CN115270840A - Lattice code counterfeit identification method and system - Google Patents

Abstract

The invention relates to a lattice code false distinguishing method, which comprises the following steps: acquiring an image to be processed, wherein the image to be processed comprises a coded graph; extracting the graphic characteristics of the coded graph, wherein the graphic characteristics comprise at least one of code point size characteristics, code point shape characteristics, code point brightness characteristics, code point edge gradient characteristics, code point density characteristics and code point spacing characteristics; pre-constructing and training a counterfeit identification model; and inputting the graph characteristics to an authentication model to obtain an authentication result. The invention extracts the graphic characteristics reflecting the printing characteristics in the coded graphic, and judges the authenticity of the coded graphic through the counterfeit identification model, thereby being suitable for the anti-counterfeiting requirements of small batch and high precision; in addition, the invention does not need to store the image or the characteristic information related to the coded graph in the background, and has small required storage space and stronger real-time property.

Description

Lattice code counterfeit identification method and system

Technical Field

The invention relates to a dot matrix code authenticity identification method and a dot matrix code authenticity identification system, and belongs to the field of dot matrix code authenticity identification.

Background

The two-dimensional code, the dot code and other coding patterns can conveniently transmit anti-counterfeiting information to consumers, so that the anti-counterfeiting label is widely applied to various scenes and commodities as the anti-counterfeiting label. However, these coding patterns are very easy to copy and counterfeit, so that it is impossible to effectively prevent others from copying and forging, and there is no security. Therefore, a method for identifying the authenticity of the encoded pattern is required.

Currently, there are several types of prior art: 1) The time and the times of scanning and reading the code image are recorded, so that whether the code image is secondarily used or not, namely whether the code image is copied or not is judged. 2) The copying difficulty of the coded image is improved by using the techniques of special ink, paper lines, engraved intaglio printing and the like, but the cost is high, and special equipment and professionals are required for identification. 3) Pre-storing the image or characteristic information of the generated coding graph; and when the image is identified, acquiring the image or the characteristic information of the coded graph to be identified, comparing the image or the characteristic information with pre-stored data, and if the comparison is consistent, determining that the coded graph is not authentic. See the following two patents:

the patent publication No. CN110533704A, namely 'identification and verification method, device, equipment and medium of ink labels', discloses: identifying the ink label image to extract ink characteristic information; comparing the characteristics of the ink with the characteristics of a pre-stored verification template to obtain a characteristic comparison result; and returning a false checking result according to the characteristic comparison result.

The technical scheme of the patent needs to store a large amount of code pattern related data in the background of the system.

Patent publication No. CN113888198A, "anti-counterfeiting method based on anti-counterfeiting feature correction" discloses: after carrying out anti-counterfeiting characteristic updating correction on the printed anti-counterfeiting code image, forming an anti-counterfeiting detection comparison file serving as a commodity authenticity identification basis and storing the anti-counterfeiting detection comparison file; binding the anti-counterfeiting detection comparison file with the code-assigning parameters of the printed anti-counterfeiting code image to form a binding relationship and storing the binding relationship; collecting an anti-counterfeiting code image printed on a printed matter, and performing anti-counterfeiting feature decoding and coding parameter analysis; and acquiring an anti-counterfeiting detection comparison file corresponding to the acquired anti-counterfeiting code image which has a binding relationship with the encoding parameter from a database according to the analyzed encoding parameter, performing similarity calculation on the anti-counterfeiting feature obtained by decoding and the anti-counterfeiting feature recorded in the anti-counterfeiting detection comparison file, and judging whether the product printed with the anti-counterfeiting code image is true or not according to the calculation result.

According to the technical scheme, the anti-counterfeiting characteristic correction is performed on the printed anti-counterfeiting code image on the basis of a patent CN110533704A, the influence of different coding processes on the anti-counterfeiting characteristic identification accuracy is reduced, and a large amount of code image related data still need to be stored in a system background.

Disclosure of Invention

In order to overcome the problems in the prior art, the invention designs a dot matrix code counterfeit identification method and a dot matrix code counterfeit identification system, which are used for extracting the pattern characteristics reflecting the printing characteristics in the coded pattern and judging the authenticity of the coded pattern through a counterfeit identification model, and are suitable for the anti-counterfeiting requirements of small batch and high precision; in addition, the invention does not need to store the image or the characteristic information related to the coded graph in the background, and has small required storage space and stronger real-time property.

