Method for authenticating printed products

By analyzing a limited sample of printed products to create a batch-specific fingerprint using a neural network, the method addresses the inefficiencies of high-resolution imaging, reducing costs and workload, and enabling reliable batch assignment and offline authentication.

WO2026233754A1PCT designated stage Publication Date: 2026-11-12SCHREINER GRP GMBH & CO KG
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/EP2026/061897
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-05
Filing Date
2026-04-24
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

Existing methods for authenticating printed products to determine their production batch require high-resolution imaging and dedicated equipment, increasing workload and costs, and are not suitable for standard printing presses.

Method used

A method that determines batch-specific characteristics based on a limited sample of printed products, using a neural network to analyze images and create a characteristic fingerprint for the entire production batch, allowing reliable assignment without high-resolution imaging of every product.

Benefits of technology

Reduces the need for high-resolution imaging and dedicated equipment, lowering costs and workload while ensuring reliable batch assignment, enabling offline authentication and efficient identification of production batches.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure EP2026061897_12112026_PF_FP_ABST
    Figure EP2026061897_12112026_PF_FP_ABST
Patent Text Reader

Abstract

A method for authenticating printed products comprises producing a production batch of printed products and analyzing at least one random sample of the printed products of the production batch. On the basis of the analysis of the at least one random sample of the printed products, production-batch-specific features of the printed products that are characteristic of the production batch are determined in order to identify an association of the printed products with the production batch.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] P2025,0111 WO N / 24033P-WO April 24, 2026

[0002] - 1 -

[0003] Description

[0004] Methods for authenticating printed materials

[0005] The invention relates to a method for authenticating printed products, for example labels, in particular to determine which production batch the printed products are to be assigned to or whether printed products belong to a specific production batch.

[0006] A production batch of printed products, especially labels, refers to a number of printed products that have been produced in the same manufacturing process. For example, all printed products belonging to a specific production batch were produced in the same printing process using the same printing plate. Common printing processes used include flexographic printing, screen printing, and inkjet printing.

[0007] In flexographic printing, the printing plate used serves as the printing form. All printed materials belonging to a production batch produced using flexographic printing were therefore produced with the same printing plate. In screen printing, the screen serves as the printing form. Flatbed screen printing typically uses a metal frame covered with a mesh fabric. Rotary screen printing uses rotary screens. All printed materials belonging to a production batch where screen printing was used were produced with the same screen. When inkjet printing is used to produce printed materials, the same printhead is used for all printed materials belonging to a specific production batch.

[0008] Manufacturers of printed products, such as labels, are interested for various reasons in determining which production batch printed products already delivered and in use on the market belong to. Individual production batches of printed products may, for example, be intended by a manufacturer for different sales regions. The ability to assign printed products to a production batch would, for example, allow a manufacturer to [P2025,0111 WO N / 24033P-WO 24 April 2026]

[0009] - 2 -

[0010] Check whether printed materials circulating in a market were actually intended by the manufacturer for that market.

[0011] To assign labels to a specific production batch, for example, the same code could be printed on every label of that batch. Each label would then have the same characteristic fingerprint. However, printing a batch-specific code reduces the available printing area on the label for the actual information to be printed.

[0012] In addition to this additive approach, in which each printed product in a production batch receives a special code that enables authentication and thus assignment of the printed product to a specific production batch, so-called non-clonable physical features (physically unclonable functions, PUF) can also be used to authenticate printed products.

[0013] Security features based on PUF (Polymerized Fiber Optic) utilize, for example, the characteristic fiber arrangement of a label within a paper substrate, the presence of air bubbles in a thick printed layer, or the arrangement of colored pigment granules on the label to uniquely identify it. These non-clonable security features can be intentionally added to printed materials using an additive approach, or they can arise from existing characteristics, such as natural variations in the contour of existing text or a printed 2D code. The resulting appearance of a printed product is always random, meaning that even the manufacturer cannot create a copy.

