Vision-based grabbing system

Abstract

The invention discloses a grabbing system based on vision, and belongs to the field of vision grabbing systems. The grabbing system comprises a composite actuator, the composite actuator is installed at the output end of the motion module, and a receding notch is formed in a machine shell of the composite actuator and used for receding for the camera. The material grabbing and collecting unit comprises a grabbing station and a grabbing camera, the grabbing station is located within the execution range of the composite actuator, and the grabbing camera is fixed to the upper portion of the grabbing station and matched with the receding notch; the discharging collection unit comprises a placement station and a placement camera, the placement station is located in the execution range of the composite actuator, and the placement camera is fixed to the upper portion of the placement station and matched with the receding notch; and the upward shooting camera is arranged in the execution range of the composite executor and located between the grabbing station and the placing station. The visual grabbing system provided by the invention not only improves the efficiency of wafer grabbing work, but also ensures high precision of work.

Description

[0001] Divisional Application Instructions

[0002] This application is a divisional application based on Chinese patent application 202411348693.3. The application date of the original application is September 26, 2024. The name of the invention is "A vision-based grasping system and grasping method thereof". Technical Field

[0003] The invention belongs to the field of vision grabbing systems, in particular to a vision-based grabbing system. Background Art

[0004] A visual grasping system is an automated system that combines a visual sensor (such as a camera) and a manipulator (such as a composite actuator) to identify, locate, and grasp objects. This system is widely used in many industries such as manufacturing, logistics, and packaging to improve production efficiency and automation levels.

[0005] In practical applications, in order to reduce the cost of the visual acquisition device, the composite actuator and the visual acquisition device are installed together on the motion module to achieve accurate grasping of the material. Taking the visual grasping system grasping wafers as an example, the existing technology has the following problems:

[0006] 1. Efficiency issue: In each grasping cycle, the acquisition device and the composite actuator need to alternately enter the wafer grasping area, which will cause additional actions and waiting time, thus limiting work efficiency.

[0007] 2. Accuracy problem: The operating accuracy of the motion module directly affects the positioning accuracy and execution accuracy of the composite actuator and acquisition device, which in turn causes operational errors.

[0008] In order to solve the above problems, the mechanical structure of the visual grasping system needs to be optimized to achieve high-precision and high-efficiency wafer grasping. Summary of the invention

[0009] Purpose of the invention: To provide a vision-based grasping system to solve the above-mentioned problems existing in the prior art.

[0010] Technical solution: A vision-based grasping system, characterized by comprising:

[0011] A composite actuator, wherein the composite actuator is installed at the output end of the motion module, and a housing of the composite actuator forms a gap for making way for the camera;

[0012] A material grabbing and collecting unit, the material grabbing and collecting unit comprising a grabbing station and a grabbing camera, the grabbing station is located within the execution range of the composite actuator, the grabbing camera is fixed above the grabbing station and cooperates with the clearance gap;

[0013] A material discharging and collecting unit, the material discharging and collecting unit comprises a placing station and a placing camera, the placing station is located within the execution range of the composite actuator, the placing camera is fixed above the placing station and cooperates with the clearance gap;

[0014] An upward-shooting camera is arranged within the execution range of the compound actuator and is located between the grabbing station and the placing station.

[0015] Furthermore, the composite actuator includes a housing, a linear motion module, an output assembly and a rotary motor. The housing is an L-shaped shell structure. The linear motion module is arranged in a long cavity of the housing. The output assembly is arranged in a short cavity of the housing and is connected to the output end of the linear motion module through a guide rail assembly. The output assembly is located below the clearance gap. The rotary motor is mounted on the output assembly and is transmission-connected to the output shaft of the output assembly.

[0016] Furthermore, the linear motion module includes a stator and a mover, the stator is installed in a long cavity of the casing, and the mover is connected to the output assembly through a guide rail assembly; the guide rail assembly includes a slide rail installed in the long cavity, an air path block installed at the movable end of the slide rail, and a connecting plate connecting the air path block and the mover.

