Anti-counterfeit label film coating tension control method and system

Abstract

The invention provides an anti-counterfeit label film coating tension control method and system, and relates to the technical field of film coating tension control. The method comprises the following steps: acquiring surface image data of a substrate film; identifying and acquiring real-time coordinates of the feature points; calculating the local strain of the substrate film according to the real-time coordinates of the feature points; calculating to obtain a strain deviation; calculating the time required by the film section corresponding to the strain deviation to reach the film coating station as the adjustment time; calculating a tension compensation amount according to the strain deviation; and within the adjusting time, the tension control unit adjusts the tension applied to the substrate film according to the tension compensation amount. According to the method, through accurate local strain measurement and prospective tension adjustment, it is ensured that each section of film is kept in an ideal strain state at the moment of film coating, so that the integrity and functional effectiveness of optical microstructures such as diffraction gratings are thoroughly protected, and the technical problems in anti-counterfeit label film coating production are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of lamination tension control, and in particular to a method and system for controlling the lamination tension of an anti-counterfeiting label. Background Art

[0002] In the roll-to-roll production process for security labels, a key step is laminating the substrate film printed with security graphics to enhance the label's durability and protect the security features within. This production line typically includes an unwinder, a tension control unit, a laminating station, a traction unit, and a rewinder. The substrate roll bearing the security information and the transparent protective film roll are each fed from the unwinder. After passing through a series of guide rollers and a tension control unit, they enter the laminating station for lamination. Finally, the rewinder takes up the finished rolls. To ensure lamination quality and prevent wrinkles, bubbles, or stretching, the tension of the substrate film and protective film must be controlled before lamination.

[0003] The core value of anti-counterfeiting labels lies in the optical anti-counterfeiting features integrated into them, such as microtext, diffraction grating structures, or holographic images. These features are tiny in size, and their optical effect depends on their physical structure. During the lamination process, any slight tensile deformation may cause the size or relative position of these microstructures to change, thereby degrading or even rendering the anti-counterfeiting features ineffective. For example, a change in the spacing between the lines of a diffraction grating may cause the color presented at a specific angle to shift, resulting in the product being mistakenly identified as a counterfeit. Therefore, maintaining constant strain on the substrate film throughout the lamination process is a prerequisite for ensuring the effectiveness of the anti-counterfeiting label.

[0004] However, in actual production, the substrate film used to manufacture anti-counterfeiting labels is not a completely uniform material. Due to the limitations of the manufacturing process of the raw materials or the subsequent anti-counterfeiting feature processing technology, the substrate film will have periodic or non-periodic physical property fluctuations in the length direction. These fluctuations are specifically manifested as slight differences in the local thickness, density or elastic modulus of the material. For example, when producing a batch of substrate film, the upstream casting equipment may cause the film to have thickness fluctuations in the length direction due to slight periodic temperature fluctuations. Alternatively, when a holographic anti-counterfeiting area is formed on the substrate through a hot pressing process, the high temperature and high pressure will cause the crystallization state of the material in this area and its adjacent areas to change, resulting in a difference in its elastic modulus compared to the untreated area.

[0005] These physical property inhomogeneities along the length of the film, imperceptible to the naked eye, present a challenge for tension control. Traditional tension control systems, whether open-loop, closed-loop, or simple feedback control, typically aim to maintain a constant tension value—that is, a constant tensile force applied to the film. However, under constant tension, the elongation of a material is inversely proportional to its elastic modulus. This means that when a section of film with physical inhomogeneities passes through a tension zone, even with a constant applied tensile force, areas with a lower elastic modulus will experience greater tensile deformation than those with a higher elastic modulus. This uneven deformation can directly affect the security features carried on it, causing overstretching and damage in some areas while potentially under-tensioning others. Feedback control systems based on rear-mounted cameras to detect wrinkles or deformation in finished products can correct the average tension, but they are unable to intervene in the localized, transient over-deformation caused by inherent material inhomogeneities during high-speed operation. By the time deformation is detected, the film section has already been laminated, and the damage has already occurred.

[0006] In order to solve this problem, there is an urgent need for a feedforward control strategy that can predict the changes in the physical properties of the material before it enters the laminating station and adjust the tension in advance. Summary of the Invention

[0007] The purpose of the present invention is to provide a method and system for controlling the lamination tension of an anti-counterfeiting label. Through precise local strain measurement and forward-looking tension adjustment, it is ensured that each section of the film maintains an ideal strain state at the moment of lamination, thereby thoroughly protecting the integrity and functional effectiveness of optical microstructures such as diffraction gratings, and effectively solving the technical problems faced in the production of anti-counterfeiting label lamination.

[0008] In a first aspect, the present invention provides a method for controlling the lamination tension of an anti-counterfeiting label, comprising the following steps: S1 obtains substrate film surface image data at an upstream position of the laminating station; S2. Identify the inherent optical microstructures on the substrate film as feature points from the substrate film surface image data, and obtain the real-time coordinates of the feature points along the direction of movement of the substrate film by tracking the feature points; S3. Calculate the local strain of the substrate film based on the real-time coordinates of the feature points; S4. Calculating a strain deviation based on the local strain and a preset target strain; S5. According to the conveying speed of the substrate film and the distance between the substrate film surface image data acquisition position and the laminating station, the time required for the film segment corresponding to the strain deviation to reach the laminating station is calculated as the adjustment time; S6. Calculate the tension compensation amount based on the strain deviation and the relationship between the preset strain deviation and the tension adjustment amount; S7. During the adjustment time, the tension compensation amount is sent to the tension control unit so that the tension control unit adjusts the tension applied to the substrate film according to the tension compensation amount when the corresponding film segment enters the laminating station.

[0009] The present invention provides a method for controlling lamination tension on anti-counterfeit labels. This method utilizes the inherent optical microstructures of the label's substrate film as visual feature points. A visual system captures the relative positional changes of these feature points during high-speed film motion in real time, thereby calculating the film's local strain. This local strain data is used as a feedforward signal to predict the strain state of the film segment about to enter the lamination station. This information is then used to preemptively adjust the output of the tension control unit, ensuring uniform strain across the film at the moment of lamination and protecting the integrity of the optical microstructures.