[0014] To uniquely authenticate printed materials, for example, an image of a defined area of ​​the printed material or characteristic features derived from this image could be stored in a central database. During the actual authentication process, the image of a sample to be examined, or again, the features derived from it, is then compared with the information stored in the central database for precisely this object in order to verify the authenticity of a single printed product. P2025,0111 WO N / 24033P-WO 24 April 2026

[0015] - 3 -

[0016] or to assess the assignment of an individual printed product to a specific production batch.

[0017] However, such a solution has the disadvantage that images of all individual items in a production batch must be taken to determine the characteristics typical of a printed product. To capture these individual images, either a dedicated machine or a printing press specifically equipped for image capture, such as one with an inline camera, must be used.

[0018] This disadvantage is particularly significant because the technology for reading and verifying non-clonable physical features cannot utilize the standard camera resolutions used in the printing industry (300 dpi or sometimes even 600 dpi), but requires higher resolutions, typically exceeding 1000 dpi. Since high-resolution cameras are not standard features in conventional printing presses, such image capture devices must be specifically procured and installed to capture images of each individual printed product and extract its characteristic features.

[0019] One objective of the present invention is to provide a method for authenticating printed products that makes it possible to assign printed products to a specific production batch without requiring a significantly increased workload.

[0020] An embodiment of a method for authenticating printed products, which makes it possible to determine with relatively little effort and yet reliably whether printed products can be assigned to a specific production batch or to which production batch an examined printed product can be assigned, is specified in claim 1.

[0021] The proposed method for authenticating printed materials, particularly labels, involves producing a production batch of printed materials. All printed materials in the production batch are thus in the same production or P2025,0111 WO N / 24033P-WO 24 April 2026

[0022] - 4 -

[0023] The printing process was carried out using the same printing form, for example, the same printing plate for flexographic printing, the same screen for screen printing, or the same printhead for inkjet printing or drop-on-demand inkjet printing.

[0024] At least one sample of the printed products from the production batch is analyzed. Furthermore, batch-specific characteristics of the printed products—that is, characteristics that are typical for the production batch—are determined based on the analysis of this at least one sample of printed products to identify whether the printed products belong to the production batch.

[0025] Batch-specific characteristics of printed products, especially labels, are those characteristics that are typical of a production batch or that allow for the unambiguous assignment of printed products to a specific production batch. Batch-specific characteristics of printed products can, for example, be characteristics that are common to all printed products in a production batch or that occur in all printed products of the production batch and differ from batch-independent characteristics. If a production batch comprises printed products, especially labels, arranged on a large number of printed sheets, then batch-specific characteristics can be those characteristics that are common to at least the printed products or labels arranged at a specific position on the printed sheets.In contrast, the batch-independent characteristics of labels of a production batch are all characteristics of the labels of a production batch that change within the production batch and thus do not remain constant or unchanged during the production of a production batch.

[0026] Instead of establishing the characteristic properties of a production batch based on images of all individual printed products or labels, according to the invention, only a limited sample of printed products from a production batch is used to determine the production batch-specific characteristics of the printed products of the production batch. The systematic differentiation and delimitation of one production batch of printed products from another production batch is described in P2025,0111 WO N / 24033P-WO 24 April 2026.

[0027] - 5 -

[0028] The printing process is based on the printing form, so printed products produced with a different printing form are assigned to a different production batch.

[0029] A production batch of printed materials is described by a set of batch-specific characteristics, that is, certain physical characteristics that represent a characteristic fingerprint of the printed materials from a particular production batch. This fingerprint allows for the verification of the authenticity of individual printed materials or labels. In particular, it makes it possible to determine whether individual printed materials or labels can be attributed to a specific production batch.

[0030] The determined batch-specific characteristics of a production batch can be stored in a data repository. To later determine whether a printed product, especially a label, originates from a specific production batch or to which production batch a printed product belongs, at least one characteristic of the printed product is extracted. This characteristic is compared with the batch-specific characteristics determined for a production batch. If the extracted characteristic matches one of the batch-specific characteristics determined for the production batch, the examined printed product can be assigned to that production batch.

