Visual guidance-based automatic weighing and labeling method for metal ingot

Abstract

This invention provides a vision-guided automatic weighing and labeling method for metal ingots, comprising the following steps: a transfer module transfers a group of metal ingots onto an electronic scale; the electronic scale measures the weight of the metal ingot group and sends it to a control module; the control module acquires the weight data, generates label data, and sends it to a labeling module, which then prints the labels; a labeling robot moves the labeling module to a label-collecting position according to a preset path, and the labeling module collects the labels; the labeling robot moves the vision positioning module to a distance measuring position, and a distance sensor measures the distance between the labeling robot and the metal ingot group; based on the difference between the measured distance and the distance between the set shooting position and the metal ingot group, the labeling robot moves the vision positioning module to a shooting position, and an industrial camera captures a partial photograph of the metal ingot group; the control module acquires the photograph and performs image processing to obtain feature points of the metal ingots and calculates the offset between the shooting position and the set labeling position; the labeling robot moves the labeling module to the labeling position based on the offset to complete the labeling.

Description

Technical Field

[0001] This invention belongs to the field of non-ferrous metal smelting technology, and particularly relates to a vision-guided automatic weighing and labeling method for metal ingots. Background Technology

[0002] In non-ferrous metal smelting workshops, the surface temperature of metal ingots is typically maintained at 80-100℃ when they first come off the production line. Weighing, information binding, label printing, accurate labeling, and transfer must be completed within a short time to meet subsequent logistics traceability requirements. Currently, there are two main implementation methods for labeling metal ingots, both of which have significant drawbacks.

[0003] The first method is manual weighing and labeling. Operators use forklifts to move metal ingots onto a platform scale for weighing, and then manually label and affix the labels. Operators need to face the high-temperature metal ingots, which poses a risk of burns. This method is labor-intensive, inefficient, and prone to errors in labeling or omissions, directly affecting downstream logistics tracking and the company's profitability.

[0004] The second type is the traditional fixed coordinate labeling machine operation method. This method relies on preset fixed coordinates for labeling, which cannot adapt to the size dispersion problem after metal ingots are stacked. After 96 metal ingots are stacked into 11 layers, the overall height deviation can reach 5-10cm, and the edge offset is up to 7cm. This makes it impossible for traditional labeling machines to achieve dynamic compensation, resulting in a high labeling rework rate and poor compatibility. It cannot adapt to metal ingots of different specifications and stacking postures.

[0005] In addition, the existing equipment has not achieved a fully automated closed loop for the entire process of transfer, weighing, marking, labeling, and visual positioning compensation, which makes it difficult to meet the high-efficiency and precision requirements of large-scale metal ingot production.

[0006] Therefore, it is necessary to provide a vision-guided automatic weighing and labeling method for metal ingots to solve or at least partially solve the above problems. Summary of the Invention

[0007] This invention provides a vision-guided automatic weighing and labeling method for metal ingots, which enables automatic weighing and high-precision automatic labeling of metal ingots.

[0008] To achieve the above objectives, the present invention provides the following technical solution.

[0009] A vision-guided automatic weighing and labeling method for metal ingots includes a transfer module, a weighing module, a marking module, a control module, and a labeling robot. The weighing module includes a weighing platform and an electronic scale, with the electronic scale mounted on the weighing platform. The transfer module is used to transfer groups of metal ingots between a conveyor line and the electronic scale. The marking module and the control module are both housed in a control cabinet. The control cabinet, the labeling robot, and the weighing platform are sequentially arranged on one side of the conveyor line. The labeling robot's arm end is equipped with a vision positioning module and a labeling module. The control module connects to and controls the electronic scale, the marking module, the labeling robot, the vision module, and the labeling module. The vision positioning module includes an industrial camera and a distance sensor. The automatic weighing and labeling method includes the following steps:

[0010] S1: The transfer module transfers the metal ingot group that is stopped at the weighing position on the conveyor line to the electronic scale; the electronic scale measures the weight of the metal ingot group and sends the measured weight value to the control module;

[0011] S2: The control module acquires and stores the weight data of the metal ingot group, generates label data and sends it to the marking module. The marking module completes label printing according to the label data and sends a label printing completion signal back to the control module.