[0010] In a second aspect, the present invention provides an anti-counterfeiting label lamination tension control system, comprising: A first acquisition module is used to acquire surface image data of the substrate film at an upstream position of the laminating station; The second acquisition module is used to identify the inherent optical microstructures on the substrate film as feature points from the surface image data of the substrate film, and obtain the real-time coordinates of the feature points along the movement direction of the substrate film by tracking the feature points; A calculation module is used to calculate the local strain of the substrate film based on the real-time coordinates of the feature points; A first calculation module is used to calculate the strain deviation based on the local strain and the preset target strain; The second calculation module is used to calculate the time required for the film segment corresponding to the strain deviation to reach the laminating station according to the conveying speed of the substrate film and the distance between the position where the substrate film surface image data is obtained and the laminating station as the adjustment time; a third calculation module, configured to calculate a tension compensation amount based on the strain deviation and a preset relationship between the strain deviation and the tension adjustment amount; The control module is used to send the tension compensation amount to the tension control unit within the adjustment time, so that the tension control unit adjusts the tension applied to the substrate film according to the tension compensation amount when the corresponding film section enters the laminating station.

[0011] From the above, it can be seen that the anti-counterfeiting label lamination tension control method provided by the present invention, in the high-speed roll-to-roll production process of anti-counterfeiting labels integrated with precise optical microstructures such as diffraction gratings or holograms, in view of the slight fluctuations in physical properties such as local elastic modulus or thickness caused by the manufacturing or processing technology of the substrate film itself, and the inability of the existing constant tension control system to ensure strain uniformity, and the feedback control system based on finished product detection has a response lag, under the premise of not damaging the optical microstructure, the local physical property change area on the film that is about to cause uneven tensile strain is identified in real time and accurately, and the lamination tension is adjusted dynamically and proactively accordingly to ensure that the strain of the film at the moment of lamination always remains uniform, thereby fundamentally protecting the integrity and functional effectiveness of the optical microstructure in the anti-counterfeiting label.

[0012] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the embodiments of the present invention. The purposes and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A flow chart of a method for controlling lamination tension of an anti-counterfeiting label provided in an embodiment of the present invention.

[0014] Figure 2 A schematic structural diagram of the anti-counterfeiting label lamination tension control system provided by an embodiment of the present invention.

[0015] Description of labels: 100, first acquisition module; 200, second acquisition module; 300, solution module; 400, first calculation module; 500, second calculation module; 600, third calculation module; 700, control module. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present invention.

[0017] Actual research has found that these microscopic changes in the physical properties of substrate films, while difficult to discern with the naked eye, are accompanied by changes in optical characteristics that can be captured by visual systems. For example, slight fluctuations in thickness can cause corresponding periodic changes in the transmittance or reflectance of the film under specific lighting conditions. Furthermore, areas where the elastic modulus changes due to the hot pressing process can have surface microtexture or responses to polarized light that differ from surrounding areas. These subtle changes in optical characteristics serve as a "fingerprint" for predicting the material's physical properties.

[0018] Therefore, a line scan camera and lighting system can be installed upstream of the lamination station, before the primary film tension application area. This vision system continuously scans the passing substrate film, capturing image data from its surface. By analyzing the image data stream in real time, the system can identify specific optical "fingerprints" associated with changes in the material's physical properties. For example, Fourier transform analysis of the image spectrum can detect the frequency of light intensity variations caused by periodic fluctuations in film thickness. Or, texture analysis algorithms can identify surface texture patterns corresponding to areas of varying elastic modulus. Once visual features indicating impending material changes are detected, the system calculates the required tension compensation based on pre-established relationships. This compensation command is fed forward to the downstream tension control unit. For example, if a region indicating a low elastic modulus is detected approaching the lamination station, the system instructs the tension controller to momentarily reduce the tension applied to that section of film, ensuring that the actual strain at the moment of lamination is consistent with that in other areas. In this way, the control system no longer passively responds to deformation that has already occurred, but actively foresees and eliminates the root cause of the deformation, thereby fundamentally ensuring the integrity of the anti-counterfeiting feature during the lamination process.

[0019] Based on this, refer to the attached Figure 1 The present invention provides a method for controlling the lamination tension of an anti-counterfeiting label, comprising the following steps: S1 obtains substrate film surface image data at an upstream position of the laminating station; S2. Identify the inherent optical microstructures on the substrate film as feature points from the substrate film surface image data, and obtain the real-time coordinates of the feature points along the direction of movement of the substrate film by tracking the feature points; S3. Calculate the local strain of the substrate film based on the real-time coordinates of the feature points; S4. Calculating a strain deviation based on the local strain and a preset target strain; S5. According to the conveying speed of the substrate film and the distance between the substrate film surface image data acquisition position and the laminating station, the time required for the film segment corresponding to the strain deviation to reach the laminating station is calculated as the adjustment time; S6. Calculate the tension compensation amount based on the strain deviation and the relationship between the preset strain deviation and the tension adjustment amount; S7. During the adjustment time (i.e., before the film segment corresponding to the strain deviation reaches the laminating station), the tension compensation amount is sent to the tension control unit so that the tension control unit adjusts the tension applied to the substrate film according to the tension compensation amount when the corresponding film segment enters the laminating station.

[0020] Intrinsic optical microstructure refers to microscopic optical features formed during the manufacturing or processing of the substrate film itself, which are associated with changes in physical properties. These features can be identified using image processing techniques, such as by analyzing the film's surface microtexture, local variations in transmittance or reflectance, or differences in response to polarized light. They serve as a "fingerprint" reflecting the film's local physical state, providing traceable reference points for subsequent strain calculations. Feature points are optical microstructures identified from surface image data that represent local motion and deformation of the film. These features can be extracted using image recognition, pattern matching, or feature extraction algorithms, such as those based on SIFT, SURF, or ORB. They provide precise displacement information for quantifying local deformation of the film. Local strain refers to the relative degree of deformation within a specific local region of the substrate film. It can be calculated using methods such as digital image correlation (DIC), geometric deformation calculation based on feature point tracking, or finite element analysis. It quantifies the actual deformation of the film due to nonuniform physical properties, serving as a basis for tension adjustment. Strain deviation refers to the difference between the actual local strain of the substrate film and the preset target strain. It can be calculated using numerical subtraction or weighted averaging algorithms. Its primary purpose is to clarify the gap between the deformation state of the current film segment and the ideal state, providing quantitative input for calculating the tension compensation. Adjustment time refers to the time required for the film segment corresponding to the strain deviation to reach the laminating station from the image data acquisition location. It can be calculated by dividing the distance by the speed or dynamically calculated using the real-time speed sensor data of the conveyor system. It is primarily used to achieve time synchronization for feedforward control and ensure that the tension adjustment command takes effect at the appropriate time. Tension compensation refers to the adjustment required to the tension applied to the substrate film to eliminate strain deviation. It can be calculated using a proportional-integral-derivative (PID) control algorithm, a lookup table, or a machine learning model. Its primary purpose is to convert strain deviation into specific tension adjustment commands to achieve the target strain during lamination.

