TW202309818A

TW202309818A - Automatic teller machine and money mule alert method

Info

Publication number: TW202309818A
Application number: TW111129054A
Authority: TW
Inventors: 廖芳育
Original assignee: 華南商業銀行股份有限公司
Priority date: 2021-08-16
Filing date: 2022-08-03
Publication date: 2023-03-01

Abstract

An automatic teller machine (ATM) includes an input interface, a communication device, a camera device, and a processor. The communication device sends an input content obtained from the input interface to a transaction server, and receives card information from the transaction server. After the input interface is operated, the camera device captures the first image during an initial period, and captures the second image after a specified time interval. The processor calculates the difference degree between the first and second images, and triggers a first judgment event and a second judgment event when the difference degree is greater than a threshold value. The first judgment event is to determine whether the card information corresponds to one single card, and the second judgment event is to determine whether the input content corresponds to a withdrawal operation or a transfer operation. When the first and second judgment events are both true, the processor controls the communication device to send an alert signal to the external server.

Description

自動櫃員機及詐騙車手警報方法ATM and Scam Rider Alert Method

本發明係關於自動櫃員機，特別是一種針對詐騙車手發出警報的自動櫃員機及警報方法。The invention relates to an automatic teller machine, in particular to an automatic teller machine and an alarm method for issuing an alarm against fraudulent drivers.

自動櫃員機（Automatic Teller Machine，ATM）可提供一般民眾提款、存款、繳費及轉帳等服務。自動櫃員機的據點也從金融機構所在地的週邊增設至民眾住所附近的便利商店。自動櫃員機現行採用的安全機制主要係以監視攝像機拍攝，當發生盜領案件或是有其他金融操作上的問題需要釐清時，就需要調閱監視器請專人逐一比對。Automatic teller machine (Automatic Teller Machine, ATM) can provide general public with withdrawal, deposit, payment and transfer services. The bases of ATMs have also expanded from the vicinity of financial institutions to convenience stores near people's residences. The current security mechanism used by ATMs is mainly based on surveillance cameras. When there is a case of theft or other financial operations that need to be clarified, it is necessary to consult the monitors and ask someone to compare them one by one.

然而，金融卡盜領案件層出不窮，詐騙集團車手（簡稱詐騙車手）猖獗，且其犯案工具及手法日益精良，現有的自動櫃員機的安全機制根本難以防範，造成金融機構及客戶的損失，亦形成社會治安的嚴重問題。However, financial card theft cases are emerging one after another, fraudulent group drivers (referred to as fraudulent drivers) are rampant, and their crime tools and methods are becoming more and more sophisticated. The existing ATM security mechanism is difficult to prevent, causing financial institutions and customers. Serious problems of law and order.

雖然監視攝影機雖能夠拍攝到使用自動櫃員機的民眾，但是等到盜領案件發生後，需要查明某個特定使用者時，仍需要以人力耗費大量時間觀看影像。此外，雖然現在已可透過臉部辨識的方式驗證使用者，然而這種傳統的方法仍需回傳大量的影像資訊至金融機構的伺服器，對於遍佈在各地且提估全天候服務的自動櫃員機而言，勢必需要布設大量的伺服器才能滿足提供臉部辨識服務給每一台自動櫃員機的需求，這樣一來，大幅提高了硬體成本和維修上的成本。Although surveillance cameras can capture people using ATMs, when a specific user needs to be identified after a theft case occurs, it still takes a lot of time to watch the images manually. In addition, although it is now possible to verify users through facial recognition, this traditional method still needs to return a large amount of image information to the server of the financial institution. In other words, it is necessary to deploy a large number of servers to meet the needs of providing facial recognition services to each ATM, which greatly increases the cost of hardware and maintenance.

有鑑於此，本發明提出一種自動櫃員機及詐騙車手警報方法，藉此解決上述問題。In view of this, the present invention proposes an automatic teller machine and a method for alerting fraudulent drivers to solve the above problems.

依據本發明一實施例的一種詐騙車手警報方法，適用於一自動櫃員機，該自動櫃員機具有一輸入介面、一通訊裝置、一攝像裝置及一處理器，該詐騙車手警報方法包含以處理器進行以下步驟：透過該輸入介面取得一輸入內容；控制該通訊裝置發送該輸入內容至一交易伺服器；透過該通訊裝置從該交易伺服器接收一卡片資訊；控制該攝像裝置於一起始期間拍攝一第一影像，並於該起始期間後控制該攝像裝置隔一指定時間拍攝一第二影像，其中該起始期間開始於該輸入介面***作之後；依據該第一影像及該第二影像計算一差異度；當該差異度大於一門檻值時，觸發一第一判斷事件及一第二判斷事件，其中該第一判斷事件係判斷該卡片資訊是否對應於單一張卡片，該第二判斷事件係判斷該輸入內容是否對應於提款操作或轉帳操作；以及當該第一判斷事件及該第二判斷事件皆為真時，控制該通訊裝置發送一警報訊號至一外部伺服器。According to an embodiment of the present invention, a method for alerting a fraudulent driver is applicable to an automatic teller machine, the automatic teller machine has an input interface, a communication device, a camera device and a processor, and the method for alerting a fraudulent driver includes performing the following steps with the processor Steps: Obtain an input content through the input interface; control the communication device to send the input content to a transaction server; receive a card information from the transaction server through the communication device; control the camera device to take pictures of a first An image, and after the initial period, control the camera device to take a second image at a specified time interval, wherein the initial period starts after the input interface is operated; calculate a degree of difference; when the degree of difference is greater than a threshold value, a first judgment event and a second judgment event are triggered, wherein the first judgment event is to judge whether the card information corresponds to a single card, and the second judgment event is judging whether the input content corresponds to a cash withdrawal operation or a transfer operation; and controlling the communication device to send an alarm signal to an external server when both the first judging event and the second judging event are true.

依據本發明一實施例的一種詐騙車手警報方法，另包含以該處理器進行以下步驟：在該差異度小於或等於該門檻值時，累加一計數值；以及當該計數值大於一設定值時，觸發該第二判斷事件及一第三判斷事件，其中該第三判斷事件係判斷該卡片資訊中的一行為範本與該輸入內容的一輸入行為是否相異，該輸入行為對應於多個輸入值，且該輸入內容記錄該些輸入值之間的時間差；以及當該第二判斷事件及該第三判斷事件皆為真時，控制該通訊裝置發送該警報訊號至該外部伺服器。A method for alerting fraudulent drivers according to an embodiment of the present invention further includes the following steps by using the processor: when the difference is less than or equal to the threshold value, accumulating a count value; and when the count value is greater than a set value , triggering the second judging event and a third judging event, wherein the third judging event is judging whether a behavior template in the card information is different from an input behavior of the input content, and the input behavior corresponds to multiple inputs value, and the input content records the time difference between these input values; and when both the second judgment event and the third judgment event are true, control the communication device to send the alarm signal to the external server.

依據本發明一實施例的一種詐騙車手警報方法，另包含以該處理器進行以下步驟：在該差異度小於或等於該門檻值時，累加一計數值；以及當該計數值大於一設定值時，觸發該第一判斷事件及一第四判斷事件，其中該第四判斷事件係判斷該輸入內容對應於非指定帳戶之轉帳操作；當該第一判斷事件及該第四判斷事件皆為真時，控制該通訊裝置發送該警報訊號至該外部伺服器。A method for alerting fraudulent drivers according to an embodiment of the present invention further includes the following steps by using the processor: when the difference is less than or equal to the threshold value, accumulating a count value; and when the count value is greater than a set value , triggering the first judging event and a fourth judging event, wherein the fourth judging event is judging that the input content corresponds to a transfer operation of an unspecified account; when both the first judging event and the fourth judging event are true , controlling the communication device to send the alarm signal to the external server.

