Banknote image recording and anomaly detection system

Abstract

This invention provides a banknote image recording and anomaly detection system, in which a camera captures images of banknotes. A banknote detector verifies the banknotes to generate a verification result, analyzes the image to identify the corresponding banknote serial number, and records the banknote status. A server processor checks if the stored serial number in the database matches the banknote serial number to generate a first confirmation result, and decides whether to perform a banknote status verification operation based on the first confirmation result. When the first confirmation result is yes, the server processor performs a banknote status verification operation based on the stored banknote status to generate a second confirmation result, and decides whether to generate an anomaly notification based on the second confirmation result and the verification result. This effectively records banknote serial numbers and detects counterfeit banknotes, avoiding disputes arising from customer reports of counterfeit banknotes.

Description

Technical field

[0001] The invention relates to a paper image detection system and, in particular, to a banknote image recording and anomaly detection system. Prior technology

[0002] In a general financial institution, if the public reports the withdrawal of money as counterfeit banknotes, they can only emphasize to the customer several times that the banknotes withdrawn have passed through the banknote checking machine and that the banknotes leaving the financial institution will not be counterfeit banknotes. However, customers still insist that there are counterfeit banknotes inside, leaving financial institutions and operating agents undecided. Such disputes often occur repeatedly, causing distress and numerous inconveniences to financial institutions. It follows that there is currently a lack of a banknote image recording and anomaly detection system on the market that can effectively record the banknote serial number (banknote serial number) or pattern and detect its authenticity. Contents of the invention

[0003] The purpose of the present invention is to provide a banknote image recording and anomaly detection system that can record banknote serial numbers and has a network module to connect the database to the server to compare the banknote serial number transaction records of the blockchain and in series with the counterfeit banknote serial number published by the central bank. Furthermore, if the customer reports the withdrawal of money as counterfeit, the serial number of the banknote where the customer received cash can be retrieved from the system to prove that it is different from the so-called counterfeit banknote (such as toy banknote) in the customer's hand.

[0004] According to one embodiment of the present invention, a banknote image recording and anomaly detection system is provided, comprising a camera, a banknote detector, and a server. The camera captures an image of a banknote. The banknote detector is connected to the camera and receives the image. The banknote detector verifies the banknote to generate a verification result, analyzes the image to identify the banknote serial number, and records the banknote status. The server is connected to the banknote detector and includes a database and a server processor. The database stores a plurality of retained serial numbers and a plurality of retained banknote statuses corresponding to these retained serial numbers. The server processor is connected to the database and receives these retained serial numbers, these retained banknote statuses, and the verification result, banknote serial number, and banknote status from the database and the banknote detector. The server processor confirms whether one of these retained serial numbers matches a banknote serial number, generates a first confirmation result, and decides whether to perform a banknote status confirmation operation based on the first confirmation result. When the first confirmation result is yes, the banknote serial number is the same as one of these retained serial numbers. The servo processor performs a banknote status confirmation operation based on the status of these retained banknotes to generate a second confirmation result, and decides whether to generate an exception notification based on the second confirmation result and the banknote verification result.

[0005] Other embodiments of the aforementioned implementation are as follows: The aforementioned banknote detector includes a memory, a processor, and a network module. The memory is used to store images, banknote verification results, banknote serial numbers, and banknote status. The processor is connected to the memory and is used to verify banknotes to generate verification results, and analyze images to identify the corresponding banknote serial number. The network module is connected to the processor and is used to transmit banknote verification results, banknote serial numbers, and banknote status through a network.

[0006] Other embodiments of the aforementioned implementation are as follows: The aforementioned banknote image recording and anomaly detection system further includes a blockchain, which is connected to a server and used to receive banknote serial numbers and banknote status in order to generate a transaction result.

[0007] Other embodiments of the aforementioned implementation are as follows: When the aforementioned first confirmation result is negative, the server processor transmits the banknote serial number and banknote status to the blockchain. The blockchain generates and sends back a transaction serial number and a transaction hash value to the server's database based on the banknote serial number and banknote status. The transaction result includes the transaction serial number and the transaction hash value.

[0008] Other embodiments of the aforementioned implementation are as follows: Each of the aforementioned retained serial numbers corresponds to one of the retained banknote states. The banknote state confirmation operation includes the server processor confirming whether the retained banknote state is the same as the banknote state to generate a second confirmation result. When the second confirmation result is yes, the server processor generates an exception notification; otherwise, when the second confirmation result is no, the server processor performs a state change operation on the banknote state, which includes changing the banknote state to either entry information or exit information.

