Intelligent compound eye vision sensor and control method thereof

Abstract

The invention relates to the technical field of visual sensors, and discloses an intelligent compound eye visual sensor and a control method thereof.The intelligent compound eye visual sensor comprises an annular shell, a plurality of sub-visual modules evenly and annularly distributed in the annular shell and a lens visual angle adjusting module rotationally arranged on the annular shell; the lens visual angle adjusting module comprises an annular lens plate and a plurality of deflection lenses rotationally arranged on the annular lens plate, and the deflection lenses are used for deflecting the observation direction of the sub-vision module. Visual recognition is completed through the small-size compound eye recognition structure, the industrial robot is installed on the industrial robot, the stroke range of the industrial robot cannot be affected, meanwhile, the observation visual angle can be greatly avoided through recognition completed through the compound eye structure, and the whole sensor achieves comprehensive and clear observation; a more accurate front effect can be achieved on visual identification control of the industrial robot, and a comprehensive protection effect can also be achieved on the sub-visual modules.

Description

Technical Field

[0001] This invention relates to the technical field of visual sensors, and more particularly to an intelligent compound eye visual sensor and its control method. Background Technology

[0002] With the comprehensive advancement of intelligent manufacturing, industrial robots are increasingly being used at workstations on production lines. In the intelligent manufacturing industry, industrial robots with visual recognition capabilities are evolving from "robotic arms" into "cognitive collaborative partners." Through the cooperation of robotic arms and visual sensors, robotic arms can automatically capture real-time working conditions to analyze and locate materials, achieving various manufacturing needs such as flexible grasping, precise assembly, and defect detection. This upgrades the production line from "blind operation" to "adaptive operation," significantly improving efficiency and yield, and supporting a future of intelligent manufacturing characterized by small-batch, multi-variety production and even personalized customization.

[0003] Currently, there are two main ways to install vision sensors for industrial robots. One is to fix the vision sensor directly near the work area to complete the visual recognition operation. The advantage of this fixing method is that the fixing and installation operation is simple, and the installation of the vision sensor will not affect the movement of the robotic arm. However, the disadvantage is that the observation position of the vision sensor usually needs to be adjusted specifically when performing different processing operations to ensure that the vision sensor can complete the visual recognition clearly and comprehensively. After moving the position of the vision sensor, recalibration is required, which is troublesome.

[0004] Another installation method involves using vision sensors mounted on the robotic arm. One or more vision sensors are mounted on a bracket at the forearm near the end effector position to observe the workstation. The advantage is that the vision follows the movement of the robotic arm, requiring almost no recalibration when the robot is used in different operations. However, this installation method can obstruct the vision sensors' line of sight. To ensure clear and comprehensive observation of the robotic arm's working position, multiple cameras are typically mounted on the arm, and each camera's bracket is extended a considerable distance to prevent the vision sensors from being blocked. While this solves the obstruction problem, it also results in excessive installation space, potentially restricting the robotic arm's movement. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent compound eye vision sensor and its control method to solve the problem of inconvenience in using existing vision sensors.

[0006] This invention is achieved through the following technical solution: On one hand, an intelligent compound eye vision sensor is provided, including an annular housing, a plurality of sub-vision modules uniformly arranged within the annular housing, and a lens angle adjustment module rotatably mounted on the annular housing. The lens angle adjustment module includes an annular lens plate and a plurality of deflecting lenses rotatably mounted on the annular lens plate. The deflecting lenses are used to deflect the observation direction of the sub-vision modules. The lens angle adjustment module further includes a first driving mechanism and a second driving mechanism. The first driving mechanism is fixed within the annular housing and is used to drive the entire annular lens plate to rotate. The second driving mechanism is fixed to the sub-vision modules and is used to drive the deflecting lenses located at the lens end of the sub-vision modules to rotate.

[0007] In one possible design, the annular lens plate is divided into multiple lens areas by the ribs, the number of lens areas being the same as the number of sub-vision modules, and each lens area is provided with deflecting lenses of multiple deflection angles.

[0008] In one possible design, the lens area includes three types of deflecting lenses: one is a deflecting lens without deflection angle, and when one sub-vision module is aligned with the deflecting lens without deflection angle, the remaining sub-vision modules are also aligned with the non-deflecting sub-vision modules; another is a panoramic deflecting lens, in which the deflecting angle of the deflecting lens in front of each sub-vision module is the same, thereby achieving a wide range of observation; and the third is that the deflecting angle of the deflecting lens in front of each sub-vision module is different.

