WO2013163915A1 - 人脸身份认证方法和***、计算机存储介质 - Google Patents

人脸身份认证方法和***、计算机存储介质 Download PDF

Info

Publication number
WO2013163915A1
WO2013163915A1 PCT/CN2013/073807 CN2013073807W WO2013163915A1 WO 2013163915 A1 WO2013163915 A1 WO 2013163915A1 CN 2013073807 W CN2013073807 W CN 2013073807W WO 2013163915 A1 WO2013163915 A1 WO 2013163915A1
Authority
WO
WIPO (PCT)
Prior art keywords
face
interaction
authentication
track
trajectory
Prior art date
Application number
PCT/CN2013/073807
Other languages
English (en)
French (fr)
Inventor
喻欣
秦雷
胡育辉
郝允允
Original Assignee
腾讯科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 腾讯科技（深圳）有限公司 filed Critical 腾讯科技（深圳）有限公司
Priority to BR112014026846-0A priority Critical patent/BR112014026846B1/pt
Publication of WO2013163915A1 publication Critical patent/WO2013163915A1/zh
Priority to US14/449,793 priority patent/US9524441B2/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31—User authentication
    • G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/167—Detection; Localisation; Normalisation using comparisons between temporally consecutive images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G06V40/169—Holistic features and representations, i.e. based on the facial image taken as a whole
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/40—Spoof detection, e.g. liveness detection

