Method and device for identifying currency authenticity

Abstract

This application provides a method and device for currency authenticity identification, which relates to the field of artificial intelligence and can also be used in the financial field. The method comprises: constructing a currency information knowledge graph based on collected images of real currency; and using the currency information knowledge graph to verify the basic information and texture information of the currency to be verified, thereby obtaining a currency authenticity identification result. This application can construct a currency information knowledge graph and use it to identify currency authenticity.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence and can be used in the financial field. Specifically, it provides a method and device for identifying the authenticity of currency. Background Art

[0002] Authenticating banknotes and coins is a crucial aspect of bank counter services. Before internet technology became widespread, bank deposits and withdrawals were typically handled through the counter. Bank staff relied on experience and knowledge to identify genuine counterfeit banknotes, while also relying on counterfeit verification machines to verify authenticity. Today, all major banks are equipped with automated teller machines (ATMs) both inside and outside their branches. ATMs allow customers to deposit and withdraw cash. In daily life, ATMs often fail to process banknotes due to wrinkles, folds, damage, or stains on banknotes. In some cases, cash withdrawals can be difficult to re-deposit. This can lead to a significant decline in customer deposit experience, forcing customers to flock to counter services, which in turn increases the workload for bank staff.

[0003] Especially when customers use foreign currency to conduct related business at the bank, the authenticity of the currency can only be determined by the experience of the counter staff. In this context, there is an urgent need for an intelligent, convenient, and fast intelligent genuine and counterfeit currency identification system to quickly assist the staff in completing related business and improve the service efficiency of cash business. Summary of the Invention

[0004] In response to the problems in the existing technology, the present application provides a method and device for identifying the authenticity of currency, which can construct a currency information knowledge graph and identify the authenticity of currency based on it.

[0005] To solve the above technical problems, this application provides the following technical solutions:

[0006] In a first aspect, the present application provides a method for identifying the authenticity of currency, comprising:

[0007] Build a currency information knowledge graph based on the collected real currency images;

[0008] The currency information knowledge graph is used to verify the basic information and texture information of the currency to be tested, and a currency authenticity identification result is obtained.

[0009] Furthermore, the process of constructing a currency information knowledge graph based on the collected real currency images includes:

[0010] Extracting real currency attributes based on the real currency image; the real currency attributes include currency type, denomination, serial number, fineness grade, real plane texture features, and real three-dimensional texture features;

[0011] generating a relationship between real currencies according to the real currency attributes;

[0012] The currency information knowledge graph is generated according to the real currency, the real currency attributes and the relationship between the real currencies.

[0013] Furthermore, the method of verifying the basic information of the currency to be verified by using the currency information knowledge graph includes:

[0014] According to the currency type and denomination of the currency to be verified, the real currency of the corresponding currency type and denomination in the currency information knowledge graph is retrieved to obtain the serial number and fineness grade of the real currency of the corresponding currency type and denomination;

[0015] The serial number and fineness of the real currency of the corresponding currency type and denomination are compared with the serial number and fineness of the currency to be verified to obtain the basic information verification result of the currency to be verified.

[0016] Furthermore, the texture information includes plane texture information; and the verifying the basic information and texture information of the currency to be verified by using the currency information knowledge graph includes:

[0017] Fusing the plane texture features of the currency to be verified with the real plane texture features of the real currency of the corresponding currency type and denomination to obtain the plane fusion features to be judged;

[0018] Fusing the real plane texture features of real currency of corresponding currency type and denomination to obtain the real plane fusion features;

[0019] The deviation between the plane fusion feature to be judged and the real plane fusion feature is calculated to generate a plane texture verification result corresponding to the plane texture information.

[0020] Furthermore, the texture information includes three-dimensional texture information; and the verifying the basic information and texture information of the currency to be verified by using the currency information knowledge graph includes:

[0021] Fusing the 3D texture features of the currency to be verified with the real 3D texture features of the real currency of the corresponding currency type and denomination to obtain a 3D fusion feature to be judged;

[0022] Fusing the real 3D texture features of real currency of corresponding currency type and denomination to obtain real 3D fusion features;

[0023] The information entropy between the to-be-determined stereo fusion feature and the real stereo fusion feature is calculated to generate a stereo texture verification result corresponding to the stereo texture information.

[0024] Furthermore, the currency authenticity identification method further includes:

[0025] Performing early warning processing according to the basic information verification result and the plane texture verification result;

[0026] If the basic information verification result fails and / or the plane texture verification result fails, a high-risk warning process is performed.

[0027] Furthermore, the currency authenticity identification method further includes:

[0028] Performing early warning processing according to the basic information verification result, the plane texture verification result, and the three-dimensional texture verification result;

[0029] If the basic information verification result fails, the plane texture verification result and the three-dimensional texture verification result all fail, a high-risk warning process is performed;

[0030] If the basic information verification result passes, but the plane texture verification result and the three-dimensional texture verification result both fail, an intermediate warning process is performed;

[0031] If both the basic information verification result and the plane texture verification result pass, but the three-dimensional texture verification result fails, a low-risk warning process is performed.

[0032] In a second aspect, the present application provides a currency authenticity identification device, comprising:

[0033] A knowledge graph construction unit, used to construct a currency information knowledge graph based on the collected real currency images;

[0034] The recognition result generating unit is used to verify the basic information and texture information of the currency to be tested by using the currency information knowledge graph to obtain the currency authenticity recognition result.

[0035] Furthermore, the knowledge graph construction unit includes:

[0036] a real attribute extraction module, configured to extract real currency attributes based on the real currency image; the real currency attributes include currency type, denomination, serial number, fineness grade, real plane texture features, and real three-dimensional texture features;

[0037] A real relationship extraction module, configured to generate a relationship between real currencies based on the real currency attributes;

[0038] The knowledge graph construction module is used to generate the currency information knowledge graph based on the real currency, the real currency attributes and the relationship between the real currencies.

[0039] Furthermore, the recognition result generating unit includes:

[0040] A real currency selection module is used to retrieve the real currency of the corresponding currency and denomination in the currency information knowledge graph according to the currency and denomination of the currency to be verified, and obtain the serial number and fineness grade of the real currency of the corresponding currency and denomination;

[0041] The basic information verification module is used to compare the serial number and fineness of the real currency of the corresponding currency type and denomination with the serial number and fineness of the currency to be verified, and obtain the basic information verification result of the currency to be verified.

[0042] Furthermore, the texture information includes plane texture information; and the recognition result generating unit includes:

[0043] a plane feature generation module for determining the currency to be determined, configured to fuse the plane texture features of the currency to be determined with the real plane texture features of the real currency of the corresponding currency type and denomination to obtain a fused plane feature to be determined;

[0044] A real plane feature generation module is used to fuse the real plane texture features of real currency of corresponding currency type and denomination to obtain real plane fusion features;

[0045] The plane texture verification module is used to calculate the deviation between the plane fusion feature to be judged and the real plane fusion feature, and generate a plane texture verification result corresponding to the plane texture information.

