Quality control equipment based on AI machine vision
Patent Information
- Application Number
- CN202611175005.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-04
- Publication Date
- 2026-10-30
AI Technical Summary
[0004]针对现有技术存在的问题，本发明提供了一种基于AI机器视觉的品控设备，具备集成AI机器视觉、智能翻面与差异化分拣的一体化品控的优点，解决了现有技术的问题
1、通过在环面转盘上设置并排的第一承载区和第二承载区，分别对应需上下表面检测的第一产品和仅需上表面检测的第二产品，同时搭配AI机器视觉模块的双摄像头设计，实现两类配套产品的同步流转、分别检测，无需额外增设多台设备，有效解决了传统品控设备只能单一检测一种产品、无法适配配套产品差异化检测需求的问题，大幅提升了设备的兼容性和通用性，适配多品种、小批量的柔性生产场景。
Smart Images
Figure CN122999541A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of identification and conveying technology, and in particular relates to a quality control device based on AI machine vision. Background Technology
[0002] In modern automated production processes, product surface quality inspection is a crucial step in ensuring the pass rate of finished products. Traditional manual inspection suffers from problems such as low efficiency, strong subjectivity, and susceptibility to fatigue and missed inspections. Conventional machine vision inspection equipment is mostly a single-station, single-pass inspection mode, which cannot simultaneously meet the differentiated inspection needs of supporting products (such as the first product requiring full inspection of both top and bottom surfaces, while the second product only requires sampling inspection of the top surface). Furthermore, it lacks efficient integrated solutions for flipping inspection and intelligent sorting, resulting in lengthy inspection processes, complex equipment structures, poor compatibility, and difficulty in adapting to flexible production scenarios with multiple varieties and small batches.
[0003] Meanwhile, existing flipping mechanisms mostly use clamping or suction cup flipping, which can easily cause scratches or deformation to the product surface. Furthermore, the high coupling between sorting and inspection actions makes it impossible to quickly remove defective products and ensure continuous flow of qualified products, thus hindering the simultaneous improvement of production efficiency and inspection accuracy. Therefore, there is an urgent need for an integrated quality control device that combines AI machine vision, intelligent flipping, and differentiated sorting to address these technical pain points. Summary of the Invention
[0004] To address the problems existing in the prior art, this invention provides a quality control device based on AI machine vision, which has the advantages of integrated quality control that combines AI machine vision, intelligent flipping, and differentiated sorting, thus solving the problems of the prior art.
[0005] This invention is implemented as follows: a quality control device based on AI machine vision includes a turntable mechanism. The turntable mechanism has several parallel first and second bearing areas on its annular turntable, and the annular turntable is provided with sorting holes. A bearing frame is rotatably connected to an upper frame, which has a through hole in its center and clamping plates fixedly connected to both sides. A turntable is provided under the bearing frame, and a first electric push rod is fixedly connected to the lower side of the turntable. The first electric push rod passes through the through hole, and a spring is fixedly connected to its lower end. The two sides of the spring are fixedly connected to the clamping plates. The first bearing area is located between the two clamping plates. A wheel mechanism is provided on the side of the clamping plates that are close to each other. An AI machine vision module includes a first camera and a second camera, which can be aligned with the first and second bearing areas respectively, and are used to detect the first and second products respectively.
[0006] In a preferred embodiment of the present invention, the turntable mechanism includes a first motor, the output shaft of the first motor is fixedly connected to a bearing disk, the edge of the bearing disk is fixedly connected to a connecting rod, the end of the connecting rod is fixedly connected to an annular turntable, and the first bearing area, the second bearing area and the sorting hole are all disposed on the annular turntable.
[0007] As a preferred embodiment of the present invention, the wheel mechanism includes a second motor and a wheel, the wheel being fixedly connected to the output shaft of the second motor; the second motor is fixedly connected to the side of the clamping plate that is far away from each other, and the wheel is located on the side of the clamping plate that is close to each other.
[0008] As a preferred embodiment of the present invention, the first camera is disposed on the lower surface of the spring sheet, and the second camera is disposed on the side of the clamping plate.
[0009] As a preferred embodiment of the present invention, the support frame includes an upright frame, a horizontal plate is fixedly connected to the upright frame, a second electric push rod is provided on the horizontal plate, and a lifting frame is fixedly connected to the second electric push rod.
