacco sorting gripper and system based on artificial intelligence and flexible gripping technology. Background Technology
[0002] In the traditional tobacco processing industry, sorting is a crucial and labor-intensive step. This process involves identifying and removing non-tobacco substances from tobacco leaves, such as plastic fragments, paper scraps, and small stones. These impurities not only affect the quality of tobacco products but may also pose potential health risks to consumers. Therefore, ensuring that tobacco products contain no or minimal amounts of these impurities is of paramount importance.
[0003] In existing technologies, the impurity removal process mainly relies on manual visual inspection. Workers need to carefully inspect each tobacco leaf on the production line and manually remove impurities. This method has the following significant problems: low efficiency, insufficient accuracy, high cost, and poor working environment. To solve these problems, automated and intelligent impurity removal technologies have emerged. Early automated impurity removal equipment was mainly based on mechanical sieving and simple image processing technology, but these devices often could not accurately identify and remove all types of impurities, especially those that are similar in color and shape to tobacco leaves.
[0004] With the development of artificial intelligence technology, especially the advancements in machine vision and deep learning, the accuracy and efficiency of automated waste sorting equipment have been significantly improved. These technologies can be trained to identify various types of waste and guide robotic arms to precisely remove them. However, existing automated waste sorting equipment still has some limitations, such as:
[0005] 1. Poor system integration: Many automated sorting devices are added to the production line as separate modules, resulting in poor integration with existing production line equipment and control systems.
[0006] 2. Limited adaptability: Some automated equipment can only handle impurities of specific types or sizes, and is not adaptable to impurities of different types and shapes.
[0007] 3. Cost and maintenance issues: High-end automated sorting equipment is expensive and requires professional maintenance and calibration, which is a challenge for many tobacco processing companies.
[0008] Furthermore, existing intelligent impurity removal systems use solenoid valves to generate suction at the nozzle of a flexible gripper, sucking up and conveying impurities to a collection container for removal. However, this method has drawbacks: while the solenoid valve generates strong suction at the nozzle, quickly removing impurities from the conveyor belt, excessive suction can also draw out too much tobacco, clogging the nozzle and resulting in significant waste. Summary of the Invention
[0009] To address the aforementioned issues, this utility model discloses a tobacco impurity-removing gripper and system based on artificial intelligence and flexible gripping technology. It aims to overcome the limitations of existing technologies and provide a highly efficient, accurate, and cost-effective automated impurity-removing solution. By integrating the latest machine vision technology, deep learning algorithms, and intelligent robotic arm technology, it achieves rapid and accurate identification and removal of impurities from tobacco leaves, thereby improving the quality of tobacco products, reducing production costs, and improving the working environment for workers.
[0010] The specific plan is as follows:
[0011] On one hand, a tobacco sorting gripper based on artificial intelligence and flexible grasping technology is provided, including a flexible gripper set at the end of a robot. The flexible gripper includes a central fixed pneumatic conveyor, a central actuator, a drive linkage assembly, and flexible finger modules. The central fixed pneumatic conveyor is equipped with a compressed air inlet and is precisely controlled by a solenoid valve. The central actuator is sleeved outside the central fixed pneumatic conveyor and is equipped with a piston rod. After compressed air enters the air chamber below the central fixed pneumatic conveyor, it can push the piston rod to move downward. The flexible finger modules are evenly distributed around the outside of the central actuator and are connected to the central actuator by the drive linkage assembly. The drive linkage assembly includes drive linkage a and drive linkage b that are connected vertically.
[0012] Furthermore, the central fixed pneumatic conveyor has a debris outlet and a debris inlet connected vertically in the middle, and an annular air chamber on the side. The annular air chamber has a compressed air inlet and a compressed air outlet on the side, and the bottom is evenly distributed with lower lifting lugs that connect to the inner end of the drive connecting rod b.
[0013] Furthermore, the central actuator is a ring structure, with the same number of drive connecting rods as the flexible finger modules evenly distributed along its side in the circumferential direction, and piston rods evenly distributed along its bottom in the circumferential direction. The outer end of the drive connecting rod is provided with an upper lifting lug connected to the upper end of the drive connecting rod a, and the lower end of the piston rod is connected to a ring piston that can move up and down in the ring air chamber.