[0031] At least one of the production batch-specific characteristics can be a non-cloneable physical characteristic or a non-cloneable security feature. Such non-cloneable physical characteristics arise randomly during the production of a printed product, for example, due to manufacturing tolerances in the printing plate, and therefore cannot be copied.

[0032] According to the present approach for authenticating printed products or labels, and in particular for assigning individual printed products / labels to a production batch, the characteristic fingerprint for a complete production batch is created solely based on a limited sample of the printed products or labels. P2025,0111 WO N / 24033P-WO 24 April 2026

[0033] - 6 -

[0034] According to the inventive approach, the batch-specific features can be determined as features that are characteristic of the manufacturing process used to produce the printed products. The batch-specific features can, for example, be features that are characteristic of a printing form used to produce the printed products.

[0035] According to one possible embodiment of the method, individual images of print samples from at least one sample of the printed products can be analyzed. The production batch-specific characteristics are determined based on the analysis of the respective images of the print samples.

[0036] The distinction between one production batch and another can be specifically influenced by the printing plate used. Thus, print samples produced with printing plate A, for example, can invariably be assigned to the corresponding production batch A, whereas all print samples produced with a different printing plate B can be assigned to production batch B, even if printing plate B has an identical layout to printing plate A.

[0037] According to one possible embodiment of the method for authenticating printed materials, a first sample of the printed materials, which may contain a specific number of items, is analyzed. Subsequently, at least one second sample of the printed materials, which is different from the first sample, is analyzed. Based on the analysis of the first and the at least one second sample of the printed materials, the batch-specific characteristics of the printed materials are determined. The first sample and the at least one second sample of printed materials from a production batch may contain the same or different numbers of printed materials. The batch-specific characteristics represent common features of printed materials that occur unchanged in both the first and the at least one second sample.

[0038] To ensure that the production batch-specific characteristics are as robust as possible, so that all printed products of a production batch reliably correspond to the associated fingerprint P2025,0111 WO N / 24033P-WO 24 April 2026

[0039] - 7 -

[0040] Since the samples can be assigned to specific groups, they can be designed to include a certain degree of variation.

[0041] This can be achieved by taking samples distributed throughout the production process. For example, the printed products or selected print samples for analysis can be taken at different times, such as from the beginning and end of production and from evenly spaced points during production. Alternatively, the samples can cover different positions of the printed product on a single sheet, for example, the various positions of labels on a sheet produced with a single flexographic printing plate or screen, or side-by-side across the web with the same inkjet printhead.

[0042] According to one possible embodiment of the method for authenticating printed products, the first sample of the printed products is therefore taken at a first point in time during the production of the batch. The at least one second sample of the labels is taken at a second point in time during the production of the batch, different from the first point in time.

[0043] According to another possible embodiment of the method for authenticating printed products, the first and at least one second sample can be drawn in such a way that the first and at least one second sample contain printed products that are arranged at different positions on a printed sheet.

[0044] This makes the determined production batch-specific characteristics very robust, so that all printed products or labels of a production batch can be reliably assigned to the determined associated fingerprint, which is represented by the production-specific characteristics.

[0045] Furthermore, the sample can be specifically enriched by varying printing parameters, for example by increasing or decreasing the viscosity of the printing ink, which simultaneously simulates heating of the printing ink for longer production times. P2025,0111 WO N / 24033P-WO 24 April 2026

[0046] - 8 -

[0047] Furthermore, for example, the contact pressure, the register setting, the printing speed, the intensity of the corona pretreatment, the drying parameters, the line spacing and / or the ink volume of the anilox roller can be varied.

[0048] The invention is explained clearly below using figures. These show:

[0049] Figure 1 shows a feature space of all features, which can be divided into production batch-specific features and production batch-independent features;

[0050] Figure 2 shows a graphically illustrative distinction between different production batches based on their production batch-specific characteristics;

[0051] Figure 3 shows a decision space for whether a feature is present, depending on two exemplary properties of the feature, e.g., the width and depth of an indentation on the side of a letter line; and

[0052] Figure 4 shows an enlarged area for the presence of a feature when pressure parameters are varied in the same decision space as in Figure 3.