Definitions

  • the present invention relates to identity authentication technologies, and in particular, to a face identity authentication method and system, and a computer storage medium.
  • Biometric authentication is a means of using human biometrics such as sound, fingerprints, irises, and signatures for identity authentication. As an accurate, fast and efficient authentication method, biometric authentication is being used more and more widely in various fields requiring identity authentication. Among them, face authentication is one of the widely used authentication technologies.
  • a standard face image is collected, and feature data is extracted from the standard face image and stored in a database of the authentication server.
  • the identity authentication is performed, the feature in the currently input face image is extracted, and compared with the feature data stored in the database in advance. If the matching is performed, the authentication is passed, otherwise the authentication fails.
  • the face identity authentication method When the face identity authentication method is applied to the Internet domain, since the user side of the Internet and the authentication server are not in a real real environment, the authentication server cannot distinguish whether the authenticated object is a real person, and the third party can often pass a fake photo or has The recorded video spoofing authentication causes the real user's account to be lost or the virtual property to be lost, which brings great security risks to the user's Internet application.
  • a method for authenticating a face identity includes the following steps: initiating an interactive authentication request, the interaction authentication request includes a requested interaction track, receiving face interaction information sent by the authentication client, and extracting a face according to the face interaction information An activity track; determining whether the face activity track is consistent with the requested interaction track.
  • the interaction authentication request includes a requested face movement interaction trajectory; and the step of determining whether the face activity trajectory is consistent with the requested interaction trajectory comprises: determining a face in the face activity trajectory Whether the movement trajectory coincides with the requested face movement interaction trajectory.
  • the interaction authentication request includes a requested face rotation interaction trajectory; and the step of determining whether the face activity trajectory is consistent with the requested interaction trajectory comprises: determining a face rotation in the face activity trajectory Whether the trajectory is consistent with the requested face rotation interaction trajectory.
  • the step of determining whether the face activity track is consistent with the requested interaction track comprises: determining whether the face motion track includes a face rotation track.
  • the method further includes: the authentication client generates an interaction authentication operation indication according to the interaction authentication request, and displays the interaction authentication operation indication.
  • the method further includes the step of authenticating the client to obtain facial interaction information, the step comprising: tracking a facial motion trajectory performed by the user according to the interactive authentication operation indication; according to the facial motion trajectory Collecting a face image; extracting a face image collected within a preset time or a face video recorded within a preset time according to the interactive authentication request.
  • the method before the step of initiating the interactive authentication request, further includes: receiving a first facial image acquired by the authentication client, and extracting a facial feature in the first facial image; The face feature is compared with the pre-stored standard feature data, and if the face feature is consistent with the standard feature data, the step of initiating the interactive authentication request is entered.
  • the method further includes: receiving a second face image when the user ends the face interaction operation acquired by the authentication client, and extracting a face feature in the second face image; The face feature in the two-face picture is compared with the pre-stored standard feature data, and it is determined whether the face feature in the second face picture is consistent with the standard feature data.
  • the method before the step of receiving the first face image acquired by the authentication client, the method further includes: receiving a user registration request; collecting a standard face image; and extracting the standard face image Standard feature data is stored and stored.
  • the interaction authentication request further includes a face interaction operation time
  • the method further includes: acquiring a time when the user starts the face interaction operation until the user ends the face interaction operation, and determining the acquired time Whether during the face interaction operation time.
  • a face identity authentication system includes an authentication server and an authentication client, the authentication server includes an interaction authentication module, and the interaction authentication module is configured to initiate an interaction authentication request, where the interaction authentication request includes the requested interaction track;
  • the authentication client is configured to obtain facial interaction information and send the information to the authentication server.
  • the interaction authentication module is further configured to extract a facial activity trajectory according to the facial interaction information, and determine whether the facial activity trajectory interacts with the request. The trajectory is consistent.
  • the interaction authentication request includes a requested face movement interaction track; the interaction authentication module is configured to determine whether the face movement track in the face activity track and the requested face movement interaction track Consistent.
  • the interaction authentication request includes a requested face rotation interaction track
  • the interaction authentication module is configured to determine whether the face rotation track in the face activity track and the requested face rotation interaction track Consistent.
  • the interaction authentication module is configured to determine whether a face rotation track is included in the face activity track.
  • the authentication client is further configured to generate an interaction authentication operation indication according to the interaction authentication request, and display the interaction authentication operation indication.
  • the authentication client includes: a face detection tracking module, configured to track a face motion trajectory performed by the user according to the interaction authentication operation instruction; and an information collection module, configured to perform motion according to the face The trajectory collects a face image; the interaction information extraction module is configured to extract a face image collected within a preset time or a face video recorded within a preset time according to the interaction authentication request.
  • the authentication server further includes: a feature extraction module, configured to receive a first face image acquired by the authentication client, and extract a face feature in the first face image; and the feature verification module And comparing the face feature with the pre-stored standard feature data, and if the face feature is consistent with the standard feature data, notifying the interaction authentication module to initiate an interaction authentication request.
  • a feature extraction module configured to receive a first face image acquired by the authentication client, and extract a face feature in the first face image
  • the feature verification module And comparing the face feature with the pre-stored standard feature data, and if the face feature is consistent with the standard feature data, notifying the interaction authentication module to initiate an interaction authentication request.
  • the feature extraction module is further configured to: receive a second face image when the user ends the face interaction operation acquired by the authentication client, and extract a face feature in the second face image;
  • the feature verification module is further configured to compare the face feature in the second face image with pre-stored standard feature data, and determine whether the face feature in the second face image and the standard feature are The data is consistent.
  • the authentication client is further configured to receive a user registration request, and collect a standard face image;
  • the feature extraction module is further configured to extract standard feature data in the standard face image;
  • the authentication The server also includes a feature database for storing the standard feature data.
  • the interaction authentication request further includes a face interaction operation time
  • the interaction authentication module is further configured to acquire a time when the user starts the face interaction operation until the user ends the face interaction operation, and determines the acquisition. Whether the time is within the face interaction time.
  • An authentication server comprising a storage medium and at least one processor, the storage medium storing one or more programs, the one or more programs being executed by the at least one processor, the one or more programs include:
  • a first module configured to initiate an interaction authentication request, where the interaction authentication request includes a requested interaction trajectory
  • a second module configured to receive face interaction information sent by the authentication client
  • a third module configured to extract a facial activity trajectory according to the facial interaction information
  • the fourth module is configured to determine whether the face activity track is consistent with the requested interaction track.
  • the above-mentioned face authentication method and system, and a computer storage medium acquire facial interaction information when initiating an interactive authentication request, and extract a facial activity trajectory according to the face interaction information, if the extracted facial activity trajectory is consistent with the requested interaction trajectory , the certification is passed, otherwise the certification does not pass.
  • the face needs to generate a certain activity track to pass the authentication, and the face provided by the third party or the recorded video is often static and fixed, and cannot be based on the interactive authentication request. Different activity trajectories are generated. Therefore, the above-described face authentication method and system are authenticated with respect to the conventional only face images, which can prevent fraudulent authentication and ensure the authenticity of the identity passed by the authentication, thereby improving each The security of an Internet application.
  • FIG. 1 is a schematic flow chart of a method for authenticating a face identity in an embodiment
  • FIG. 2 is a schematic view of a human face turning track in an embodiment
  • FIG. 3 is a schematic structural diagram of a face identity authentication system in an embodiment
  • FIG. 4 is a schematic diagram of an authentication client in an embodiment
  • FIG. 5 is a schematic diagram of an authentication server in an embodiment
  • FIG. 6 is a schematic diagram of an application of a face identity authentication system in an embodiment.
  • FIG. 8 is a timing diagram of face identity authentication in an embodiment
  • a face identity authentication method runs on an authentication server, and the authentication server is based on a computer system.
  • the method includes the following steps:
  • Step S102 Initiating an interaction authentication request, where the interaction authentication request includes the requested interaction track.
  • Step S104 Receive face interaction information sent by the authentication client.
  • Step S106 extracting a face activity track according to the face interaction information.
  • step S108 it is determined whether the face activity track is consistent with the requested interaction track, and if yes, the process goes to step S110, otherwise the process goes to step S112.
  • step S110 the authentication is passed.
  • step S112 the authentication fails.
  • the interaction authentication request may be initiated when the application operation request is received, where the application operation request may be a user login request, a payment request, a download request, and the like.
  • the interaction authentication request includes interaction request information, and the interaction trajectory of the request may be randomly specified in the interaction request information, including the rotation, movement, and movement of the target area of the face.
  • the face activity track is feature data indicating a track generated by physical movement of the face, and may include a face rotation track and a face movement track, which are feature data indicating face rotation and face movement, respectively.
  • determining whether the face activity track is consistent with the requested interaction track does not mean that the face activity track must be exactly the same as the requested interaction track, but when the face activity track and the request are If the similarity of the interaction trajectory is within a preset range, it is determined that the trajectory of the human face is consistent with the requested interaction trajectory.
  • the “consistent” in this embodiment and the following embodiments is “consistent” in the general technical sense in the field of image recognition.
  • the face activity may be performed according to the requested interaction track.
  • the face activity track may be extracted from the face interaction information, if the face activity track and the requested face are If the interaction trajectory is consistent, the authentication is passed, otherwise the authentication fails. Since the face activity trajectory is dynamic, compared with the conventional only face image for authentication, the face identity authentication method can prevent fraudulent authentication and ensure the authenticity of the identity passed by the authentication, thereby improving each The security of an Internet application.
  • the authentication client may receive an interactive authentication request sent by the authentication server and obtain the input face interaction information. Before obtaining the input face interaction information, the authentication client may generate an interaction authentication operation indication according to the interaction authentication request, and display the interaction authentication operation indication. In this way, the user performs a corresponding face interaction operation according to the interactive authentication operation instruction. For example, the interactive authentication operation instruction prompts the user to move the face to the left to the specified target area or the like.
  • the authentication client when the user performs the corresponding face interaction operation according to the interaction authentication operation instruction, the authentication client obtains the face interaction information, and the step may include: the authentication client tracking user performs the interaction authentication operation instruction.
  • the face motion track, the face image is collected according to the face motion track, and the interaction information is extracted according to the interaction authentication request, and the interaction information may be a face image collected within a preset time or a face recorded in a preset time. video.
  • the face motion trajectory is a physical trajectory formed by the user moving the face in the real space according to the interactive authentication operation knowledge, and is a moving process of the face. The moving process can be captured in a certain time to collect the movement process. A series of face pictures.
  • the face needs to be moved to the left and rotated, and finally the face falls into the specified target area.
  • the face picture may be collected according to a preset time interval (such as 1S), and the face motion track is tracked during the face interaction operation.
  • the interaction information generated during the face interaction operation may be extracted, which may be multiple face images collected within a preset time, or may be a face recorded within a preset time period.
  • Video where each frame in the face video is a face picture taken every 1 second.
  • the interactive authentication request further includes a face interaction operation time.
  • the time when the user starts the face interaction operation to the time when the user ends the face interaction operation can be obtained, and whether the acquired time is within the face interaction operation time, and if yes, the authentication is passed, and the face interaction track can be further performed. Judge, otherwise, the certification does not pass.
  • the time for the face interaction operation is specified in the interactive authentication operation instruction, for example, the movement of the face to the specified target area is completed in 10 seconds, and the extracted interaction information is collected within the specified time. Face image or recorded face video.
  • the time for the face interaction operation is specified in the mutual authentication request, that is, the user is required to complete the face interaction operation within the specified time, and the situation of authentication by video spoofing can be further avoided.
  • each face image in the face image or the face video may be analyzed, and the face activity is extracted therefrom.
  • the inter-authentication request includes the requested face movement interaction trajectory
  • the interactive authentication request includes the requested face rotation track, it may be determined whether the face rotation track in the face activity track is consistent with the requested face rotation interaction track, and if so, the authentication is passed. Otherwise the certification does not pass.
