KR100564372B1 - System and method for automatically taking a human face front-view photograph - Google Patents

Abstract

The present invention is applied to an automatic cash dispenser equipped with an external device such as a card reader, a touch monitor, an automatic cash dispenser computer, and the like, in the automatic photographing camera system for checking the identity of a user using the automatic cash dispenser, the face of the user. A camera for capturing an image, receiving a face image from the camera, determining whether the input image is a front face image, generating an output signal for whether a front face image is acquired, and a unique password for generating a face image and a face password A client terminal including a computer storing a pattern and a switch receiving the output signal from the computer and intermitting a connection with the external device; and a management server storing unique information of the client terminal and an encryption pattern corresponding thereto. Face automatic shooting for automatic cash dispenser It relates to a camera system. The present invention is functionally separated from an automatic cash dispenser and has an effect of protecting a user's information through a stable and encrypted process.

Description

System and method for automatically taking a human face front-view photograph}

Figure 1a is a block diagram of a front face automatic photographing camera system for an automatic cash dispenser of the present invention

1B is a block diagram of a front face automatic photographing camera system for an ATM machine according to another embodiment of the present invention.

2 is a block diagram of an automatic cash dispenser to which the present invention is applied.

Figure 3 is a flow chart of the front face automatic shooting method associated with the automatic cash dispenser

4 is a flowchart for finding a front face in an arbitrary image image.

5 is a flowchart illustrating a process of encrypting and decrypting a face image determined to be a front face.

6 is a data structure diagram when storing and transmitting the front face image;

7A is a video image including a front face

7B is a schematic diagram of an encryption pattern

8 is a flowchart illustrating a method of generating a face password using cryptographic pattern location points.

9 is a key structure diagram of a server system for face password authentication

10 is a schematic diagram showing a central control model by the network connection of the present invention

11A illustrates a face image before shuffling of a user front face image.

11B shows a face image after shuffling part of the face image.

11C illustrates a case in which a header portion of data of a front face image is modulated

<Description of the symbols for the main parts of the drawings>

10: camera 20: computer

21: intelligent camera 22: input and output unit

23: image sensor 24: central processing unit

25: image signal processor 26: storage unit

28: communication unit 29: external storage unit

30: switch 42: card reader

44: touchmonitor 46: automatic teller machine

The present invention relates to a front face automatic photographing camera system and method thereof for an automatic cash dispenser. In particular, when the face image of a user is input from a camera attached to the front of an automatic cash dispenser (ATM or CD), the input image is a front face image. Automatic face cash camera for automatic cash dispenser that can protect the user's biometric information by controlling the use of the automatic cash dispenser according to the result of the determination and controlling the use of the automatic cash dispenser. A system and method thereof are provided.

This is a method to prevent crime using a conventional automatic cash dispenser. When a user's own card is lost or stolen, a method of preventing illegal use by a password or a camera automatic recording (DVR) at a place where an automatic cash dispenser is installed. A method was used to set up a device to keep records of people withdrawing cash so that it could be verified later.

However, the prevention of fraud through passwords cannot act to prevent crimes if the password is known and the criminal is withdrawn, wearing a helmet, wearing a mask or wearing dark glasses when the criminal withdraws the cash. In the case of covering a conventional camera automatic recording (DVR) device it is difficult to verify the identity of the person withdrawing cash.

In addition, the above methods are insufficient as a solution when the user deducts the cash from the card even though the user does not come out of the cash, and there is no means for leaving the user record of the ATM, which makes it difficult to detect the criminal.

Therefore, in order to overcome the above problems, there has been an attempt to install a user identification system using a camera in an automatic cash dispenser. Specifically, unlike the conventional method of reading the information on the magnetic tape, a smart card having a semiconductor chip embedded inside the credit card is generally used. Therefore, a credit card is stored by storing personal biometric information in the smart card. Attempts to use as a means of authentication or to add a fingerprint recognition system to an automatic cash dispenser and to use it as a means for user authentication in addition to a credit card, or to install a user identification system using a camera in an automatic cash dispenser. Yes.

However, attempts to install a user identification system using a camera in the automatic cash dispenser are complicated in combination with an ATM and a biometric recognition rate is applied when a face recognition system is used to verify a user's identity using a user's face. There was a problem that can be inconvenient to use does not reach 100%.

This problem is the same in the case of fingerprint recognition, and in particular, face recognition is known that the accuracy is much lower than the fingerprint recognition due to the change in illumination or pose, the perspective effect of the face size.