[0021] The operating principle of this invention is to utilize the inherent optical microstructures of the security label substrate film as natural visual features that require no additional marking. By deploying a visual acquisition unit upstream of the laminating station, the relative displacement of these microstructural features during film motion is captured in real time and at high speed. This relative displacement directly reflects the instantaneous tensile deformation (i.e., strain) of a localized region of the film. The system compares the real-time calculated local strain with a preset ideal strain to determine the strain deviation. Combining this with the film's conveying speed and physical distance, the system accurately predicts when the deviated section of film will reach the laminating station. Based on this, the system calculates the required tension compensation according to a preset strain-tension adjustment relationship and, in the form of a "feedforward" command, preemptively adjusts the tension output of the tension control unit before the corresponding film section enters the laminating station. This predictive and proactive adjustment mechanism ensures that the actual strain of the film remains uniform at the moment of lamination, fundamentally protecting the integrity of the precise optical microstructure integrated into the security label and preventing deformation damage caused by material inhomogeneities.

[0022] The core innovation of this application lies in obtaining surface image data of the substrate film at a position upstream of the laminating station, and identifying and tracking the inherent optical microstructures on the film as feature points from the image, thereby solving the local strain of the film, thereby being able to proactively calculate the required tension compensation amount and adjust the tension applied to the film in advance before the film segment reaches the laminating station, effectively solving the problem of inconsistent local strain caused by uneven physical properties of the substrate film, and ensuring the integrity of the anti-counterfeiting features during the laminating process.

[0023] Specifically, this method acquires surface image data of the substrate film upstream of the laminating station, capturing its surface optical features before the film enters the laminating station. The system then identifies inherent optical microstructures on the film as feature points from this image data, and continuously tracks the real-time coordinates of these feature points along the film's motion direction. These optical microstructures serve as "fingerprints" of local variations in the film's physical properties, and changes in their motion coordinates directly reflect the film's microscopic deformation. Based on the real-time coordinates of these feature points, the system calculates the local strain of the substrate film, quantifying the degree of stretch or compression in different regions of the film. The calculated local strain is compared with a preset target strain to calculate the strain deviation, clarifying the difference between the actual deformation of the current film segment and the desired deformation. To achieve precise feedforward control, the system calculates the time required for the film segment corresponding to the strain deviation to reach the laminating station based on the substrate film's conveying speed and the distance between the image data acquisition location and the laminating station. This time is used as the adjustment time. Furthermore, the specific tension compensation amount is calculated based on the strain deviation and the pre-established relationship between the strain deviation and the tension adjustment amount. Finally, within the calculated adjustment time—that is, before the film segment corresponding to the strain deviation reaches the laminating station—this tension compensation value is transmitted to the tension control unit. Upon receiving this instruction, the tension control unit adjusts the tension applied to the substrate film based on the received tension compensation value at the precise moment the corresponding film segment enters the laminating station. Through this series of steps, the system achieves real-time perception, quantification, prediction, and proactive adjustment of local film strain, ensuring that the film strain at the moment of lamination remains within the target range, effectively preventing damage to the anti-counterfeiting feature due to material inhomogeneity.

[0024] In some embodiments, the specific steps in step S2 include: S21. Performing image preprocessing on the substrate film surface image data; image preprocessing includes: enhancing the contrast of the optical microstructure and suppressing reflection interference by adaptively adjusting the substrate film surface image data according to the local brightness distribution and reflection characteristics of the substrate film surface image data; S22. Extracting candidate feature points corresponding to the optical microstructures from the preprocessed substrate film surface image data based on the geometric morphology and optical response characteristics of the optical microstructures; S23. According to the preset optical microstructure template or its spatial distribution pattern, by matching and screening the candidate feature points, determining the effective feature points corresponding to the inherent optical microstructure on the substrate film; S24. According to the position change of the effective feature point in the continuous image frames and in combination with the conveying speed of the substrate film, the effective feature point is tracked to obtain the real-time coordinates of the effective feature point along the moving direction of the substrate film.

[0025] Optical response characteristics refer to the unique optical behaviors exhibited by these microstructures under specific lighting conditions, such as light absorption, reflection, transmission, diffraction, or polarization, resulting in specific brightness, color, or texture patterns in the image. These characteristics are the basis for identifying microstructures. A preset optical microstructure template refers to a pre-established standard image pattern or feature descriptor representing a real optical microstructure. It can be a pixel template, a set of feature vectors, or a mathematical model, and is used to compare similarity with candidate feature points extracted from an image. Spatial distribution patterns refer to the specific macroscopic arrangement patterns of the inherent optical microstructures on the substrate film. These patterns can be, for example, a periodic array, a random distribution with a specific density, or a linear arrangement along a specific direction. These patterns can be used to verify the authenticity of candidate feature points, for example, by calculating the distance, angle, or topological relationship between candidate points to determine whether they conform to the preset pattern. The position change of valid feature points in consecutive image frames refers to the changes in the pixel coordinates corresponding to optical microstructures identified as real and stable in the time-series image data stream over time. These changes reflect the actual movement of the substrate film. Combining the conveying speed of the substrate film means introducing the known film production line speed information into the feature point tracking algorithm, for example, to predict the approximate position of the feature point in the next frame, or to correct the tracking results, thereby improving the accuracy and robustness of tracking, especially in high-speed production environments.

[0026] This solution utilizes a series of sophisticated image processing and feature point screening steps that work together to ensure accurate and stable identification and tracking of the inherent optical microstructures on substrate films from complex image data. First, the raw substrate film surface image data acquired by the visual acquisition unit undergoes image preprocessing. This process adaptively adjusts the image's local brightness distribution and reflective properties to enhance the contrast of the optical microstructures and effectively suppress interference such as specular reflections. This fundamental and critical step transforms the original low-signal-to-noise ratio image into a clear, feature-rich image, laying a high-quality data foundation for subsequent accurate identification. Based on this preprocessed image data, the system further extracts potential candidate feature points based on the inherent geometric morphology and optical response characteristics of the optical microstructures. This process leverages prior knowledge of the microstructure, avoids blind searches, and improves the efficiency and accuracy of initial identification. Subsequently, to ensure the authenticity and stability of the identified feature points, this solution introduces a rigorous matching and screening mechanism. By comparing these candidate feature points against a pre-established standard template representing a real optical microstructure and verifying their spatial distribution, false or unstable feature points can be effectively eliminated, ultimately identifying highly reliable valid feature points. This screening process is crucial for ensuring subsequent tracking accuracy. Finally, based on these rigorously screened valid feature points, the system dynamically tracks them by analyzing their positional changes in successive image frames and, combined with the known conveyor speed of the substrate film, obtaining their precise coordinates along the film's motion in real time. Incorporating conveyor speed optimizes the tracking algorithm and improves robustness in high-speed production environments. Overall, this series of steps forms a complete chain from coarse to fine, from preprocessing to precise tracking. Preprocessing addresses image quality issues, feature extraction and screening ensure the authenticity and reliability of feature points, and precise tracking provides real-time, accurate motion data. Through this synergistic operation, this solution overcomes the challenges of traditional methods in feature point identification and tracking under complex lighting conditions and unique optical effects, significantly improving the accuracy and stability of feature point data. These high-precision, high-stability real-time coordinates of feature points serve as key input for the subsequent calculation of local strain in the substrate film, directly improving the accuracy of local strain calculations. This in turn enables the entire anti-counterfeiting label lamination tension control method to more accurately and timely adjust the tension applied to the substrate film, thereby effectively avoiding damage to anti-counterfeiting features due to material inhomogeneity and ensuring the quality and functional integrity of the anti-counterfeiting label.

