A holographic image anti-counterfeiting pattern generation and verification method

Abstract

The application provides a holographic image anti-counterfeiting pattern generation and verification method, extracts a spatial feature set through a preset geometric structure, adopts subgraph matching and time sequence sorting to construct a time sequence framework, further extracts interference characteristic information and constructs a binding constraint model, obtains diffraction effect data for light field reconstruction and intensity distribution calculation when the matching degree is higher than a threshold, combines an enhanced sequence logic and a gradient descent method to iteratively optimize parameter boundaries, finally realizes verification of consistency of an activation sequence and a spatial arrangement, and outputs a verification result. Through double constraints of space and time, the application ensures high precision and reliability of anti-counterfeiting verification, and significantly improves the anti-counterfeiting effect of the holographic pattern.

Description

Technical Field

[0001] This invention belongs to the field of anti-counterfeiting and security technology, and relates to a method for generating and verifying holographic anti-counterfeiting patterns. Background Technology

[0002] Holographic anti-counterfeiting technology, as an important branch of optical information security, has been widely used in recent years for the protection of high-end documents such as passports and ID cards, valuable goods such as luxury packaging, and financial instruments such as bank cards and lottery tickets. It generates unique visual patterns through the interference and diffraction properties of light and is regarded as a core means to improve the reliability of anti-counterfeiting.

[0003] Most current mainstream holographic anti-counterfeiting solutions rely on a single static pattern or a simple dynamic flashing effect. While these methods offer some visual distinctiveness, they reveal significant shortcomings in practical applications. Attackers can capture pattern details using high-precision optical replication equipment, or gradually reconstruct spatial interference structures using computational holography algorithms, and even simulate limited dynamic changes, significantly weakening the uniqueness of the originally distinctive anti-counterfeiting pattern. More importantly, existing verification methods often focus only on the final visual result, ignoring the inherent patterns hidden during pattern formation. This allows highly similar-looking counterfeits to be easily mistaken for genuine products through ordinary visual inspection or examination with a handheld magnifying glass.

[0004] Holographic anti-counterfeiting relies heavily on two fundamental attributes: spatial distribution and temporal control. Spatial distribution requires the precise arrangement of multiple independent holographic sub-images into a specific geometric array, such as forming asymmetrical multi-layered nested patterns within the holographic area to produce a recognizable composite image. Temporal control, on the other hand, requires these sub-images to appear one by one or in groups according to a strictly predetermined sequence, such as displaying the organizer's logo first on a concert ticket before switching to the artist's portrait. If the geometric features of the spatial arrangement and the sequential logic of temporal activation are not tightly bound, a fatal flaw will occur: attackers can successfully forge the product simply by copying the final static image, or even if temporal variations are incorporated, the verification system will be unable to distinguish between genuine and counterfeit products if the sequence pattern is reverse-analyzed. This disconnect between spatial and temporal dimensions is particularly prominent in business scenarios. For example, once holographic anti-counterfeiting labels are obtained by criminals, they can use optical imaging equipment to record all visible patterns and then replay the dynamic sequence using programmable display devices such as small projectors. Therefore, there is a close relationship between spatial distribution and temporal control. Without spatial constraints, temporal sequences are easily simulated by irrelevant patterns; conversely, without temporal constraints, spatial patterns degenerate into static, easily copied images. This contradiction of interdependence and lack of interdependence means that holographic anti-counterfeiting faces a severe challenge of verification failure in high-risk business traceability labels, seriously threatening supply chain security and consumer rights.

[0005] Therefore, in the process of generating and verifying holographic anti-counterfeiting patterns, the key issue is how to make the spatially arranged geometric structure and the temporally played activation sequence form an interdependent and inseparable overall constraint. Summary of the Invention

[0006] This invention proposes a method for generating and verifying holographic anti-counterfeiting patterns. It addresses the problem of accurately matching spatial features and temporal characteristics in holographic pattern anti-counterfeiting in business scenarios. By integrating spatial arrangement data extraction, time series framework construction, and interference characteristic analysis, it solves the problem of consistency verification in spatial and temporal dimensions.

