Image shell defect identification equipment based on visual inspection

Abstract

The invention discloses image shell defect identification equipment based on visual inspection. The equipment comprises a control system, a feeding mechanism, a transparent rotary detection platform, an upstream conveying mechanism, a first camera visual inspection assembly, a first discharging sorting mechanism, a second camera visual inspection assembly and a second discharging sorting mechanism; the first camera vision detection assembly is arranged along the upstream conveying mechanism and comprises an inner wall detection group and an upper surface detection group; the first blowing mechanism blows defective products detected by the first camera vision detection assembly into the first material channel; the second camera vision detection assembly is arranged along the transparent rotary detection platform and comprises an outer wall detection group, a lower surface detection group and an inner opening detection group; the second discharging sorting mechanism is arranged along the transparent rotary detection platform and comprises three blowing mechanisms and corresponding material channels, and the blowing mechanisms sort the image shells detected by the second camera vision detection assembly to the corresponding material channel outer wall detection sets.

Description

Technical Field

[0001] This invention relates to the technical field of inspection equipment, and in particular to an image-based shell defect identification device based on visual inspection. Background Technology

[0002] In the contemporary automotive manufacturing field, with the rapid development of advanced driver assistance systems and autonomous driving technologies, in-vehicle imaging systems have become a core component of modern automobiles. As the physical protection and functional carrier of these precision sensors, the quality and accuracy of the imaging housing directly affect the performance, reliability, and lifespan of the imaging system. Die-cast aluminum automotive imaging housings are key components widely used to meet these high standards.

[0003] Aluminum die casting is the preferred method for manufacturing such housings due to its ability to produce complex geometries, high dimensional accuracy, excellent mechanical properties, and suitability for mass production. Although the aluminum die casting process is mature, various factors during production, such as mold wear, fluctuations in process parameters, material purity, or improper post-processing, can easily lead to various defects in the housing, including surface defects such as porosity, slag inclusions, shrinkage cavities, cold shuts, flow marks, scratches, sticking to the mold, burrs, and flash. More serious defects such as defects like missing parts, inclusions, cracks, and deformation can also occur. If these defects are not effectively detected and flow into subsequent assembly lines and vehicles, they can cause minor issues like malfunctions in individual imaging modules, requiring costly rework or replacement; or even lead to misjudgments by ADAS systems, posing a serious threat to driving safety. Therefore, conducting 100% comprehensive, high-precision defect inspection of aluminum die-cast imaging housings before shipment is an absolutely essential step in ensuring the reliability of automotive active safety systems.

[0004] Machine vision inspection technology, as a non-contact, high-speed, and high-precision automated inspection method, has become a pillar technology of industrial quality control. Its basic principle is to simulate human visual function by using an optical imaging system to convert the object under test into a digital image signal, which is then transmitted to a dedicated image processing system. Based on information such as pixel distribution, brightness, and color, specific algorithms are used to perform calculations to obtain information such as the target's shape, size, position, and surface condition. Finally, the inspection results and data reports are output according to preset judgment criteria.

[0005] While machine vision inspection technology is widely used in industry, its application to precision components such as aluminum die-cast automotive imaging housings—components with complex three-dimensional geometry, highly reflective surfaces, and diverse defect types—faces significant challenges. A significant contradiction exists between the high-quality inspection requirements of aluminum die-cast automotive imaging housings and the current inspection methods, which rely on multiple devices, are inefficient, and still require manual intervention. The market urgently needs an innovative solution that is more integrated, flexible, and capable of rapid and accurate identification of multiple types of defects in the housing from multiple angles using a single device in a single operation. Summary of the Invention

[0006] In order to solve the problems in the prior art, the present invention provides an image shell defect recognition device based on visual inspection.