[0009] In one possible design, the first drive mechanism includes a main servo motor, a first drive gear, and a gear ring. The motor body of the main servo motor is fixed inside the annular housing. The first drive gear is fixed on the motor shaft of the main servo motor. The gear ring is fixed on the annular lens plate, and the first drive gear meshes with the gear ring.

[0010] In one possible design, the second drive mechanism includes an angle-switching servo, a second drive gear, and a clutch connector. The body of the angle-switching servo is fixed inside the annular housing. The second drive gear is fixed on the motor shaft of the angle-switching servo. The clutch connector includes a driven gear, a magnetic sleeve, and a coil sleeve. The coil sleeve is fixed to the sub-vision module. The driven gear is rotatably mounted on the coil sleeve and meshes with the second drive gear. The magnetic sleeve and the driven gear slide along the axial direction of the coil sleeve. An end ring is fixed on the magnetic sleeve, and a rubber gasket is fixed on the end ring. Several conical spikes are fixed on the side of the deflection lens facing the sub-vision module.

[0011] In one possible design, the magnetic sleeve has a groove and the driven gear has a ridge. The sliding fit between the magnetic sleeve and the driven gear is achieved through the cooperation of the groove and the ridge. The magnetic sleeve is made entirely of magnets.

[0012] In one possible design, the deflecting lens includes a lens barrel, a lens fixedly disposed within the lens barrel, a magnetic ring, and a limiting ring. The annular lens plate has a through hole for mounting the deflecting lens, and the lens barrel passes through the through hole to rotate with the annular lens plate. The limiting ring and the magnetic ring are located on opposite sides of the annular lens plate.

[0013] In one possible design, the annular lens plate is made of iron, the magnetic ring is made of magnet, a first friction surface is provided on the side of the magnetic ring near the wire rope annular lens plate, and a second friction surface is provided on the annular lens plate. When the deflecting lens is not pushed by the magnetic sleeve, the mutual attraction between the magnetic ring and the annular lens plate causes the magnetic ring to be tightly attached to the annular lens plate.

[0014] In one possible design, the annular shell includes an inner cylinder, a base plate, and an outer cylinder, wherein the inner cylinder and the outer cylinder are coaxially fitted, and the inner cylinder and the outer cylinder are connected and fixed through the base plate, and a mounting flange is fixed inside the inner cylinder.

[0015] On the other hand, a control method for an intelligent compound eye vision sensor is provided, the control method comprising the following control steps; S1. Power-on self-test, detect the main servo motor's return to zero, check whether there is any obstruction within the field of view of each sub-vision module, ensure that the sub-vision module is aligned with the deflection lens, and control all magnetic sleeves to disconnect from power.

[0016] S2. Switch the corresponding observation mode according to the actual use scenario, including three modes: normal, panoramic and fine. Control the main servo to switch the deflection lens in front of the sub-vision module according to different use modes, and then control the deflection lens to rotate and switch the observation angle through the angle switching servo. If it is fine mode, control each deflection lens to rotate towards the coordinates of the receiving focus point.

[0017] S3, Ring Lens Board Control: The main servo drives the first drive gear to mesh with the gear ring, causing the ring lens board to rotate as a whole, switching the deflection lens type.

[0018] S4. Adjusting the lens angle: The second drive mechanism controls the rotation of the deflecting lens to the corresponding angle state according to different observation modes.

[0019] S5, visual acquisition and feedback, parallel acquisition of images by each sub-vision module, transmission of acquired visual images to the internal system, and calculation of depth or defect results by GPU.

[0020] Compared with the prior art, the present invention has the following advantages and beneficial effects: This invention achieves visual recognition through a small-volume compound eye recognition structure. When installed on an industrial robot, it will not affect the robot's travel range. At the same time, the compound eye structure can greatly reduce the observation angle, allowing the entire sensor to observe comprehensively and clearly. This can have a more accurate and positive effect on the visual recognition control of the industrial robot, and can also provide comprehensive protection for the sub-vision module. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of the structure of the present invention; Figure 2 This is a schematic diagram of the structure of the present invention after removing the outer cylinder; Figure 3 This is a schematic diagram of the sub-vision module within the annular housing in an embodiment of the present invention; Figure 4 This is a schematic diagram of the sub-vision module and angle switching servo motor within the annular housing in an embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of the ring lens plate and the deflecting lens in an embodiment of the present invention; Figure 6 This is a schematic diagram of the interaction between the ring lens plate and the deflecting lens and the main servo motor in an embodiment of the present invention; Figure 7 This is a front view in an embodiment of the present invention; Figure 8 This is an embodiment of the present invention. Figure 7 Sectional view at point AA; Figure 9 for Figure 8 Enlarged view at point B in the middle; Figure 10 This is a schematic diagram of the mating cross-section of the driven gear, magnet sleeve, and coil sleeve in this embodiment; Figure 11 for Figure 9 Enlarged view at point C; Figure 12 This is a schematic diagram of the deflection lens and sub-vision module in an embodiment of the present invention.