[0046] Furthermore, the texture information includes three-dimensional texture information; and the recognition result generating unit includes:

[0047] a module for generating 3D features to be judged, configured to fuse the 3D texture features of the currency to be judged with the real 3D texture features of the real currency of the corresponding currency type and denomination to obtain a 3D fusion feature to be judged;

[0048] A real 3D feature generation module, used to fuse the real 3D texture features of real currency of corresponding currency type and denomination to obtain real 3D fusion features;

[0049] The stereo texture verification module is used to calculate the information entropy between the stereo fusion feature to be judged and the real stereo fusion feature, and generate a stereo texture verification result corresponding to the stereo texture information.

[0050] Furthermore, the currency authenticity identification device further includes:

[0051] A basic plane warning unit, configured to perform warning processing based on the basic information verification result and the plane texture verification result;

[0052] If the basic information verification result fails and / or the plane texture verification result fails, a high-risk warning process is performed.

[0053] Furthermore, the currency authenticity identification device further includes:

[0054] A basic plane and three-dimensional early warning unit, configured to perform early warning processing according to the basic information verification result, the plane texture verification result, and the three-dimensional texture verification result;

[0055] If the basic information verification result fails, the plane texture verification result and the three-dimensional texture verification result all fail, a high-risk warning process is performed;

[0056] If the basic information verification result passes, but the plane texture verification result and the three-dimensional texture verification result both fail, an intermediate warning process is performed;

[0057] If both the basic information verification result and the plane texture verification result pass, but the three-dimensional texture verification result fails, a low-risk warning process is performed.

[0058] In a third aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the currency authenticity identification method when executing the program.

[0059] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the currency authenticity identification method when executed by a processor.

[0060] In a fifth aspect, the present application provides a computer program product, comprising a computer program / instruction, which implements the steps of the currency authenticity identification method when executed by a processor.

[0061] In response to the problems in the existing technology, the currency authenticity identification method and device provided in this application can fully collect the image features of real currency, establish corresponding knowledge graphs and data sets, and use relational model search algorithms to find effective evidence to identify the authenticity of each currency to be tested. This can reduce the workload of relevant business personnel to a certain extent, speed up the efficiency of bank counter business processing, and improve the service experience of bank counter business. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0063] Figure 1This is a flow chart of the currency authenticity identification method in an embodiment of the present application;

[0064] Figure 2 A flowchart for constructing a currency information knowledge graph in an embodiment of the present application;

[0065] Figure 3 This is a flowchart of verifying the basic information of the currency to be verified in an embodiment of the present application;

[0066] Figure 4 This is one of the flow charts for verifying the texture information of the currency to be verified in an embodiment of the present application;

[0067] Figure 5 This is the second flow chart of verifying the texture information of the currency to be verified in the embodiment of the present application;

[0068] Figure 6 This is a structural diagram of a currency authenticity identification device in an embodiment of the present application;

[0069] Figure 7 This is a structural diagram of the knowledge graph construction unit in the embodiment of this application;

[0070] Figure 8 This is one of the structural diagrams of the recognition result generation unit in the embodiment of the present application;

[0071] Figure 9 This is the second structural diagram of the knowledge graph construction unit in the embodiment of this application;

[0072] Figure 10 This is the third structural diagram of the knowledge graph construction unit in the embodiment of this application;

[0073] Figure 11 This is an organizational chart of the currency authenticity identification system based on image recognition in an embodiment of the present application;

[0074] Figure 12 This is a schematic diagram of the basic knowledge system of a certain currency in an embodiment of this application;

[0075] Figure 13 This is a flowchart of information collection for different scenarios in the embodiments of this application;

[0076] Figure 14 This is one of the schematic diagrams showing the effect of processing currency image information in an embodiment of the present application;

[0077] Figure 15 This is a second schematic diagram of the effect of processing currency image information in an embodiment of the present application;

[0078] Figure 16 A schematic diagram of the key pattern and texture information of currency in an embodiment of the present application;

[0079] Figure 17 This is a schematic diagram of image fusion elimination in an embodiment of the present application;

[0080] Figure 18 This is a schematic diagram of the relationship between the overall comparison deviation and the quantity in the embodiment of this application;

[0081] Figure 19 This is a flowchart of currency warning in the embodiment of this application;

[0082] Figure 20 Schematic diagram of the structure of the electronic device in the embodiment of the present application. DETAILED DESCRIPTION

[0083] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0084] It should be noted that the currency authenticity identification method and device provided in this application can be used in the financial field, and can also be used in any field other than the financial field. The application field of the currency authenticity identification method and device provided in this application is not limited.

[0085] The acquisition, storage, use and processing of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0086] In one embodiment, see Figure 1 In order to construct a currency information knowledge graph and identify the authenticity of currency based on it, this application provides a currency authenticity identification method, including:

[0087] S101: Constructing a currency information knowledge graph based on the collected real currency images;

[0088] S102: Using the currency information knowledge graph, verify the basic information and texture information of the currency to be verified to obtain a currency authenticity identification result.

[0089] It is understood that this application is an efficient method for identifying the authenticity of currency using image recognition technology. This method can reduce the error rate of relevant business personnel when distinguishing the authenticity of currency and provide customers with an effective currency identification method. The corresponding system of this method can fully utilize the image characteristics of currency, construct an original data set (corresponding to a knowledge graph) based on the existing issued currency, and use a relational model search algorithm to find effective evidence to verify the authenticity of each currency. By deploying this system, the workload of business personnel can be reduced to a certain extent, the efficiency of bank counter business can be accelerated, and the service experience of bank counter business can be improved.

[0090] The main functions of the "currency authenticity identification system based on image recognition" corresponding to the currency authenticity identification method provided in this application are as follows:

[0091] First, we collect extensive image information of real currency in circulation (corresponding to images of real currency). Based on this information, we assign important parameters such as denomination, weight, and texture to each currency (corresponding to the attributes of real currency). This information is used to build a relational model of currency and store it in a database (corresponding to a knowledge graph). Because the currency information in the database is real currency information, it provides an effective and accurate basis for verifying the authenticity of the currency to be verified.

[0092] Secondly, using incremental storage technology, newly issued currency information is imported into the database, and old currency information recovered from various counters and ATM channels is archived and backed up to reduce the number of search engine visits and improve comparison efficiency.

[0093] Third, deploying this system provides counter services and convenient services. By providing counter staff with smart identification glasses and / or other hardware devices equipped with this system, they can easily identify the authenticity of currency, improving work efficiency. Of course, the method provided in this application can also be deployed on a server in the form of a software program and / or virtual device, and this application is not limited to this.

[0094] The "Currency Authenticity Identification System Based on Image Recognition" is mainly divided into the following software functional components: (The organizational structure diagram can be found in Figure 11 shown)

[0095] 1. Information Collection Component 101

[0096] 2. Data Management Component 102

[0097] 3. Smart glasses recognition component 103

[0098] 4. ATM intelligent identification component 104

[0099] 5. APP intelligent recognition component 105

[0100] 6. Intelligent early warning alarm component 106

[0101] To achieve the above objectives, the method provided by this application may be implemented by the following technical steps:

[0102] Step 1: Collect real currency image information;

[0103] Step 2: construct a currency information dataset;

[0104] Step 3: Establish a currency information relational model database;

[0105] Step 4: Collect image information of the currency to be verified;

[0106] Step 5: Extract key symbol information of the currency to be verified;

[0107] Step 6: extracting the pattern and texture information of the currency to be tested;

[0108] Step 7: Extract the depth information of the currency to be verified;

[0109] Step 8: Complete currency identification;

[0110] Step 9: Alarm warning.