[0027] In some embodiments, the step of enhancing the contrast of the optical microstructure and suppressing reflection interference by adaptively adjusting the substrate film surface image data based on the local brightness distribution and reflection characteristics of the substrate film surface image data includes: S211. Obtaining grayscale information reflecting the local brightness distribution and polarization information reflecting the reflection characteristics from the substrate film surface image data; S212. According to the grayscale information, the contrast of the optical microstructure is enhanced by performing local contrast enhancement processing on the image data; S213. Separate the specular reflection component and the diffuse reflection component in the substrate film surface image data according to the polarization information, and suppress the reflection interference caused by the specular reflection by enhancing the diffuse reflection component.

[0028] This application refines the image preprocessing steps. Through step-by-step, targeted processing, it effectively improves the contrast of optical microstructures and suppresses reflection interference, thus laying the foundation for subsequent accurate identification and tracking of feature points. Specifically, first, grayscale information reflecting the local brightness distribution and polarization information reflecting the reflection characteristics are obtained from the image data of the substrate film surface. This acquisition process is the basis for subsequent refined image processing. By obtaining grayscale information and polarization information separately from the original image data, the system can decompose the image processing task into issues related to brightness and contrast, and issues related to reflection interference. Grayscale information directly reflects the local brightness distribution of the image and is the direct basis for contrast enhancement. Polarization information can reveal the characteristics of the interaction between light and the film surface, especially distinguishing between specular reflection and diffuse reflection, which is crucial for effectively suppressing reflection interference. This method of separately obtaining different types of information provides data support for subsequent targeted processing and avoids the limitations that may be brought about by a single processing method. On this basis, based on the acquired grayscale information, the contrast of the optical microstructure is enhanced by performing local contrast enhancement processing on the image data. Optical microstructures on substrate films are often tiny and have subtle differences in brightness from the surrounding background, resulting in low contrast and difficulty in accurate recognition. By utilizing grayscale information for local contrast enhancement, it is possible to selectively enhance brightness differences in specific areas of the image, making them more visually prominent. This local enhancement effectively highlights subtle optical features, providing clearer and easier-to-process image input for subsequent feature point recognition, thereby improving recognition accuracy. Simultaneously, based on acquired polarization information, the system separates the specular and diffuse reflection components of the substrate film surface image data. Enhancement of the diffuse reflection component suppresses reflective interference caused by specular reflection. In actual production environments, specular reflections from light sources on film surfaces are often very strong, obscuring or blurring the underlying optical microstructures and severely impacting image quality. Polarization information distinguishes light with different polarization states. Specular reflections typically have a higher degree of polarization, while diffuse reflections have a lower degree of polarization. By analyzing polarization information, the system can decompose image data into specular and diffuse reflection components. Specular reflection is the primary source of interference. By separating and suppressing it, we can effectively eliminate the obstruction of optical microstructures by highlight areas. Furthermore, optical microstructures often exhibit their characteristics through diffuse reflection or diffraction. Therefore, enhancing the diffuse reflection component can further highlight these microstructures, making them more clearly visible in the image after removing interference. This refined processing based on polarization information can fundamentally suppress reflective interference, ensuring image purity and feature visibility, and providing a high-quality image foundation for subsequent feature point extraction and tracking.Through the synergistic effect of the above steps, the image preprocessing solution of this application is no longer a general adaptive adjustment, but a divide-and-conquer refined processing of the brightness and reflective characteristics of the image. Grayscale information processing focuses on improving the visibility of the microstructure itself, while polarization information processing focuses on eliminating external interference. This two-pronged strategy can effectively enhance the contrast of optical microstructures and suppress reflective interference even in complex environments, thereby significantly improving the accuracy of subsequent feature point recognition and tracking, thereby ensuring the accuracy and robustness of the entire anti-counterfeiting label lamination tension control method.

[0029] In some embodiments, the specific steps in step S22 include: S221. Performing local area analysis on the pre-processed substrate film surface image data to obtain local brightness, contrast or texture information reflecting the optical response characteristics of the optical microstructure; S222. Based on local brightness, contrast, or texture information, and in combination with preset geometric features of the optical microstructure, identify the boundaries, corners, or specific shape areas of the optical microstructure by performing structural element recognition on the pre-processed substrate film surface image data; S223. Based on the identified boundaries, corner points or specific shape areas, and combined with their optical response characteristics, the image areas in the pre-processed substrate film surface image data are screened and aggregated to determine the candidate feature points corresponding to the optical microstructure.

[0030] Local region analysis refers to processing a small set of pixels within image data to reveal information specific to that region. This can be achieved using sliding window techniques, patch-based feature extraction algorithms, or multi-scale analysis methods. Optical response features refer to observable properties of an optical microstructure under specific lighting conditions, such as the reflection, transmission, absorption, or scattering of light on its surface or within its interior. These can manifest as information such as local brightness, contrast, color, polarization state, or diffraction pattern. Predefined geometric morphological features of optical microstructures refer to predefined recognition patterns or templates based on known or expected geometric properties of the optical microstructure, such as its shape, size, and arrangement. These can include the microstructure's boundary outline, internal corners, center points, and the area or aspect ratio of a specific region. Structural element recognition refers to identifying pixel sets with specific geometric shapes or topological structures from image data using image processing algorithms. This can be achieved using morphological operations (such as erosion, dilation, opening, and closing), edge detection algorithms (such as Canny and Sobel), corner detection algorithms (such as Harris and Shi-Tomasi), or template matching algorithms. Screening and aggregation refer to the process of further evaluation, filtering and integration of the initially identified image areas or feature points. Screening can be based on a set threshold, confidence level or matching degree to exclude areas that do not meet the conditions; aggregation can be the merging of scattered feature points or areas belonging to the same microstructure into a whole, for example, through clustering algorithms or connected domain analysis.

