Product identification attachment identification system and equipment

Abstract

The invention relates to the field of product identifier identification, and discloses a product identifier attachment identification system and equipment, the equipment comprises an information flow control module, a pre-labeling station, labeling equipment, an identifier printing defect abnormity processing module and an equipment hardware fault detection module; meanwhile, the system is applied to the equipment and comprises a product information acquisition module, a labeling instruction issuing module, a labeling execution module and a detection module; in the working process, after products flow into a pre-labeling station, after detection and information collection, an upper computer is linked with an MES to obtain printing information, a printing system is controlled to print labels, after pre-detection is qualified, the labels are attached through a mechanical arm, then secondary detection is conducted through multiple sets of detection modules, and a repairing mechanism is started when the products are abnormal; according to the invention, full-process automation can be realized, the production efficiency is improved, the detection reliability is enhanced, flexible production requirements can be met, and the technical problems of low efficiency and poor detection quality of a traditional process and insufficient adaptability of existing equipment are solved.

Description

Technical Field

[0001] The present invention relates to the field of product identification and recognition, and in particular to a product identification attachment and recognition system and equipment. Background Art

[0002] In the field of industrial product production, labeling of packaging is a key link in ensuring product traceability and information transmission. It involves various carriers such as cartons, laminated films, metal plates / films, etc., and labeling is the current mainstream implementation method.

[0003] The traditional process system has significant technical limitations: First, it relies on manual labor to complete the attachment and inspection operations, which are affected by factors such as operating proficiency and physical load. The production rhythm is unstable and it is difficult to break through the bottleneck of processing volume per unit time; second, label content verification adopts a visual method, and the judgment consistency of key indicators such as character clarity and barcode integrity is insufficient. It is easy to cause false detection and missed detection due to human negligence, and there is a lack of error and fool-proofing mechanisms; third, faced with flexible production scenarios such as multi-variety switching and customized orders, manual adjustment of operation parameters takes a long time and is difficult to adapt to rapidly changing production needs.

[0004] Some leading manufacturing industries have also introduced automatic labeling equipment or visual inspection devices, but these can only meet the needs of label application and basic inspection, such as barcode information and print quality. However, once the production sequence of products changes or with the flexible production of customized products, the reliability of inspection will seriously decline and the stability of the equipment will be poor. Summary of the Invention

[0005] The purpose of the present invention is to provide a product labeling and identification system and equipment to solve the above technical problems.

[0006] The purpose of the present invention can be achieved through the following technical solutions: Product labeling and identification equipment, including: Information flow control module, used to communicate with MES, QMS and IOT to obtain product information and confirm the status of work permits; The pre-labeling station is equipped with: Photoelectric sensors are used to detect product inflow and control access to workstations; Length sensor, used to determine product size; The movable camera module can adjust its position according to the product size and collect the physical status information; Labeling equipment includes: The host computer is connected to the movable camera module for receiving and analyzing the physical status information of the product, extracting the model information and sending it to the MES, and receiving the corresponding printing information returned by the MES; At least two printing systems arranged in parallel, communicatively connected to the host computer, for receiving the printing information and executing label printing, and configured with a redundancy mechanism for automatic switching in the event of an abnormality; Pre-inspection module, including a movable camera and light source module, for initial quality inspection of printed labels; A robotic arm device, equipped with a bionic wiping suction cup, performs labeling actions based on three-dimensional space path planning according to pre-detection results; Multiple sets of movable inspection modules are used to perform secondary quality inspection of attached labels from multiple angles.

[0007] As a further technical solution, the working process of the pre-detection module is as follows: S20, receiving a printing completion signal from the host computer; S21, according to the position of the printing system and the label status, the movable camera light source module moves to the detection position; S22. Perform an initial quality inspection on the printed labels to identify defects such as incomplete printing, broken lines, uneven thickness, abnormal printing depth, and character adhesion. S23, if the detection result is OK, output an OK signal to the host computer; If the test result is NG, the following actions are triggered: The printing system discharges defective labels into a waste area; Notify the host computer, and the host computer instructs the printing system to reprint the label.

