KR20000050230A - Method for authentication security on network by face recognition - Google Patents

Abstract

PURPOSE: A method for performing authenticating security on a network is provided to prevent unauthentic user accesses even when identifications(IDs) and passwords are leaked, by identifying the inputted IDs and the passwords to recognize inputted face images, and by finally going through authenticating procedures. CONSTITUTION: A method for performing authenticating security on a network comprises the steps of: if a client connects with a server, inputting authentication data based on texts to secure authenticating procedures from the client, and inputting a face image of the client; making the server receive authenticating security data from the client to identify the authentication data, and compare the face image of the client with a face image stored in the server, then deciding whether the client is registered; if so, allowing the client to connect with the network, and if not, prohibiting the client from connecting with the network.

Description

Method for authentication security on network by face recognition}

The present invention relates to an authentication security method on a network, and more particularly, to an authentication security method on a network by face recognition, which is suitable for performing a strict authentication process by performing a face recognition process.

In general, ID and password have been used for authentication security on a network.

Therefore, conventionally, when accessing a server through a network, the user can perform work on the server after first authenticating the user using an ID and password.

In the conventional ID and PASSWORD authentication method, the ID may be opened. In this case, the password may be the only protection target.

However, if a password is leaked, the reality is that it cannot be protected. That is, in the conventional password security system using a character input method such as ID and password, there is a problem that security cannot be achieved when the ID and password are leaked.

Accordingly, the present invention has been proposed to solve the conventional problems as described above, and an object of the present invention is to perform a face recognition process to perform a more rigorous authentication process to perform an authentication process that can prevent unintended access. By checking the ID and password input, finding and recognizing the face for the input face image, and finally undergoing the authentication process, face recognition can be prevented from access by unauthorized users when ID and password are leaked. It provides a security method for authentication on the network.

1 is a block diagram of a network system to which the present invention is applied;

FIG. 2 is a detailed block diagram of a client in FIG. 1;

3 is a flow chart showing an authentication security method on a network by face recognition according to the present invention;

4 is a diagram illustrating authentication security between a client and a server according to FIG. 3;

5 is a view showing in detail the authentication security performed in the client in FIG.

FIG. 6 is a diagram illustrating a detailed operation when logging in between a client and a server in FIG. 4.

Explanation of symbols on the main parts of the drawings

10: client 20: network

30: server

Hereinafter, with reference to the accompanying drawings the configuration and operation of the embodiment according to the technical spirit of the authentication security method on the network by the face recognition according to the present invention as follows.

First, the present invention is to supplement the conventional authentication security system by ID and password, by checking the input ID and password, by finding and recognizing the face for the input face image and finally undergoes the authentication procedure, Even if a password is leaked, we want to prevent unauthorized users from accessing it.

Therefore, in order to process the authentication process via the network, the face image input confirms abnormal access such as random data input that does not go through the dedicated input system through internal information confirmation between the server side and the client side. . Accordingly, it is possible to block the problem due to hacking from the beginning by checking whether a face image is input by a normal method.

1 is a block diagram of a network system to which the present invention is applied.

As shown therein, the client 10 inputs authentication information to access the server 30; A network (20) for connecting the client (10) to the server (30); The server 30 is configured to receive the user information of the client 10 through the network 20 to process the authentication.

FIG. 2 is a detailed block diagram of a client in FIG. 1.

As shown therein, a PC camera 11 for photographing a face image of a user; An input unit 12 for receiving a face image photographed by the PC camera 11; An image processor 13 for performing image processing on the face image input by the input unit 12; A data storage unit 14 for storing information on the face image processed by the image processor 13; And a data transmitter 15 for transmitting the face image information processed by the image processor 13 to the server 30 through the network 20.

Thus, the input unit 12 of the client 10 first receives a face image image of the user through the PC camera 11. In addition, the image processing unit 13 performs image processing required for data transmission format and face image recognition during authentication security at the server 30 through the network 20. The image processed data is then stored in the data storage unit 14. In addition, after transmitting authentication data based on text such as ID and password through the data transmission unit 15, image information about the face image of the user is transmitted.

Then, if the authentication data and image information based on the text is transmitted to the server 30 through the network 20, the server 30 receives the data based on the authentication data based on the internal data storage unit (not shown in the drawing). It is determined whether or not to allow access to the server 30 by comparing with the image information of the corresponding ID user stored in the).

3 is a flowchart illustrating a method for authenticating security on a network by face recognition according to the present invention.

As shown in the drawing, when the client 10 accesses the server 30, the client 10 inputs a text-based authentication data for authentication security and then inputs a face image image of the user. ST11) (ST12); The server 30 receives the authentication security data from the client 10 and checks the text-based authentication data, and then the face image image of the user received from the client 10 and the user's data stored in the server 30. Comparing the face image image to determine whether the user is a registered user (ST13 to ST16); If the user is a registered user, the client 10 is allowed to access the client 10, and if the user is not a registered user, a third step (ST17) (ST18) of disallowing access to the client 10 is performed.

In the first step, the face feature value is extracted when the face image image is input, and the server 30 determines the face feature value when the authentication security is determined by the face image image.

The facial feature value is a user-specific value extracted by analyzing the distance information between the facial feature elements (eye, nose, mouth), the relationship with the face outline, the face outline shape information, and the face face pattern information.

In the first step, when the face image image data in the form of a picture or a file is input at the time of initial registration, the client 10 repeatedly inputs different face image image data several times to repeat user-specific face feature values. To extract the facial feature values.

In the first step, when the face image image data is input from the client 10 through the PC camera 11 at the time of initial registration, the user of the client 10 faces the PC camera 11 for a predetermined time. By extracting the facial feature value unique to the user with the face image of the PC camera 11 taken for a certain time.

