TWI661398B - Transaction system based on face recognitioin for verification and method thereof - Google Patents

Abstract

A trading system based on face recognition for verification, including an automated teller machine and a verification server. The automatic teller machine includes: a camera module, which obtains the image to be verified of the user's face, a card reading module for inserting the financial card, an input module for inputting the card password and the transaction content, and the control module prompting the user to input the transaction content and indicating The communication module sends the image to be verified and the transaction content. The verification server, the communication module connected to the automatic teller machine, includes: an account database, having a plurality of account information, each account information includes at least a first reference image, and an operation module for comparing the verification image and the first A reference image generates a continuable signal based on the comparison result. The trading module of the ATM performs trading operations based on the renewable signal and the transaction content.

Description

Transaction system and method based on face recognition verification

The present invention relates to a transaction system and method thereof, and more particularly to a transaction system and method for verifying based on facial features and expression recognition.

Nowadays, there are frequent cases of social fraud. The money laundering incident through the head account is also known. The former is the fraud group to defraud the victim's financial card and the withdrawal password after the victim has obtained the trust. The driver then carries the card to the ATM to withdraw the money. The latter is a fraudulent group that defrauds the general public's account number, financial card and account password, then tricks the victim into remitting the money into these head accounts, and then dispatches the driver to the ATM to withdraw the amount of the head account.

Observing the above-mentioned types of piracy cases, the similarity is that the person who operates the ATM to withdraw cash is not the account holder. Due to the withdrawal procedure of the traditional ATM, only the financial card and the withdrawal password are used as the basis for account security verification. Therefore, once the card is lost and the password is leaked, the risk of the stolen deposit will be greatly increased. In addition, even if the account person personally withdraws, it cannot be ruled out that it is forced to collect money from the ATM by being threatened by the culprits. However, the current ATM has not yet provided any mechanism to prevent the burglary incident.

In view of this, the present invention discloses a transaction system for authenticating by face recognition and a method thereof, thereby reducing the potential for the automatic teller machine to be taken by others or forced to withdraw involuntarily.

A transaction system for verifying based on face recognition according to an embodiment of the invention includes: an automatic teller machine and a verification server. The automatic teller machine comprises: a camera module, a card reading module, an input module, a control module, a communication module and a transaction module. The camera module, the card reading module, the input module, the communication module and the transaction module are all electrically connected to the control module. The camera module is configured to obtain an image to be verified including a user's face. The card reader module is configured to place a financial card, send a verification request, and receive a verification reply sent by the financial card. The input module is used to input the card password and transaction content. The control module is configured to receive the image to be verified, generate a verification request according to the card password, selectively generate a prompt message according to the verification reply, and prompt the user to input the transaction content, and the control module is further configured to be received by the input module. Transaction content. The communication module is configured to send the image to be verified and the transaction content according to the instruction of the control module. The transaction module is configured to perform a transaction operation, and the transaction operation outputs cash that matches the amount set in the transaction content. The verification server communication is connected to the communication module of the automatic teller machine, and the verification server includes an account database and an operation module. The account database includes a plurality of account information, and each account information includes at least a first reference image. The computing module is electrically connected to the account database, and the computing module is configured to compare the verified image and the at least one first reference image, and selectively generate the renewable signal according to the comparison result, and then send the renewable signal to the automatic Teller machine.

According to another embodiment of the present invention, a transaction system for verifying based on facial features, wherein each account information further includes a second reference image, and the verification server further includes an expression recognition module electrically connected to the account database. And an operation module, the expression recognition module is configured to analyze the expression of the face in the image to be verified, and the operation module further comprises: comparing the image to be verified and the at least one second reference image according to the analyzed expression.