However, due to the recent advances in computer engineering, in the case of face recognition, the pattern recognition technology, which simply finds the eyes of the face or discriminates whether it is the front face under various lighting environments, has almost 100% technical completion and the computational amount is also the face recognition. Relatively small compared to technology. However, when storing the record of the user's identity as a moving picture, the larger the amount of data, the more the number of the automatic cash dispenser, the more administrative problems may occur.

In order to overcome this problem, there have been attempts to record a still image. In this case, unlike a moving image, the still image has a data characteristic that can be easily forged and tampered with. This may be a serious problem in terms of privacy since the user's face contains the biometric characteristics of the individual.

In particular, when sending and receiving face image information that can confirm the user's identity using the TCP / IP communication protocol of the Internet, the security becomes more vulnerable, and when exposed to hacking, the reliability of the face image for user identity authentication is greatly damaged. There is a problem that the system may not achieve its intended purpose.

An object of the present invention is to solve the above problems, and when the face image of the user is input from the camera attached to the front of the automatic cash dispenser, it is determined whether the input image is a front face image and the use of the automatic cash dispenser based on the determination result. The present invention provides a front face automatic photographing system and method for an automatic cash dispenser having a stable structure independent of the automatic cash dispenser.

In addition, another object of the present invention is to provide a front face automatic photographing system and method for an automatic cash dispenser that protects a user's information by performing an encryption process on an image determined as a front face and storing or transmitting the image.

Front face automatic photographing system for an automatic cash dispenser according to the present invention for achieving the above object is applied to an automatic cash dispenser equipped with an external device such as a card reader, a touch monitor, an automatic cash dispenser computer to use the automatic cash dispenser. An automatic photographing camera system for checking a user's identity, comprising: a camera photographing a face image of the user, a face image received from the camera to determine whether an input image is a front face image, and whether a front face image is obtained; A client terminal for generating an output signal and storing a unique encryption pattern, and a switch for receiving the output signal from the computer and intermitting a connection with the external device; And a management server storing unique information of the client terminal and an encryption pattern corresponding thereto.

The management server may store the unique information and the encryption pattern in a database, and the camera may be made of a solid state imaging device or a complementary field effect device image sensor.

In addition, the present invention is applied to an automatic cash dispenser equipped with an external device, such as a card reader, a touch monitor, an automatic cash dispenser computer, an automatic shooting camera system for checking the identity of a user using the automatic cash dispenser, the face of the user An intelligent camera for capturing an image and determining whether the captured image is a front face image and generating an output signal for acquiring a front face image, and a switch that receives the output signal from the intelligent camera and controls the connection with the external device. It may include.

In this case, the intelligent camera may further include an image sensor for detecting an image of the user and an image signal processor for transmitting the image of the user detected by the image sensor.

The computer or the intelligent camera may include an input / output unit for transmitting and receiving an electrical signal with the external device, a central processing unit for determining whether a face image transmitted from the camera or the image signal processor is a front face image, the unique information, and the face image. It may include a storage unit for storing a, a communication unit for transmitting and receiving the face image and the unique information stored in the storage unit via a network.

The storage unit is an EEPROM, a nonvolatile memory, a volatile memory, or a hard disk. The central processing unit searches for an eye in an input face image and normalizes two image regions estimated by both eyes in consideration of the size of the image and the position of the eye target region. The image processing is then performed to normalize the selected image, and to mask and project the eye and mouth parts to analyze the frequency characteristics.

The CPU may determine whether the front face image is a front face image, and output a signal for operating the switch in a connected state when the front face image is determined.

If the determination result is not a front face image, the switch outputs a signal for operating the switch in a blocking state, and the image processing and front face image are applied to the other one of the received input images until the front face image is acquired. It is characterized by repeating the determination of the authorization.

In addition, the central processing unit generates a face image file by using the encryption pattern on the image determined as the front face image and generates a face image file, and an encryption process of scrambling information corresponding to each pixel of the face image image. Characterized in that it comprises a. The face image file is composed of a header portion, a data portion, and an information portion, wherein the information portion includes a face password, an image capturing time, and a unique number of a client system acquiring an image.

In addition, the password pattern is defined by the administrator's prior appointment for the purpose of security authentication and stored in the storage, the defined password pattern is generated in the face password generation engine software in the central processing unit to generate a face password It is done.

In addition, the central processing unit generates a face password with a value obtained by performing a mathematical operation by a promised one-way function using two or more location point values obtained by matching by similar shape deformation of both eyes and the encryption pattern. It is done.

The present invention further includes a management server connected to the communication unit through a network, receiving a face image file including the unique information from the communication unit, and retrieving and decrypting an encryption pattern corresponding to the unique information from the DB. It may be characterized in that the configuration.