[0111] From the above description, it can be seen that the currency authenticity identification method and device provided by this application can fully collect the image features of real currency, establish corresponding knowledge graphs and data sets, and use relational model search algorithms to find effective evidence to identify the authenticity of each currency to be tested, thereby reducing the workload of relevant business personnel to a certain extent, speeding up the efficiency of bank counter business, and improving the service experience of bank counter business.

[0112] In one embodiment, see Figure 2 , said constructing a currency information knowledge graph based on the collected real currency images includes:

[0113] S201: Extracting real currency attributes based on the real currency image; the real currency attributes include currency type, denomination, serial number, fineness grade, real plane texture features, and real three-dimensional texture features;

[0114] S202: generating a relationship between real currencies according to the real currency attributes;

[0115] S203: Generate the currency information knowledge graph according to the real currency, the real currency attributes and the relationship between the real currencies.

[0116] It is understood that in step 1 above, the specific method for collecting image information of real currency may include, but is not limited to: counting currency during the bank's currency issuance and / or cash withdrawal process; during the currency counting process, the information collector in the "currency authenticity identification system based on image recognition" collects images of the currency. The information collector includes, but is not limited to, an image collector, a weighing machine, and a multi-function camera.

[0117] An information collector is used to fully collect information about images of real currency, thereby collecting real currency image information (including real currency images and corresponding real currency attributes) and reducing human intervention. Because currency image information is of great importance and the consequences of leakage are disastrous, encryption algorithms such as RSA are used throughout the collection process to encrypt data transmission and storage, ensuring that encryption is performed at the source from the moment the image information is generated. By equipping the embodiment of the present application with an automated information collector, errors caused by manual operation are reduced, the bank's existing work rhythm is not affected, and the smooth progress of the collection work is ensured.

[0118] In step 2, the specific method for constructing a currency information dataset is to use the currency image information collected in step 1 and store the data in a database. Because step 1 employs encryption during information transmission, the currency image information stored in the database is encrypted data and cannot be directly used to construct the dataset. In a specific implementation, three key authentication USB flash drives belonging to relevant business personnel can be simultaneously inserted into three corresponding locations on the corresponding device, thereby activating the device to execute the dataset construction algorithm. The dataset can be constructed using various common dataset construction and statistical methods, such as classification, aggregation, permutation, and combination. This allows for orderly storage of currency of different types, denominations, and (manufactured) fineness grades. The fineness grade can be pre-set. It should be noted that the real currency attributes stored in the dataset include, but are not limited to, currency type, denomination, serial number, fineness grade, real planar texture features, and real three-dimensional texture features. Real planar texture features reflect the planar (surface) texture of real currency, while real three-dimensional texture features reflect the three-dimensional (convex and concave) texture of real currency.

[0119] Furthermore, in the aforementioned step 3, the specific method of establishing a currency information relationship model database (actually building a currency information knowledge graph) is:

[0120] Based on the currency image dataset obtained in step 2 above, we mapped the relationships between currencies using various common mapping methods, including mining, analysis, construction, and mapping, to construct a relationship model between the currencies. This model broadly categorizes the currencies by series, year of issue, denomination, and material.

[0121] like Figure 12 The following is the basic knowledge system of a certain currency. When each set of currency is printed and issued, the currencies of the same material and denomination have the same fineness grade. For banknotes, each currency of the same denomination has its own unique number, and this number is continuous and uninterrupted. In this way, the position of each banknote in the entire relationship map and its relationship with the banknotes with adjacent numbers can be known. For coins, the metal composition used in the casting of currencies of the same denomination is very similar. Although coins do not have unique properties such as numbers, coins of the same denomination in the same year have very similar properties. By constructing such a knowledge map, the embodiment of the present application can greatly speed up the search speed of information, especially when the currency image information presents a massive pattern, it can better play its advantages.

[0122] In summary, the three elements of knowledge graph construction, namely entities, relationships, and attributes, in this embodiment of the application are the real currency, the relationship between the real currencies, and the real currency attributes. These three elements can all be obtained through the above steps.

[0123] From the above description, it can be seen that the currency authenticity identification method provided by this application can construct a currency information knowledge graph based on the collected real currency images.

[0124] In one embodiment, see Figure 3 , the use of the currency information knowledge graph to verify the basic information of the currency to be verified includes:

[0125] S301: searching the currency information knowledge graph for real currencies of corresponding currency and denomination according to the currency and denomination of the currency to be verified, and obtaining the serial number and fineness grade of the real currency of corresponding currency and denomination;

[0126] S302: Compare the serial number and fineness of the real currency of the corresponding currency type and denomination with the serial number and fineness of the currency to be verified, and obtain a basic information verification result of the currency to be verified.

[0127] It is understandable that the process of collecting the image information of the currency to be verified (corresponding to the currency to be verified) in the aforementioned step 4 is as follows:

[0128] When a bank teller or user provides currency for verification, there are three possible approaches, each requiring different modes for capturing image information of the currency. First, when a bank teller inspects the currency, they simply place the currency directly in front of their line of sight. Using smart glasses equipped with an image-based currency authentication system, they can collect image information of the currency. Second, when a user inspects the currency at an ATM, they simply place the currency into the ATM. A laser scanner inside the ATM, equipped with an image-based currency authentication system, generates a flat color image and a 3D depth image of the currency, and also captures the currency's weight. Third, when a user uses a mobile phone app to authenticate the currency, the camera on the phone can capture the currency image. If the user's phone has a depth camera, a corresponding 3D depth image can also be captured. This embodiment of the present application utilizes existing on-site equipment to perform conditional currency image capture, tailored to different application scenarios. This reduces the additional investment required for system construction and minimizes disruption to existing currency counting and storage processes. The algorithm flow chart can be found in Figure 13 shown.

[0129] Furthermore, step 5, extracting key symbol information of the currency to be tested, involves performing symbol detection on the currency image information collected in step 4. First, using the character recognition algorithm within the image recognition algorithm, the collected currency image information is subjected to a series of operations, including grayscale conversion, binarization, and digit outline detection, to extract currency attributes such as the denomination, year of issue, serial number, and issuing bank.

[0130] Among them, grayscale refers to grayscale processing of the image. The original RGB image is as follows: Figure 14 As shown on the left, formula (1) can be used to calculate each pixel in the image to obtain the gray value of each pixel. Since the original RGB image is a three-dimensional information set, it is difficult to effectively compare the corresponding information during comparative analysis. Gray information is a dimensionality reduction process of three-dimensional information. The value range of Gray is usually [0, 255]. The processed image is as follows Figure 14 This formula is only an example, and the parameter settings are not limited in this application.

[0131] Gray=R×0.3+G×0.59+B×0.11 (1)

[0132] In formula (1), R is the red pixel value of the pixel; G is the green pixel value of the pixel; and B is the blue pixel value of the pixel.

[0133] Binarization is based on grayscale and uses 0 or 1 to describe the entire image. This processing method can better show the contour features of the image. Usually a threshold ξ can be set as the binarization coefficient to dynamically adjust the binarization image effect. For example, under the condition of ξ = 127.5, the binarized image can be obtained as follows Figure 14 (right).

