TWI819765B - Cardless finance transaction system and its host server - Google Patents

Abstract

A cardless financial transaction system includes an ATM and a host server. The ATM includes a camera device and a touch screen. The camera device captures a facial image of an operator, the touch screen displays an unlocking array including a plurality of areas, and receives a trajectory pattern input by the operator from the unlocking array. The host server includes a database unit, an image recognition module, a symbol generating unit and a processing unit. The database unit has a plurality of user data corresponding to user members, the image recognition module is used for recognizing facial images, and the symbol generation unit is used for randomly distributing the plurality of symbol elements in the areas of the unlocking array. The processing unit sequentially determines whether the facial image and the trajectory pattern respectively match the user information of the user member so as to be a basis for whether to verify the identity certification of the operator.

Description

Cardless financial transaction system and host server

The invention relates to a financial transaction system, in particular to a cardless financial transaction system and a host server.

Generally speaking, in the existing automatic teller machine (ATM) cash withdrawal system, the withdrawer needs to insert a debit card and use his or her hands to enter a numeric password on the ATM keyboard. However, entering the numeric password may cause harm to others. Risk of leakage due to skimming or peeping.

In addition, ATMs also include a cardless withdrawal mechanism. One of the identity verification methods of this mechanism is through vein identification of the withdrawer for identity verification. However, ATMs that use vein recognition must be equipped with specific vein sensing devices, so additional hardware equipment is required, increasing hardware costs, and thus failing to increase the willingness of operators to install it.

It can be seen that the above-mentioned technology obviously still has inconveniences and defects, and needs to be further improved. Therefore, how to effectively solve the above inconveniences and defects is indeed one of the current important research and development topics, and it has also become an urgent need for improvement in related fields.

One object of the present invention is to provide a cardless financial transaction system and a host server to solve the above difficulties mentioned in the prior art.

One embodiment of the present invention provides a cardless financial transaction system. The cardless financial transaction system includes a host server and at least one automatic teller machine. The automatic teller machine includes a first transmission unit, a first camera device and a touch screen. The first camera device is used to capture a first facial image of an operator, and the touch screen is used to display an unlocking array including a plurality of array blocks, and receive a trajectory input by the operator dragging from the unlocking array. pattern. The host server includes a second transmission unit, a database unit, an image recognition module, a symbol generation unit and a processing unit. The second transmission unit is electrically connected to the first transmission unit through a network architecture. The database unit stores multiple pieces of user information corresponding to different user members. The image recognition module is used to recognize the first facial image. The symbol generating unit is used to randomly allocate multiple symbol elements into these array blocks. The processing unit is electrically connected to the second transmission unit, the database unit, the image recognition module and the symbol generation unit, and is used to sequentially determine whether the first facial image and the trajectory pattern match the user information of one of the user members to determine whether the authentication is passed. The basis for identity authentication of operators.

According to one or more embodiments of the present invention, the above-mentioned cardless financial transaction system further includes a bank server. The bank server includes a second camera device, a third transmission unit and a registration unit. The second camera device is used to capture a second facial image of a person. The third transmission unit is electrically connected to the second transmission unit through the network architecture. The registration unit is electrically connected to the second camera device and the third transmission unit, and is used to require the image recognition module to recognize the second facial image, parse the second facial image into a facial feature group, and store the facial feature group in the database unit.

According to one or more embodiments of the present invention, in the above-mentioned cardless financial transaction system, the image recognition module includes a segmentation unit, a marking unit and a calculation unit. The segmentation unit is used to perform an image segmentation technology on the first facial image, and segment the first facial image into a first square matrix including a plurality of first squares of the same size. The marking unit is electrically connected to the dividing unit, and is used to respectively mark a part of the first squares corresponding to all edges of the first square matrix, and another part of the first squares at other positions. The calculation unit is electrically connected to the marking unit, and is used to calculate a plurality of first feature vectors corresponding to other parts of the first squares of the first facial image, and these first feature vectors are arranged into a first feature matrix.

According to one or more embodiments of the present invention, in the above-mentioned cardless financial transaction system, the segmentation unit further performs image segmentation technology on the second facial image, and segments the second facial image into a plurality of second facial images of the same size. One of the squares and the second square. The marking unit further marks a part of the second square grid corresponding to all edges of the second square matrix, and another part of the second square grid at other positions. The calculation unit further calculates a plurality of second feature vectors corresponding to other parts of the second squares in the second facial image, and these second feature vectors are arranged into a second feature matrix. Calculate the standard deviations corresponding to the second eigenvectors in the second eigenmatrix.