On the other hand, the image reconstructed by the software that reduces the scrambled image from the face image file to the original image, extracts only the front face image, generates a face code from the extracted front face image, and is generated in the encryption process. The method may include determining whether the face image is forgery or forgery compared to the face password, and the method for generating a face code in the decryption process may include two or more obtained by matching by similar deformation of both eyes and a password pattern. The face cipher may be generated by performing a mathematical operation using a promised one-way function using the location point values.

In addition, the present invention is an automatic shooting method for checking the identity of a user who uses the automatic cash dispenser as a client terminal interlocked with an automatic cash dispenser having an external device such as a card reader, a touch monitor, an automatic cash dispenser computer, (a) receiving an input signal of a card from the card reader, (b) inputting a face image of the user in real time, (c) an image processing step of processing an input image, and (d) a processed image Determining whether the face image is obtained and (e) controlling the connection with the external device based on whether a face image is obtained.

Step (c) is a step of processing the image by driving a software engine implemented by a pattern recognition algorithm that can distinguish the front face of a person by capturing the input image, (c1) finding an eye in the input image, (c2) normalizing the two image areas, which are assumed to be both eyes, in consideration of the size of the image and the position of the eye target area, and (c3) masking and projecting the eye and mouth areas with respect to the selected image. Characterized in that it comprises a.

In addition, in the step (e), when the front face image is acquired, the switch operates in a connected state, and when the front face image is not obtained, the switch operates in a blocked state. (B), (c), (d) is repeated until it is obtained.

And (f) generating a face cipher, (g) generating a face image file including information such as the face cipher and a face image, for the image determined as the front face image, (h) And scramble processing the information corresponding to each pixel of the face image file, and (i) storing or transmitting the face image file.

The facial image file may include a header portion, a data portion corresponding to the contents of the image, a face password, an image collection time, and an information portion including a unique number of the client system that acquired the image.

Step (f) is a step of finding an eye from the front face image (f1), and (f2) a unique encryption code promised with a reference point based on both eyes as a reference point in the target area of the eye found by the pattern recognition technique in step (c1). Matching a pattern, (f3) extracting a vertex position of the cryptographic pattern with respect to the reference point, (f4) extracting a video image coordinate value for the extracted vertex, (f5) one or more image image coordinates And generating a face password using the value and the image coordinate value of the reference point. At this time, the encryption pattern is included in the software for generating the encryption.

On the other hand, step (i) is characterized in that for transmitting the face image file to the management server that stores the unique information and the corresponding encryption pattern of the client terminal in the database (DB) and the management server (j) The method may further include receiving the face image file, and (k) retrieving and decrypting an encryption pattern corresponding to the unique information from the database.

In addition, the step (k) is (k1) a decoding step of restoring the image with software for reducing the scrambled image to the original image in step (h), (k2) extracting only the restored front face image Step (k3) generating a face password for the extracted front face image, (k4) comparing the face password included in the face image file with the face password generated in step (k3) of the transmitted face image It may be characterized by further comprising the step of determining whether the forgery.

In the step (k3), the eye is searched in the (k31) front face image, and (k32) two reference points and an encryption pattern of a predetermined form are defined based on any two eyes among the target areas of the found eye. Matching, (k33) extracting encryption pattern vertex positions for the reference points, (k34) extracting image image coordinate values for the extracted vertices, (k35) one or more image image coordinate values and reference points It may be characterized in that it comprises a step of generating a face password using the image coordinate value for.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1A is a diagram illustrating a configuration of a front face automatic photographing camera system for an automatic cash dispenser as an embodiment of the present invention.

Referring to FIG. 1A, the computer 20 includes an input / output unit 22, a central processing unit 24, a storage unit 26, and a communication unit 28.

The input / output unit 22 plays a role of transmitting and receiving electrical signals from an external device, and the central processing unit 24 performs all other arithmetic processing including an image image processing process transmitted from the camera 10.

 In addition, the storage unit 26 stores contents including the unique information of the computer 20, and may be implemented as various storage devices such as an EEPROM, a nonvolatile memory, a volatile memory, and a hard disk. The storage unit 26 is provided inside the computer 20 to provide a means for storing a large amount of information.

In addition, the communication unit 28 serves to connect to the network. The communication unit 28 is connected to a network through an application such as TCP / IP, PSTN, LAN, WAN, and the like to transmit and receive data.