A transaction method for verifying based on facial features according to an embodiment of the present invention includes: placing a financial card into a card reading module of an automatic teller machine; after the financial card is placed in the card reading module, the automatic teller machine The camera module obtains a to-be-verified image including the face of the user; the control module of the automatic teller machine instructs the communication module of the automatic teller machine to send the image to be verified to the verification server; after the financial card is placed in the card reading module, The input module of the automatic teller machine inputs a card password; the control module of the automatic teller machine generates a verification request according to the card password and sends a verification request to the financial card through the card reading module; the card reading module receives the verification reply of the financial card and sends a verification reply. To the control module; the control module selectively prompts the user to input the transaction content based on the verification reply; the control module instructs the communication module to send the transaction content to the verification server; and the verification module of the verification server is compared according to the image to be verified An account database, wherein the account database includes a plurality of account information and each account information includes at least one first parameter The computing module selectively generates a renewable signal according to the comparison result of the image to be verified; the verification server sends the renewable signal to the ATM; and the transaction module of the automatic teller machine is based on the renewable signal and the control module The indication is to perform a transaction operation.

According to another embodiment of the present invention, a transaction method for verifying based on facial features, after the verification server receives the image to be verified, further includes: an expression recognition module of the verification server analyzes a face in the image to be verified. An expression; the computing module compares the expression of the face with a plurality of second reference images in the account database; and the computing module performs the verification image and the plurality of first reference images and the plurality of second The comparison of the reference images selectively produces a renewable signal.

With the above structure, the transaction system and method for authenticating face recognition disclosed by the present invention further increases the face of the user who operates the automatic teller machine based on the original withdrawal of the financial card and the password. Identification as a mechanism for verification. In addition, expression recognition can be added as a further verification method, which can effectively reduce the risk of non-accounts to the ATM, and can immediately detect that the account itself is coerced rather than voluntarily In the case of early, avoid the loss of money in the account itself.

The above description of the disclosure and the following description of the embodiments of the present invention are intended to illustrate and explain the spirit and principles of the invention, and to provide further explanation of the scope of the invention.

The detailed features and advantages of the present invention are set forth in the Detailed Description of the Detailed Description of the <RTIgt; </ RTI> <RTIgt; </ RTI> </ RTI> </ RTI> <RTIgt; The objects and advantages associated with the present invention can be readily understood by those skilled in the art. The following examples are intended to describe the present invention in further detail, but are not intended to limit the scope of the invention.

Please refer to FIG. 1 , which is a structural diagram of a transaction system based on face recognition for verification according to an embodiment of the present invention. The system mainly includes an ATM 1 and a verification server 5 , and the ATM 1 is connected to the verification. Server 5, in practice, the "communication connection" refers to connection by wired or wireless connection; the wired communication connection method is, for example, a telephone line, a universal serial bus (USB), a twisted pair cable, a coaxial cable Or a fiber-optic cable or the like for communication connection; the wireless communication connection method is, for example, a Global System for Mobile communications (GSM), Worldwide Interoperability for Microwave Access (WiMax), code division multiple access (Code Division Multiple Access, CDMA), Near Field Communication (NFC), Bluetooth, or Wi-Fi. The present invention does not limit the type of communication connection.

Please refer to FIG. 1 , the automatic teller machine 1 includes a camera module 10 , a card reading module 11 , an input module 12 , a control module 13 , a communication module 14 , and a transaction module 15 , wherein the camera module 10 and the card reader module The group 11, the input module 12, the communication module 14, and the transaction module 15 are all electrically connected to the control module 13, as shown in FIG.

The camera module 10 is used to obtain a to-be-verified image. In practice, the camera module is, for example, a camera including a photosensitive-coupled device (CCD) or a complementary metal-oxide semiconductor (CMOS). Camera. However, the hardware device of the camera module 10 of the present invention is not limited thereto; in addition, the present invention does not limit the number of the camera modules 10. The image to be verified must include a facial image, that is, a user's facial image that operates the automated teller machine 1. In general, the camera module 10 should be disposed at the position of the automated teller machine 1 as viewed by the user. Moreover, considering that the height of the user is different, the camera module 10 of the fixed height may not be able to take the face of the user whose height is much higher or lower than the fixed height; therefore, it is practical to set more on the ATM 1 The camera module 10 at different heights or a wide-angle lens or a fisheye lens is added to the single camera module 10, thereby increasing the shooting range of the camera module 10.

The card reader module 11 is configured to operate a user of the automated teller machine 1 to place a financial card. After the user inserts the financial card, the card reading module 11 is further configured to send a verification request to the financial card or to receive the verification sent by the financial card. Reply. In practice, the card reading module 11 is, for example, a card reader or other hardware device that can read a wafer financial card or a magnetic stripe financial card.