[0014] On the other hand, a tobacco cleaning system based on artificial intelligence and flexible gripping technology is provided, including a conveyor belt, an industrial camera, an industrial computer, a controller, an encoder, a solenoid valve, and a flexible gripper. The conveyor belt is used to feed materials into the working area. The industrial camera is placed above the conveyor belt to continuously photograph the materials and transmit the captured material images to the industrial computer. The industrial computer is used to process and analyze the image data to obtain the real coordinate information of the debris on the conveyor belt and send it to the controller. The encoder is installed on the drive shaft of the conveyor belt to obtain the motion data of the conveyor belt and send it to the controller. The controller dynamically adjusts the robot's speed and direction based on the debris coordinate information provided by the industrial computer and the conveyor belt motion data provided by the encoder, so that the robot reaches the coordinate position of the debris. The flexible gripper is used to grip and suck up the debris, and the solenoid valve is used to control the air supply of the flexible gripper.
[0015] Furthermore, the flexible gripper includes a central fixed pneumatic conveyor, a central actuator, a drive linkage assembly, and flexible finger modules. The central fixed pneumatic conveyor is equipped with a compressed air inlet and is precisely controlled by a solenoid valve. The central actuator is sleeved outside the central fixed pneumatic conveyor and is equipped with a piston rod. After compressed air enters the air chamber below the central fixed pneumatic conveyor, it can push the piston rod to move downward. The flexible finger modules are evenly distributed around the outside of the central actuator and are connected to the central actuator by the drive linkage assembly. The drive linkage assembly includes drive linkage a and drive linkage b that are connected vertically.
[0016] Furthermore, the central fixed pneumatic conveyor has a debris outlet and a debris inlet connected vertically in the middle, and an annular air chamber on the side. The annular air chamber has a compressed air inlet and a compressed air outlet on the side, and the bottom is evenly distributed with lower lifting lugs that connect to the inner end of the drive connecting rod b.
[0017] Furthermore, the central actuator is a ring structure, with the same number of drive connecting rods as the flexible finger modules evenly distributed along its side in the circumferential direction, and piston rods evenly distributed along its bottom in the circumferential direction. The outer end of the drive connecting rod is provided with an upper lifting lug connected to the upper end of the drive connecting rod a, and the lower end of the piston rod is connected to a ring piston that can move up and down in the ring air chamber.
[0018] This application combines flexible gripping with pneumatic conveying: by installing a flexible finger module in front of the flexible gripper, precise grasping of debris is achieved, avoiding the problems of tobacco waste and mouthpiece blockage caused by excessive suction in traditional flexible grippers. Furthermore, the system integrates and automates control: by integrating machine vision, deep learning algorithms, flexible gripping technology, and a pneumatic conveying system, the entire process from debris identification to removal is automated, improving the overall efficiency and stability of the system.
[0019] The beneficial effects of this utility model are as follows:
[0020] 1. Efficient debris removal: The precise gripping of the robotic arm and the coordinated work of the flexible gripper ensure that debris is removed quickly and accurately;
[0021] 2. Reduce tobacco waste: The flexible finger module grips the tobacco leaves, preventing them from being over-absorbed;
[0022] 3. Prevent nozzle clogging: The optimized conveyor design and clogging monitoring system effectively reduce clogging.
[0023] 4. Automation and intelligence: The entire system achieves efficient operation through automated control, reducing manual intervention. Attached Figure Description
[0024] Figure 1 is a schematic diagram of a flexible gripper.
[0025] Figure 2 is a cross-sectional view of AA in Figure 1.
[0026] Figure 3 is an enlarged view of B in Figure 2.
[0027] Figure 4 is a top view of Figure 1.
[0028] Figure 5 is a schematic diagram of a centrally fixed pneumatic conveyor.
[0029] Figure 6 is a front view of Figure 5.
[0030] Figure 7 is a cross-sectional view of CC in Figure 6.