[0010] In a preferred embodiment of the present invention, the lifting frame includes a straight plate and an annular plate, the annular plate has an annular track on its lower side, the upper frame includes a vertical block, the vertical block has a slot, and the slot is slidably connected to the annular track.
[0011] As a preferred embodiment of the present invention, the spring is an arc-shaped elastic metal sheet, the middle part of the spring is fixedly connected to the lower end of the first electric push rod, and the two ends of the spring are respectively fixedly connected to the inner sidewalls of the two side clamps.
[0012] As a preferred embodiment of the present invention, the AI machine vision module includes: Image acquisition unit: consists of a first camera, a second camera and a corresponding ring light. The first camera is used to acquire images of the upper and lower surfaces of the first product, and the second camera is used to acquire images of the upper surface of the second product. Data preprocessing unit: electrically connected to the image acquisition unit, used to perform noise reduction, enhancement, and normalization processing on the acquired images, and output standardized image data; AI Inference Unit: Loaded with a pre-trained surface defect detection model, used to receive standardized image data, identify defects such as scratches, stains, and deformations on the product surface, and output information on the defect type, location, and severity. Result Judgment Unit: Based on the comparison between defect information and preset quality threshold, outputs a judgment result of "qualified" or "unqualified" and synchronizes the result to the equipment control unit; Data storage unit: Used to store original images, defect data and judgment results, supporting production traceability and model iterative optimization.
[0013] As a preferred embodiment of the present invention, the AI machine vision module further includes: Human-computer interaction unit: Equipped with a display screen and operation buttons, it is used to display inspection images, defect annotations and pass rate statistics in real time, and supports manual review and parameter adjustment; Communication unit: It adopts Ethernet or MQTT protocol to interface with the production line MES system to realize the uploading of test data, the issuance of production instructions and remote monitoring; Model update unit: Supports importing new defect detection models via wired / wireless methods, enabling algorithm iteration without downtime and adapting to new product detection needs.
[0014] Building upon the above, human-computer interaction, data communication, and model update functions are added: the human-computer interaction unit provides real-time display and operation; the communication unit uploads data to the production line system; and the model update unit enables online algorithm upgrades. The system displays the inspection screen via a touchscreen; communicates with the MES system via Ethernet; and imports new models via USB or network, with the system automatically switching to and retaining the old version, supporting rollback.
[0015] As a preferred embodiment of the present invention, the result determination unit establishes closed-loop control logic with the equipment control unit: For the first product: if it is determined to be "qualified", the control wheel mechanism drives the product to flip 180° and triggers a second inspection of the lower surface; if it is determined to be "unqualified", the control turntable rotates 90° to align with the sorting hole and performs a rejection action. For the second product: if it is determined to be "unqualified", the first electric push rod is extended to drive the spring to open the clamp and push the product to the sorting area; if it is determined to be "qualified", there is no sorting action and the product continues to flow with the toroidal turntable; the equipment control unit customizes the graded processing strategy according to the severity of the defect.
[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. By setting up a first and second bearing area side by side on the toroidal turntable, corresponding to the first product that needs to be inspected on both the upper and lower surfaces and the second product that only needs to be inspected on the upper surface, and with the dual-camera design of the AI machine vision module, the synchronous flow and separate inspection of the two types of supporting products can be realized without the need to add multiple additional devices. This effectively solves the problem that traditional quality control equipment can only inspect one type of product and cannot adapt to the differentiated inspection needs of supporting products, greatly improving the compatibility and versatility of the equipment and adapting to flexible production scenarios with multiple varieties and small batches.