[0031] Through a series of refined steps, this solution accurately extracts candidate feature points corresponding to optical microstructures on substrate films, providing high-quality input for subsequent feature point tracking and local strain calculation. First, by performing localized analysis on preprocessed substrate film surface image data, the system can obtain local brightness, contrast, or texture information that reflects the optical response characteristics of the optical microstructure. This step recognizes that the characteristics of optical microstructures are often reflected in the subtle optical response of local regions. By focusing on local analysis, these critical features can be effectively captured, avoiding interference from global information or background, laying the foundation for subsequent recognition. Based on this localized optical response information and combined with pre-defined geometric features of the optical microstructure, the system performs structural element recognition on the image data to identify boundaries, corners, or specific shaped regions of the optical microstructure. This step combines the optical response features with pre-defined geometric features. The localized optical response information indicates potential microstructure regions, while the pre-defined geometric features provide a template for identifying these microstructures. Through structural element recognition, image regions that meet dual criteria can be precisely located and identified, effectively extracting structural features related to optical microstructures from complex image data, and improving recognition accuracy. Finally, the system screens and aggregates the image regions based on the identified boundaries, corners, or specific shapes, and further combines their optical response characteristics to ultimately determine candidate feature points corresponding to the optical microstructure. This step, based on the identification of structural elements, further incorporates optical response characteristics for verification and optimization. By further screening and aggregating the identified structural regions, it ensures that the final candidate feature points not only meet the geometric requirements but also conform to the actual optical microstructure in terms of optical response. This multi-verification mechanism effectively eliminates misidentified areas and aggregates scattered feature information, thereby determining candidate feature points corresponding to the optical microstructure with higher accuracy and reliability. Overall, this solution decomposes the optical microstructure feature point extraction process into the acquisition of local optical response information, the identification of structural elements combining optical response and geometric morphology, and the screening and aggregation based on dual features, forming a step-by-step, mutually verified recognition process. This multi-dimensional, multi-stage feature extraction strategy significantly enhances the recognition of small and weak optical features and overcomes the influence of image noise and background interference. Compared to extraction methods that rely solely on general geometric morphology or optical response, this solution can provide a set of candidate feature points with higher accuracy, completeness, and robustness. These high-quality candidate feature points, as input to subsequent steps, can significantly improve the accuracy of valid feature point determination and the stability of feature point tracking, thereby directly improving the accuracy and reliability of local strain calculations of the substrate film. Ultimately, this improves the overall performance of lamination tension control for anti-counterfeiting labels and ensures the integrity of anti-counterfeiting features during the production process.

[0032] In some embodiments, the specific steps in step S23 include: S231. Obtain the local image area corresponding to the candidate feature point; S232. Based on the preset optical microstructure template, similarity calculation is performed on the local image area to obtain the matching degree between the candidate feature points and the template; S233. Determine whether the candidate feature points conform to a preset optical microstructure spatial distribution pattern based on the spatial relative positions between the candidate feature points; S234. Based on the judgment results of the matching degree and the spatial distribution law, the candidate feature points are screened to determine the effective feature points corresponding to the inherent optical microstructure on the substrate film.

[0033] A local image region refers to a set of pixels of a certain size surrounding the center of each candidate feature point. This can be captured using a fixed-size rectangular or circular window or an area adaptively resized based on image content. A preset optical microstructure template refers to a pre-established image pattern or mathematical model that represents the standard morphology and optical response characteristics of a real optical microstructure. This template can be constructed using image blocks of typical microstructures extracted from a standard sample image or a synthetic image generated based on the optical microstructure design parameters. Spatial relative position refers to the relative positional relationship of candidate feature points in the image coordinate system, such as their distance, angle, or arrangement order. This can be described using Euclidean distance, angular difference, or topological relationships. A preset spatial distribution pattern of optical microstructures refers to the specific geometric pattern or periodic characteristics of the optical microstructures on the substrate film surface. This pattern can be defined as a periodic array, a regular grid, a specific pattern, or a random but statistically consistent distribution pattern. Valid feature points are those that have been rigorously verified to be authentic, reliable, and representative of the inherent optical microstructure of the substrate film, serving as the basis for subsequent precise measurement and control.

[0034] This solution details how to accurately and robustly select valid feature points that truly represent the inherent optical microstructure of a substrate film from a preliminary extraction of candidate feature points, thereby addressing the problem of subsequent measurement inaccuracies caused by impure candidate points. First, by acquiring the local image regions corresponding to the candidate feature points, the necessary data foundation is provided for subsequent refined analysis and verification. This is because the identification and verification of optical microstructures often require detailed analysis of the surrounding local texture, shape, and brightness information, not just the isolated point itself. Local image regions provide more comprehensive contextual information. Based on this, similarity calculations are performed on these local image regions against a preset optical microstructure template to determine the match between the candidate feature points and the template. This process incorporates a direct comparison mechanism based on known, standard optical microstructure features. By comparing the local image of the candidate point with the pre-set standard template, the degree of similarity in morphology and optical response can be effectively quantified, thereby preliminarily eliminating candidate points that appear to be inconsistent with the actual optical microstructure and improving recognition accuracy. A higher match indicates a more likely candidate point to be a genuine and reliable optical microstructure. At the same time, candidate feature points are verified from another important perspective: their relative spatial positions are used to determine whether they conform to the predetermined spatial distribution pattern of the optical microstructure. The optical microstructures on many anti-counterfeiting labels are not randomly distributed, but rather arranged in a specific array, pattern, or periodic pattern. By examining the relative positional relationships between candidate points, isolated, randomly occurring noise points or non-structural features can be further eliminated, thereby enhancing the robustness of valid feature point identification. For example, if the predetermined optical microstructure is periodically arranged, only candidate points that conform to this periodic distribution are considered valid. This helps filter out false features that may resemble the original locally but not conform to the overall structural pattern. Ultimately, by combining the results of the matching degree and the spatial distribution pattern, candidate feature points are screened and identified as valid feature points corresponding to the inherent optical microstructure of the substrate film. This dual verification mechanism—which considers both the local morphology and optical properties of individual feature points (via template matching) and the macroscopic spatial relationships between feature points (via distribution pattern determination)—significantly improves the accuracy and reliability of valid feature point identification. Through this rigorous screening process, it can be ensured that subsequent feature point tracking and local strain solution are based on high-quality, high-confidence feature points, thereby providing a solid data foundation for precise tension control, effectively avoiding tension adjustment deviations caused by inaccurate feature point identification, and ultimately ensuring the quality of the anti-counterfeiting label lamination and the integrity of the anti-counterfeiting features.This refinement and improvement in the robustness of feature point recognition enables the entire anti-counterfeiting label lamination tension control method to more accurately sense the local strain of the film, thereby achieving more precise tension feedforward adjustment, thereby fundamentally solving the problem of anti-counterfeiting feature damage caused by material unevenness.