The input module 12 is configured to input the card password and the transaction content by the user operating the ATM 1. In practice, the input module 12 is, for example, a physical keyboard or a virtual keyboard displayed through a screen. The card password refers to the withdrawal password of the financial card placed in the card reading module 11. The transaction content is, for example, a financial transaction item such as a deposit amount, a withdrawal amount, an amount of transfer, or an account for transfer.

The control module 13 is configured to receive the image to be verified captured by the camera module 10; generate a verification request according to the card password input by the user through the input module 12, and send the verification request to the financial system through the card reading module 11 The card receives the verification reply sent by the financial card from the card reading module 11 at the same time. If the verification reply indicates that the card password is correct, the control module 13 generates a prompt message for prompting the user to input the transaction content, and then the control module 13 receives the transaction content completed by the user input, and the control module 13 further instructs the communication. Module 14 sends the transaction content. If the verification reply indicates that the card password is incorrect, the control module 13 prompts the user to re-enter the password or instruct the communication module to send a warning message. In practice, the control unit 13 is, for example, a central processing unit (CPU), a microprocessor (Microprocessor), a digital signal processor (DSP), and a special application integrated circuit (Application- Specific Integrated Circuit (ASIC), other similar components, or a combination of the above components, the present invention does not limit the hardware type of the control unit 13.

The communication module 14 is, for example, a network communication chip for transmitting the image to be verified and the transaction content according to the instruction of the control module 13.

The transaction module 15 is configured to perform a transaction operation according to the instruction of the control module 13, for example, providing the same amount of cash to the user according to the withdrawal amount set in the transaction content. The control module 13 indicates the basis for verifying the renewable signal sent by the server 5 and the transaction content, and the continuation signal will be described later.

In another embodiment of the present invention, the automated teller machine 1 further includes a display module. The display module is electrically connected to the control module 13 and is used to instantly display the image to be verified captured by the camera module 10 for confirmation by the user. The captured image. In addition, the "control module 13 generates a prompt message" and the "control module 13 prompts the user to re-enter the password" as described above are also presented by the display module in text or image for viewing by the user. In practice, the display module is, for example, a Plasma Display Panel, a Liquid-Crystal Display (LCD), a Thin Film Transistor Liquid Crystal Display (TFT-LCD), and a light-emitting diode. Light-Emitting Diode Display (LED Display) Organic Light-Emitting Diode (OLED) or other display device with similar functions.

In an embodiment of the present invention, when the card reading module 11 detects that the card is placed, the control module 13 can be notified to enable the camera module 10 to perform shooting, and the camera module 14 will immediately shoot through the communication module 14. The image to be verified is sent to the verification server 5. In another embodiment of the present invention, after the card password is correctly input, the control module 13 enables the camera module 10 to perform shooting, and sends the transaction content together with the image to be verified after the user inputs the transaction content. To the verification server 5. However, in the present invention, the timing at which the control module 13 enables the camera module 10 to be photographed, and the timing at which the communication module 14 transmits the image to be verified is not limited to the above embodiment. In practice, the financial institution of the ATM 1 is set. According to the execution efficiency of the verification server 5 and the transmission efficiency of the communication connection, it is possible to decide when to shoot and transmit the image.

Please refer to Figure 1. The verification server 5 is communicatively coupled to the communication module 14 of the automated teller machine 1. In an embodiment of the invention, the verification server 5 includes an account database 50 and an operation module 51. In practice, the verification server 5 is, for example, a Pedestal server, a Rack server, or a Blade server. The present invention does not limit the hardware type of the verification server 5.

The account database 50 includes a plurality of account information, each account information representing a user's profile. Personal data, for example, is authorized by the user to provide financial institutions with the basis for personal identity verification when opening an account. Specifically, the account information includes at least one first reference image; the first reference image is, for example, a facial image that is taken by the user when the account is opened, or a photo provided by the user, such as an identity card or a driver's license. Face image.