  • the interactive authentication request includes the requested face movement interaction trajectory and the requested face rotation trajectory, and then it can be determined whether the face movement trajectory is consistent with the requested face movement interaction trajectory, and the face rotation is determined. Whether the trajectory is consistent with the requested face rotation interaction trajectory. If both are consistent, the authentication passes, otherwise the authentication fails.
  • the real face may show signs of rotation during the movement
  • the face motion track is tracked and the range of the face is recognized, as shown in FIG. 2, the outline of the face is square, and the vertical axis is a face.
  • the center line, the horizontal axis is the left and right range of the face, when the face is rotated, the center line will be shifted left and right.
  • the face rotation track may be an offset track of the center line extracted according to the face interaction information. When the center line is shifted left and right, it can be judged that there is an indication of the face turning.
  • the face movement track may be a moving path of the face extracted according to the face interaction information.
  • the first facial image acquired by the authentication client may be received, and the facial feature in the first facial image is extracted; the facial feature and the pre-stored standard feature are extracted. The data is compared. If the face feature is consistent with the standard feature data, the authentication is passed, and the step of initiating the interactive authentication request is entered, otherwise the authentication fails.
  • the user registration request may be received before the step of acquiring the first face picture; the standard face picture is collected, preferably the face front image; the standard feature data in the standard face picture is extracted and stored. Standard feature data is used to verify face images.
  • the authentication request is initiated when the application operation request of the user is received, and when the authentication request is received, the current face image of the user, that is, the first facial image is collected.
  • the face feature in the first face image is extracted and compared with the standard feature data stored at the time of registration. If the two are consistent, the authentication passes and the process of interactive authentication is entered, otherwise the authentication fails.
  • the second face image when the user ends the face interaction operation acquired by the authentication client may be further received, and the face feature in the second face image is extracted; the person in the second face image is The face feature is compared with the pre-stored standard feature data. If the face feature in the second face image is consistent with the standard feature data, the authentication passes, otherwise the authentication fails.
  • the user's face image is collected for authentication, which ensures the authenticity of the user who performs the interactive authentication, and further prevents fraudulent authentication, thereby further improving security.
  • a computer readable storage medium storing one or more programs, the one or more programs being used by one or more processors to perform a
  • the specific steps of the face authentication method, the computer executable instructions in the computer storage medium, and the method for performing the face identity authentication are described in the above method, and are not described herein again.
  • an authentication server comprising a storage medium and at least one processor having one or more programs stored therein, the one or more programs being at least one processor Execution, the one or more programs include:
  • the first module is configured to initiate an interaction authentication request, where the interaction authentication request includes the requested interaction track.
  • the second module is configured to receive the face interaction information sent by the authentication client.
  • the third module is configured to extract a face activity track according to the face interaction information.
  • the fourth module is configured to determine whether the face activity track is consistent with the requested interaction track.
  • first module, the second module, the third module, and the fourth module in this embodiment are program modules corresponding to the foregoing steps S102 to S112, and implementation details thereof are not described herein again.
  • a face identity authentication system includes an authentication server 100 and an authentication client 200.
  • the authentication server 100 includes an interaction authentication module 102, and the interaction authentication module 102 is configured to initiate an interaction authentication.
  • the request, the interactive authentication request includes the requested interaction track; the authentication client 200 is configured to obtain the face interaction information and send it to the authentication server 100.
  • the interaction authentication module 102 is further configured to extract the face according to the face interaction information.
  • the activity track determines whether the face activity track is consistent with the requested interaction track. If yes, the authentication passes, otherwise the authentication fails.
  • the interaction authentication module 102 is configured to initiate an interaction authentication request when receiving an application operation request, where the application operation request may be a user login request, a payment request, a download request, and the like.
  • the interaction authentication request includes interaction request information, and the interaction trajectory of the request may be randomly specified in the interaction request information, including the rotation, movement, and movement of the target area of the face.
  • the interaction authentication module 102 determines whether the face activity trajectory is consistent with the requested interaction trajectory, and does not mean that the face activity trajectory must be exactly the same as the requested interaction trajectory, and the face activity trajectory extracted by the interaction authentication module 102 and If the similarity of the requested interaction trajectory is within a preset range, it is determined that the face activity trajectory is consistent with the requested interaction trajectory.
  • the user may perform the face activity according to the requested interaction track after receiving the interaction authentication request, and the authentication client 200 obtains the face interaction information by tracking the face image, and the interaction authentication module 102 interacts from the face.
  • the face activity track is extracted from the information. If the face activity track is consistent with the requested interaction track, the authentication passes, otherwise the authentication fails. Since the face activity trajectory is dynamic, compared with the conventional only face image for authentication, the face identity authentication method can prevent fraudulent authentication and ensure the authenticity of the identity passed by the authentication, thereby improving each The security of an Internet application.
  • the authentication client 200 is further configured to generate an interaction authentication operation indication according to the interaction authentication request, and display the interaction authentication operation indication.
  • the authentication client 200 includes a face detection tracking module 202, an information collection module 204, and an interaction information extraction module 206, wherein:
  • the face detection tracking module 202 is configured to track a face motion trajectory performed by the user according to the interaction authentication operation instruction; the information collection module 204 is configured to collect a face image according to the face motion trajectory; and the interaction information extraction module 206 is configured to extract according to the interaction authentication request.
  • the interaction information may be a face image collected within a preset time or a face video recorded within a preset time.
  • the interactive authentication request further includes a face interaction operation time.
  • the interaction authentication module 102 is further configured to obtain a time when the user starts the face interaction operation until the user ends the face interaction operation, and determines whether the acquired time is within the face interaction operation time. If yes, the authentication is passed, and further The judgment of the face interaction trajectory is performed, otherwise, the authentication does not pass.
  • the time of the face interaction operation is specified in the interaction authentication operation instruction, for example, the movement of the face to the left is completed in 10 seconds, and the interaction information extracted by the interaction information extraction module 206 is A face image or a recorded face video captured during the specified time.
  • the time for the face interaction operation is specified in the mutual authentication request, that is, the user is required to complete the face interaction operation within the specified time, and the situation of authentication by video spoofing can be further avoided.
  • the plurality of face images or the recorded face videos extracted by the interaction information extraction module 206 are submitted to the authentication server 100, and the interaction authentication module 102 in the authentication server 100 analyzes multiple face images or Each frame of the face image in the face video extracts the face activity track, including the face movement track and the face rotation track.
  • the interaction authentication request includes the requested face movement interaction trajectory; the interaction authentication module 102 is configured to determine whether the face movement trajectory in the face activity trajectory is consistent with the requested face movement interaction trajectory, and if so, the authentication Pass, otherwise the certification will not pass.
  • the interaction authentication request includes the requested face rotation interaction track; the interaction authentication module 102 is configured to determine whether the face rotation track in the face activity track is consistent with the requested face rotation interaction track, and if so, The certification is passed, otherwise the certification will not pass.
  • the interaction authentication request may include the requested face movement track and the requested face rotation interaction track
  • the interaction authentication module 102 is configured to determine whether the face movement track in the face activity track is related to the requested person. If the face movement interaction trajectory is consistent and the face rotation trajectory in the broken face activity trajectory is consistent with the requested face rotation interaction trajectory, if both are consistent, the authentication passes, otherwise the authentication fails.
  • the interaction authentication module 102 is further configured to determine whether the face rotation track is included in the face activity track, and if yes, the authentication is passed, otherwise the authentication fails. Since the real face will show signs of rotation during the movement, it is simple and reliable to realize the identity of the face by directly judging whether there is any sign of face rotation, and can ensure the authenticity of the identity passed by the authentication.
  • the face detection tracking module 202 is configured to track the face motion track and identify the range of the face, and the face rotation track may be based on the face interaction.
  • the offset trajectory of the centerline extracted from the information. When the center line is shifted left and right, it can be judged that there is an indication of the face turning.
  • the face movement track may be a moving path of the face extracted according to the face interaction information.
  • the authentication server 100 can further include a feature extraction module 104, a feature verification module 106, and a feature database 108, wherein:
  • the feature extraction module 104 is configured to receive the first face image acquired by the authentication client 200, and extract the face feature in the first face image; the feature verification module 106 is configured to perform the face feature and the pre-stored standard feature data. In contrast, if the face feature is consistent with the standard feature data, the authentication is passed, and the interactive authentication module 102 is notified to initiate an interactive authentication request, otherwise the authentication fails.
  • the authentication client 200 is configured to receive a user registration request and collect a standard face picture.
  • the feature extraction module 106 is further configured to extract standard feature data in a standard face picture.
  • the feature database 108 is used to store standard feature data.
  • the feature extraction module 104 is further configured to receive a second face image when the user ends the face interaction operation acquired by the authentication client 200, and extract a face feature in the second face image.
  • the feature verification module 106 is further configured to compare the face feature in the second face image with the standard feature data pre-stored in the feature database 108, if the face feature in the second face image is consistent with the standard feature data, The certification is passed, otherwise the certification will not pass.
  • the authentication server includes a face authentication server and an identity authentication server, and the identity authentication server is provided with an identity authentication module.
  • Face authentication includes two processes of registered face authentication and online face authentication. among them:
  • the specific process of registering face authentication is:
  • the user submits the registration face authentication request by using the PC or the mobile device, and the authentication client starts the video device driver by the information collection module after receiving the registration face authentication request, and is photographed by a video device such as a camera carried by the PC or the mobile device.
  • the photo is tracked by the face detection and tracking module to identify the face motion range, and the face range is recognized.
  • the information collection module acquires the face image and submits it to the face authentication server.
  • the feature extraction module extracts the face feature from the face image, and the face feature is standard feature data and is stored in the feature database.
  • the face authentication server establishes registration information, including a correspondence between the user identifier and the standard feature data, and submits the registration information to the identity authentication server.
  • the identity authentication server returns the registration result to the authentication client, for example, prompting the registration to succeed or fail, and the authentication client displays the registration result to the user.
  • the specific process of online face authentication includes:
  • the information collection module of the authentication client starts the video device driver, and the video device such as the camera carried by the PC or the mobile device The photo is taken, and the face motion tracking track is tracked by the face detection tracking module to identify the face range.
  • the authentication client obtains the first face image, and the authentication client submits the first face image to the face authentication server.
  • the feature extraction module in the face authentication server extracts the face feature in the first face image, compares the face feature with the pre-stored standard feature data, and if the two are consistent, the authentication passes, and the The authentication result (such as the authentication result a in FIG. 4) is submitted to the identity authentication server, and a face interaction authentication request is initiated to the authentication client.
  • the authentication client receives the face interaction authentication request, generates a face interaction authentication operation instruction, and performs the presentation, and the user can perform the face interaction operation under the instruction.
  • the information collection module in the authentication client records the face interaction process, including tracking the face motion track, and collecting the face image every preset time (such as 1S).
  • the interaction information extraction module in the authentication client extracts the interaction information after receiving the confirmation operation of the user (the face image collected within the preset time or recorded in the preset time)
  • the face video, the preset time may be the time specified in the face interaction authentication request, and obtain the second face image when the user ends the face interaction operation, and submit the interaction information and the second face image to the face authentication. server.
  • the interaction authentication module in the face authentication server extracts the face activity trajectory according to the interaction information, and determines whether the interaction trajectory is consistent with the request. If the two are consistent, the authentication is passed, and the authentication result is obtained (the authentication result in FIG. 4) b) Submit to the authentication server.
  • the feature extraction module in the face authentication server extracts the face feature in the second face image, and the feature verification module compares the face feature with the pre-stored standard feature data. If the two are consistent, the authentication is passed.
  • the authentication result (the authentication result c in Fig. 4) is submitted to the identity authentication server.
  • the identity authentication server determines that the user identity authentication fails. Finally, the identity authentication server returns the identity authentication result to the PC, mobile device, etc. used by the user.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Human Computer Interaction (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Security & Cryptography (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Computer Hardware Design (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Collating Specific Patterns (AREA)