[0031] Figure 8 is a schematic diagram of the central driver.
[0032] Figure 9 is a front view of Figure 8.
[0033] Figure 10 is a top view of Figure 8.
[0034] Figure 11 is a structural diagram of the system.
[0035] Figure 12 is a structural diagram of Figure 1 without the protective cover and door.
[0036] Figure 13 is a structural diagram of Figure 2 without the video cabinet.
[0037] List of reference numerals in the attached diagram:
[0038] 1-Center fixed pneumatic conveyor, 2-Center driver, 3-Drive link a, 4-Drive link b, 5-Flexible finger module, 6-Piston rod, 101-Debris outlet, 102-Compressed air inlet, 103-Annular air chamber, 104-Compressed air outlet, 105-Debris inlet, 202-Drive connecting rod, 7-Conveyor belt, 8-Industrial camera, 9-Robot, 10-Light source, 11-Protective cover, 12-Protective door, 13-Video cabinet, 14-Air conditioner. Detailed Implementation
[0039] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention.
[0040] As shown in Figures 1-4, this utility model provides a tobacco sorting gripper based on artificial intelligence and flexible gripping technology, including a flexible gripper set at the end of a robot. The flexible gripper includes a central fixed pneumatic conveyor 1, a central driver 2, a drive linkage assembly, and a flexible finger module 5. The central fixed pneumatic conveyor is equipped with a compressed air inlet and is precisely controlled by a solenoid valve. The central driver is sleeved outside the central fixed pneumatic conveyor and is equipped with a piston rod 6. After compressed air enters the air chamber below the central fixed pneumatic conveyor, it can push the piston rod to move downward. The flexible finger module is evenly distributed around the outside of the central driver and is connected to the central driver by the drive linkage assembly. The drive linkage assembly includes a drive linkage a3 and a drive linkage b4 that are connected vertically.
[0041] As shown in Figures 5-7, the central fixed pneumatic conveyor has a debris outlet 101 and a debris inlet 105 connected vertically in the middle, and an annular air chamber 103 on the side. The annular air chamber is provided with a compressed air inlet 102 and a compressed air outlet 104 on the side, and the bottom is evenly distributed with lower lifting lugs that are connected to the inner end of the drive connecting rod b.
[0042] As shown in Figures 8-10, the central actuator is a ring structure. The same number of drive connecting main rods 202 as the flexible finger modules are evenly distributed along the circumference on its side. Piston rods are evenly distributed along the circumference on its bottom. The outer end of the drive connecting main rod is provided with an upper lifting lug connected to the upper end of the drive connecting rod a. The lower end of the piston rod is connected to a ring piston that can move up and down in the ring air chamber.
[0043] This application allows the object to be grasped and then directly transported away by pneumatic conveying. Compared with traditional suction cups, grippers, and pneumatic conveying structures, it not only saves grasping and releasing time but also prevents blockage.
[0044] As shown in Figures 11-13, this utility model also provides a tobacco sorting system based on artificial intelligence and flexible gripping technology, including a conveyor belt 7, an industrial camera 8, an industrial computer, a controller, an encoder, a solenoid valve, and a flexible gripper. The conveyor belt is used to send materials into the working area. The industrial camera is placed above the conveyor belt to continuously photograph the materials and transmit the photographed material images to the industrial computer. A light source 10 is provided at the shooting position, and a video cabinet 13 is installed outside. An air conditioner 14 is installed on the top of the video cabinet. The industrial computer is used to process and analyze the image data to obtain the actual coordinate information of the debris on the conveyor belt and send it to the controller. The encoder is installed on the drive shaft of the conveyor belt to obtain the motion data of the conveyor belt and send it to the controller. The controller dynamically adjusts the movement speed and direction of the robot 9 according to the debris coordinate information provided by the industrial computer and the conveyor belt motion data provided by the encoder, so that the robot 9 reaches the coordinate position of the debris. A protective cover 11 and a protective door 12 are provided on the outside of the robot. The flexible gripper is placed at the end of the robot 9 to grip and suck up the debris. The solenoid valve is used to control the air supply of the flexible gripper.
