A method and device for processing and identifying a forgery-proof image

Abstract

This invention provides a method, method, and apparatus for processing and recognizing anti-counterfeiting images, relating to the field of anti-counterfeiting technology. It solves the problem that existing anti-counterfeiting images generated by existing technologies are easily copied by ultra-high-precision scanning and professional printing equipment, making it impossible to distinguish genuine from counterfeit. The method for processing the anti-counterfeiting image includes: acquiring parameters and graphic content information of a target anti-counterfeiting image; generating a halftone image matrix based on the parameters of the target anti-counterfeiting image; fusing the halftone image matrix and the graphic content information to obtain a binarized halftone dot map; generating a target anti-counterfeiting image based on the binarized halftone dot map; and performing anti-counterfeiting recognition on the anti-counterfeiting image to be recognized based on the halftone image matrix in the target anti-counterfeiting image. The solution of this invention improves the anti-counterfeiting accuracy and protection capability of anti-counterfeiting images, while achieving accurate anti-counterfeiting recognition of existing anti-counterfeiting images and preventing high-precision illegal copying of existing anti-counterfeiting images.

Description

Technical Field

[0001] This invention relates to the field of anti-counterfeiting technology, and in particular to a method, identification method and device for processing anti-counterfeiting images. Background Technology

[0002] In today's business world, the problem of counterfeit and substandard goods has always been a difficult issue for businesses and consumers. To protect brand reputation and safeguard consumer rights, packaging anti-counterfeiting technology has emerged and continues to develop and improve. While existing packaging anti-counterfeiting technologies have a certain degree of anti-counterfeiting performance, meaning that it is difficult to counterfeit similar products without the relevant software, materials, or process parameters, packaging products inevitably rely on physical media, and most are produced through printing. Therefore, with the rapid development of image copying equipment and technology, most packaging products can be illegally copied through ultra-high precision scanning and professional printing equipment (such as high-definition gravure printing), thereby reducing the protective capabilities of existing anti-counterfeiting images and making it impossible to achieve accurate anti-counterfeiting. Summary of the Invention

[0003] This invention provides a method, method, and apparatus for processing and recognizing anti-counterfeiting images. It solves the problem that anti-counterfeiting images generated using existing anti-counterfeiting technologies are easily copied by ultra-high-precision scanning and professional printing equipment, making it impossible to distinguish genuine from counterfeit images.

[0004] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0005] Embodiments of the present invention provide a method for processing anti-counterfeiting images, comprising:

[0006] Obtain the parameters and graphic content information of the target anti-counterfeiting image;

[0007] Based on the parameters of the target anti-counterfeiting image, a net image matrix is ​​generated;

[0008] The halftone image matrix and the text and image content information are fused to obtain a binarized halftone image;

[0009] Based on the binarized dot pattern, generate the target anti-counterfeiting image;

[0010] Based on the mesh image matrix in the target anti-counterfeiting image, anti-counterfeiting recognition is performed on the anti-counterfeiting image to be identified.

[0011] Embodiments of the present invention also provide a method for identifying anti-counterfeiting images, comprising:

[0012] Obtain the anti-counterfeiting image to be identified;

[0013] Based on the net image matrix of the target anti-counterfeiting image, determine the preset recognition template matrix;

[0014] Based on the identification template, the anti-counterfeiting image to be identified is subjected to anti-counterfeiting identification.

[0015] Embodiments of the present invention also provide an anti-counterfeiting image processing apparatus, comprising:

[0016] The first acquisition module is used to acquire the parameters and graphic content information of the target anti-counterfeiting image;

[0017] The first processing module is used to generate a halftone image matrix based on the parameters of the target anti-counterfeiting image; fuse the halftone image matrix and the graphic content information to obtain a binarized halftone image; generate a target anti-counterfeiting image based on the binarized halftone image; and perform anti-counterfeiting recognition on the anti-counterfeiting image to be identified based on the halftone image matrix in the target anti-counterfeiting image.

[0018] Embodiments of the present invention also provide an anti-counterfeiting image recognition device, comprising:

[0019] The second acquisition module is used to acquire the anti-counterfeiting image to be identified;

[0020] The second processing module is used to determine a preset recognition template matrix based on the net image matrix of the target anti-counterfeiting image; and to perform anti-counterfeiting recognition on the anti-counterfeiting image to be recognized based on the recognition template.

[0021] The above-described solution of the present invention has at least the following beneficial effects:

[0022] The anti-counterfeiting image processing method of this invention obtains the parameters and graphic content information of the target anti-counterfeiting image; generates a halftone image matrix based on the parameters of the target anti-counterfeiting image; fuses the halftone image matrix and the graphic content information to obtain a binarized halftone image; generates the target anti-counterfeiting image based on the binarized halftone image; and performs anti-counterfeiting identification on the anti-counterfeiting image to be identified based on the halftone image matrix in the target anti-counterfeiting image. This improves the anti-counterfeiting accuracy and protection capability of existing anti-counterfeiting images, while simultaneously achieving accurate anti-counterfeiting identification of existing anti-counterfeiting images and preventing high-precision illegal copying of existing anti-counterfeiting images. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the anti-counterfeiting image processing method of the present invention;

[0024] Figure 2 This is a schematic diagram of the structure of the mesh image matrix in the anti-counterfeiting image processing method of the present invention;

[0025] Figure 3 This is a schematic diagram of the structure of the preset recognition template matrix in the anti-counterfeiting image processing method of the present invention;

[0026] Figure 4This is a schematic diagram of the structure of the target anti-counterfeiting image in the anti-counterfeiting image processing method of the present invention;

[0027] Figure 5 This is a schematic diagram of the structure of the preset recognition template matrix recognition process in the anti-counterfeiting image processing method of the present invention;

[0028] Figure 6 This is a fusion effect diagram of the single dot recognition result in the anti-counterfeiting image processing method of the present invention;

[0029] Figure 7 This is a schematic diagram of dot recognition result matching in the anti-counterfeiting image processing method of the present invention;

[0030] Figure 8 This is a flowchart illustrating the anti-counterfeiting image recognition method of the present invention;

[0031] Figure 9 This is a schematic diagram of the module of the anti-counterfeiting image processing device of the present invention;

[0032] Figure 10 This is a schematic diagram of the module of the anti-counterfeiting image recognition device of the present invention. Detailed Implementation

[0033] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0034] like Figures 1 to 4 As shown, an embodiment of the present invention provides a method for processing anti-counterfeiting images, including:

[0035] Step 1: Obtain the parameters and graphic content information of the target anti-counterfeiting image;

[0036] Step 2: Generate a mesh image matrix based on the parameters of the target anti-counterfeiting image;

[0037] Step 3: The halftone image matrix and the text and image content information are fused to obtain a binarized halftone image;

[0038] Step 4: Generate the target anti-counterfeiting image using the binarized dot pattern;

[0039] Step 5: Based on the mesh image matrix in the target anti-counterfeiting image, perform anti-counterfeiting identification on the anti-counterfeiting image to be identified.