[0012] S3: The control module sends a label printing completion signal to the labeling robot, and the labeling robot drives the labeling module to the label picking position according to the preset path, and the labeling module picks up the label;

[0013] S4: The labeling robot moves the visual positioning module to a distance measuring position close to the top of the metal ingot group according to a preset path, and measures the distance to the metal ingot group through the distance measuring sensor;

[0014] S5: The labeling robot moves the vision positioning module to the shooting position based on the difference between the measured distance value and the set shooting position and the distance value between the metal ingot group, and the industrial camera takes a partial photo of the metal ingot group.

[0015] S6: After acquiring the photo, the control module performs image processing to obtain the feature points of the metal ingot and calculates the offset between the shooting position and the set labeling position; the labeling robot moves the labeling module to the labeling position according to the offset to complete the labeling.

[0016] S7: The labeling robot returns to its initial safe position, the transfer module transfers the metal ingot group to the conveyor line, and the conveyor line moves until the next metal ingot group reaches the weighing position.

[0017] Preferably, the control module and the electronic scale are connected via the ModBus_TCP protocol; step S1 further includes: after the transfer module transfers the metal ingot group to the electronic scale, it sends a weighing and labeling request signal to the control module; after receiving the request signal, the control module continuously obtains the real-time weight from the electronic scale; and the control module obtains stable and accurate weight data of the metal ingot group through a deployed filtering algorithm.

[0018] Preferably, the marking module includes a printer and a label peeling mechanism. The control module communicates with the printer via TCP / IP protocol. The label data includes weight data, shift number, serial number, and fixed information. Step S2 further includes: after receiving the label data, the printer adds it to the printing queue; the printer sends a printing ready signal to the control module; the control module controls the printer to print the label; when the label passes through the label peeling mechanism, the label is peeled off from the adhesive backing paper; and the printer sends a printing queue cleared signal to the control module.

[0019] Preferably, the control module communicates with the labeling robot via the PROFINET protocol, determines the label picking position and the distance measuring position based on the positions of the marking module, the labeling robot, and the weighing platform, and sets the path for the labeling robot to move from the initial safe position to the label picking position and from the label picking position to the distance measuring position.

[0020] Preferably, the labeling module includes a vacuum generator, a label suction cup, and a label positioning sensor. Step S3 further includes: the control module controls the vacuum generator to work, the label suction cup picks up the label, and after the label positioning sensor detects the label for a set time, the labeling module completes the label picking up.

[0021] Preferably, the shooting position is set at a fixed shooting distance from the surface of the metal ingot group. The distance value between the metal ingot group measured by the ranging sensor is compared with the set fixed shooting distance, the deviation between the distance value and the fixed shooting distance is calculated, and the deviation is stored in the register of the control module. This deviation is used as the focal length compensation value of the subsequent industrial camera.

[0022] Preferably, step S5 further includes the control module analyzing the features of the metal ingot within the field of view of the industrial camera based on the photo taken by the industrial camera. If no features are found within the field of view, the total height of the metal ingot group is determined to be too low. The control module then moves the labeling robot down a set distance relative to the photo position and re-analyzes the features of the metal ingot within the field of view of the industrial camera. If a feature is found, further image processing is performed on the photo. If no feature is found, the labeling robot returns to the initial safe position and issues an alarm, indicating that the metal ingot group is abnormal and awaiting confirmation. If the feature is offset to the edge within the field of view of the industrial camera, the total height of the metal ingot group is determined to be too high. The control module then moves the labeling robot up a set distance relative to the photo position and re-analyzes the features of the metal ingot within the field of view of the industrial camera. If the feature is close to the center, further image processing is performed on the photo. If no feature is found, the labeling robot returns to the initial safe position and issues an alarm, indicating that the metal ingot group is abnormal and awaiting confirmation. If the feature point is close to the center within the field of view of the camera, further image processing is performed on the photo.