[0017] Furthermore, the output assembly includes an output kit, a sealed bearing and an output shaft, the output kit is installed on the air path carrier, the end of the output kit extends to the outside of the casing, the output shaft is arranged in the output kit and is connected to the rotating motor through a coupling, the two sealed bearings are arranged between the output kit and the output shaft, and a sealed cavity is formed between the sealed bearings, the output kit and the output shaft, a vent hole is opened on the output shaft, the vent hole is located in the sealed cavity, an output air hole is opened on the output kit, and the output air hole is communicated with the sealed cavity.

[0018] Furthermore, the compound actuator also includes an air circuit assembly, the output shaft is connected to the external air source through the air circuit assembly, the air circuit assembly includes a ventilation block, a bent air pipe and a connecting air pipe, the ventilation block is installed in the long cavity of the casing, one end of the ventilation block is connected to the bent air pipe, and the other end is connected to the air nozzle extending to the outside of the casing, the ventilation block is connected to the carrier block through hole opened in the connecting air circuit carrier block, and the carrier block through hole is used to output the air hole to be connected through the connecting air pipe.

[0019] Furthermore, the composite actuator also includes a grating sensor, and the grating sensor includes a reader mounted on the housing and a grating ruler mounted on the air path carrier.

[0020] Furthermore, the composite actuator 6 also includes a circuit assembly 67, which includes a wiring terminal mounted on the housing and a wiring harness connected to the wiring terminal, and the wiring harness is electrically connected to the mover, the rotating motor and the reader of the grating sensor respectively.

[0021] A grasping method of a grasping system based on vision, comprising the following steps:

[0022] Step 1: Use a grabbing camera to obtain a wafer placement image of the grabbing station, and process the wafer placement image to obtain positioning information of the wafer to be grabbed; use a placement camera to obtain a tray image of the placement station, and process the tray image to obtain positioning information of the tray;

[0023] Step 2: The motion module moves the output end of the composite actuator to face the wafer according to the positioning information of the wafer to be grasped, the composite actuator grasps the wafer, and then moves the output end of the composite actuator to face the upward camera;

[0024] Step 3: When the composite actuator passes directly above the upward camera, the upward camera is used to obtain a grasping image of the grasped wafer, and the grasping image of the grasped wafer is processed to obtain positioning information of the grasped wafer;

[0025] Step 4: Use the positioning information of the captured wafer and the positioning information of the receiving tray to obtain the execution information of the motion module and the composite actuator, and place the wafer into the tray of the placement station according to the execution information.

[0026] Furthermore, the step 1 comprises the following steps:

[0027] Step 11: Apply image enhancement technology and filtering denoising technology to pre-process the wafer placement image and the tray image to obtain a wafer pre-processing image and a tray pre-processing image;

[0028] Step 12: Using Hough transform to obtain edge lines of the grabbing station and the wafer in the wafer preprocessing image, and using Hough transform to obtain edge lines of the placing station and the tray in the tray preprocessing image;

[0029] Step 13: Obtain the geometric center point P of the grasping station according to the edge line of the grasping station, and obtain the geometric center point p of the wafer according to the edge line of the wafer; based on the coordinates of the geometric center point P in the execution coordinate system of the motion module, obtain the execution coordinates of the geometric center point p in the motion module, and use it as the positioning information of the wafer to be grasped;

[0030] Among them, the coordinates of the geometric center point P in the motion module execution coordinate system are (X1, Y1), and the coordinates of the geometric center point p in the motion module execution coordinate system are (X1+Δx1, Y1+Δy1), Δx1 and Δy1 are the offsets of the geometric center point p in the motion module execution coordinate system relative to the geometric center point P in the X and Y directions respectively;

[0031] Step 14: Obtain the geometric center point Q of the placement station according to the edge line of the placement station, and obtain the geometric center point q of the pallet according to the edge line of the pallet; based on the coordinates of the geometric center point Q in the motion module execution coordinate system, obtain the execution coordinates of the geometric center point q in the motion module (5) execution coordinate system; use the geometric center point q and the midpoint of any pallet edge line to obtain the pallet vector; use the execution coordinates of the geometric center point q in the motion module execution coordinate system and the pallet vector as the positioning information of the pallet;

[0032] Among them, the coordinates of the geometric center point Q in the motion module execution coordinate system are (X2, Y2), and the coordinates of the geometric center point q in the motion module execution coordinate system are (X2+Δx2, Y2+Δy2), Δx2 and Δy2 are the offsets of the geometric center point q in the motion module execution coordinate system relative to the geometric center point Q in the X and Y directions, respectively.