[0040] In this embodiment, the parameters of the target anti-counterfeiting image are the length and width of the target anti-counterfeiting image; the graphic content information is the text and pattern information within the target anti-counterfeiting image, such as... Figure 4 The apple pattern in the image; Step 3 specifically involves performing binarized image halftone processing on the halftone image matrix and the graphic content information to obtain a binarized halftone dot image; Step 4 specifically involves transferring the halftone dot image to an output device, and transferring the graphic content of the binarized halftone dot image to a physical printing substrate through an output transfer process to generate a target anti-counterfeiting image, thus completing the substrate transfer of the anti-counterfeiting graphic content. This type of transfer process includes various traditional printing processes, laser holographic imaging processes, etc. The graphic transfer method involved in this invention is not limited to the above-mentioned graphic transfer processes. This invention not only uses traditional printing processes to create binarized anti-counterfeiting graphics using digital or analog methods. The printing plate is designed so that the image areas are oleophilic and the blank areas are hydrophilic. During printing, the printing plate is first dampened with a dampening solution and then the ink is applied. Only the image areas absorb the ink, and then the ink is transferred to the substrate such as paper by a blanket roller, achieving precise transfer and replication of the image. Furthermore, holograms are recorded on photosensitive materials using laser interference technology. After development and fixing, a master plate (holographic original) is made. Then, a metal mold (nickel plate) is made by electroforming. The fine interference fringe structure is then thermally transferred or UV-cured and replicated to the surface of plastic film or paper by imprinting or molding, achieving large-scale batch transfer of holographic images. Based on the above transfer and replication process, the final transferred substrate product is obtained.

[0041] The anti-counterfeiting image processing method described in this embodiment solves the problem that anti-counterfeiting images generated by existing anti-counterfeiting technologies are easily copied by ultra-high precision scanning and professional printing equipment, making it impossible to identify the authenticity of existing anti-counterfeiting images. By using specific printing processes and reading equipment, multi-level anti-counterfeiting identification methods are achieved without changing existing production processes and costs, and a truly effective anti-counterfeiting effect is achieved. The anti-counterfeiting accuracy and protection capability of existing counterfeit images are improved, while accurate anti-counterfeiting identification of existing anti-counterfeiting images is achieved, preventing high-precision illegal copying of existing anti-counterfeiting images.

[0042] In an optional embodiment of the present invention, step 2, generating a net image matrix based on the parameters of the target anti-counterfeiting image, includes:

[0043] Step 21: Based on the parameters of the target anti-counterfeiting image, determine an initial image matrix of a preset size;

[0044] Step 22: Generate multiple randomly distributed target center points within the initial image matrix region;

[0045] Step 23: Based on the target center point, determine the triangular mesh that satisfies the empty circle property corresponding to each target center point within the initial image matrix;

[0046] Step 24: Map the triangular grids within the initial image matrix to obtain the threshold corresponding to each triangular grid within the initial image matrix;

[0047] Step 25: Determine the meshing image matrix based on the threshold corresponding to each of the triangular grids in the initial image matrix.

[0048] In this embodiment, step 21 can specifically be: generating an initial image matrix with an area equal to a preset multiple of the target anti-counterfeiting image based on the parameters of the target anti-counterfeiting image; step 23, determining a triangular mesh that satisfies the empty circle characteristic for each target center point within the initial image matrix based on the target center point, including: determining a super triangle containing all target center points as the initial mesh within the initial image matrix based on the target center point; inserting a new target center point into the initial mesh each time, and deleting all triangles whose circumcircles contain the new target center point; connecting the new target center point with the vertices of the affected area to form new triangles, ensuring that the empty circle criterion is met, that is, each newly generated triangle does not contain any target center point, that is, no other target center point exists within a triangle formed by three target center points; in this embodiment, the design of generating a triangular mesh that satisfies the empty circle characteristic based on the target center point can effectively avoid the occurrence of elongated triangles.

[0049] In an optional embodiment of the present invention, step 22, generating a plurality of randomly distributed target center points within the initial image matrix region, includes:

[0050] Step 221: Randomly generate an initial point within the initial image matrix region, and store the initial point in the target point set and the candidate point set respectively;

[0051] Step 222: Based on the estimated radius of a single dot in the target anti-counterfeiting image, the region of the initial image matrix is ​​divided into grids to obtain a gridded initial image matrix.

[0052] Step 223: Based on the initial points in the candidate point set and the gridded initial image matrix, determine a preset number of initial candidate points corresponding to the initial points in the candidate point set;

[0053] Step 224: Based on the location information of each initial candidate point, filter all initial candidate points to obtain target candidate points;

[0054] Step 225: Based on the target candidate points, determine a set of target points including multiple target center points;

[0055] Step 226: Determine the target center point within the initial image matrix region based on the set of target points.

[0056] In this embodiment, the set of target points refers to the set of target center points;

[0057] Specifically, in step 221, a random and uniformly distributed initial point P0 is generated within the matrix region, and the coordinate information of the initial point is determined. Then, P0 is added to the set of target points as the target center point, and at the same time, P0 is added to the set of candidate points as a candidate point;

[0058] Specifically, in step 222, the region of the initial image matrix is divided into a grid by the formula , to obtain a meshed initial image matrix, where h is the side length of the grid and r is the estimated radius of a single grid point; the estimated radius of a single grid point refers to the estimated radius of the shape of the grid points in the target anti-counterfeiting image, and each grid point in the target anti-counterfeiting image represents a polygon;

[0059] Specifically, in step 223, based on the coordinate information of the initial point, k initial candidate points are randomly generated by ; where represents the i-th initial candidate point; represents the candidate point selected from the set of candidate points, and if the selected candidate point here is P0, then here represents P0; represents a random direction vector within the unit circle (polar angle ), d represents the radius, ;

[0060] Specifically, in step 224, for each initial candidate point , check whether it is within the initial image matrix region of [0, N] x [0, N]. If it is out of bounds, discard it. If it meets the requirements, retain it to obtain the first screened point . For all the first screened points , perform the following processing. According to the grid coordinates where is located, check all the grid cells within the 3x3 or 5x5 neighborhood centered on (usually check the 5x5 neighborhood to ensure security); traverse each existing point in these neighborhood grid cells, and calculate the Euclidean distance || - || between and ; If there exists any point such that || - || < r, then the first screened point Invalid and discarded if there is no point such that || - || < r, then is retained and added to the candidate point set as a target candidate point, and at the same time added to the target point set as a target center point;

[0061] Specifically, in step 225, based on the new candidate point set obtained in step 224, a candidate point is randomly selected from the new candidate point set to execute steps 223 to 224, that is, new initial candidate points are generated around the candidate points in the new candidate point set, and new target candidate points are determined through the initial candidate points to determine the target center points in the target point set; until the candidate point set is empty, that is, no candidate point can be found for all points in the candidate point set, which means that no new position can be found to insert points that satisfy the minimum distance r; at the same time, a random point set that satisfies the minimum spacing r constraint can be obtained at this time, that is, a target point set including multiple target center points; during the loop process, if after After generating all k candidate points, no valid point is successfully found, then it is considered that the space around the point has been filled and needs to be removed from the candidate point set; if there is any point such that || - || < r, then the candidate point is invalid and discarded.