In the second step, a problem key randomly generated by the server 30 is sent to the client 10, and the client 10 sends the face image image data of the user together with the answer key to the problem key. Transmission to the server 30 to determine the authentication security.

FIG. 4 is a diagram illustrating authentication security between a client and a server according to FIG. 3.

Therefore, as shown in FIG. 4, the present invention complements the security system by analyzing the ID, password, and face image of an initial input user, the ID, password, and face image input next time. .

The detailed view of the client 10 side of FIG. 4 is a key (problem key to the client) transmitted from the server 30 side to the client 10 side through the network 20 to prevent hacking, and the server of FIG. 4. The detailed diagram on the 30 side shows the user data (ID, password, face image and face image feature values) for login authentication and the key capable of decrypting the encrypted data on the client side and the problem key transmitted from the server 30. The answer is for. Both keys are randomly generated every time, and based on this information sent from the client, hacking is judged, and if it matches with the user data on the server 30 side, login is allowed, if not, login is not. It is carried out in a system that does not allow.

The present invention can be summarized into two steps: initial user information registration and authentication through user information verification.

FIG. 5 is a diagram illustrating details of performing authentication security on a client in FIG. 4.

Thus, during initial user registration, the user's ID, password, and other user's information are input. Then, the front face image of the user is input as a moving image or a still image through the PC camera 11, and the image processor 13 extracts a face region. In addition, the image processor 13 analyzes a facial feature element on the extracted region and extracts a facial feature value of the corresponding user. Here, water-marking refers to a technique for adding special information to a face image in order to prepare for a leak problem caused by hacking of user's face image information registered in a server. If you don't know the technique, you can check it, so you can't interpret the image information normally unless it is a client for authentication.

In addition, face image input during initial user registration supports direct input and file input through a PC camera 11 and other video capturing equipment (not shown).

In addition, the facial feature value is a user-specific facial feature value through the analysis of the distance information between the facial feature elements (eye, nose, mouth), the relationship with the outside of the face, the shape of the face shape, and the face surface information (Texture).

In addition, when inputting the face image data in the form of photos or files during initial registration, different face image data is inputted several times to learn the unique facial feature values and extract the facial feature values. When inputting image data, sufficient input can be performed with a frontal gaze of several seconds. Both methods automatically check whether the face is in the front and whether the face is located within the camera's range by automatically looking at the front of the camera. The user will be notified by message.

FIG. 6 is a diagram illustrating a detailed operation when logging in between a client and a server in FIG. 4.

Thus, as shown in FIG. 5, the user authentication is performed on the registered user information as shown in FIG. 5. Therefore, the ID and password input from the local or remote place is checked, the face is detected and recognized for the input face image, and finally the authentication process is performed. This prevents unauthorized users from accessing ID and passwords.

In addition, when performing user authentication, the user's face image is input only through a video capture device such as a PC camera 11 and the like, and input by a file is blocked.

In addition, in order to prevent abnormal access such as input of arbitrary face image data that does not go through a dedicated input system when inputting face image for authentication, a key from the server 30 side (the side which receives user information and processes authentication) to the client 10 side. Randomly generate and send (1 in Figure 6). At this time, this key becomes a kind of problem key to the client 10. This key is received by the client 10 to calculate the answer key for this key and to be sent back to the server 30 side. Then, the client 10 re-encrypts the data input by the user and the answer key for the problem key sent by the server 30, and sends it back to the server 30 together with the decryption key (2 in FIG. 6). Will be blocked from the beginning.

As such, the present invention performs a more stringent authentication process by performing a face recognition process.

Although the preferred embodiment of the present invention has been described above, the present invention may use various changes, modifications, and equivalents. It is clear that the present invention can be applied in the same manner by appropriately modifying the above embodiments. Accordingly, the above description does not limit the scope of the invention as defined by the limitations of the following claims.

As described above, the authentication security method on the network by face recognition according to the present invention uses an input medium, such as a PC camera, which is becoming popular, and analyzes a user's face image information to analyze a security system based on an existing ID and password. By supplementing it, it is possible to prevent unintended access even when ID and password are leaked.

Claims (6)

A first step of inputting a face image image of a user after inputting text-based authentication data for authentication security at the client when the client accesses the server; The server receives the authentication security data from the client, checks text-based authentication data, and then compares the face image image of the user received from the client with the face image image of the user stored in the server to determine whether the user is a registered user. Performing a second step; And if the user is a registered user, permits access of the client, and if the user is not a registered user, performs a third step of disallowing access of the client. The method of claim 1, wherein the first step, And extracting a facial feature value when inputting a facial image image so that the server distinguishes the facial feature value by using the facial feature value when discriminating the authentication security by the face image image. The method of claim 2, wherein the facial feature value, Authentication on the network using face recognition, which is a user-specific value extracted through analysis of distance information between face feature elements (eye, nose, mouth), relationship with face outline, face outline shape information, and face pattern information Security method. The method of claim 1, wherein the first step, When inputting the face image image data in the form of a photo or a file at the time of initial registration, the client extracts the face feature value by repeatedly inputting different face image image data on the client several times to perform repetitive learning on the user's unique face feature value. Authentication security method on the network by face recognition. The method of claim 1, wherein the first step, When the client inputs the face image image data through the PC camera during initial registration, the client's user stares directly at the PC camera for a predetermined time and uses the face image of the PC camera taken for a predetermined time. Authentication security method on the network by facial recognition, characterized in that the extraction. The method of claim 1, wherein the second step, The server randomly generates a problem key generated by the server, and the client transmits the face image image data of the user together with the answer key for the problem key to the server for face recognition. Authentication security method on the network.