Abstract

一种人脸身份认证方法，包括以下步骤：发起交互认证请求，所述交互认证请求中包括请求的交互轨迹；获取人脸交互信息；根据所述人脸交互信息提取人脸活动轨迹；判断所述人脸活动轨迹是否与请求的交互轨迹一致，若是，则认证通过，否则认证不通过。通过从人脸交互信息去提取人脸活动轨迹，根据人脸活动轨迹进行判断，实现将人脸身份认证中的动态密码取代了传统的静态密码，能够有效防止欺骗认证，从而能够确保认证通过的身份的真实性。此外，还提供了一种人脸身份认证***和计算机存储介质。

Description

人脸身份认证方法和***、计算机存储介质
【技术领域】
本发明涉及身份认证技术，特别是涉及一种人脸身份认证方法和***、计算机存储介质。
【背景技术】
生物特征认证是利用人的生物特征，如声音、指纹、虹膜及签名等，来进行身份认证的一种手段。生物特征认证作为一种准确、快速和高效的身份认证方法正被越来越广泛地应用于各种需要身份认证的领域，其中，人脸身份认证就是其中一种应用广泛的认证技术。
传统的人脸身份认证方法中，需采集标准人脸图片，从标准人脸图片中提取特征数据并存储在认证服务器的数据库中。在进行身份认证时，提取当前输入的人脸图片中的特征，与事先存储在数据库中的特征数据进行对比，若相符，则认证通过，否则认证不通过。
这种人脸身份认证方法应用到互联网领域时，由于互联网的用户侧与认证服务器不处于现实真实环境，认证服务器无法辨别认证的对象是否是真实的人物，第三方往往可通过虚假的照片或已录制的视频欺骗认证，从而导致真实用户的帐号丢失或虚拟财产丢失，给用户的互联网应用带来极大的安全隐患。
【发明内容】
基于此，有必要针对上述易被第三方欺骗认证的问题，提供一种能防止欺骗认证、确保身份真实性的人脸身份认证方法和***、计算机存储介质。
一种人脸身份认证方法，包括以下步骤：发起交互认证请求，所述交互认证请求中包括请求的交互轨迹；接收认证客户端发送的人脸交互信息；根据所述人脸交互信息提取人脸活动轨迹；判断所述人脸活动轨迹是否与请求的交互轨迹一致。
在其中一个实施例中，所述交互认证请求包含请求的人脸移动交互轨迹；所述判断人脸活动轨迹是否与请求的交互轨迹一致的步骤包括：判断所述人脸活动轨迹中的人脸移动轨迹是否与请求的人脸移动交互轨迹一致。
在其中一个实施例中，所述交互认证请求包含请求的人脸转动交互轨迹；所述判断人脸活动轨迹是否与请求的交互轨迹一致步骤包括：判断所述人脸活动轨迹中的人脸转动轨迹是否与请求的人脸转动交互轨迹一致。
在其中一个实施例中，所述判断人脸活动轨迹是否与请求的交互轨迹一致的步骤包括：判断所述人脸活动轨迹中是否包含人脸转动轨迹。
在其中一个实施例中，所述方法还包括：认证客户端根据交互认证请求生成交互认证操作指示，展示所述交互认证操作指示。
在其中一个实施例中，所述方法还包括认证客户端获取人脸交互信息的步骤，该步骤包括：跟踪用户根据所述交互认证操作指示进行的人脸运动轨迹；根据所述人脸运动轨迹采集人脸图片；根据交互认证请求提取预设时间内采集到的人脸图片或预设时间内录制的人脸视频。
在其中一个实施例中，所述发起交互认证请求的步骤之前，还包括：接收认证客户端获取到的第一人脸图片，提取所述第一人脸图片中的人脸特征；将所述人脸特征与预先存储的标准特征数据进行对比，如果所述人脸特征与所述标准特征数据一致，则进入所述发起交互认证请求的步骤。
在其中一个实施例中，所述方法还包括：接收认证客户端获取到的用户结束人脸交互操作时的第二人脸图片，提取第二人脸图片中的人脸特征；将所述第二人脸图片中的人脸特征与预先存储的标准特征数据进行对比，判断所述第二人脸图片中的人脸特征是否与所述标准特征数据一致。
在其中一个实施例中，所述接收认证客户端获取到的第一人脸图片的步骤之前，该方法还包括：接收用户注册请求；采集标准人脸图片；提取所述标准人脸图片中的标准特征数据并进行存储。
在其中一个实施例中，所述交互认证请求中还包括人脸交互操作时间，所述方法还包括：获取用户开始人脸交互操作到用户结束人脸交互操作的时间，判断所述获取的时间是否在所述人脸交互操作时间内。
一种人脸身份认证***，包括认证服务器和认证客户端，所述认证服务器包括交互认证模块，所述交互认证模块用于发起交互认证请求，所述交互认证请求中包含请求的交互轨迹；所述认证客户端用于获取人脸交互信息并发送给所述认证服务器；所述交互认证模块还用于根据人脸交互信息提取人脸活动轨迹，判断所述人脸活动轨迹是否与请求的交互轨迹一致。
在其中一个实施例中，所述交互认证请求包含请求的人脸移动交互轨迹；所述交互认证模块用于判断所述人脸活动轨迹中的人脸移动轨迹是否与请求的人脸移动交互轨迹一致。
在其中一个实施例中，所述交互认证请求包含请求的人脸转动交互轨迹，所述交互认证模块用于判断所述人脸活动轨迹中的人脸转动轨迹是否与请求的人脸转动交互轨迹一致。
在其中一个实施例中，所述交互认证模块用于判断所述人脸活动轨迹中是否包含人脸转动轨迹。
在其中一个实施例中，所述认证客户端还用于根据交互认证请求生成交互认证操作指示，展示所述交互认证操作指示。
在其中一个实施例中，所述认证客户端包括：人脸检测跟踪模块，用于跟踪用户根据所述交互认证操作指示进行的人脸运动轨迹；信息采集模块，用于根据所述人脸运动轨迹采集人脸图片；交互信息提取模块，用于根据交互认证请求提取预设时间内采集到的人脸图片或预设时间内录制的人脸视频。
在其中一个实施例中，所述认证服务器还包括：特征提取模块，用于接收认证客户端获取到的第一人脸图片，提取所述第一人脸图片中的人脸特征；特征验证模块，用于将所述人脸特征与预先存储的标准特征数据进行对比，如果所述人脸特征与所述标准特征数据一致，则通知所述交互认证模块发起交互认证请求。
在其中一个实施例中，所述特征提取模块还用于接收认证客户端获取到的用户结束人脸交互操作时的第二人脸图片，提取所述第二人脸图片中的人脸特征；所述特征验证模块还用于将所述第二人脸图片中的人脸特征与预先存储的标准特征数据进行对比，判断所述第二人脸图片中的人脸特征是否与所述标准特征数据一致。
在其中一个实施例中，所述认证客户端还用于接收用户注册请求，采集标准人脸图片；所述特征提取模块还用于提取所述标准人脸图片中的标准特征数据；所述认证服务器还包括用于存储所述标准特征数据的特征数据库。
在其中一个实施例中，所述交互认证请求中还包括人脸交互操作时间，所述交互认证模块还用于获取用户开始人脸交互操作到用户结束人脸交互操作的时间，判断所述获取的时间是否在所述人脸交互操作时间内。
一种认证服务器，包括存储介质和至少一个处理器，所述存储介质中存储有一个或一个以上的程序，该一个或一个以上程序被所述至少一个处理器执行，该一个或一个以上的程序包括：
第一模块，用于发起交互认证请求，所述交互认证请求中包括请求的交互轨迹；
第二模块，用于接收认证客户端发送的人脸交互信息；
第三模块，用于根据所述人脸交互信息提取人脸活动轨迹；
第四模块，用于判断所述人脸活动轨迹是否与请求的交互轨迹一致。
一种用于存储计算机可执行指令的计算机存储介质，所述计算机可执行指令用于控制计算机执行人脸身份认证方法，其特征在于，所述方法包括：
发起交互认证请求，所述交互认证请求中包括请求的交互轨迹；
获取人脸交互信息；
根据所述人脸交互信息提取人脸活动轨迹；
判断所述人脸活动轨迹是否与请求的交互轨迹一致。
上述人脸身份认证方法和***、计算机存储介质，在发起交互认证请求时，获取人脸交互信息，根据人脸交互信息提取人脸活动轨迹，如果提取的人脸活动轨迹与请求的交互轨迹一致，则认证通过，否则认证不通过。这样，在交互认证的过程中，人脸需产生一定的活动轨迹才能通过认证，而第三方采用的照片或者录制的视频所提供的人脸往往是静态且固定不变的，不能根据交互认证请求产生不同的活动轨迹，因此上述人脸身份认证方法和***，相对于传统的仅采用静态的人脸图片来进行认证，能够防止欺骗认证，能够确保认证通过的身份的真实性，从而能够提高各种互联网应用的安全性。
【附图说明】
图1为一个实施例中人脸身份认证方法的流程示意图；
图2为一个实施例中人脸转动轨迹的示意图；
图3为一个实施例中人脸身份认证***的结构示意图；
图4为一个实施例中认证客户端的示意图；
图5为一个实施例中认证服务器的示意图；
图6为一个实施例中人脸身份认证***的应用示意图。
图7为一个实施例中注册人脸认证的时序图；
图8为一个实施例中人脸身份认证的时序图；
【具体实施方式】