[0007] The technical solution adopted by the present invention to solve the above-mentioned technical problems is: a visual inspection-based image shell defect recognition device, comprising: The control system is used to coordinate the actions of various mechanisms and the visual inspection process. The feeding mechanism is used for automatic or manual feeding. A transparent rotating inspection platform is used to support and rotate the image housing to the corresponding position for inspection. The upstream conveying mechanism, including the conveying track and guide assembly, is used to convey the image housing to be inspected to the transparent rotating inspection platform; The first camera vision inspection assembly is arranged along the upstream conveying mechanism and includes an inner wall inspection group and an upper surface inspection group. The first feeding and sorting mechanism is arranged along the upstream conveying mechanism and includes a first blowing mechanism and a first material channel. The first blowing mechanism blows the defective products detected by the first camera vision inspection component into the first material channel. The second camera vision inspection assembly is arranged along the transparent rotating inspection platform and includes an outer wall inspection group, a lower surface inspection group, and an inner opening inspection group. The second material sorting mechanism is set along the transparent rotating detection platform and includes a second air blowing mechanism, a third air blowing mechanism, a fourth air blowing mechanism, and corresponding second, third, and fourth material channels. Each air blowing mechanism sorts the image housing detected by the second camera vision detection component to the corresponding material channel outer wall detection group.

[0008] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: both the inner wall detection group and the outer wall detection group include a first column and a 360° visual imaging group that moves up and down along the first column; the 360° visual imaging group includes a first base, a first ring light source disposed above the first base, four first cameras evenly distributed in a ring below the first base, and four reflectors disposed corresponding to the first cameras.

[0009] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: the first base is provided with a downwardly extending vertical mounting plate, and the side of the vertical mounting plate is provided with a support plate at a 90-degree angle to the vertical mounting plate. The support plate can move back and forth with a limited range relative to the vertical mounting plate, and the reflector is inclinedly arranged on the support plate with the reflective surface facing upward.

[0010] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: the bracket plate is provided with an arc-shaped groove, the reflector is installed laterally in the arc-shaped groove and can move along the arc-shaped groove to adjust the tilt angle.

[0011] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: the upper surface detection group includes an upper side detection mechanism and an upper front detection mechanism; the second camera of the upper side detection mechanism is tilted downward to capture an image of the upper side of the image housing; the third camera of the upper front detection mechanism is vertically downward to capture an image of the upper front of the image housing.

[0012] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: the upper-level lateral detection mechanism includes a second column, a second base on the second column, a second ring light source on the second base, and three second cameras. The second base has three horizontal support arms evenly distributed in a ring. The second cameras are mounted on the horizontal support arms in an inward tilting and angle-adjustable manner.

[0013] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: the inner wall detection group is located upstream of the upper lateral detection mechanism; an upper forward detection mechanism is provided between the inner wall detection group and the upper lateral detection mechanism and downstream of the upper lateral detection mechanism; The upper-level forward detection mechanism includes a third column, a third base mounted on the third column, a third camera vertically mounted under the third base, and a third ring light source mounted under the third camera.

[0014] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: the lower surface detection group includes a fourth column, a fourth base set on the fourth column, a fourth camera and a fourth ring light source set vertically upward on the fourth base, and a flat reflective surface set on the fourth column. The flat reflective surface is located above the transparent rotating detection platform, and the fourth camera and the fourth ring light source are located below the transparent rotating detection platform.

[0015] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: the inner opening detection group includes a fifth column, a fifth base set on the fifth column, a fifth camera set vertically upward on the fifth base, and a fifth ring light source set on the fifth column. The fifth ring light source is located above the transparent rotating detection platform, and the fifth camera is located below the transparent rotating detection platform.

[0016] The preferred technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: the feeding mechanism includes a feeding plate that is inclined downward toward the conveying track, and the guiding component includes a plurality of baffles distributed at different positions along the conveying track.

[0017] Compared with existing technologies, the advantages of this invention are: the control system directs the feeding mechanism to supply materials according to a predetermined program. After the shell enters the conveying track, it is continuously photographed and inspected by the first vision component during movement. If a defect is found, it is immediately rejected by the first air blowing mechanism. The image shells that meet the first-level requirements are sent to a transparent rotating platform. The platform rotates so that the shells sequentially reach the optimal inspection positions of different inspection groups for high-precision photography. The second vision component feeds back the results to the control system, which controls the corresponding air blowing valves in the second sorting mechanism to accurately blow the shells into the corresponding material channels, completing the final sorting. In this vision-based image shell defect recognition device, all mechanisms are coordinated by the control system, realizing full automation of the entire process from feeding, conveying, positioning, inspection to sorting, reducing manual intervention, and ensuring uniform inspection standards and stable results. Attached Figure Description

[0018] The present invention will be further described in detail below with reference to the accompanying drawings and preferred embodiments. However, those skilled in the art will understand that these drawings are drawn only for the purpose of explaining the preferred embodiments and therefore should not be construed as limiting the scope of the invention. Furthermore, unless specifically indicated, the drawings are only schematic representations of the composition or structure of the described objects and may contain exaggerated depictions, and the drawings are not necessarily drawn to scale.