[0053] To authenticate the belonging of printed products, such as labels, to a specific production batch, it is proposed to create a characteristic fingerprint of a production batch based on a limited sample of all printed products or labels. If this characteristic fingerprint is identified in a printed product or label, the printed product can be unambiguously assigned to the corresponding production batch.

[0054] Figure 1 illustrates a feature space of features of a production batch of printed products, for example, labels. The feature space includes production batch-specific features that occur in all printed products of a particular production batch, that is, in all printed products that were produced within the same manufacturing or printing process, or at least in all printed products that are arranged in the same position on different printing sheets. P2025,0111 WO N / 24033P-WO 24 April 2026

[0055] - 9 -

[0056] Furthermore, the feature space includes other features that are independent of the production batch.

[0057] Batch-independent characteristics are features that the printed products of a production batch do not have in common. Such characteristics can change during the production or printing process of a production batch, so that printed products produced at the beginning of a production batch, for example, have different characteristics than printed products produced at the end of the production or printing process of the production batch.

[0058] To determine the batch-specific characteristics of printed products that are characteristic of a particular production batch and thus of the manufacturing or printing process used, at least one sample of the printed products from a production batch is analyzed. Based on the analysis of this sample, the batch-specific characteristics can then be determined and subsequently used to identify which printed products belong to the production batch.

[0059] Batch-specific characteristics are those characteristics that are typical for a printing form that has been used to produce the printed products during the manufacturing or printing process.

[0060] When printed products, especially labels, are produced in a production batch using a flexographic printing process, the printing form is a printing plate used in the flexographic printing process. The printing plate can be flexible and made of photopolymer or rubber.

[0061] When printed products, especially labels, are produced in a production batch using a screen printing process, the printing form is represented by a screen used in the screen printing process. This can be, for example, a metal frame covered with a mesh (flatbed screen printing) or rotary screens for rotary screen printing. P2025,0111 WO N / 24033P-WO 24 April 2026

[0062] - 10 -

[0063] When printed products, especially labels, from a production batch are manufactured using an inkjet printing process, batch-specific characteristics are those characteristics that are typical of a printhead used in the inkjet printing process. The printhead represents the specific characteristics of the printing system, for example, through the orientation of the print nozzles relative to each other and through the individual thickness of the ink application by the individual print nozzles.

[0064] Different production batches are distinguished from one another based on the printing form, i.e., the flexographic printing plate for flexographic printing, the screen for screen printing, or a specific printhead for inkjet printing or drop-on-demand inkjet printing. The typical variations resulting from manufacturing tolerances in the production of a printing form ensure that a printed image contains sufficient batch-specific characteristics to allow for the identification of a characteristic fingerprint for the printed products of that production batch.

[0065] For example, letters that are identically placed side-by-side on a flexographic printing plate and printed together exhibit many small differences that only occur with the printing form used, in this example, the printing plate. Such differences can be analyzed to determine the characteristic fingerprint of a production batch.

[0066] To determine batch-specific characteristics, images of print samples from at least one sample can be analyzed. The evaluation of the print samples, typically based on a neural network, extracts characteristics from the images of the print samples of the at least one sample of printed products. These characteristics fall into the two categories shown in Figure 1: those characteristics that change with a production batch (batch-specific characteristics), and the remaining characteristics that are independent of production batches and thus occur in all production batches or already vary within a production batch. The neural network primarily distinguishes between the batch-specific characteristics based on P2025,0111 WO N / 24033P-WO April 24, 2026.

[0067] - 11 -

[0068] between the different production batches and thus separates the production batches from each other.

[0069] Figure 2 shows a schematic representation of the subspace of batch-specific features A and B of a feature space, plotted on the X and Y axes of the subspace shown in Figure 2. Feature A could, for example, be a slight indentation on one side of a printed letter. Feature B could be an indentation on a nearly right-angled contour segment of a printed letter.