[0022] The reference numerals in the attached drawings represent: 1-mounting flange, 2-annular housing, 201-inner cylinder, 202-base plate, 203-outer cylinder, 3-annular lens plate, 301-rib plate, 302-second friction surface, 4-deflecting lens, 401-lens cylinder, 402-lens, 403-magnetic ring, 4031-first friction surface, 404-limiting ring, 5-tooth ring, 6-sub-vision module, 7-angle switching servo, 8-main servo, 9-first drive gear, 10-second drive gear, 11-driven gear, 1101-protruding ridge, 12-magnetic sleeve, 1201-groove, 13-coil sleeve, 14-end ring, 15-rubber gasket, 16-cone spike. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0024] Example 1, as Figures 1 to 12 As shown, an intelligent compound-eye vision sensor is provided, including an annular housing 2, a plurality of sub-vision modules 6 evenly arranged within the annular housing 2, and a lens angle adjustment module rotatably mounted on the annular housing 2. The lens angle adjustment module includes an annular lens plate 3 and a plurality of deflecting lenses 4 rotatably mounted on the annular lens plate 3. The deflecting lenses 4 are used to deflect the observation direction of the sub-vision modules 6. By rotating the deflecting lenses 4, the observation angle of the sub-vision modules 6 can be adjusted without the sub-vision modules 6 remaining stationary. This effectively concentrates the sub-vision modules 6 within the annular housing 2, providing comprehensive protection for the sub-vision modules 6. Furthermore, the annular housing 2 houses a plurality of sub-vision modules 6, and the compound-eye structure of the multiple sub-vision modules 6, in conjunction with the lens angle adjustment module, can comprehensively improve the observation angle and avoid blind spots.

[0025] In this embodiment, the lens angle adjustment module further includes a first driving mechanism and a second driving mechanism. The second driving mechanism is fixed to the sub-vision module 6 and is used to drive the deflection lens 4 located at the lens end of the sub-vision module 6 to rotate, thereby adjusting the viewing angle direction of the sub-vision module 6. The first driving mechanism is fixed inside the annular housing 2 and is used to drive the entire annular lens plate 3 to rotate, thereby causing each deflection lens 4 on the entire annular lens plate 3 to rotate together, and is used to switch the deflection lens 4 at the lens end of the sub-vision module 6. By equipping the sub-vision module 6 with deflection lenses 4 of different deflection angles, the observation angle of the sub-vision module 6 can be further changed. The above layout facilitates clear visual recognition by industrial robots, effectively avoiding visual obstruction. Furthermore, when adjusting the viewing angle, compared to the traditional method of adjusting the viewing angle by rotating or moving the sub-vision module 6, this embodiment ensures that the sub-vision module 6 remains within the annular housing 2, providing comprehensive protection for the sub-vision module 6. This effectively prevents damage to the sub-vision module 6 caused by accidental bumps during industrial robot operation, thereby improving the overall safety of the vision sensor.

[0026] In this embodiment, the annular lens plate 3 is divided into multiple lens areas by the rib plate 301. The number of lens areas is the same as the number of sub-vision modules 6. Each lens area is provided with a deflecting lens 4 with multiple deflection angles. By rotating the annular lens plate 3, the deflecting lens 4 in front of the sub-vision module 6 is switched, thereby further changing the observation range of the entire compound eye vision sensor.