如图1所示，在一个实施例中，一种人脸身份认证方法，运行于认证服务器，认证服务器基于计算机***，该方法包括以下步骤：
步骤S102，发起交互认证请求，交互认证请求中包括请求的交互轨迹。
步骤S104，接收认证客户端发送的人脸交互信息。
步骤S106，根据人脸交互信息提取人脸活动轨迹。
步骤S108，判断人脸活动轨迹是否与请求的交互轨迹一致，若是，则进入步骤S110，否则进入步骤S112。
步骤S110，认证通过。
步骤S112，认证不通过。
具体的，可在接收到应用操作请求时发起交互认证请求，该应用操作请求可以是用户登录请求、支付请求、下载请求等。交互认证请求中包含交互请求信息，在交互请求信息中可随机指定请求的交互轨迹，包括人脸的转动、移动和移动的目标区域等。人脸活动轨迹为表示人脸部进行物理运动所产生的轨迹的特征数据，可包括人脸转动轨迹和人脸移动轨迹，分别为表示脸部转动和脸部移动的特征数据。
应当说明的是，在图像识别领域中，判断人脸活动轨迹是否与请求的交互轨迹一致，并不是指人脸活动轨迹必须与请求的交互轨迹完全相同，而是当人脸活动轨迹与请求的交互轨迹的相似度在预设的范围内，则确定为人脸活动轨迹与请求的交互轨迹一致。本实施例及下述实施例中的“一致”即为图像识别领域中一般技术意义上的“一致”。
本实施例中，可在收到交互认证请求后按照请求的交互轨迹进行人脸活动，通过跟踪人脸图片，可从人脸交互信息中提取人脸活动轨迹，如果人脸活动轨迹与请求的交互轨迹一致，则认证通过，否则认证不通过。由于人脸活动轨迹是动态的，相对于传统的仅采用静态的人脸图片来进行认证，这种人脸身份认证方法能够防止欺骗认证，能够确保认证通过的身份的真实性，从而能够提高各种互联网应用的安全性。在一个实施例中，认证客户端可接收认证服务器发送的交互认证请求，并获取输入的人脸交互信息。且认证客户端在获取输入的人脸交互信息之前，可根据交互认证请求生成交互认证操作指示，并展示该交互认证操作指示。这样，用户根据交互认证操作指示进行相应的人脸交互操作。例如，交互认证操作指示中提示用户需将人脸向左移动到指定的目标区域等。
进一步的，在一个实施例中，用户根据交互认证操作指示进行相应的人脸交互操作过程中，认证客户端获取人脸交互信息，该步骤可包括：认证客户端跟踪用户根据交互认证操作指示进行的人脸运动轨迹，再根据人脸运动轨迹采集人脸图片，以及根据交互认证请求提取交互信息，该交互信息可以是预设时间内采集到的人脸图片或预设时间内录制的人脸视频。人脸运动轨迹是用户根据交互认证操作知识移动脸部在真实空间中形成的物理轨迹，是人脸的移动过程，在一定时间内对该移动过程进行拍摄即可采集到体现这一移动过程的一系列的人脸图片。
例如，交互认证操作指示中显示需将人脸左移，并转动，最后将人脸落入到指定的目标区域内。在用户完成这一系列的人脸交互操作过程中，可按照预先设定的时间间隔（如1S）采集人脸图片，并在该人脸交互操作过程中，跟踪人脸运动轨迹。在接收到用户的确认完成操作时，则提取该人脸交互操作过程中产生的交互信息，可以是预设时间内采集到的多张人脸图片，也可以是预设时间内录制的人脸视频，其中人脸视频中的每一帧为每1秒钟采集的人脸图片。
进一步的，在一个实施例中，交互认证请求中还包括人脸交互操作时间。本实施例中，可获取用户开始人脸交互操作到用户结束人脸交互操作的时间，判断获取的时间是否在人脸交互操作时间内，若是，则认证通过，可进一步进行人脸交互轨迹的判断，否则，认证不通过。
本实施例中，在交互认证操作指示中规定人脸交互操作的时间，例如规定在10秒完成将人脸左移到指定的目标区域等，则提取的交互信息为在该规定时间内所采集到的人脸图片或录制的人脸视频。在交互认证请求中规定人脸交互操作的时间，即要求用户在规定时间内完成人脸交互操作，可进一步避免通过视频欺骗认证的情形。
进一步的，在获取到预设时间内的多张人脸图片或录制的人脸视频后，可分析多张人脸图片或人脸视频中的每一帧人脸图片，从中提取出人脸活动轨迹，包括人脸移动轨迹和人脸转动轨迹。
在一个实施例中，交互认证请求中包含请求的人脸移动交互轨迹，则可判断人脸活动轨迹中的人脸移动轨迹是否与请求的人脸移动交互轨迹一致，如是，则认证通过，否则认证不通过。
进一步的，在一个实施例中，交互认证请求中包含请求的人脸转动轨迹，则可判断人脸活动轨迹中的人脸转动轨迹是否与请求的人脸转动交互轨迹一致，若是，则认证通过，否则认证不通过。
在其他实施例中，交互认证请求中同时包含请求的人脸移动交互轨迹和请求的人脸转动轨迹，则可判断人脸移动轨迹是否与请求的人脸移动交互轨迹一致，以及判断人脸转动轨迹是否与请求的人脸转动交互轨迹一致，若两者都一致，则认证通过，否则认证不通过。
由于真实的人脸在移动过程中都会存在转动的迹象，进一步的，在一个实施例中，可判断人脸活动轨迹中是否包含人脸转动轨迹，若是，则认证通过，否则认证不通过。通过直接判断是否存在人脸转动迹象，实现起来简单可靠，且能够确保认证通过的身份的真实性。
具体的，在用户根据交互认证操作指示进行人脸交互操作的过程中，跟踪人脸运动轨迹并识别出人脸的范围，如图2所示，人脸的轮廓为方形，竖轴为人脸的中心线，横轴为人脸的左右范围，当人脸转动时，中心线会左右偏移。则人脸转动轨迹可以是根据人脸交互信息提取出的中心线的偏移轨迹。当中心线发生左右偏移时，可判断得到存在人脸转动迹象。而人脸移动轨迹可以是根据人脸交互信息提取出的人脸的移动路径。
在一个实施例中，在发起交互认证请求之前，可接收认证客户端获取到的第一人脸图片，提取第一人脸图片中的人脸特征；将该人脸特征与预先存储的标准特征数据进行对比，如果该人脸特征与标准特征数据一致，则认证通过，进入发起交互认证请求的步骤，否则认证不通过。
在一个实施例中，可在获取第一人脸图片的步骤之前，接收用户注册请求；采集标准人脸图片，优选为人脸正面图像；提取该标准人脸图片中的标准特征数据并进行存储。标准特征数据用于对人脸图片进行验证。
具体的，可在接收到用户的应用操作请求时发起认证请求，在接收到认证请求时，采集用户当前的人脸图片，即第一人脸图片。提取第一人脸图片中的人脸特征，与注册时存储的标准特征数据对比，如果两者一致，则认证通过，进入交互认证的流程，否则认证不通过。
在一个实施例中，可进一步接收认证客户端获取到的用户结束人脸交互操作时的第二人脸图片，提取第二人脸图片中的人脸特征；将第二人脸图片中的人脸特征与预先存储的标准特征数据进行对比，如果第二人脸图片中的人脸特征与标准特征数据一致，则认证通过，否则认证不通过。
本实施例中，在交互认证之前和交互认证结束后，分别采集用户的人脸图片进行认证，能够确保进行交互认证的用户的真实性，进一步防止欺骗认证，从而进一步提高安全性。
此外，在一个实施例中，还提供了一种计算机可读存储介质，该计算机可读存储介质存储有一个或者一个以上程序，该一个或者一个以上程序被一个或者一个以上的处理器用来执行一种人脸身份认证方法，计算机存储介质中的计算机可执行指令执行人脸身份认证方法的具体步骤如上述方法描述，在此不再赘述。
此外，在一个实施例中，还提供了一种认证服务器，包括存储介质和至少一个处理器，该存储介质中存储有一个或一个以上的程序，该一个或一个以上的程序被至少一个处理器执行，该一个或一个以上的程序包括：
第一模块，用于发起交互认证请求，交互认证请求中包括请求的交互轨迹。
第二模块，用于接收认证客户端发送的人脸交互信息。
第三模块，用于根据人脸交互信息提取人脸活动轨迹。
第四模块，用于判断人脸活动轨迹是否与请求的交互轨迹一致。
需要说明的是，本实施例中的第一模块、第二模块、第三模块和第四模块为与前述的步骤S102至步骤S112对应的程序模块，其实施细节在此不再赘述。
如图3所示，在一个实施例中，一种人脸身份认证***，包括认证服务器100和认证客户端200，其中，认证服务器100包括交互认证模块102，交互认证模块102用于发起交互认证请求，交互认证请求中包括请求的交互轨迹；认证客户端200用于获取人脸交互信息并发送给认证服务器100；本实施例中，交互认证模块102还用于根据人脸交互信息提取人脸活动轨迹，判断人脸活动轨迹是否与请求的交互轨迹一致，若是，则认证通过，否则认证不通过。
具体的，交互认证模块102可用于在接收到应用操作请求时发起交互认证请求，该应用操作请求可以是用户登录请求、支付请求、下载请求等。交互认证请求中包含交互请求信息，在交互请求信息中可随机指定请求的交互轨迹，包括人脸的转动、移动和移动的目标区域等。