[0009] Other embodiments of the aforementioned implementation are as follows: The aforementioned banknote detector is configured to perform the following operations: using a standard banknote issued by a central bank as a comparison template; performing a grayscale processing on the comparison template to generate a grayscale comparison template; marking a serial number area range in the grayscale comparison template; and analyzing the comparison template based on the serial number area range and a text recognition model to identify the corresponding retained serial numbers of the standard banknote.

[0010] Other embodiments of the aforementioned implementation are as follows: The aforementioned banknote detector is configured to perform the following operations: using a standard banknote issued by a central bank as a comparison template; performing grayscale processing on the comparison template to generate a grayscale comparison template; performing an image subtraction operation on the image and the grayscale comparison template to generate a difference image feature; calculating the area ratio of the difference image feature relative to the area of ​​a banknote in the image, and confirming whether the area ratio exceeds a preset ratio threshold to generate a soiling confirmation result; and determining whether the banknote in the corresponding image needs to be exchanged with the central bank based on the soiling confirmation result.

[0011] Other embodiments of the aforementioned implementation are as follows: The aforementioned banknote detector is configured to perform operations including: marking the key icon position of one of the grayscale comparison templates, wherein the key icon position includes an amount position or a pattern position; and determining whether there is icon displacement at the key icon position based on the difference image features and a preset offset threshold value, thereby generating a displacement judgment result. When the offset of one of the difference image features is greater than or equal to the preset offset threshold value, the banknote detector determines that the banknote has icon displacement; conversely, when the offset of the difference image features is less than the preset offset threshold value, the banknote detector determines that the banknote has no icon displacement.

[0012] Other embodiments of the aforementioned implementation are as follows: The aforementioned server processor updates a counterfeit banknote serial number group published by the central bank using a web crawler, compares whether the banknote serial number matches the counterfeit banknote serial number group to generate a counterfeit banknote comparison result, and decides whether to generate an anomaly notification based on the counterfeit banknote comparison result. The counterfeit banknote serial number group contains at least one counterfeit banknote serial number. When the counterfeit banknote comparison result is yes, the server processor generates an anomaly notification; otherwise, when the counterfeit banknote comparison result is no, the server processor does not generate an anomaly notification.

[0013] Other embodiments of the aforementioned implementation are as follows: The aforementioned servo processor is connected to a police unit. When an abnormal notification is generated, the servo processor retrieves the banker's name and a customer's name corresponding to the banknote serial number from the database and sends an alert email to either a department head's email address or a currently processing banker's email address. The servo processor also generates a notification document based on the banker's name and the customer's name and transmits the notification document to the police unit.

[0014] Therefore, the banknote image recording and anomaly detection system of this invention can establish a history of valuable paper by collecting banknote serial numbers and their condition, and upload the records of their entry and exit from circulation to the blockchain. The immutability and transparency of blockchain technology ensure the integrity of the data and prevent unauthorized modification. Since one of the core characteristics of blockchain technology is that the data uploaded to the chain cannot be tampered with, this provides a powerful protective mechanism for banknote management within financial institutions, effectively preventing malicious operations by internal personnel, enabling the tracing and identification of abnormal situations, and avoiding disputes arising from customer reports of counterfeit banknotes. Simple Explanation of the Diagram

[0015] Figure 1 is a schematic diagram illustrating a banknote image recording and anomaly detection system according to a first embodiment of the present invention; Figure 2 is a flowchart illustrating the banknote image recording and anomaly detection method according to a second embodiment of the present invention; Figure 3A illustrates a flowchart of the banknote image recording and anomaly detection method according to a third embodiment of the present invention; Figure 3B illustrates the flowchart of the banknote image recording and anomaly detection method shown in Figure 3A after an anomaly notification is issued; Figure 4 illustrates a flowchart of the banknote image recording and anomaly detection method according to the fourth embodiment of the present invention; and Figure 5 is a flowchart illustrating the banknote image recording and anomaly detection method according to the fifth embodiment of the present invention. Implementation

[0016] Several embodiments of the present invention will now be described with reference to the drawings. For clarity, many practical details will be set forth in the following description. However, it should be understood that these practical details are not intended to limit the invention. That is, in some embodiments of the invention, these practical details are not essential. Furthermore, for the sake of simplicity, some conventional structures and elements will be depicted in a simple schematic manner in the drawings; and repeated elements may be denoted by the same number.