[0035] In some embodiments, the specific steps in step S3 include: S31. According to the real-time coordinates of the effective feature points along the direction of movement of the substrate film, a plurality of adjacent effective feature points are determined, and a local feature point group is constructed based on the spatial relative position relationship between the plurality of adjacent effective feature points; S32. Calculate the strain tensor of the area covered by the local feature point group based on the relative displacement change of the effective feature points in the continuous image frames and the preset mechanical parameters of the substrate film material; S33. Extract the principal strain component along the movement direction of the substrate film from the strain tensor and use it as the local strain of the substrate film.

[0036] A local feature point group is a collection of valid feature points that are spatially close to each other and have a specific relative positional relationship. It can be defined using geometric configurations such as triangles, quadrilaterals, or arbitrary polygons to represent a local area on the substrate film. Pre-set mechanical parameters of the substrate film material are physical quantities obtained before the lamination process or through experimental measurement to describe the deformation response of the substrate film material under load. These parameters may include elastic modulus, Poisson's ratio, shear modulus, and other parameters. These parameters can be pre-stored in the control system. The strain tensor is a second-order tensor that comprehensively describes the deformation state of a specific point or region within a material under load. It can contain a normal strain component (tensile or compressive) and a shear strain component, reflecting the material's deformation in different directions and angular changes. The principal strain component refers to the maximum or minimum strain value along a specific direction within the deformation state described by the strain tensor. This corresponds to tensile or compressive deformation of the material along its principal axis.

[0037] This solution, in its method for controlling the lamination tension of substrate films, specifically refines the local strain calculation process, aiming to more accurately and comprehensively capture the microscopic deformations of the substrate film during the lamination process. First, after obtaining the real-time coordinates of valid feature points along the substrate film's motion direction, the system uses these real-time coordinates to identify and determine multiple spatially proximal valid feature points. Based on the spatial relative positional relationships between these adjacent points, a local feature point group is constructed to represent a specific local area. This construction process transcends the traditional limitation of focusing solely on the displacement of isolated feature points, instead considering multiple points within a region as a whole, laying the foundation for subsequent regional deformation analysis. Furthermore, the system further analyzes the relative displacement changes of each valid feature point in the local feature point group across successive image frames. These relative displacement changes directly reflect the deformation of the local area over a very short period of time. Furthermore, combined with pre-defined mechanical parameters of the substrate film material, the system can convert the observed geometric displacements into physical strains. By calculating the strain tensor for the area covered by the local feature point group, information on the normal and shear strains in all directions within that area can be obtained. This provides a comprehensive mathematical description of the complex deformation pattern, overcoming the information deficit that can arise from calculating strain in only a single direction. Ultimately, while the strain tensor provides comprehensive deformation information, given that the key deformation that most directly affects the size and spacing of security features during roll-to-roll lamination is typically tension or compression along the film's transport direction, the system accurately extracts the principal strain components along the substrate film's motion from the calculated strain tensor and uses these as the local strains of the substrate film. This focus on extracting these principal strain components enables precise and targeted tension control, directly corresponding to the degree of deformation of the security feature in these critical directions. This logical progression from isolated point tracking to regional feature point group construction, from single-direction strain to comprehensive strain tensor calculation, and finally to the extraction of key principal strain components enables this solution to more accurately and comprehensively reflect the actual deformation state of the substrate film. This precise local strain calculation, combined with the identification and tracking of optical microstructure feature points in upstream steps, provides highly accurate input for downstream tension adjustment. This approach effectively addresses the complex physical property fluctuations and multimodal deformations of substrate film materials, ensuring precise tension adjustment and effectively protecting the integrity of the optical microstructure, improving the production quality and reliability of anti-counterfeiting labels.

[0038] In some embodiments, the specific steps in step S32 include: S321. Obtaining the real-time coordinates of each valid feature point in the local feature point group in the continuous image frames, and calculating the displacement vector of each valid feature point between the continuous image frames; S322. Constructing a deformation gradient tensor describing the deformation of the local area based on the displacement vector of each valid feature point and its initial spatial position in the local feature point group; S323. Based on the deformation gradient tensor and in combination with the preset mechanical parameters of the substrate film material, the strain tensor of the area covered by the local feature point group is solved by considering the anisotropic or nonlinear deformation characteristics of the material.

[0039] The deformation gradient tensor is a mathematical tool used in continuum mechanics to comprehensively describe the local deformation of an object. It can simultaneously reflect multiple deformation modes of a material, such as tension, shear, and rotation. It can be constructed using shape function interpolation based on the finite element method or by fitting the displacement data of discrete characteristic points using the least squares method. Anisotropic or nonlinear deformation properties refer to materials exhibiting different mechanical responses in different directions, or their stress-strain relationship is not a simple linear proportional relationship. This can be described using anisotropic elastic constitutive models (such as the orthotropic model) or nonlinear elastic / elastoplastic constitutive models (such as the Mooney-Rivlin model and the Ogden model). The strain tensor is a physical quantity that describes the relative displacement and degree of deformation at each point within an object. It can comprehensively reflect the tensile, compressive, and shear deformation of a material under load. It can be represented using the Green-Lagrange strain tensor, the Euler-Almansi strain tensor, or the engineering strain tensor.

[0040] This solution aims to improve the accuracy of identifying the local deformation state of a substrate film by refining the strain tensor calculation process. This ensures accurate strain calculation, particularly when dealing with material inhomogeneities, anisotropy, or nonlinear deformation properties, thereby providing a reliable basis for subsequent precise tension adjustment. Specifically, the solution's operational logic begins at step S321. By acquiring the real-time coordinates of each valid feature point in a local feature point group in consecutive image frames and calculating the displacement vectors of each valid feature point between consecutive image frames, this provides the fundamental data for describing the local motion and deformation of the material. This real-time coordinate acquisition and displacement vector calculation based on consecutive image frames can capture subtle, instantaneous deformation information of the substrate film during high-speed motion, laying the data foundation for subsequent precise local strain analysis. Based on this, step S322 constructs a deformation gradient tensor describing the local deformation based on the displacement vectors of each valid feature point and its initial spatial position within the local feature point group. The deformation gradient tensor, a mathematical tool used in continuum mechanics to comprehensively describe the local deformation of an object, can simultaneously reflect multiple deformation modes, including tension, shear, and rotation. By constructing a continuous deformation gradient tensor from discrete feature point displacement data, this solution can transform the microscopic feature point motion into a macroscopic description of local regional deformation, providing a rigorous mathematical model for more accurate strain calculation. Finally, step S323, based on the constructed deformation gradient tensor and combined with the preset mechanical parameters of the substrate film material, calculates the strain tensor for the area covered by the local feature point group by taking into account the anisotropic or nonlinear deformation properties of the material. This organic combination of steps eliminates the simplistic treatment of the material as a homogeneous or linear elastic body. Instead, it uses a more realistic mechanical model based on the actual physical inhomogeneities of the substrate film and the complex mechanical behavior it may exhibit under stress. Precisely because S321 and S322 provide accurate and comprehensive local deformation information, S323 is able to build on this foundation by considering the anisotropic or nonlinear deformation properties of the material, achieving a refined strain tensor solution. This refined solution is closely linked to the aforementioned step S31 (constructing a local feature point group). S31 ensures that strain calculations are performed within a clearly defined and representative local region, while S321-S323 ensure highly accurate and comprehensive deformation analysis within that region. This combination enables the system to accurately assess strain based on localized inhomogeneities in the substrate film, providing a reliable basis for subsequent precise tension adjustments.