[0008] As a further technical solution, the working process of multiple groups of movable detection modules includes: S30, receiving a labeling completion signal from the robotic arm device; S31, three groups of detection modules automatically plan collaborative inspection paths based on the product's spatial posture and label position; S32. Each detection module collects multi-angle label images in real time during movement, covering at least three appearance surfaces of the product; S33. Comprehensive evaluation through visual analysis algorithm: Label pasting position accuracy; Surface bubble / wrinkle area ratio; Print quality: contrast and clarity; S34, output the judgment result: If all indicators are qualified, mark OK and upload to MES, triggering product discharge; If any indicator exceeds the tolerance, it will be marked as NG and executed synchronously; Upload defect type and location coordinates to MES; Send an activation instruction to the logo printing defect exception handling module.

[0009] As a further technical solution, the device further includes: The marking printing defect abnormality processing module is communicatively connected to the host computer and the multiple groups of movable detection modules, and is started when an NG judgment of the multiple groups of movable detection modules is received.

[0010] As a further technical solution, the working process of the marking printing defect exception processing module includes: An anomaly detection unit receives NG determinations and associated defect coordinate data sets from the plurality of movable detection modules; re-inspects the defect area, extracts the curve contours using a highlight segmentation algorithm, and determines the defect type; The repair unit determines whether the defect is a trend defect or a sudden defect based on defect type analysis, and then selects the corresponding repair mode; Secondary verification unit re-inspects the repaired area and verifies that the color difference ΔE between the repaired area and the original label is ≤ 1.5; If the verification is qualified, it is marked as repair OK, triggering product discharge; if the verification is unqualified, it is marked as permanent NG, triggering product abandonment.

[0011] As a further technical solution, the first repair mode is selected for trend defects, and the printing parameters are adjusted when the defects meet the historical prediction model; the historical prediction model is generated based on the historical defect data training of the pre-detection module and multiple groups of detection modules. For sudden defects, select the second repair mode and start the ink jet system: control the three-dimensional positioning of the micro-nozzle according to the defect coordinates; and perform local ink filling according to the original label color formula.

[0012] As a further technical solution, the device further includes a device hardware fault detection module, including: The data acquisition unit collects parameters in real time based on a multi-source sensor array, including: equipment three-dimensional spatial position coordinates, equipment vibration spectrum data, key mechanical component temperatures, motor operating current waveforms, and ambient temperature and humidity parameters; The health status calculation unit calculates the equipment health status value in real time based on a weighted combination of the vibration effective value, temperature rise rate, current harmonic distortion rate, and equipment life attenuation coefficient. The weighting factor is automatically adjusted according to the ambient temperature and humidity. The early warning unit sends an early warning signal to the security personnel terminal when the equipment health status value enters the early warning threshold range [m,n]; when the equipment health status value enters the shutdown threshold range [M,N], it triggers the equipment to stop suddenly and issues an audible and visual alarm.

[0013] As a further technical solution, the device hardware fault detection module further includes: Dynamic threshold generation unit, which dynamically calculates the warning threshold range [m,n] based on the equipment's cumulative operating time, regional climate conditions, and factory environmental factors; The shutdown threshold range [M, N] is dynamically calculated based on the equipment age, load intensity, and power supply stability.

[0014] A product identification attachment and identification system, using the product identification attachment and identification device, includes: Product information acquisition module, used to obtain product information; Labeling instruction sending module, the host computer sends the barcode information instruction to the printer for printing; The labeling execution module cooperates with the mechanical structure to complete the labeling. The labeling execution module implements labeling through the mechanical arm device of the equipment. Detection module, which uses visual depth analysis algorithm for automatic detection; Product inflow control module: Products flow into the pre-labeling station, and products are allowed to flow in when there is no product in the station; After the product flows in, the length sensor is used to determine the length of the product. The camera at the pre-labeling station moves to the corresponding position to take a photo, analyzes the photo content and sends it to the host computer. Then the product flows into the labeling station. Information interaction module: The host computer extracts the required machine model information based on the received information and sends the model information to MES; MES retrieves the printing information required by the machine model from the server and sends it to the printing software; After the printing software completes the typesetting, it controls the printer to print; Label detection and processing module: The pre-inspection module pre-inspects the label. If the inspection result is OK, the machine will grab it and stick it on the product. If the inspection result is NG, it will be placed on the waste board and reprinted. After labeling is completed, the inspection module will inspect the labeled labels again, determine whether the product is OK or NG, upload it to the MES server, and then discharge the product.