[0017] 2. The system integrates the turntable mechanism's circulation function, the carrier frame and wheel mechanism's flipping function, the spring and clamp plate's sorting function, and the AI machine vision module's detection function into one unit, achieving full automation of the "product circulation, surface inspection, qualified flipping, and unqualified rejection" process without manual intervention. Compared to the traditional manual inspection, separate flipping, and separate sorting methods, this significantly shortens the inspection cycle and improves production efficiency. Furthermore, the flipping and inspection actions of the first product are seamlessly connected, and the sorting and inspection actions of the second product are synchronized, avoiding efficiency losses caused by process disconnects. Attached Figure Description
[0018] Figure 1 This is a first-view three-dimensional structural diagram of the quality control equipment based on AI machine vision provided in an embodiment of the present invention; Figure 2 This is provided by the embodiments of the present invention. Figure 1 A magnified structural diagram of part A in the middle; Figure 3 This is a second-view three-dimensional structural diagram of the quality control equipment based on AI machine vision provided in an embodiment of the present invention; Figure 4 This is provided by the embodiments of the present invention. Figure 3 A magnified structural diagram of part B in the middle section; Figure 5 This is provided by the embodiments of the present invention. Figure 3 A magnified structural diagram of section C; Figure 6 This is provided by the embodiments of the present invention. Figure 3 A magnified structural diagram of section D in the middle; Figure 7 This is a module block diagram provided in an embodiment of the present invention.
[0019] In the diagram: 101, First motor; 102, Carrying plate; 103, Annular turntable; 104, First carrying area; 105, Second carrying area; 106, Sorting hole; 107, Connecting rod; 201, Upright frame; 202, Horizontal plate; 203, Turntable; 204, Upper frame; 205, Through hole; 206, Clamping plate; 207, First electric push rod; 208, Spring; 209, Second electric push rod; 210, Lifting frame; 211, Straight plate; 212. Circular plate; 213. Circular track; 214. Vertical block; 215. Slot; 301. Second motor; 302. Wheel; 401. First camera; 402. Second camera; 403. Circular fill light; 404. Data preprocessing unit; 405. AI inference unit; 406. Result judgment unit; 407. Data storage unit; 408. Human-computer interaction unit; 409. Communication unit; 410. Model update unit. Detailed Implementation
[0020] To further understand the invention's content, features, and effects, the following embodiments are provided, and detailed descriptions are given in conjunction with the accompanying drawings.
[0021] The structure of the present invention will now be described in detail with reference to the accompanying drawings.
[0022] like Figures 1 to 7 As shown in the figure, an AI machine vision-based quality control device provided by an embodiment of the present invention includes: a turntable mechanism, wherein the annular turntable 103 of the turntable mechanism has a plurality of parallel first bearing areas 104 and second bearing areas 105, and the annular turntable 103 is provided with sorting holes 106; a support frame, wherein an upper frame 204 is rotatably connected to the lower side of the support frame, the upper frame 204 is provided with a through hole 205 in the middle, and clamping plates 206 are fixedly connected to both sides of the upper frame 204; a turntable 203 is provided on the lower side of the support frame, and a turntable 203 is fixedly connected to the lower side of the turntable 203. A first electric push rod 207 passes through the through hole 205. A spring piece 208 is fixedly connected to the lower end of the first electric push rod 207. Both sides of the spring piece 208 are fixedly connected to the clamping plate 206. A first bearing area 104 is located between the two clamping plates 206. A wheel mechanism is provided on the side of the clamping plates 206 that are close to each other. The AI machine vision module includes a first camera 401 and a second camera 402, which can be aimed at the first bearing area 104 and the second bearing area 105 respectively, and are used to detect the first product and the second product respectively.
[0023] Specifically, the turntable mechanism includes a first motor 101, the output shaft of which is fixedly connected to a carrier plate 102. A connecting rod 107 is fixedly connected to the edge of the carrier plate 102, and a toroidal turntable 103 is fixedly connected to the end of the connecting rod 107. The first carrier area 104, the second carrier area 105, and the sorting hole 106 are all located on the toroidal turntable 103. The first motor 101 drives the carrier plate 102 to rotate, and the connecting rod 107 drives the toroidal turntable 103 to perform indexing movements, causing the first carrier area 104, the second carrier area 105, and the sorting hole 106 to enter the inspection station, achieving continuous feeding, inspection, and discharge. The first motor 101 is a stepper motor, and its angle is controlled by a PLC. The carrier plate 102 and the toroidal turntable 103 are rigidly connected by multiple connecting rods 107.