[0033] Furthermore, the step 3 comprises the following steps:

[0034] Step 31: Apply image enhancement technology and filtering denoising technology to pre-process the captured image of the wafer to obtain a captured pre-processed image;

[0035] Step 32: Using Hough transform to obtain edge lines of the housing and the captured wafer in the captured pre-processed image;

[0036] Step 33: obtaining the geometric center point S of the housing according to the edge line of the housing, and obtaining the execution coordinates of the geometric center point S in the motion module execution coordinate system based on the coordinates of the geometric center point S in the motion module execution coordinate system; obtaining the actuator vector using the geometric center point s and the midpoint of any edge line of the housing (61); using the execution coordinates of the geometric center point s in the motion module execution coordinate system and the actuator vector as positioning information of the captured wafer;

[0037] Among them, the coordinates of the geometric center point S in the motion module execution coordinate system are (X3, Y3), and the coordinates of the geometric center point s in the motion module execution coordinate system are (X3+Δx3, Y3+Δy3), Δx3 and Δy3 are the offsets of the geometric center point s in the motion module execution coordinate system relative to the geometric center point S in the X and Y directions, respectively.

[0038] Further, the step 4 comprises the following steps:

[0039] Step 41: Obtain the deflection angle using the tray vector and the actuator vector;

[0040] Step 42: Obtain the execution route of the motion module (5) using the coordinates (X3+Δx3, Y3+Δy3) and the coordinates (X2+Δx2, Y2+Δy2);

[0041] Step 43: The deflection angle and the execution route are used as execution information.

[0042] Beneficial effects: The composite actuator provided by the present invention is installed at the output end of the motion module, and the clearance notch on its housing allows the camera to be directly aimed at the target object, without the need for the composite actuator and the camera to alternately enter the working area, thus reducing unnecessary actions and waiting time and improving work efficiency. In terms of execution accuracy, the camera is fixedly installed on the external end face and is no longer affected by the motion module, and the setting of the overhead camera can provide additional visual information during the wafer grabbing and releasing process, ensuring the positioning accuracy and stability during the wafer grabbing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a structural schematic diagram of the grabbing system in the present invention;

[0044] Figure 2 It is a schematic diagram of the working condition when the composite actuator and the grabbing camera are co-located in the present invention;

[0045] Figure 3 is a front view of the composite actuator of the present invention;

[0046] Figure 4 It is a schematic diagram of the internal structure of the composite actuator in the present invention;

[0047] Figure 5 It is a schematic diagram of the connection structure between the linear motion module and the output component in the present invention;

[0048] Figure 6 is a schematic diagram of the structure of the circuit assembly in the present invention;

[0049] Figure 7 It is a structural schematic diagram of the gas path component in the present invention;

[0050] Figure 8 It is a schematic diagram of the internal structure of the output component in the present invention;

[0051] Fig. 9 It is a flow chart of the grasping method based on the grasping system structure in the present invention.

[0052] The accompanying drawings are marked as follows: 1, grabbing station; 2, grabbing camera; 3, placing station; 4, placing camera; 5, motion module; 6, compound actuator; 60, clearance gap; 61, housing; 62, linear motion module; 621, stator; 622, mover; 63, guide rail assembly; 631, slide rail; 632, air path carrier; 633, connecting plate; 64, output assembly; 641, output kit; 642, seal Sealed bearing; 643, output shaft; 644, end gasket; 645, output air hole; 65, rotating motor; 651, coupling; 66, grating sensor; 661, reader; 662, grating ruler; 67, circuit assembly; 671, terminal block; 672, wiring; 68, air circuit assembly; 681, air nozzle; 682, ventilation block; 683, bent air pipe; 684, connecting air pipe; 7, overhead camera. DETAILED DESCRIPTION

[0053] In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is apparent to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features well known in the art are not described.