[0023] Preferably, image processing of the photograph includes: converting the photograph to grayscale to obtain a black and white image; performing binarization processing on the black and white image and adjusting the image threshold to separate the background and features; performing morphological operations on the image to reduce noise, segmenting features and interference points, filtering out interference points and retaining features through particle filtering, capturing feature intersections to find feature points, obtaining the coordinates of the feature points, and comparing the coordinates of the feature points with the coordinates of the set label position to calculate the offset of the X-axis, Y-axis and angle A.

[0024] Preferably, the labeling module further includes a labeling position sensor, and step S6 further includes: the labeling robot moves the labeling module to the labeling position according to the offset, moves the labeling along the labeling linear direction until the labeling position sensor is triggered, sends a labeling position signal, shuts off the vacuum generator, waits for a set time, and then moves the labeling robot linearly a set distance away from the label, confirms that the labeling position sensor has not detected the label, and determines that the labeling is complete.

[0025] Preferably, the control module includes a controller, a touch screen, and an alarm. The control module monitors the operating status of the electronic scale, the marking module, the labeling robot, the vision module, and the labeling module in real time. When an abnormality occurs, it immediately issues an alarm, controls the alarm to sound, and displays the corresponding fault information on the touch screen.

[0026] Compared with the prior art, the technical solutions of the embodiments of the present invention have beneficial effects.

[0027] For example, the vision-guided automatic weighing and labeling method for metal ingots provided by this invention forms a closed-loop control through ranging, photography, calculation, compensation, labeling, and detection. The control module acquires the ranging deviation and visual offset, and the labeling robot performs coordinate transformation based on these offsets to adjust the spatial position and posture angle of the labeling module so that the label can be accurately aligned with the target position on the surface of the metal ingot. After the labeling position sensor is triggered, the labeling success is verified by detecting whether the label position sensor signal disappears, thus realizing fully automated closed-loop control; improving labeling accuracy and efficiency.

[0028] Furthermore, it enables the automated and standardized integration of weighing data and labeling information, avoiding errors caused by manual data entry. Attached Figure Description

[0029] Figure 1 This is a flowchart of a vision-guided automatic weighing and labeling method for metal ingots according to an embodiment of the present invention;

[0030] Figure 2 This is a schematic diagram of the layout of the transfer module, weighing module, marking module, control module, and labeling robot according to an embodiment of the present invention;

[0031] Figure 3 This is a schematic diagram of the visual positioning module and labeling module according to an embodiment of the present invention;

[0032] Figure 4 This is a schematic diagram of the control cabinet structure according to an embodiment of the present invention;

[0033] Figure 5 This is a schematic diagram of a conveyor assembly line consisting of a transfer module, a weighing module, a marking module, a control module, and a labeling robot, according to an embodiment of the present invention. Detailed Implementation

[0034] To make the objectives, features, and beneficial effects of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It is to be understood that the specific embodiments described below are merely for explaining the present invention and are not intended to limit the present invention. Furthermore, the same or similar reference numerals may be used in the drawings to refer to the same or similar elements in different embodiments, and descriptions of the same or similar elements in different embodiments, as well as descriptions of prior art elements, features, effects, etc., may be omitted.

[0035] Reference Figures 1-5 This invention provides a vision-guided automatic weighing and labeling method for metal ingots.