[0007] In one aspect, the present invention provides a method for generating and verifying holographic image anti-counterfeiting patterns, the method comprising: S1, extract spatial arrangement data from the composite image through a preset geometric structure, perform coordinate mapping and neighborhood association processing on the spatial arrangement data, and obtain a set of spatial features; S2, based on the spatial feature set, the subgraph matching method is used to determine the activation order parameters, and the activation order parameters are sorted in time and analyzed for dependencies to obtain the time series framework; S3. Interference characteristic information is extracted from the manifestation process through a time series framework. The interference characteristic information is then subjected to spectral decomposition and phase comparison to obtain a binding constraint model. If the matching degree between the interference characteristic in the binding constraint model and the preset geometric structure is higher than the threshold, the preliminary verification is deemed to have passed. The matching degree result is then subjected to threshold filtering to obtain the verification signal. S4. Based on the verification signal, diffraction effect data is obtained from the holographic pattern. The light field is reconstructed and the intensity distribution is calculated from the diffraction effect data to obtain the enhanced sequence logic. The binding link is iteratively optimized through the enhanced sequence logic. The gradient descent method is incorporated into the iterative optimization process to adjust the parameter boundaries and obtain the final anti-counterfeiting verification result. S5. If the final anti-counterfeiting verification result shows that the activation order is consistent with the spatial arrangement, then the real pattern is determined, the consistency index is confirmed by threshold, and the authentication output is obtained.

[0008] Preferably, in step S2, the attribute matching metric between the spatial feature set and the template subgraph library is calculated to characterize the geometric similarity of the topological structure; if the attribute matching metric meets the isomorphism judgment condition, the active template subgraph nodes are locked and the activation order parameters are assigned. The activation order parameters are determined according to the flow mark order of each template subgraph and the temporal sorting list is determined; the dependency weights between adjacent nodes in the temporal sorting list are analyzed. If the dependency weights exceed a preset threshold, logical constraint edges are established between the corresponding node pairs to construct a time series framework containing the node evolution path.

[0009] Preferably, in step S3, the trajectory deviation value is calculated to obtain interference characteristic information and spectral decomposition is performed according to the data stream mapped from the time series framework to the manifestation process, and the phase spectrum and energy distribution state are generated; the main frequency phase spectrum of the energy distribution state is extracted, the phase difference and coupling coefficient are calculated, a strong interaction relationship is established based on the coupling coefficient to define the motion equation, and the motion equation is aggregated to construct a binding constraint model; Interference characteristic data in the binding constraint model is analyzed, and topological mapping is used to convert the interference characteristic data into spatial distribution feature vectors. The Hausdorff distance between the spatial distribution feature vectors and the preset geometric structure point cloud data is calculated to generate a structural similarity metric. The matching degree value is normalized to obtain a matching degree value. If the matching degree value is higher than the preset verification judgment threshold, the preliminary verification is determined to be passed and high confidence matching segments are extracted. Threshold filtering is performed on the high confidence matching segments to remove outliers and noise, thereby obtaining the verification signal.

[0010] Preferably, in step S4, diffraction effect data is demodulated from the holographic pattern based on the verification signal. The diffraction effect data contains high-frequency spectral components. Complex domain decomposition and weighted fusion are performed to generate complex amplitude wavefront data. The complex amplitude wavefront data is processed using the angular spectrum propagation algorithm to output a three-dimensional light field voxel set. The three-dimensional light field voxel set is sliced ​​along the depth axis and the modulus square value is calculated to construct an intensity distribution model. If the intensity distribution model meets the preset sharpness condition, the feature vector is extracted and encoded to obtain the enhanced sequence logic. The binding request data stream is acquired, the sequence feature vector is parsed and generated, and the sequence feature vector is reconstructed based on the self-attention mechanism to obtain the enhanced logical sequence data. The deviation loss value of the enhanced logical sequence data on the decision plane is calculated, and the initial boundary parameters are corrected according to the deviation loss value using the gradient descent method to obtain the updated boundary set. The updated boundary set is used to define the decision region, and the enhanced logical sequence data is mapped to the decision region for legality judgment. The final anti-counterfeiting verification result is output.

[0011] Preferably, in step S5, the activation timestamp sequence and spatial coordinate point data are obtained, a spatiotemporal trajectory mapping matrix is ​​constructed, and the structural mapping correlation degree is obtained by parsing the matrix. The structural mapping correlation degree is calculated from the sequence feature vector and the permutation topology. If the structural mapping correlation degree satisfies the mapping condition, the real pattern texture distribution data is extracted and the consistency index value is calculated. The consistency index value is obtained by weighted fusion of the real pattern texture distribution data. The consistency index value is compared with the security authentication threshold. If it is greater than the security authentication threshold, an authentication output signal is generated.

[0012] In another aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the method described thereon.

[0013] In another aspect, the present invention also provides a computer program product, comprising a computer program, characterized in that the computer program, when executed by a processor, implements the method described herein.