[0070] According to the schematic representation in Figure 2, the presence of feature A in a printed sample of a printed product, for example, a label, indicates that the printed product originates from production batch 1. Conversely, the presence of feature B in a printed sample of a printed product indicates that the printed product belongs to production batch 2. The two features A and B shown are examples of production batch-specific features that can be used to distinguish between different production batches of printed products. Besides features A and B, there are countless other production batch-specific features that are not shown in Figure 2 for the sake of clarity.

[0071] Figure 3 illustrates whether a feature is present depending on two exemplary properties: in this example, the width of the indentation (plotted on the x-axis) and the depth of the indentation (plotted on the y-axis). For a certain suitable combination of these two parameters, the feature is considered present (dashed area). In this example, both the width and the depth of the indentation must exceed a certain minimum value. There is also an upper limit for these two parameters; if the indentation becomes too wide, for example, it no longer represents an indentation, but rather the corresponding letter segment is simply thicker throughout. If the combination of these two exemplary properties lies outside the marked area, then the feature is considered not present. P2025,0111 WO N / 24033P-WO April 24, 2026

[0072] - 12 -

[0073] Figure 4 represents the same feature space as in Figure 3, with the addition of varying pressure parameters. This variation of pressure parameters results in different, yet characteristic, batch-specific values ​​for a given production batch during manufacturing. Consequently, during the analysis phase, this feature is more robustly detected (schematically visualized in Figure 4 by the larger area being recognized as feature A), and unintended fluctuations in pressure parameters do not lead to a batch-specific feature no longer being detected.

[0074] To train the neural network to define the catch area around each production batch along the direction of the batch-specific characteristics, it can orient itself to the typical fluctuations of these characteristics. If the training sample exhibits very similar patterns, the neural network will define a comparatively small catch area, as shown in Figure 3. By varying different pressure parameters, this catch area can be selectively enlarged, as shown in Figure 4, making the assignment of a pressure pattern to a production batch more robust against unintended but typical production variations.

[0075] Varying the manufacturing or printing parameters can be achieved by changing the viscosity of a printing ink, for example by adding a thickener or thinner, which corresponds to the behavior of the printing ink during longer production times when the printing ink becomes more liquid with increasing temperature, or by changing the registration alignment of different printing units relative to each other, or by changing, for example by increasing and / or decreasing, a printing speed, or by varying the strength of a corona pretreatment, or by changing, in particular by increasing and / or decreasing, the contact pressure of a printing unit, or by printing process-specific changes, such as by varying the line spacing and / or the ink volume of the anilox roller in flexographic printing.

[0076] An advantage of the inventive method for authenticating printed products with regard to their belonging to a specific production batch is that P2025,0111 WO N / 24033P-WO 24 April 2026

[0077] - 13 -

[0078] Only images of a limited number, i.e., a sample, of printed products from a production batch are required. This significantly reduces the demands on the camera system, which no longer needs to be designed to deliver an extremely high throughput of hundreds of thousands of images per hour. This would require inline cameras across the entire print width with the necessary high resolution of over 1000 dpi and, among other things, extremely powerful lighting for high production speeds. On the contrary, in extreme cases, even photos taken with a good smartphone camera are sufficient for creating the images used to analyze the sample.

[0079] In addition to capturing images for the entire production batch, the effort required to create batch-specific characteristics—that is, fingerprints—for each individual printed product or label is also reduced, especially given the typically high production volumes, for example in label printing, which often reach into the millions or even tens of millions for a single product. Creating a large number of fingerprints requires a correspondingly high level of computing power, which can be significantly reduced with the inventive method.

[0080] Using a single, typical fingerprint for an entire production batch, which can exhibit several batch-specific characteristics, reduces not only manufacturing costs but also operating costs. Conventional visual scanning of all printed materials requires storing all images, or at least all fingerprints, of the various printed materials within a production batch to make them available for authentication. This demands significant storage capacity, especially given the aforementioned high volumes, and also requires considerable computing power to precisely identify the stored image or fingerprint that corresponds to the current authentication request from this large number of items.