[0036] Specifically, the vision-guided automatic weighing and labeling method for metal ingots includes a transfer module 7, a weighing module 6, a marking module 2, a control module 1, and a labeling robot 5. The weighing module 6 includes a weighing platform 61 and an electronic scale 62, with the electronic scale 62 mounted on the weighing platform 61. The transfer module 7 is used to transfer the metal ingot group 100 between the conveyor line 200 and the electronic scale 62. The marking module 2 and the control module 1 are both located in a control cabinet 8. The control cabinet 8, the labeling robot 5, and the weighing platform 61 are arranged sequentially on one side of the conveyor line 200. The end of the arm of the labeling robot 5 is equipped with a vision positioning module 3 and a labeling module 4. The control module 1 connects to and controls the electronic scale 62, the marking module 2, the labeling robot 5, the vision module 3, and the labeling module 4. The vision positioning module 3 includes an industrial camera 31 and a distance sensor 32. The automatic weighing and labeling method includes the following steps:

[0037] S1: The transfer module 7 transfers the metal ingot group 100 stopped at the weighing position on the conveyor line 200 to the electronic scale 62; the electronic scale 62 measures the weight of the metal ingot group 100 and sends the measured weight value to the control module 1.

[0038] S2: Control module 1 acquires and stores the weight data of metal ingot group 100, generates label data and sends it to marking module 2. Marking module 2 completes label printing according to the label data and sends a label printing completion signal back to control module 1.

[0039] S3: Control module 1 sends the label printing completion signal to labeling robot 5. Labeling robot 5 drives labeling module 4 to the label picking position according to the preset path, and labeling module 4 picks up the label.

[0040] S4: The labeling robot 5 moves the visual positioning module 3 to a distance measuring position close to the top of the metal ingot group 100 according to the preset path, and measures the distance to the metal ingot group 100 through the distance measuring sensor 32.

[0041] S5: The labeling robot 5 moves the vision positioning module 3 to the shooting position based on the difference between the measured distance value and the set shooting position and the distance value between the metal ingot group 100, and the industrial camera 31 takes a partial photo of the metal ingot group 100.

[0042] S6: After the control module 1 acquires the photo, it performs image processing to obtain the feature points of the metal ingot and calculates the offset between the shooting position and the set labeling position; the labeling robot 5 drives the labeling module 4 to move to the labeling position according to the offset to complete the labeling.

[0043] S7: Labeling robot 5 returns to the initial safe position, and transfer module 7 transfers metal ingot group 100 to conveyor line 200. Conveyor line 200 moves until the next group of metal ingots 100 reaches the weighing position.

[0044] In some embodiments, the control module 1 and the electronic scale 62 are connected via the ModBus_TCP protocol; step S1 further includes: after the transfer module 7 transfers the metal ingot group 100 onto the electronic scale 62, it sends a weighing and labeling request signal to the control module 1. After receiving the request signal, the control module 1 continuously obtains the real-time weight from the electronic scale 62. The control module 1 obtains stable and accurate weight data of the metal ingot group 100 through the deployed filtering algorithm.

[0045] Specifically, the control module 1 continuously acquires real-time weight data and uses a stable filtering algorithm, rather than simple instantaneous sampling, to reduce mechanical vibration interference and obtain stable and accurate weighing data for the metal ingot group. This solves the problem of weight reading fluctuations caused by mechanical vibration and airflow disturbance after the metal ingot group 100 is transported to the electronic scale 62. Through continuous sampling and filtering algorithms (such as moving average or Kalman filtering), instantaneous interference is eliminated, and the subsequent marking process is triggered only after the weight data is stable, avoiding product information mismatch caused by incorrect weight data and keeping the weighing error within the allowable range.

[0046] In some embodiments, the marking module 2 includes a printer 21 and a label peeling mechanism 22. The control module 1 communicates with the printer 21 via TCP / IP protocol. The label data includes weight data, shift number, serial number, and fixed information. Step S2 also includes: after receiving the label data, the printer 21 adds it to the printing queue. The printer 21 sends a printing ready signal to the control module 1. The control module 1 controls the printer 21 to print the label. When the label passes through the label peeling mechanism 22, the label is peeled off from the backing paper. The printer 21 sends a printing queue cleared signal to the control module.