[0014] This invention extracts a set of spatial features through a pre-defined geometric structure, constructs a time-series framework using subgraph matching and temporal sorting, and then extracts interference characteristic information and constructs a binding constraint model. When the matching degree is higher than a threshold, diffraction effect data is acquired for light field reconstruction and intensity distribution calculation. The parameter boundaries are iteratively optimized using enhanced sequence logic and gradient descent methods, ultimately verifying the consistency between the activation order and spatial arrangement, and outputting the verification result. This invention, through dual spatial and temporal constraints, ensures high precision and reliability in anti-counterfeiting verification, significantly improving the anti-counterfeiting effect of holographic patterns. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] A method for generating and verifying holographic anti-counterfeiting patterns, the method comprising: S1 extracts spatial arrangement data from the composite image through a preset geometric structure, performs coordinate mapping and neighborhood association processing on the spatial arrangement data, and obtains a set of spatial features.

[0019] Specifically, the process involves acquiring a composite image containing multi-source visual information, generating spatial arrangement data containing position indices and pixel attributes based on a preset geometric structure, mapping the arrangement data using a transformation matrix to obtain a mapped coordinate sequence after eliminating geometric deformation, calculating association weights based on the mapped coordinate sequence to construct a neighborhood association topology, and extracting the topological dimension data of the neighborhood association topology, and combining the topological dimension data with the pixel attributes to perform tensor concatenation to obtain a high-dimensional spatial feature set.

[0020] Composite images typically refer to images formed by superimposing data from multiple sources. Spatial arrangement data is extracted from the images, including the position, shape, and relative relationships of objects. The preset geometric structure can be a rectangular grid, used to divide the image region for easier subsequent analysis.

[0021] Neighborhood association processing is a core step used to establish relationships between data points. Neighborhood association refers to connecting adjacent points through distance thresholds or topological rules to form a network structure. When processing spatially arranged data, the k-nearest neighbor algorithm can be used to find the k nearest neighbors for each data point and establish edges connecting them. The specific steps of association processing include calculating Euclidean distance, setting a threshold such as considering those within 5 meters as neighbors, and then generating an adjacency matrix to represent the relationships. This processing method not only captures local structure but also reveals global arrangement patterns. Through neighborhood association, spatially arranged data is transformed into a richer feature representation.

[0022] The spatial feature set is a summary of all processing results, including mapped coordinates, relationships, and derived features. The feature set can contain node coordinates, edge lengths, and path connectivity.

[0023] Data is extracted using a fine grid, and then a projection transformation is introduced into the coordinate mapping to adapt to curved terrain. Neighborhood association can be extended to three-dimensional space, taking into account height information to form a three-dimensional network. The effectiveness of this method lies in the rapid generation of feature sets, which help identify the arrangement changes of affected areas.

[0024] S2. Based on the spatial feature set, the subgraph matching method is used to determine the activation order parameters. The activation order parameters are then sorted temporally and their dependencies are analyzed to obtain the time series framework.

[0025] Specifically, the attribute matching metric between the computational spatial feature set and the template subgraph library is used to characterize the geometric similarity of the topological structure. If the attribute matching metric meets the isomorphism judgment condition, the active template subgraph nodes are locked and activation order parameters are assigned. The activation order parameters are determined based on the flow mark order of each template subgraph and the temporal sorting list is determined. The dependency weights between adjacent nodes in the temporal sorting list are analyzed. If the dependency weights exceed a preset threshold, logical constraint edges are established between the corresponding node pairs to construct a time series framework containing the node evolution path.

[0026] The subgraph matching method involves representing a set of spatial features as a graph structure, where nodes correspond to feature points and edges correspond to neighborhood associations. First, a template subgraph is defined, based on a preset activation pattern. A matching algorithm searches the main graph for substructures similar to the template, calculating the matching degree. The template subgraph may include chain-like structures formed by multiple nodes, representing a pattern sequence. The matching process scans the entire graph, finding all matching subgraph instances. Each instance corresponds to an activation order parameter, which records the node order and activation weight of the matched subgraph. This approach achieves the transformation from static spatial data to dynamic parameters, providing a foundation for subsequent time-series analysis.

[0027] Temporal sorting refers to arranging parameters into an ordered sequence based on their intrinsic temporal attributes or external timestamps. The activation order parameters may originate from multi-temporal images. The sorting process first extracts the time label for each parameter, such as the image acquisition time, and then applies a sorting algorithm, such as quicksort, to arrange them in ascending chronological order. Specifically, this involves constructing a parameter list, where each parameter contains an order value and a timestamp; then iterating through the list, comparing timestamps and swapping them until an ordered sequence is reached. After this processing, the parameter sequence reflects the evolution of spatial features over time, which helps in tracking pattern changes.