[0054] Embodiment 1: This embodiment proposes a new visual capture system to solve the problems of the existing visual capture system in terms of execution efficiency and execution accuracy.

[0055] like Figure 1-Figure 8 As shown, a vision-based grasping system is characterized by comprising:

[0056] The compound actuator 6 is installed at the output end of the motion module 5, and the housing 61 of the compound actuator 6 forms a clearance gap 60 for making way for the camera; the compound actuator 6 is installed at the output end of the motion module 5, and the clearance gap 60 on its housing 61 allows the camera to be directly aimed at the target object, without the need for the compound actuator 6 and the camera to alternately enter the working area, thereby reducing unnecessary movements and waiting time. By reasonably arranging the positional relationship between the camera and the compound actuator 6, mutual occlusion between the camera and the actuator is avoided, reducing unnecessary movements caused by avoidance, and significantly improving the speed of the grasping cycle.

[0057] The material grabbing and collecting unit includes a grabbing station 1 and a grabbing camera 2. The grabbing station 1 is located within the execution range of the compound actuator 6. The grabbing camera 2 is fixed above the grabbing station 1 and cooperates with the clearance gap 60; the material unloading and collecting unit includes a placement station 3 and a placement camera 4. The placement station 3 is located within the execution range of the compound actuator 6. The placement camera 4 is fixed above the placement station 3 and cooperates with the clearance gap 60; the camera configuration of the grabbing station and the placement station: the grabbing station 1 and the placement station 3 are equipped with dedicated cameras (grabbing camera 2 and placement camera 4). These cameras are fixed on the external end surface and can quickly and accurately capture the position information of the target object. The shooting accuracy is not affected by the motion module 5, which helps to improve the positioning accuracy of the entire system.

[0058] The upward camera 7 is arranged within the execution range of the compound actuator 6 and between the grabbing station 1 and the placing station 3. The upward camera 7 is located on the working path of the compound actuator 6, and can provide additional visual information during the process of grabbing and releasing the wafer, thereby ensuring the positioning accuracy and stability during the wafer grabbing process.

[0059] In summary, this vision-based grasping system not only improves the efficiency of wafer grasping, but also ensures high precision of the operation. It is very suitable for use in application scenarios with high requirements for speed and precision.

[0060] like Figure 3 and Figure 4 As shown, the compound actuator 6 includes a housing 61, a linear motion module 62, an output assembly 64 and a rotary motor 65. The housing 61 is an L-shaped shell structure. The linear motion module 62 is arranged in the long cavity of the housing 61. The output assembly 64 is arranged in the short cavity of the housing 61 and is connected to the output end of the linear motion module 62 through a guide rail assembly 63. The output assembly 64 is located below the clearance notch 60. The rotary motor 65 is installed on the output assembly 64 and is transmission-connected to the output shaft 643 of the output assembly 64.

[0061] The design of the clearance notch 60 on the housing 61 of the compound actuator 6 allows the camera to be directly aligned with the wafer, without the need for the compound actuator 6 and the camera to alternately enter the working area, thus reducing unnecessary movements and waiting time. The compound actuator 6 is composed of multiple independent components, which is convenient for disassembly and maintenance, reducing maintenance costs and time. The unique design of the compound actuator 6 lies in the clearance notch 60 of the housing 61. The position design of the clearance notch 60 allows the camera to be directly aligned with the wafer, providing convenient conditions for the camera, the output shaft 643 and the wafer to be perpendicular to each other, which is conducive to the linkage of visual acquisition and the output of the compound actuator 6.

[0062] like Figure 5-Figure 8As shown, the linear motion module 62 includes a stator 621 and a mover 622. The stator 621 is installed in the long cavity of the housing 61. The mover 622 is connected to the output assembly 64 through a guide rail assembly 63. The guide rail assembly 63 includes a slide rail 631 installed in the long cavity, an air path block 632 installed at the movable end of the slide rail 631, and a connecting plate 633 connecting the air path block 632 and the mover 622. In particular, the air path block 632 and the connecting plate 633 can be integrally processed to facilitate transmission connection with the mover 622. The guide rail assembly 63 can provide stable support for the transmission of the mover 622 to maintain the output accuracy of the linear motion module 62.