According to one or more embodiments of the present invention, in the above-mentioned cardless financial transaction system, the processing unit further compares these first feature vectors and these second feature vectors; determines each first feature vector and the corresponding second feature Whether the difference of the vectors is within the corresponding standard deviation; and when the processing unit determines that each difference is within the corresponding standard deviation, the processing unit determines that the first facial image matches the user profile of the user member.

According to one or more embodiments of the present invention, in the above-mentioned cardless financial transaction system, the host server further includes a question collection unit. The question collection unit is electrically connected to the registration unit and the processing unit, and is used to collect at least one authentication question and one authentication answer corresponding to the authentication question sent by one of the user members from the bank server, and store the authentication question and the corresponding authentication answer. into the database unit.

According to one or more embodiments of the present invention, in the above-mentioned cardless financial transaction system, when the processing unit determines that the first facial image matches the user information of the user member, the processing unit is further configured to respond to the user member's user information. User information displays authentication questions and unlocking arrays simultaneously on the touch screen, and allocates authentication answers to a portion of these array blocks. The symbol generating unit is further used to generate a plurality of random characters and distribute the random characters to the remaining parts of the array blocks. The processing unit is further used to parse an input data according to the drag sequence of the trajectory pattern, and when it is determined that the input data matches the authentication answer, the identity authentication of the operator is passed.

According to one or more embodiments of the present invention, in the above cardless financial transaction system, when the processing unit determines that the first facial image matches the user information of the user member, the processing unit is further configured to respond to the user member's user information. User information, the authentication questions and the unlocking array are displayed on the touch screen at the same time, and the authentication answers are allocated in a part of these array blocks. The symbol generating unit is further used to generate a plurality of random characters and distribute the random characters into another part of the array blocks. The processing unit is further used to parse an input data according to the drag sequence of the trajectory pattern, and when it is determined that the input data contains these random characters, the operator's identity is authenticated.

According to one or more embodiments of the present invention, in the above-mentioned cardless financial transaction system, the touch screen further outputs a numeric key interface to receive an operator's input of a numeric group from the numeric key interface. When the processing unit determines that the first facial image matches the user information of the user member, the processing unit determines whether the number group input by the operator matches the user information of one of the user members, as a result of whether the identity authentication of the operator is passed. basis.

An embodiment of the present invention provides a host server. The host server is suitable for a cardless financial transaction system, and the host server includes a second transmission unit, a database unit, an image recognition module, a symbol generation unit and a processing unit. The database unit stores multiple pieces of user information corresponding to different user members. The transmission unit is electrically connected to an automatic teller machine through a network architecture, and is used to receive an operator's facial image from the automatic teller machine and a trajectory pattern input on a touch screen of the automatic teller machine. The image recognition module is used to recognize the facial image. The symbol generating unit is used to randomly allocate a plurality of symbol elements into an unlocking array displayed on the touch screen. The unlocking array includes a plurality of array blocks. The processing unit is electrically connected to the transmission unit, image recognition module, symbol generation unit and database unit, and is used to sequentially determine whether the facial image and trajectory pattern match the user information of one of the user members, as a basis for identity authentication of the operator. basis.

According to one or more embodiments of the present invention, in the above-mentioned host server, the image recognition module includes a segmentation unit, a marking unit and a computing unit. The segmentation unit is used to perform an image segmentation technology on the facial image, and segment the facial image into a square matrix including a plurality of squares of the same size. The marking unit is electrically connected to the dividing unit, and is used to respectively mark a part of the squares corresponding to all edges of the square matrix, and another part of the squares at other positions. The calculation unit is electrically connected to the segmentation unit and the marking unit, and is used to calculate a plurality of feature vectors corresponding to other parts of the facial images of the squares, and these feature vectors are arranged into a feature matrix.

According to one or more embodiments of the present invention, the above-mentioned host server further includes a question collecting unit. The question receiving unit is electrically connected to the processing unit, and is used to collect at least one authentication question and one authentication answer corresponding to the authentication question sent by the user member from a bank server, and store the authentication question and the corresponding authentication answer in the data. within the library unit.

According to one or more embodiments of the present invention, in the above host server, when the processing unit determines that the facial image matches the user information of the user member, the processing unit is further configured to simultaneously display the authentication question and the unlock array on the touch screen. and distribute the authentication answer within one of these array blocks. The symbol generating unit is further used to generate a plurality of random characters and distribute the random characters to the remaining parts of the array blocks. The processing unit is further used to parse an input data according to the drag sequence of the trajectory pattern, and when it is determined that the input data matches the authentication answer, the authentication passes the operator's identity authentication.

According to one or more embodiments of the present invention, in the above-mentioned cardless financial transaction system, when the processing unit determines that the facial image matches the user information of the user member, the processing unit is further configured to simultaneously display the authentication question and the unlock array on the touch screen and distribute the authentication answer within one of these array blocks. The symbol generating unit is further used to generate a plurality of random characters and distribute the random characters to the remaining parts of the array blocks. The processing unit is further used to parse an input data according to the drag sequence of the trajectory pattern, and when it is determined that the input data contains these random characters, the operator's identity is authenticated.