[0019] Figure 1 A three-dimensional image shell defect recognition device based on vision inspection Figure 1 ; Figure 2 A three-dimensional image shell defect recognition device based on vision inspection Figure 2 ; Figure 3 This is a schematic diagram of the inner wall detection group of an image shell defect recognition device based on vision inspection; Figure 4 This is a schematic diagram of the upper-level forward inspection mechanism of a vision-based image shell defect recognition device; Figure 5This is a schematic diagram of the upper-level lateral inspection mechanism of an image shell defect recognition device based on vision inspection; Figure 6 This is a schematic diagram of the outer wall detection group of an image shell defect recognition device based on vision inspection; Figure 7 This is a schematic diagram of the lower surface detection group of an image shell defect recognition device based on vision inspection; Figure 8 This is a schematic diagram of the inner opening detection group of an image shell defect recognition device based on vision inspection.

[0020] Figure label: 1. Feeding mechanism; 3. Transparent rotating inspection platform; 4. Upstream conveying mechanism; 5. First camera vision inspection assembly; 6. First unloading and sorting mechanism; 7. Second camera vision inspection assembly; 8. Second unloading and sorting mechanism; 41. Conveying track; 42. Guide assembly; 51. Inner wall inspection group; 52. Upper surface inspection group; 62. First material channel; 71. Outer wall inspection group; 72. Lower surface inspection group; 73. Inner opening inspection group; 81. Second air blowing mechanism; 82. Third air blowing mechanism; 83. Fourth air blowing mechanism; 91. First column; 92. First base; 93. First ring light source; 94. First camera; 95. Reflector; Vertical Mounting plate 96; bracket plate 97; arc groove 98; upper side detection mechanism 10; upper forward detection mechanism 20; second column 12; second base 13; second ring light source 14; second camera 11; horizontal support arm 15; third column 22; third base 23; third camera 21; third ring light source 24; fourth column 25; fourth base 26; fourth camera 27; flat reflective surface 28; fourth ring light source 29; fifth column 16; fifth base 17; fifth camera 18; fifth ring light source 19; feeding plate 100; push plate 400; liner plate 99. Detailed Implementation

[0021] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings. Those skilled in the art will appreciate that these descriptions are merely descriptive and exemplary and should not be construed as limiting the scope of the invention.

[0022] It should be noted that similar labels in the following figures indicate similar items; therefore, once an item is defined in one figure, it will not be further defined and explained in subsequent figures.

[0023] In the description of this invention, it should be noted that the terms "upper," "lower," "front," "rear," "left," "right," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the die-cast parts of this invention are used. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0024] like Figure 1-2 As shown, this embodiment provides a vision-based image shell defect recognition device, including a control system, a feeding mechanism 1, a transparent rotating detection platform 3, an upstream conveying mechanism 4, a first camera vision detection component 5, a first unloading and sorting mechanism 6, a second camera vision detection component 7, and a second unloading and sorting mechanism 8.

[0025] The control system coordinates the actions of each mechanism and the visual inspection process. The feeding mechanism 1 is used for automatic or manual feeding.

[0026] The transparent rotating inspection platform 3 is located downstream of the upstream conveying mechanism 4 and upstream of the second unloading and sorting mechanism 8. It is used to carry and rotate the image housing to the corresponding position for inspection.

[0027] The upstream conveying mechanism 4 is located downstream of the feeding mechanism 1 and includes a conveying track 41 and a guide assembly 42, used to convey the image housing to be inspected to the transparent rotating inspection platform 3.

[0028] The first camera vision inspection assembly 5 is arranged along the upstream conveying mechanism 4 and includes an inner wall inspection group 51 and an upper surface inspection group 52.

[0029] The first material sorting mechanism 6 is located downstream of the first camera vision inspection component 5 and along the upstream conveying mechanism 4. It includes a first air blowing mechanism and a first material channel 62. The first air blowing mechanism blows the defective products detected by the first camera vision inspection component 5 into the first material channel 62.