依據本發明一實施例的一種詐騙車手警報方法，另包含以該處理器進行以下步驟：在該差異度小於或等於該門檻值時，累加一計數值；以及當該計數值大於一設定值時，判斷該第一判斷事件及一第五判斷事件，其中該第五判斷事件係該卡片資訊中的前次提領地點與本次提領地點相異；以及當該第一判斷事件及該第五判斷事件皆為真時，控制該通訊裝置發送該警報訊號至該外部伺服器。A method for alerting fraudulent drivers according to an embodiment of the present invention further includes the following steps by using the processor: when the difference is less than or equal to the threshold value, accumulating a count value; and when the count value is greater than a set value , judging the first judging event and a fifth judging event, wherein the fifth judging event is that the previous withdrawal location in the card information is different from the current withdrawal location; and when the first judging event and the second 5. When judging that the events are all true, controlling the communication device to send the alarm signal to the external server.

依據本發明一實施例的一種自動櫃員機，包括：一輸入介面，用以取得一輸入內容；一通訊裝置，用以與一交易伺服器交換資訊；一攝像裝置，用以拍攝影像；以及一處理器，電性連接該輸入介面、該通訊裝置及該攝像裝置，其中該處理器用以控制該通訊裝置發送該輸入內容至該交易伺服器，透過該通訊裝置從該交易伺服器接收該卡片資訊；該處理器控制該攝像裝置於一起始期間拍攝一第一影像，並於該起始期間後控制該攝像裝置隔一指定時間拍攝一第二影像，其中該起始期間開始於該輸入介面***作之後；該處理器依據該第一影像及該第二影像計算一差異度，當該差異度大於一門檻值時，觸發一第一判斷事件及一第二判斷事件，其中該第一判斷事件係判斷該卡片資訊是否對應於單一張卡片，該第二判斷事件係判斷該輸入內容是否對應於提款操作或轉帳操作；以及當該第一判斷事件及該第二判斷事件皆為真時，控制該通訊裝置發送一警報訊號至一外部伺服器。An automatic teller machine according to an embodiment of the present invention includes: an input interface for obtaining an input content; a communication device for exchanging information with a transaction server; a camera device for shooting images; and a processing a device electrically connected to the input interface, the communication device and the camera device, wherein the processor is used to control the communication device to send the input content to the transaction server, and receive the card information from the transaction server through the communication device; The processor controls the imaging device to capture a first image during an initial period, and controls the imaging device to capture a second image at a specified time interval after the initial period, wherein the initial period begins when the input interface is operated Afterwards; the processor calculates a degree of difference based on the first image and the second image, and when the degree of difference is greater than a threshold, a first judgment event and a second judgment event are triggered, wherein the first judgment event is Judging whether the card information corresponds to a single card, the second judging event is judging whether the input content corresponds to a cash withdrawal operation or a transfer operation; and when both the first judging event and the second judging event are true, the control The communication device sends an alarm signal to an external server.

依據本發明一實施例的一種自動櫃員機，其中：在該差異度小於或等於該門檻值時，該處理器累加一計數值；當該計數值大於一設定值時，該處理器判斷該第二判斷事件及一第三判斷事件，其中該第三判斷事件係比對該卡片資訊中的一行為範本與該輸入內容的一輸入行為是否相異，該輸入行為對應於多個輸入值，且輸入內容記錄該些輸入值之間的時間差；以及當該第二判斷事件及該第三判斷事件皆為真時，該處理器控制該通訊裝置發送該警報訊號至該外部伺服器。An automatic teller machine according to an embodiment of the present invention, wherein: when the difference is less than or equal to the threshold value, the processor accumulates a count value; when the count value is greater than a set value, the processor judges that the second A judging event and a third judging event, wherein the third judging event compares whether a behavior template in the card information is different from an input behavior of the input content, the input behavior corresponds to a plurality of input values, and the input The content records the time difference between the input values; and when the second judgment event and the third judgment event are both true, the processor controls the communication device to send the alarm signal to the external server.

依據本發明一實施例的一種自動櫃員機，其中：在該差異度小於或等於該門檻值時，該處理器累加一計數值；以及當該計數值大於一設定值時，該處理器判斷該第一判斷事件及一第四判斷事件，其中該第四判斷事件係判斷該輸入內容對應於非指定帳戶之轉帳操作；當該第一判斷事件及該第四判斷事件皆為真時，該處理器控制該通訊裝置發送該警報訊號至該外部伺服器。An automatic teller machine according to an embodiment of the present invention, wherein: when the difference degree is less than or equal to the threshold value, the processor accumulates a count value; and when the count value is greater than a set value, the processor judges that the first A judging event and a fourth judging event, wherein the fourth judging event is judging that the input content corresponds to a transfer operation of an unspecified account; when both the first judging event and the fourth judging event are true, the processor controlling the communication device to send the alarm signal to the external server;

依據本發明一實施例的一種自動櫃員機，其中：在該差異度小於或等於該門檻值時，該處理器累加一計數值；以及當該計數值大於一設定值時，該處理器判斷該第一判斷事件及一第五判斷事件，其中該第五判斷事件係該卡片資訊中的前次提領地點與本次提領地點相異；以及當該第一判斷事件及該第五判斷事件皆為真時，該處理器控制該通訊裝置發送該警報訊號至該外部伺服器。An automatic teller machine according to an embodiment of the present invention, wherein: when the difference degree is less than or equal to the threshold value, the processor accumulates a count value; and when the count value is greater than a set value, the processor judges that the first A judging event and a fifth judging event, wherein the fifth judging event is that the previous withdrawal location in the card information is different from the current withdrawal location; and when the first judging event and the fifth judging event are both When true, the processor controls the communication device to send the alarm signal to the external server.

依據本發明一實施例的一種詐騙車手警報方法，適用於一自動櫃員機，該自動櫃員機具有一輸入介面、一通訊裝置、一攝像裝置及一處理器，該詐騙車手警報方法包含以處理器進行以下步驟：透過該輸入介面取得一輸入內容；控制該通訊裝置發送該輸入內容至一交易伺服器；透過該通訊裝置從該交易伺服器接收一卡片資訊；控制該攝像裝置於一起始期間拍攝一第一影像，並於該起始期間後控制該攝像裝置隔一指定時間拍攝一第二影像，其中該起始期間開始於該輸入介面***作之後；以及依據該第一影像及該第二影像計算一差異度，當該差異度大於一門檻值時，觸發以下步驟：判斷該卡片資訊是否對應於單一張卡片，以產生一第一權重值；判斷該輸入內容是否對應於提款操作或轉帳操作，以產生一第二權重值；以及當該第一權重值以及該第二權重值的加總值或乘積值大於一預設值時，控制該通訊裝置發送一警報訊號至一外部伺服器。According to an embodiment of the present invention, a method for alerting a fraudulent driver is applicable to an automatic teller machine, the automatic teller machine has an input interface, a communication device, a camera device and a processor, and the method for alerting a fraudulent driver includes performing the following steps with the processor Steps: Obtain an input content through the input interface; control the communication device to send the input content to a transaction server; receive a card information from the transaction server through the communication device; control the camera device to take pictures of a first an image, and control the imaging device to take a second image at a specified time interval after the initial period, wherein the initial period starts after the input interface is operated; and calculate based on the first image and the second image A degree of difference, when the degree of difference is greater than a threshold value, trigger the following steps: determine whether the card information corresponds to a single card to generate a first weight value; determine whether the input content corresponds to a withdrawal operation or a transfer operation , to generate a second weight value; and when the sum or product of the first weight value and the second weight value is greater than a preset value, control the communication device to send an alarm signal to an external server.

綜上所述，本發明無須透過人工方式以肉眼作人物比對，並且能夠在短時間內鎖定部分可疑人物，相較於現行技術能夠節省大量時間和成本。當民眾意外受騙而匯款時，本發明能夠減少調閱監視器逐一比對所耗費的時間，加快逮捕詐騙車手減少民眾損失，也藉由自動櫃員機提供的日誌檔（log）判別其行為，當行為異常時能夠從之前拍攝的第一影像，盡快發現詐騙車手。To sum up, the present invention does not need to manually compare persons with the naked eye, and can locate some suspicious persons in a short time, which saves a lot of time and cost compared with the current technology. When the public is accidentally deceived and remits money, the invention can reduce the time it takes to check the monitors for one-by-one comparison, speed up the arrest of fraudulent drivers and reduce the loss of the public, and also use the log file (log) provided by the automatic teller machine to identify their behavior. When there is an abnormality, the fraudulent driver can be found as soon as possible from the first image taken before.