The computing module 51 is electrically connected to the account database 50. The computing module 51 is configured to match whether the image to be verified matches the first reference image in the account database 50. Specifically, the computing module 51 is, for example, a microprocessor, a digital signal processor, a special application integrated circuit, other similar components, or a combination of the above. For the manner in which the computing module 51 compares, for example, the feature points of the image to be verified are first captured by image processing, and matched with the feature points of the first reference image stored in the account information to determine the similarity between the two images.

It must be specifically noted here that, in an embodiment, the image to be verified is compared with all first reference images in the account database. In another embodiment, the card reading module 11 further includes an account for obtaining the financial card in advance, and the control module 13 receives the account and instructs the communication module 14 to send the account to the verification image and the transaction content to be verified. Server 5, and each account information in the account database uses the account number as the primary key of the database, so the target account with the same card account can be selected by comparing the account number with the account number. The information, and the first reference image and the image to be verified for this account information, can speed up the comparison speed of the image to be verified.

The feature point identification part can be implemented by one of the overall features or the local features; the former compares the entire facial image as a single feature, and the latter first finds the local features of the face, such as the eyes, nose, The mouth, the corner of the mouth or the corner of the eye, etc., then compare these local features separately, and finally integrate the results of the individual local feature comparisons to obtain the final identification result. If the computing module 51 finds the first reference image in the account database that is similar to the image to be verified by a specified ratio, the user who operates the ATM 1 is the account owner. Therefore, the computing module 51 generates a renewable signal and sends the renewable signal to the automated teller 1 by the verification server 5. The continuation signal is a verification procedure for the user to pass the face recognition. Therefore, after receiving the continuation signal, the ATM 1 can instruct the transaction module 15 to perform a transaction operation according to the transaction content set by the user. Based on the above-mentioned face recognition mechanism, the transaction system disclosed in an embodiment of the present invention can prevent the fraudulent group from dispatching a driver to take the initiative in comparison with the traditional automatic teller machine transaction system which only uses the card password as the security verification.

Please refer to Figure 1. In another embodiment of the present invention, in order to improve the security of the verification based on the face recognition, the verification server 5 further includes an expression recognition module 52 electrically connected to the account database 50 and the operation module 51, and the expression recognition module 52 is used to analyze facial expressions in the image to be verified, such as angry, disgusting, scared, happy, surprised, and sad, and provide identification results to the computing module 51. In practice, in order to prevent the expression recognition module 52 from erroneously recognizing the expression of the image to be verified in order to prevent each person's expression change elements from being identical, when the account information of the account database 51 is established, in addition to storing the user's face image as the first In addition to a reference image, the method further includes storing a facial expression determined by the user as a second reference image. Therefore, when the verification server 5 receives the transaction content and the image to be verified transmitted by the automated teller machine 1, in addition to the first reference image in the account database 50 through the computing module 51, it is determined whether the image is the account owner. In addition, the expression recognition module 52 analyzes the expression of the image to be verified and the expression of the second reference image, and the expression recognition module 52 transmits the analysis result to the operation module 51 to compare the image to be verified with the second reference image. The similarity. In addition, the expression recognition module 52 may further include a machine learning function. In practice, when the image to be verified can correspond to one of the account information, the computing module 51 further includes storing the image to be verified and comparing the result to the corresponding account information. In other words, by adding the image to be verified that is retained each time the user performs the transaction operation to the ATM to the training network of the neural network of the expression recognition module, the richness and return of the recognizable type of the expression recognition module can be improved. The accuracy of the image to be verified to an expression classification. On the other hand, the present invention does not limit the second reference image to have only one set of expressions. The user can set a plurality of corresponding expressions according to the commonly used transaction items and store them in the second reference image; for example, the expression of 10,000 yuan is attached to the left eye, and the expression of 5,000 yuan is used to pick the right eyebrow. In general, when the user is coerced or involuntarily to the ATM, the user usually presents an expression of anger, disgust, or fear, and the expression recognition module 51 disclosed in the above embodiment is pre-retained. The design of the second reference image will assist the computing module 51 in detecting the condition that the user is being coerced and unloaded, and then performing countermeasures corresponding to the situation.