[0134] After the above algorithm processing, the computer will output the result, which can be the information frame diagram of the front and back of a banknote. Figure 15 The embodiment of the present application utilizes this processing mode to conveniently obtain basic image information of the currency to be verified. By utilizing this basic image information, especially the serial number information, the corresponding currency information can be found in the currency information knowledge graph in the aforementioned step 3, thereby achieving a one-to-one comparison between the two sets of currency information and providing a precise comparison basis.

[0135] From the above description, it can be seen that the currency authenticity identification method provided by this application can use the currency information knowledge graph to verify the basic information of the currency to be verified.

[0136] In one embodiment, see Figure 4 , the texture information includes plane texture information; the verification of the basic information and texture information of the currency to be verified by using the currency information knowledge graph includes:

[0137] S401: Fusing the plane texture features of the currency to be verified with the real plane texture features of the real currency of the corresponding currency type and denomination to obtain the plane fusion features to be judged;

[0138] S402: Fusing the real plane texture features of the real currency of the corresponding currency type and denomination to obtain the real plane fusion features;

[0139] S403: Calculate the deviation between the to-be-judged plane fusion feature and the real plane fusion feature, and generate a plane texture verification result corresponding to the plane texture information.

[0140] It is understandable that the process of extracting the texture information of the currency to be verified in the aforementioned step 6 is as follows:

[0141] According to the currency attributes such as the issuing bank, issuing year and denomination obtained in the above step 5, searching in the currency information knowledge graph can determine the country to which the currency belongs and the key pattern information that the currency of that denomination should have when issued in that year. When the currency serial number information cannot be used to find the issuance and release information of the currency from the database in the above step 5, it is necessary to merge and compare the extracted key pattern information and texture information with the same key information of the adjacent currencies to find the similarities and differences. Among them, the currency pattern and texture information is as follows: Figure 16As shown in Figure 2. This grid-based fusion method significantly shortens fusion time. Once an anomaly is detected, it is identified as a suspect coin. A deviation analysis is then performed based on the overall binary histogram values ​​obtained after fusion and the overall binary histogram values ​​of the original images of neighboring currencies in the currency information knowledge graph. The deviation calculation formula is as follows.

[0142]

[0143] Where yj represents the metric value of a binary histogram calculated for a particular currency under test, C1 and C2 represent the metric values ​​of the binary histograms of adjacent currencies, and D1 and D2 represent the dimensions of the adjacent currencies being compared, which are generally considered to be one-dimensional. When the maximum deviation value is greater than a specified value, β, the coin is considered problematic. When the maximum deviation value is less than β, the coin is considered questionable and requires further evaluation. This approach can, to a certain extent, reduce the probability of misjudgment due to contamination and other factors, and can also accommodate subtle variations in the fineness of currency during production.

[0144] From the above description, it can be seen that the currency authenticity identification method provided in this application can use the currency information knowledge graph to verify the basic information and texture information of the currency to be verified.

[0145] In one embodiment, see Figure 5 , the texture information includes three-dimensional texture information; the verification of the basic information and texture information of the currency to be verified by using the currency information knowledge graph includes:

[0146] S501: Fusing the 3D texture features of the currency to be verified with the real 3D texture features of the real currency of the corresponding currency type and denomination to obtain a 3D fusion feature to be judged;

[0147] S502: Fusing the real 3D texture features of real currency of corresponding currency type and denomination to obtain real 3D fusion features;

[0148] S503: Calculate the information entropy between the to-be-determined stereo fusion feature and the real stereo fusion feature, and generate a stereo texture verification result corresponding to the stereo texture information.

[0149] It is understood that extracting depth information of the currency under inspection in step 7 means that if, after a series of steps, including steps 1 through 6, the authenticity of the suspect currency cannot be determined, further verification of the currency's authenticity is necessary based on the depth information (three-dimensional texture features) contained in the currency information. According to currency identification rules, the texture information on a genuine currency is not flat, but rather has a certain degree of unevenness, and this unevenness is clearly perceptible to the touch. The embodiments of the present application utilize a depth camera to capture the depth information (three-dimensional texture features) of the currency, specifically, to capture this unevenness. A depth image of a currency differs from an ordinary visible light image in that it is a grayscale image that cannot be interpreted by the naked eye. The grayscale values ​​within the image represent the depth information. Preliminary processing is performed using common mathematical methods, including but not limited to obtaining the average, maximum, minimum, deviation, and variance values, to characterize the overall digitized depth information of the image. By comparing the depth information with the depth information of neighboring currencies (nearby on the currency information knowledge graph), it is possible to determine whether the overall metric of the texture depth information is consistent or similar.

[0150] If the measurement deviation is large, it is necessary to further fuse the depth information of the image to be detected with one or more nearby images in the information library to find the local inconsistency location, so as to determine the authenticity of the currency.

[0151] like Figure 17 As mentioned above, the leftmost part of the figure is the depth image information of a real coin, the middle part is the depth image information of a problem coin, and the rightmost part is the fusion result information. Figure 17 It can be seen that the fused image still has obvious image features, so it can be concluded that the banknote to be detected is a problem banknote.

[0152] In theory, the judgment basis of this fusion comparison is mainly achieved by measuring the information entropy in the image. The calculation formula is as follows: the fused pixel information is summed to obtain the corresponding information entropy H(A).

[0153] H(A)=-∑ a P A (a)log2P A (a) (3)

[0154] A represents an event, H(A) represents the information entropy of event A, Alpha represents the error loss of the fused image, and P A (Alpha) indicates the probability that the error in the image fusion event is alpha. Since the image processing is binary, log2P is used. A(Alpha) is used as a dimensionality reduction coefficient to reduce fluctuations in calculated values. When the information entropy is high, meaning the error loss of the fused image is less than α, the fused image is highly similar to the original image and can be identified as a genuine coin. When the information entropy is low, meaning the error loss of the fused image is greater than or equal to α, the fused image is highly similar to the original image and can be identified as a questionable coin. Depth image comparison effectively leverages the texture information provided by currency printing, including its concave and convex textures. This is a key method for currency authentication, effectively preventing significant errors in identification results. Figure 18 The graph shows the relationship between the overall comparison deviation and the number of comparisons. Curve 1 is the lower limit of the deviation, curve 2 is the currency to be verified, and curve 3 is the upper limit of the deviation. The specific method is: Figure 18 The relationship between the overall comparison deviation and the comparison quantity is shown in Figure 1. Curve 1 is the lower limit of the deviation, curve 2 is the currency to be tested, and curve 3 is the upper limit of the deviation. When the number of inventory currencies that can be compared with the currency to be tested is determined, Figure 18 In the figure, draw a vertical straight line l. The intersection of this straight line with the upper limit can obtain the upper limit grayscale value G1, and the intersection with the lower limit can obtain the lower limit grayscale value G2. That is, when the calculated grayscale value of the coin to be tested is between [G1, G2], it cannot be determined that the coin to be tested does not meet the requirements.

[0155] From the above description, it can be seen that the currency authenticity identification method provided in this application can use the currency information knowledge graph to verify the basic information and texture information of the currency to be verified.