[0015] Beneficial effects of the present invention: (1) In order to solve the problem that the traditional process relies on manual attachment and visual inspection, which is not only time-consuming and labor-intensive, but also prone to efficiency fluctuations due to human fatigue, the present invention links the MES, QMS and IOT systems through the information flow control module to realize automatic acquisition of product information and operation permission judgment, eliminating the manual verification link; the parallel printing system cooperates with the redundant switching mechanism to ensure uninterrupted operation of the printing link; the robotic arm device is equipped with a bionic wiping suction cup, which completes labeling according to three-dimensional path planning, replacing manual repetitive labor. From product inflow inspection, label printing to attachment and quality verification, the entire process does not require manual intervention, completely getting rid of dependence on manual operation, improving the product processing volume per unit time, and avoiding efficiency fluctuations caused by manual operation, providing stable and efficient technical support for large-scale production.

[0016] (2) The present invention addresses the problem that existing automatic equipment can only complete basic inspections, has insufficient accuracy in identifying label defects, and lacks a systematic error prevention mechanism. By constructing a dual inspection system through a pre-inspection module and multiple sets of movable inspection modules, the pre-inspection module conducts a comprehensive screening of printed labels for the first time, identifying minor defects such as incomplete printing and character adhesion; multiple sets of inspection modules conduct secondary verification from multiple angles, comprehensively evaluating the accuracy of pasting position, surface condition and printing quality, covering multiple appearance surfaces of the product; the label printing defect abnormality processing module further takes repair measures according to the defect type, intervenes in advance in trend defects through parameter adjustment, and accurately repairs sudden defects with the help of an ink jet system, and cooperates with secondary verification to ensure the quality of repair; compared with traditional visual inspection and basic equipment inspection, it realizes closed-loop quality control of the entire process from label printing to attachment, which can reduce the risk of missed inspections and false inspections, and improve inspection reliability and error prevention capabilities.

[0017] (3) The present invention can automatically adapt the detection position according to the product size through the dynamic adjustment mechanism of the movable camera module and the detection module, without the need for manual re-debugging; the upper computer extracts the model information in real time based on the physical status information of the product, and links the MES to retrieve the corresponding printing parameters to ensure a rapid response when the production sequence changes; at the same time, the equipment hardware fault detection module collects data through multi-source sensors, dynamically generates warning and shutdown thresholds, and combines health status calculations to achieve early warning and automatic repair of equipment abnormalities, reducing production interruptions caused by equipment failures; whether it is small-batch multi-variety customized production or large-scale continuous production, it can maintain a stable operating state and consistent inspection accuracy, completely solving the adaptability problem of traditional equipment in flexible production scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention will be further described below with reference to the accompanying drawings.