Meanwhile, the switch 30 interacts with the computer 20 to block or connect an external device for the purpose of controlling small signal transmission between the external devices. When the card reader 42, the touch monitor 44, the automatic cash dispenser computer 46 are installed as an external device, and the external device is connected to the computer 20 through a switch 30 located between the external devices. When the card is input to the card reader 42, the card reader 42 sends a start signal to the input / output unit 22 of the computer 20.

Then, the input / output unit 22 transmits a start signal to the central processing unit 24, and the central processing unit 24 outputs an output signal corresponding to the start signal. At this time, the switch 30 outputs a signal ("1" or "0") for controlling small signal transmission between the touch monitor 44 or the automatic teller machine 46, which is another external device in the input / output unit 22. Received from the central processing unit 24 to block or connect the input signal of the automatic cash dispenser user.

Figure 1b is another embodiment of the front face automatic photographing camera system for an automatic cash dispenser as another embodiment of the present invention.

Referring to FIG. 1B, the automatic face photographing camera system of the present invention includes an intelligent camera system 21 including a camera 10 and a computer 20 as a single product, and a switch 30. In this case, the camera system 21 further includes an image sensor 23 and an image signal processor 25 capable of processing signals of the image sensor.

In addition, the front face automatic shooting camera system of the present invention may be configured to include a separate external storage unit 29 in the outside as well as the camera system 21. In this case, the external storage unit 29 may accommodate various types of storage devices. In addition, the operation principle of the front face automatic photographing camera system of the present invention shown in FIG. 1B is the same as that of FIG. 1A.

Figure 2 shows the configuration of the automatic cash dispenser to which the present invention is applied.

Referring to FIG. 2, the automatic cash dispenser includes a camera 10, a computer 20, and a switch 30. The system includes a card reader 42, a touch monitor 44, and an automatic cash dispenser computer 46. ), It is applied to a general automatic cash dispenser consisting of a touch panel 48 is implemented as a whole system.

The computer 20 of the present invention receives the starting point of use of the automatic cash dispenser of the user (customer) from the card reader 42 and receives the user's face image photographed from the camera 10, and then the front face of the computer 20. After the determination is made, the operation state (switching state) of the switch 30 is changed according to the determination result.

3 is a flowchart illustrating an operation procedure of a front face automatic photographing camera linked with an automatic cash dispenser.

In the present invention, between the credit card input step (S210) and the transaction selection step (S260) to install a switch for the purpose of simply opening and closing the small signal transmission passage of the electrical signal, the operation timing and operation results of the switch (opening and closing state) ) Control the use of the ATM machine by controlling the electrical signal (eg, "1" or "0").

As shown in FIG. 3, when the user (customer) inserts the card into the ATM (S210), the ATM automatically starts to operate. When the signal indicating that the card is input is transmitted to the input / output unit 22 of the present invention, the central processing unit 24 generates a start signal, and the switch 30 is switched to the blocking state by the control signal from the central processing unit 24. . If no additional release signal is input, the current state is maintained.

A process for outputting "Yes" (or a connected state) of the small signal control switch state will be described in detail below.

In the state in which the external video signal is continuously input from the camera 10 to the computer 20, the step S220 is performed for processing one video signal captured by the start signal and the front face of the person can be distinguished. In step S230, a software engine implemented by a pattern recognition algorithm is processed to perform signal processing on the input image. Next, it is determined whether the image is a front face image (S240).

As a result of the determination, if it is not the front face image (No in S240), the face image input step S220 is repeated to select another one of images input in real time from the camera, and then the image signal processing step is performed. (S230), the front face image determination step (S240) is repeatedly performed. As a result of the determination, in the case of the front face image (YES in S240), after outputting a control signal for connecting the switch 30 (S250), the user may select a transaction type to be used later (S260). .

After that, the operation sequence of the ATM is the same as using a conventional ATM. In other words, after entering the cash withdrawal amount and password through the transaction selection step (S260) and withdraw the cash, a series of processes are completed. Meanwhile, in the present invention, an encryption step (S300) of an image determined as a front face image is performed. In the present invention, when it is necessary to store the image determined as the front face image or to transmit using the Internet, the step of storing or transmitting may be additionally performed (S350).

4 is a flowchart for finding a front face in an arbitrary image image. In detail, the input image signal processing step S230 and the front face image determination step S240 of FIG. 3 are described in detail. .

If a single image is captured from the images continuously input from the camera along with the image processing start signal, an eye searching step S232 is performed from the captured input image. In this case, the eye finding method extracts a plurality of image areas corresponding to the eye target area from the RGB or YCbCr image considering the various environmental conditions in which the input image is taken, and then correlates the eye and eyebrow hairs. Estimate and judge.