[0156] In one embodiment, the currency authenticity identification method further includes:

[0157] Performing early warning processing according to the basic information verification result and the plane texture verification result;

[0158] If the basic information verification result fails and / or the plane texture verification result fails, a high-risk warning process is performed.

[0159] In one embodiment, the currency authenticity identification method further includes:

[0160] Performing early warning processing according to the basic information verification result, the plane texture verification result, and the three-dimensional texture verification result;

[0161] If the basic information verification result fails, the plane texture verification result and the three-dimensional texture verification result all fail, a high-risk warning process is performed;

[0162] If the basic information verification result passes, but the plane texture verification result and the three-dimensional texture verification result both fail, an intermediate warning process is performed;

[0163] If both the basic information verification result and the plane texture verification result pass, but the three-dimensional texture verification result fails, a low-risk warning process is performed.

[0164] Furthermore, completing currency identification in step 8 means that after steps 1 through 7, the system ultimately generates an identification result. For the system, the identification result is expressed as a probability, P. Assuming the number of currency comparisons is N, and the sum of the overall comparison weights is 1, then P = a1 × P1 + a2 × P2 + a3 × P3…, a1 + a2 + a3 +… = 1, where a is the comparison weight of each image, decreasing by a multiple k based on proximity, i.e., a follows a geometric progression. P1, P2, etc. are the similarity probabilities obtained by comparing each image with the image to be tested.

[0165] When P is greater than or equal to the set value γ, the user is prompted with the message "The currency is real"; when P is less than the set value γ and greater than or equal to ξ, the user is prompted with the message "The currency may be a problem currency, please flatten the currency and retest it." When P is less than ξ, the user is directly prompted with the message "The currency is a problem currency, please hand it over to the bank for processing." The judgment flow chart is as follows: Figure 19 shown.

[0166] Furthermore, the alarm warning in the aforementioned step 9 refers to a comprehensive analysis of the authenticity and circulation status of currencies in different regions based on the historical data generated in the aforementioned steps 1 to 8 after the system has been running for a period of time.

[0167] In order to more clearly illustrate the method provided by this application, two examples are given below.

[0168] Example 1

[0169] An intelligent currency analysis system is provided to counter staff Xiao A, who is equipped with smart recognition glasses to determine the authenticity of the 100-yuan RMB in his hand.

[0170] This application is a currency authenticity identification system based on image recognition. It will use on-site equipment to provide users with currency identification services based on existing scene judgment. The specific steps are as follows:

[0171] Step 1: Collecting real currency image information: Banks need to count currency during currency issuance and disbursement. During this process, the currency can flow through an information collector equipped with this system. The collector utilizes an onboard weighing device, a flipping mechanism, and a multi-function camera to collect comprehensive information about the currency. This automatically collects real currency information, reducing manual intervention.

[0172] Part of the sources of the obtained currency information is shown in Table 1. To protect the confidentiality of the currency source, this is only an example and is not authentic. x and y in the table represent quantities.

[0173] Table 1 Currency source information

[0174]

[0175] Step 2', constructing a currency information dataset: Using the currency information collected in step 1', store the data in the database. Since step 1' uses encryption during information transmission, the information set in the database is encrypted data and cannot be directly constructed.

[0176] Step 3', establishing a currency information relationship model database: Based on the currency information dataset obtained in step 2', the relationship between currencies is displayed through mining, analysis, construction, and drawing to construct a relationship model between currencies.

[0177] Step 4', collecting image information of the currency to be verified: counter clerk Xiao A is wearing smart recognition glasses to identify the 100-yuan banknote in his hand.

[0178] Step 5', extract key mark information of the currency to be tested: perform mark detection on the data information collected in step 4. First, use the character recognition algorithm in the image recognition algorithm to perform a series of grayscale, binarization, digital contour search, cutting and other operations on the collected currency image information to extract the face value, issue year, number and issuing bank of the currency. The invention can easily obtain the basic information of the currency to be tested by using this processing mode. Using this basic information, especially the number information, the corresponding currency information can be found in the relational model database in step 3, so as to achieve a one-to-one comparison of the two sets of currency information and provide a basis for accurate comparison. According to the algorithm, it can be known that Xiao A's recognition mark of the currency is such as Figure 15 As shown, the currency with this number was not found in the database, which means that the currency was already circulating in the market before it was collected.

[0179] Step 6', extract the pattern and texture information of the currency to be tested: Based on the currency issuing bank, issuance year, and denomination obtained in step 5, the key pattern information of the country to which the currency belongs and the currency of the denomination issued in that year can be easily determined. If the currency serial number information cannot be used to find the issuance and delivery information of the currency from the database in step 5, it is necessary to fuse and compare the extracted key pattern information and texture information with the same key information of its neighboring currencies to find the similarities and differences. Using this grid-division pattern fusion method, the fusion time can be greatly shortened. Once an anomaly is found, it can be determined as a suspicious currency, and a deviation analysis is performed based on the overall binary histogram value obtained after fusion and the overall binary histogram value of the original multiple neighboring currency images. The deviation calculation formula is as follows.

[0180]

[0181] When the maximum deviation value is greater than the specified value β, it is considered a problem coin. When the maximum deviation value is less than β, it is considered a questionable coin and needs further judgment. This method can reduce the probability of misjudgment of currency due to contamination and other reasons to a certain extent, and can also adapt to the extremely subtle color differences in the currency during the production process. Here we can see that Xiao A extracts the key pattern and texture information of the currency. Figure 6 As shown, the calculated deviation value is 2.91, which meets the set requirements and can be determined to be a real currency.

[0182] Step 7': Extracting Depth Information of the Currency to be Verified: After completing steps 1 through 6 above, assuming that Xiao A still doesn't believe the suspect currency is genuine, further verification of the currency's authenticity will be based on the depth information contained in the currency. According to currency authentication rules, the texture of a genuine currency is not flat, but rather has a certain degree of unevenness, which is noticeable when touched. This application uses a depth camera to capture depth information of the currency, specifically this unevenness. Unlike ordinary visible light images, depth images are grayscale images that cannot be interpreted by the naked eye. The grayscale values ​​within the image represent depth information. By calculating the average, maximum, minimum, deviation, and variance values, the overall depth information of the image can be characterized. By comparing the depth information with the depth information of neighboring currencies, the overall consistency or similarity of the texture depth information can be determined. If the metric deviation is large, the two compared depth images must be further fused to identify local inconsistencies and determine the currency's authenticity. This fusion comparison is primarily based on measuring the information entropy within the image, calculated as follows.

[0183] H(A)=-∑ a P A (a)log2P A (a) (2)

[0184] When the information entropy is high, that is, when the error loss of the fused image is less than α, the fused image is highly similar to the original image and can be identified as a genuine currency. When the information entropy is low, that is, when the error loss of the fused image is greater than or equal to α, the fused image is very similar to the original image and can be identified as a counterfeit currency. Depth image comparison can effectively leverage the texture information of the concave and convex properties of currency printing, which is a key method for currency identification and can effectively avoid significant errors in the identification results. Calculation shows that H(A) = 99.9999%, so the currency in Xiao A's hand can be determined to be genuine.