应当说明的是，交互认证模块102判断人脸活动轨迹是否与请求的交互轨迹一致，并不是指人脸活动轨迹必须与请求的交互轨迹完全相同，当交互认证模块102提取的人脸活动轨迹与请求的交互轨迹的相似度在预设的范围内，则确定为人脸活动轨迹与请求的交互轨迹一致。
本实施例中，用户可在收到交互认证请求后按照请求的交互轨迹进行人脸活动，认证客户端200通过跟踪人脸图片，获取人脸交互信息中，交互认证模块102则从人脸交互信息中提取人脸活动轨迹，如果人脸活动轨迹与请求的交互轨迹一致，则认证通过，否则认证不通过。由于人脸活动轨迹是动态的，相对于传统的仅采用静态的人脸图片来进行认证，这种人脸身份认证方法能够防止欺骗认证，能够确保认证通过的身份的真实性，从而能够提高各种互联网应用的安全性。
在一个实施例中，认证客户端200还用于根据交互认证请求生成交互认证操作指示，展示该交互认证操作指示。如图4所示，在一个实施例中，认证客户端200包括人脸检测跟踪模块202、信息采集模块204和交互信息提取模块206，其中：
人脸检测跟踪模块202用于跟踪用户根据交互认证操作指示进行的人脸运动轨迹；信息采集模块204用于根据人脸运动轨迹采集人脸图片；交互信息提取模块206用于根据交互认证请求提取预设时间内采集到的人脸图片或预设时间内录制的人脸视频。其中，交互信息可以是预设时间内采集到的人脸图片或预设时间内录制的人脸视频。
进一步的，在一个实施例中，交互认证请求中还包括人脸交互操作时间。本实施例中，交互认证模块102还用于获取用户开始人脸交互操作到用户结束人脸交互操作的时间，判断获取的时间是否在人脸交互操作时间内，若是，则认证通过，可进一步进行人脸交互轨迹的判断，否则，认证不通过。
本实施例中，在交互认证操作指示中规定人脸交互操作的时间，例如规定在10S秒完成将人脸左移到指定的目标区域等，则交互信息提取模块206提取的交互信息为在该规定时间内所采集到的人脸图片或录制的人脸视频。在交互认证请求中规定人脸交互操作的时间，即要求用户在规定时间内完成人脸交互操作，可进一步避免通过视频欺骗认证的情形。
进一步的，交互信息提取模块206提取到的预设时间内的多张人脸图片或录制的人脸视频提交至认证服务器100，由认证服务器100中的交互认证模块102分析多张人脸图片或人脸视频中的每一帧人脸图片，从中提取出人脸活动轨迹，包括人脸移动轨迹和人脸转动轨迹。
在一个实施例中，交互认证请求包含请求的人脸移动交互轨迹；交互认证模块102用于判断人脸活动轨迹中的人脸移动轨迹是否与请求的人脸移动交互轨迹一致，若是，则认证通过，否则认证不通过。
在另一个实施例中，交互认证请求包含请求的人脸转动交互轨迹；交互认证模块102用于判断人脸活动轨迹中的人脸转动轨迹是否与请求的人脸转动交互轨迹一致，若是，则认证通过，否则认证不通过。
在其他实施例中，交互认证请求可同时包含请求的人脸移动轨迹和请求的人脸转动交互轨迹，则交互认证模块102用于判断人脸活动轨迹中的人脸移动轨迹是否与请求的人脸移动交互轨迹一致以及断人脸活动轨迹中的人脸转动轨迹是否与请求的人脸转动交互轨迹一致，两者都一致时，则认证通过，否则认证不通过。
在另一个实施例中，交互认证模块102还用于判断人脸活动轨迹中是否包含人脸转动轨迹，若是，则认证通过，否则认证不通过。由于真实的人脸在移动过程中都会存在转动的迹象，通过直接判断是否存在人脸转动迹象，实现起来简单可靠，且能够确保认证通过的身份的真实性。
具体的，在用户根据交互认证操作指示进行人脸交互操作的过程中，人脸检测跟踪模块202用于跟踪人脸运动轨迹并识别出人脸的范围，人脸转动轨迹可以是根据人脸交互信息提取出的中心线的偏移轨迹。当中心线发生左右偏移时，可判断得到存在人脸转动迹象。而人脸移动轨迹可以是根据人脸交互信息提取出的人脸的移动路径。
如图5所示，在一个实施例中，认证服务器100还可包括特征提取模块104、特征验证模块106和特征数据库108，其中：
特征提取模块104用于接收认证客户端200获取到的第一人脸图片，提取第一人脸图片中的人脸特征；特征验证模块106用于将人脸特征与预先存储的标准特征数据进行对比，如果人脸特征与标准特征数据一致，则认证通过，通知交互认证模块102发起交互认证请求，否则认证不通过。
在一个实施例中，认证客户端200用于接收用户注册请求，采集标准人脸图片。特征提取模块106还用于提取标准人脸图片中的标准特征数据。本实施例中，特征数据库108用于存储标准特征数据。
在一个实施例中，特征提取模块104还用于接收认证客户端200获取到的用户结束人脸交互操作时的第二人脸图片，提取第二人脸图片中的人脸特征。特征验证模块106还用于将第二人脸图片中的人脸特征与特征数据库108中预先存储的标准特征数据进行对比，如果第二人脸图片中的人脸特征与标准特征数据一致，则认证通过，否则认证不通过。
下面结合一个具体的应用场景来阐述上述人脸身份认证方法和***的原理。在一个实施例中，认证服务器包括人脸认证服务器和身份认证服务器，身份认证服务器中设有身份认证模块。人脸身份认证包括注册人脸认证和在线人脸认证两个过程。其中：
在一个实施例中，如图6和7所示，注册人脸认证的具体过程为：
（1）用户使用PC或移动设备提交注册人脸认证请求，认证客户端收到该注册人脸认证请求后由信息采集模块启动视频设备驱动，由PC或移动设备带有的摄像头等视频设备拍摄照片，并由人脸检测跟踪模块跟踪人脸运动轨迹，识别出人脸范围，待用户确认人脸范围后信息采集模块获取人脸图片，并提交至人脸认证服务器。
（2）人脸认证服务器收到人脸图片后，特征提取模块从人脸图片中提取人脸特征，该人脸特征为标准特征数据，存储在特征数据库中。人脸认证服务器建立注册信息，包括用户标识与标准特征数据之间的对应关系，并将该注册信息提交至身份认证服务器。
（3）身份认证服务器向认证客户端返回注册结果，例如提示注册成功或失败，认证客户端向用户展示注册结果。
在一个实施例中，如图6和图8所示，在线人脸认证的具体过程包括：
（1）当用户使用PC或移动设备进行应用操作时（如登录应用、完成支付、请求下载）等，认证客户端的信息采集模块启动视频设备驱动，由PC或移动设备带有的摄像头等视频设备拍摄照片，并由人脸检测跟踪模块跟踪人脸运动轨迹，识别出人脸范围。在接收到用户确认人脸范围的操作后，认证客户端获取第一人脸图片，认证客户端将第一人脸图片提交至人脸认证服务器。
（2）人脸认证服务器中的特征提取模块提取第一人脸图片中的人脸特征，将该人脸特征与预先存储的标准特征数据进行对比，若两者一致，则认证通过，将该认证结果（如图4中的认证结果a）提交至身份认证服务器，并向认证客户端发起人脸交互认证请求。
（3）认证客户端接收到人脸交互认证请求，生成人脸交互认证操作指示，并进行展示，则用户可在该指示下进行人脸交互操作。在用户按照指示进行人脸交互操作的过程中，认证客户端中的信息采集模块记录人脸交互过程，包括跟踪人脸运动轨迹、每隔预设时间（如1S）采集人脸图片等。用户完成人脸交互操作后确认操作完成，则认证客户端中的交互信息提取模块在接收到用户的确认操作后提取交互信息（预设时间内采集到的人脸图片或预设时间内录制的人脸视频，该预设时间可为人脸交互认证请求中规定的时间），并获取用户结束人脸交互操作时的第二人脸图片，将交互信息和第二人脸图片提交至人脸认证服务器。
（4）人脸认证服务器中的交互认证模块根据交互信息提取人脸活动轨迹，判断与请求的交互轨迹是否一致，如果两者一致，则认证通过，将该认证结果（图4中的认证结果b）提交至身份认证服务器。人脸认证服务器中的特征提取模块再提取第二人脸图片中的人脸特征，特征验证模块将该人脸特征与预先存储的标准特征数据进行对比，如果两者一致，则认证通过，将该认证结果（图4中的认证结果c）提交至身份认证服务器。
（5）身份认证服务器接收到的认证结果a、认证结果b和认证结果c都为认证通过时，判定用户身份认证通过，允许用户进行应用操作。如果其中有一个认证结果不通过，则身份认证服务器判定用户身份认证不通过。最后身份认证服务器将身份认证结果返回至用户所使用的PC、移动设备等。
以上所述实施例仅表达了本发明的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对本发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进，这些都属于本发明的保护范围。因此，本发明专利的保护范围应以所附权利要求为准。