以上之關於本揭露內容之說明及以下之實施方式之說明係用以示範與解釋本發明之精神與原理，並且提供本發明之專利申請範圍更進一步之解釋。The above description of the disclosure and the following description of the implementation are used to demonstrate and explain the spirit and principle of the present invention, and provide a further explanation of the patent application scope of the present invention.

以下在實施方式中詳細敘述本發明之詳細特徵以及特點，其內容足以使任何熟習相關技藝者了解本發明之技術內容並據以實施，且根據本說明書所揭露之內容、申請專利範圍及圖式，任何熟習相關技藝者可輕易地理解本發明相關之構想及特點。以下之實施例係進一步詳細說明本發明之觀點，但非以任何觀點限制本發明之範疇。The detailed features and characteristics of the present invention are described in detail below in the implementation mode, and its content is enough to enable any person familiar with the relevant art to understand the technical content of the present invention and implement it accordingly, and according to the content disclosed in this specification, the scope of the patent application and the drawings , anyone who is familiar with the related art can easily understand the ideas and features related to the present invention. The following examples are to further describe the concept of the present invention in detail, but not to limit the scope of the present invention in any way.

圖1是本發明一實施例的自動櫃員機的方塊架構圖，如圖1所示，自動櫃員機20包括輸入介面1、通訊裝置3、攝像裝置5及處理器7。處理器7電性連接輸入介面1、通訊裝置3及攝像裝置5。通訊裝置3通訊連接此自動櫃員機20所屬之金融機構的交易伺服器40，且通訊裝置3通訊連接屬於警政單位的外部伺服器60。FIG. 1 is a block diagram of an automatic teller machine according to an embodiment of the present invention. As shown in FIG. 1 , the automatic teller machine 20 includes an input interface 1 , a communication device 3 , a camera device 5 and a processor 7 . The processor 7 is electrically connected to the input interface 1 , the communication device 3 and the camera device 5 . The communication device 3 is connected to the transaction server 40 of the financial institution to which the automatic teller machine 20 belongs, and the communication device 3 is connected to the external server 60 belonging to the police department.

輸入介面1例如為鍵盤或觸控式螢幕，用以取得輸入內容。通訊裝置3用以與交易伺服器40交換資訊。攝像裝置5用以拍攝操作自動櫃員機20的人員的影像。處理器7則用以控制上述元件執行本發明一實施例的詐騙車手警報方法。The input interface 1 is, for example, a keyboard or a touch screen for obtaining input content. The communication device 3 is used for exchanging information with the transaction server 40 . The camera device 5 is used to capture images of people operating the ATM 20 . The processor 7 is used to control the above components to execute the method for alerting a fraudulent driver according to an embodiment of the present invention.

處理器7例如採用下列範例中的一或數者進行實作：處理器，例如中央處理器單元（central processor unit，CPU）、圖形處理器單元（graphic processing unit）及/或應用處理器（application processor，AP）；邏輯晶片，例如現場可程式化閘陣列（field programmable gate array，FPGA）及特殊應用IC（application-specific integrated circuit，ASIC）。The processor 7 is, for example, implemented by one or more of the following examples: a processor, such as a central processor unit (central processor unit, CPU), a graphics processing unit (graphic processing unit), and/or an application processor (application processor, AP); logic chip, such as field programmable gate array (field programmable gate array, FPGA) and application-specific IC (application-specific integrated circuit, ASIC).

以下配合圖2~圖5說明處理器7執行的內容，圖2~圖5是本發明一實施例的詐騙車手警報方法的流程圖，所述的詐騙車手警報方法適用於圖1所示的自動櫃員機20。如圖2所示，步驟S1為輸入介面1取得輸入內容。在一實施例中，輸入介面1包括設置於自動櫃員機20的讀卡裝置以及觸控式螢幕或鍵盤。當使用者操作自動櫃員機20的輸入介面1時，例如在讀卡裝置***金融卡片以及透過觸控式螢幕或鍵盤輸入資料，輸入介面1便可取得卡片號碼以及使用者輸入的資料等作為輸入內容。輸入內容包括：卡片號碼、卡片密碼的多個數值以及每一個數值的發送時間。關於卡片密碼的多個數值，例如：密碼為123，其中數值「1」在12時00分00秒發送，數值「2」在12時00分01秒發送，數值「3」在12時00分03秒發送，則輸入內容將包含上述這些密碼數值「1」、「2」、「3」、時間數值「12:00:00」、「12:00:01」、「12:00:03」及密碼數值與時間數值之間的對應關係等資訊，但本發明不以上述數值舉例為限。此外，輸入內容還包括使用者透過觸控式螢幕或鍵盤所輸入的操作項目（如提款或轉帳），轉帳帳號（若操作項目為轉帳），提款金額等。在一實施例中，使用者（或是詐騙車手）可能在自動櫃員機20前連續使用多張金融卡片進行操作。因此，輸入內容包括多張卡片的每一者的卡片號碼、卡片密碼數值、數值發送時間、操作項目、轉帳帳號、提款金額等。本發明不限制輸入內容包括一張卡片的相關資訊或是多張卡片的相關資訊。The following describes the content executed by the processor 7 in conjunction with FIGS. 2 to 5. FIGS. Teller machine 20. As shown in FIG. 2 , step S1 obtains input content for the input interface 1 . In one embodiment, the input interface 1 includes a card reading device and a touch screen or a keyboard disposed on the automatic teller machine 20 . When the user operates the input interface 1 of the automatic teller machine 20, such as inserting a financial card into the card reader and inputting data through the touch screen or keyboard, the input interface 1 can obtain the card number and the data input by the user as input content. The input content includes: the card number, multiple values of the card password and the sending time of each value. Regarding the multiple values of the card password, for example: the password is 123, where the value "1" is sent at 12:00:00, the value "2" is sent at 12:00:01, and the value "3" is sent at 12:00 03 seconds, the input content will include the above password values "1", "2", "3", time values "12:00:00", "12:00:01", "12:00:03" And information such as the corresponding relationship between the password value and the time value, but the present invention is not limited to the above numerical values. In addition, the input content also includes the operation items (such as withdrawal or transfer), the transfer account number (if the operation item is transfer), the withdrawal amount, etc. input by the user through the touch screen or the keyboard. In one embodiment, the user (or fraudulent driver) may use multiple financial cards in front of the automatic teller machine 20 to operate continuously. Therefore, the input content includes the card number of each of the plurality of cards, the card password value, the value transmission time, operation items, transfer account number, withdrawal amount, and the like. The present invention does not limit the input content to include related information of one card or related information of multiple cards.

步驟S2為處理器7控制通訊裝置3發送輸入內容至交易伺服器40，所發送的輸入內容包括一或多張卡片的每一者的卡片號碼、卡片密碼數值、數值發送時間、操作項目、轉帳帳號、提款金額等。步驟S3為處理器7透過通訊裝置從交易伺服器40接收卡片資訊。在一實施例中，交易伺服器40可依據輸入內容中的一或多個卡片號碼，在交易伺服器40本地端所連接的資料庫查找對應於每一個卡片號碼所對應的相關資訊，然後將每一個卡片號碼的相關資訊（簡稱為卡片資訊）回傳至自動櫃員機20中的通訊裝置3，處理器7再從通訊裝置3取得卡片資訊。在一實施例中，卡片資訊包括通過交易伺服器40驗證後的卡片號碼，前次使用所述一或多張卡片的地點，本次使用所述一或多張卡片所對應的卡片的地點等。詳言之，通訊裝置3發送卡片號碼及卡片密碼至交易伺服器40，交易伺服器在驗證卡片號碼與卡片密碼相符之後，將通過此驗證的卡片號碼回傳至自動櫃員機20。Step S2 is for the processor 7 to control the communication device 3 to send the input content to the transaction server 40. The input content sent includes the card number of each of one or more cards, the card password value, the value sending time, the operation item, the transfer account number, withdrawal amount, etc. Step S3 is that the processor 7 receives card information from the transaction server 40 through the communication device. In one embodiment, the transaction server 40 can search the database connected to the local end of the transaction server 40 for relevant information corresponding to each card number according to one or more card numbers in the input content, and then The relevant information of each card number (referred to as card information) is sent back to the communication device 3 in the automatic teller machine 20 , and the processor 7 obtains the card information from the communication device 3 . In one embodiment, the card information includes the card number verified by the transaction server 40, the place where the one or more cards were used last time, the place where the card corresponding to the one or more cards is used this time, etc. . In detail, the communication device 3 sends the card number and card password to the transaction server 40 , and the transaction server returns the verified card number to the ATM 20 after verifying that the card number matches the card password.