In this way, through the structure described in the above embodiments, using the surveillance cameras and specific identification mechanisms on the existing ATMs, the cardless withdrawal mechanism can quickly and effectively authenticate the identity of the operator. In addition, for banks, there is no need to add additional hardware equipment, which can greatly increase the bank's willingness to set up.

The above is only used to describe the problems to be solved by the present invention, the technical means to solve the problems, the effects thereof, etc. The specific details of the present invention will be introduced in detail in the following embodiments and related drawings.

The following will disclose multiple embodiments of the present invention in the drawings. For the sake of clarity, many practical details will be explained in the following description. However, it will be understood that these practical details should not limit the invention. That is to say, in various embodiments of the present invention, these practical details are not necessary. In addition, for the sake of simplifying the drawings, some commonly used structures and components will be illustrated in a simple schematic manner in the drawings.

Figure 1 is a block diagram of a cardless financial transaction system 10 according to an embodiment of the present invention. As shown in FIG. 1 , the cardless financial transaction system 10 includes a host server 200 and one or more automatic teller machines 100 (ATM). The host server 200 is connected to these ATMs 100 respectively through a network architecture N.

Figure 2 is a detailed block diagram of the ATM 100 in Figure 1 . As shown in Figures 1 and 2, each ATM 100 allows an operator to perform card-present or card-less financial transactions in person, such as withdrawals, deposits, transfers or inquiries. Each ATM 100 has a first transmission unit 110, a first camera device 120, a touch screen 130, an input unit 140 and a control unit 150. The control unit 150 is electrically connected to the first transmission unit 110, the first camera device 120, the touch screen 130 and the input unit 140 respectively, and controls each of the above units to operate appropriately. The first transmission unit 110 is used to electrically connect to the corresponding external mechanism through the above-mentioned network architecture N. The first camera device 120 is used to capture an image of an operator, for example, to capture a facial image of the operator's face (ie, face or face) located at the front edge of the ATM 100 . The touch screen 130 is used to display a prompt screen 500 to the above-mentioned operator, and the prompt screen 500 is used to prompt the above-mentioned operator to perform the above-mentioned financial transaction operation. For example, when performing identity authentication, the prompt screen 500 includes an unlock array 520 of a plurality of array blocks 521 (for example, nine), and receives input from the operator (i.e., dragging) on the unlock array 520. A trajectory pattern 523. The input unit 140 is used to allow the operator to subsequently input execution instructions. For example, the first transmission unit 110 and the control unit 150 are respectively located in the machine of the ATM 100, and the first camera device 120, the touch screen 130 and the input unit 140 are respectively located in the front edge of the machine of the ATM 100.

Figure 3 is a detailed block diagram of the host server 200 of Figure 1 . As shown in Figures 2 and 3, the host server 200 includes a second transmission unit 210, a database unit 220, an image recognition module 230 and a processing unit 260. The second transmission unit 210 is electrically connected to the first transmission unit 110 of the ATM 100 through the above-mentioned network architecture N (FIG. 1). The database unit 220 stores multiple pieces of user information 221 corresponding to different user members. Each user member's user profile 221 is unique or exclusive, that is, unique and non-duplicate. The image recognition module 230 is used to recognize and analyze the facial image of the operator. In this embodiment, the image recognition module 230 analyzes the operator's facial features based on the operator's facial image. The processing unit 260 is electrically connected to the second transmission unit 210, the database unit 220 and the image recognition module 230, and is used to sequentially determine whether the operator's facial image and trajectory pattern match the user information 221 of one of the user members, as a basis for authentication. The basis for passing the identity authentication of the operator.

Figure 4 is a detailed block diagram of the bank server 300 in Figure 1 . Figure 5 is a schematic diagram of image segmentation of the facial image 400 in this embodiment. As shown in Figures 1 and 4, the above-mentioned cardless financial transaction system 10 further includes one or more bank servers 300. Each bank server 300 is located, for example, in one of the bank branches, and each bank server 300 includes a second camera 310 , a third transmission unit 320 and a registration unit 330 . The third transmission unit 320 is electrically connected to the second transmission unit 210 of the host server 200 through the network architecture N (FIG. 3). The second camera device 310 is used to photograph the face (ie, face or face) of a person (eg, a registered person) to capture a facial image 400 . The registration unit 330 electrically connects the second camera device 310 and the third transmission unit 320.