[0062] Specifically, in step 226, based on the position parameter information of the target center points in the target point set, marking is performed within the initial image matrix region, so as to obtain an initial image matrix including multiple randomly distributed target center points.

[0063] In an optional embodiment of the present invention, in step 24, mapping processing is performed on the triangular meshes in the initial image matrix to obtain the thresholds corresponding to each triangular mesh in the initial image matrix, including:

[0064] Step 241, determining the area of each triangular mesh within the initial image matrix region;

[0065] Step 242, based on the area of each triangular mesh, performing normalization processing on each triangular mesh to obtain the normalization value corresponding to each triangular mesh;​​​​​In this embodiment, step 241 specifically involves calculating the area of ​​each triangular grid by using the coordinates of the target center point corresponding to the triangular grid.

[0068] Step 242 specifically involves, through Normalize each triangular mesh to obtain the normalized value corresponding to each triangular mesh, where, This represents the normalized value corresponding to the normalization process of the i-th triangular mesh; This represents the area of ​​the i-th triangular mesh. This represents the area of ​​the largest triangle among all triangle meshes; in a preferred embodiment, the triangle meshes can also be normalized using the centroid coordinates to obtain a normalized value for each triangle mesh; step 243 specifically maps the normalized values ​​to 8-bit deep integers, calculated as follows:

[0069] Through formula The normalized value corresponding to each triangular grid is mapped to an 8-bit depth integer value to obtain the threshold corresponding to each triangular grid; where round() means rounding; This represents the threshold corresponding to the i-th triangular grid.

[0070] In an optional embodiment of the present invention, step 25, determining the mesh image matrix based on the threshold corresponding to each of the triangular grids within the initial image matrix, includes:

[0071] Step 251: Perform Gaussian noise jitter processing on each of the thresholds in the initial image matrix to obtain the first threshold processing data;

[0072] Step 252: Within the initial image matrix, the first threshold processing data is overlaid on the triangular grid area corresponding to each first threshold processing data to obtain the first processed image matrix;

[0073] Step 253: Obtain each pixel of the first processed image matrix;

[0074] Step 254: Based on the first threshold processing data and the parameters of the pixels, determine the triangular grid corresponding to each pixel;

[0075] Step 255: Based on the center point of each triangular grid, determine the pixel threshold corresponding to each pixel within each triangular grid;

[0076] Step 256: Determine the halftone image matrix based on the pixel threshold corresponding to each pixel.

[0077] In this embodiment, step 251 specifically involves using the formula... Gaussian noise jitter processing is applied to each threshold in the initial image matrix to obtain the first threshold processing data. ;in This is a general Gaussian calculation function;

[0078] Step 252 specifically involves directly covering the triangular grid area with the first threshold processing data based on the triangular grid corresponding to the first threshold processing data.

[0079] Step 253 specifically involves scanning the first processed image matrix to obtain each pixel of the first processed image matrix.

[0080] Step 254 specifically involves determining the distance between each pixel and the first threshold processing data based on the first threshold processing data and the coordinate parameters of the pixel, and determining the triangular grid corresponding to the pixel based on the distance.

[0081] Step 255 specifically involves determining the pixel threshold corresponding to each pixel based on the center point of each triangular grid and the distance between the center point and each pixel within each triangular grid. The greater the distance, the larger the value, and the closer the distance, the smaller the value, thereby achieving precise positioning of each pixel.

[0082] Step 256 specifically involves generating the corresponding halftone image matrix directly based on the determined threshold for each pixel.

[0083] In an optional embodiment of the present invention, step 3 involves fusing the halftone image matrix and the text and image content information to obtain a binarized halftone image, including:

[0084] Step 31: Obtain all corresponding pixels of the image matrix and text content information;

[0085] Step 32: Compare the halftone image matrix with all corresponding pixels of the text and image content information to obtain a binarized halftone image.

[0086] In this embodiment, step 31 specifically involves tiling the halftone image matrix onto the text and image content information, determining all pixels corresponding to the halftone image matrix and the text and image content information, and simultaneously obtaining the first threshold corresponding to all corresponding pixels of the text and image content information, as well as the pixel threshold corresponding to all corresponding pixels of the halftone image matrix. Step 32 specifically involves comparing the pixel threshold on the halftone image matrix with the first threshold corresponding to all pixels of the text and image content information. If the first threshold (grayscale value) corresponding to a pixel of the text and image content information is less than the grayscale value of the pixel threshold (grayscale value) on the halftone image matrix, the position of the pixel is output as 1; otherwise, it is output as 0. This process is repeated to obtain the binarized halftone dot image after halftone application.

[0087] In an optional embodiment of the present invention, step 5, based on the mesh image matrix in the target anti-counterfeiting image, performs anti-counterfeiting identification on the anti-counterfeiting image to be identified, including:

[0088] Step 51: Obtain the anti-counterfeiting image to be identified;

[0089] Step 52: Determine the preset recognition template matrix based on the mesh image matrix of the target anti-counterfeiting image;

[0090] Step 53: Perform anti-counterfeiting identification on the anti-counterfeiting image to be identified according to the identification template.