[0028] Dependency analysis is used to reveal the interdependencies between activation order parameters. It involves constructing a dependency graph where nodes are the ordered parameters and edges represent dependency strengths. One possible implementation employs causal reasoning to compute dependencies, such as assessing the correlation between parameters using the Pearson correlation coefficient. Specifically, for two parameters A and B, the correlation coefficient of their ordered value sequences is calculated. If the coefficient exceeds a threshold such as 0.7, a dependency edge from A to B is established, which allows analysis of the dependencies in the activation order of the pattern. The analysis process involves iteratively traversing all parameter pairs, calculating coefficients, and updating the dependency graph. Furthermore, a topological sort is introduced to ensure acyclic dependencies, forming a directed acyclic graph. This analysis not only captures local dependencies but also reveals global temporal patterns. Through dependency analysis, parameters are transformed into a more structured representation, supporting complex tasks.

[0029] Combining the above steps, a time series framework is obtained, integrating ranking parameters and dependencies into a framework structure to represent the dynamic evolution of data. First, activation order parameters are determined by matching subgraphs from the spatial feature set; then, temporal ranking is performed; next, dependencies are analyzed. This integration forms a framework including a time axis, parameter nodes, and dependency edges, which can be used to monitor future changes.

[0030] S3. Interference characteristic information is extracted from the manifestation process through a time series framework. The interference characteristic information is then subjected to spectral decomposition and phase comparison to obtain a binding constraint model. If the matching degree between the interference characteristic in the binding constraint model and the preset geometric structure is higher than the threshold, the preliminary verification is deemed to have passed. The matching degree result is then subjected to threshold filtering to obtain the verification signal.

[0031] Specifically, based on the data flow mapped from the time series framework to the manifestation process, trajectory deviation values ​​are calculated to obtain interferometric characteristic information and spectral decomposition is performed to generate phase spectra and energy distribution states. The dominant frequency phase spectrum of the energy distribution state is extracted, phase difference and coupling coefficient are calculated, and a strong interaction relationship is established based on the coupling coefficient to define the motion equations. The motion equations are aggregated to construct a binding constraint model. The interferometric characteristic data in the binding constraint model is analyzed, and the interferometric characteristic data is converted into spatial distribution feature vectors using topological mapping. The Hausdorff distance between the spatial distribution feature vectors and the preset geometric structure point cloud data is calculated to generate a structural similarity metric. The matching degree value is normalized to obtain a matching degree value. If the matching degree value is higher than the preset verification judgment threshold, the preliminary verification is determined to be passed and high-confidence matching segments are extracted. Threshold filtering is performed on the high-confidence matching segments to remove outliers and noise, thereby obtaining the verification signal.

[0032] The core of extracting interferometric information from the manifestation process using a time-series framework lies in: utilizing the previously obtained time-series framework, a structured representation integrating ordered activation order parameters and dependencies for dynamic data evolution analysis; and the manifestation process, which refers to the dynamic changes in spatial features in an image over time. First, relevant parameter sequences are selected from the time-series framework, recording the temporal evolution of features. Then, an extraction algorithm is applied to the manifestation process to identify interferometric information, which refers to the phase shift and amplitude fluctuation characteristics caused by these changes. The extraction process involves scanning the time axis of the framework, calculating the manifestation difference at each time point, and generating an interferometric information vector containing phase and amplitude values. This extraction achieves the transformation from time-series data to interference features, providing foundational data for subsequent processing.

[0033] Furthermore, spectral decomposition of the interferometric characteristics is the process of converting the interferometric characteristics into a frequency domain representation, with the aim of separating different frequency components to reveal hidden patterns. Specifically, this process involves applying a variant of the Fourier transform to decompose the time-series interferometric signal into low-frequency and high-frequency components. The low-frequency component corresponds to slow changes, while the high-frequency component corresponds to sudden interference.

[0034] Phase comparisons are performed to further analyze the decomposed information. These comparisons involve contrasting the phase differences between different components in the spectral decomposition results, aiming to quantify the synchronicity and delay relationships of the interference. For two spectral components, their phase value differences are calculated; if the difference is less than a threshold, they are considered to be synchronously manifested. By iteratively comparing all component pairs, a phase relationship matrix is ​​established, which records the temporal consistency of the manifestation process.

[0035] Based on the above steps, a binding constraint model is obtained. This model is a constraint structure that integrates spectral decomposition and phase comparison results, used to describe the binding rules of the manifestation process. For example, it defines the minimum phase difference between features as a constraint condition. The model construction process first collects the phase relationship matrix, then applies a graph model representation, where nodes represent interference components and edges represent constraint strengths. Through topology optimization, the model is ensured to be acyclic, forming a reliable binding framework that constrains the order in which the patterns manifest.