[0047] Specifically, the control module 1 realizes the automated and standardized integration of weighing data and labeling information to avoid manual input errors; through print queue management (ready signal, clear signal feedback) it coordinates the working rhythm of the printer 21 and the labeling robot 5 to prevent label printing delays from causing the labeling robot 5 to wait, or printing too fast from causing labels to accumulate and curl at the label peeling mechanism 22, ensuring that the timing of label feeding and labeling robot 5 label picking is matched.

[0048] In some embodiments, the control module 1 communicates with the labeling robot 5 via the PROFINET protocol, determines the label picking position and the distance measuring position based on the positions of the labeling module 2, the labeling robot 5 and the weighing platform 61, and sets the path for the labeling robot 5 to move from the initial safe position to the label picking position and from the label picking position to the distance measuring position.

[0049] In some embodiments, the labeling module 4 includes a vacuum generator 41, a label suction cup 42, and a label positioning sensor 43. Step S3 further includes: the control module 1 controls the vacuum generator 41 to work, the label suction cup 42 picks up the label, and after the label positioning sensor 43 detects the label for a set time, the labeling module 4 completes the label picking up.

[0050] Specifically, by continuously setting time detection (rather than instantaneous detection), false judgments caused by vacuum build-up delays, label jitter, or sensor malfunctions are eliminated. Labels are relatively hard and may curl up after peeling. Instantaneous detection may generate false signals due to label jitter. Setting time detection ensures that the vacuum suction is stably established and the label is firmly adsorbed, preventing the labeling robot 5 from carrying half-detached labels or running to the labeling position without a load, thus avoiding missed labels or labels falling off.

[0051] In some embodiments, the shooting position is set at a fixed shooting distance from the surface of the metal ingot group 100. The distance value between the metal ingot group 100 measured by the distance sensor 32 is compared with the set fixed shooting distance, the deviation between the distance value and the fixed shooting distance is calculated, and the deviation is stored in the register of the control module 1. The deviation is used as the focal length compensation value of the subsequent industrial camera 31.

[0052] Specifically, the visual positioning module 3 also includes a lens 33 and a light source 34 that are matched with the industrial camera 31.

[0053] In some embodiments, step S5 further includes: the control module 1 analyzes the features of the metal ingot within the field of view of the industrial camera 31 based on the photo taken by the industrial camera 31. If no features are found within the field of view, the total height of the metal ingot group 100 is determined to be too low. The control module 1 then moves the labeling robot 5, which in turn moves the vision positioning module 3 down a set distance relative to the photo position, and analyzes the features of the metal ingot within the field of view of the industrial camera 31 again. If a feature is found, further image processing is performed on the photo. If no feature is found, the labeling robot 5 returns to the initial safe position and issues an alarm, indicating that the metal ingot group 100 is abnormal and awaiting confirmation. If the feature is offset to the edge within the field of view of the industrial camera 31, the total height of the metal ingot group 100 is determined to be too high. The control module 1 then moves the labeling robot 5, which in turn moves the vision positioning module 3 up a set distance relative to the photo position, and analyzes the features of the metal ingot within the field of view of the industrial camera again. If the feature is close to the center, further image processing is performed on the photo. If no feature is found, the labeling robot 5 returns to the initial safe position and issues an alarm, indicating that the metal ingot group 100 is abnormal and awaiting confirmation. If the feature point is close to the center within the field of view of the camera, further image processing is performed on the photo.

[0054] Specifically, to address the issues of height deviation of 5cm-10cm and edge offset of 0-7cm for metal ingot group 100 after palletizing, a layered compensation strategy of first measuring distance and then using vision is adopted. The distance sensor 32 first detects the actual distance to the surface of the metal ingot group 100 and calculates the difference between the distance value of the set shooting position and the distance value of the metal ingot group 100. This deviation value is used to adjust the distance between the shooting position and the distance measuring position of the labeling robot to ensure that the industrial camera can form a clear image within the focal plane. Then, by acquiring features through visual imaging and judging whether the features are within the field of view of the industrial camera 31, the height coordinates (Z-axis compensation) of the robot shooting position and the labeling position are adjusted so that the labeling head can accurately reach the target height. This method of combining physical distance measurement and visual positioning breaks through the limitations of the fixed coordinate system and achieves self-adaptation for workpieces with discrete heights.