In another embodiment of the present invention, the camera module 10 further includes an infrared temperature sensing component, such as an infrared thermal imaging device, for obtaining the body temperature of the user when operating the automated teller machine 1, and the communication module 14 transmits the same. This temperature information is further included in the image to be verified. After receiving the temperature information, the operation module 51 of the verification server 5 determines whether the variation range exceeds a reasonable temperature difference range. In general, when a person is in a state of stress or agitation, the body temperature will rise involuntarily. Therefore, if the operation module 51 finds this phenomenon from the temperature information, it will improve its own similarity judgment standard for the image to be verified and the first reference image, and also improve the expression recognition module 52 for the image to be verified and the second. The standard value of the reference image is judged as a standard value. Through the above mechanism, it is possible to effectively detect the situation in which the culprits camouflage the original account users through masks or easy-to-capacity methods, and also detect abnormal changes in body temperature caused by the gangsters due to the tension of the illegal acts, thereby further enhancing the present invention. The described security of the transaction system based on face recognition verification.

In still another embodiment of the present invention, the verification server 5 further includes an alert module electrically connected to the computing module 51. The warning module is configured to selectively send an alert signal according to the comparison result of the computing module 51, for example, by calling the telephone to contact the police unit or sending the short message to the emergency contact set by the account.

Referring to FIG. 2, the following is a flow chart of the automatic teller machine 1 in the transaction system for verifying the face recognition according to an embodiment of the present invention. Verified transaction method.

Please refer to step S0: placing the financial card; specifically, the user inserts the financial card into a card reading module 11 of the ATM 1.

Please refer to step S1: capturing the image to be verified of the user's face; in detail, after the financial card is placed in the card reading module 11, the control module 13 of the ATM 1 then enables the camera module 10 to obtain the to-be-verified Image, and the image to be verified includes the face of the user.

Please refer to step S2: inputting a password; after the financial card is placed in the card reading module 11, the user inputs the password of the financial card through the input module 11 of the automatic teller machine.

Please refer to step S3: password verification; in detail, after the user inputs the password in step S2, the control module 13 of the ATM 1 generates a verification request according to the password and sends the verification request to the financial through the card reading module 11. The card, the financial card will generate a verification reply according to the verification request, and the card reader module 11 can receive the verification reply and transmit it to the control module 13 to determine whether the entered password is correct.

Please refer to step S4: input transaction content; after the control module 13 determines in step S3 that the password is verified, the control module 13 prompts the user to input a transaction content, for example, a withdrawal amount.

Please refer to step S5: prompting the password error; if the control module 13 determines that the password verification has not passed according to the verification reply, the user is prompted to input a password error, and returns to step S2 to allow the user to input the password again. In practice, after the user enters the wrong password for a specified number of times (for example, three times), the financial card will be locked, which is the protection mode that the automatic teller machine 1 already has, and will not be described here or related. The steps are in the flow chart of this embodiment.

Continuing to refer to step S6, the transaction content and the image to be verified are sent. Specifically, the control module 13 instructs the communication module 15 of the automatic teller machine to send the transaction content input in step S4 and the verification to be verified in step S1. Image to verification server 5. It should be particularly noted that, in another embodiment of the present invention, step S1 may be performed again in step S4. In other words, after the user passes the password verification, the control module 13 enables the camera module 10 to be enabled. The image to be verified is taken, which ensures that the face image is captured when the user password is input correctly, thereby reducing unnecessary power consumption of the camera module 10. In another embodiment of the present invention, the communication module 15 can send the image to be verified to the verification server 5 immediately after obtaining the image to be verified in step S1, so that the server can be verified during the input of the transaction content by the user. 5 Simultaneously performing the comparison work of the images to be verified, thereby reducing the time for the user to wait for the completion of the face verification program.