[0019] Figure 1 This is the system logic diagram of the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] See also Figure 1 As shown, the present invention is a product identification attachment and identification device, comprising: The information flow control module is used to communicate with the manufacturing execution system (MES), the quality management system (QMS), and the Internet of Things (IoT) to obtain product information and confirm the status of work permits; In one implementation, the information flow control module communicates with the MES, QMS, and IOT using industrial Ethernet or wireless LAN, exchanging data via a preset communication protocol. The acquired product information includes product model, specifications, production batch, etc. Confirming the work permit status means checking whether the product meets the current production plan and quality requirements. The pre-labeling station is equipped with: Photoelectric sensors are used to detect product inflow and control access to workstations; Length sensor, used to determine product size; The movable camera module can adjust its position according to the product size and collect the physical status information; In one embodiment, the photoelectric sensor at the pre-labeling station is a diffuse reflection photoelectric sensor. When it detects a product, it sends a signal to control the blocking mechanism at that station to prevent subsequent products from flowing in. After the current product leaves, the blocking mechanism resets to allow the next product to flow in. The length sensor is a laser ranging sensor that calculates the length of the product by emitting a laser and receiving a reflected signal. The movement of the movable camera module is driven by a servo motor. Based on the product size measured by the length sensor, the host computer sends a command to control the servo motor to drive the camera module to a preset photo position. The resolution of the camera module is not less than 5 million pixels to ensure that the collected physical state information is clear and legible. The blocking mechanism adopts existing technology and can achieve the above functions and is not subject to protection. Labeling equipment includes: The host computer is connected to the mobile camera module and is used to receive and analyze the physical status information of the product, extract the model information and send it to the MES, and receive the corresponding printed information returned by the MES. The host computer is an industrial-grade computer equipped with dedicated control software to process the image information collected by the mobile camera module and extract the product model information through image recognition algorithms. For example, an improved YOLOv5 object detection algorithm is used, which is pre-trained with a dataset containing various product appearance features to achieve rapid recognition of product models. At least two printing systems are arranged in parallel and are communicatively connected to the host computer, for receiving the printing information and executing label printing, and are configured with a redundancy mechanism for automatic switching in the event of an abnormality. Each of the printing systems arranged in parallel has an independent control unit and paper supply mechanism. When an abnormality such as a paper jam or lack of ink occurs in one of the printing systems, the host computer automatically issues a command to switch to the other printing system. The switching process is completed within 1 second, ensuring continuous printing work. The pre-inspection module includes a movable camera and light source module, which is used to perform the initial quality inspection of printed labels. The movable camera and light source module of the pre-inspection module is also driven by a servo motor. The camera resolution is not less than 3 million pixels, and the light source is a ring-shaped LED light source with adjustable brightness according to the label material and color. A robotic arm device, equipped with a bionic wiping suction cup, performs labeling actions based on three-dimensional space path planning according to pre-detection results; Multiple sets of movable inspection modules are used to perform multi-angle secondary quality inspection on attached labels. Each set of movable inspection modules includes a camera and a light source. The camera resolution is not less than 3 million pixels, and the light source is a strip LED light source. Its movement is driven by a linear module and can be flexibly moved on a preset track to realize the inspection of different appearance surfaces of the product.

[0022] The working process of the pre-detection module is as follows: S20, receiving a printing completion signal from the host computer; S21. Based on the printing system position and label status, the movable camera light source module is moved to the detection position. The moving position of the movable camera light source module is pre-calculated and determined by the host computer based on the coordinate parameters of the printing system and the size information of the label. During the movement, the encoder provides real-time feedback of the position information to ensure accurate arrival at the detection position. Specific calculation process example: The coordinate system is established based on the center of the label outlet of the printing system. The X axis is along the label outlet direction and the Y axis is perpendicular to the label outlet surface. The coordinates are set as The label size parameters include length L (along the X-axis) and width W (along the Y-axis). The optimal detection distance of the camera light module is D (preset according to the lens focal length, such as 300mm). The detection field of view must cover a redundant area of ​​10% outside the label. Calculation steps: 1. Determine the label center coordinates: ; 2. Calculate the coordinates of the camera detection center: offset the distance D along the direction perpendicular to the label surface (Z axis), that is, ; 3. Field of view adjustment compensation: according to the label tilt angle (Feedback from the printing system), the coordinates are corrected using the rotation transformation formula: ; ; ; 4. Finally output the moving coordinates of the camera light source module , which is converted into a servo motor pulse signal by the host computer and controls the module to move to that position.

[0023] S22. Perform an initial quality inspection on the printed labels to identify defects such as incomplete printing, broken lines, uneven thickness, abnormal printing depth, and character adhesion. An image comparison algorithm is used to compare and analyze the collected label image with a preset standard image to identify various defects. For example, an image matching algorithm based on ORB features is used to compare the label image to be inspected with a standard template image. S23, if the detection result is OK, output an OK signal to the host computer; If the test result is NG, the following actions are triggered: The printing system discharges defective labels into a waste area; Notify the host computer, and the host computer instructs the printing system to reprint the label.