In order to achieve the purpose, the invention adopts the following technical scheme:

the first technical scheme is as follows:

a lattice code false distinguishing method comprises the following steps:

acquiring an image to be processed, wherein the image to be processed comprises a coded graph;

extracting the graphic characteristics of the coded graph, wherein the graphic characteristics comprise at least one of code point size characteristics, code point shape characteristics, code point brightness characteristics, code point edge gradient characteristics, code point density characteristics and code point spacing characteristics;

pre-constructing and training a counterfeit identification model;

and inputting the graph characteristics to an authentication model to obtain an authentication result.

Further, the coding graph is a one-dimensional code, a two-dimensional code, a three-dimensional code or a dot matrix code.

Further, the counterfeit identification model comprises a plurality of decision trees; inputting the image characteristics to each decision tree respectively, and outputting voting results by each decision tree; and obtaining an authenticity result according to a plurality of voting results.

Further, the method for constructing and training the counterfeit identification model comprises the following specific steps:

s1, constructing a sample set, wherein the sample set comprises a plurality of positive samples and a plurality of negative samples;

s2, randomly selecting N samples from the sample set to train a decision tree; in the training process, randomly selecting M graphic features as decision tree nodes;

and S3, repeating the step S2 to obtain the counterfeit identification model comprising a plurality of decision trees.

Further, the extraction step of the code point shape feature is as follows:

calculating the mean or variance of at least one diameter of the code point; and comparing the mean value or the variance of the diameter with a preset threshold value, and assigning the shape characteristics of the code points according to the comparison result.

Further, the extraction step of the code point edge gradient feature is as follows:

extracting code point edges; and taking two pixel points in the diameter direction of the edge of the code point, and calculating the ratio of the gray scale variation of the pixel points to the distance of the pixel points.

Further, the coding graph comprises a positioning module and a data module; the code point spacing characteristic is specifically a code point spacing ratio or a code point spacing average value of the positioning module.

The second technical scheme is as follows:

a lattice code counterfeit discrimination system comprises:

the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring an image to be processed, and the image to be processed comprises a coding graph;

the characteristic extraction unit is used for extracting the graphic characteristics of the coded graph, and the graphic characteristics comprise at least one of code point size characteristics, code point shape characteristics, code point brightness characteristics, code point edge gradient characteristics, code point density characteristics and code point spacing characteristics;

and the counterfeit identifying unit is provided with a counterfeit identifying model, and inputs the graph characteristics to the counterfeit identifying model to obtain a counterfeit identifying result.

Compared with the prior art, the invention has the following characteristics and beneficial effects:

1. the invention extracts the graphic characteristics reflecting the printing characteristics in the coded graphic, judges the authenticity of the coded graphic through the authenticity identification model, and is suitable for the anti-counterfeiting requirements of small batch and high precision; in addition, the invention does not need to store images or characteristic information related to the coded graphics in the background, and has small required storage space and stronger real-time property.

2. The false distinguishing model constructed by the invention comprises a plurality of weak classifiers (namely decision trees), each decision tree independently judges to obtain a voting result and takes the voting result with the largest quantity as a false distinguishing result, so that the influence of the randomness of printing characteristics on the false distinguishing result can be effectively reduced.

3. The invention takes the characteristics of constructing the size of code points, the shape of code points, the brightness of code points, the edge gradient of code points, the density of code points and the space of code points as the graphic characteristics, and has the advantages that: a. the extraction speed of the graphic features is high, and the real-time performance is strong. b. The graphic characteristics are closely combined with the printing process, and compared with an imitation product and a genuine product, the characteristics of the imitation product and the genuine product are greatly different in the brightness, size, shape and the like of code points. c. Compared with the one-to-one corresponding relation between the characteristics such as code point position characteristics, code point texture characteristics and the like and the coding graph (once the coding data is changed, the code point arrangement of the coding graph is changed), the graph characteristics extracted by the method do not need to update the model due to the change of the coding graph, and the counterfeit identification method has strong generalization capability.

Drawings

FIG. 1 is a flow chart of the present invention.

Detailed Description

The present invention will be described in more detail with reference to examples.

In one embodiment, as shown in fig. 1, a method for authenticating a lattice code includes the following steps:

and generating a coding graph. Preferably, the coding pattern is a dot matrix code; for the encoding rule of the lattice code, reference may be made to "a method for generating and decoding lattice codes and an anti-counterfeit method" in patent publication No. CN 113435556A. The dot matrix code comprises a positioning module and a data module. The code point spacing or the code point spacing ratio in the positioning module is different from the code point spacing or the code point spacing ratio in the data module.

The printing resolution of the ink jet printer was set to 812dpi, and a code pattern was printed, each code dot in the code pattern having a diameter of 2 pixels.

And acquiring the image to be processed containing the coding graph.