[0055] In some embodiments, image processing of a photograph includes: converting the photograph to grayscale to obtain a black and white image; performing binarization on the black and white image and adjusting the image threshold to separate the background and features; performing morphological operations on the image to reduce noise, segmenting features and interference points, filtering out interference points and retaining features through particle filtering, capturing feature intersections to find feature points, obtaining the coordinates of the feature points, and comparing the coordinates of the feature points with the coordinates of a set label position to calculate the offset of the X-axis, Y-axis, and angle A.

[0056] Specifically, the photo is converted to grayscale, the contrast is adjusted to make the black and white images distinct, and the brightness is adjusted to make the black and white contrast more obvious.

[0057] In some embodiments, the labeling module 4 further includes a labeling position sensor 44, and step S6 further includes: the labeling robot 5 moves the labeling module 4 to the labeling position according to the offset, moves the labeling along the labeling linear direction until the labeling position sensor 44 is triggered, sends a labeling position signal, turns off the vacuum generator 41, waits for a set time, and then moves the labeling robot 5 linearly away from the label direction by a set distance. After confirming that the labeling position sensor 43 has not detected the label, the labeling is considered complete.

[0058] Specifically, the labeling sensor 44 is triggered, detecting whether the labeling sensor 43 signal has disappeared (verifying that the label has detached from the label suction cup 42 and is adhered to the metal ingot surface); this prevents labeling failure caused by incomplete vacuum release, insufficient adhesive strength, or interference from the oxide layer on the metal ingot surface, which could lead to poor adhesion and cause the labeling robot 5 to retract. Through this secondary verification of the reverse movement over a set distance and the disappearance of the labeling sensor 43 signal, it is ensured that the label is indeed on the metal ingot surface and not still adsorbed on the label suction cup 42, minimizing the risk of missed labeling and preventing defective products from flowing into the next process.

[0059] Specifically, a closed-loop control system is formed through ranging, photographing, calculation, compensation, labeling, and detection. Control module 1 acquires ranging deviations and visual offsets (including X-axis and Y-axis translational offsets and angle A rotational offset). Labeling robot 5 performs coordinate transformations based on these offsets, adjusting the spatial position and orientation angle of labeling module 4 to ensure the label is precisely aligned with the target position on the metal ingot surface (offset controlled within 0.5mm, deflection angle controlled within 0.3°). After labeling placement sensor 44 is triggered, the success of labeling is verified by detecting whether the signal from label placement sensor 43 disappears, achieving fully automated closed-loop control.

[0060] In some embodiments, the control module 1 includes a controller (not shown), a touch screen 11, and an alarm 12. The control module 1 monitors the operating status of the electronic scale 62, the marking module 2, the labeling robot 5, the visual positioning module 3, and the labeling module 4 in real time. When an abnormality occurs, it immediately issues an alarm to control the alarm 12 to sound an alarm and displays the corresponding fault information on the touch screen 11.

[0061] Specifically, through the dual prompts of the alarm 12 and the touch screen 11, maintenance personnel can quickly locate the fault point (such as checking on the touch screen 11 whether it is "tag adsorption failure" or "photo taking failure"), greatly reducing the troubleshooting time and improving efficiency.

[0062] The vision-guided automatic weighing and labeling method for metal ingots according to embodiments of the present invention has the following advantages:

[0063] Significantly improved safety: The entire process of labeling 100 metal ingots is fully automated, eliminating the risk of burns for operators who do not need to directly contact the high-temperature metal ingots, reducing the intensity of manual labor, saving two operators per shift, and solving the safety pain points of traditional manual operation.