[0024] Specifically, the wheel mechanism includes a second motor 301 and a wheel 302. The wheel 302 is fixedly connected to the output shaft of the second motor 301. The second motor 301 is fixedly connected to the side of the clamping plate 206 that is furthest away from it, and the wheel 302 is located on the side of the clamping plate 206 that is closest to it. The second motor 301 drives the wheel 302 to rotate, using the friction between the wheel 302 and the first product to achieve clamping and 180° rotation, thereby completing the switching of upper and lower surface detection. The second motor 301 is fixed to the outside of the clamping plate 206, and its output shaft passes through the clamping plate 206 and connects to the wheel 302. The wheel 302 is made of flexible and wear-resistant material. When rotation is required, the second motor 301 drives the wheel 302 to rotate 180°, causing the first product to rotate synchronously.
[0025] Specifically, the first camera 401 is disposed on the lower surface of the spring piece 208, and the second camera 402 is disposed on the side of the clamping plate 206. The first camera 401, mounted on the lower surface of the spring piece 208, can directly capture images of the upper and lower surfaces of the first product; the second camera 402, mounted on the side of the clamping plate 206, is used to capture images of the upper surface of the second product. The two cameras do not interfere with each other. The first camera 401 faces downwards towards the first bearing area 104; the second camera 402 faces vertically towards the second bearing area 105; both are equipped with supplementary lighting structures to ensure clear imaging of the upper and lower surfaces.
[0026] Specifically, the support frame includes an upright frame 201, on which a horizontal plate 202 is fixedly connected. A second electric push rod 209 is mounted on the horizontal plate 202, and a lifting frame 210 is fixedly connected to the second electric push rod 209. The second electric push rod 209 drives the lifting frame 210 to rise and fall as a whole, thereby causing the upper frame 204, clamping plate 206, and wheel mechanism to rise or fall, achieving the clamping of the product and its release. The second electric push rod 209 is fixed to the horizontal plate 202, and its lower end is connected to the lifting frame 210. The lifting frame 210 moves up and down along the guide of the upright frame 201; when the push rod extends, it lowers to clamp; when the push rod retracts, it rises to release.
[0027] Specifically, the lifting frame 210 includes a straight plate 211 and an annular plate 212. An annular track 213 is provided on the lower side of the annular plate 212. The upper frame 204 includes a vertical block 214 with a slot 215 slidably connected to the annular track 213. The annular track 213 of the lifting frame 210 slidably engages with the slot 215 of the upper frame 204, allowing the upper frame 204 to rotate horizontally around its center, providing freedom of movement for the turntable 203 to rotate the upper frame 204 and align it with the sorting hole 106. The annular track 213 is located under the annular plate 212; the vertical block 214 of the upper frame 204 has a slot 215 that slidably engages with the track; the turntable 203 drives the upper frame 204 to rotate, aligning defective products with the sorting hole 106.
[0028] Specifically, the spring piece 208 is an arc-shaped elastic metal sheet. The middle part of the spring piece 208 is fixedly connected to the lower end of the first electric push rod 207, and the two ends of the spring piece 208 are respectively fixedly connected to the inner sidewalls of the two side clamping plates 206. When the first electric push rod 207 extends, the arc-shaped spring piece 208 is pressed down, generating elastic deformation, causing the two side clamping plates 206 to open outward, realizing the rapid rejection of the second product; when the push rod retracts, the spring piece 208 returns to its original position, and the clamping plates 206 close. The spring piece 208 is made of an elastic metal sheet; the middle part is fixedly connected to the first electric push rod 207, and the two ends are fixedly connected to the clamping plates 206; when the push rod extends, the spring piece deforms, and the clamping plates open; when the push rod shortens, the spring piece returns to its original position, and the clamping plates close.
[0029] Specifically, the AI machine vision module includes: Image acquisition unit: Composed of a first camera 401, a second camera 402 and a corresponding ring light 403. The first camera 401 is used to acquire images of the upper and lower surfaces of the first product, and the second camera 402 is used to acquire images of the upper surface of the second product. Data preprocessing unit 404: electrically connected to the image acquisition unit, used to perform noise reduction, enhancement, and normalization processing on the acquired image, and output standardized image data; AI Inference Unit 405: Loaded with a pre-trained surface defect detection model, used to receive standardized image data, identify defects such as scratches, stains, and deformations on the product surface, and output information on the defect type, location, and severity. Result Judgment Unit 406: Based on the comparison between the defect information and the preset quality threshold, output the judgment result of "qualified" or "unqualified" and synchronize the result to the equipment control unit; Data storage unit 407: Used to store original images, defect data and judgment results, supporting production traceability and model iterative optimization.