Therefore, when the registered person wants to apply for the cardless withdrawal service, the registration unit 330 requires the registered person to perform a plurality of input steps to complete the above registration process for the cardless withdrawal service. In addition to personal data in these input steps, the registration unit 330 also requires the image recognition module 230 to recognize the facial image 400 of the registered person and parse the facial image 400 of the registered person into a facial feature group, thereby creating this facial feature group. It is stored in the database unit 220 as part of the user information 221 corresponding to the registered person.

As shown in Figures 3 and 5, the image recognition module 230 includes a segmentation unit 231, a marking unit 232 and a calculation unit 233. The segmentation unit 231 is used to perform an image segmentation technology on the facial image 400 of the registered person, and segment the facial image 400 of the registered person into a square matrix 410 containing a plurality of squares of the same size (for example, 468 or more). The marking unit 232 is electrically connected to the dividing unit 231, and is used to respectively mark a part of the squares corresponding to all edges of the square matrix 410 (such as dot squares 411), and other parts of the squares in other positions (such as non-dot squares). 412). The calculation unit 233 is electrically connected to the marking unit 232, and is used to calculate a plurality of second feature vectors corresponding to other parts of the squares of the facial image 400 of the registered person, so as to obtain a second feature matrix. In addition, the calculation unit 233 further calculates the standard deviations respectively corresponding to the second feature vectors in the second feature matrix, and the calculation result is, for example, 468 feature vectors.

More specifically, for example, each face is represented as a list of 468 facial landmarks (i.e., facial feature vectors), each landmark consisting of an X-axis, a Y-axis, and a Z-axis. The X-axis and Y-axis are normalized to [0.0~1.0] by the image width and height respectively. The Z-axis represents the depth of the landmark with the depth of the head center as the origin. The smaller the value, the closer the landmark is to the camera. The size of the Z-axis uses roughly the same scale as the X-axis.

In this way, when performing identity authentication on the operator on the front side of the ATM 100, the ATM 100 transmits the operator's facial image to the image recognition module 230, so that the segmentation unit 231 further performs image segmentation technology on the facial image 400, and The facial image 400 is divided into a square matrix 410 including a plurality of squares of the same size. Then, the calculation unit 233 is further used to calculate a first feature vector corresponding to each square of the facial image 400. These first feature vectors are Arranged into a first characteristic matrix.

The processing unit 260 compares these first feature vectors and these second feature vectors, and determines whether the difference between each first feature vector and the corresponding second feature vector is within the corresponding standard deviation. When the processing unit 260 determines that each The differences are all within the corresponding standard deviation, and the processing unit 260 determines that the facial image 400 of the operator on the front side of the ATM 100 matches the user profile 221 of the user member.

In this way, since the personal feature matrix of each user member is unique and non-duplicate in the system 10, when the processing unit 260 compares the facial image of the operator with the facial feature records of each customer in the database unit 220, whether a match is found At this time, the processing unit 260 calculates the difference between the facial features of each user member in the database unit 220 and the first feature matrix of the operator. If the 468 feature vectors of the operator are all within the gap range, it is determined to be the most suitable customer. .

As shown in Figures 3 and 4, the host server 200 further includes a symbol generating unit 240 and a question collecting unit 250. The question receiving unit 250 is electrically connected to the processing unit 260 and the registration unit 330 of the bank server 300. The question collecting unit 250 further collects one or more authentication question banks sent by the registered person from the bank server 300 when the registered person applies for the cardless withdrawal service. Each authentication question bank includes an authentication question 511 and a certification answer 512 corresponding to the authentication question 511, and the question collection unit 250 stores these authentication question banks in the database unit 220 as part of the user information 221 corresponding to the registered person. , to complete the registration process. The authentication questions 511 of the authentication question bank are, for example, ID number, birthday, zodiac sign, favorite movie, pet name, lucky number or credit card number, etc. However, the present invention is not limited thereto. The symbol generation unit 240 is electrically connected to the processing unit 260, and the symbol generation unit 240 is used to randomly allocate a plurality of symbol elements (such as text or numerical values) into the array blocks 521 of the unlocking array 520.

More specifically, when the above-mentioned operator is determined to be the most suitable customer, the processing unit 260 randomly extracts an authentication question 511 from the authentication question bank of the user information 221 stored by the above-mentioned operator in the database unit 220, and authenticates this The authentication questions 511 of the question bank and the above-mentioned unlocking array 520 are displayed on the touch screen 130 at the same time. The processing unit 260 randomly places the authentication answers 512 into a part of these array blocks 521, and the symbol generation unit 240 generates a plurality of random words. elements 522, and distribute these random characters 522 within the rest of the array block 521. Next, the processing unit 260 parses an input data according to the drag sequence of the trajectory pattern 523 and determines whether the input data matches the authentication answer 512 . Therefore, when the processing unit 260 determines that the input data matches the authentication answer 512, the processing unit 260 authenticates the operator's identity authentication; otherwise, the processing unit 260 denies the operator's identity authentication.