[0063] like Figure 8 As shown, the output assembly 64 includes an output kit 641, a sealed bearing 642 and an output shaft 643. The output kit 641 is installed on the gas path carrier 632. The end of the output kit 641 extends to the outside of the housing 61. The output shaft 643 is arranged in the output kit 641 and connected to the rotating motor 65 through a coupling 651. Two sealed bearings 642 are arranged between the output kit 641 and the output shaft 643. A sealed cavity is formed between the sealed bearings 642, the output kit 641 and the output shaft 643. A vent hole is provided on the output shaft 643, and the vent hole is located in the sealed cavity. An output air hole 645 is provided on the output kit 641, and the output air hole 645 is connected with the sealed cavity. In order to improve the internal sealing of the output kit 641 and the stable operation of the output shaft 643, an end gasket 644 is installed at the end of the output kit 641. The design of the sealed cavity can smoothly transmit the gas source to the output shaft 643, which is ingeniously designed and has high transmission quality.

[0064] like Figure 7 As shown, the compound actuator 6 also includes an air circuit component 68, and the output shaft 643 is connected to the external air source through the air circuit component 68. The air circuit component 68 includes a ventilation block 682, a bent air pipe 683 and a connecting air pipe 684. The ventilation block 682 is installed in the long cavity of the housing 61. One end of the ventilation block 682 is connected to the bent air pipe 683, and the other end is connected to the air nozzle 681 extending to the outside of the housing 61. The ventilation block 682 is connected to a carrier block through hole opened in the connecting air circuit carrier block 632, and the carrier block through hole is used to connect the output air hole 645 through the connecting air pipe 684.

[0065] like Figure 6 As shown, the compound actuator 6 also includes a grating sensor 66, which includes a reader 661 installed on the housing 61 and a grating ruler 662 installed on the air path carrier 632. The grating sensor 66 can monitor the output of the mover 622. The compound actuator 6 also includes a circuit component 67, which includes a terminal 671 installed on the housing 61 and a flat cable 672 connected to the terminal 671, and the flat cable 672 is electrically connected to the mover 622, the rotating motor 65 and the reader 661 of the grating sensor 66 respectively.

[0066] Embodiment 2: This embodiment is based on the visual grasping system provided in Embodiment 1 and proposes a grasping method to improve the transfer accuracy of wafers.

[0067] like Fig. 9 As shown, a grasping method of a vision-based grasping system is characterized by comprising the following steps:

[0068] Step 1: Use the grabbing camera 2 to obtain the wafer placement image of the grabbing station 1, and process the wafer placement image to obtain the positioning information of the wafer to be grabbed; use the placement camera 4 to obtain the tray image of the placement station 3, and process the tray image to obtain the tray positioning information; use the grabbing camera 2 and the placement camera 4 to obtain the image information of the wafer and the tray, and obtain accurate positioning information through image processing technology. And the camera (grabbing camera 2 and placement camera 4) is fixedly installed on the external end surface and will not be affected by the motion module, which helps to improve the positioning accuracy of the entire system.

[0069] Step 2: The motion module 5 carries the output end of the compound actuator 6 to face the wafer according to the positioning information of the wafer to be grasped, the compound actuator 6 grasps the wafer, and then carries the output end of the compound actuator 6 to face the overhead camera 7; the motion module 5 carries the compound actuator 6 according to the positioning information, and accurately controls the movement of the compound actuator 6, thereby reducing unnecessary waiting time and movement, thereby improving the speed of the grasping cycle.

[0070] Step 3: When the composite actuator 6 passes directly above the overhead camera 7, the overhead camera 7 is used to obtain a grasping image of the grasped wafer, and the grasping image of the grasped wafer is processed to obtain the positioning information of the grasped wafer; during the process of grasping and releasing the wafer, the overhead camera 7 can provide additional visual information to ensure the positioning accuracy and stability of the wafer during the entire grasping process.

[0071] Step 4: Utilize the positioning information of the captured wafer and the positioning information of the receiving tray to obtain the execution information of the motion module 5 and the composite actuator 6, and place the wafer into the tray of the placement station 3 according to the execution information.