And extracting the graphic characteristics of the coded graph, wherein the graphic characteristics comprise code point size characteristics, code point shape characteristics, code point brightness characteristics, code point edge gradient characteristics, code point density characteristics and code point spacing characteristics.

And constructing and training the counterfeit identification model in advance. Preferably, the authentication model is a model comprising a plurality of decision trees.

Inputting the graphic features into each decision tree respectively, and outputting voting results by each decision tree; and obtaining the voting result with the largest quantity as the authenticity identifying result.

In one embodiment, the method for extracting the graphic features of the encoded graphic specifically includes the following steps:

quickly identifying all code points in the image to be processed:

determining a scanning line; searching a first code point along a scanning line; calculating LBP characteristics of pixel points at the centroid of the first code point to obtain first LBP characteristics; continuously calculating the LBP characteristics of each pixel point along the scanning line to obtain second LBP characteristics; and calculating the Hamming distance between the first LBP characteristic and the second LBP characteristic, and if the Hamming distance is smaller than a distance threshold value, considering that the pixel point corresponding to the second LBP characteristic is the centroid of the code point, thereby completing the positioning identification of the new code point.

Randomly selecting a plurality of code points; preferably, five code points are chosen. The following graphical features were extracted:

code point size characteristics: and calculating the average value of the pixel numbers occupied by the diameters or the side lengths of a plurality of code points. As in the present embodiment, the code point size characteristics are extracted for five code points, and the average value of the number of code point diameter pixels is calculated to be 2.

Code point shape characteristics: taking four diameters passing through the transverse and longitudinal directions and plus and minus 45 degrees of the centroid of the code point, calculating the mean value and the variance of the four diameters, and if the mean value and the variance are both smaller than a preset mean value threshold value and a preset variance threshold value, considering that the shape is circular and the shape characteristic value is 1; otherwise, the shape feature value is 0. As in the present embodiment, the code point shape features are extracted from five code points to obtain code point shape features 11110, which show that four are circular and one is non-circular.

Code point luminance characteristics: converting an image to be processed into a YUV color space; and calculating the average value of Y components of a plurality of code points, namely the average value of brightness components. As in this embodiment, the luminance features of the code points are extracted from five code points, and the average value of the calculated luminance components is 10.

Code point edge gradient feature: and calculating the ratio t of the gray variation and the distance of the pixel points at two sides of the edge of the code point in different diameter directions (such as the horizontal and vertical directions and the directions of plus and minus 45 degrees), wherein the ratio t is expressed as t = delta y/delta x by a formula, delta y represents the gray variation, and delta x represents the distance. And taking the edge gradient characteristic of the code point with the maximum ratio. In this embodiment, the edge gradient feature of the code point is extracted from five code points, and the maximum value (assumed to be 80) of the edge gradient feature in the five code points is taken as the final edge gradient feature of the code point.

Code point density characteristics: the number of code points in a unit area is calculated. In this embodiment, the number of code points calculated in a unit area (square decimeter) is 30, which is a code point density characteristic.

Code point spacing characteristics: and calculating the code point spacing proportion or the average value of the code point spacing in the lattice code positioning module. If the average code dot spacing is 10 pixels, the code dot spacing feature 10 is obtained.

In summary, in this embodiment, the feature vectors of the finally obtained patterns are <2,11110,10,80,30,10>.

In one embodiment, the method for constructing and training the counterfeit identification model comprises the following specific steps:

constructing a sample set: shooting a plurality of coded graphic prints to obtain a group A of pictures; and extracting graphic features of each picture to obtain a group A of positive samples. Shooting a plurality of code pattern copies to obtain a group B of pictures; and extracting the graphic characteristics of each picture to obtain a group B of negative samples.

And the replaced randomly drawn N samples are used as training samples of a decision tree, and A is far larger than N.

And (3) constructing a decision tree according to the samples, randomly extracting m graphic features before splitting the decision tree each time, and taking the most preferable one of the m graphic features as a node (namely as a partition attribute). Supposing that m graphic characteristics such as code point size characteristics and code point density characteristics are randomly extracted, wherein the information gain of the code point size characteristics is maximum; then the code point size characteristic of the positive sample is taken as a node, the code point size characteristic of the positive sample is taken as a first branch of a first node in the decision tree, and the code point size characteristic of the negative sample is taken as a second branch of the first node in the decision tree. And in the rest (m-1) graph features, if the information gain of the code point density feature is maximum, taking the code point density feature of the positive sample as a first branch of a second node in the decision tree, and taking the code point density feature of the negative sample as a second branch of the second node in the decision tree. And repeating the steps until the samples belong to the same category under the current partition attribute, and the samples cannot be continuously divided downwards, and marking the current nodes as leaf nodes to obtain at least one decision tree, wherein each decision tree in the at least one decision tree comprises a plurality of nodes. And combining at least one decision tree to obtain the authenticity identification model.