[0030] The second camera vision inspection assembly 7 is arranged along the transparent rotating inspection platform 3 and includes an outer wall inspection group 71, a lower surface inspection group 72 and an inner opening inspection group 73.

[0031] The second material sorting mechanism 8 is set along the transparent rotating detection platform 3 and includes a second air blowing mechanism 81, a third air blowing mechanism 82, a fourth air blowing mechanism 83, and corresponding second, third, and fourth material channels. Each air blowing mechanism sorts the image shell detected by the second camera vision detection component 7 to the corresponding material channel outer wall detection group 71.

[0032] In this vision-based image-based shell defect identification device, all mechanisms are coordinated by the control system, achieving full automation from feeding, conveying, positioning, detection to sorting, reducing manual intervention, and ensuring consistent detection standards and stable results. The control system directs the feeding mechanism 1 to supply materials according to a predetermined program. After the shells enter the conveyor track 41, they are continuously photographed and detected by the first vision component during movement. If a defect is detected, it is immediately rejected by the first air blowing mechanism. Image-based shells that meet the first-level requirements are sent to a transparent rotating platform. The platform rotates, causing the shells to sequentially reach the optimal detection positions of different detection groups for high-precision imaging. The second vision component feeds back the results to the control system, which controls the corresponding air blowing valves in the second sorting mechanism to precisely blow the shells into the corresponding material channels, completing the final sorting.

[0033] like Figure 3 , 6 As shown, both the inner wall detection group 51 and the outer wall detection group 71 include a first column 91 and a 360° visual imaging group that moves up and down along the first column 91. The 360° visual imaging group includes a first base 92, a first ring light source 93 disposed above the first base 92, four first cameras 94 evenly distributed in a ring below the first base 92, and four reflectors 95 disposed corresponding to the first cameras 94. The outer wall detection group 71 also includes a liner 99 located below the transparent rotating detection platform 3.

[0034] By using a circular arrangement of four cameras and reflectors 95, 360° circumferential images of the inner or outer wall of a cylinder can be acquired simultaneously at a single workstation with a single trigger. This significantly improves detection efficiency compared to solutions involving single-camera rotation scanning or multiple rotations of the housing. Simultaneous acquisition by multiple cameras avoids image stitching misalignment issues caused by equipment vibration, lighting changes, or slight housing movements that can result from time-sharing shooting, ensuring the integrity and accuracy of the panoramic image. In this solution, the overall imaging group can be raised and lowered along the first column 91, adapting to imaging housings of different sizes and offering strong versatility.

[0035] like Figure 3 , 6As shown, the first base has a downwardly extending vertical mounting plate 96. A support plate 97, at a 90-degree angle to the vertical mounting plate 96, is located on the side of the vertical mounting plate 96. The support plate 97 can move back and forth relative to the vertical mounting plate 96 with a limited range. A reflector 95 is tilted on the support plate 97 with its reflective surface facing upwards. Moving the support plate 97 and the reflector 95 back and forth essentially adjusts the optical path length and reflection angle of the camera's optical system. This allows the imaging assembly to adapt to image housings of different sizes, ensuring that housings of different sizes can form appropriately sized and clear images on the camera sensor. This is a key mechanical adjustment method for achieving flexible equipment inspection, ensuring clear images and accurate field-of-view coverage. This adjustment mechanism provides mechanical fine-tuning capabilities, reducing the stringent requirements for absolute precision during equipment assembly and facilitating later replacement of reflectors 95 of different specifications or recalibration after repairs.

[0036] Preferably, the support plate 97 is provided with an arc-shaped groove 98, and the reflector 95 is laterally installed in the arc-shaped groove 98 and can move along the arc-shaped groove 98 to adjust the tilt angle. Fine adjustment of the angle of the reflector 95 can precisely control the reflected light path, ensuring that the image of a specific area of ​​the inner or outer wall can be accurately projected into the center of the corresponding camera, optimizing the image edge quality and reducing distortion.