[0041] In some embodiments, the specific steps in step S323 include: S3231. Obtain the deformation gradient tensor of the area covered by the local feature point group, and the tension information applied to the area by the laminating station; S3232. Evaluate the deviation between the preset mechanical parameters of the substrate film material and the actual mechanical response of the current film segment based on the deformation gradient tensor, the tension information, and the relationship between the historically calculated strain tensor and the corresponding tension information; S3233. Based on the deviation, the preset mechanical parameters of the substrate film material are corrected; the correction includes adjusting the parameters related to the anisotropic or nonlinear deformation characteristics of the material; S3234. Based on the deformation gradient tensor and combined with the corrected mechanical parameters of the substrate film material, the strain tensor of the area covered by the local feature point group is solved.

[0042] The tensile force applied by the laminating station to the region refers to the external force exerted on the substrate film by the laminating station, causing it to deform. This information can be acquired in real time using tension sensors or force sensors installed before and after the laminating station. The relationship between the historically calculated strain tensor and the corresponding tensile force information refers to a database, lookup table, or mathematical model, established through long-term operation and data accumulation, that describes the strain response patterns of the substrate film under different tensile forces. This information can be obtained by training historical data using machine learning algorithms or establishing empirical formulas through regression analysis. Evaluating the deviation between the preset mechanical parameters of the substrate film and the actual mechanical response of the current film segment involves comparing the real-time deformation gradient tensor and tensile force information of the current film segment with the response predicted by historical data or models to quantify the difference between the preset material parameters and the actual situation. This can be accomplished through error analysis, residual calculation, or statistical methods. Correction refers to adjusting the preset material mechanical parameters based on the assessed deviation to align them with the actual mechanical properties of the current film segment. This can be achieved through iterative optimization algorithms, adaptive control algorithms, or parameter correction methods based on lookup tables. Adjustment of parameters related to the anisotropic or nonlinear deformation characteristics of materials refers to the modification of parameters that describe the differences in the mechanical responses of the material in different directions or the nonlinear behavior exhibited when the deformation degree reaches a certain level. It may include adjustment of the elastic modulus, shear modulus, Poisson's ratio of the orthotropic materials, or the yield criterion parameters and hardening law parameters of the elastoplastic materials.

[0043] This solution addresses the problem of insufficient strain calculation accuracy, which arises from the inability of preset parameters to accurately reflect the actual mechanical response when the physical properties of the substrate film are non-uniform, by introducing a dynamic evaluation and correction mechanism for the mechanical parameters of the substrate film. Specifically, by obtaining the deformation gradient tensor for the area covered by the local feature point group and the tensile force applied by the laminating station to that area, real-time deformation and force data are provided for subsequent material mechanical parameter evaluation. This data is fundamental to understanding the current mechanical state of the film. Secondly, based on this real-time data and the relationship between the system's long-term accumulated historical strain tensor and corresponding tensile force information, the deviation between the preset mechanical parameters of the substrate film and the actual mechanical response of the current film segment is assessed. This evaluation mechanism not only relies on instantaneous measurements but also incorporates the historical behavior of the film material under different operating conditions, enabling accurate and comprehensive assessment of parameter deviations. Based on this estimated deviation, the preset mechanical parameters of the substrate film are corrected, with particular attention paid to adjusting parameters related to anisotropic or nonlinear deformation characteristics of the material. This dynamic correction allows the material parameters used for strain calculation to adapt in real time to changes in the film's local physical properties, thus overcoming the limitations of traditional fixed-parameter models in dealing with material inhomogeneities. Ultimately, the strain tensor for the area covered by the local feature point group is calculated based on the acquired deformation gradient tensor and combined with the corrected mechanical parameters of the substrate film material. By using correction parameters that closely reflect the actual physical properties of the film segment, the strain tensor calculation ensures that the resulting strain tensor truly reflects the actual strain state of the film segment before the lamination station.

[0044] Reference Attachment Figure 2 The present invention provides an anti-counterfeiting label laminating tension control system, comprising: The first acquisition module 100 is used to acquire surface image data of the substrate film at an upstream position of the laminating station; The second acquisition module 200 is used to identify the inherent optical microstructures on the substrate film as feature points from the substrate film surface image data, and obtain the real-time coordinates of the feature points along the movement direction of the substrate film by tracking the feature points; A calculation module 300 is used to calculate the local strain of the substrate film according to the real-time coordinates of the feature points; A first calculation module 400 is configured to calculate a strain deviation based on a local strain and a preset target strain; The second calculation module 500 is used to calculate the time required for the film segment corresponding to the strain deviation to reach the laminating station based on the conveying speed of the substrate film and the distance between the position where the substrate film surface image data is obtained and the laminating station, as the adjustment time; A third calculation module 600 is configured to calculate a tension compensation amount based on the strain deviation and a preset relationship between the strain deviation and the tension adjustment amount; The control module 700 is used to send the tension compensation amount to the tension control unit within the adjustment time, so that the tension control unit adjusts the tension applied to the substrate film according to the tension compensation amount when the corresponding film section enters the laminating station.

[0045] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any actual relationship or order between these entities or operations.

[0046] The foregoing description is merely an embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for controlling the lamination tension of an anti-counterfeiting label, characterized in that: The following steps are involved: S1 obtains substrate film surface image data at an upstream position of the laminating station; S2. Identify the inherent optical microstructures on the substrate film as feature points from the substrate film surface image data, and obtain the real-time coordinates of the feature points along the direction of movement of the substrate film by tracking the feature points; S3. Calculate the local strain of the substrate film based on the real-time coordinates of the feature points; S4. Calculating a strain deviation based on the local strain and a preset target strain; S5. According to the conveying speed of the substrate film and the distance between the substrate film surface image data acquisition position and the laminating station, the time required for the film segment corresponding to the strain deviation to reach the laminating station is calculated as the adjustment time; S6. Calculate the tension compensation amount based on the strain deviation and the relationship between the preset strain deviation and the tension adjustment amount; S7. During the adjustment time, the tension compensation amount is sent to the tension control unit so that the tension control unit adjusts the tension applied to the substrate film according to the tension compensation amount when the corresponding film segment enters the laminating station.