[0072] Specifically, step 1 includes the following steps:

[0073] Step 11: Apply image enhancement technology and filtering denoising technology to pre-process the wafer placement image and the tray image to obtain a wafer pre-processing image and a tray pre-processing image;

[0074] Step 12: Using Hough transform to obtain edge lines of grabbing station 1 and the wafer in the wafer preprocessing image, and using Hough transform to obtain edge lines of placing station 3 and the tray in the tray preprocessing image;

[0075] Step 13: Obtain the geometric center point P of the grabbing station 1 according to the edge line of the grabbing station 1, and obtain the geometric center point p of the wafer according to the edge line of the wafer; based on the coordinates of the geometric center point P in the execution coordinate system of the motion module 5, obtain the execution coordinates of the geometric center point p in the motion module 5, and use it as the positioning information of the wafer to be grabbed;

[0076] The coordinates of the geometric center point P in the execution coordinate system of the motion module 5 are (X1, Y1), and the coordinates of the geometric center point p in the execution coordinate system of the motion module 5 are (X1+Δx1, Y1+Δy1), Δx1 and Δy1 are the offsets of the geometric center point p in the execution coordinate system of the motion module 5 relative to the geometric center point P in the X and Y directions, respectively;

[0077] Step 14: Obtain the geometric center point Q of the placement station 3 according to the edge line of the placement station 3, and obtain the geometric center point q of the pallet according to the edge line of the pallet; based on the coordinates of the geometric center point Q in the execution coordinate system of the motion module 5, obtain the execution coordinates of the geometric center point q in the execution coordinate system of the motion module 5; obtain the pallet vector using the geometric center point q and the midpoint of any pallet edge line; use the execution coordinates of the geometric center point q in the execution coordinate system of the motion module 5 and the pallet vector as the positioning information of the pallet;

[0078] Among them, the coordinates of the geometric center point Q in the execution coordinate system of the motion module 5 are (X2, Y2), and the coordinates of the geometric center point q in the execution coordinate system of the motion module 5 are (X2+Δx2, Y2+Δy2), Δx2 and Δy2 are the offsets of the geometric center point q in the execution coordinate system of the motion module 5 relative to the geometric center point Q in the X and Y directions, respectively.

[0079] Specifically, step 3 includes the following steps:

[0080] Step 31: Apply image enhancement technology and filtering denoising technology to pre-process the captured image of the wafer to obtain a captured pre-processed image;

[0081] Step 32: using Hough transform to obtain edge lines of the housing 61 and the captured wafer in the captured pre-processed image;

[0082] Step 33: Obtain the geometric center point S of the housing 61 according to the edge line of the housing 61, and obtain the execution coordinates of the geometric center point s in the execution coordinate system of the motion module 5 based on the coordinates of the geometric center point S in the execution coordinate system of the motion module 5; obtain the actuator vector using the geometric center point s and the midpoint of any edge line of the housing 61; use the execution coordinates of the geometric center point s in the execution coordinate system of the motion module 5 and the actuator vector as the positioning information of the captured wafer;

[0083] Among them, the coordinates of the geometric center point S in the execution coordinate system of the motion module 5 are (X3, Y3), and the coordinates of the geometric center point s in the execution coordinate system of the motion module 5 are (X3+Δx3, Y3+Δy3), Δx3 and Δy3 are the offsets of the geometric center point s in the execution coordinate system of the motion module 5 relative to the geometric center point S in the X and Y directions respectively.

[0084] Further, step 4 includes the following steps:

[0085] Step 41: Obtain the deflection angle using the tray vector and the actuator vector;

[0086] Step 42: Obtain the execution route of the motion module 5 using the coordinates (X3+Δx3, Y3+Δy3) and the coordinates (X2+Δx2, Y2+Δy2);

[0087] Step 43: The deflection angle and the execution route are used as execution information.

[0088] In summary, the installation positions of the grabbing station 1, the placing station 3 and the overhead camera 7 are fixed, so their coordinates in the execution coordinate system of the motion module 5 are determined.

[0089] In particular, in order to reduce the difficulty of obtaining the tray vector and the actuator vector, identification points can be directly set on the housing 61 and the tray to avoid the operation of capturing the midpoint of the contour line during data processing.