[0081] The use of a single typical fingerprint for an entire production batch also enables authentication in offline mode, thus ensuring accessibility to a central server with images or P2025,0111 WO N / 24033P-WO April 24, 2026

[0082] - 14 -

[0083] Image data is not required. On the contrary, a few production batch-specific characteristics can be stored locally in an app to enable authentication without access to a central server. This is particularly relevant when the authentication of printed materials is to take place in a location with low internet availability or a poor internet connection, for example, in commercial areas with remote warehouses.

[0084] Linking a specific printing plate to its corresponding production batch enables the systematic and deliberate switching between different production batches. These distinct production batches can be used for different purposes without requiring a different layout. For example, such production batches can be designated for different sales regions, thus helping to identify gray markets or uncover diversion—that is, the redirection of products to other markets without the knowledge and consent of the brand owner.

[0085] A further advantage of the method according to the invention is that, to authenticate the belonging of a printed product to a specific production batch, no special color needs to be printed for a security feature specifically designated for this purpose, nor do any other special layout elements need to be included on the printed product. The printed image of the sample itself, i.e., for example, of a font or other graphic elements, provides the properties or characteristics tested within the framework of the security feature.

Claims

P2025,0111 WO N / 24033P-WO April 24, 2026 - 15 - Patent claims 1. Procedures for authenticating printed materials, comprising: - Producing a production batch of printed materials, - Analyzing at least one sample of the printed products from the production batch, - Determining production batch-specific characteristics of the printed products that are characteristic of the production batch, based on the analysis of at least one sample of the printed products to indicate that the printed products belong to the production batch.

2. The method of claim 1, comprising: - Analyzing an initial sample of the printed materials, - Analyzing at least one second sample of the printed materials that differs from the first sample, - Determining the production batch-specific characteristics of the printed products based on the analysis of the first and at least one second sample of the printed products.

3. Method according to claim 2, - where the first sample of the printed products is taken at an initial point in time during the production of the manufacturing batch, - wherein at least one second sample of the printed products is taken at a second point in time of production of the production batch, different from the first point in time.

4. Method according to claim 2 or 3, wherein the first and at least one second sample contain printed products arranged in different positions on a printed sheet.

5. Method according to any one of claims 1 to 4, where at least one of the production batch-specific characteristics is a non-clonable physical characteristic.

6. Method according to any one of claims 1 to 5, comprising: P2025,0111 WO N / 24033P-WO 24 April 2026 - 16 - - Analysis of respective images of print samples from at least one sample of printed products, - Determining the production batch-specific characteristics based on the analysis of the respective images of the printed samples.

7. Method according to any one of claims 1 to 6, comprising: Determining the production batch-specific characteristics as characteristics that are characteristic of the manufacturing process used to produce the printed products.

8. Method according to any one of claims 1 to 7, where the production batch-specific characteristics are characteristic of a printing form used to produce the printed products.

9. Method according to claim 8, wherein the printed products are produced using a flexographic printing process and the printing form is a printing plate used in the flexographic printing process.

10. Method according to claim 8 or 9, wherein the printed products are produced using a screen printing process and the printing form is a screen used in the screen printing process.

11. Method according to any one of claims 1 to 10, wherein the printed products are manufactured using an inkjet printing process and the production batch-specific characteristics are characteristic of a printhead used in the inkjet printing process.

12. Method according to any one of claims 1 to 11, comprising: Variations in manufacturing parameters that influence the manufacturing process used for the printed products.

13. Method according to claim 12, where varying the manufacturing parameters includes changing the viscosity of a printing ink or changing the registration alignment of different printing units relative to each other. P2025,0111 WO N / 24033P-WO 24 April 2026 - 17 - or includes changing a printing speed, varying the strength of a corona pretreatment, changing the contact pressure of a printing tool, or changes specific to the printing process.

14. Method according to any one of claims 1 to 13, comprising: Storing the production batch-specific characteristics in a data storage device.

15. Method according to any one of claims 1 to 14, comprising: - Extracting at least one feature from one of the printed products, - Comparing the characteristic with the batch-specific characteristics determined for the production batch, - Assigning the printed product to the production batch if the extracted characteristic matches one of the production batch-specific characteristics determined for the production batch.