步驟S4及S5分別為處理器7控制攝像裝置5於起始期間拍攝第一影像，並於起始期間後控制攝像裝置5隔指定時間拍攝第二影像，其中起始期間開始於輸入介面1***作之後。換言之，當使用者開始操作輸入介面1時，攝像裝置5便可拍攝到包含使用者的第一影像（可為全身或臉部影像）。而在使用者操作自動櫃員機20的期間，攝像裝置5每隔一段時間便拍攝包含使用者的第二影像。在一實施例中，本發明係以第一影像為基礎，比對第一影像與第二影像的差異度，藉此判斷兩個影像中的使用者是否為同一人。在另一實施例中，則可採用第N次拍攝的第二影像及第N+1次拍攝的第二影像兩者進行比對。由於本實施例是透過比對影像差異度作為判斷是否有詐騙車手的依據之一，因此本發明不用上傳包含第一影像及第二影像的整個影片檔，故可大幅降低傳輸上的成本和硬體上的成本。此外，由於所上傳的資料量相對於整個影片檔較小，故可在短時間內上傳完畢，加快警方辦案的效率。Steps S4 and S5 are respectively for the processor 7 to control the camera device 5 to take a first image during the initial period, and to control the camera device 5 to take a second image at a specified time interval after the initial period, wherein the initial period begins when the input interface 1 is activated after operation. In other words, when the user starts to operate the input interface 1 , the camera device 5 can capture the first image (which can be a whole body or face image) including the user. While the user is operating the automatic teller machine 20 , the imaging device 5 captures a second image containing the user at regular intervals. In one embodiment, based on the first image, the present invention compares the degree of difference between the first image and the second image, thereby judging whether the users in the two images are the same person. In another embodiment, the second image taken at the Nth time and the second image taken at the N+1th time can be used for comparison. Since the present embodiment is one of the bases for judging whether there is a fraudulent driver by comparing the degree of image difference, the present invention does not need to upload the entire video file including the first image and the second image, so the cost and hardware of transmission can be greatly reduced. body costs. In addition, since the amount of uploaded data is relatively small compared to the entire video file, it can be uploaded in a short time, which speeds up the efficiency of police handling cases.

在一實施例中，比對第一影像與第二影像的差異度的步驟如下：將第一影像與第二影像調整為相同尺寸；將第一影像的每個像素的RGB值分別減去第二影像的對應位置的像素的RGB值，包括將第一影像中座標為(x, y)的像素的RGB值各自減去第二影像中座標為(x,y)的像素的RGB值，由此可以得到三個差值(分別對應R，G，B)；然後計算這三個差值的平均作為座標(x, y)的像素代表的像素差異值；然後將第一影像中所有座標的像素差異值進行加總，作為第一影像和第二影像的差異度。In one embodiment, the step of comparing the difference between the first image and the second image is as follows: adjusting the first image and the second image to the same size; subtracting the RGB value of each pixel of the first image The RGB value of the pixel at the corresponding position of the second image includes subtracting the RGB value of the pixel with the coordinate (x, y) in the second image from the RGB value of the pixel whose coordinate is (x, y) in the first image, respectively, obtained by This can get three difference values (respectively corresponding to R, G, B); then calculate the average of these three differences as the pixel difference value represented by the pixel of coordinates (x, y); and then take all coordinates in the first image The pixel difference values are summed up to be the difference degree between the first image and the second image.

在另一實施例中，類似於上述的實施例，而差別在於最後一步驟加總所有座標的像素差異值時，是進行加權加總。權重包括相對較大的第一權重和相對較小的第二權重。第一權重所對應的像素符合下列條件：在第一影像中此像素的RGB值屬於特定範圍之間，且此像素的座標屬於特定區域內。換言之，第一權重所代表的像素例如具有人臉特徵所具有的顏色，例如皮膚的膚色，嘴唇的唇色，頭髮的髮色等。所述特定區域則是對應到第一影像中預期將拍攝到人臉的位置，包括五官的位置，頭髮分布的位置等。透過限制像素在符合特定區域的條件以及符合特定顏色的條件，可進一步提升人臉差異度的評估準確度。In another embodiment, similar to the above embodiment, the difference is that the last step is to perform weighted summation when summing up the pixel difference values of all coordinates. The weights include relatively large first weights and relatively small second weights. The pixel corresponding to the first weight meets the following conditions: in the first image, the RGB value of the pixel belongs to a specific range, and the coordinates of the pixel belong to a specific area. In other words, the pixels represented by the first weight have, for example, the color of the facial features, such as the skin color of the skin, the lip color of the lips, and the hair color of the hair. The specific area is corresponding to the position in the first image where the human face is expected to be photographed, including the position of facial features, the position of hair distribution and so on. By restricting the pixels to meet the conditions of a specific area and the conditions of a specific color, the accuracy of the evaluation of the face difference can be further improved.

此外，值得注意的是，當使用者開始操作輸入介面1時，步驟S2及步驟S4可同時開始進行，或是由步驟S2、S4其中一者先進行。In addition, it is worth noting that when the user starts to operate the input interface 1, step S2 and step S4 can be performed simultaneously, or one of steps S2 and S4 can be performed first.

請參考步驟S6，處理器7依據第一影像及第二影像計算差異度，並判斷差異度是否大於門檻值。在若步驟S6的判斷結果為是，代表第一影像及第二影像拍攝到不同使用者，此狀況代表具有詐騙車手的可能性，因此移至步驟S7，處理器7觸發第一判斷事件及一第二判斷事件，其中第一判斷事件係判斷卡片資訊是否對應於單一張卡片，第二判斷事件係判斷輸入內容是否對應於提款操作或轉帳操作。如前文所述，卡片資訊例如包括：通過交易伺服器40驗證後的卡片號碼。因此，處理器可判斷通過交易伺服器40驗證後的卡片號碼是否具有多個，如果判斷為是，代表使用者在自動提款機進行操作時使用了多張卡片，第一判斷事件的結果為假。反之，如果處理器取得的卡片資訊中只有單一個卡片號碼，則第一判斷事件的結果為真。另一方面，處理器7可依據步驟S1取得的輸入內容中的操作項目進行第二判斷事件的判斷。若操作項目不是提款或轉帳操作，則第二判斷事件的結果為假。若操作項目是提款操作或轉帳操作，則第二判斷事件的結果為真。Please refer to step S6, the processor 7 calculates the degree of difference according to the first image and the second image, and determines whether the degree of difference is greater than a threshold value. If the judgment result in step S6 is yes, it means that the first image and the second image are captured by different users, and this situation represents the possibility of defrauding the driver. Therefore, it moves to step S7, and the processor 7 triggers the first judgment event and a The second judging event, wherein the first judging event is judging whether the card information corresponds to a single card, and the second judging event is judging whether the input content corresponds to a cash withdrawal operation or a transfer operation. As mentioned above, the card information includes, for example: the card number verified by the transaction server 40 . Therefore, the processor can judge whether there are multiple card numbers verified by the transaction server 40. If it is judged to be yes, it means that the user has used multiple cards when operating the automatic teller machine. The result of the first judgment event is Fake. On the contrary, if there is only a single card number in the card information obtained by the processor, the result of the first judgment event is true. On the other hand, the processor 7 can judge the second judging event according to the operation item in the input content obtained in step S1. If the operation item is not a withdrawal or transfer operation, the result of the second judging event is false. If the operation item is a withdrawal operation or an account transfer operation, the result of the second judging event is true.