[0027] Advantageously, each lens area in this embodiment includes three types of deflecting lenses 4. One type is a deflecting lens 4 without deflection angle, and when one sub-vision module 6 is aligned with the deflecting lens 4 without deflection angle, the remaining sub-vision modules 6 are also aligned with the non-deflecting sub-vision modules 6. Another type is a panoramic deflecting lens 4, in which the deflection angle of the deflecting lens 4 in front of each sub-vision module 6 is the same, thereby achieving a wide range of observation. A third type is where the deflection angle of the deflecting lens 4 in front of each sub-vision module 6 is different, but by rotating the deflecting lens 4, the observation areas of multiple sub-vision modules 6 can be converged to a small area, thereby achieving fine observation of the area. The above different observation methods can enable industrial robots to be applied in different working scenarios and complete corresponding observation operations. For example, the first observation mode can be applied to low-precision operations, such as simple palletizing and handling. The second observation mode is used for working in areas with complex surrounding environments, and the wide field of view can effectively prevent industrial robots from colliding with the surrounding environment. The third observation mode can be applied to more defect detection or some more precise assembly operations to achieve precise operation.

[0028] In this embodiment, the first drive mechanism includes a main servo motor 8, a first drive gear 9, and a gear ring 5. The motor body of the main servo motor 8 is fixed inside the annular housing 2. The first drive gear 9 is fixed on the motor shaft of the main servo motor 8. The gear ring 5 is fixed on the annular lens plate 3. The first drive gear 9 meshes with the gear ring 5, thereby controlling the rotation of the annular lens plate 3 through the main servo motor 8.

[0029] In this embodiment, the second drive mechanism includes an angle-switching servo 7, a second drive gear 10, and a clutch connector. The body of the angle-switching servo 7 is fixed inside the annular housing 2. The second drive gear 10 is fixed on the motor shaft of the angle-switching servo 7. The clutch connector includes a driven gear 11, a magnet sleeve 12, and a coil sleeve 13. The coil sleeve 13 is fixed to the sub-vision module 6. The driven gear 11 is rotatably mounted on the coil sleeve 13 and meshes with the second drive gear 10. The magnetic sleeve 12 and the driven gear 11 slide in axial engagement along the coil sleeve 13. An end ring 14 is fixed on the magnetic sleeve 12, and a rubber pad 15 is fixed on the end ring 14. A plurality of cones 16 are fixed on the side of the deflection lens 4 facing the sub-vision module 6. The magnetic sleeve 12 drives the rubber pad 15 to move closer to the deflection lens 4. After the cones 16 are inserted into the rubber pad 15, the magnetic sleeve 12 can drive the deflection lens 4 to rotate as a whole, thereby adjusting the entire deflection angle.

[0030] Beneficial reference Figure 10 The magnetic sleeve 12 has a groove 1201 and the driven gear 11 has a protrusion 1101. The sliding engagement between the magnetic sleeve 12 and the driven gear 11 is achieved through the cooperation of the groove 1201 and the protrusion 1101. The magnetic sleeve 12 is made entirely of magnets. The extension and retraction of the magnetic sleeve 12 is controlled by changing the current flowing through the coil sleeve 13, thereby realizing the clutch operation.

[0031] In this embodiment, the deflecting lens 4 includes a lens barrel 401, a lens 402 fixedly disposed within the lens barrel 401, a magnetic ring 403, and a limiting ring 404. The annular lens plate 3 has a through hole for mounting the deflecting lens 4. The lens barrel 401 passes through the through hole and rotates with the annular lens plate 3. The limiting ring 404 and the magnetic ring 403 are located on both sides of the annular lens plate 3 to limit the position of the lens barrel 401. The lens 402 deflects the observation angle of the sub-vision module 6 by refracting light. By using lenses 402 with different refractive indices, the observation deflection angle of the sub-vision module 6 can be different.

[0032] Beneficial reference Figure 11The annular lens plate 3 is made of iron, and the magnetic ring 403 is made of magnet. A first friction surface 4031 is provided on the side of the magnetic ring 403 near the steel wire rope annular lens plate 3, and a second friction surface 302 is provided on the annular lens plate 3. When the deflecting lens 4 is not pushed by the magnetic sleeve 12, the mutual attraction between the magnetic ring 403 and the annular lens plate 3 makes the magnetic ring 403 stick tightly to the annular lens plate 3. The cooperation of the first friction surface 4031 and the second friction surface 302 can effectively prevent the free rotation of the deflecting lens 4 and ensure the stability of the viewing angle of the deflecting lens 4.

[0033] In this embodiment, both the main servo motor 8 and the angle switching servo motor 7 are small servo geared motors with encoders, which can record the rotation angles of the ring lens plate 3 and the deflection lens 4, so that the robot system can perform whole-machine feedback control on the observation angle of the entire compound eye vision sensor.

[0034] In this embodiment, the compound eye vision sensor is installed at the end effector flange of the industrial robot via mounting flange 1, thereby cooperating with the industrial robot to complete various automated processing operations.