[0091] In this embodiment, step 51 specifically involves acquiring the anti-counterfeiting image to be identified using photographic equipment such as a mobile phone or a magnifying glass; step 52, based on the net image matrix of the target anti-counterfeiting image, determines the preset identification template matrix by processing data at the first threshold of the net image matrix. A second threshold t is set and binarized to obtain a preset recognition template matrix, labeled T. In this embodiment, the second threshold t = 82, which is expressed as: Where N represents the width and height dimensions of the halftone image matrix; Let represent the set of recognition template matrices, where a and b represent the horizontal and vertical coordinates of pixels within the template, respectively. Step 53, based on the preset recognition template matrix, specifically involves performing anti-counterfeiting recognition on the anti-counterfeiting image to be recognized by sliding the preset recognition template matrix across the anti-counterfeiting image to be recognized, and calculating the similarity between the anti-counterfeiting image to be recognized and the preset recognition template matrix at each sliding position during the sliding process. The similarity matching degree is calculated as follows:

[0092] ,in, This represents the local color value at the position of row x+e and column y+v in the anti-counterfeiting image to be identified. x and y represent the local coordinates of the anti-counterfeiting image to be identified. These coordinates refer to the coordinates of the local center point of the anti-counterfeiting image to be identified; that is, the sliding position of the current preset identification template matrix on the anti-counterfeiting image to be identified. This represents the average color value of a local area in the anti-counterfeiting image to be identified. This represents the mean of the preset recognition template matrix; The matching degree value is the similarity score; the larger the value, the higher the matching degree. N represents the width and height dimensions of the image matrix. The similarity score is compared with the first preset threshold τ to determine the target image segment on the anti-counterfeiting image to be identified. Specifically, if C(x, y) > τ, it is determined that there is a preset identification template dot at that position, i.e., there is a dot in the image matrix. The center (x1, y1) of all dots in the preset identification template matrix corresponding to the current local coordinate position of the anti-counterfeiting image to be identified is obtained and calibrated. Based on the center of all dots in the preset identification template matrix, the centroid of the dots corresponding to the center of the dots in the preset identification template matrix at all locations of the current local coordinate position of the anti-counterfeiting image to be identified is determined. The precise centroid of the dots can be calculated using the following formula. ,in, Where r is the approximate radius of a single halftone dot, i.e., the estimated radius of a single halftone dot; m and n represent the index values ​​of each pixel in the bounding rectangle of a single halftone dot; in this embodiment, =0.8, r is set to 10, and the single-site identification result is referenced. Figure 6 By sequentially sliding and comparing, all corresponding target image blocks on the anti-counterfeiting image to be identified are finally determined.

[0093] like Figure 5 As shown, during the sliding process, the anti-counterfeiting image to be identified is divided into multiple image blocks according to its size. Figure 5 In this model, each grid represents an image segment; based on the dots within the preset recognition template matrix, each target image segment is compared to obtain the comparison result for each target image segment.

[0094] like Figure 7 As shown, the coordinates (x1, y1) of the halftone dots in the preset recognition template matrix are respectively compared with the coordinates of the halftone dots in the target image block. A comparison is performed, and halftone dots with coordinate errors within a preset range are identified as successfully matched halftone dots, while halftone dots exceeding the preset range are identified as unmatched halftone dots. The matching degree is calculated as the number of successfully matched halftone dots divided by the sum of the number of successfully matched halftone dots and the number of unmatched halftone dots. The matching result of each target image segment is determined based on the matching degree of each target image segment and a preset matching value. When the matching degree is greater than the preset matching value (the preset matching value = 90%), the current target image segment is identified as halftone dot matched; when the matching degree is less than or equal to the preset matching value, the current target image segment is identified as halftone dot mismatched.

[0095] Based on the comparison results of each target image segment, the authenticity of the current anti-counterfeiting image to be identified is determined. Specifically, when the comparison results of all target image segments are dot mismatch, the current anti-counterfeiting image to be identified is determined to be a fake anti-counterfeiting image or a copy. When the comparison result of any target image segment among all target image segments is dot match, the current anti-counterfeiting image to be identified is determined to be a genuine target anti-counterfeiting image.

[0096] The anti-counterfeiting image processing method described in this invention utilizes supercellular frequency modulation and amplitude modulation (FMAM) halftone screen technology. It employs special, non-standard dot arrangement rules, encompassing irregular dot distributions, unique polygonal dot shape combinations, and adaptive dot sizes, among other unique FMAM halftone screen generation methods. Combined with specific printing methods and processes, this invention uses specialized detection equipment such as mobile phones. Based on the high-precision dot images already printed and transferred to the substrate, it employs an efficient, adaptive, multimodal dot template precise matching algorithm to achieve digitally quantifiable authentication of the dot shape and distribution characteristics of packaging printed materials. This invention realizes a dedicated high-precision microscopic dot feature creation and precise identification method, effectively resisting both ordinary copying and illegal copying by ultra-high precision scanning and professional printing equipment, resulting in extremely high anti-counterfeiting performance.

[0097] The specific implementation process of the anti-counterfeiting image processing method of the present invention is as follows:

[0098] The first step is the anti-counterfeiting image generation process:

[0099] Obtain the parameters and graphic content information of the target anti-counterfeiting image;

[0100] Based on the parameters of the target anti-counterfeiting image, an initial image matrix of a preset size is determined, with the matrix width and height dimensions being N. In this embodiment, N=320.

[0101] An initial point P0 is randomly and uniformly generated within the matrix region, and the coordinate information of the initial point is determined. P0 is added to the target point set as the target center point, and P0 is also added to the candidate point set as a candidate point.

[0102] Through formula The region of the initial image matrix is ​​divided into grids to obtain a gridded initial image matrix, where h is the side length of the grid and r is the estimated radius of a single dot; the estimated radius of a single dot refers to the estimated radius of the dot shape in the target anti-counterfeiting image, where each dot represents a polygon in the target anti-counterfeiting image.

[0103] Based on the coordinate information of the initial point, through To randomly generate k initial candidate points; where, This represents the i-th initial candidate point; Denote the candidate point selected from the set of candidate points. If the selected candidate point here is P0, then here represents P0; Denote the random direction vector within the unit circle (polar angle ), and d represents the radius. ;

[0104] For each initial candidate point , check whether it is within the region of the initial image matrix [0, N] x [0, N]. If it is out of bounds, discard it. If it meets the requirements, retain it to obtain the first screened point . For all the first screened points , perform the following operations. According to the grid coordinates where it is located, check all the grid cells within the 3x3 or 5x5 neighborhood centered on (usually check the 5x5 neighborhood to ensure safety); traverse each existing point in these neighborhood grid cells , calculate the Euclidean distance || and || between them; if there exists any point A- ||; if there exists any point such that || - || < r, then the first screened point is invalid and is discarded. If there does not exist any point such that || - || < r, then retain and add it to the set of candidate points as a target candidate point, and at the same time add it to the set of target points as a target center point; based on the obtained new set of candidate points, randomly select a candidate point from the new set of candidate points to loop and execute the above steps, that is, generate new initial candidate points around the candidate points in the new set of candidate points, and determine new target candidate points through the initial candidate points to determine the target center points in the set of target points; terminate until the set of candidate points is empty, that is, all points in the set of candidate points cannot find any candidate point. At this time, it means that no new position can be found to insert points that satisfy the minimum distance r; at the same time, a random point set that satisfies the minimum spacing r constraint can be obtained at this time, that is, the set of target points including multiple target center points; during the loop process, if after generating all k candidate points around , no valid point is successfully found, it is considered that the space around the point has been filled, and needs to be removed from the set of candidate points; if there exists any point such that || - || If r, then the candidate point is invalid and discarded;