[0030] The AI machine vision module completes the entire intelligent inspection process through image acquisition, preprocessing, AI inference, result judgment, and data storage. The first camera 401 and the second camera 402 acquire images; the data preprocessing unit 404 optimizes the images; the AI inference unit 405 identifies defects; the result judgment unit 406 outputs pass / fail; and the data storage unit 407 records the data. An embedded AI processor is used; images are input into the defect detection model after denoising and enhancement; the model outputs the defect location and level; after comparison with a preset threshold, a judgment signal is output; all images and results are automatically saved for traceability.
[0031] Furthermore, the AI machine vision module also includes: Human-computer interaction unit 408: Equipped with a display screen and operation buttons, it is used to display inspection images, defect markings and pass rate statistics in real time, and supports manual review and parameter adjustment; Communication unit 409: It adopts Ethernet or MQTT protocol to interface with the production line MES system to realize the uploading of detection data, the issuance of production instructions and remote monitoring; Model Update Unit 410: Supports importing new defect detection models via wired / wireless methods, enabling algorithm iteration without downtime and adapting to new product detection needs.
[0032] Building upon the above, human-computer interaction, data communication, and model update functions are added: the human-computer interaction unit 408 provides real-time display and operation; the communication unit 409 uploads data to the production line system; and the model update unit 410 enables online algorithm upgrades. The system displays the detection screen via a touchscreen; communicates with the MES system via Ethernet; and imports new models via USB or network, with the system automatically switching to and retaining the old version, supporting rollback.
[0033] Working principle: The result determination unit 406 establishes closed-loop control logic with the equipment control unit: For the first product: if it is determined to be "qualified", the control wheel mechanism drives the product to flip 180° and triggers a second inspection of the lower surface; if it is determined to be "unqualified", the control turntable 203 rotates 90° to align with the sorting hole 106 and performs a rejection action. For the second product: if it is determined to be "unqualified", the first electric push rod 207 is extended, and the spring 208 is driven to open the clamp 206, and the product is dropped into the sorting area; if it is determined to be "qualified", there is no sorting action, and the product continues to flow with the toroidal turntable 103. The equipment control unit can customize a graded processing strategy based on the severity of the defect, such as marking minor defects and directly removing severe defects.
[0034] The result determination unit 406 sends a signal to the equipment control unit, forming a closed loop of detection, decision-making, and execution: if the first product is qualified, it is rotated 180°; if the first product is unqualified, the turntable 203 rotates 90° to reject it; if the second product is unqualified, the first electric push rod 207 pushes open the spring 208 to reject it and can be processed according to defect level. The equipment control unit adopts a PLC or microcontroller; it controls the second motor 301, turntable 203, first electric push rod 207, and second electric push rod 209 respectively according to the AI signal; multiple defect thresholds can be set to achieve accurate sorting.
[0035] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0036] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A quality control device based on AI machine vision, characterized in that, include: The turntable mechanism has a number of first bearing areas (104) and second bearing areas (105) arranged side by side on the annular turntable (103), and the annular turntable (103) is provided with sorting holes (106). The support frame is rotatably connected to the lower side of the support frame (204), the upper frame (204) is provided with a through hole (205) in the middle, and the upper frame (204) is fixedly connected to the two sides of the upper frame (206). A turntable (203) is provided on the lower side of the support frame. A first electric push rod (207) is fixedly connected to the lower side of the turntable (203). The first electric push rod (207) passes through the through hole (205). A spring piece (208) is fixedly connected to the lower end of the first electric push rod (207). The two sides of the spring piece (208) are fixedly connected to the clamping plate (206). The first bearing area (104) is located between two clamping plates (206); a wheel mechanism is provided on the side of the clamping plates (206) that are close to each other; The AI machine vision module includes a first camera (401) and a second camera (402), which can be aimed at the first carrier area (104) and the second carrier area (105) respectively, and are used to detect the first product and the second product respectively.