2. The method for controlling the lamination tension of an anti-counterfeiting label according to claim 1, characterized in that: The specific steps in step S2 include: S21. Perform image preprocessing on the substrate film surface image data; S22. Extracting candidate feature points corresponding to the optical microstructures from the preprocessed substrate film surface image data based on the geometric morphology and optical response characteristics of the optical microstructures; S23. According to the preset optical microstructure template or its spatial distribution pattern, by matching and screening the candidate feature points, determining the effective feature points corresponding to the inherent optical microstructure on the substrate film; S24. According to the position change of the effective feature point in the continuous image frames and in combination with the conveying speed of the substrate film, the effective feature point is tracked to obtain the real-time coordinates of the effective feature point along the moving direction of the substrate film.

3. The method for controlling the lamination tension of an anti-counterfeiting label according to claim 2, characterized in that: The specific steps in step S21 include: According to the local brightness distribution and reflection characteristics of the substrate film surface image data, the contrast of the optical microstructure is enhanced and the reflection interference is suppressed by adaptively adjusting the substrate film surface image data.

4. The method for controlling the lamination tension of an anti-counterfeiting label according to claim 3, characterized in that: The steps of enhancing the contrast of the optical microstructure and suppressing reflection interference by adaptively adjusting the substrate film surface image data based on the local brightness distribution and reflection characteristics of the substrate film surface image data include: S211. Obtaining grayscale information reflecting the local brightness distribution and polarization information reflecting the reflection characteristics from the substrate film surface image data; S212. According to the grayscale information, the contrast of the optical microstructure is enhanced by performing local contrast enhancement processing on the image data; S213. Separate the specular reflection component and the diffuse reflection component in the substrate film surface image data according to the polarization information, and suppress the reflection interference caused by the specular reflection by enhancing the diffuse reflection component.

5. The method for controlling the lamination tension of an anti-counterfeiting label according to claim 2, characterized in that: The specific steps in step S22 include: S221. Performing local area analysis on the pre-processed substrate film surface image data to obtain local brightness, contrast or texture information reflecting the optical response characteristics of the optical microstructure; S222. Based on local brightness, contrast, or texture information, and in combination with preset geometric features of the optical microstructure, identify the boundaries, corners, or specific shape areas of the optical microstructure by performing structural element recognition on the pre-processed substrate film surface image data; S223. Based on the identified boundaries, corner points or specific shape areas, and combined with their optical response characteristics, the image areas in the pre-processed substrate film surface image data are screened and aggregated to determine the candidate feature points corresponding to the optical microstructure.

6. The method for controlling the lamination tension of an anti-counterfeiting label according to claim 2, characterized in that: The specific steps in step S23 include: S231. Obtain the local image area corresponding to the candidate feature point; S232. Based on the preset optical microstructure template, similarity calculation is performed on the local image area to obtain the matching degree between the candidate feature points and the template; S233. Determine whether the candidate feature points conform to a preset optical microstructure spatial distribution pattern based on the spatial relative positions between the candidate feature points; S234. Based on the judgment results of the matching degree and the spatial distribution law, the candidate feature points are screened to determine the effective feature points corresponding to the inherent optical microstructure on the substrate film.

7. The method for controlling the lamination tension of an anti-counterfeit label according to claim 2, characterized in that: The specific steps in step S3 include: S31. According to the real-time coordinates of the effective feature points along the direction of movement of the substrate film, a plurality of adjacent effective feature points are determined, and a local feature point group is constructed based on the spatial relative position relationship between the plurality of adjacent effective feature points; S32. Calculate the strain tensor of the area covered by the local feature point group based on the relative displacement change of the effective feature points in the continuous image frames and the preset mechanical parameters of the substrate film material; S33. Extract the principal strain component along the movement direction of the substrate film from the strain tensor and use it as the local strain of the substrate film.

8. The method for controlling the lamination tension of an anti-counterfeiting label according to claim 7, characterized in that: The specific steps in step S32 include: S321. Obtaining the real-time coordinates of each valid feature point in the local feature point group in the continuous image frames, and calculating the displacement vector of each valid feature point between the continuous image frames; S322. Constructing a deformation gradient tensor describing the deformation of the local area based on the displacement vector of each valid feature point and its initial spatial position in the local feature point group; S323. Based on the deformation gradient tensor and in combination with the preset mechanical parameters of the substrate film material, the strain tensor of the area covered by the local feature point group is solved by considering the anisotropic or nonlinear deformation characteristics of the material.

9. The method for controlling the lamination tension of an anti-counterfeiting label according to claim 8, characterized in that: The specific steps in step S323 include: S3231. Obtain the deformation gradient tensor of the area covered by the local feature point group, and the tension information applied to the area by the laminating station; S3232. Evaluate the deviation between the preset mechanical parameters of the substrate film material and the actual mechanical response of the current film segment based on the deformation gradient tensor, the tension information, and the relationship between the historically calculated strain tensor and the corresponding tension information; S3233. Based on the deviation, the preset mechanical parameters of the substrate film material are corrected; the correction includes adjusting the parameters related to the anisotropic or nonlinear deformation characteristics of the material; S3234. Based on the deformation gradient tensor and combined with the corrected mechanical parameters of the substrate film material, the strain tensor of the area covered by the local feature point group is solved.

10. A film tension control system for anti-counterfeiting labels, characterized in that: include: A first acquisition module is used to acquire surface image data of the substrate film at an upstream position of the laminating station; The second acquisition module is used to identify the inherent optical microstructures on the substrate film as feature points from the surface image data of the substrate film, and obtain the real-time coordinates of the feature points along the movement direction of the substrate film by tracking the feature points; A calculation module is used to calculate the local strain of the substrate film based on the real-time coordinates of the feature points; A first calculation module is used to calculate the strain deviation based on the local strain and the preset target strain; The second calculation module is used to calculate the time required for the film segment corresponding to the strain deviation to reach the laminating station according to the conveying speed of the substrate film and the distance between the position where the substrate film surface image data is obtained and the laminating station as the adjustment time; a third calculation module, configured to calculate a tension compensation amount based on the strain deviation and a preset relationship between the strain deviation and the tension adjustment amount; The control module is used to send the tension compensation amount to the tension control unit within the adjustment time, so that the tension control unit adjusts the tension applied to the substrate film according to the tension compensation amount when the corresponding film section enters the laminating station.

Images