[0024] The working process of multiple sets of movable detection modules includes: S30, receiving a labeling completion signal from the robotic arm device; S31: The three detection modules automatically plan a collaborative detection path based on the product's spatial pose and tag location. Information about the product's spatial pose and tag location is provided by the host computer, and the A* search algorithm is used to automatically plan the collaborative detection path, ensuring that the detection modules do not collide during movement and that detection efficiency is maximized. S32. Each detection module collects multi-angle label images in real time during movement, covering at least three appearance surfaces of the product; S33. Comprehensive evaluation through visual analysis algorithm: Label pasting position accuracy; Surface bubble / wrinkle area ratio; Print quality: contrast and clarity; Among them, the visual analysis algorithm includes edge detection algorithm and grayscale analysis algorithm. The label pasting position accuracy is evaluated by calculating the deviation between the actual label position and the preset standard position. For example, the specific calculation is: Physical coordinates of standard label center ( is the label rotation angle, preset value); detect the label center pixel coordinates (x, y) in the image, and the image pixel spacing (e.g. 0.02mm / pixel, calibrated by camera resolution and object distance); slope of the label edge line in the image (used to calculate the actual rotation angle θ); Calculation steps: Actual physical coordinates ;in The physical coordinates corresponding to the image origin are pre-calibrated; Translational deviation: , ; Rotational deviation: ; Comprehensive position accuracy: using weighted deviation value ; (The weight is set according to the process requirements, the unit is unified in mm, and θ is converted to the mm deviation corresponding to radians, such as 1°≈0.017mm / mm radius).

[0025] Surface bubble / wrinkle area ratio: The ratio of the bubble / wrinkle area to the total label area is calculated using an image segmentation algorithm. Example calculation process: The total number of pixels in the label area in the image ; Number of pixels in bubble / wrinkle area , binary mask statistics output by the U-Net++ segmentation model; Calculation steps: Area share ; To eliminate edge noise, Perform morphological filtering, such as removing connected domains with an area less than 5 pixels, and retain two decimal places in the final R.

[0026] Contrast of printing quality: based on the grayscale difference between label characters and background; The average grayscale of the character area is Gchar, and the average grayscale of the background area is Gbg; Formula: C=(Gchar-Gbg) / (Gchar+Gbg), normalized to 0-1. The higher the value, the better the contrast. Usually C≥0.3 is required. Clarity D: based on edge gradient energy; Use the Laplace operator to convolve the label image to obtain the edge response matrix ; formula: ,matrix The higher the value, the clearer the edge. Usually D≥500 is required.

[0027] S34, output the judgment result: If all indicators are qualified, mark OK and upload to MES, triggering product discharge; If any indicator exceeds the tolerance, it will be marked as NG and executed synchronously; Upload defect type and location coordinates to MES; Send an activation instruction to the logo printing defect exception handling module.

[0028] The device further comprises: The marking printing defect abnormality processing module is communicatively connected to the host computer and the multiple groups of movable detection modules, and is started when an NG judgment of the multiple groups of movable detection modules is received.

[0029] The working process of the logo printing defect exception processing module includes: An anomaly detection unit receives NG determinations and associated defect coordinate data sets from the plurality of movable detection modules; re-inspects the defect area, extracts the curve contours using a highlight segmentation algorithm, and determines the defect type; The highlight segmentation algorithm uses a deep learning-based image segmentation model to accurately extract the curved contours of defect areas. In addition to the previously mentioned defect types, it also identifies damaged labels, stains, and other types. The secondary verification unit also uses image acquisition and analysis to re-inspect the repaired area. The color difference ΔE is calculated based on the color difference formula in the CIELab color space. For example, the U-Net++ deep learning segmentation model is used to extract defect area features through an encoder-decoder structure.