請參考步驟S8，若第一及第二判斷事件皆為真（單一張卡片，且為提款或轉帳操作），代表此時自動櫃員機20前出現不同的人對同一張卡片進行提款或轉帳等操作，因此執行步驟S9，處理器7控制通訊裝置3發送警報訊號至外部伺服器60，藉此通報警察局或是金融機構有詐騙車手正在進行盜領或非法轉帳等交易行為。Please refer to step S8, if the first and second judgment events are both true (single card, and it is a cash withdrawal or transfer operation), it means that different people appear in front of the ATM 20 at this time to withdraw money or transfer money to the same card Wait for the operation, so step S9 is executed, the processor 7 controls the communication device 3 to send an alarm signal to the external server 60, thereby notifying the police station or financial institutions that fraudulent drivers are conducting transactions such as stealing or illegal transfers.

另一方面，若第一判斷事件及第二判斷事件中有一者不成立，則流程回到步驟S5，繼續本發明一實施例的詐騙車手警報方法的偵測流程。On the other hand, if one of the first judging event and the second judging event is not established, the process returns to step S5 to continue the detection process of the fraudulent driver alert method according to an embodiment of the present invention.

上述內容及圖2已說明當不同人操作自動櫃員機20時，如何判斷這些人是否為詐騙車手。以下內容及圖3~圖5將說明書當同一人操作自動櫃員機20時，如何判斷此人為詐騙車手。The above content and FIG. 2 have explained how to judge whether these people are fraudulent drivers when different people operate the automatic teller machine 20 . The following content and Figures 3 to 5 will describe how to determine that the person is a fraudulent driver when the same person operates the automatic teller machine 20 .

請參考圖3的步驟S6，當差異度小於或等於門檻值時，代表此時操作自動櫃員機的使用者為同一人，因此執行步驟S31，處理器累加計數值，藉此評估此人在自動櫃員機前操作的時間長度。Please refer to step S6 in Figure 3, when the degree of difference is less than or equal to the threshold value, it means that the user operating the ATM is the same person at this time, so step S31 is executed, and the processor accumulates the count value, so as to evaluate whether this person is in the ATM The length of time before the operation.

根據統計，一般民眾使用自動櫃員機20通常只會執行小於指定次數的金融操作，因此，如步驟S32所示，當處理器7累加的計數值大於設定值時，代表有詐騙車手盜領或非法轉帳的可能，故繼續執行步驟S33，處理器7觸發第二判斷事件及第三判斷事件，其中第三判斷事件係判斷卡片資訊中的行為範本與輸入內容的輸入行為是否相異。According to statistics, ordinary people usually only perform financial operations less than the specified number of times when using the automatic teller machine 20. Therefore, as shown in step S32, when the count value accumulated by the processor 7 is greater than the set value, it means that there is fraudulent driver stealing or illegal transfer Therefore, continue to execute step S33, the processor 7 triggers the second judgment event and the third judgment event, wherein the third judgment event is to judge whether the behavior template in the card information is different from the input behavior of the input content.

在一實施例中，行為範本記錄兩個時間數值之間的時間差為X秒，因此若輸入行為中兩個時間數值之間的時間差為Y秒，且X不等於Y，即視為行為範本與輸入行為相異，且處理器7將第三判斷事件設定為真。在一實施例中，預先收集多筆正常交易操作的案例，並由處理器計算這些交易操作對應的輸入內容中連續兩筆時間數值之間的時間差，然後將此時間差記錄在時間範本中。In one embodiment, the time difference between the two time values recorded in the behavior template is X seconds, so if the time difference between the two time values in the input behavior is Y seconds, and X is not equal to Y, it is regarded as the behavior template and The input behavior is different, and the processor 7 sets the third judgment event as true. In one embodiment, multiple cases of normal transaction operations are collected in advance, and the processor calculates the time difference between two consecutive time values in the input content corresponding to these transaction operations, and then records the time difference in the time template.

在一實施例中，輸入行為對應於多個輸入值，這些輸入值例如對應於密碼數值，且輸入內容記錄這些密碼數值之間的時間差。對於金融卡片的帳戶本人而言，其輸入密碼的節奏或是執行金融操作時的點按順序具有一定的規律及高度的相似性可循，而該規律及相似性則可從帳戶本人在初期操作自動櫃員機20的若干次操作紀錄中學習而得，並將學習到的行為範本記錄到交易伺服器40中。舉例來說，處理器7收集帳戶本人前N次的密碼輸入記錄，並記錄每個正確的密碼數值被輸入的時間，從而計算每兩個正確的密碼數值被輸入之間的時間差。例如，密碼數值1和密碼數值2之間的時間差包括0.13秒、0.15秒、…以及0.16秒。處理器7記錄N次的時間差中的最大值和最小值並儲存在行為範本之中。除此之外，行為範本也可記錄用於比對上述時間差的一指定閾值。該指定閾值可以根據過去使用者輸入密碼時經常出現的時間差長度決定之，本發明對此不予限制。In an embodiment, the input behavior corresponds to a plurality of input values, such as corresponding to password values, and the input content records the time difference between these password values. For the account owner of the financial card, the rhythm of entering the password or the order of clicking when performing financial operations has certain rules and a high degree of similarity to follow. The automatic teller machine 20 is learned from several operation records, and the learned behavior template is recorded in the transaction server 40 . For example, the processor 7 collects the previous N password input records of the account itself, and records the time when each correct password value is entered, so as to calculate the time difference between every two correct password values being entered. For example, the time difference between password value 1 and password value 2 includes 0.13 seconds, 0.15 seconds, . . . and 0.16 seconds. The processor 7 records the maximum value and the minimum value among the time differences of N times and stores them in the behavior template. In addition, the behavior template can also record a specified threshold for comparing the above-mentioned time difference. The specified threshold may be determined according to the length of the time difference that often occurred when the user entered the password in the past, which is not limited by the present invention.