[0045] During operation, materials (such as tobacco flakes) on the conveyor belt are fed into the system's working area. At this time, a high-resolution industrial camera mounted above the conveyor belt begins to continuously capture images of the material. The camera lens is carefully designed and adjusted to ensure clear images are captured under various lighting conditions. After the camera captures images of the material, this image data is sent to the industrial control computer via a high-speed data transmission interface. The industrial control computer runs specially developed image processing and analysis software with powerful image processing capabilities and advanced deep learning algorithms. The software first preprocesses the image, including denoising, grayscale conversion, and binarization, to highlight the features of debris in the image and eliminate unnecessary interference. Next, the software uses a deep learning model to extract features and classify the preprocessed image, accurately identifying debris such as plastic fragments, paper scraps, and small stones. After identifying the debris, the software calculates the X-axis and Y-axis coordinates of the debris on the image. These coordinates are determined based on the position of the image pixels, representing the specific location of the debris in the image. To convert image coordinates into actual physical coordinates, the software performs a series of coordinate transformation calculations based on the camera's imaging parameters (such as focal length and aperture) and the size and position of the conveyor belt. This allows the software to obtain the actual coordinates of the debris on the conveyor belt, providing accurate target location information for subsequent removal operations.
[0046] After obtaining the physical coordinates of the object, the industrial control machine sends this coordinate information to the robot controller. Upon receiving the coordinate information, the controller calculates the robot's trajectory and action commands based on the robot's kinematic model and the real-time position of the conveyor belt. An encoder mounted on the conveyor belt's drive shaft monitors the conveyor belt's speed and position in real time. The robot controller dynamically adjusts the robot's speed and direction based on the real-time position information provided by the encoder, ensuring the robot accurately reaches the designated coordinate position of the object.
[0047] Once the robot precisely reaches the designated target position, the robot controller immediately sends a specific signal, triggering the solenoid valve to a precise action. As a key control component of the pneumatic system, the solenoid valve can quickly and accurately control the air supply to the flexible gripper upon receiving the controller's command. The central fixed pneumatic conveyor 1 is equipped with a compressed air interface. With the precise control of the solenoid valve, compressed air enters the air chamber below the central fixed pneumatic conveyor 1 in an orderly manner, thereby pushing the piston rod 6 of the central actuator 2 downwards steadily. This process, through the ingenious connection and transmission of the drive linkage, enables the flexible gripper to quickly and accurately complete the grasping action of the object.
[0048] Following this, under the precise control of the solenoid valve, the centrally fixed pneumatic conveyor 1 efficiently transports the captured debris to a pre-set collection container, thus successfully completing the entire rejection process. The centrally fixed pneumatic conveyor is essentially an advanced device that cleverly utilizes compressed air to achieve material suction and conveying functions. During operation, it exhibits numerous significant advantages such as rapid response, flexible operation, and zero pollution. When the solenoid valve is in operation, the suction nozzle of the centrally fixed pneumatic conveyor can instantly generate a strong suction force, easily and quickly sucking up all the debris on the conveyor belt.
[0049] To further optimize system performance, a flexible finger module 5 was carefully installed at the front end of the central fixed pneumatic conveyor. This design plays a crucial role in extracting impurities from tobacco leaves. On one hand, the flexible finger module effectively assists the flexible grippers in accurately grasping impurities; on the other hand, it cleverly solves the tricky problem of excessive suction leading to the absorption of too many tobacco leaves and subsequent blockage of the suction port when the conveyor is started to directly suck up impurities, ensuring that the entire system can operate continuously, stably, and efficiently.
[0050] The technical means disclosed in this utility model are not limited to those disclosed in the above embodiments, but also include technical solutions composed of any combination of the above technical features. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of this utility model, and these improvements and modifications are also considered within the scope of protection of this utility model.