[0017] Furthermore, in this document, when a component (or unit or module, etc.) is "connected" to another component, it can mean that the component is directly connected to the other component, or that the component is indirectly connected to the other component, meaning that there is another component between the component and the other component. Only when it is explicitly stated that a component is "directly connected" to another component does it indicate that there is no other component between the component and the other component. The terms "first," "second," and "third" are only used to describe different components and do not limit the components themselves; therefore, "first component" can also be referred to as "second component." Moreover, the combinations of components / units / circuits in this document are not combinations generally known, conventional, or customary in this field. Whether the components / units / circuits themselves are customary cannot be used to determine whether their combination relationships are easily accomplished by someone with ordinary knowledge in the art.

[0018] Please refer to Figure 1, which is a schematic diagram illustrating a banknote image recording and anomaly detection system 100 according to a first embodiment of the present invention. The banknote image recording and anomaly detection system 100 includes a camera 200, a banknote detector 300, and a server 400. The camera 200 is used to capture an image of a banknote. The banknote detector 300 is connected to the camera 200 and receives the image. The banknote detector 300 is used to inspect the banknote to generate a banknote verification result, analyze the image to identify the banknote serial number corresponding to the banknote, and record the banknote status of the corresponding banknote. The server 400 is connected to the banknote detector 300 and includes a database 410 and a server processor 420. The database 410 stores a plurality of retained serial numbers and a plurality of retained banknote statuses corresponding to these retained serial numbers. The servo processor 420 connects to the database 410 and receives stored serial numbers, stored banknote statuses, and banknote verification results, banknote serial numbers, and banknote statuses from the database 410 and the banknote verification machine 300. The servo processor 420 checks if one of the stored serial numbers matches a banknote serial number, generating a first confirmation result, and decides whether to perform a banknote status confirmation operation based on the first confirmation result. When the first confirmation result is yes, the banknote serial number matches one of the stored serial numbers, and the servo processor 420 performs a banknote status confirmation operation based on the stored banknote statuses, generating a second confirmation result, and decides whether to generate an exception notification based on the second confirmation result and the banknote verification result.

[0019] Please refer to Figures 1 and 2 together, where Figure 2 is a flowchart illustrating the banknote image recording and anomaly detection method S0 according to the second embodiment of the present invention. The banknote image recording and anomaly detection method S0 can be used in the banknote image recording and anomaly detection system 100 and includes steps S02, S04, S06, and S08. Step S02 includes storing the retained serial numbers and the corresponding banknote states in the database 410 of the server 400. Step S04 includes capturing images of the banknotes using the camera 200. Step S06 includes inspecting the banknotes using a banknote detector 300 to generate a banknote verification result, analyzing the image to identify the banknote serial number of the corresponding banknote, and recording the banknote state of the corresponding banknote. Step S08 includes using the server processor 420 of the server 400 to confirm whether one of the retained serial numbers is the same as the banknote serial number to generate a first confirmation result, and determining whether to perform a banknote state confirmation operation based on the first confirmation result. When the first confirmation result is yes, the banknote serial number is the same as one of the retained serial numbers. The servo processor 420 performs a banknote status confirmation operation based on the status of these retained banknotes to generate a second confirmation result, and decides whether to generate an abnormal notification based on the second confirmation result and the banknote verification result.

[0020] Therefore, the banknote image recording and anomaly detection system 100 and the banknote image recording and anomaly detection method S0 of the present invention can record the banknote serial numbers entering the bank (banknotes passing through the banknote detector 300 and ready to be deposited into the bank) or leaving the bank (banknotes passing through the banknote detector 300 and ready to be withdrawn from the bank). A network module 330 connects to a database 410 and a server 400 to compare the banknote serial number transaction records on the blockchain, and links them with the counterfeit banknote serial numbers published by the central bank. If there is suspicion of counterfeit banknotes, it can immediately alert bank staff and notify the police. Furthermore, if a customer reports that the withdrawn cash is counterfeit, the system can retrieve the banknote serial number of the cash withdrawn by the customer at that time to prove that it is different from the alleged counterfeit banknote in the customer's possession. It can also prove that the counterfeit banknote was switched by the customer or their cohabitant, and that the financial institution is not involved. This protects the rights and interests of the financial institution and safeguards its image, thereby solving the problem of being unable to prove the authenticity of counterfeit banknotes reported by customers in the past.

[0021] Specifically, the aforementioned banknote detector 300 includes a memory 310, a processor 320, and a network module 330. The memory 310 stores images, banknote verification results, banknote serial numbers, and banknote status. The processor 320 is connected to the memory 310 and is used to verify banknotes to generate verification results and analyze images to identify the corresponding banknote serial numbers. The network module 330 is connected to the processor 320 and is used to transmit the banknote verification results, banknote serial numbers, and banknote status to the server 400 via a network for subsequent transaction blockchain uploading and status verification, and can also report to police units or police systems online. Furthermore, the banknote detector 300 includes an input interface and a screen, both of which are connected to the processor 320. The input interface records staff information, customer information, and banknote status, while the screen displays the entered amount and transaction status.