Then, in the next step, normalization is performed on an image having a size of 128 * 128 pixels for two image regions estimated by both eyes (S234). In the normalization process, the size of the image and the position of the eye target region are considered. .

Next, masking and projection of the eye and mouth target regions of the previous stage image and analysis of the image frequency components of the corresponding region are performed (S236). It is determined whether a front face exists (S240). In the step of determining whether the front face exists, the position of the eye is comprehensively determined again in the eye finding step, and if the front face does not exist ("No" in S240), the other image inputted from the camera is determined. The same image processing process is repeated for the subject.

In the above face discrimination method, it is determined by reflecting the frequency characteristics of the image constituting the face image, and it is different from other face recognition technologies such as the feature point extraction method and the eigenface method applied in the conventional face recognition technology. There is a significant difference.

In the present invention, after setting the eye target area using both eyes, eyebrows and hairs in the face image, by analyzing the frequency characteristics of the eyes and mouth region image by the method of determining whether the front face is sunglass, Distinguishing a face wearing a helmet or mask and providing a means for controlling the use authority of the ATM user (S250 switch control).

5 is a diagram illustrating an encryption and decryption procedure of a face image determined as a front face.

Referring to FIG. 5, the encryption process of the face image is described in detail. A face password is generated using the position points of both eyes determined in the process of finding the face in FIG. 4 from the acquired face image as a reference point (S310). . This will be described in detail later with reference to FIGS. 6 and 7.

After generating the face password, the next step is to construct a face image file. In this case, the face shape file includes a header part having information on the size of the compressed image of the image image, a data part corresponding to the image content of the image, and a face password and image collection, which are additionally generated and included for security authentication purposes in the present invention. It consists of an information part including information such as the time, the unique number of the system from which the image was obtained, and the information of the information part is generated in a new image file format that may be included in the header part (S320).

6 is a view for explaining the structure of the face image file in the user front face image storage and transmission of the camera device.

As shown in FIG. 6, the facial image file is made of a file standard with a header portion 80, a data portion 82 corresponding to image contents, and an information portion 84 further included in the present invention. The newly added information part of the present invention includes information such as a face password, an image acquisition time, a unique number of a system for acquiring an image, and the like.

In particular, the file structure of FIG. 6 sets the face information generated by the image processing process from the face and the image information containing the face image as a single file standard so that the face image and the face code have a correlation with each other. It provides a new means for improving the security of the face image by the method of judging the face image. In addition, by using a combination of the face password, time, and the unique number of the camera device of the information portion, it is possible to prevent and verify forgery modulation of the face image transmitted through the network, thereby increasing the security of the present invention.

After generating the face image file as described above, the face image encoding step of scrambles the information corresponding to each pixel of the face image image in a promised manner so that a part or the whole of the image cannot be identified. In operation S 330, a face image of which the encryption process is completed is stored in the same standard as that of the data portion of the face image file or transmitted to the Internet or a network (S 350).

In the present invention, the face password is used as an authenticity determination tool for the face image of the individual judged as the front face, and the user is allowed to use the whole or two means by mixing the personal information so that an unauthorized person cannot see it. By selectively using the method, it is possible to improve the effectiveness of security while improving the convenience of management.

Hereinafter, a process of decrypting an image file generated by the encryption process will be described in detail.

First, a face image decoding step (S410) of restoring an image with software that descrambles the scrambled image back to the original image is performed.

In addition, thereafter, only the front face image is extracted from the reconstructed image (S420) and the face password generation step (S430) is sequentially performed. And the same software engine algorithm used to determine the front face image in FIG.

The next step is to compare the face code included in the transmitted face image file with the newly generated face code through the front face image step (S420) and the face code generation step (S430). In step S440, a determination is made as to whether or not a forgery is performed, and the process proceeds to an original face image step S450 or a fake or modulated face image step S455.

7A and 7B are diagrams for explaining a method of generating a face password in the above step. FIG. 7A shows an image image 60 including a front face 61, and FIG. 7B shows an encryption pattern.

In order to generate a face password from the image image of FIG. 7A, the two eyes of FIG. 7A are reference points (P1, P2) (62, 63), and two predetermined position points (C1, C2) using the encryption pattern of FIG. 7B. ), 64,65. The encryption pattern is known and manageable only to the device installer or a specific person who wants to manage the camera of the present invention, and the encryption pattern is determined in a predetermined form. The cryptographic pattern is included in the software for cryptographic generation.