[0064] High labeling accuracy and low rework rate: By combining machine vision positioning with labeling robot 5, the machine achieves accurate detection and dynamic compensation of 100 poses of metal ingots. It can adapt to the size dispersion problem after metal ingot stacking (height deviation 5-10cm, edge offset 0-7cm). The labeling offset in the X-axis and Y-axis directions is controlled within 1mm, the deflection angle is controlled within 0.5°, and the rework rate is reduced to less than 1%, which is far superior to traditional fixed coordinate labeling machines and manual labeling.

[0065] Highly adaptable and compatible: It can adapt to metal ingots of different specifications and stacking postures. The labeling module can be adjusted to adapt to different sized labels, without the need for major equipment adjustments, reducing equipment debugging costs and making it suitable for large-scale metal ingot production scenarios.

[0066] High efficiency and low error: The entire process of weighing, marking, labeling and compensation is integrated to achieve closed-loop control. The labeling time for a single stack of metal ingots is reduced to less than 30 seconds, the information mislabeling rate is reduced to less than 0.1%, and the problems of missing labels and duplicate labels are avoided, thereby improving production efficiency and product traceability accuracy.

[0067] High stability and good maintainability: Real-time monitoring of the operating status of each module facilitates staff to troubleshoot faults and trace production data. At the same time, the modular design of each module makes later maintenance convenient and reduces equipment operation and maintenance costs.

[0068] Although specific embodiments have been described above, these embodiments are not intended to limit the scope of the invention, even when only a single embodiment is described with respect to a particular feature. The feature examples provided in this disclosure are intended to be illustrative and not limiting, unless otherwise stated. In practice, one or more technical features of the dependent claims may be combined with the technical features of the independent claims as needed and where technically feasible, and may be derived from the technical features of the respective independent claims in any suitable manner rather than solely by the specific combinations listed in the claims.

Claims

1. A vision-guided automatic weighing and labeling method for metal ingots, characterized in that, The system includes a transfer module, a weighing module, a marking module, a control module, and a labeling robot. The weighing module includes a weighing platform and an electronic scale, with the electronic scale mounted on the weighing platform. The transfer module is used to transfer metal ingots between the conveyor line and the electronic scale. The marking module and the control module are both housed in a control cabinet. The control cabinet, the labeling robot, and the weighing platform are sequentially arranged on one side of the conveyor line. The labeling robot's arm end is equipped with a vision positioning module and a labeling module. The control module connects to and controls the electronic scale, the marking module, the labeling robot, the vision module, and the labeling module. The vision positioning module includes an industrial camera and a distance sensor. The automatic weighing and labeling method includes the following steps: S1: The transfer module transfers the metal ingot group that is stopped at the weighing position on the conveyor line to the electronic scale; the electronic scale measures the weight of the metal ingot group and sends the measured weight value to the control module; S2: The control module acquires and stores the weight data of the metal ingot group, generates label data and sends it to the marking module. The marking module completes label printing according to the label data and sends a label printing completion signal back to the control module. S3: The control module sends a label printing completion signal to the labeling robot, and the labeling robot drives the labeling module to the label picking position according to the preset path, and the labeling module picks up the label; S4: The labeling robot moves the visual positioning module to a distance measuring position close to the top of the metal ingot group according to a preset path, and measures the distance to the metal ingot group through the distance measuring sensor; S5: The labeling robot moves the vision positioning module to the shooting position based on the difference between the measured distance value and the set shooting position and the distance value between the metal ingot group, and the industrial camera takes a partial photo of the metal ingot group. S6: After acquiring the photo, the control module performs image processing to obtain the feature points of the metal ingot and calculates the offset between the shooting position and the set labeling position; the labeling robot moves the labeling module to the labeling position according to the offset to complete the labeling. S7: The labeling robot returns to its initial safe position, the transfer module transfers the metal ingot group to the conveyor line, and the conveyor line moves until the next metal ingot group reaches the weighing position.

2. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, The control module and the electronic scale are connected via the ModBus_TCP protocol; step S1 further includes: after the transfer module transfers the metal ingot group to the electronic scale, it sends a weighing and labeling request signal to the control module. After receiving the request signal, the control module continuously obtains the real-time weight from the electronic scale. The control module obtains stable and accurate weight data of the metal ingot group through a deployed filtering algorithm.

3. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, The marking module includes a printer and a label peeling mechanism. The control module is connected to the printer via TCP / IP protocol. The label data includes weight data, shift number, serial number and fixed information. Step S2 further includes: after receiving the label data, the printer adds it to the printing queue; the printer sends a printing ready signal to the control module; the control module controls the printer to print the label; when the label passes through the label peeling mechanism, the label is peeled off from the adhesive backing paper; and the printer sends a printing queue cleared signal to the control module.

4. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, The control module communicates with the labeling robot via the PROFINET protocol. It determines the label picking position and the distance measuring position based on the positions of the marking module, the labeling robot, and the weighing platform, and sets the path for the labeling robot to move from the initial safe position to the label picking position and from the label picking position to the distance measuring position.

5. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, The labeling module includes a vacuum generator, a label suction cup, and a label positioning sensor. Step S3 further includes: the control module controls the vacuum generator to work, the label suction cup picks up the label, and after the label positioning sensor detects the label for a set time, the labeling module completes the label picking up.

6. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, The shooting position is set at a fixed shooting distance from the surface of the metal ingot group. The distance value between the metal ingot group measured by the ranging sensor is compared with the set fixed shooting distance, the deviation between the distance value and the fixed shooting distance is calculated, and the deviation is stored in the register of the control module. This deviation is used as the focal length compensation value of the subsequent industrial camera.

7. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, Step S5 further includes the control module analyzing the characteristics of the metal ingot within the field of view of the industrial camera based on the photos taken by the industrial camera. If no characteristics are found within the field of view, it is determined that the total height of the metal ingot group is too low. The control module then moves the labeling robot down a set distance relative to the photo position and analyzes the characteristics of the metal ingot within the field of view of the industrial camera again. If a characteristic is found, the photo is further processed. If no characteristic is found, the labeling robot returns to the initial safe position and issues an alarm, indicating that the metal ingot group is abnormal and awaiting confirmation and processing. If the feature is offset to the edge within the field of view of the industrial camera, the total height of the metal ingot group is considered too high. The labeling robot is then controlled to move the vision positioning module upwards by a set distance relative to the shooting position. The situation of the metal ingot feature within the field of view of the industrial camera is analyzed again. If the feature is close to the center, further image processing is performed on the photo. If the feature cannot be found, the labeling robot returns to the initial safe position and issues an alarm, indicating that the metal ingot group is abnormal and awaiting confirmation and processing. If the feature point is close to the center within the field of view of the camera, further image processing is performed on the photo.

8. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, Image processing of the photo includes: converting the photo to grayscale to obtain a black and white image; binarizing the black and white image and adjusting the image threshold to separate the background and features; performing morphological operations to reduce noise, segmenting features and interference points, filtering out interference points and retaining features through particle filtering, finding feature points by capturing feature intersections, obtaining the coordinates of the feature points, and comparing the coordinates of the feature points with the coordinates of the set label position to calculate the offset of the X-axis, Y-axis and angle A.

9. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, The labeling module also includes a labeling position sensor. Step S6 further includes: the labeling robot moves the labeling module to the labeling position according to the offset, moves the labeling along the labeling linear direction until the labeling position sensor is triggered, sends a labeling position signal, shuts off the vacuum generator, waits for a set time, and then moves the labeling robot linearly away from the label by a set distance. Once it is confirmed that the labeling position sensor has not detected the label, the labeling is considered complete.

10. The method for automatic weighing and labeling of metal ingots based on vision guidance according to claim 1, characterized in that, The control module includes a controller, a touch screen, and an alarm. The control module monitors the operating status of the electronic scale, the marking module, the labeling robot, the vision module, and the labeling module in real time. When an abnormality occurs, it immediately issues an alarm, controls the alarm to sound, and displays the corresponding fault information on the touch screen.

Images