[0036] It should be noted that the binding constraint model is used for remote sensing data analysis. This model integrates spectral decomposition and phase comparison results to form a constraint structure for the visualization process. The binding constraint model describes the binding rules between interferometric properties, such as defining phase difference constraints for features in an image to support the monitoring of changes. Interferometric properties refer to phase shifts and amplitude fluctuations, while the preset geometry is a predefined geometric template, such as a grid-like or curved structure, used for the spatial distribution of features.

[0037] This matching process is achieved by calculating the similarity between interference characteristics and geometric structures. For example, the cosine similarity method is used to quantify the angular differences between vectors, thereby assessing whether they meet preset conditions. Furthermore, if the matching degree between the interference characteristics and the preset geometric structure in the constraint model is higher than a threshold, the preliminary verification is considered successful. The core of this step lies in the calculation of the matching degree and the threshold comparison, with the threshold typically set to 0.8 to ensure high reliability. In monitoring scenarios, such as pattern display processes, interference characteristic vectors are first extracted from the model and then compared with a preset geometric structure template. If the matching degree exceeds the threshold, it indicates that the interference characteristics conform to the expected geometric distribution, and the preliminary verification is successful, avoiding misjudgment.

[0038] Matching degree calculation involves mapping interference characteristics to a geometric coordinate system, comparing positional deviations point-by-point, generating matching scores, and applying a threshold filter to the matching degree results to obtain a verification signal. Threshold filtering refers to applying a quadratic threshold, such as 0.9, to further filter the initial matching degree results and eliminate noise. The specific process includes sorting the matching degree values ​​and then retaining those above the filtering threshold to form the verification signal, which is represented by a binary or continuous value. In monitoring, such as verifying pattern display processes, the matching degree result sequence is first calculated, and then low-matching items are removed through filtering, outputting a verification signal to confirm the geometric consistency of the flood path. This filtering step iteratively checks each matching degree value; if it is below the threshold, it is marked as invalid, thereby generating a clean signal sequence and ensuring the stability of the operational process.

[0039] S4. Based on the verification signal, diffraction effect data is obtained from the holographic pattern. The light field is reconstructed and the intensity distribution is calculated from the diffraction effect data to obtain the enhanced sequence logic. The binding link is iteratively optimized through the enhanced sequence logic. The gradient descent method is incorporated into the iterative optimization process to adjust the parameter boundaries and obtain the final anti-counterfeiting verification result.

[0040] Specifically, based on the verification signal, diffraction effect data is demodulated from the holographic pattern. This diffraction effect data contains high-frequency spectral components. Complex domain decomposition and weighted fusion are performed to generate complex amplitude wavefront data. The complex amplitude wavefront data is processed using an angular spectrum propagation algorithm to output a three-dimensional light field voxel set. The three-dimensional light field voxel set is sliced ​​along the depth axis and the modulus squared value is calculated to construct an intensity distribution model. If the intensity distribution model meets a preset sharpness condition, feature vectors are extracted and encoded to obtain enhanced sequence logic. The binding request data stream is acquired, and sequence feature vectors are parsed and generated. Based on a self-attention mechanism, the sequence feature vectors are reconstructed to obtain enhanced logic sequence data. The deviation loss value of the enhanced logic sequence data on the decision plane is calculated. The gradient descent method is used to correct the initial boundary parameters based on the deviation loss value to obtain an updated boundary set. The updated boundary set is used to define the decision region. The enhanced logic sequence data is mapped to the decision region for legality judgment, and the final anti-counterfeiting verification result is output.

[0041] The verification signal, a key result of the preprocessing, is used to extract relevant data from the holographic pattern. The verification signal is typically a reliable identifier generated from the matching degree and threshold filtering steps, indicating that the interference properties match the pre-defined geometry as expected. This signal is stored in binary or continuous value form to ensure versatility across different business scenarios.

[0042] Acquiring diffraction effect data is a crucial step in extracting data from holographic patterns. It involves the separation and quantization of specific signal features within the spatial distribution pattern, typically referring to signal components related to spatial fluctuations that reflect the dynamic changes of a specific phenomenon. The process for acquiring diffraction effect data includes frequency domain decomposition of the holographic pattern to separate frequency components related to the target features, followed by inverse transformation mapping these components back to the spatial domain to form a quantifiable data sequence. This data can be used to identify trends; specifically, by layer-by-layer analysis of the spatial components of the holographic pattern, signal components related to geometric features are separated.

[0043] Light field reconstruction is the core step in spatially reconstructing diffraction data. It aims to integrate discrete signal components into a continuous spatial distribution representation. The principle lies in restoring the distribution pattern of remotely sensed targets in three-dimensional space through multidimensional mapping of diffraction data. The light field reconstruction process includes spatial interpolation of the diffraction data to generate a continuous signal field. Then, by superimposing signal components of different dimensions, a complete spatial distribution model is formed. Specifically, the diffraction data is first gridded, mapping discrete signal points onto a regular grid. Then, interpolation methods are used to fill in data gaps. Finally, temporal information is superimposed to form a dynamic distributed light field. This transforms abstract signal data into an intuitive spatial distribution, providing a reliable visual basis for subsequent analysis.