[0185] Step 8: Complete currency authentication: After steps 1 through 7, the system uses a series of authentication methods to ultimately produce an authentication result. For the system, this authentication result is expressed as a probability P. When P is greater than or equal to the set value γ, the user is prompted with the message "The currency is genuine." When P is less than the set value γ and greater than or equal to ξ, the user is prompted with the message "This currency may be a problem coin. Please flatten it and retest it." When P is less than ξ, the user is directly prompted with the message "This currency is a problem coin. Please hand it over to the bank for processing." This approach reduces psychological pressure on users, preventing them from being startled by sudden alarms. After a series of actions, counter clerk Xiao A confirms the currency is genuine and provides feedback to the user, who expresses his appreciation for Xiao A's expertise.

[0186] Step 9', alarm and warning: After the system has been running for a period of time, based on the historical data generated from steps 1 to 8 above, the authenticity and circulation status of currencies in different regions can be comprehensively analyzed.

[0187] Example 2

[0188] An intelligent currency analysis system is provided to user Xiao B, who actively uses the intelligent recognition ATM to perform currency identification operations. He holds a certain country's currency with a face value of 100,000.

[0189] This application is a currency authenticity identification system based on image recognition. It will use on-site equipment to provide users with currency identification services based on existing scene judgment. The specific steps are as follows:

[0190] Step 1: Collect real currency image information: Part of the currency information obtained is from the source shown in Table 2. To protect the confidentiality of the currency source, this is only an example and is not authentic. In the table, x and y represent the quantity.

[0191] Table 2 Currency source information

[0192]

[0193] Step 2”, construct the currency information dataset: use the currency information collected in step 1” to store the data in the database.

[0194] Step 3: Establish a currency information relationship model database: Based on the currency information dataset obtained in Step 2, the relationship between currencies is displayed through mining, analysis, construction, and drawing to construct a relationship model between currencies.

[0195] Step 4: Collect the image information of the currency to be verified: User Xiao B uses the intelligent recognition ATM to identify the currency.

[0196] Step 5: Extract key identifiers of the currency to be verified: Based on the data collected in Step 4, perform identifier verification. The algorithm determines that the currency provided by B is a foreign currency, originating from the region of ××, with a unit of ×× and a denomination of ××. A database search reveals that no circulation information for this currency is found.

[0197] Step 6” is to extract the pattern and texture information of the currency to be tested: Based on the issuing bank, year of issuance, and denomination of the currency obtained in Step 5”, the key pattern information of the country to which the currency belongs and the currency of the denomination issued in that year can be easily determined. When the currency serial number information cannot be used to find the issuance and delivery information of the currency from the database in Step 5”, it is necessary to fuse and compare the extracted key pattern information and texture information with the same key information of its neighboring currencies to find the similarities and differences. Using this grid division pattern fusion method, the fusion time can be greatly shortened. Once an anomaly is found, it can be determined as a suspicious currency, and a deviation analysis is performed based on the overall binary histogram value obtained after fusion and the overall binary histogram value of the original multiple neighboring currency images. The deviation calculation formula is as follows.

[0198]

[0199] When the maximum deviation value is greater than the specified value β, the coin is considered a problem coin. When the maximum deviation value is less than β, the coin is considered questionable and requires further evaluation. This approach can reduce the probability of misjudgment due to contamination and other factors, and can also accommodate subtle variations in the fineness of the currency during the production process. Here, the system calculates a deviation value of 10.91 for the currency held by Xiao B, which does not meet the specified requirements and is therefore considered a problem coin.

[0200] Step 7: Extracting Depth Information of the Currency to be Verified: After completing steps 1 through 6 above, assuming that Xiao B still does not believe the suspected currency is fake, further verification of the currency's authenticity is based on the depth information contained within the currency. According to currency authentication rules, the texture of a genuine currency is not flat, but rather has a certain degree of unevenness, which is noticeable to the touch. This application utilizes a depth camera to capture depth information of the currency, specifically this unevenness. Depth images differ from ordinary visible light images in that they are grayscale images that cannot be interpreted by the naked eye. The grayscale values ​​within the image represent depth information. The average, maximum, minimum, deviation, and variance values ​​are calculated to characterize the overall depth information of the image. By comparing the depth information with the depth information of neighboring currencies, the overall consistency or similarity of the texture depth information can be determined. If the metric deviation is significant, the two compared depth images must be fused to identify local inconsistencies and determine the currency's authenticity. This fusion comparison is primarily based on measuring the information entropy within the image, calculated as follows.

[0201] H(A)=-∑ a P A (a)log2P A (a) (2)

[0202] When the information entropy is high, that is, when the error loss of the fused image is less than α, the fused image is highly similar to the original image and can be identified as a genuine currency. When the information entropy is low, that is, when the error loss of the fused image is greater than or equal to α, the fused image is very similar to the original image and can be identified as a counterfeit currency. Depth image comparison can effectively leverage the texture information of the concave and convex properties of currency printing, which is a key method for currency identification and can effectively avoid significant errors in the identification results. Calculation shows that H(A) = 39.8793%, and the currency in Xiao B's hand can be determined to be a counterfeit currency.

[0203] Step 8: Complete currency identification: After steps 1 through 7, the system uses a series of identification methods to ultimately produce an identification result. For the system, this identification result is expressed as a probability P. When P is greater than or equal to the set value γ, the user is prompted with the message "The currency is genuine." When P is less than the set value γ and greater than or equal to ξ, the user is prompted with the message "This currency may be a problem coin. Please flatten the coin and retest." When P is less than ξ, the user is directly prompted with the message "Problem coin, please hand it over to the bank for processing." This approach reduces psychological pressure on users, preventing them from being startled by sudden alarms. The intelligent analysis system determines that the currency held by Xiao B is a problem coin and provides this information to the user, prompting them to proactively hand it over to the bank for processing.

[0204] Step 9”, Alarm and warning: After the system has been running for a period of time, based on the historical data generated from steps 1 to 8 above, the authenticity and circulation status of currencies in different regions can be comprehensively analyzed.

[0205] Based on the same inventive concept, the embodiments of the present application also provide a currency authenticity identification device, which can be used to implement the method described in the above embodiments, as described in the following embodiments. Since the principle of solving the problem by the currency authenticity identification device is similar to that of the currency authenticity identification method, the implementation of the currency authenticity identification device can refer to the implementation of the method based on software performance benchmark determination, and the repetitions will not be repeated. As used below, the term "unit" or "module" can implement a combination of software and / or hardware for a predetermined function. Although the system described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0206] In one embodiment, see Figure 6 In order to construct a currency information knowledge graph and identify the authenticity of the currency based on it, the present application provides a currency authenticity identification device, including: a knowledge graph construction unit 601 and an identification result generation unit 602.

[0207] A knowledge graph construction unit 601 is configured to construct a currency information knowledge graph based on the collected real currency images;

[0208] The recognition result generating unit 602 is used to verify the basic information and texture information of the currency to be verified by using the currency information knowledge graph to obtain a currency authenticity recognition result.

[0209] In one embodiment, see Figure 7 The knowledge graph construction unit 601 includes: a real attribute extraction module 701, a real relationship extraction module 702 and a knowledge graph construction module 703.