It should be understood that the input data parsed by the above-mentioned processing unit 260 according to the drag sequence of the trajectory pattern 523 can be a text string or the trajectory pattern 523, so that the processing unit 260 can compare the text of the authentication answer 512 on the same basis. string, or the trajectory pattern 523 formed by the authentication answer 512 on the unlocking array 520.

Figures 6A and 6B are respectively schematic diagrams showing the authentication questions 511 displayed in different orders on the touch screen 130 in Figure 2 . As shown in Figure 6A, for example, when the authentication question 511 of the extracted authentication question bank is "What elementary school did you graduate from?" and the authentication answer 512 corresponding to the authentication question 511 is "Beimen Elementary School", the processing unit 260 The four characters "北", "门", "国" and "小" are randomly allocated to a part of these array blocks 521, and the symbol generating unit 240 will be, for example, "East", "South", "West". The random characters 522 of ", "new" and "中" are randomly distributed in the remaining parts of the array block 521. In this way, when there are more authentication questions 511, the difficulty of counterfeiting will be increased and the person who made the withdrawal can be confirmed, thereby improving the reliability and security of the system.

As shown in Figures 6A and 6B, even if the same authentication question bank is drawn again, compared to Figure 6A, the symbol generation unit 240 will generate different random characters 522 in Figure 6A, such as "new", The random characters 522 of "大", "安", "日" and "三", and the symbol generating unit 240 will not allocate these random characters 522 to the same position of the array block 521 to reduce the risk of illegal activities. Opportunities for people to figure out a specific authentication answer 512.

FIG. 6C and FIG. 6D are schematic diagrams of the touch screen 130 displaying the authentication question 511 in different embodiments. This embodiment is roughly the same as the above-mentioned Figure 6A. The difference is that, as shown in Figure 6C, when the processing unit 260 determines that the input data contains the random characters 522 in the unlock array 520, the processing unit 260 will authenticate the operation. The identity authentication of the operator, otherwise, the processing unit 260 denies the identity authentication of the operator.

For example, in Figure 6A, when the operator drags through "North", "Gate", "Country" and "Small" in sequence from the unlocking array 520, the processing unit 260 determines that the input data matches the authentication answer. 512, therefore, the processing unit 260 authenticates the operator's identity authentication; on the other hand, in Figure 6C, when the operator drags the trajectory pattern 524 from the unlocking array 520 through "west", "east", "south", All random characters 522 of "new" and "中", that is, when "North", "Gate", "国" and "小" are left without dragging through, because the processing unit 260 determines that the input data includes unlocking The random characters 522 in the array 520 are, therefore, the processing unit 260 authenticates the operator's identity.

As another example, this embodiment is roughly the same as the above-mentioned Figure 6A. The difference is that, as shown in Figure 6D, when the processing unit 260 determines that the input data contains at least one of the plurality of characters in the authentication answer 512, The processing unit 260 will authenticate the identity authentication of the operator. Otherwise, the processing unit 260 will deny the identity authentication of the operator.

For example, in Figure 6A, the operator must drag through "North", "Gate", "Guo" and "Small" in order from the unlocking array 520 before he or she can be authenticated and the operator's identity is authenticated; on the other hand, in In Figure 6D, the processing unit 260 will only display the three characters "北", "国" and "小" in the unlocking array 520. It is not necessary to display "北", "门", "国" and "小". "Four words, thus reducing the chance of criminals guessing the number of specific authentication answers 512."

Figures 7A and 7B are respectively schematic diagrams showing different authentication questions 511 displayed on the touch screen 130 in Figure 2 . As shown in Figure 7A, when the operator on the front side of the ATM 100 passes the matching step of the facial image 400, the touch screen 130 further outputs a numeric key interface 530 to receive the operator's input from the numeric key interface 530. Number group 531. Next, the processing unit 260 determines whether the number group 531 input by the operator matches the user profile 221 of one of the user members, as a basis for whether the identity authentication of the operator is passed.

In addition to the certification question bank being "graduated from elementary school", the certification question bank also contains the birthdays of specific user members. As shown in Figure 7A, when the authentication question 511 of the extracted authentication question bank is "What is your birthday?" and the authentication answer 512 corresponding to the authentication question 511 is "02/06", the processing unit 260 will "0", Three numbers "2" and "6" are randomly allocated to a part of these array blocks 521, and the symbol generation unit 240 selects from "1", "3", "4", "5", "7", " Six of the random characters 522 of "8" and "9" are randomly selected (9 array blocks minus 3 numbers) and randomly distributed in the remaining parts of the array block 521.