[0044] Furthermore, intensity distribution calculation is a crucial step in the quantitative analysis of the light field reconstruction results. It is used to assess the signal intensity variation characteristics in the spatial distribution. By statistically analyzing the signal values ​​at each point in the light field, a distribution map reflecting the intensity changes is generated. Intensity distribution calculation can be used to quantify signal intensity differences within a region. Specifically, by scanning the light field reconstruction results point by point, the signal values ​​at each location are extracted and normalized to form an intensity distribution map.

[0045] Enhanced sequence logic, as the final output, is a logical representation generated after structuring the intensity distribution calculation results. It typically manifests as an ordered signal sequence, indicating the spatial and temporal variation patterns of the target. Enhanced sequence logic can generate a logical sequence reflecting changing trends by performing time-series processing on the intensity distribution results.

[0046] To iteratively optimize the binding process by enhancing sequence logic, it's essential to first understand its role in anti-counterfeiting verification. Sequence logic refers to a set of ordered logical rules used to process the binding data between anti-counterfeiting labels and products, such as the generation and verification sequence of anti-counterfeiting codes. In the field of anti-counterfeiting verification, this logic can be applied to the binding process of product packaging, ensuring that each product corresponds to a unique anti-counterfeiting identifier.

[0047] For an anti-counterfeiting label pattern, enhanced sequence logic can optimize the binding process by adding redundant checks, such as introducing multi-layer sequence verification during binding to reduce the risk of false matches. This enhancement approach involves the collection of initial binding data, the definition and iterative application of sequence rules, thereby improving binding accuracy in anti-counterfeiting verification. Furthermore, gradient descent is incorporated into the iterative optimization process to adjust parameter boundaries. Gradient descent is an optimization algorithm used to minimize a loss function, gradually adjusting parameters by calculating the direction of the gradient. This method can be integrated into the iterative optimization of the binding process, for example, adjusting binding parameters such as threshold boundaries. The specific process is as follows: First, define initial parameter boundaries, for example, setting the similarity threshold for binding matching to 0.8; then, calculate the loss value under the current parameters during iteration, such as the binding error rate; next, update the parameters according to the gradient direction, for example, if the loss increases, decrease the boundary value to tighten the verification standard. This adjustment process is iterated multiple times in the anti-counterfeiting system until the loss converges. In this way, the parameter boundaries are optimized, ensuring that the binding process is more adaptable to the variability of different product anti-counterfeiting scenarios, such as label binding in high-noise environments.

[0048] It should be noted that obtaining the final anti-counterfeiting verification result depends on the output of the aforementioned optimization. After completing the iteration, the system generates a verification judgment based on the adjusted parameter boundaries. For example, if the binding matching degree exceeds the optimization threshold, the system outputs a "genuine" result. This result generation process involves summarizing the output of the sequence logic and evaluating the parameters after adjustment. For example, during anti-counterfeiting queries, the system compares the user-input code with the optimized binding records to arrive at a final conclusion, enabling more reliable authenticity identification. For example, in the verification of pharmaceutical packaging, the enhanced sequence logic iteratively optimizes the binding process of pharmaceutical labels. For instance, the sequence rules include ordered checks of batch numbers, dates, and unique codes. Then, gradient descent is incorporated to adjust the parameter boundaries, such as gradually reducing the boundary value of the matching error from an initial 0.1 to 0.05. The business details of this process include multiple iterations to calculate the loss and update the boundaries, ultimately obtaining a verification result such as "verification passed."

[0049] S5. If the final anti-counterfeiting verification result shows that the activation order is consistent with the spatial arrangement, then the real pattern is determined, the consistency index is confirmed by threshold, and the authentication output is obtained.

[0050] Specifically, the activation timestamp sequence and spatial coordinate point data are acquired, a spatiotemporal trajectory mapping matrix is ​​constructed, and the structural mapping correlation degree is obtained by analyzing the matrix. The structural mapping correlation degree is calculated from the sequence feature vector and the permutation topology. If the structural mapping correlation degree satisfies the mapping condition, the real pattern texture distribution data is extracted and the consistency index value is calculated. The consistency index value is obtained by weighted fusion of the real pattern texture distribution data. The consistency index value is compared with the security authentication threshold. If it is greater than the security authentication threshold, an authentication output signal is generated.