2. The quality control equipment based on AI machine vision as described in claim 1, characterized in that: The turntable mechanism includes a first motor (101), the output shaft of the first motor (101) is fixedly connected to a bearing plate (102), the edge of the bearing plate (102) is fixedly connected to a connecting rod (107), the end of the connecting rod (107) is fixedly connected to an annular turntable (103), and the first bearing area (104), the second bearing area (105) and the sorting hole (106) are all provided on the annular turntable (103).
3. The quality control equipment based on AI machine vision as described in claim 1, characterized in that: The wheel mechanism includes a second motor (301) and a wheel (302), the wheel (302) being fixedly connected to the output shaft of the second motor (301); the second motor (301) being fixedly connected to the side of the clamping plate (206) that is far away from each other, and the wheel (302) being located on the side of the clamping plate (206) that is close to each other.
4. The quality control equipment based on AI machine vision as described in claim 1, characterized in that: The first camera (401) is disposed on the lower surface of the spring sheet (208), and the second camera (402) is disposed on the side of the clamping plate (206).
5. The quality control equipment based on AI machine vision as described in claim 1, characterized in that: The support frame includes a vertical frame (201), a horizontal plate (202) is fixedly connected to the vertical frame (201), a second electric push rod (209) is provided on the horizontal plate (202), and a lifting frame (210) is fixedly connected to the second electric push rod (209).
6. The quality control equipment based on AI machine vision as described in claim 5, characterized in that: The lifting frame (210) includes a straight plate (211) and an annular plate (212). An annular track (213) is provided on the lower side of the annular plate (212). The upper frame (204) includes a vertical block (214). A slot (215) is provided on the vertical block (214). The slot (215) is slidably connected to the annular track (213).
7. The quality control equipment based on AI machine vision as described in claim 1, characterized in that: The spring sheet (208) is an arc-shaped elastic metal sheet. The middle part of the spring sheet (208) is fixedly connected to the lower end of the first electric push rod (207), and the two ends of the spring sheet (208) are fixedly connected to the inner sidewalls of the two side clamps (206).
8. The quality control equipment based on AI machine vision as described in claim 1, characterized in that: The AI machine vision module includes: Image acquisition unit: composed of a first camera (401), a second camera (402) and a corresponding ring light (403). The first camera (401) is used to acquire images of the upper and lower surfaces of the first product, and the second camera (402) is used to acquire images of the upper surface of the second product. Data preprocessing unit (404): electrically connected to the image acquisition unit, used to perform noise reduction, enhancement and normalization processing on the acquired image, and output standardized image data; AI inference unit (405): Loaded with a pre-trained surface defect detection model, used to receive standardized image data, identify defects such as scratches, stains, and deformations on the product surface, and output information on the type, location, and severity of defects; Result Judgment Unit (406): Based on the comparison between the defect information and the preset quality threshold, output the judgment result of "qualified" or "unqualified" and synchronize the result to the equipment control unit; Data storage unit (407): Used to store original images, defect data and judgment results, supporting production traceability and model iterative optimization.
9. The quality control equipment based on AI machine vision as described in claim 8, characterized in that: The AI machine vision module also includes: Human-computer interaction unit (408): Equipped with a display screen and operation buttons, it is used to display inspection images, defect markings and pass rate statistics in real time, and supports manual review and parameter adjustment; Communication unit (409): It adopts Ethernet or MQTT protocol to interface with the production line MES system to realize the uploading of detection data, the issuance of production instructions and remote monitoring; Model update unit (410): Supports importing new defect detection models via wired or wireless means, and can complete algorithm iteration without downtime to adapt to the detection needs of new products.
10. A quality control device based on AI machine vision as described in claim 8, characterized in that: The result determination unit (406) establishes closed-loop control logic with the equipment control unit: For the first product: if it is determined to be "qualified", the control wheel mechanism drives the product to flip 180° and triggers a second inspection of the lower surface; if it is determined to be "unqualified", the control turntable (203) rotates 90° to align with the sorting hole (106) and performs a rejection action; For the second product: if it is determined to be "unqualified", the first electric push rod (207) is extended, and the spring (208) is driven to open the clamp (206) and push the product to the sorting area; if it is determined to be "qualified", there is no sorting action, and the product continues to flow with the toroidal turntable (103); The equipment control unit defines a tiered processing strategy based on the severity of the defect.