[0105] According to the position parameter information of the target center points in the target point set, mark within the initial image matrix area, so as to obtain an initial image matrix including multiple randomly distributed target center points;

[0106] Triangulating the initial image matrix to obtain a triangular mesh specifically means that based on the target center points, determine a super triangle containing all the target center points within the initial image matrix as the initial mesh;

[0107] Each time insert a new target center point within the initial mesh, and delete all triangles whose circumcircles contain this new target center point;

[0108] Connect the new target center point with the vertices of the influence area to form new triangles, ensuring that the empty circle criterion is satisfied, that is, no target center point is contained within each generated new triangle, that is, no other target center point exists within the triangle formed by every three target center points; During the triangulation process, the mathematical properties that the initial image matrix needs to satisfy are as follows: The spatial property of the initial image matrix needs to satisfy that the circumcircle of any triangle ∆ABC does not contain other target center points, and its expression is as follows: For ∆ABC, ∀W i ∉Circle(A, B, C); W represents the set of all target center points; W i represents the i-th target center point in the set of all target center points; The circumcircle of ∆ABC is expressed as follows: ; where, ( ) represents the points within the circumcircle of ∆ABC; The circumcenter of ∆ABC ( , ) that is, the centroid coordinates of the triangle are obtained by solving the following formula: ;

[0109] ;

[0110] where, ( , ) represents the coordinates of point A in ∆ABC; ( , ) represents the coordinates of point B in ∆ABC; ( , ) represents the coordinates of point C in ∆ABC;

[0111] Calculate the area of each triangular mesh through the coordinates of the target center points corresponding to the triangular mesh;

[0112] Through Normalize each triangular mesh to obtain the normalized value corresponding to each triangular mesh, where, This represents the normalized value corresponding to the normalization process of the i-th triangular mesh; This represents the area of ​​the i-th triangular mesh. This represents the area of ​​the largest triangle among all triangular meshes;

[0113] Mapping the normalized value to an 8-bit deep integer is calculated as follows: using the formula The normalized value corresponding to each triangular grid is mapped to an 8-bit depth integer value to obtain the threshold corresponding to each triangular grid; where round() means rounding; This represents the threshold corresponding to the i-th triangular grid.

[0114] Through formula Gaussian noise jitter processing is applied to each threshold in the initial image matrix to obtain the first threshold processing data. ;in This is a general Gaussian calculation function;

[0115] Based on the triangular grid corresponding to the first threshold processing data, the first threshold processing data is directly overlaid onto the triangular grid area;

[0116] By scanning the first processed image matrix, each pixel of the first processed image matrix is ​​obtained;

[0117] Based on the first threshold processing data and the coordinate parameters of the pixels, the distance between each pixel and the first threshold processing data is determined, and the triangular grid corresponding to the pixel is determined based on the distance.

[0118] Based on the center point of each triangular grid, the pixel threshold corresponding to each pixel is determined according to the distance between the center point and each pixel in each triangular grid. The farther the distance, the larger the value, and the closer the distance, the smaller the value, thereby achieving accurate positioning of each pixel.

[0119] Based on the determined threshold for each pixel, the corresponding halftone image matrix can be directly generated.

[0120] The mesh image matrix is ​​laid flat onto the text and image content information to determine all pixels corresponding to the mesh image matrix and the text and image content information. At the same time, the first threshold corresponding to all corresponding pixels of the text and image content information and the pixel threshold corresponding to all corresponding pixels of the mesh image matrix are obtained.

[0121] The pixel threshold on the halftone image matrix is ​​compared with the first threshold corresponding to all pixels of the text and image content information. When the first threshold (gray value) corresponding to the pixel of the text and image content information is less than the gray value of the pixel threshold (gray value) on the halftone image matrix, the output is 1; otherwise, the output is 0. This process is repeated to obtain the binarized halftone dot image after halftone.

[0122] The binarized dot matrix after being screened is transmitted to the output device. The graphic content of the binarized dot map is transferred to the physical printing medium through the output transfer process to generate the target anti-counterfeiting image.

[0123] Anti-counterfeiting image recognition process:

[0124] The anti-counterfeiting images to be identified are captured using photographic equipment such as mobile phones and magnifying glasses;

[0125] Data is processed at the first threshold of the image matrix. A second threshold t is set, which in this embodiment is 82, and binarization is performed to obtain a preset recognition template matrix, labeled as T, and represented as: Where N represents the width and height dimensions of the halftone image matrix;

[0126] A preset recognition template matrix is ​​slid across the anti-counterfeiting image to be identified, and the similarity between the anti-counterfeiting image to be identified and the preset recognition template matrix at each sliding position is calculated. The similarity matching degree is calculated as follows:

[0127] ;

[0128] in, This represents the local color value at the position of row x+e and column y+v in the anti-counterfeiting image to be identified. x and y represent the local coordinates of the anti-counterfeiting image to be identified. These coordinates refer to the coordinates of the local center point of the anti-counterfeiting image to be identified; that is, the sliding position of the current preset identification template matrix on the anti-counterfeiting image to be identified. This represents the average color value of a local area in the anti-counterfeiting image to be identified. This represents the mean of the preset recognition template matrix; The matching score is the similarity score; the higher the score, the better the matching score.

[0129] The similarity is compared with the first preset threshold τ to determine the target image segment on the anti-counterfeiting image to be identified. Specifically, if C(x, y) > τ, it is determined that a preset identification template dot exists at that location, i.e., a dot in the hanging image matrix exists. The center (x1, y1) of all dots in the preset identification template matrix corresponding to the current local coordinate position of the anti-counterfeiting image to be identified is obtained and calibrated. Based on the center of all dots in the preset identification template matrix, the centroid of the dots corresponding to the center of the dots in the preset identification template matrix at all locations of the current local coordinate position of the anti-counterfeiting image to be identified is determined. The precise centroid of the dots can be calculated using the following formula. ,in, Where r is the approximate radius of a single halftone dot, i.e., the estimated radius of a single halftone dot; m and n represent the index values ​​of each pixel in the bounding rectangle of a single halftone dot; in this embodiment, =0.8, r is set to 10, and the single-site identification result is referenced. Figure 6 By sequentially sliding and comparing, all corresponding target image blocks on the anti-counterfeiting image to be identified are finally determined.

[0130] like Figure 5 As shown, during the sliding process, the anti-counterfeiting image to be identified is divided into multiple image blocks according to its size. Figure 5 In this model, each grid represents an image segment; based on the dots within the preset recognition template matrix, each target image segment is compared to obtain the comparison result for each target image segment.