[0051] The anti-counterfeiting pattern consists of multiple functional points that can be excited by a specific light source or magnetic field. These functional points are distributed according to a preset spatial position and have specific response timing characteristics. The anti-counterfeiting verification device first acquires a sequence of dynamic response images of the pattern during the excitation process using an imaging sensor. The activation order refers to the sequence in which each functional point produces an observable response on the time axis. For example, under ultraviolet light, some fluorescent points emit light first, while others emit light with a delay of hundreds of milliseconds; this temporal relationship constitutes the activation order. The verification device performs temporal analysis on the acquired image sequence, extracts the first response time of each functional point, and sorts all functional points according to the magnitude of the time, thus obtaining the actual observed activation order. Further, spatial arrangement refers to the geometrical layout of each functional point on a two-dimensional plane. A coordinate system is typically used to locate the pattern; for example, the center of the pattern is set as the origin, and each functional point is assigned horizontal and vertical coordinate values, forming a set of spatial positions. The verification device uses image processing algorithms to identify the center position of each functional point and records its coordinate information, obtaining the observed spatial arrangement.

[0052] The consistency index is used to quantify the correspondence between the activation order and the preset spatial arrangement. First, the preset spatial arrangement of the actual pattern is converted into a reference activation order, that is, the spatial positions are sorted from smallest to largest according to the expected response delay during design. Then, the actual observed activation order is compared with the reference order. The Spearman rank correlation coefficient between the two orders can be calculated, and this coefficient value can be used as the consistency index; alternatively, the number of mismatched point pairs in the order can be counted, and the reciprocal of the mismatch ratio can be used as the consistency index.

[0053] When the consistency index is expressed numerically, for example, within the range of 0 to 1, the threshold confirmation process is as follows: In testing a large number of known genuine and counterfeit samples, the lower limit of the distribution of the consistency index for genuine products and the upper limit for counterfeit products are statistically analyzed, and a value between the two is selected as the decision threshold. For example, if the threshold is set to 0.85, when the actually calculated consistency index is greater than or equal to 0.85, the activation order and spatial arrangement are considered consistent, and the pattern is determined to be a genuine pattern; otherwise, it is determined to be counterfeit.

[0054] For anti-counterfeiting patterns on tickets, the verification equipment uses a portable ultraviolet exciter to illuminate the ticket surface at the ticket window, while a camera continuously captures 50 frames of images. This process completes the consistency verification, ultimately displaying "Authentication Passed" or "Suspected Counterfeit" on the screen. For anti-counterfeiting labels on high-value product packaging, fixed detection equipment is used to automatically scan each product label at the end of the production line. The results are then judged based on a consistency index threshold of 0.88. If the index exceeds the threshold, the product is allowed to proceed to the next process. It should be noted that by confirming the consistency index threshold, the level of verification strictness can be flexibly adjusted in different scenarios, thus ensuring the reliability of anti-counterfeiting measures while adapting to the protected objects of varying value.

[0055] In another aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0056] In another aspect, the present invention also proposes a computer program product, comprising a computer program, characterized in that the computer program implements the above-described method when executed by a processor.

[0057] In particular, according to some embodiments of this disclosure, the processes described above can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processing device, it performs the functions defined above in the methods of some embodiments of this disclosure.

[0058] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a task data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated task data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0059] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol, such as HTTP (Hypertext Transfer Protocol), and can interconnect with digital task data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0060] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: determine the network connection status of the switch production line management application in response to detecting a query operation on a production collaboration document in the switch production line management application; replace the webpage entry information corresponding to the production collaboration document with target entry file information and load target webpage resource information in response to determining that the network connection status of the switch production line management application indicates an offline state, so as to display the webpage of the production collaboration document offline in the switch production line management application, wherein the target entry file information is the file information of the entry file corresponding to the webpage of the production collaboration document downloaded in advance, and the target webpage resource information is the resource information corresponding to the webpage stored locally; in response to determining that the network connection status of the switch production line management application indicates an online state and that the webpage resource information corresponding to the production collaboration document is not stored locally, download the webpage resource information of the webpage from the production line document server, wherein the webpage resource information includes an entry file and resource information; display the webpage of the production collaboration document in the switch production line management application according to the webpage resource information, and store the webpage resource information in a local database.