The cryptographic pattern of FIG. 7B allows both eyes to be the two vertices of a triangle and simultaneously acts as a reference line between the reference points (62, 63) and both eyes, and based on triangulation In order to derive the same position point for various sizes and rotations of the face image, the reference point and the reference line are matched to deform the shape of the encryption pattern.

FIG. 8 illustrates a flowchart of generating a face password using the encryption pattern location points of FIG. 7A.

First, an operation of searching an eye from an input front face image (S302) is performed. At this time, in the step of finding both eyes, any one point (for example, the center of the eye target area) among the eye target areas found by the pattern recognition technology can be recognized as the eye and used as a reference point. After the step of finding the eye, a step of matching the reference point and the encryption pattern in the face is performed (S304), and then the other one vertex position extracted as the encryption pattern is extracted (S306).

Then, the image image coordinate value corresponding to the vertex position extracted in the next step is extracted based on the image coordinate system having the upper left as the (0,0) reference point (S308). In the next step, one or more image image coordinate values and the reference point are extracted. A face password is generated using the image coordinate value for (S310).

More specifically, at this time, the result value obtained by calculating the coordinate values of two arbitrary position points C1 and C2 as a previously promised one-way function is used as the face password. For example, when multiplying numbers of coordinate values corresponding to one location point (C1) and another location point (C2) with each other and using the value as a face password, the two location points (C1) used to generate a password are used. Since the coordinate value of, C2) is an arbitrary value, of course, the result value is also generated as an arbitrary value, and even after performing the same operation as a simple multiplication, the calculated result values are inversely estimated from the two position point (C1, C2) values. There is an advantage that can not. Therefore, the face password generated by the above method is a separate numerical value different from the face image and has a meaning as an authentication code generated for each face image.

In addition, since the position of the reference point of two eyes and the position of the remaining vertices of the encryption pattern according to the movement of the position of two eyes for an arbitrary face image are always changed according to the movement of the face in the image coordinate system, it is possible to make an encryption system using the feature points of the face. Provide means.

In addition, forgery and modulation of image signals can be prevented by applying RGB image values of the same position together with the coordinate values corresponding to two position points (C1, C2) as arbitrary variable values adopted in the face code generation process. It is possible to provide a means to improve the security of the entire face image by implementing the number of location points to two or more points.

In addition, in the image processing process of the face engine that finds the front face and the eye, even for the photographs of the face images taken sequentially under the same environmental conditions, the eye position may be moved finely due to the slight change of the face image. Also, the position of the vertices extracted as the encryption pattern is changed by moving or changing the pose finely, and at the same time, the generation and duplication of the face password for an arbitrary face image is very difficult because of applying the same encryption pattern.

In addition, the above-mentioned face password can be easily produced and produced, but since it has a structural characteristic that cannot be restored, it is difficult to generate and duplicate the face password without replicating the entire security system. Therefore, the face cipher represents a new cipher generation system combining the biometric value of the face and the principle of the artificial cipher pattern promised in advance.

9 shows a key structure of a server system for face password authentication.

FIG. 9 shows an image image 60 including the front face 61, reference points 62 and 63 representing both eyes in the face, and several vertices 64 generated by encryption patterns previously promised from the reference point. , 65, 66, 67). Automatic face photographing for multiple automatic cash dispensers Each different face password pattern embedded in the camera system can be configured and managed in the server system with a password pattern corresponding to the total sum. By the method of determining the respective camera system can be authenticated or determined.

In the case of the conventional network management system, the Internet IP or the hardware unique number of the terminal is used as a means for recognizing the system of each end. In the present invention, in addition to the hardware means, the operation of the software solution processed at each end is performed. There is an effect of providing a security means that can recognize the hardware of the end by using the biometric data including the front face of the individual transmitted from each end to the central control system of the server. The above method provides a security means that is superior and differentiated from the conventional method as a security means of biological characteristics using a face image of a person to be secured.

Figure 10 shows a central control model by the network connection of the present invention device.

Referring to FIG. 10, a plurality of automatic photographing camera systems 74, 76, and 78 of the present invention, which are composed of a camera 10 and a computer 20, are connected to the Internet 70. A model for authenticating and controlling facial images transmitted from camera apparatuses is shown. The control system of FIG. 10 has the server system key of FIG. 9, so that each camera device can be distinguished according to the form and combination of the previously-provided encryption patterns.

11 illustrates an encryption method by shuffling a user front face image.