In addition, in addition to the "birthday" in the authentication question bank, the authentication question bank also contains the ATM card password of a specific user member. As shown in Figure 7A, when the authentication question 511 of the extracted authentication question bank is "What is the ATM card password?", and the authentication answers 512 corresponding to this authentication question 511 are "0", "1", "3", "2", "9", "6", "7" and "8", the processing unit 260 converts "0", "1", "3", "2", "9", "6", " The eight numbers "7" and "8" are randomly assigned to a part of these array blocks 521, and the symbol generation unit 240 generates a random character 522 of "5" and distributes it to the remaining parts of the array block 521.

Figure 8 is a schematic flow sequence diagram of the cardless financial transaction system 10 of this embodiment. As shown in Figures 3 and 8, when an operator performs a cardless financial transaction at the ATM 100, in step A, the ATM 100 first allows the first camera device 120 to capture the operator's face. Image 400; then, when the processing unit 260 determines that the facial image 400 matches the user information 221 corresponding to the operator in the database unit 220, step B is entered. In step B, the ATM 100 randomly selects an authentication question bank (for example, what is a primary school), and simultaneously displays the authentication questions 511 and the unlock array 520 of the authentication question bank on the touch screen 130 (FIG. 6A). The operator drags on the touch screen 130 and sequentially inputs the authentication answers 512; conversely, when it is determined that the facial image 400 does not match the user information 221, the host server 200 denies the operator's identity authentication and removes the authentication answer from the touch screen 130. Raise an error message.

Next, when the processing unit 260 determines that the input authentication answer 512 matches the user information 221 corresponding to the operator in the database unit 220, step C is entered. In step C, the ATM 100 randomly selects another authentication question bank (for example, what is your birthday), and displays the authentication question 511 and the unlocking array 520 of this authentication question bank on the touch screen 130 at the same time (FIG. 7A). For the operator to drag on the touch screen 130 to sequentially input the authentication answer 512; on the contrary, when it is determined that the input authentication answer 512 does not match the user information 221, the host server 200 denies the operator's identity authentication and starts from the touch screen. Screen 130 presents a recognition error message.

Next, when the processing unit 260 determines that the input authentication answer 512 matches the user information 221 corresponding to the operator in the database unit 220, step D is entered. In step D, the ATM 100 randomly selects another authentication question bank (for example, what is the ATM card password), and displays the authentication question 511 and the unlock array 520 of this authentication question bank on the touch screen 130 at the same time (FIG. 7B ) for the operator to drag on the touch screen 130 to sequentially input the authentication answer 512; conversely, when it is determined that the input authentication answer 512 does not match the user information 221, the host server 200 denies the operator's identity authentication, and A recognition error message is raised from the touch screen 130 .

Then, when the processing unit 260 determines that the input authentication answer 512 matches the user information 221 corresponding to the operator in the database unit 220, the host server 200 sends a correct identity verification message from the touch screen 130 and agrees to allow Operators continue to conduct card-not-present financial transactions.

It should be understood that in the above embodiments, the host server 200 and the bank server 300 are, for example, server devices or other similar components. The first transmission unit 110, the second transmission unit 210 and the third transmission unit 320 are, for example, a network interface controller (NIC) or other similar components. The first camera device 120 and the second camera device 310 are, for example, a camera device or other similar components. The input unit 140 is, for example, a keyboard, a touch panel, or other similar components. The database unit 220 may be, for example, any type of fixed or removable random access memory (Random Access Memory, RAM), read-only memory (Read-Only Memory, ROM), flash memory (flash memory), Hard disk (Hard Disk Drive, HDD), solid state drive (Solid State Drive, SSD) or similar components or combinations of the above components, the present invention is not limited thereto. The control unit 150, the processing unit 260, the symbol generation unit 240 and the registration unit 330 are, for example, a central processing unit 260 (Central Processing Unit, CPU), or other programmable general-purpose or special-purpose microprocessors (Microprocessors), Digital signal processor (Digital Signal Processor, DSP), programmable controller, application specific integrated circuit (Application Specific Integrated Circuit, ASIC) or other similar components or combinations of the above components, the present invention is not limited thereto. The automatic teller machine is not limited to a conventional hardware device or an online ATM located in the terminal device. Since the automatic teller machine 100 (ATM) is a conventional device, it will not be described in detail here.

In this way, through the structure described in the above embodiments, using the surveillance cameras and specific identification mechanisms on the existing ATMs, the cardless withdrawal mechanism can quickly and effectively authenticate the identity of the operator. In addition, for banks, there is no need to add additional hardware equipment, which can greatly increase the bank's willingness to set up.