[0131] like Figure 7 As shown, the coordinates (x1, y1) of the halftone dots in the preset recognition template matrix are respectively compared with the coordinates of the halftone dots in the target image block. A comparison is performed, and halftone dots with coordinate errors within a preset range are identified as successfully matched halftone dots, while halftone dots exceeding the preset range are identified as unmatched halftone dots. The matching degree is calculated as the number of successfully matched halftone dots divided by the sum of the number of successfully matched halftone dots and the number of unmatched halftone dots. The matching result of each target image segment is determined based on the matching degree of each target image segment and a preset matching value. When the matching degree is greater than the preset matching value (the preset matching value = 90%), the current target image segment is identified as halftone dot matched; when the matching degree is less than or equal to the preset matching value, the current target image segment is identified as halftone dot mismatched.

[0132] Based on the comparison results of each target image segment, the authenticity of the current anti-counterfeiting image to be identified is determined. Specifically, when the comparison results of all target image segments are dot mismatch, the current anti-counterfeiting image to be identified is determined to be a fake anti-counterfeiting image or a copy. When the comparison result of any target image segment among all target image segments is dot match, the current anti-counterfeiting image to be identified is determined to be a genuine target anti-counterfeiting image.

[0133] like Figure 8 As shown, embodiments of the present invention also provide a method for identifying anti-counterfeiting images, including:

[0134] Step 81: Obtain the anti-counterfeiting image to be identified;

[0135] Step 82: Determine the preset recognition template matrix based on the net image matrix of the target anti-counterfeiting image;

[0136] Step 83: Perform anti-counterfeiting identification on the anti-counterfeiting image to be identified according to the identification template.

[0137] In this embodiment, step 81 specifically involves acquiring the anti-counterfeiting image to be identified using photographic equipment such as a mobile phone or a magnifying glass; step 82 specifically involves determining the preset identification template matrix based on the net image matrix of the target anti-counterfeiting image, specifically by processing data at the first threshold of the net image matrix. A second threshold t is set and binarized to obtain a preset recognition template matrix, labeled T. In this embodiment, the second threshold t = 82, which is expressed as: Where N represents the width and height dimensions of the halftone image matrix; Let represent the set of recognition template matrices, where a and b represent the horizontal and vertical coordinates of pixels within the template, respectively. Step 83, based on the preset recognition template matrix, specifically involves performing anti-counterfeiting recognition on the anti-counterfeiting image to be recognized by sliding the preset recognition template matrix across the anti-counterfeiting image to be recognized, and calculating the similarity between the anti-counterfeiting image to be recognized and the preset recognition template matrix at each sliding position during the sliding process. The similarity matching degree is calculated as follows:

[0138] ,in, This represents the local color value at the position of row x+e and column y+v in the anti-counterfeiting image to be identified. x and y represent the local coordinates of the anti-counterfeiting image to be identified. These coordinates refer to the coordinates of the local center point of the anti-counterfeiting image to be identified; that is, the sliding position of the current preset identification template matrix on the anti-counterfeiting image to be identified. This represents the average color value of a local area in the anti-counterfeiting image to be identified. This represents the mean of the preset recognition template matrix; The matching score is the similarity score; the higher the score, the better the matching score.

[0139] The similarity is compared with the first preset threshold τ to determine the target image segment on the anti-counterfeiting image to be identified. Specifically, if C(x, y) > τ, it is determined that a preset identification template dot exists at that location, i.e., a dot in the hanging image matrix exists. The center (x1, y1) of all dots in the preset identification template matrix corresponding to the current local coordinate position of the anti-counterfeiting image to be identified is obtained and calibrated. Based on the center of all dots in the preset identification template matrix, the centroid of the dots corresponding to the center of the dots in the preset identification template matrix at all locations of the current local coordinate position of the anti-counterfeiting image to be identified is determined. The precise centroid of the dots can be calculated using the following formula. ,in, Where r is the approximate radius of a single halftone dot, i.e., the estimated radius of a single halftone dot; m and n represent the index values ​​of each pixel in the bounding rectangle of a single halftone dot; in this embodiment, =0.8, r is set to 10, and the single-site identification result is referenced. Figure 6 By sequentially sliding and comparing, all corresponding target image blocks on the anti-counterfeiting image to be identified are finally determined.

[0140] like Figure 5 As shown, during the sliding process, the anti-counterfeiting image to be identified is divided into multiple image blocks according to its size. Figure 5 In this model, each grid represents an image segment; based on the dots within the preset recognition template matrix, each target image segment is compared to obtain the comparison result for each target image segment.

[0141] like Figure 7 As shown, the coordinates (x1, y1) of the halftone dots in the preset recognition template matrix are respectively compared with the coordinates of the halftone dots in the target image block. A comparison is performed, and halftone dots with coordinate errors within a preset range are identified as successfully matched halftone dots, while halftone dots exceeding the preset range are identified as unmatched halftone dots. The matching degree is calculated as the number of successfully matched halftone dots divided by the sum of the number of successfully matched halftone dots and the number of unmatched halftone dots. The matching result of each target image segment is determined based on the matching degree of each target image segment and a preset matching value. When the matching degree is greater than the preset matching value (the preset matching value = 90%), the current target image segment is identified as halftone dot matched; when the matching degree is less than or equal to the preset matching value, the current target image segment is identified as halftone dot mismatched.

[0142] Based on the comparison results of each target image segment, the authenticity of the current anti-counterfeiting image to be identified is determined. Specifically, when the comparison results of all target image segments are dot mismatch, the current anti-counterfeiting image to be identified is determined to be a fake anti-counterfeiting image or a copy. When the comparison result of any target image segment among all target image segments is dot match, the current anti-counterfeiting image to be identified is determined to be a genuine target anti-counterfeiting image.

[0143] like Figure 9 As shown, embodiments of the present invention also provide an anti-counterfeiting image processing device 90, comprising:

[0144] The first acquisition module 91 is used to acquire the parameters and graphic content information of the target anti-counterfeiting image;

[0145] The first processing module 92 is used to generate a halftone image matrix based on the parameters of the target anti-counterfeiting image; fuse the halftone image matrix and the graphic content information to obtain a binarized halftone image; generate a target anti-counterfeiting image based on the binarized halftone image; and perform anti-counterfeiting recognition on the anti-counterfeiting image to be identified based on the halftone image matrix in the target anti-counterfeiting image.

[0146] Optionally, based on the parameters of the target anti-counterfeiting image, a mesh image matrix is ​​generated, including:

[0147] Based on the parameters of the target anti-counterfeiting image, determine an initial image matrix of a preset size;

[0148] Multiple target center points are generated randomly within the initial image matrix region;

[0149] Based on the target center point, a triangular mesh satisfying the empty circle property is determined within the initial image matrix for each target center point;

[0150] The triangular grid within the initial image matrix is ​​mapped to obtain the threshold corresponding to each triangular grid within the initial image matrix;

[0151] The meshing image matrix is ​​determined based on the threshold corresponding to each of the triangular grids in the initial image matrix.