[0037] like Figure 1-2 As shown, the upper surface detection group 52 includes an upper lateral detection mechanism 10 and an upper forward detection mechanism 20. The second camera 11 of the upper lateral detection mechanism 10 is tilted downward to capture an image of the upper side of the image housing; the third camera 21 of the upper forward detection mechanism 20 is vertically downward to capture an image of the upper side of the image housing. The tilted camera can detect non-horizontal features such as the side edges, chamfers, and vertical surfaces of the upper surface; the vertically positioned camera is responsible for detecting the main horizontal planes. The combination of the two enables comprehensive detection of the three-dimensional morphology of the upper surface, and can detect various defects such as planar scratches, edge chips, and sidewall stains.

[0038] like Figure 5 As shown, the upper-level lateral detection mechanism 10 includes a second column 12, a second base 13 mounted on the second column 12, a second ring light source 14 mounted on the second base 13, and three second cameras 11. The second base 13 has three horizontal support arms 15 evenly distributed in a ring. The second cameras 11 are mounted on the horizontal support arms 15 in an inwardly tilted and angle-adjustable manner. This structure creates a stable "triangulation" observation point. The three cameras acquire images of the same upper surface side edge region from different directions. The software can compare and comprehensively analyze these images to fully reflect the non-horizontal features of the upper surface side edge, chamfers, and vertical surfaces of the image shell.

[0039] like Figure 4As shown, the upper forward detection mechanism 20 includes a third column 22, a third base 23 disposed on the third column 22, a third camera 21 vertically disposed under the third base 23, and a third ring light source 24 disposed under the third camera 21.

[0040] like Figure 1-2 As shown, the inner wall detection group 51 is located upstream of the upper lateral detection mechanism 10; an upper forward detection mechanism 20 is provided between the inner wall detection group 51 and the upper lateral detection mechanism 10, as well as downstream of the upper lateral detection mechanism 10. The lateral detection mechanism in the middle has forward detection mechanisms before and after it, which can perform two detections on the upper surface planar area of ​​the same shell, playing a role in repeated verification and improving reliability.

[0041] like Figure 7 As shown, the lower surface detection group 72 includes a fourth column 25, a fourth base 26 mounted on the fourth column 25, a fourth camera 27 vertically mounted on the fourth base 26, a fourth ring light source 29, and a flat reflective surface 28 mounted on the fourth column 25. The flat reflective surface 28 is located above the transparent rotating detection platform 3, while the fourth camera 27 and the fourth ring light source 29 are located below the transparent rotating detection platform 3. This scheme utilizes the principle of backlighting. The flat reflective surface 28 serves as a background, projecting the outline of the lower surface of the housing in the form of a silhouette. The camera receives light passing through the platform and the edge of the housing from below. Any defects affecting the integrity of the outline, such as missing corners or burrs, or foreign objects attached between the transparent platform and the housing, will produce abnormal shadows or light spots in the silhouette image, thus being captured with high sensitivity.

[0042] like Figure 8 As shown, the internal opening detection group 73 includes a fifth column 16, a fifth base 17 mounted on the fifth column 16, a fifth camera 18 vertically mounted on the fifth base 17, and a fifth ring light source 19 mounted on the fifth column 16. The fifth ring light source 19 is located above the transparent rotating detection platform 3, and the fifth camera 18 is located below the transparent rotating detection platform 3. The fifth ring light source 19 is positioned directly above the opening of the image housing, and its light enters the internal opening at a certain angle, effectively outlining the microscopic morphology of the internal opening's sidewalls. The camera captures images from directly below through the transparent platform, obtaining uniformly illuminated images of the bottom and sidewalls of the internal opening. This combination of vertical imaging and top ring illumination is particularly suitable for detecting internal hole defects with rotational symmetry characteristics, producing uniform imaging without eccentric shadows.

[0043] like Figure 1-2 As shown, the feeding mechanism 1 includes a feeding plate 100 that is inclined downward toward the conveying track 41. It relies on gravity to achieve the initial conveying and buffering of materials. The structure is simple, reliable, and low in cost.

[0044] The guide assembly 42 includes multiple baffles 400 distributed at different positions along the conveyor track 41, which can guide and correct the housing in stages and gradually.