Claims (22)

  1. 一种人脸身份认证方法，运行于认证服务器，该方法包括以下步骤：
    发起交互认证请求，所述交互认证请求中包括请求的交互轨迹；
    接收认证客户端发送的人脸交互信息；
    根据所述人脸交互信息提取人脸活动轨迹；
    判断所述人脸活动轨迹是否与请求的交互轨迹一致。
  2. 根据权利要求1所述的人脸身份认证方法，其特征在于，所述交互认证请求包含请求的人脸移动交互轨迹；所述判断人脸活动轨迹是否与请求的交互轨迹一致的步骤包括：
    判断所述人脸活动轨迹中的人脸移动轨迹是否与请求的人脸移动交互轨迹一致。
  3. 根据权利要求1或2所述的人脸身份认证方法，其特征在于，所述交互认证请求包含请求的人脸转动交互轨迹；所述判断人脸活动轨迹是否与请求的交互轨迹一致步骤包括：
    判断所述人脸活动轨迹中的人脸转动轨迹是否与请求的人脸转动交互轨迹一致。
  4. 根据权利要求1所述的人脸身份认证方法，其特征在于，所述判断人脸活动轨迹是否与请求的交互轨迹一致步骤包括：
    判断所述人脸活动轨迹中是否包含人脸转动轨迹。
  5. 根据权利要求1所述的人脸身份认证方法，其特征在于，所述方法还包括认证客户端根据交互认证请求生成交互认证操作指示，展示所述交互认证操作指示。
  6. 根据权利要求5所述的人脸身份认证方法，其特征在于，所述方法还包括认证客户端获取人脸交互信息的步骤，且该步骤包括：
    跟踪用户根据所述交互认证操作指示进行的人脸运动轨迹；
    根据所述人脸运动轨迹采集人脸图片；
    根据交互认证请求提取预设时间内采集到的人脸图片或预设时间内录制的人脸视频。
  7. 根据权利要求1所述的人脸身份认证方法，其特征在于，所述发起交互认证请求的步骤之前，还包括：
    接收认证客户端获取到的第一人脸图片，提取所述第一人脸图片中的人脸特征；
    将所述人脸特征与预先存储的标准特征数据进行对比，如果所述人脸特征与所述标准特征数据一致，则进入所述发起交互认证请求的步骤。
  8. 根据权利要求1或7所述的人脸身份认证方法，其特征在于，所述方法还包括：
    接收认证客户端获取到的用户结束人脸交互操作时的第二人脸图片，提取第二人脸图片中的人脸特征；
    将所述第二人脸图片中的人脸特征与预先存储的标准特征数据进行对比，判断所述第二人脸图片中的人脸特征是否与所述标准特征数据一致。
  9. 根据权利要求8所述的人脸身份认证方法，其特征在于，所述接收认证客户端获取到的第一人脸图片的步骤之前，该方法还包括：
    接收用户注册请求；
    采集标准人脸图片；
    提取所述标准人脸图片中的标准特征数据并进行存储。
  10. 根据权利要求1所述的人脸身份认证方法，其特征在于，所述交互认证请求中还包括人脸交互操作时间，所述方法还包括：
    获取用户开始人脸交互操作到用户结束人脸交互操作的时间，判断所述获取的时间是否在所述人脸交互操作时间内。
  11. 一种人脸身份认证***，其特征在于，包括认证服务器和认证客户端，所述认证服务器包括交互认证模块，所述交互认证模块用于发起交互认证请求，所述交互认证请求中包括请求的交互轨迹；
    所述认证客户端用于获取人脸交互信息并发送给所述认证服务器；
    所述交互认证模块还用于根据人脸交互信息提取人脸活动轨迹，判断所述人脸活动轨迹是否与请求的交互轨迹一致。
  12. 根据权利要求11所述的人脸身份认证***，其特征在于，所述交互认证请求包含请求的人脸移动交互轨迹；所述交互认证模块用于判断所述人脸活动轨迹中的人脸移动轨迹是否与请求的人脸移动交互轨迹一致。
  13. 根据权利要求11或12所述的人脸身份认证***，其特征在于，所述交互认证请求包含请求的人脸转动交互轨迹，所述交互认证模块用于判断所述人脸活动轨迹中的人脸转动轨迹是否与请求的人脸转动交互轨迹一致。
  14. 根据权利要求11所述的人脸身份认证***，其特征在于，所述交互认证模块用于判断所述人脸活动轨迹中是否包含人脸转动轨迹。
  15. 根据权利要求11所述的人脸身份认证***，其特征在于，所述认证客户端模块还用于根据交互认证请求生成交互认证操作指示，展示所述交互认证操作指示。
  16. 根据权利要求15所述的人脸身份认证***，其特征在于，所述认证客户端模块包括：
    人脸检测跟踪模块，用于跟踪用户根据所述交互认证操作指示进行的人脸运动轨迹；
    信息采集模块，用于根据所述人脸运动轨迹采集人脸图片；
    交互信息提取模块，用于根据交互认证请求提取预设时间内采集到的人脸图片或预设时间内录制的人脸视频。
  17. 根据权利要求11所述的人脸身份认证***，其特征在于，所述认证服务器还包括：
    特征提取模块，用于接收认证客户端获取到的第一人脸图片，提取所述第一人脸图片中的人脸特征；
    特征验证模块，用于将所述人脸特征与预先存储的标准特征数据进行对比，如果所述人脸特征与所述标准特征数据一致，则通知所述交互认证模块发起交互认证请求。
  18. 根据权利要求17所述的人脸身份认证***，其特征在于，所述特征提取模块还用于接收认证客户端获取到的用户结束人脸交互操作时的第二人脸图片，提取所述第二人脸图片中的人脸特征；所述特征验证模块还用于将所述第二人脸图片中的人脸特征与预先存储的标准特征数据进行对比，判断所述第二人脸图片中的人脸特征是否与所述标准特征数据一致。
  19. 根据权利要求18所述的人脸身份认证***，其特征在于，所述认证客户端模块还用于接收用户注册请求，采集标准人脸图片；所述特征提取模块还用于提取所述标准人脸图片中的标准特征数据；所述认证服务器还包括用于存储所述标准特征数据的特征数据库。
  20. 根据权利要求11所述的人脸身份认证***，其特征在于，所述交互认证请求中还包括人脸交互操作时间，所述交互认证模块还用于获取用户开始人脸交互操作到用户结束人脸交互操作的时间，判断所述获取的时间是否在所述人脸交互操作时间内。
  21. 一种认证服务器，包括存储介质和至少一个处理器，所述存储介质中存储有一个或一个以上的程序，该一个或一个以上程序被至少一个处理器执行，该一个或一个以上的程序包括：
    第一模块，用于发起交互认证请求，所述交互认证请求中包括请求的交互轨迹；
    第二模块，用于接收认证客户端发送的人脸交互信息；
    第三模块，用于根据所述人脸交互信息提取人脸活动轨迹；
    第四模块，用于判断所述人脸活动轨迹是否与请求的交互轨迹一致。
  22. 一种计算机可读存储介质，所述计算机可读存储介质存储有一个或者一个以上程序，所述一个或者一个以上程序被一个或者一个以上的处理器用来执行一种人脸身份认证方法，其特征在于，该计算机可读存储介质设置于认证服务器上，且所述方法包括：
    发起交互认证请求，所述交互认证请求中包括请求的交互轨迹；
    接收认证客户端发送的人脸交互信息；
    根据所述人脸交互信息提取人脸活动轨迹；
    判断所述人脸活动轨迹是否与请求的交互轨迹一致。
PCT/CN2013/073807 2012-05-04 2013-04-07 人脸身份认证方法和***、计算机存储介质 WO2013163915A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
BR112014026846-0A BR112014026846B1 (pt) 2012-05-04 2013-04-07 Sistema e método para autenticação de identidade baseados em reconhecimento de rosto
US14/449,793 US9524441B2 (en) 2012-05-04 2014-08-01 System and method for identity authentication based on face recognition, and computer storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201210137251.5A CN103384234B (zh) 2012-05-04 2012-05-04 人脸身份认证方法和***
CN201210137251.5 2012-05-04