Finally, the above disclosed embodiments are not intended to limit the present invention. Anyone skilled in the art can make various modifications and modifications without departing from the spirit and scope of the present invention, and all of them are protected by the present invention. Inventing. Therefore, the protection scope of the present invention shall be determined by the appended patent application scope.

10:Card financial transaction system 100:ATM 110: First transmission unit 120: First camera device 130:Touch screen 140:Input unit 150:Control unit 200: Host server 210: Second transmission unit 220: Database unit 221:User information 230:Image recognition module 231: Split unit 232: Mark unit 233:Computing unit 240: Symbol generation unit 250: Question collection unit 260: Processing unit 300:Bank server 310: Second camera device 320: Third transmission unit 330:Registration unit 400:Facial image 410:Phalanx 411: dot grid 412: No dot grid 500: Prompt screen 511:Authentication problem 512:Authentication answer 520:Unlock array 521:Array block 522: Random characters 523, 524: trajectory pattern 530:Number key interface 531: Number group A~D: steps N: Network architecture

In order to make the above and other objects, features, advantages and embodiments of the present invention more apparent and understandable, the accompanying drawings are described as follows: Figure 1 is a block diagram of a cardless financial transaction system according to an embodiment of the present invention; Figure 2 is a detailed block diagram of the automatic teller machine of Figure 1; Figure 3 is a detailed block diagram of the host server in Figure 1; Figure 4 is a detailed block diagram of the bank server in Figure 1; Figure 5 is a schematic diagram of facial image segmentation in this embodiment; Figures 6A and 6B are respectively schematic diagrams showing the authentication questions displayed in different orders on the touch screen in Figure 2; Figures 6C and 6D are schematic diagrams of touch screens displaying authentication questions in different embodiments; Figures 7A and 7B are respectively schematic diagrams showing different authentication questions displayed on the touch screen in Figure 2; and Figure 8 is a schematic flow sequence diagram of the cardless financial transaction system of this embodiment.

Domestic storage information (please note in order of storage institution, date and number) without Overseas storage information (please note in order of storage country, institution, date, and number) without

200: Host server

210: Second transmission unit

220: Database unit

221:User information

230:Image recognition module

231: Split unit

232: Mark unit

233:Computing unit

240: Symbol generation unit

250: Question collection unit

260: Processing unit

Claims (14)