[0030] The repair unit determines whether the defect is a trend defect or a sudden defect based on defect type analysis, and then selects the corresponding repair mode; Secondary verification unit re-inspects the repaired area and verifies that the color difference ΔE between the repaired area and the original label is ≤ 1.5; If the verification is qualified, it is marked as repair OK, triggering product discharge; if the verification is unqualified, it is marked as permanent NG, triggering product abandonment.

[0031] As a further technical solution, trending defects select the first repair mode and adjust printing parameters when the defect meets the historical prediction model. The historical prediction model is generated based on historical defect data trained on pre-detection modules and multiple detection modules. For example, the historical prediction model uses an LSTM recurrent neural network model, which uses the past three months of pre-detection and secondary detection defect data, including defect type, frequency, and environmental parameters, as input to predict defect trends for the next seven days and output the probability of each defect. For example, if the model predicts that the probability of a print line break defect increases for three consecutive days and exceeds a threshold, it is determined to be a trending defect and the print head pressure parameters are automatically adjusted.

[0032] For sudden defects, select the second repair mode and start the ink jet system: control the three-dimensional positioning of the micro-nozzle according to the defect coordinates; and perform local ink filling according to the original label color formula.

[0033] The device also includes a device hardware fault detection module, including: The data acquisition unit collects parameters in real time based on a multi-source sensor array, including: equipment three-dimensional spatial position coordinates, equipment vibration spectrum data, key mechanical component temperatures, motor operating current waveforms, and ambient temperature and humidity parameters; In one implementation scheme, in a multi-source sensor array of a data acquisition unit, the three-dimensional spatial coordinates of the equipment are acquired by a laser tracker, the vibration spectrum data of the equipment are acquired by an acceleration sensor, the temperature of key mechanical components are acquired by an infrared temperature sensor, the motor operating current waveform is acquired by a current sensor, and the ambient temperature and humidity parameters are acquired by a temperature and humidity sensor. The data acquisition frequency of all sensors is no less than 10 Hz. The health status calculation unit calculates the equipment health status value in real time based on a weighted combination of the vibration effective value, temperature rise rate, current harmonic distortion rate, and equipment life attenuation coefficient. The weighting factor is automatically adjusted according to the ambient temperature and humidity. In the weighted combination of the health status calculation unit, the weight factor of each parameter is pre-set through experiments and data analysis, and stored in the host computer. The corresponding weight factor is automatically called according to the changes in ambient temperature and humidity; the health status value The expression is: ; Each parameter is normalized and mapped to the range of 0-1. The closer the F value is to 1, the worse the device status is. =Accumulated operating time / design life; The early warning unit sends an early warning signal to the security personnel terminal when the equipment health status value enters the early warning threshold range [m,n]. When the equipment health status value enters the shutdown threshold range [M,N], the equipment is triggered to stop suddenly and an audible and visual alarm is issued. In one embodiment, the warning signal sent by the warning unit to the security personnel terminal includes the equipment name, abnormal parameters, and warning level information. The sound and light alarm device is installed in a conspicuous position of the equipment. The sound alarm volume is not less than 80 decibels, and the light alarm is a red flashing light.

[0034] The device hardware fault detection module also includes: Dynamic threshold generation unit, which dynamically calculates the warning threshold range [m,n] based on the equipment's cumulative operating time, regional climate conditions, and factory environmental factors; The shutdown threshold range [M, N] is dynamically calculated based on the equipment age, load intensity, and power supply stability.