Please refer to step S7 to compare the reference image of the account database 50 with the image to be verified; in detail, the computing module 51 of the verification server 5 compares the account database 50 according to the image to be verified, wherein the account database 50 includes A plurality of account information and each account information includes at least a first reference image. In an embodiment of the present invention, the operation module 51 performs a facial recognition-related algorithm (for example, the Lucas-Kanade method) to calculate the image value to be verified and the first reference image one by one to obtain the degree of similarity between the two and determine Whether the similar procedure exceeds a predetermined threshold. In another embodiment of the present invention, the computing module 51 searches for a target account information having the same account number as the card account from the account database in advance (the account is obtained by the card reading module 11 in step S0). The method is sent to the control module 13 and transmitted to the verification server 5) through the communication module 14 in step S6. Then, the operation module 51 only produces a comparison result with the first reference image and the image to be verified for the target account information.

In another embodiment of step S7, the verification server 5 has an expression recognition module 52 for determining an expression of the image to be verified and determining the account database 51, in addition to determining whether the account is the user according to the first reference image and the image to be verified. The expression corresponding to the second reference image of the account owner. The second reference image is also a facial image that the user has previously retained in the financial institution, and the facial image includes an expression determined by the user, so that the second reference image can be transmitted through the expression recognition module. The judgment of the expression determines the expression of the image to be verified, and confirms whether the user is in a state of being forced to withdraw.

Please refer to step S8 "comparison pass" and step S9 "send renewable signal", in step S7, if the operation module 51 can find an account information in the account database, the first reference image and the image to be verified If the degree of similarity is above a predetermined threshold, it can be determined that the user who is to take the image to be verified is the account holder. Therefore, the operation module 51 can continue the signal according to the image to be verified and pass the verification server 5 The continuation signal is sent by the communication connection with the ATM 1.

Referring to step S10, the transaction operation is performed; in detail, when the control module 13 of the ATM 1 receives the renewable signal, the transaction module 15 is instructed to execute the item of the transaction content setting, for example, issuing five thousand cash. Yuan to the user.

Referring to step S11, the warning signal is sent. If the operation module 51 cannot find the first reference image corresponding to the image to be verified in the account database, it means that the current ATM 1 is likely to be fraudulent. The verification server 1 can send an alert signal to the police unit, the financial institution, the account holder, or the emergency contact set by the account person to achieve the effect of instant notification.

In summary, the transaction system and method for verifying based on facial recognition disclosed in the present invention, by photographing a user's face image at the ATM end and at the remote end of the verification server for the facial image and its expression Analyze the comparison to determine whether the user is abnormal in the facial expression of the card holder and the user, thereby effectively reducing the possibility of the fraud group taking the victim account or stealing the head account through the driver. It also instantly reflects the fact that the account itself is threatened rather than voluntarily. In other words, the present invention adds a check procedure to the user's account security according to the user's face recognition mechanism, and the facial surface disclosed by the present invention is disclosed without excessively affecting the user's original operation process on the automatic teller machine. Identifying the verification of the trading system and its methods is undoubtedly more secure for individual account security; for financial institutions, it can achieve the effect of immediate anti-blocking fraud or money laundering and other illegal financial activities.

Although the present invention has been disclosed above in the foregoing embodiments, it is not intended to limit the invention. It is within the scope of the invention to be modified and modified without departing from the spirit and scope of the invention. Please refer to the attached patent application for the scope of protection defined by the present invention.

1‧‧‧Automatic Teller Machine

10‧‧‧ camera module

11‧‧‧Reading module

12‧‧‧ Input Module

13‧‧‧Control Module

14‧‧‧Communication module

15‧‧‧Transaction module

5‧‧‧Verification server

50‧‧‧ Account Database

51‧‧‧ Computing Module

52‧‧‧Expression recognition module

S0~S11‧‧‧Steps

FIG. 1 is a structural diagram of a transaction system for verifying based on face recognition according to an embodiment of the invention. 2 is a flow chart of a transaction method for verifying based on face recognition according to an embodiment of the invention.