[0022] The aforementioned memory 310 may include random access memory (RAM) or other types of dynamic storage devices for storing information and instructions to be executed by the processor 320. The processor 320 may be a digital signal processor (DSP), microprocessor (MPU), central processing unit (CPU), or other electronic processor. The processor 320 may include a single device (e.g., single-core) or a group of devices (e.g., multi-core). The network module 330 may be connected to the server 400 via a wired or wireless network. Furthermore, the database 410 may include RAM or other types of dynamic storage devices for storing information and instructions to be executed by the server processor 420. The server processor 420 may include any type of processor, network processor, or cloud processor. The server processor 420 may include a single device (e.g., single-core) or a group of devices (e.g., multi-core). This invention is not limited to the above.

[0023] Please refer to Figures 1 and 3A, where Figure 3A is a flowchart illustrating the banknote image recording and anomaly detection method S2 according to the third embodiment of the present invention. The banknote image recording and anomaly detection method S2 represents the entry process (i.e., banknotes being deposited into a bank) and can be used in the banknote image recording and anomaly detection system 100. The banknote image recording and anomaly detection system 100 further includes a blockchain 500, which is connected to the server 400 and used to receive banknote serial numbers and banknote status to generate a transaction result. The immutability of the blockchain 500 ensures that once data is uploaded to the blockchain, it cannot be modified. This means that no bank employee (BA) can directly modify the data uploaded to the blockchain 500. Even if a BA attempts to tamper with the database 410, it will still be detected because the transaction records on the blockchain 500 cannot be changed. Furthermore, the blockchain 500 has transaction immutability; even if a BA modifies the records in the database 410, the transaction data on the blockchain 500 remains unalterable. Any modifications will result in a mismatch with the transaction records returned by Blockchain 500, thus exposing the tampering. Furthermore, the system periodically compares the data in Database 410 and Blockchain 500 to ensure that the status of each transaction is completely consistent with the transaction sequence number and transaction hash value on Blockchain 500. This is a robust verification method that ensures the data has not been altered.

[0024] The banknote image recording and anomaly detection method S2 includes steps S202, S204, S206, S208, S210, S212, S214, S216, S218, S220, S222, and S224. Step S202 involves the customer's CA (Cash on Delivery) depositing money into the bank, processed by the bank clerk (BA). Step S204 involves the BA inputting the customer's name, account, and clerk information into the banknote verification machine 300. Step S206 involves recording the serial number and status of each banknote, i.e., recording the corresponding banknote serial number and status. Step S208 involves uploading the banknote serial number and status to the server 400 for verification. Step S210 involves comparing the banknote serial number and status with the stored serial numbers and statuses in the database 410.

[0025] Step S212 includes confirming whether there are existing banknote serial numbers, that is, the servo processor 420 confirms whether one of these stored serial numbers is the same as a banknote serial number to generate a first confirmation result. When the first confirmation result is yes (i.e., there is an existing banknote serial number), this of these stored serial numbers corresponds to one of these stored banknote states, and the servo processor 420 executes step S214; when the first confirmation result is no (i.e., there is no existing banknote serial number), the servo processor 420 executes step S220.

[0026] Step S214 includes confirming whether the banknote status is consistent, that is, performing a banknote status confirmation operation. The banknote status confirmation operation includes the server processor 420 confirming whether the current status of these stored banknotes is the same as the banknote status to generate a second confirmation result. When the second confirmation result is yes (meaning that both are entering or leaving the banknotes, and there are two banknotes with the same serial number in the banknotes), the server processor 420 executes step S216; when the second confirmation result is no (meaning that one is entering the banknotes and the other is leaving the banknotes, the system continues the previously set status, and there are no banknotes with the same serial number in the banknotes), the server processor 420 executes step S218. Step S216 includes generating and issuing an exception notification. Step S218 includes performing a status change operation on the banknote status, which includes changing the banknote status to entering information.

[0027] Step S220 includes uploading banknote information to blockchain 500, that is, server processor 420 transmits banknote serial number and banknote status to blockchain 500. Step S222 includes blockchain 500 sending back the transaction result, that is, blockchain 500 generates and sends back a transaction serial number and a transaction hash value to server 400 based on banknote serial number and banknote status. The transaction result includes transaction serial number and transaction hash value. Step S224 includes storing in database 410, that is, storing transaction serial number and transaction hash value in database 410.