[0035] Specifically: Warning threshold [m,n]: Based on the cumulative operating time t of the equipment, the regional humidity H, and the temperature T; 、 、 They are time influence coefficient, humidity influence coefficient, and temperature influence coefficient, all of which are obtained based on the analysis of historical data and operating data; ; , is the width of the warning interval, which is a fixed value of 0.2; Shutdown threshold [M, N]: Based on service life y, load intensity S (actual load / rated load), power supply voltage fluctuation P (±%), and benchmark shutdown lower limit That is, the initial value when the equipment is new and the working conditions are ideal, and the influence coefficient of each variable 、 、 , obtained through empirical data analysis and is non-negative; ; N=1.0, the maximum value of the equipment health status, exceeding which the equipment will be shut down; A product identification attachment and identification system, using the product identification attachment and identification device, includes: Product information acquisition module, used to obtain product information; Labeling instruction sending module, the host computer sends the barcode information instruction to the printer for printing; The labeling execution module cooperates with the mechanical structure to complete the labeling. The labeling execution module implements labeling through the mechanical arm device of the equipment. Detection module, which uses visual depth analysis algorithm for automatic detection; Product inflow control module: The product flows into the pre-labeling station. When there is no product in the station, the product is allowed to flow in. The blocking mechanism in the product inflow control module is a pneumatic baffle, and its lifting action is controlled by a solenoid valve; After the product flows in, the length sensor is used to determine the length of the product. The camera at the pre-labeling station moves to the corresponding position to take a photo, analyzes the photo content and sends it to the host computer. Then the product flows into the labeling station. Information interaction module: The host computer extracts the required machine model information based on the received information and sends the model information to MES; MES retrieves the printing information required by the machine model from the server and sends it to the printing software; After the printing software completes the typesetting, it controls the printer to print; Label detection and processing module: The pre-inspection module pre-inspects the label. If the inspection result is OK, the machine will grab it and stick it on the product. If the inspection result is NG, it will be placed on the waste board and reprinted. After labeling is completed, the inspection module will inspect the labeled labels again, determine whether the product is OK or NG, upload it to the MES server, and then discharge the product.

[0036] It should be noted that the calculation formulas and various parameters involved in the calculations in the present invention have been dimensionally processed in advance, and the process of dimensionless processing is well known in the industry and will not be described here.

[0037] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. Product identification device, characterized in that: include: Information flow control module, used to communicate with MES, QMS and IOT to obtain product information and confirm the status of work permits; The pre-labeling station is equipped with: Photoelectric sensors are used to detect product inflow and control access to workstations; Length sensor, used to determine product size; The movable camera module can adjust its position according to the product size and collect the physical status information; Labeling equipment includes: The host computer is connected to the movable camera module for receiving and analyzing the physical status information of the product, extracting the model information and sending it to the MES, and receiving the corresponding printing information returned by the MES; At least two printing systems arranged in parallel, communicatively connected to the host computer, for receiving the printing information and executing label printing, and configured with a redundancy mechanism for automatic switching in the event of an abnormality; Pre-inspection module, including a movable camera and light source module, for initial quality inspection of printed labels; A robotic arm device, equipped with a bionic wiping suction cup, performs labeling actions based on three-dimensional space path planning according to pre-detection results; Multiple sets of movable inspection modules are used to perform secondary quality inspection of attached labels from multiple angles.

2. The product identification device according to claim 1, characterized in that: The working process of the pre-detection module is as follows: S20, receiving a printing completion signal from the host computer; S21, according to the position of the printing system and the label status, the movable camera light source module moves to the detection position; S22. Perform an initial quality inspection on the printed labels to identify defects such as incomplete printing, broken lines, uneven thickness, abnormal printing depth, and character adhesion. S23, if the detection result is OK, output an OK signal to the host computer; If the test result is NG, the following actions are triggered: The printing system discharges defective labels into a waste area; Notify the host computer, and the host computer instructs the printing system to reprint the label.

3. The product identification device according to claim 2, characterized in that: The working process of multiple sets of movable detection modules includes: S30, receiving a labeling completion signal from the robotic arm device; S31, three groups of detection modules automatically plan collaborative inspection paths based on the product's spatial posture and label position; S32. Each detection module collects multi-angle label images in real time during movement, covering at least three appearance surfaces of the product; S33. Comprehensive evaluation through visual analysis algorithm: Label pasting position accuracy; Surface bubble / wrinkle area ratio; Print quality: contrast and clarity; S34, output the judgment result: If all indicators are qualified, mark OK and upload to MES, triggering product discharge; If any indicator exceeds the tolerance, it will be marked as NG and executed synchronously; Upload defect type and location coordinates to MES; Send an activation instruction to the logo printing defect exception handling module.