[0210] A real attribute extraction module 701 is configured to extract real currency attributes based on the real currency image; the real currency attributes include currency type, denomination, serial number, fineness grade, real plane texture features, and real three-dimensional texture features;

[0211] A real relationship extraction module 702, configured to generate a relationship between real currencies based on the real currency attributes;

[0212] The knowledge graph construction module 703 is used to generate the currency information knowledge graph according to the real currency, the real currency attributes and the relationship between the real currencies.

[0213] In one embodiment, see Figure 8The recognition result generating unit 602 includes: a real currency selection module 801 and a basic information verification module 802.

[0214] The real currency selection module 801 is configured to retrieve the real currency of the corresponding currency and denomination in the currency information knowledge graph according to the currency and denomination of the currency to be verified, and obtain the serial number and fineness grade of the real currency of the corresponding currency and denomination;

[0215] The basic information verification module 802 is used to compare the serial number and fineness of the real currency of the corresponding currency type and denomination with the serial number and fineness of the currency to be verified, and obtain the basic information verification result of the currency to be verified.

[0216] In one embodiment, see Figure 9 , the texture information includes plane texture information; the recognition result generating unit 602 includes: a plane feature generating module 901 to be judged, a real plane feature generating module 902 and a plane texture verification module 903.

[0217] The plane feature generation module 901 is configured to fuse the plane texture features of the currency to be verified with the real plane texture features of the real currency of the corresponding currency type and denomination to obtain the plane fusion features to be determined;

[0218] A real plane feature generation module 902 is used to fuse the real plane texture features of real currency of corresponding currency type and denomination to obtain a real plane fusion feature;

[0219] The plane texture verification module 903 is used to calculate the deviation between the plane fusion feature to be determined and the real plane fusion feature, and generate a plane texture verification result corresponding to the plane texture information.

[0220] In one embodiment, see Figure 10 , the texture information includes three-dimensional texture information; the recognition result generating unit includes: a three-dimensional feature generating module 1001 to be judged, a real three-dimensional feature generating module 1002 and a three-dimensional texture verification module 1003.

[0221] The module 1001 for generating 3D features to be judged is used to fuse the 3D texture features of the currency to be judged with the real 3D texture features of the real currency of the corresponding currency type and denomination to obtain the 3D fusion features to be judged;

[0222] A real 3D feature generation module 1002 is used to fuse the real 3D texture features of real currency of corresponding currency type and denomination to obtain a real 3D fusion feature;

[0223] The stereo texture verification module 1003 is used to calculate the information entropy between the to-be-determined stereo fusion feature and the real stereo fusion feature, and generate a stereo texture verification result corresponding to the stereo texture information.

[0224] In one embodiment, the currency authenticity identification device further includes:

[0225] A basic plane warning unit, configured to perform warning processing based on the basic information verification result and the plane texture verification result;

[0226] If the basic information verification result fails and / or the plane texture verification result fails, a high-risk warning process is performed.

[0227] In one embodiment, the currency authenticity identification device further includes:

[0228] A basic plane and three-dimensional early warning unit, configured to perform early warning processing according to the basic information verification result, the plane texture verification result, and the three-dimensional texture verification result;

[0229] If the basic information verification result fails, the plane texture verification result and the three-dimensional texture verification result all fail, a high-risk warning process is performed;

[0230] If the basic information verification result passes, but the plane texture verification result and the three-dimensional texture verification result both fail, an intermediate warning process is performed;

[0231] If both the basic information verification result and the plane texture verification result pass, but the three-dimensional texture verification result fails, a low-risk warning process is performed.

[0232] From a hardware perspective, in order to construct a currency information knowledge graph and identify currency authenticity based on it, this application provides an embodiment of an electronic device for implementing all or part of the currency authenticity identification method. The electronic device specifically includes the following:

[0233] A processor, a memory, a communications interface, and a bus; wherein the processor, the memory, and the communications interface communicate with each other via the bus; the communications interface is used to transmit information between the currency authenticity identification device and related devices such as a core business system, a client terminal, and a related database; the logic controller can be a desktop computer, a tablet computer, a mobile terminal, etc., but the present embodiment is not limited thereto. In this embodiment, the logic controller can be implemented with reference to the embodiments of the currency authenticity identification method and the embodiments of the currency authenticity identification device in the embodiments, the contents of which are incorporated herein, and repeated parts are not repeated.

[0234] It is understandable that the client terminal may include a smart phone, a tablet electronic device, a network set-top box, a portable computer, a desktop computer, a personal digital assistant (PDA), a vehicle-mounted device, a smart wearable device, etc. Among them, the smart wearable device may include smart glasses, a smart watch, a smart bracelet, etc.

[0235] In practical applications, portions of the currency authenticity identification method may be performed on the electronic device as described above, or all operations may be performed on the client device. The specific method may be selected based on the processing capabilities of the client device and the limitations of the client's usage scenario. This application does not impose any restrictions on this. If all operations are performed on the client device, the client device may also include a processor.

[0236] The client device may include a communication module (i.e., a communication unit) that can establish a communication connection with a remote server to implement data transmission with the server. The server may include a server on the task scheduling center side, and in other implementation scenarios, may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, a server cluster consisting of multiple servers, or a server structure of a distributed device.

[0237] Figure 20 Schematic block diagram of the system structure of the electronic device 9600 according to an embodiment of the present application. Figure 20 As shown, the electronic device 9600 may include a central processing unit 9100 and a memory 9140; the memory 9140 is coupled to the central processing unit 9100. It is worth noting that the Figure 20 is exemplary; other types of structures may also be used to supplement or replace this structure to implement telecommunication functions or other functions.

[0238] In one embodiment, the currency authenticity identification method function may be integrated into the central processing unit 9100. The central processing unit 9100 may be configured to perform the following control:

[0239] S101: Constructing a currency information knowledge graph based on the collected real currency images;

[0240] S102: Using the currency information knowledge graph, verify the basic information and texture information of the currency to be verified to obtain a currency authenticity identification result.

[0241] From the above description, it can be seen that the currency authenticity identification method and device provided by this application can fully collect the image features of real currency, establish corresponding knowledge graphs and data sets, and use relational model search algorithms to find effective evidence to identify the authenticity of each currency to be tested, thereby reducing the workload of relevant business personnel to a certain extent, speeding up the efficiency of bank counter business, and improving the service experience of bank counter business.

[0242] In another embodiment, the currency authenticity identification device can be configured separately from the central processor 9100. For example, the data composite transmission device currency authenticity identification device can be configured as a chip connected to the central processor 9100, and the function of the currency authenticity identification method can be realized through the control of the central processor.

[0243] like Figure 20 As shown, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is worth noting that the electronic device 9600 does not necessarily have to include Figure 20 In addition, the electronic device 9600 may also include all components shown in Figure 20 For components not shown, reference may be made to the prior art.

[0244] like Figure 20 As shown, the central processing unit 9100 is sometimes also referred to as a controller or operation control, and may include a microprocessor or other processor device and / or logic device. The central processing unit 9100 receives input and controls the operation of various components of the electronic device 9600.

[0245] Memory 9140 can be, for example, one or more of a cache, flash memory, hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It can store the aforementioned failure-related information and also store programs that execute the relevant information. The CPU 9100 can execute the programs stored in memory 9140 to implement information storage or processing.