[0045] This article uses specific examples to describe the image shell defect recognition device based on visual inspection provided by the present invention. The above description of the embodiments is only for the purpose of helping to understand the present invention and its core ideas. It should be noted that for those skilled in the art, several improvements and modifications can be made to the present invention without departing from the principle of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A visual inspection-based image shell defect recognition device, characterized in that... include: The control system is used to coordinate the actions of various mechanisms and the visual inspection process. The feeding mechanism is used for automatic or manual feeding. A transparent rotating inspection platform is used to support and rotate the image housing to the corresponding position for inspection. The upstream conveying mechanism, including the conveying track and guide assembly, is used to convey the image housing to be inspected to the transparent rotating inspection platform; The first camera vision inspection assembly is arranged along the upstream conveying mechanism and includes an inner wall inspection group and an upper surface inspection group. The first feeding and sorting mechanism is arranged along the upstream conveying mechanism and includes a first blowing mechanism and a first material channel. The first blowing mechanism blows the defective products detected by the first camera vision inspection component into the first material channel. The second camera vision inspection assembly is arranged along the transparent rotating inspection platform and includes an outer wall inspection group, a lower surface inspection group, and an inner opening inspection group. The second material sorting mechanism is set along the transparent rotating detection platform and includes a second air blowing mechanism, a third air blowing mechanism, a fourth air blowing mechanism, and corresponding second, third, and fourth material channels. Each air blowing mechanism sorts the image housing detected by the second camera vision detection component to the corresponding material channel outer wall detection group.

2. The image shell defect recognition device based on visual inspection according to claim 1, characterized in that: Both the inner wall detection group and the outer wall detection group include a first column and a 360° visual imaging group that moves up and down along the first column; the 360° visual imaging group includes a first base, a first ring light source disposed above the first base, four first cameras evenly distributed in a ring below the first base, and four reflectors disposed corresponding to the first cameras.

3. The image shell defect recognition device based on visual inspection according to claim 2, characterized in that: The first base is provided with a downwardly extending vertical mounting plate. The side of the vertical mounting plate is provided with a support plate at a 90-degree angle to the vertical mounting plate. The support plate can move back and forth relative to the vertical mounting plate with a limited range. The reflector is inclinedly mounted on the support plate with the reflective surface facing upward.

4. The image shell defect recognition device based on visual inspection according to claim 3, characterized in that: The support plate is provided with an arc-shaped groove, and the reflector is installed laterally in the arc-shaped groove and can move along the arc-shaped groove to adjust the tilt angle.

5. The image shell defect recognition device based on visual inspection according to claim 1, characterized in that: The upper surface detection group includes an upper lateral detection mechanism and an upper front detection mechanism; the second camera of the upper lateral detection mechanism is tilted downward to capture an image of the upper side of the image housing; the third camera of the upper front detection mechanism is vertically downward to capture an image of the top of the image housing.

6. The image shell defect recognition device based on visual inspection according to claim 5, characterized in that: The upper-level lateral detection mechanism includes a second column, a second base on the second column, a second ring light source on the second base, and three second cameras. The second base has three horizontal support arms evenly distributed in a ring, and the second cameras are mounted on the horizontal support arms in an inward tilt and adjustable manner.

7. The image shell defect recognition device based on visual inspection according to claim 5, characterized in that: The inner wall detection group is located upstream of the upper lateral detection mechanism; an upper forward detection mechanism is provided between the inner wall detection group and the upper lateral detection mechanism, as well as downstream of the upper lateral detection mechanism. The upper-level forward detection mechanism includes a third column, a third base mounted on the third column, a third camera vertically mounted under the third base, and a third ring light source mounted under the third camera.

8. The image shell defect recognition device based on visual inspection according to claim 1, characterized in that: The lower surface detection group includes a fourth column, a fourth base mounted on the fourth column, a fourth camera and a fourth ring light source mounted vertically upward on the fourth base, and a flat reflective surface mounted on the fourth column. The flat reflective surface is located above the transparent rotating detection platform, and the fourth camera and the fourth ring light source are located below the transparent rotating detection platform.

9. The image shell defect recognition device based on visual inspection according to claim 1, characterized in that: The inner opening detection group includes a fifth column, a fifth base mounted on the fifth column, a fifth camera mounted vertically upward on the fifth base, and a fifth ring light source mounted on the fifth column. The fifth ring light source is located above the transparent rotating detection platform, and the fifth camera is located below the transparent rotating detection platform.

10. The image shell defect recognition device based on visual inspection according to claim 1, characterized in that: The feeding mechanism includes a feeding plate that tilts downward toward the conveying track, and the guiding assembly includes multiple baffles distributed at different positions along the conveying track.

Images