4. The product identification device according to claim 3, characterized in that: The device further comprises: The marking printing defect abnormality processing module is communicatively connected to the host computer and the multiple groups of movable detection modules, and is started when an NG judgment of the multiple groups of movable detection modules is received.

5. The product identification device according to claim 4, characterized in that: The working process of the logo printing defect exception processing module includes: An anomaly detection unit receives NG determinations and associated defect coordinate data sets from the plurality of movable detection modules; re-inspects the defect area, extracts the curve contours using a highlight segmentation algorithm, and determines the defect type; The repair unit determines whether the defect is a trend defect or a sudden defect based on defect type analysis, and then selects the corresponding repair mode; Secondary verification unit re-inspects the repaired area and verifies that the color difference ΔE between the repaired area and the original label is ≤ 1.5; If the verification is qualified, it is marked as repair OK, triggering product discharge; if the verification is unqualified, it is marked as permanent NG, triggering product abandonment.

6. The product identification device according to claim 5, characterized in that: The first repair mode is selected for trend defects. When the defect meets the historical prediction model, the printing parameters are adjusted. The historical prediction model is generated based on the historical defect data training of the pre-detection module and multiple groups of detection modules. For sudden defects, select the second repair mode and start the ink jet system: control the three-dimensional positioning of the micro-nozzle according to the defect coordinates; and perform local ink filling according to the original label color formula.

7. The product identification device according to claim 1, characterized in that: The device also includes a device hardware fault detection module, including: The data acquisition unit collects parameters in real time based on a multi-source sensor array, including: equipment three-dimensional spatial position coordinates, equipment vibration spectrum data, key mechanical component temperatures, motor operating current waveforms, and ambient temperature and humidity parameters; The health status calculation unit calculates the equipment health status value in real time based on a weighted combination of the vibration effective value, temperature rise rate, current harmonic distortion rate, and equipment life attenuation coefficient. The weighting factor is automatically adjusted according to the ambient temperature and humidity. The early warning unit sends an early warning signal to the security personnel terminal when the equipment health status value enters the early warning threshold range [m,n]; when the equipment health status value enters the shutdown threshold range [M,N], it triggers the equipment to stop suddenly and issues an audible and visual alarm.

8. The product identification device according to claim 7, characterized in that: The device hardware fault detection module also includes: Dynamic threshold generation unit, which dynamically calculates the warning threshold range [m,n] based on the equipment's cumulative operating time, regional climate conditions, and factory environmental factors; The shutdown threshold range [M, N] is dynamically calculated based on the equipment age, load intensity, and power supply stability.

9. Product labeling and identification system, characterized in that: The product identification device according to any one of claims 1 to 8 is used, wherein the system comprises: Product information acquisition module, used to obtain product information; Labeling instruction sending module, the host computer sends the barcode information instruction to the printer for printing; The labeling execution module cooperates with the mechanical structure to complete the labeling. The labeling execution module implements labeling through the mechanical arm device of the equipment. Detection module, which uses visual depth analysis algorithm for automatic detection; Product inflow control module: Products flow into the pre-labeling station, and products are allowed to flow in when there is no product in the station; After the product flows in, the length sensor is used to determine the length of the product. The camera at the pre-labeling station moves to the corresponding position to take a photo, analyzes the photo content and sends it to the host computer. Then the product flows into the labeling station. Information interaction module: The host computer extracts the required machine model information based on the received information and sends the model information to MES; MES retrieves the printing information required by the machine model from the server and sends it to the printing software; After the printing software completes the typesetting, it controls the printer to print; Label detection and processing module: The pre-inspection module pre-inspects the label. If the inspection result is OK, the machine will grab it and stick it on the product. If the inspection result is NG, it will be placed on the waste board and reprinted. After labeling is completed, the inspection module will inspect the labeled labels again, determine whether the product is OK or NG, upload it to the MES server, and then discharge the product.

Images