[0028] Please refer to Figures 1, 3A, and 3B. Figure 3B illustrates the flowchart of the banknote image recording and anomaly detection method S2 in Figure 3A after an anomaly notification is issued. After the anomaly notification is issued, the banknote image recording and anomaly detection system 100 executes steps S226, S228, S230, S232, S234, S236, S238, S240, and S242. Step S226 involves confirming the customer's name, account, and staff information. Step S228 involves manually verifying the customer's CA banknote. Step S230 involves determining the banknote is genuine, and then executing steps S218, S220, S222, and S224. Step S232 involves locating the customer's CB banknote within the bank. Step S234 involves manually verifying the customer's CB banknote. Step S236 involves determining the banknote is counterfeit. Step S238 involves sending an alert email to the department head and the staff member handling the issue, BA. Step S240 includes connecting to the online reporting system (connecting to the police unit), that is, reporting that there is currently a banknote with the same serial number "XND0339911" and the banknote status has been entered. Step S242 includes updating the database 410 status (counterfeit banknote), that is, the database 410 updates the data according to the counterfeit banknote and then stores the counterfeit banknote serial number (step S224).

[0029] Please refer to Figures 1, 3A, 3B, and 4, where Figure 4 is a flowchart illustrating the banknote image recording and anomaly detection method S4 according to the fourth embodiment of the present invention. The banknote image recording and anomaly detection method S4 represents the process of banknotes being withdrawn from the bank and can be used in the banknote image recording and anomaly detection system 100. The banknote image recording and anomaly detection method S4 includes steps S402, S404, S406, S408, S410, S412, S414, S416, S418, S420, S422, and S424. Step S402 involves the customer's CA (Cash on Delivery Card) withdrawing cash from the bank, which is processed by the bank teller (BA). Steps S404, S406, S408, S410, S412, S414, and S416 are the same as steps S204, S206, S208, S210, S212, S214, and S216 in Figure 3A, and will not be described again. Step S418 involves performing a state change operation on the banknote's state, which includes changing the banknote's state to travel information. Steps S420, S422, and S424 are the same as steps S220, S222, and S224 in Figure 3A, and will not be described again.

[0030] In other embodiments, the banknote image recording and anomaly detection method S4 may only include steps S402, S404, S406, S408, S418, S420, S422, and S424; in other words, the banknotes withdrawn by the customer's CA are banknotes already recorded on the server 400, so steps S410, S412, S414, and S416 (which can be regarded as confirmation actions of banknote serial number and banknote status) do not need to be executed during withdrawal, which can improve withdrawal efficiency. The present invention is not limited thereto.

[0031] Please refer to Figures 1, 3A, 3B, 4, and 5, where Figure 5 is a flowchart illustrating the banknote image recording and anomaly detection method S6 of the fifth embodiment of the present invention. The banknote image recording and anomaly detection method S6 represents the process of banknotes being withdrawn from the bank and can be used in the banknote image recording and anomaly detection system 100. The banknote image recording and anomaly detection method S6 includes steps S602, S604, S606, S608, S610, S612, S614, S616, S618, S620, S622, and S624. Step S602 includes the customer (CA) withdrawing cash from an automated teller machine (ATM). Step S604 includes the bank teller (BA) inputting teller information into the banknote verification machine 300. Step S606 includes recording the serial number and status of each banknote, that is, recording the banknote serial number and banknote status of the corresponding banknote. The ATM records all information from the customer's cash withdrawal via CA, as well as bank employee information, banknote serial number, and banknote status from the banknote detector 300. Steps S608, S610, S612, S614, S616, S618, S620, S622, and S624 are the same as steps S408, S410, S412, S414, S416, S418, S420, S422, and S424 in Figure 4, and will not be repeated here.

[0032] In other embodiments, the banknote image recording and anomaly detection method S6 may only include steps S602, S604, S606, S608, S618, S620, S622, and S624; in other words, the banknotes withdrawn by the customer's CA are banknotes already recorded on the server 400, so steps S610, S612, S614, and S616 (which can be regarded as confirmation actions of banknote serial number and banknote status) do not need to be executed again when withdrawing money, which can improve withdrawal efficiency. The present invention is not limited thereto.