當本次使用者操作自動櫃員機20時，若其輸入密碼的節奏或是點按觸控式螢幕的位置及順序與平時行為範本中儲存的操作模式不同，則代表本次使用者為詐騙車手的機率極高。舉例來說，在行為範本中儲存的密碼數值1和密碼數值2的輸入時間差的範圍是0.13~0.16秒，密碼數值2和密碼數值3的輸入時間差的範圍是0.15~0.17秒。若本次使用者操作自動櫃員機20時，輸入密碼數值1和密碼數值2的時間差位在0.13~0.16秒的範圍之外，而且超過行為範本中記錄的指定閾值（如0.05秒），而且輸入密碼數值2和密碼數值3的時間差也位在0.15~0.17秒的範圍之外，而且超過0.05秒的指定閾值；換言之，當出現多次數入密碼數值的時間差沒有落在行為範本中記錄的範圍之中時，則處理器7將第二判斷事件設定為真。故在步驟S34中，當處理器7判斷第二判斷事件（輸入內容對應於提款操作或轉帳操作）及第三判斷事件（使用者行為模式與平常不符）皆為真時，則移至步驟S9，處理器7控制通訊裝置3發送7警報訊號至外部伺服器60。上面的範例適用於使用實體鍵盤輸入密碼，若是使用觸控式螢幕輸入密碼，則行為範本除了記錄如同上面的範例所舉例的時間差的範圍之外，更記錄了每個密碼數值被輸入時，觸控式螢幕被觸碰到的座標範圍。舉例來說，在行為範本中，記錄第一個密碼數值的座標範圍為：水平10~20像素，垂直5~15像素；第二個密碼數值的座標範圍為：水平50~60像素，垂直5~15像素；第三個密碼數值的座標範圍為：水平30~40像素，垂直5~15像素。因此，當使用者操作自動櫃員機20時，可依據行為範本中記錄的密碼數值輸入的時間差，以及密碼數值透過觸控式螢幕輸入的座標的範圍，處理器可以判斷行為範本中記錄的這些範圍與處理器7從使用者本次輸入行為中取得的多個範圍，逐一比較對應的範圍之間是否相符，從而產生第三判斷事件為真或是為假的結果。舉例來說，若是密碼數值輸入的時間差與行為範本記錄的不符的數量超過兩個，而且密碼數值輸入的座標範圍與行為記錄的不符的數量超過三個，則第三判斷事件為真，代表本次操作極有可能是詐騙車手所為。When the user operates the automatic teller machine 20 this time, if the rhythm of entering the password or the position and order of clicking the touch screen is different from the operation mode stored in the usual behavior template, it means that the user is defrauding the driver. The probability is extremely high. For example, the range of input time difference between password value 1 and password value 2 stored in the behavior template is 0.13~0.16 seconds, and the range of input time difference between password value 2 and password value 3 is 0.15~0.17 seconds. If the user operates the automatic teller machine 20 this time, the time difference between entering the password value 1 and the password value 2 is outside the range of 0.13~0.16 seconds, and exceeds the specified threshold (such as 0.05 seconds) recorded in the behavior template, and the password is entered The time difference between the value 2 and the password value 3 is also outside the range of 0.15~0.17 seconds, and exceeds the specified threshold of 0.05 seconds; in other words, when there are multiple times to enter the password value, the time difference does not fall within the range recorded in the behavior template , the processor 7 sets the second judgment event as true. Therefore, in step S34, when the processor 7 judges that the second judging event (the input content corresponds to the withdrawal operation or the transfer operation) and the third judging event (the user's behavior pattern is not consistent with the usual) are both true, then move to step S34 S9, the processor 7 controls the communication device 3 to send 7 an alarm signal to the external server 60 . The above example is applicable to inputting passwords using a physical keyboard. If using a touch screen to input passwords, the behavior template not only records the range of time difference as exemplified in the above example, but also records when each password value is entered. The coordinate range that the control screen is touched. For example, in the behavior template, the coordinate range for recording the first password value is: horizontal 10~20 pixels, vertical 5~15 pixels; the coordinate range for the second password value is: horizontal 50~60 pixels, vertical 5 pixels ~15 pixels; the coordinate range of the third password value is: horizontal 30~40 pixels, vertical 5~15 pixels. Therefore, when the user operates the automatic teller machine 20, according to the time difference of inputting the password value recorded in the behavior template, and the range of the coordinate input of the password value through the touch screen, the processor can determine whether these ranges recorded in the behavior template are consistent with The processor 7 compares the multiple ranges obtained from the user's current input behavior one by one to see if the corresponding ranges are consistent, so as to generate a third result of judging whether the event is true or false. For example, if there are more than two discrepancies between the time difference of the password value input and the behavior record, and the number of discrepancies between the coordinate range of the password value input and the behavior record is more than three, then the third judgment event is true, representing this This operation is very likely to be done by defrauding the driver.

另一方面，若第二判斷事件及第三判斷事件中有一者不成立，則回到步驟S5，繼續本發明一實施例的詐騙車手警報方法的偵測流程。On the other hand, if one of the second judging event and the third judging event is not established, go back to step S5 and continue the detection process of the method for alerting a fraudulent driver according to an embodiment of the present invention.

在本實施例的變化例中，第三判斷事件的警報閾值可隨著嘗試次數提升而進行調整。換言之，在步驟S34中，若第二判斷事件通過（為真），但第三判斷事件判斷不通過時，用於決定第三判斷事件是否通過的閾值將被動態地調整，藉此可避免初始閾值設定過高造成誤判斷使用者為車手之頻率過高的問題，又可應對車手之操作時間較長之情況下，未能正確判斷出車手正在使用而發送警報訊號之問題。In a variation of this embodiment, the alarm threshold of the third judging event may be adjusted as the number of attempts increases. In other words, in step S34, if the second judging event is passed (true), but the third judging event is not passed, the threshold used to determine whether the third judging event is passed will be dynamically adjusted, thereby avoiding the initial If the threshold is set too high, the frequency of misjudging the user as a driver is too high, and it can also deal with the problem that the driver cannot correctly determine that the driver is using it and send an alarm signal when the driver's operation time is long.

舉例來說，在執行步驟S34時，更包括：累計步驟S34被執行的次數、依據步驟S34被執行的次數調整對應於第三判斷事件中所用的閾值。舉例來說，假設第一次執行步驟S34，處理器計算第三判斷事件中使用者行為範本與輸入行為的相似度，假設對應此第三判斷事件的預設閾值為75%，當相似度若超過75%時，則判定為合法使用者。若未超過75%相似且第二判斷事件為真時，因為並非第二判斷事件及第三判斷事件皆為真，按圖2及圖3所示的流程，將執行步驟S5…S6…S31~S34的流程。當再次執行步驟S34時，假設步驟S34目前已被執行第n次，則處理器將第三判斷事件採用的操作模式相似度閾值調整為為預設操作模式度閾值*(1+0.05*n)。換言之，本發明藉由累計「第二判斷事件為真且第三判斷事件為假」的次數，藉此提高下一次第三判斷事件的判斷用的閾值，藉此可避免初始閾值設定過高造成誤判斷使用者為車手之頻率過高的問題，又可應對車手之操作時間較長之情況下，未能正確判斷出車手正在使用而發送警報訊號之問題。需特別注意的是，除了第二判斷事件及第三判斷事件可採用上述的方式調整第三判斷事件的閾值，在圖4及5中的步驟S44及步驟S54亦可採用上述方式，在一個判斷事件為真，另一個判斷事件為假的條件下，累計步驟S44及步驟S54的執行次數，並且動態地調整所述另一個判斷事件的閾值，本發明亦不限制將閾值向上或向下調整。For example, when step S34 is executed, it further includes: accumulating the number of times step S34 is executed, and adjusting the threshold corresponding to the third judging event according to the number of times step S34 is executed. For example, assuming that step S34 is executed for the first time, the processor calculates the similarity between the user behavior template and the input behavior in the third judging event, assuming that the preset threshold corresponding to the third judging event is 75%, when the similarity is When it exceeds 75%, it is judged as a legitimate user. If there is no more than 75% similarity and the second judging event is true, because not both the second judging event and the third judging event are true, according to the flow shown in Figure 2 and Figure 3, steps S5...S6...S31~ will be executed The process of S34. When step S34 is executed again, assuming that step S34 has been executed for the nth time, the processor adjusts the operating mode similarity threshold used by the third judging event to be the preset operating mode threshold*(1+0.05*n) . In other words, the present invention increases the threshold for judging the next third judging event by accumulating the number of times "the second judging event is true and the third judging event is false", thereby avoiding the occurrence of excessively high initial threshold settings. The frequency of misjudging that the user is a driver is too high, and it can also deal with the problem that the driver is using it for a long time, and the problem of sending an alarm signal cannot be correctly judged that the driver is using it. It should be noted that, in addition to the second judgment event and the third judgment event, the threshold value of the third judgment event can be adjusted in the above-mentioned manner, and the above-mentioned method can also be used in steps S44 and S54 in FIGS. 4 and 5 . Under the condition that the event is true and the other judgment event is false, the execution times of step S44 and step S54 are accumulated, and the threshold of the other judgment event is dynamically adjusted, and the present invention does not limit the adjustment of the threshold upward or downward.

在步驟S34的另一實施例中，除了根據第二及第三判斷事件，可更包含處理器7判斷第一判斷事件（卡片資訊對應於單一張卡片）是否不成立（代表卡片資訊對應於多張卡片）。換言之，當判斷為同一人並操作多張卡片時（例如從多張卡片中領取現金），本發明將判定此金融操作屬於詐騙車手的行為模式，然後移至步驟S9發送警報訊號。另一方面，若第一判斷事件為真，且第二及第三判斷事件中有一者不成立，則回到步驟S5繼續本發明一實施例的詐騙車手警報方法的偵測流程。由於本實施例是透過比對金融操作的行為模式來識別是否有詐騙車手出現，因此僅需上傳行為模式資訊，而不用上傳所拍攝的整個影片檔，故可大幅降低傳輸上和硬體上的成本。此外，由於所上傳的資料遠小於先前技術，故可在短時間內上傳完畢，加快警方辦案的效率。In another embodiment of step S34, in addition to the second and third judging events, it may further include the processor 7 judging whether the first judging event (the card information corresponds to a single card) is not established (representing that the card information corresponds to multiple card). In other words, when it is judged that the same person operates multiple cards (such as receiving cash from multiple cards), the present invention will determine that this financial operation is a behavior pattern of defrauding the driver, and then move to step S9 to send an alarm signal. On the other hand, if the first judging event is true, and one of the second and third judging events is not true, return to step S5 to continue the detection process of the fraudulent driver alert method according to an embodiment of the present invention. Since this embodiment identifies whether there are fraudulent drivers by comparing the behavior patterns of financial operations, it is only necessary to upload the behavior pattern information instead of uploading the entire video file, so the transmission and hardware costs can be greatly reduced. cost. In addition, since the uploaded data is much smaller than the previous technology, it can be uploaded in a short time, which speeds up the efficiency of police handling cases.