[0035] In this embodiment, the annular housing 2 includes an inner cylinder 201, a base plate 202, and an outer cylinder 203. The inner cylinder 201 and the outer cylinder 203 are coaxially fitted and connected and fixed to each other through the base plate 202. The mounting flange 1 is fixed inside the inner cylinder 201 to facilitate cooperation with the end effector of the industrial robot. The outer cylinder 203 is used to provide safety protection for the entire compound eye vision sensor, especially to protect the internal sub-vision module 6, preventing both collisions and dust.

[0036] Example 2: Based on Example 1, this example further provides a control method for an intelligent compound eye vision sensor to complete the operation control of the compound eye vision sensor, including the following control steps.

[0037] S1. Power-on self-test, check the main servo motor 8 to return to zero, check whether there is any obstruction in the field of view of each sub-vision module 6, ensure that the sub-vision module 6 is aligned with the deflection lens 4, and control all magnetic sleeves 12 to disconnect from power.

[0038] S2. Switch the corresponding observation mode according to the actual usage scenario, including three modes: normal, panoramic and fine. Control the main servo motor 8 to switch the deflection lens 4 in front of the sub-vision module 6 according to different usage modes. Then control the deflection lens 4 to rotate and switch the observation angle through the angle switching servo motor 7. If it is fine mode, control each deflection lens 4 to rotate towards the coordinates of the receiving focus point.

[0039] S3, the ring lens plate control, the main servo motor 8 drives the first drive gear 9 to mesh with the gear ring 5, causing the ring lens plate 3 to rotate as a whole, switching the deflection lens 4 type.

[0040] S4. Adjusting the lens angle: The second drive mechanism controls the deflection lens 4 to rotate. According to different observation modes, the deflection lens 4 is controlled to rotate to the corresponding angle state. The coil sleeve 13 is energized to generate a magnetic field. The magnet sleeve 12 extends out, and the rubber pad 15 and the cone 16 engage to deflect the lens 4. The angle switching servo 7 drives the second drive gear 10 to drive the driven gear 11, so that the magnet sleeve 12 rotates the deflection lens 4 to the target angle. After the power is cut off, the magnet sleeve 12 retracts, and the magnetic ring 403 attracts the annular lens plate 3 to fix the deflection lens 4.

[0041] S5, visual acquisition and feedback, each sub-vision module 6 acquires images in parallel, transmits the acquired visual images to the internal system, and calculates depth or defect results through the GPU.

[0042] S6. The encoder records the angles of the ring lens plate 3 and the deflection lens 4 in real time and feeds them back to the closed-loop control of the robot system.

[0043] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An intelligent compound eye vision sensor, comprising an annular housing (2), a plurality of sub-vision modules (6) uniformly arranged within the annular housing (2), and a lens viewing angle adjustment module rotatably mounted on the annular housing (2), characterized in that, The lens angle adjustment module includes a ring lens plate (3) and multiple deflecting lenses (4) rotatably mounted on the ring lens plate (3). The deflecting lenses (4) are used to deflect the observation direction of the sub-vision module (6). The lens angle adjustment module includes a first driving mechanism and a second driving mechanism. The first driving mechanism is fixed inside the annular housing (2) and is used to drive the entire annular lens plate (3) to rotate. The second driving mechanism is fixed to the sub-vision module (6) and is used to drive the deflection lens (4) located at the lens end of the sub-vision module (6) to rotate.

2. The intelligent compound eye vision sensor according to claim 1, characterized in that, The annular lens plate (3) is divided into multiple lens areas by ribs (301). The number of lens areas is the same as the number of sub-vision modules (6). Each lens area is provided with deflecting lenses (4) with multiple deflection angles.

3. The intelligent compound eye vision sensor according to claim 2, characterized in that, The lens area includes three types of deflection lenses (4). One type is a deflection lens (4) without deflection angle, and when a sub-vision module (6) is aligned with the deflection lens (4) without deflection angle, the other sub-vision modules (6) are also aligned with the non-deflection sub-vision modules (6). Another type is a panoramic deflection lens (4), in which the deflection angle of the deflection lens (4) in front of each sub-vision module (6) is the same, thereby achieving a wide range of observation. The third type is where the deflection angle of the deflection lens (4) in front of each sub-vision module (6) is different.