[0033] In one embodiment, the banknote detector 300 of the present invention is configured to perform the following operations: (1) using a standard banknote (one of a plurality of valuable banknotes) issued by a central bank as a comparison template, wherein these valuable banknotes have different denominations; (2) performing a grayscale processing on the comparison template to generate a grayscale comparison template, which can avoid the influence of different ambient light sources; (3) marking a serial number area in the grayscale comparison template. Since the serial number area of ​​a valuable banknote is a fixed range, marking the serial number area helps with subsequent image object recognition; and (4) analyzing the grayscale comparison template based on the serial number area and a text recognition model to identify the corresponding standard banknote's stored serial number. The stored serial number is transmitted to the database 410 of the server 400 for subsequent comparison analysis.

[0034] In one embodiment, the banknote detector 300 of the present invention is configured to perform the following operations: (1) using a standard banknote issued by a central bank as a comparison template; (2) performing grayscale processing on the comparison template to generate a grayscale comparison template; (3) performing an image subtraction operation on the image captured by the camera 200 and the grayscale comparison template to generate a difference image feature, which helps to determine the degree of banknote wear and soiling and image displacement; (4) calculating the area ratio of the difference image feature relative to the area of ​​a banknote in the image, and confirming whether the area ratio exceeds a preset ratio threshold to generate a soiling confirmation result, and recording the area ratio (soiling ratio) and the coordinates of the soiled points in the database 410. Assuming that the soiling exceeds 50% (i.e., the preset ratio threshold is equal to 50%, and the soiling confirmation result is yes), it can be automatically noted in the database 410 as a procedure for subsequent banknote exchange, which can also increase the bank's revenue; and (5) determining whether the banknote in the corresponding image needs to be exchanged with the central bank based on the soiling confirmation result. According to the Central Bank's regulations, in order to improve the quality of circulating banknotes, dirty and worn banknotes are collected and sent to the Bank of Taiwan during the collection process. The Central Bank provides a certain reward to financial institutions for returning damaged banknotes. In addition, after the banknotes are exchanged, their serial numbers are recorded in the database to prevent the same serial numbers from entering the banking system in the future.

[0035] In one embodiment, the banknote detector 300 of the present invention is configured to perform operations including: (1) using a standard banknote issued by a central bank as a comparison template; (2) performing grayscale processing on the comparison template to generate a grayscale comparison template; (3) marking a key icon position in the grayscale comparison template, wherein the key icon position includes a monetary amount position or a pattern position; and (4) determining whether there is icon displacement at the key icon position based on the difference image features and a preset offset threshold value (e.g., 3%) to generate a displacement judgment result. When the offset of one of the difference image features is greater than or equal to the preset offset threshold value, the banknote detector 300 determines that the banknote has icon displacement (the banknote is counterfeit); conversely, when the offset of the difference image features is less than the preset offset threshold value, the banknote detector 300 determines that the banknote has no icon displacement (the banknote is genuine).

[0036] In one embodiment, the valuable banknote field of the database 410 may include: (1) Transaction ID, which uniquely identifies the detection process of each banknote; (2) Employee ID, which records the information of the bank staff operating the banknote detection machine (i.e., banknote detector 300); (3) Customer ID, which records the information of the customer who submits banknotes for detection; (4) Machine ID, which records the equipment number of the counterfeit banknote reporting machine (i.e., banknote detector 300) used; (5) Detection timestamp, which records the time of counterfeit banknote detection; and (6) Wear Level, which is calculated by the storage model. The degree of wear or soiling of the banknote is expressed as a percentage; (7) Wear Area Percentage: records the proportion of the worn area relative to the whole banknote; (8) Wear Area Coordinates: records the coordinate position of the worn area; (9) Wear Threshold Exceeded: if the wear exceeds 50%, it is marked as True; (10) Displacement Status: marks whether the displacement of the key icon position has been detected; (11) Displacement Coordinates: records the icon's position. (12) Displacement Percentage: Records the proportion of displacement compared to the template; (13) Displacement Threshold Exceeded: Marked as True if the offset exceeds 3%; (14) Blockchain Status: Records whether the transaction has been successfully uploaded to Blockchain 500; (15) Blockchain Transaction ID; (16) Blockchain Transaction Hash; (17) Police Notification Status (18) Serial Number, which records whether the banknote serial number output by the image recognition is automatically reported to the police unit; (19) Serial Number Match Status, which marks whether the banknote serial number matches the serial number stored in the database 410; (20) Transaction Type, which records the banknote's entry or exit status; (21) Anomaly Detected, which marks whether anomalies are detected in the serial number or other aspects; and (22) Remarks, which records any other special circumstances or manually entered information during the detection process.