FIG. 11A is a face image before the shuffling process, FIG. 11B is a face image when a part of the data portion image of FIG. 10 is shuffled, and FIG. 11C shows that the face image itself cannot be identified by modulating the header portion of the compressed image of FIG. One case is shown. Accordingly, the method of FIG. 11 may implement a security system that protects personal information and does not cause administrative inconvenience by modifying part or all of the face image according to a need of customer or user personal information management.

The system of the present invention as described above is a camera system capable of intelligently judging and photographing only the front face of a person, providing a control action to an external device, and storing and transmitting a face image. It can also be used as an unmanned surveillance device.

The present invention is not limited to the above-described embodiments and may be variously modified and changed by those skilled in the art, which are included in the spirit and scope of the present invention as defined in the appended claims.

As described above, according to the present invention and a method for face automatic photographing for an automatic cash dispenser according to the present invention, if a face image of a user is input from a camera attached to the front of the automatic cash dispenser, the input image is determined to be a front face image. By controlling the use of the automatic teller machine according to the determination result, the automatic photographing system of the present invention has a structure that is functionally separated from the automatic teller machine. It has the effect of having a structure.

The present invention also provides a front face automatic photographing system and a method for an automatic cash dispenser for protecting a user's information by performing an encryption process on an image determined as a front face and storing or transmitting the front image.

Claims (27)