[0090] The preferred embodiments of the invention are described in detail above in conjunction with the accompanying drawings; however, the invention is not limited to the specific details in the above embodiments; within the technical concept of the invention, various equivalent transformations may be made to the technical solutions of the invention, and all of these equivalent transformations fall within the protection scope of the invention.

Claims

1. A vision-based grasping system, characterized in that: include: A composite actuator (6), the composite actuator (6) being mounted at the output end of the motion module (5), the housing (61) of the composite actuator (6) forming a clearance notch (60) for making way for the camera; A material grabbing and collecting unit, the material grabbing and collecting unit comprising a grabbing station (1) and a grabbing camera (2), the grabbing station (1) being located within the execution range of the composite actuator (6), the grabbing camera (2) being fixed above the grabbing station (1) and cooperating with the clearance notch (60); A material discharging and collecting unit, the material discharging and collecting unit comprising a placement station (3) and a placement camera (4), the placement station (3) being located within the execution range of the composite actuator (6), the placement camera (4) being fixed above the placement station (3) and cooperating with the clearance notch (60); An upward-shooting camera (7), wherein the upward-shooting camera (7) is arranged within the execution range of the composite actuator (6) and is located between the grabbing station (1) and the placing station (3). The composite actuator (6) comprises a housing (61), a linear motion module (62), an output assembly (64) and a rotary motor (65); the housing (61) is an L-shaped housing structure; the linear motion module (62) is arranged in a long cavity of the housing (61); the output assembly (64) is arranged in a short cavity of the housing (61) and is connected to the output end of the linear motion module (62) through a guide rail assembly (63); the output assembly (64) is located below the clearance notch (60); the rotary motor (65) is mounted on the output assembly (64) and is transmission-connected to an output shaft (643) of the output assembly (64).

2. A vision-based grasping system according to claim 1, characterized in that: The linear motion module (62) comprises a stator (621) and a mover (622); the stator (621) is installed in a long cavity of the housing (61); the mover (622) is connected to an output assembly (64) via a guide rail assembly (63); the guide rail assembly (63) comprises a slide rail (631) installed in the long cavity, an air path carrier block (632) installed at a movable end of the slide rail (631), and a connecting plate (633) connecting the air path carrier block (632) and the mover (622).

3. A vision-based grasping system according to claim 2, characterized in that: The output assembly (64) comprises an output assembly (641), a sealed bearing (642) and an output shaft (643); the output assembly (641) is mounted on the air path carrier (632); the end of the output assembly (641) extends to the outside of the housing (61); the output shaft (643) is arranged in the output assembly (641) and connected to the rotating motor (65) through a coupling (651); the two sealed bearings (642) are arranged between the output assembly (641) and the output shaft (643); a sealed cavity is formed between the sealed bearings (642), the output assembly (641) and the output shaft (643); a vent hole is provided on the output shaft (643); the vent hole is located in the sealed cavity; the output assembly (641) is provided with an output air hole (645); the output air hole (645) is communicated with the sealed cavity.

4. A vision-based grasping system according to claim 3, characterized in that: The composite actuator (6) further comprises an air circuit assembly (68), wherein the output shaft (643) is connected to an external air source via the air circuit assembly (68), wherein the air circuit assembly (68) comprises a ventilation block (682), a bent air pipe (683) and a connecting air pipe (684), wherein the ventilation block (682) is installed in a long cavity of the housing (61), wherein one end of the ventilation block (682) is connected to the bent air pipe (683), and the other end is connected to an air nozzle (681) extending to the outside of the housing (61), wherein the ventilation block (682) is connected to a carrier block through hole provided in a connecting air circuit carrier block (632), wherein the carrier block through hole is used to communicate with the output air hole (645) via the connecting air pipe (684).

5. A vision-based grasping system according to claim 4, characterized in that: The composite actuator (6) further comprises a circuit assembly (67), wherein the circuit assembly (67) comprises a wiring terminal (671) mounted on the housing (61) and a flat cable (672) connected to the wiring terminal (671), wherein the flat cable (672) is electrically connected to the mover (622), the rotary motor (65) and the reader (661) of the grating sensor (66), respectively.

Images