[0246] The input unit 9120 provides input to the central processing unit 9100. The input unit 9120 may be, for example, a keypad or touch input device. The power supply 9170 is used to provide power to the electronic device 9600. The display 9160 is used to display objects such as images and text. The display may be, for example, an LCD display, but is not limited thereto.

[0247] The memory 9140 may be a solid-state memory, such as a read-only memory (ROM), a random access memory (RAM), or a SIM card. Alternatively, it may be a memory that retains information even when power is off, can be selectively erased, and is provided with more data. Examples of such memory are sometimes referred to as EPROMs. The memory 9140 may also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application / function storage unit 9142 for storing application programs and function programs or processes for executing the operation of the electronic device 9600 by the central processing unit 9100.

[0248] The memory 9140 may also include a data storage unit 9143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 may include various driver programs for the electronic device's communication functions and / or for executing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0249] The communication module 9110 is a transmitter / receiver 9110 that sends and receives signals via an antenna 9111. The communication module (transmitter / receiver) 9110 is coupled to the central processor 9100 to provide input signals and receive output signals, which may be the same as in a conventional mobile communication terminal.

[0250] Based on different communication technologies, multiple communication modules 9110 can be provided in the same electronic device, such as a cellular network module, a Bluetooth module, and / or a wireless local area network module. The communication module (transmitter / receiver) 9110 is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide audio output via the speaker 9131 and receive audio input from the microphone 9132, thereby implementing common telecommunication functions. The audio processor 9130 may include any suitable buffer, decoder, amplifier, etc. In addition, the audio processor 9130 is also coupled to the central processing unit 9100, enabling local recording via the microphone 9132 and playback of stored audio via the speaker 9131.

[0251] The embodiments of the present application also provide a computer-readable storage medium capable of implementing all steps of the currency authenticity identification method in the above-mentioned embodiment, where the execution subject is a server or a client. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the computer program implements all steps of the currency authenticity identification method in the above-mentioned embodiment, where the execution subject is a server or a client. For example, when the processor executes the computer program, the following steps are implemented:

[0252] S101: Constructing a currency information knowledge graph based on the collected real currency images;

[0253] S102: Using the currency information knowledge graph, verify the basic information and texture information of the currency to be verified to obtain a currency authenticity identification result.

[0254] From the above description, it can be seen that the currency authenticity identification method and device provided by this application can fully collect the image features of real currency, establish corresponding knowledge graphs and data sets, and use relational model search algorithms to find effective evidence to identify the authenticity of each currency to be tested, thereby reducing the workload of relevant business personnel to a certain extent, speeding up the efficiency of bank counter business, and improving the service experience of bank counter business.

[0255] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, devices, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0256] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (apparatus), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0257] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0258] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0259] Specific embodiments are used in this application to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for identifying the authenticity of currency, characterized in that: include: Build a currency information knowledge graph based on the collected real currency images; The currency information knowledge graph is used to verify the basic information and texture information of the currency to be tested, and obtain a currency authenticity identification result, wherein the texture information includes: plane texture information; The method of using the currency information knowledge graph to verify the basic information and texture information of the currency to be verified includes: If no information on the issuance and release of the currency to be verified is detected in the currency information knowledge graph, the texture features of the plane to be verified of the currency to be verified are fused with the real plane texture features of multiple adjacent real currencies in the currency information knowledge graph to obtain the fused features of the plane to be judged; Fusing the real plane texture features of multiple real currencies adjacent to the currency information knowledge graph to obtain a real plane fusion feature; wherein the to-be-determined plane fusion feature and the real plane fusion feature are both metric values ​​of a two-dimensional histogram; The deviation between the plane fusion feature to be judged and the real plane fusion feature is calculated to generate a plane texture verification result corresponding to the plane texture information.

2. The method for identifying currency authenticity according to claim 1, wherein: The method of constructing a currency information knowledge graph based on the collected real currency images includes: Extracting real currency attributes based on the real currency image; the real currency attributes include currency type, denomination, serial number, fineness grade, real plane texture features, and real three-dimensional texture features; generating a relationship between real currencies according to the real currency attributes; The currency information knowledge graph is generated according to the real currency, the real currency attributes and the relationship between the real currencies.

3. The method for identifying currency authenticity according to claim 2, wherein: The method of verifying the basic information of the currency to be verified by using the currency information knowledge graph includes: According to the currency type and denomination of the currency to be verified, the real currency of the corresponding currency type and denomination in the currency information knowledge graph is retrieved to obtain the serial number and fineness grade of the real currency of the corresponding currency type and denomination; The serial number and fineness of the real currency of the corresponding currency type and denomination are compared with the serial number and fineness of the currency to be verified to obtain the basic information verification result of the currency to be verified.

4. The method for identifying currency authenticity according to claim 1, wherein: The texture information includes three-dimensional texture information; and the verification of the basic information and texture information of the currency to be verified by using the currency information knowledge graph includes: Fusing the 3D texture features of the currency to be verified with the real 3D texture features of the real currency of the corresponding currency type and denomination to obtain a 3D fusion feature to be judged; Fusing the real 3D texture features of real currency of corresponding currency type and denomination to obtain real 3D fusion features; The information entropy between the to-be-determined stereo fusion feature and the real stereo fusion feature is calculated to generate a stereo texture verification result corresponding to the stereo texture information.

5. The method for identifying currency authenticity according to claim 1, wherein: Also includes: Performing early warning processing according to the basic information verification result and the plane texture verification result; If the basic information verification result fails and / or the plane texture verification result fails, a high-risk warning process is performed.

6. The method for identifying currency authenticity according to claim 4, wherein: Also includes: Performing early warning processing according to the basic information verification result, the plane texture verification result, and the three-dimensional texture verification result; If the basic information verification result fails, the plane texture verification result and the three-dimensional texture verification result all fail, a high-risk warning process is performed; If the basic information verification result passes, but the plane texture verification result and the three-dimensional texture verification result both fail, an intermediate warning process is performed; If both the basic information verification result and the plane texture verification result pass, but the three-dimensional texture verification result fails, a low-risk warning process is performed.

7. A currency authenticity identification device, characterized in that: include: A knowledge graph construction unit, used to construct a currency information knowledge graph based on the collected real currency images; An identification result generating unit is used to verify the basic information and texture information of the currency to be verified by using the currency information knowledge graph to obtain a currency authenticity identification result, wherein the texture information includes: plane texture information; The recognition result generating unit is specifically configured to: If no information on the issuance and release of the currency to be verified is detected in the currency information knowledge, the texture features of the plane to be verified of the currency to be verified are fused with the real plane texture features of multiple adjacent real currencies in the currency information knowledge graph to obtain the fused features of the plane to be judged; Fusing the real plane texture features of multiple real currencies adjacent to the currency information knowledge graph to obtain a real plane fusion feature; wherein the to-be-determined plane fusion feature and the real plane fusion feature are both metric values ​​of a two-dimensional histogram; The deviation between the plane fusion feature to be judged and the real plane fusion feature is calculated to generate a plane texture verification result corresponding to the plane texture information.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the currency authenticity identification method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the currency authenticity identification method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps of the currency authenticity identification method according to any one of claims 1 to 6 are implemented.

Images