As an automatic shooting camera system applied to an automatic cash dispenser equipped with an external device such as a card reader, a touch monitor, an automatic cash dispenser computer, etc., to identify a user using the automatic cash dispenser. A camera for photographing the face image of the user; A computer for receiving a face image from the camera to determine whether an input image is a front face image, generating an output signal for whether a front face image is obtained, and storing a unique encryption pattern for generating a face image and a face password; A client terminal including a switch receiving the output signal from the computer to control a connection with the external device; And A management server storing unique information of the client terminal and a corresponding encryption pattern; Including, The computer is An input / output unit for transmitting and receiving an electrical signal with the external device; Finding eyes from the face image transmitted from the camera, normalizing the two image areas estimated by both eyes in consideration of the size of the image and the position of the eye target area, and masking and projecting the eye and mouth parts to analyze frequency characteristics A central processing unit which determines whether the face image inputted through the image processing process is the front face image; A storage unit which stores the unique information and the face image; And And a communication unit for transmitting and receiving a face image and the unique information stored in the storage unit through a network. The central processing unit is promised one-way using two or more location point values defined by the administrator's prior appointment for security authentication purposes and obtained by matching by similar variations of the password pattern stored in the storage. Encryption including a process of generating a face password by performing a mathematical operation by a function, a process of generating a face image file including the face password, and a process of scrambling information corresponding to each pixel of the image image A front face automatic photographing camera system for an automated teller machine, characterized in that to carry out the process. The method of claim 1, wherein the management server Front face automatic photographing camera system for an automatic cash dispenser, characterized in that for storing the unique information and the encryption pattern in a database (DB). The method of claim 1, wherein the camera A front face automatic photographing camera system for an automatic cash dispenser, characterized by being made of a solid state image pickup device or a complementary field effect element image sensor. As an automatic shooting camera system applied to an automatic cash dispenser equipped with an external device such as a card reader, a touch monitor, an automatic cash dispenser computer, etc., to identify a user using the automatic cash dispenser. An intelligent camera that photographs a face image of a user and determines whether the captured image is a front face image and generates an output signal for whether a front face image is obtained; And A switch which receives the output signal from the intelligent camera and regulates a connection with the external device; Including, The intelligent camera An input / output unit for transmitting and receiving an electrical signal with the external device; An image sensor detecting an image of a user; An image signal processor for transmitting an image of the user sensed by the image sensor; Find the eye from the face image transmitted from the image signal processor, normalize the two image areas estimated by both eyes in consideration of the size of the image and the position of the eye target area, and mask and project the eye and mouth parts to obtain frequency characteristics. A central processing unit for determining whether the input face image is a front face image through an analyzing image processing process; A storage unit which stores the unique information and the face image; And A communication unit for transmitting and receiving a face image and the unique information stored in the storage unit through a network; More, The central processing unit is promised one-way using two or more location point values defined by the administrator's prior appointment for security authentication purposes and obtained by matching by similar variations of the password pattern stored in the storage. Encryption including a process of generating a face password by performing a mathematical operation by a function, a process of generating a face image file including the face password, and a process of scrambling information corresponding to each pixel of the image image A front face automatic photographing camera system for an automated teller machine, characterized in that to carry out the process. delete delete The method of claim 4, wherein the storage unit A front face auto-taking camera system for an automated teller machine, characterized by an EEPROM, nonvolatile memory, volatile memory or a hard disk. delete The method of claim 4, wherein the central processing unit Judging whether it is a front face image, and outputting a signal for operating the switch in a connected state when the front face image is determined, If the determination result is not a front face image, the switch outputs a signal for operating the switch in a blocking state, and the image processing and front face image are applied to the other one of the received input images until the front face image is acquired. An automatic front face photographing system for an automated teller machine, characterized in that the determination of authorization is carried out. delete The method of claim 4, wherein the facial image file And a header portion, a data portion, and an information portion, wherein the information portion includes a face password, an image collection time, and a unique number of a client system that acquires an image. delete delete The method of claim 4, wherein And a management server connected to the communication unit through a network, receiving a face image file and the unique information from the communication unit, and retrieving and decrypting an encryption pattern corresponding to the unique information from the DB. Automatic face camera system for automatic cash dispensers. The method of claim 14, Reconstruct the image with the software to reduce the scrambled image from the face image file to the original image, extract only the front face image, generate a face code from the extracted front face image, and generate the face in the encryption process. A front face automatic photographing camera system for an automated teller machine, characterized in that it determines whether a face image is forged or altered in comparison with a password. The method of claim 15, wherein the decoding process An automatic cash dispenser characterized in that a face cryptogram is generated by performing a mathematical operation by a promised one-way function using two or more location point values obtained by matching by similar variations of both eyes and a cryptographic pattern. Front face automatic shooting camera system. A client terminal interlocked with an automatic cash dispenser equipped with an external device such as a card reader, a touch monitor, an automatic cash dispenser computer, and the like, as an automatic photographing method for checking the identity of a user who uses the automatic cash dispenser, (a) receiving an input signal of a card from the card reader; (b) inputting a face image of the user in real time; (c) Normalize the two image areas estimated from the input image to be both eyes in consideration of the size of the image and the position of the eye target area, and mask and project the eye and mouth areas on the selected image. Doing; (d) determining whether the processed image is a front face image; and (e) controlling the connection with the external device based on whether a front face image is acquired; (f) generating a face cipher; (g) generating a face image file including information such as the face code and face image; (h) scramble processing information corresponding to each pixel of the image with respect to the face image file; And (i) storing or transmitting the face image file; Including, Step (f) (f1) finding eyes in the front face image; (f2) matching the reference point with the promised unique cryptographic pattern using both eyes as a reference point in the target area of the eye found by the pattern recognition technique in step (c1); (f3) extracting vertex positions of the encryption pattern with respect to the reference point; (f4) extracting image image coordinate values of the extracted vertices; And and (f5) generating a face code using one or more image image coordinate values and image coordinate values of a reference point. delete 18. The method of claim 17, wherein step (e) When the face image is acquired, the switch operates in a connected state, If the front face image is not obtained, the switch operates in a blocked state, and in this case, the automatic cash is repeated by repeating steps (b), (c), and (d) until the front face image is obtained. Automatic face-to-face shooting method for the dispenser. delete 18. The method of claim 17, wherein the facial image file Automatic face photographing method for an automatic cash dispenser comprising a header portion, a data portion corresponding to the contents of an image, a face password, an image collection time, and an information portion including a unique number of a client system that acquires the image. . delete The method of claim 17, wherein the encryption pattern is Automatic face photographing method for an ATM machine, characterized in that included in the software for generating the password. 18. The method of claim 17, wherein step (i) Automatic face photographing method for an automatic cash dispenser, characterized in that for transmitting the face image file to the management server that stores the unique information of the client terminal and the corresponding encryption pattern in the database (DB). The method of claim 24, wherein the management server (j) receiving the face image file; and (k) retrieving and decrypting an encryption pattern corresponding to the unique information from the database. The method of claim 25, wherein step (k) (k1) a decoding step of restoring the image by software for reducing the scrambled image to the original image in step (h); (k2) extracting only the restored front face image; (k3) generating a face password for the extracted front face image; (k4) comparing the face password included in the face image file with the face code generated in step (k3), and determining whether the forgery or forgery of the transmitted face image further comprises an automatic cash dispenser. Front face auto-shooting method. The method of claim 26, wherein step (k3) (k31) finding eyes in the front face image; (k32) matching two reference points with an encryption pattern of a predetermined form by using any two eyes as a reference point among the found eye target areas; (k33) extracting encryption pattern vertex positions with respect to the reference points; (k34) extracting image image coordinate values of the extracted vertices; And (k35) generating a face password using one or more image image coordinate values and image coordinate values for a reference point; Front face automatic shooting method for an automatic cash dispenser, characterized in that configured to include.

Images