請參考圖4，其中步驟S6、S41至S42的流程與圖3的步驟S6、S31至S32的流程相同，在此不重複敘述。請參考步驟S43，處理器7觸發第一判斷事件及第四判斷事件，其中第四判斷事件係判斷輸入內容對應於非指定帳戶之轉帳操作。請參考步驟S44，當第一判斷事件及第四判斷事件皆為真時，代表此狀況為同一人對同一張卡片進行轉帳，且轉入的目標帳戶為可疑帳戶，此狀況為詐騙車手的機率極高，因此移至步驟S9，處理器7控制通訊裝置3發送警報訊號至外部伺服器60。Please refer to FIG. 4 , wherein the flow of steps S6 , S41 to S42 is the same as the flow of steps S6 , S31 to S32 in FIG. 3 , and will not be repeated here. Please refer to step S43, the processor 7 triggers a first judgment event and a fourth judgment event, wherein the fourth judgment event is to judge that the input content corresponds to a transfer operation of an unspecified account. Please refer to step S44. When both the first judgment event and the fourth judgment event are true, it means that the same person transfers money to the same card, and the target account of the transfer is a suspicious account. This situation is the probability of defrauding the driver Very high, so move to step S9, the processor 7 controls the communication device 3 to send an alarm signal to the external server 60.

另一方面，若第一判斷事件及第四判斷事件中有一者不成立，則回到步驟S5，繼續本發明一實施例的詐騙車手警報方法的偵測流程。On the other hand, if one of the first judging event and the fourth judging event is not established, go back to step S5 and continue the detection process of the method for alerting a fraudulent driver according to an embodiment of the present invention.

由於本實施例是透過帳戶類型以及受操作的卡片識別碼來進行判斷是否有詐騙車手出現，故不用上傳所拍攝的整個影片檔，故可大幅降低傳輸上和硬體上的成本。此外，由於所上傳的資料遠小於先前技術，故可在短時間內上傳完畢，加快警方辦案的效率。Since this embodiment judges whether there is a fraudulent driver through the account type and the operated card identification code, there is no need to upload the entire video file taken, so the cost of transmission and hardware can be greatly reduced. In addition, since the uploaded data is much smaller than the previous technology, it can be uploaded in a short time, which speeds up the efficiency of police handling cases.

請參考圖5，其中步驟S6、S51至S52的流程與圖3的步驟S6、S31至S32的流程相同，在此不重複敘述。請參考步驟S53，處理器7觸發第一判斷事件及一第五判斷事件，其中該第五判斷事件係卡片資訊中的前次提領地點與本次提領地點相異。請參考步驟S54，當第一判斷事件及第五判斷事件皆為真時，代表此狀況為同一人或是同張卡在不同的地點遭到提領或是轉帳，此狀況為詐騙車手的機率極高，因此移至步驟S9，處理器7控制通訊裝置3發送警報訊號至外部伺服器60。Please refer to FIG. 5 , where the flow of steps S6 , S51 to S52 is the same as the flow of steps S6 , S31 to S32 in FIG. 3 , and will not be repeated here. Please refer to step S53, the processor 7 triggers a first judging event and a fifth judging event, wherein the fifth judging event is that the previous withdrawal location in the card information is different from the current withdrawal location. Please refer to step S54. When both the first judgment event and the fifth judgment event are true, it means that the situation is that the same person or the same card has been withdrawn or transferred at different places. This situation is the probability of defrauding the driver Very high, so move to step S9, the processor 7 controls the communication device 3 to send an alarm signal to the external server 60.

另一方面，若第一判斷事件及第五判斷事件中有一者不成立，則回到步驟S5，繼續本發明一實施例的詐騙車手警報方法的偵測流程。On the other hand, if one of the first judging event and the fifth judging event is not established, return to step S5 to continue the detection process of the fraudulent driver alert method according to an embodiment of the present invention.

由於本實施例是透過使用者資訊、提款機識別碼及受操作的卡片識別碼來進行判斷是否有詐騙車手出現，故不用上傳所拍攝的整個影片檔，故可大幅降低傳輸上和硬體上的成本。此外，由於所上傳的資料遠小於先前技術，故可在短時間內上傳完畢，加快警方辦案的效率。此外，由於所上傳的資料遠小於先前技術，故可在短時間內上傳完畢，加快警方辦案的效率。Since this embodiment judges whether there is a fraudulent driver through the user information, cash machine identification code and the operated card identification code, there is no need to upload the entire video file taken, so the transmission and hardware can be greatly reduced. on the cost. In addition, since the uploaded data is much smaller than the previous technology, it can be uploaded in a short time, which speeds up the efficiency of police handling cases. In addition, since the uploaded data is much smaller than the previous technology, it can be uploaded in a short time, which speeds up the efficiency of police handling cases.

雖然本發明以前述之實施例揭露如上，然其並非用以限定本發明。在不脫離本發明之精神和範圍內，所為之更動與潤飾，均屬本發明之專利保護範圍。關於本發明所界定之保護範圍請參考所附之申請專利範圍。Although the present invention is disclosed by the aforementioned embodiments, they are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, all changes and modifications are within the scope of patent protection of the present invention. For the scope of protection defined by the present invention, please refer to the appended scope of patent application.

1:輸入介面 3:通訊裝置 5:攝像裝置 7:處理器 20:自動櫃員機 40:交易伺服器 60:外部伺服器 S1~S9、S31~S34、S41~S44、S51~S54:步驟 1: Input interface 3: Communication device 5: camera device 7: Processor 20: Automatic teller machine 40: Transaction Server 60:External server S1~S9, S31~S34, S41~S44, S51~S54: steps

圖1是本發明一實施例的自動櫃員機的方塊架構圖；圖2~5是本發明不同實施例的詐騙車手警報方法的流程圖。 FIG. 1 is a block diagram of an automatic teller machine according to an embodiment of the present invention; 2 to 5 are flow charts of methods for alerting fraudulent drivers in different embodiments of the present invention.

S1~S9:步驟 S1~S9: steps

Claims

一種詐騙車手警報方法，適用於一自動櫃員機，該自動櫃員機具有一輸入介面、一通訊裝置、一攝像裝置及一處理器，該詐騙車手警報方法包含以處理器進行以下步驟：透過該輸入介面取得一輸入內容；控制該通訊裝置發送該輸入內容至一交易伺服器；透過該通訊裝置從該交易伺服器接收一卡片資訊；控制該攝像裝置於一起始期間拍攝一第一影像，並於該起始期間後控制該攝像裝置隔一指定時間拍攝一第二影像，其中該起始期間開始於該輸入介面***作之後；依據該第一影像及該第二影像計算一差異度；當該差異度大於一門檻值時，觸發一第一判斷事件及一第二判斷事件，其中該第一判斷事件係判斷該卡片資訊是否對應於單一張卡片，該第二判斷事件係判斷該輸入內容是否對應於提款操作或轉帳操作；以及當該第一判斷事件及該第二判斷事件皆為真時，控制該通訊裝置發送一警報訊號至一外部伺服器。 A method for alerting a fraudulent driver is applicable to an automatic teller machine, the automatic teller machine has an input interface, a communication device, a camera device and a processor, the method for alerting a fraudulent driver includes performing the following steps with the processor: Obtain an input content through the input interface; controlling the communication device to send the input content to a transaction server; receiving a card information from the transaction server through the communication device; controlling the imaging device to capture a first image during an initial period, and controlling the imaging device to capture a second image at a specified time interval after the initial period, wherein the initial period starts after the input interface is operated; calculating a difference degree according to the first image and the second image; When the difference is greater than a threshold, a first judgment event and a second judgment event are triggered, wherein the first judgment event is to judge whether the card information corresponds to a single card, and the second judgment event is to judge the input whether the content corresponds to a withdrawal operation or a transfer operation; and When both the first judging event and the second judging event are true, the communication device is controlled to send an alarm signal to an external server.