Claims
1. A tobacco picking and sorting gripper based on artificial intelligence and flexible gripping technology, characterized in that, The device includes a flexible gripper located at the end of a robot. The flexible gripper comprises a central fixed pneumatic conveyor, a central actuator, a drive linkage assembly, and flexible finger modules. The central fixed pneumatic conveyor is equipped with a compressed air inlet and is precisely controlled by a solenoid valve. The central actuator is fitted outside the central fixed pneumatic conveyor and is equipped with a piston rod. After compressed air enters the air chamber below the central fixed pneumatic conveyor, it can push the piston rod to move downward. The flexible finger modules are evenly distributed around the outside of the central actuator and are connected to the central actuator by the drive linkage assembly. The drive linkage assembly includes drive linkage a and drive linkage b that are connected vertically.
2. The tobacco sorting gripper based on artificial intelligence and flexible gripping technology according to claim 1, characterized in that, The central fixed pneumatic conveyor has a debris outlet and a debris inlet connected vertically in the middle, and an annular air chamber on the side. The annular air chamber has a compressed air inlet and a compressed air outlet on the side, and the bottom is evenly distributed with lower lifting lugs that connect to the inner end of the drive connecting rod b.
3. The tobacco sorting gripper based on artificial intelligence and flexible gripping technology according to claim 2, characterized in that, The central actuator is a ring structure with drive connecting rods of the same number as the flexible finger modules evenly distributed along its side circumferentially, and piston rods evenly distributed along its bottom circumferentially. The outer end of the drive connecting rod is provided with an upper lifting lug connected to the upper end of the drive connecting rod a, and the lower end of the piston rod is connected to a ring piston that can move up and down in the ring air chamber.
4. A tobacco sorting system based on artificial intelligence and flexible grasping technology, characterized in that, The gripper according to any one of claims 1-3 includes a conveyor belt, an industrial camera, an industrial computer, a controller, an encoder, a solenoid valve, and a flexible gripper. The conveyor belt is used to feed materials into the working area. The industrial camera is placed above the conveyor belt to continuously capture images of the materials and transmit the captured material images to the industrial computer. The industrial computer is used to process and analyze the image data to obtain the actual coordinate information of the debris on the conveyor belt and send it to the controller. The encoder is installed on the drive shaft of the conveyor belt to acquire the motion data of the conveyor belt and send it to the controller. The controller dynamically adjusts the robot's speed and direction of movement according to the debris coordinate information provided by the industrial computer and the conveyor belt motion data provided by the encoder, so that the robot reaches the coordinate position of the debris. The flexible gripper is used to grip and suck up the debris, and the solenoid valve is used to control the air supply of the flexible gripper.
5. A tobacco sorting system based on artificial intelligence and flexible grasping technology according to claim 4, characterized in that, The flexible gripper includes a central fixed pneumatic conveyor, a central actuator, a drive linkage assembly, and flexible finger modules. The central fixed pneumatic conveyor is equipped with a compressed air inlet and is precisely controlled by a solenoid valve. The central actuator is sleeved outside the central fixed pneumatic conveyor and is equipped with a piston rod. After compressed air enters the air chamber below the central fixed pneumatic conveyor, it can push the piston rod to move downward. The flexible finger modules are evenly distributed around the outside of the central actuator and are connected to the central actuator by the drive linkage assembly. The drive linkage assembly includes drive linkage a and drive linkage b that are connected vertically.
6. A tobacco sorting system based on artificial intelligence and flexible grasping technology according to claim 5, characterized in that, The central fixed pneumatic conveyor has a debris outlet and a debris inlet connected vertically in the middle, and an annular air chamber on the side. The annular air chamber has a compressed air inlet and a compressed air outlet on the side, and the bottom is evenly distributed with lower lifting lugs that connect to the inner end of the drive connecting rod b.
7. The tobacco picking and sorting system based on artificial intelligence and flexible gripping technology according to claim 6, characterized in that, The central actuator is a ring structure with drive connecting rods of the same number as the flexible finger modules evenly distributed along its side circumferentially, and piston rods evenly distributed along its bottom circumferentially. The outer end of the drive connecting rod is provided with an upper lifting lug connected to the upper end of the drive connecting rod a, and the lower end of the piston rod is connected to a ring piston that can move up and down in the ring air chamber.