[0037] In one embodiment, the server processor 420 updates a counterfeit banknote serial number group published by the central bank using a web crawler, compares the banknote serial numbers with the counterfeit banknote serial number group to generate a counterfeit banknote comparison result, and decides whether to generate an exception notification based on the counterfeit banknote comparison result. The counterfeit banknote serial number group contains at least one counterfeit banknote serial number. When the counterfeit banknote comparison result is yes, the server processor 420 generates an exception notification; otherwise, when the counterfeit banknote comparison result is no, the server processor 420 does not generate an exception notification.

[0038] In one embodiment, the abnormal serial number (counterfeit banknote serial number) field of the database 410 may include: (1) a unique identifier (ID) for internal tracking and management; (2) a counterfeit banknote serial number (SerialNumber) storing the serial number identified as counterfeit; (3) a face value (Denomination) recording the face value of the counterfeit banknote, such as 100 yuan, 500 yuan, etc.; (4) a report date (DateReported) recording the date the serial number was announced as counterfeit by the central bank; and (5) a source (Source) recording the source of the serial number information, such as "central bank", "financial institution", etc. In this way, the present invention establishes a database of suspected counterfeit banknote serial numbers. When the serial number of an incoming banknote matches a counterfeit banknote serial number in the database, an abnormal status is reported.

[0039] In one embodiment, the servo processor 420 is connected to a police unit. When an anomaly notification is generated, the servo processor 420 retrieves the employee name and customer name corresponding to the banknote serial number from the database 410 and sends an alert email to either a department head's email address or the email address of the employee currently handling the issue. Furthermore, the servo processor 420 generates a notification document based on the employee and customer names and transmits the notification document to the police unit. In addition to the employee and customer names, the alert email and notification document may include more details about the abnormal banknote, such as the banknote serial number, denomination, and counterfeit banknote image.

[0040] As can be seen from the above embodiments, the present invention has the following advantages: First, it can record the serial numbers of banknotes entering or leaving the bank, and has a network module connected to a database and server to compare the banknote serial number transaction records with the blockchain, and link it with the counterfeit banknote serial numbers published by the central bank. If there is any suspicion of counterfeit banknotes, it can immediately alert bank staff and notify the police. Second, if a customer reports that the cash withdrawn is counterfeit, the serial number of the banknote taken by the customer at that time can be retrieved from the system to prove that it is different from the so-called counterfeit banknote in the customer's possession, and can prove that the counterfeit banknote was switched by the customer or their cohabitant, and has nothing to do with the financial institution, thereby protecting the rights and interests of the financial institution and defending its image, thus solving the problem of not being able to prove the authenticity of the counterfeit banknotes reported by customers in the past. Third, by collecting the characteristics of banknote serial numbers and their condition, a history of valuable paper can be established, and its entry and exit records can be uploaded to the blockchain. Through the immutability and transparency of blockchain technology, the integrity of the data is ensured to prevent unauthorized modification. Since one of the core features of blockchain technology is that the data on the chain cannot be tampered with, it provides a strong protection mechanism for the management of banknotes within financial institutions. It can effectively prevent malicious operations by internal personnel, and realize the tracing and identification of abnormal situations. At the same time, it can avoid the disputes that have occurred in the past when customers reported counterfeit banknotes.

[0041] Although the present invention has been disclosed above by way of embodiments, it is not intended to limit the present invention. Anyone skilled in the art can make various modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the invention shall be determined by the claims attached.

[0042] 100: Banknote Image Recording and Anomaly Detection System 200: Camera 300: Banknote Detector 310: Memory 320: Processor 330: Network Module 400: Server 410: Database 420: Servo Processor 500: Blockchain BA: Banker CA,CB: Customers S0, S2, S4, S6: Banknote Image Recording and Anomaly Detection Methods S02, S04, S06, S08, S202, S204, S206, S208, S210, S212, S214, S216, S218, S220, S222, S224, S226, S228, S230, S232, S234, S236, S238, S240, S242, S402, S404, S406, S408, S410, S412, S414, S416, S418, S420, S422, S424, S602, S604, S606, S608, S610, S612, S614, S616, S618, S620, S622, S624: Steps