Claims (5)

A transaction system for verifying based on facial recognition includes: an automatic teller machine, comprising: a camera module for obtaining a to-be-verified image, the image to be verified includes a face of a user; For inputting a financial card, sending a verification request, and receiving a verification reply sent by the financial card; an input module is configured to input a card password and a transaction content; a control module is electrically connected to the camera a module, the card reading module and the input module, the control module is configured to receive the image to be verified, generate the verification request according to the card password, selectively generate a prompt message according to the verification reply, and selectively generate a prompt message Instructing the camera module to obtain the image to be verified, and the prompt message is used to prompt the user to input the transaction content, and the control module is further configured to receive the transaction content by the input module; a display module, electrical Connected to the control module, the display module is configured to instantly display the image to be verified captured by the camera module; a communication module is electrically connected to the control module and used to The control module sends the image to be verified and the transaction content; and a transaction module is electrically connected to the control module and configured to perform a transaction operation according to a renewable signal; and a verification server, communication The communication module connected to the ATM includes: An account database, comprising a plurality of account information, each of the account information includes at least a first reference image and a second reference image, wherein the second reference image corresponds to a predetermined transaction item; an expression recognition module, Electrically connected to the account database, the expression recognition module is configured to analyze an expression of the face in the image to be verified, and the expression recognition module has a training set of a neural network; an operation module, Electrically connecting the account database and the expression recognition module, the operation module is configured to compare at least one of the image to be verified and the first reference images, and the verification image and the number according to the expression At least one of the two reference images, and selectively generating the continuable signal according to a comparison result, wherein the continuable signal is associated with the preset transaction item, and the operation module is further configured to store the image to be verified and The comparison result is to the corresponding target account information and the training set; and a warning module is electrically connected to the operation module, and the warning module is configured to selectively send one according to the comparison result. Shows signal. The transaction system for verifying based on face recognition according to claim 1, wherein the camera module further comprises an infrared temperature sensing component for obtaining a body temperature of the user; and the communication module sends the waiting The verification image also includes information about the body temperature. The transaction system for verifying the face recognition according to claim 1, wherein the card reading module further comprises an account for obtaining the financial card; the control module further comprises receiving the account; the communication module The method further includes: sending an account number; each of the account information further includes an account number; and the computing module is further configured to find a target account information in the account database that is the same as the account number, and compare the to-be-verified The image and the first reference image of the target account information. A transaction method for verifying based on facial features, comprising: receiving a financial card by a card reader module of an automatic teller machine; after obtaining the financial card by the card reading module, obtaining the camera module by using one of the automatic teller machines The image to be verified includes a face of the user; the control module receives the image to be verified and instructs one of the communication modules of the automatic server to send the image to be verified to a verification server; After the card reader module obtains the financial card, an input module of the automatic teller machine receives a card password; and a control module of the automatic teller machine generates a verification request according to the card password and transmits the verification card through the card reader module. Sending the verification request to the financial card; receiving, by the card reading module, a verification reply of the financial card and transmitting the verification reply to the control module; and selectively generating a prompt message based on the verification response by the control module And selectively instructing the camera module to obtain the image to be verified, the prompt message is used to prompt the user to input the transaction content; The image to be verified is captured by the module; the control module receives the transaction content and instructs the communication module to send the transaction content to the verification server; and the operation module of the verification server is verified according to the verification The image comparison library stores an account database, wherein the account database includes a plurality of account information and each of the account information includes at least a first reference image and a second reference image; and the expression recognition module of the verification server is analyzed One of the faces of the face to be verified; The computing module compares the image to be verified with at least one of the second reference images in the account database according to the expression; and when the image to be verified corresponds to one of the account information, the computing module Storing the image to be verified and the comparison result to the corresponding account information; and using the computing module to compare the result and the verification image with the at least one of the image to be verified and the first reference images At least one of the second reference images aligns the result and selectively generates a continuable signal; when the computing module selects not to generate the continuable signal, the alert module sends the alert module a warning signal; when the computing module selects to generate the renewable signal, the verification server sends the renewable signal to the ATM; and the transaction module of the automated server is based on the renewable signal And the transaction content performs a transaction operation. The transaction method for verifying based on facial features according to claim 4, wherein: when the card reader module receives the financial card, the method further includes receiving, by the card reading module, an account number of the financial card; When the module instructs the communication module to send the transaction content to the verification server, the method further includes: the control module instructing the communication module to send the account; and the computing module comparing the account data according to the image to be verified Before the library, the method further includes: comparing, by the computing module, one of the account information and the account number, and the first reference image of the account information with the same account number and the account number to be verified The images are compared.