4. The intelligent compound eye vision sensor according to claim 1, characterized in that, The first drive mechanism includes a main servo motor (8), a first drive gear (9), and a gear ring (5). The motor body of the main servo motor (8) is fixed inside the annular housing (2). The first drive gear (9) is fixed on the motor shaft of the main servo motor (8). The gear ring (5) is fixed on the annular lens plate (3). The first drive gear (9) meshes with the gear ring (5).

5. The intelligent compound eye vision sensor according to claim 1, characterized in that, The second drive mechanism includes an angle switching servo (7), a second drive gear (10), and a clutch connector. The body of the angle switching servo (7) is fixed inside the annular housing (2). The second drive gear (10) is fixed on the motor shaft of the angle switching servo (7). The clutch connector includes a driven gear (11), a magnetic sleeve (12), and a coil sleeve (13). The coil sleeve (13) is fixed to the sub-vision module (6). The driven gear (11) is rotatably mounted on the coil sleeve (13). The driven gear (11) meshes with the second drive gear (10). The magnetic sleeve (12) and the driven gear (11) slide along the axial direction of the coil sleeve (13). An end ring (14) is fixed on the magnetic sleeve (12). A rubber pad (15) is fixed on the end ring (14). Several cones (16) are fixed on the side of the deflection lens (4) facing the sub-vision module (6).

6. The intelligent compound eye vision sensor according to claim 5, characterized in that, The magnet sleeve (12) has a groove (1201) and the driven gear (11) has a protrusion (1101). The sliding fit between the magnet sleeve (12) and the driven gear (11) is achieved through the cooperation of the groove (1201) and the protrusion (1101). The magnet sleeve (12) is made entirely of magnets.

7. The intelligent compound eye vision sensor according to claim 1, characterized in that, The deflecting lens (4) includes a lens barrel (401), a lens (402) fixedly disposed in the lens barrel (401), a magnetic ring (403) and a limiting ring (404). The annular lens plate (3) has a through hole for mounting the deflecting lens (4). The lens barrel (401) passes through the through hole and rotates with the annular lens plate (3). The limiting ring (404) and the magnetic ring (403) are located on both sides of the annular lens plate (3).

8. The intelligent compound eye vision sensor according to claim 7, characterized in that, The annular lens plate (3) is made of iron material, and the magnetic ring (403) is made of magnet. A first friction surface (4031) is provided on the side of the magnetic ring (403) close to the wire rope annular lens plate (3), and a second friction surface (302) is provided on the annular lens plate (3). When the deflecting lens (4) is not pushed by the magnetic sleeve (12), the mutual attraction between the magnetic ring (403) and the annular lens plate (3) makes the magnetic ring (403) stick tightly to the annular lens plate (3).

9. The intelligent compound eye vision sensor according to claim 1, characterized in that, The annular shell (2) includes an inner cylinder (201), a bottom plate (202) and an outer cylinder (203). The inner cylinder (201) and the outer cylinder (203) are coaxially fitted. The inner cylinder (201) and the outer cylinder (203) are connected and fixed through the bottom plate (202). An installation flange (1) is fixed inside the inner cylinder (201).

10. A control method for an intelligent compound eye vision sensor according to any one of claims 1-9, characterized in that, Includes the following control steps; S1. Power-on self-test, check the main servo motor (8) to return to zero, check whether there is any obstruction in the field of view of each sub-vision module (6), ensure that the sub-vision module (6) is aligned with the deflection lens (4), and control all magnetic sleeves (12) to disconnect from power. S2. Switch the corresponding observation mode according to the actual usage scenario, including three modes: normal, panoramic and fine. Control the main servo (8) to switch the deflection lens (4) in front of the sub-vision module (6) according to different usage modes. Then control the deflection lens (4) to rotate and switch the observation angle through the angle switching servo (7). If it is fine mode, control each deflection lens (4) to rotate towards the receiving focus point coordinates. S3, Ring Lens Board Control: The main servo motor (8) drives the first drive gear (9) to mesh with the gear ring (5), causing the ring lens board (3) to rotate as a whole, switching the deflection lens (4) type; S4. Adjust the lens angle by controlling the rotation of the deflecting lens (4) through the second drive mechanism. Control the deflecting lens (4) to rotate to the corresponding angle state according to different observation modes. S5, Visual acquisition and feedback, Parallel acquisition of images by each sub-visual module (6), Transmission of acquired visual images to the internal system, and calculation of depth or defect results by GPU; S6. The encoder records the angles of the ring lens plate (3) and the deflection lens (4) in real time and feeds them back to the closed-loop control of the robot system.

Images