[0152] Optionally, multiple randomly distributed target center points are generated within the initial image matrix region, including:

[0153] An initial point is randomly generated within the initial image matrix region, and the initial point is stored in the target point set and the candidate point set, respectively.

[0154] Based on the estimated radius of a single dot in the target anti-counterfeiting image, the region of the initial image matrix is ​​divided into grids to obtain a gridded initial image matrix.

[0155] Based on the initial points in the candidate point set and the gridded initial image matrix, a preset number of initial candidate points corresponding to the initial points in the candidate point set are determined;

[0156] Based on the location information of each initial candidate point, all initial candidate points are filtered to obtain target candidate points;

[0157] Based on the target candidate points, a set of target points including multiple target center points is determined;

[0158] Based on the set of target points, determine the center point of the target within the initial image matrix region.

[0159] Optionally, the triangular grid within the initial image matrix is ​​mapped to obtain a threshold corresponding to each triangular grid within the initial image matrix, including:

[0160] Determine the area of ​​each triangular grid within the initial image matrix region;

[0161] Based on the area of ​​each triangular grid, the normalization process is performed on each triangular grid to obtain the normalized value corresponding to each triangular grid;

[0162] The normalized value corresponding to each triangular grid is mapped to a preset bit depth integer value to obtain the threshold corresponding to each triangular grid.

[0163] Optionally, the meshing image matrix is ​​determined based on the threshold corresponding to each of the triangular grids within the initial image matrix, including:

[0164] Gaussian noise jitter processing is applied to each of the thresholds in the initial image matrix to obtain the first threshold processing data;

[0165] Within the initial image matrix, the first threshold processing data is overlaid on the triangular grid area corresponding to each first threshold processing data to obtain the first processed image matrix;

[0166] Obtain each pixel of the first processed image matrix;

[0167] Based on the first threshold processing data and pixel parameters, determine the triangular grid corresponding to each pixel;

[0168] Based on the center point of each triangular grid, determine the pixel threshold corresponding to each pixel within each triangular grid; and determine the mesh image matrix based on the pixel threshold corresponding to each pixel.

[0169] Optionally, the halftone image matrix and the text and image content information are fused to obtain a binarized halftone image, including: obtaining all corresponding pixels of the halftone image matrix and the text and image content information; comparing the halftone image matrix with all corresponding pixels of the text and image content information to obtain a binarized halftone image.

[0170] Optionally, based on the mesh image matrix in the target anti-counterfeiting image, anti-counterfeiting recognition is performed on the anti-counterfeiting image to be identified, including: acquiring the anti-counterfeiting image to be identified; determining a preset recognition template matrix according to the mesh image matrix of the target anti-counterfeiting image; and performing anti-counterfeiting recognition on the anti-counterfeiting image to be identified according to the recognition template.

[0171] It should be noted that this device is the same as the above-mentioned anti-counterfeiting image processing method. All implementation methods described above are applicable to the embodiments of this device and can achieve the same technical effect.

[0172] like Figure 10 As shown, embodiments of the present invention also provide an anti-counterfeiting image recognition device 10, comprising:

[0173] The second acquisition module 101 is used to acquire the anti-counterfeiting image to be identified;

[0174] The second processing module 102 is used to determine a preset recognition template matrix based on the net image matrix of the target anti-counterfeiting image; and to perform anti-counterfeiting recognition on the anti-counterfeiting image to be recognized based on the recognition template.

[0175] Specifically, acquiring the anti-counterfeiting image to be identified involves using photographic equipment such as a mobile phone or magnifying glass to capture the image. Determining the preset recognition template matrix based on the net image matrix of the target anti-counterfeiting image specifically involves processing data at a first threshold of the net image matrix. A second threshold t is set and binarized to obtain a preset recognition template matrix, labeled T. In this embodiment, the second threshold t = 82, which is expressed as: Where N represents the width and height dimensions of the halftone image matrix; Let represent the set of recognition template matrices, where a and b represent the horizontal and vertical coordinates of pixels within the template, respectively. Step 53, based on the preset recognition template matrix, specifically involves performing anti-counterfeiting recognition on the anti-counterfeiting image to be recognized by sliding the preset recognition template matrix across the anti-counterfeiting image to be recognized, and calculating the similarity between the anti-counterfeiting image to be recognized and the preset recognition template matrix at each sliding position during the sliding process. The similarity matching degree is calculated as follows:

[0176] ,in, This represents the local color value at the position of row x+e and column y+v in the anti-counterfeiting image to be identified. x and y represent the local coordinates of the anti-counterfeiting image to be identified. These coordinates refer to the coordinates of the local center point of the anti-counterfeiting image to be identified; that is, the sliding position of the current preset identification template matrix on the anti-counterfeiting image to be identified. This represents the average color value of a local area in the anti-counterfeiting image to be identified. This represents the mean of the preset recognition template matrix; The matching score is the similarity score; the higher the score, the better the matching score.

[0177] The similarity is compared with the first preset threshold τ to determine the target image segment on the anti-counterfeiting image to be identified. Specifically, if C(x, y) > τ, it is determined that a preset identification template dot exists at that location, i.e., a dot in the hanging image matrix exists. The center (x1, y1) of all dots in the preset identification template matrix corresponding to the current local coordinate position of the anti-counterfeiting image to be identified is obtained and calibrated. Based on the center of all dots in the preset identification template matrix, the centroid of the dots corresponding to the center of the dots in the preset identification template matrix at all locations of the current local coordinate position of the anti-counterfeiting image to be identified is determined. The precise centroid of the dots can be calculated using the following formula. ,in, Where r is the approximate radius of a single halftone dot, i.e., the estimated radius of a single halftone dot; m and n represent the index values ​​of each pixel in the bounding rectangle of a single halftone dot; in this embodiment, =0.8, r is set to 10, and the single-site identification result is referenced. Figure 6 By sequentially sliding and comparing, all corresponding target image blocks on the anti-counterfeiting image to be identified are finally determined.

[0178] like Figure 5 As shown, during the sliding process, the anti-counterfeiting image to be identified is divided into multiple image blocks according to its size. Figure 5 In this model, each grid represents an image segment; based on the dots within the preset recognition template matrix, each target image segment is compared to obtain the comparison result for each target image segment.

[0179] like Figure 7 As shown, the coordinates (x1, y1) of the halftone dots in the preset recognition template matrix are respectively compared with the coordinates of the halftone dots in the target image block. A comparison is performed, and halftone dots with coordinate errors within a preset range are identified as successfully matched halftone dots, while halftone dots exceeding the preset range are identified as unmatched halftone dots. The matching degree is calculated as the number of successfully matched halftone dots divided by the sum of the number of successfully matched halftone dots and the number of unmatched halftone dots. The matching result of each target image segment is determined based on the matching degree of each target image segment and a preset matching value. When the matching degree is greater than the preset matching value (the preset matching value = 90%), the current target image segment is identified as halftone dot matched; when the matching degree is less than or equal to the preset matching value, the current target image segment is identified as halftone dot mismatched.