Corresponding to the dot matrix code counterfeit distinguishing method, the dot matrix code counterfeit distinguishing system comprises an acquisition unit, a feature extraction unit and a counterfeit distinguishing unit;

the acquisition unit is used for acquiring an image to be processed, and the image to be processed comprises a coding graph;

the feature extraction unit is used for extracting the graphic features of the coded graphics, and the graphic features comprise at least one of code point size features, code point shape features, code point brightness features, code point edge gradient features, code point density features and code point spacing features;

the counterfeit identifying unit is provided with a counterfeit identifying model, and the pattern characteristics are input into the counterfeit identifying model to obtain a counterfeit identifying result.

It should be noted that, the above proposed lattice code counterfeit identification system is also used for implementing the corresponding method steps of each embodiment of the lattice code counterfeit identification method shown in fig. 1, and the description of the present application is not repeated here.

It should be noted that, functional units/modules in the embodiments of the present invention may be integrated into one processing unit/module, or each unit/module may exist alone physically, or two or more units/modules are integrated into one unit/module. The integrated units/modules may be implemented in the form of hardware, or may be implemented in the form of software functional units/modules.

From the above description of the embodiments, it is clear for a person skilled in the art that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code or any appropriate combination thereof. For a hardware implementation, a processor may be implemented in one or more of the following units: an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to perform the functions described herein, or a combination thereof. For a software implementation, some or all of the flow of the embodiments may be accomplished by a computer program instructing the associated hardware. In practice, the program may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a computer. Computer-readable media can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.

Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the protection scope of the present invention, although the present invention is described in detail with reference to the preferred embodiments, it should be analyzed by those skilled in the art that modifications or equivalent substitutions can be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims (8)

1. A dot matrix code authenticity identification method is characterized by comprising the following steps:

acquiring an image to be processed, wherein the image to be processed comprises a coded graph;

extracting the graphic characteristics of the coded graph, wherein the graphic characteristics comprise at least one of code point size characteristics, code point shape characteristics, code point brightness characteristics, code point edge gradient characteristics, code point density characteristics and code point spacing characteristics;

pre-constructing and training a counterfeit identification model;

and inputting the graph characteristics to an authenticity identification model to obtain an authenticity identification result.

2. The method as claimed in claim 1, wherein the code pattern is a one-dimensional code, a two-dimensional code, a three-dimensional code or a dot code.

3. A lattice code authentication method as claimed in claim 1, wherein said authentication model comprises a plurality of decision trees; inputting the image characteristics to each decision tree respectively, and outputting voting results by each decision tree; and obtaining an authenticity result according to a plurality of voting results.

4. The lattice code counterfeit discrimination method according to claim 1, wherein the construction and training of the counterfeit discrimination model specifically comprises the following steps:

s1, constructing a sample set, wherein the sample set comprises a plurality of positive samples and a plurality of negative samples;

s2, randomly selecting N samples from the sample set to train a decision tree; in the training process, randomly selecting M graphic features as decision tree nodes;

and S3, repeating the step S2 to obtain the counterfeit identification model comprising a plurality of decision trees.

5. A lattice code authentication method as claimed in claim 1, wherein said extracting of the shape features of the code points comprises:

calculating the mean or variance of at least one diameter of the code point; and comparing the mean value or the variance of the diameter with a preset threshold value, and assigning the shape characteristics of the code points according to the comparison result.

6. The lattice code counterfeit discrimination method according to claim 1, wherein the step of extracting the gradient feature of the code point edge is as follows:

extracting code point edges; and taking two pixel points in the diameter direction of the edge of the code point, and calculating the ratio of the gray scale variation of the pixel points to the distance of the pixel points.

7. A lattice code authentication method as claimed in claim 1, wherein said code pattern comprises a positioning module and a data module; the code point spacing characteristic is specifically a code point spacing ratio or a code point spacing average value of the positioning module.

8. A dot matrix code authenticity identification system is characterized by comprising:

the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring an image to be processed, and the image to be processed comprises a coding graph;

the characteristic extraction unit is used for extracting graphic characteristics of the coded graphics, and the graphic characteristics comprise at least one of code point size characteristics, code point shape characteristics, code point brightness characteristics, code point edge gradient characteristics, code point density characteristics and code point spacing characteristics;

and the counterfeit identifying unit is provided with a counterfeit identifying model, and inputs the graph characteristics to the counterfeit identifying model to obtain a counterfeit identifying result.

Images