A cardless financial transaction system that includes: At least one automatic teller machine includes a first transmission unit, a first camera device and a touch screen. The first camera device is used to capture a first facial image of an operator. The touch screen is used to display An unlocking array including a plurality of array blocks, and receiving a trajectory pattern input by the operator dragging from the unlocking array; and A host server includes a second transmission unit, a database unit, an image recognition module, a symbol generation unit and a processing unit, the second transmission unit is electrically connected to the first transmission unit through a network architecture, The database unit stores a plurality of user data corresponding to different user members. The image recognition module is used to recognize the first facial image. The symbol generation unit is used to randomly allocate a plurality of symbol elements to the array areas. In the block, the processing unit is electrically connected to the second transmission unit, the database unit, the image recognition module and the symbol generation unit to sequentially determine whether the first facial image and the trajectory pattern match the users. The user information of one of the members is used as the basis for whether the identity authentication of the operator is passed. The cardless financial transaction system as described in request item 1 further includes a bank server, and the bank server includes: a second camera device for capturing a second facial image of a person; a third transmission unit electrically connected to the second transmission unit through the network architecture; and A registration unit, electrically connected to the second camera device and the third transmission unit, for requesting the image recognition module to recognize the second facial image, parse the second facial image into a facial feature group, and store the data within the library unit. The cardless financial transaction system as described in claim 2, wherein the image recognition module includes: A segmentation unit used to perform an image segmentation technology on the first facial image, and segment the first facial image into a first square matrix including a plurality of first squares of the same size; A marking unit, electrically connected to the dividing unit, used to respectively mark a part of the first squares corresponding to all edges of the first square matrix, and another part of the first squares at other positions; and A calculation unit, electrically connected to the marking unit, used to calculate a plurality of first feature vectors corresponding to the other parts of the first squares of the first facial image, and the first feature vectors are arranged as a first A feature matrix. As claimed in claim 3, the cardless financial transaction system, wherein the segmentation unit further performs the image segmentation technology on the second facial image, and segments the second facial image into a plurality of second parties of the same size. A second square matrix of a grid, the marking unit further marks a part of the second squares corresponding to all edges of the second square matrix, and another part of the second squares at other positions, and the calculation unit further calculates A plurality of second feature vectors corresponding to the other parts of the second squares are obtained from the second facial image, the second feature vectors are arranged into a second feature matrix, and the second feature vector in the second feature matrix is calculated. The standard deviations corresponding to these second eigenvectors. The cardless financial transaction system of claim 4, wherein the processing unit further compares the first feature vectors with the second feature vectors; determines whether each of the first feature vectors is the same as the corresponding corresponding one. Whether the difference of the second feature vector is within the corresponding standard deviation; and when the processing unit determines that each of the differences is within the corresponding standard deviation, the processing unit determines that the first facial image matches the The user profile of one of the user members. The cardless financial transaction system as described in request item 2, wherein the host server further includes: A question collection unit, electrically connected to the registration unit and the processing unit, for collecting at least one authentication question and a pair of authentication answers sent by one of the user members from the bank server, and authenticating the The questions and the corresponding authentication answers are stored in the database unit. The cardless financial transaction system of claim 6, wherein when the processing unit determines that the first facial image matches the user information of one of the user members, the processing unit is further used to respond to one of the users. The user information of the member displays the authentication question and the unlocking array on the touch screen simultaneously, and allocates the authentication answer to a part of the array blocks. The symbol generating unit is further used to generate a plurality of random characters, and allocate the random characters to the remaining parts of the array blocks. The processing unit is further used to parse an input data according to the drag sequence of the trajectory pattern, and determine that the input data matches the When authenticating the answer, the identity of the operator is authenticated. The cardless financial transaction system of claim 6, wherein when the processing unit determines that the first facial image matches the user information of one of the user members, the processing unit is further used to respond to one of the users. The user information of the member displays the authentication question and the unlocking array on the touch screen simultaneously, and allocates the authentication answer to a part of the array blocks. The symbol generating unit is further used to generate a plurality of random characters, and allocates the random characters to the remaining parts of the array blocks. The processing unit is further used to parse an input data according to the drag sequence of the trajectory pattern, and determine that the input data contains the When these random characters are used, the authentication passes the identity authentication of the operator. The cardless financial transaction system as described in claim 1, wherein the touch screen further outputs a numeric key interface to receive the operator's input of a numeric group from the numeric key interface, When the processing unit determines that the first facial image matches the user profile of one of the user members, the processing unit determines whether the number group input by the operator matches the user of one of the user members. The information is used as the basis for whether the identity authentication of the operator has passed. A host server suitable for a cardless financial transaction system, including: A database unit that stores multiple pieces of user information corresponding to different user members; A transmission unit, electrically connected to an automatic teller machine through a network architecture, for receiving an operator's facial image from the automatic teller machine and a trajectory pattern input on a touch screen of the automatic teller machine; An image recognition module used to recognize the facial image; a symbol generating unit for randomly allocating a plurality of symbol elements into an unlocking array displayed on the touch screen, the unlocking array including a plurality of array blocks; and A processing unit, electrically connected to the transmission unit, the image recognition module, the symbol generation unit and the database unit, for sequentially determining whether the facial image and the trajectory pattern match the user of one of the user members. The information is used as the basis for whether the identity authentication of the operator has passed. The host server of claim 10, wherein the image recognition module includes: a segmentation unit for performing an image segmentation technology on the facial image, and segmenting the facial image into a plurality of squares of the same size. a square matrix of grids; a marking unit, electrically connected to the dividing unit, used to respectively mark a part of the square grids corresponding to all edges of the square matrix, and other parts of the square grids at other positions; and a calculation unit , electrically connected to the marking unit to calculate a plurality of feature vectors of the other parts of the facial image corresponding to the squares, and the feature vectors are arranged into a feature matrix. The host server as described in request item 10 further includes: A question collection unit, electrically connected to the processing unit, is used to collect at least one authentication question and a pair of authentication answers sent by one of the user members from a bank server, and to collect the authentication question and the corresponding authentication answer. The authentication answer is stored in the database unit. The host server of claim 12, wherein when the processing unit determines that the facial image matches the user information of one of the user members, the processing unit is further used to respond to the user information of one of the user members, The authentication question and the unlocking array are displayed on the touch screen simultaneously, and the authentication answer is allocated in a part of the array blocks. The symbol generating unit is further used to generate a plurality of random characters and generate the These random characters are distributed in the remaining parts of the array blocks. The processing unit is further used to parse an input data according to the drag sequence of the trajectory pattern, and when it is determined that the input data matches the authentication answer, the authentication is passed. The identity authentication of the operator. The host server of claim 12, wherein when the processing unit determines that the facial image matches the user information of one of the user members, the processing unit is further used to respond to the user information of one of the user members, The authentication question and the unlocking array are displayed on the touch screen simultaneously, and the authentication answer is allocated in a part of the array blocks. The symbol generating unit is further used to generate a plurality of random characters and generate the These random characters are distributed in the remaining parts of the array blocks. The processing unit is further used to parse an input data according to the drag sequence of the trajectory pattern, and when it is determined that the input data contains the random characters, The authentication passes the identity authentication of the operator.

Images