Claims

1. A banknote image recording and anomaly detection system, comprising: a camera for capturing an image of a banknote; a banknote detector connected to the camera and receiving the image, the banknote detector being used to inspect the banknote and generate an inspection result, and analyzing the image to identify a banknote serial number corresponding to the banknote, and recording the banknote status corresponding to the banknote; and a server connected to the banknote detector, comprising: a database storing a plurality of retained serial numbers and a plurality of retained banknote statuses corresponding to the retained serial numbers; and a server processor connected to the database and receiving the retained serial numbers, the retained banknote statuses, the inspection result, the banknote serial number, and the banknote status from the banknote detector, the server processor confirming whether one of the retained serial numbers is the same as the banknote serial number and generating a first confirmation result, and determining whether to perform a banknote status confirmation operation based on the first confirmation result; wherein... When the first confirmation result is yes, the banknote serial number is the same as one of the stored serial numbers. The server processor performs the confirmation operation on the banknote status according to the stored banknote status to generate a second confirmation result, and decides whether to generate an abnormal notification based on the second confirmation result and the banknote verification result. The banknote verification machine includes: a memory for storing the image, the banknote verification result, the banknote serial number, and the banknote status; a processor connected to the memory for verifying the banknote to generate the banknote verification result and analyzing the image to identify the banknote serial number corresponding to the banknote; and a network module connected to the processor for transmitting the banknote verification result, the banknote serial number, and the banknote status through a network. The banknote status is either entry information or exit information.

2. The banknote image recording and anomaly detection system as described in claim 1 further includes: a blockchain connected to the server and used to receive the banknote serial number and the banknote status to generate a transaction result.

3. The banknote image recording and anomaly detection system as described in claim 2, wherein when the first confirmation result is negative, the server processor transmits the banknote serial number and the banknote status to the blockchain, the blockchain generates and sends back a transaction serial number and a transaction hash value to the database of the server based on the banknote serial number and the banknote status, and the transaction result includes the transaction serial number and the transaction hash value.

4. The banknote image recording and anomaly detection system as described in claim 1, wherein the retained serial numbers correspond to one of the retained banknote states, and the banknote state confirmation operation includes generating the second confirmation result by the server processor confirming whether the retained banknote states are the same as the banknote state; wherein, When the second confirmation result is yes, the servo processor generates the exception notification; wherein, when the second confirmation result is no, the servo processor performs a status change operation on the banknote status, the status change operation including changing the banknote status to either the entry information or the exit information.

5. The banknote image recording and anomaly detection system as described in claim 1, wherein the banknote detector is configured to perform the following operations: using a standard banknote issued by a central bank as a comparison template; performing grayscale processing on the comparison template to generate a grayscale comparison template; marking a serial number area in the grayscale comparison template; and analyzing the comparison template based on the serial number area and a character recognition model to identify the corresponding retained serial numbers of the standard banknote.

6. The banknote image recording and anomaly detection system as described in claim 1, wherein the banknote detector is configured to perform the following operations: using a standard banknote issued by a central bank as a comparison template; performing grayscale processing on the comparison template to generate a grayscale comparison template; performing an image subtraction operation on the image and the grayscale comparison template to generate a difference image feature; calculating the area ratio of the difference image feature relative to the area of ​​a banknote in the image, and confirming whether the area ratio exceeds a preset ratio threshold to generate a damage confirmation result; and determining, based on the damage confirmation result, whether the banknote corresponding to the image needs to be exchanged with the central bank.

7. The banknote image recording and anomaly detection system as described in claim 6, wherein the banknote detector is configured to perform operations including: marking a key icon position of one of the grayscale comparison templates, wherein the key icon position includes a monetary amount position or a pattern position; and determining whether there is icon displacement at the key icon position based on the difference image features and a preset offset threshold value, thereby generating a displacement judgment result; wherein, When the offset of one of the difference image features is greater than or equal to the preset offset threshold, the banknote detector determines that the banknote has icon displacement; wherein, when the offset of the difference image feature is less than the preset offset threshold, the banknote detector determines that the banknote has no icon displacement.

8. The banknote image recording and anomaly detection system as described in claim 1, wherein the server processor updates a counterfeit banknote serial number group published by the central bank using a web crawler, compares the banknote serial number with the counterfeit banknote serial number group to generate a counterfeit banknote comparison result, and determines whether to generate the anomaly notification based on the counterfeit banknote comparison result; wherein, The counterfeit banknote serial number group contains at least one counterfeit banknote serial number; wherein, when the counterfeit banknote comparison result is yes, the servo processor generates the exception notification; wherein, when the counterfeit banknote comparison result is no, the servo processor does not generate the exception notification.

9. The banknote image recording and anomaly detection system as described in claim 1, wherein the server processor is connected to a police unit, and when the anomaly notification is generated, the server processor retrieves the banker's name and a customer's name corresponding to the banknote serial number from the database and sends an alert email to either a department head's email address or a currently handling banker's email address; and the server processor generates a notification document based on the banker's name and the customer's name and transmits the notification document to the police unit.

Images