[0061] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including product-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0062] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for generating and verifying holographic image anti-counterfeiting patterns, characterized in that, The method includes: S1, extract spatial arrangement data from the composite image through a preset geometric structure, perform coordinate mapping and neighborhood association processing on the spatial arrangement data, and obtain a set of spatial features; S2, based on the spatial feature set, the subgraph matching method is used to determine the activation order parameters, and the activation order parameters are sorted in time and analyzed for dependencies to obtain the time series framework; S3. Interference characteristic information is extracted from the manifestation process through a time series framework. The interference characteristic information is then subjected to spectral decomposition and phase comparison to obtain a binding constraint model. If the matching degree between the interference characteristic in the binding constraint model and the preset geometric structure is higher than the threshold, the preliminary verification is deemed to have passed. The matching degree result is then subjected to threshold filtering to obtain the verification signal. S4. Based on the verification signal, diffraction effect data is obtained from the holographic pattern. The light field is reconstructed and the intensity distribution is calculated from the diffraction effect data to obtain the enhanced sequence logic. The binding link is iteratively optimized through the enhanced sequence logic. The gradient descent method is incorporated into the iterative optimization process to adjust the parameter boundaries and obtain the final anti-counterfeiting verification result. S5. If the final anti-counterfeiting verification result shows that the activation order is consistent with the spatial arrangement, then the real pattern is determined, the consistency index is confirmed by threshold, and the authentication output is obtained.

2. The method for generating and verifying holographic image anti-counterfeiting patterns as described in claim 1, characterized in that, In step S2, the attribute matching metric between the spatial feature set and the template sub-map library is calculated to characterize the geometric similarity of the topological structure; If the attribute matching metric meets the isomorphism determination condition, then the active template subgraph node is locked and an activation order parameter is assigned. The activation order parameter is determined based on the flow mark order of each template subgraph and the temporal sorting list is determined. Analyze the dependency weights between adjacent nodes in the time-series sorting list. If the dependency weights exceed a preset threshold, establish logical constraint edges between the corresponding node pairs to construct a time-series framework containing the node evolution paths.

3. The method for generating and verifying holographic image anti-counterfeiting patterns as described in claim 1, characterized in that, In step S3, the trajectory deviation value is calculated based on the data stream mapped from the time series frame to the manifestation process to obtain the interference characteristic information and to perform spectral decomposition to generate the phase spectrum and energy distribution state; Extract the dominant frequency phase spectrum of the energy distribution state, calculate the phase difference and coupling coefficient, establish a strong interaction relationship based on the coupling coefficient, define the motion equation, and aggregate the motion equation to construct a binding constraint model; Interference characteristic data in the binding constraint model is analyzed, and topological mapping is used to convert the interference characteristic data into spatial distribution feature vectors. The Hausdorff distance between the spatial distribution feature vectors and the preset geometric structure point cloud data is calculated to generate a structural similarity metric. The matching degree value is normalized to obtain a matching degree value. If the matching degree value is higher than the preset verification judgment threshold, the preliminary verification is determined to be passed and high confidence matching segments are extracted. Threshold filtering is performed on the high confidence matching segments to remove outliers and noise, thereby obtaining the verification signal.

4. The method for generating and verifying holographic image anti-counterfeiting patterns as described in claim 1, characterized in that, In step S4, diffraction effect data is demodulated from the holographic pattern based on the verification signal. The diffraction effect data contains high-frequency spectral components. Complex domain decomposition and weighted fusion are performed to generate complex amplitude wavefront data. The complex amplitude wavefront data is processed using the angular spectrum propagation algorithm to output a three-dimensional light field voxel set. The three-dimensional light field voxel set is sliced ​​along the depth axis and the modulus square value is calculated to construct an intensity distribution model. If the intensity distribution model meets the preset sharpness condition, the feature vector is extracted and encoded to obtain the enhanced sequence logic. The binding request data stream is acquired, the sequence feature vector is parsed and generated, and the sequence feature vector is reconstructed based on the self-attention mechanism to obtain the enhanced logical sequence data. The deviation loss value of the enhanced logical sequence data on the decision plane is calculated, and the initial boundary parameters are corrected according to the deviation loss value using the gradient descent method to obtain the updated boundary set. The updated boundary set is used to define the decision region, and the enhanced logical sequence data is mapped to the decision region for legality judgment. The final anti-counterfeiting verification result is output.

5. The method for generating and verifying holographic anti-counterfeiting patterns as described in claim 1, characterized in that, In step S5, the activation timestamp sequence and spatial coordinate point data are obtained, a spatiotemporal trajectory mapping matrix is ​​constructed, and the structural mapping correlation degree is obtained by parsing the matrix. The structural mapping correlation degree is calculated from the sequence feature vector and the permutation topology. If the structure mapping correlation satisfies the mapping condition, extract the real pattern texture distribution data and calculate the consistency index value. The consistency index value is obtained by weighted fusion of the real pattern texture distribution data. Compare the consistency index value with the security authentication threshold. If it is greater than the security authentication threshold, generate an authentication output signal.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 5.

7. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 5.

Images