如請求項1所述詐騙車手警報方法，另包含以該處理器進行以下步驟：在該差異度小於或等於該門檻值時，累加一計數值；當該計數值大於一設定值時，觸發該第二判斷事件及一第三判斷事件，其中該第三判斷事件係判斷該卡片資訊中的一行為範本與該輸入內容的一輸入行為是否相異，該輸入行為對應於多個輸入值，且該輸入內容記錄該些輸入值之間的時間差；以及當該第二判斷事件及該第三判斷事件皆為真時，控制該通訊裝置發送該警報訊號至該外部伺服器。 As described in claim item 1, the fraud driver alert method further includes the following steps with the processor: When the difference degree is less than or equal to the threshold value, add a count value; When the count value is greater than a set value, the second judging event and a third judging event are triggered, wherein the third judging event is to judge whether a behavior template in the card information is different from an input behavior of the input content , the input behavior corresponds to multiple input values, and the input content records the time difference between the input values; and When both the second judging event and the third judging event are true, the communication device is controlled to send the alarm signal to the external server.

如請求項1所述詐騙車手警報方法，另包含以該處理器進行以下步驟：在該差異度小於或等於該門檻值時，累加一計數值；當該計數值大於一設定值時，觸發該第一判斷事件及一第四判斷事件，其中該第四判斷事件係判斷該輸入內容對應於非指定帳戶之轉帳操作；以及當該第一判斷事件及該第四判斷事件皆為真時，控制該通訊裝置發送該警報訊號至該外部伺服器。 As described in claim item 1, the fraud driver alert method further includes the following steps with the processor: When the difference degree is less than or equal to the threshold value, add a count value; When the count value is greater than a set value, the first judging event and a fourth judging event are triggered, wherein the fourth judging event is judging that the input content corresponds to a transfer operation of an unspecified account; and When both the first judging event and the fourth judging event are true, the communication device is controlled to send the alarm signal to the external server.

如請求項1所述詐騙車手警報方法，另包含以該處理器進行以下步驟：在該差異度小於或等於該門檻值時，累加一計數值；當該計數值大於一設定值時，判斷該第一判斷事件及一第五判斷事件，其中該第五判斷事件係該卡片資訊中的前次提領地點與本次提領地點相異；以及當該第一判斷事件及該第五判斷事件皆為真時，控制該通訊裝置發送該警報訊號至該外部伺服器。 As described in claim item 1, the fraud driver alert method further includes the following steps with the processor: When the difference degree is less than or equal to the threshold value, add a count value; When the count value is greater than a set value, judge the first judging event and a fifth judging event, wherein the fifth judging event is that the previous withdrawal location in the card information is different from the current withdrawal location; and When both the first judging event and the fifth judging event are true, the communication device is controlled to send the alarm signal to the external server.

一種自動櫃員機，包括：一輸入介面，用以取得一輸入內容；一通訊裝置，用以與一交易伺服器交換資訊；一攝像裝置，用以拍攝影像；以及一處理器，電性連接該輸入介面、該通訊裝置及該攝像裝置，其中：該處理器用以控制該通訊裝置發送該輸入內容至該交易伺服器，透過該通訊裝置從該交易伺服器接收一卡片資訊；該處理器控制該攝像裝置於一起始期間拍攝一第一影像，並於該起始期間後控制該攝像裝置隔一指定時間拍攝一第二影像，其中該起始期間開始於該輸入介面***作之後；以及該處理器依據該第一影像及該第二影像計算一差異度，當該差異度大於一門檻值時，觸發一第一判斷事件及一第二判斷事件，其中該第一判斷事件係判斷該卡片資訊是否對應於單一張卡片，該第二判斷事件係判斷該輸入內容是否對應於提款操作或轉帳操作；以及當該第一判斷事件及該第二判斷事件皆為真時，控制該通訊裝置發送一警報訊號至一外部伺服器。 An automatic teller machine, comprising: an input interface for obtaining an input content; a communication device for exchanging information with a transaction server; a camera device for capturing images; and a processor electrically connected to the input Interface, the communication device and the camera device, wherein: the processor is used to control the communication device to send the input content to the transaction server, and receive a card information from the transaction server through the communication device; the processor controls the camera The device captures a first image during an initial period, and controls the camera device to capture a second image at a specified time interval after the initial period, wherein the initial period starts after the input interface is operated; and the processor Calculate a degree of difference based on the first image and the second image , and when the degree of difference is greater than a threshold value, trigger a first judgment event and a second judgment event, wherein the first judgment event is to judge whether the card information Corresponding to a single card, the second judging event is to judge whether the input content corresponds to a cash withdrawal operation or a transfer operation; and when both the first judging event and the second judging event are true, the communication device is controlled to send a Alarm signal to an external server.

如請求項5所述自動櫃員機，其中：在該差異度小於或等於該門檻值時，該處理器累加一計數值；當該計數值大於一設定值時，該處理器判斷該第二判斷事件及一第三判斷事件，其中該第三判斷事件係比對該卡片資訊中的一行為範本與該輸入內容的一輸入行為是否相異，該輸入行為對應於多個輸入值，且輸入內容記錄該些輸入值之間的時間差；以及當該第二判斷事件及該第三判斷事件皆為真時，該處理器控制該通訊裝置發送該警報訊號至該外部伺服器。 An automatic teller machine as claimed in claim 5, wherein: When the difference degree is less than or equal to the threshold value, the processor accumulates a count value; When the count value is greater than a set value, the processor judges the second judging event and a third judging event, wherein the third judging event is compared with an input of a behavior template in the card information and the input content Whether the behavior is distinct, the input behavior corresponds to multiple input values, and the input content records the time difference between those input values; and When both the second judging event and the third judging event are true, the processor controls the communication device to send the alarm signal to the external server.

如請求項5所述自動櫃員機，其中：在該差異度小於或等於該門檻值時，該處理器累加一計數值；當該計數值大於一設定值時，該處理器判斷該第一判斷事件及一第四判斷事件，其中該第四判斷事件係判斷該輸入內容對應於非指定帳戶之轉帳操作；以及當該第一判斷事件及該第四判斷事件皆為真時，該處理器控制該通訊裝置發送該警報訊號至該外部伺服器。 An automatic teller machine as claimed in claim 5, wherein: When the difference degree is less than or equal to the threshold value, the processor accumulates a count value; When the count value is greater than a set value, the processor judges the first judging event and a fourth judging event, wherein the fourth judging event is judging that the input content corresponds to a transfer operation of an unspecified account; and When both the first judging event and the fourth judging event are true, the processor controls the communication device to send the alarm signal to the external server.

如請求項5所述自動櫃員機，其中：在該差異度小於或等於該門檻值時，該處理器累加一計數值；當該計數值大於一設定值時，該處理器判斷該第一判斷事件及一第五判斷事件，其中該第五判斷事件係該卡片資訊中的前次提領地點與本次提領地點相異；以及當該第一判斷事件及該第五判斷事件皆為真時，該處理器控制該通訊裝置發送該警報訊號至該外部伺服器。 An automatic teller machine as claimed in claim 5, wherein: When the difference degree is less than or equal to the threshold value, the processor accumulates a count value; When the count value is greater than a set value, the processor judges the first judging event and a fifth judging event, wherein the fifth judging event is that the previous withdrawal location in the card information is the same as the current withdrawal location different; and When both the first judging event and the fifth judging event are true, the processor controls the communication device to send the alarm signal to the external server.