[0180] Based on the comparison results of each target image segment, the authenticity of the current anti-counterfeiting image to be identified is determined. Specifically, when the comparison results of all target image segments are dot mismatch, the current anti-counterfeiting image to be identified is determined to be a fake anti-counterfeiting image or a copy. When the comparison result of any target image segment among all target image segments is dot match, the current anti-counterfeiting image to be identified is determined to be a genuine target anti-counterfeiting image.

[0181] It should be noted that this device is the same as the anti-counterfeiting image recognition method described above. All implementation methods described above are applicable to the embodiments of this device and can achieve the same technical effect.

[0182] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for processing anti-counterfeiting images, characterized in that, include: Obtain the parameters and graphic content information of the target anti-counterfeiting image; Based on the parameters of the target anti-counterfeiting image, a net image matrix is ​​generated; The halftone image matrix and the text and image content information are fused to obtain a binarized halftone image; Based on the binarized dot pattern, generate the target anti-counterfeiting image; Based on the mesh image matrix in the target anti-counterfeiting image, anti-counterfeiting recognition is performed on the anti-counterfeiting image to be identified. The process of generating a mesh image matrix based on the parameters of the target anti-counterfeiting image includes: Based on the parameters of the target anti-counterfeiting image, determine an initial image matrix of a preset size; Multiple target center points are generated randomly within the initial image matrix region; Based on the target center point, a triangular mesh satisfying the empty circle property is determined within the initial image matrix for each target center point; The triangular grid within the initial image matrix is ​​mapped to obtain the threshold corresponding to each triangular grid within the initial image matrix; The meshing image matrix is ​​determined based on the threshold corresponding to each of the triangular grids in the initial image matrix; The process of mapping the triangular grid within the initial image matrix to obtain the threshold corresponding to each triangular grid within the initial image matrix includes: Determine the area of ​​each triangular grid within the initial image matrix region; Based on the area of ​​each triangular grid, the normalization process is performed on each triangular grid to obtain the normalized value corresponding to each triangular grid; The normalized value corresponding to each triangular grid is mapped to a preset bit depth integer value to obtain the threshold corresponding to each triangular grid.

2. The method for processing anti-counterfeiting images according to claim 1, characterized in that, Multiple target center points are generated randomly within the initial image matrix region, including: An initial point is randomly generated within the initial image matrix region, and the initial point is stored in the target point set and the candidate point set, respectively. Based on the estimated radius of a single dot in the target anti-counterfeiting image, the region of the initial image matrix is ​​divided into grids to obtain a gridded initial image matrix. Based on the initial points in the candidate point set and the gridded initial image matrix, a preset number of initial candidate points corresponding to the initial points in the candidate point set are determined; Based on the location information of each initial candidate point, all initial candidate points are filtered to obtain target candidate points; Based on the target candidate points, a set of target points including multiple target center points is determined; Based on the set of target points, determine the center point of the target within the initial image matrix region.

3. The method for processing anti-counterfeiting images according to claim 1, characterized in that, The meshing image matrix is ​​determined based on the threshold corresponding to each of the triangular grids within the initial image matrix, including: Gaussian noise jitter processing is applied to each of the thresholds in the initial image matrix to obtain the first threshold processing data; Within the initial image matrix, the first threshold processing data is overlaid on the triangular grid area corresponding to each first threshold processing data to obtain the first processed image matrix; Obtain each pixel of the first processed image matrix; Based on the first threshold processing data and pixel parameters, determine the triangular grid corresponding to each pixel; Based on the center point of each triangular grid, determine the pixel threshold corresponding to each pixel within each triangular grid; The halftone image matrix is ​​determined based on the pixel threshold corresponding to each pixel.

4. The method for processing anti-counterfeiting images according to claim 1, characterized in that, The halftone image matrix and the text and image content information are fused to obtain a binarized halftone image, including: Obtain all corresponding pixels of the image matrix and text content information; The halftone image matrix is ​​compared with all corresponding pixels of the text and image content information to obtain a binarized halftone image.

5. A method for recognizing anti-counterfeiting images, characterized in that, include: Obtain the anti-counterfeiting image to be identified; Based on the net image matrix of the target anti-counterfeiting image, determine the preset recognition template matrix; Based on the identification template, the anti-counterfeiting image to be identified is subjected to anti-counterfeiting identification; Wherein, the anti-counterfeiting image to be identified is obtained by the method described in any one of claims 1 to 4.

6. A device for processing anti-counterfeiting images, characterized in that, include: The first acquisition module is used to acquire the parameters and graphic content information of the target anti-counterfeiting image; The first processing module is used to generate a net image matrix based on the parameters of the target anti-counterfeiting image; The halftone image matrix and the text and image content information are fused to obtain a binarized halftone image; a target anti-counterfeiting image is generated based on the binarized halftone image. Based on the mesh image matrix in the target anti-counterfeiting image, anti-counterfeiting recognition is performed on the anti-counterfeiting image to be identified. The process of generating a mesh image matrix based on the parameters of the target anti-counterfeiting image includes: Based on the parameters of the target anti-counterfeiting image, determine an initial image matrix of a preset size; Multiple target center points are generated randomly within the initial image matrix region; Based on the target center point, a triangular mesh satisfying the empty circle property is determined within the initial image matrix for each target center point; The triangular grid within the initial image matrix is ​​mapped to obtain the threshold corresponding to each triangular grid within the initial image matrix; The meshing image matrix is ​​determined based on the threshold corresponding to each of the triangular grids in the initial image matrix; The process of mapping the triangular grid within the initial image matrix to obtain the threshold corresponding to each triangular grid within the initial image matrix includes: Determine the area of ​​each triangular grid within the initial image matrix region; Based on the area of ​​each triangular grid, the normalization process is performed on each triangular grid to obtain the normalized value corresponding to each triangular grid; The normalized value corresponding to each triangular grid is mapped to a preset bit depth integer value to obtain the threshold corresponding to each triangular grid.

7. A device for recognizing anti-counterfeiting images, characterized in that, include: The second acquisition module is used to acquire the anti-counterfeiting image to be identified; The second processing module is used to determine the preset recognition template matrix based on the net image matrix of the target anti-counterfeiting image; Based on the identification template, the anti-counterfeiting image to be identified is subjected to anti-counterfeiting identification; Wherein, the anti-counterfeiting image to be identified is obtained by the method described in any one of claims 1 to 4.

Images