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US14/449,793 Continuation US9524441B2 (en) 2012-05-04 2014-08-01 System and method for identity authentication based on face recognition, and computer storage medium

Publications (1)

Publication Number Publication Date
WO2013163915A1 true WO2013163915A1 (zh) 2013-11-07

Family

ID=49491931

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/073807 WO2013163915A1 (zh) 2012-05-04 2013-04-07 人脸身份认证方法和***、计算机存储介质

Country Status (4)

Country Link
US (1) US9524441B2 (zh)
CN (1) CN103384234B (zh)
BR (1) BR112014026846B1 (zh)
WO (1) WO2013163915A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108639012A (zh) * 2018-05-14 2018-10-12 冼汉生 基于物联网的供电***启动方法、装置及计算机存储介质
CN116152936A (zh) * 2023-02-17 2023-05-23 深圳市永腾翼科技有限公司 一种带交互式活体检测的人脸身份认证***及其方法

Families Citing this family (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104901801B (zh) * 2014-03-06 2019-01-11 腾讯科技（深圳）有限公司 身份验证方法及装置
JP6369543B2 (ja) * 2014-06-19 2018-08-08 日本電気株式会社 認証装置、認証システム、認証方法およびコンピュータプログラム
CN105227310A (zh) * 2014-06-24 2016-01-06 阿里巴巴集团控股有限公司 一种身份认证方法和设备
CN106156578B (zh) * 2015-04-22 2020-02-14 深圳市腾讯计算机***有限公司 身份验证方法和装置
CN104794465B (zh) * 2015-05-13 2019-06-07 上海依图网络科技有限公司 一种基于姿态信息的活体检测方法
CN106302330B (zh) 2015-05-21 2021-01-05 腾讯科技（深圳）有限公司 身份验证方法、装置和***
CN105100097A (zh) * 2015-07-17 2015-11-25 百度在线网络技术（北京）有限公司 身份认证方法和装置
CN105407098A (zh) * 2015-11-26 2016-03-16 小米科技有限责任公司 身份验证方法及装置
CN105426730A (zh) * 2015-12-28 2016-03-23 小米科技有限责任公司 登录验证处理方法、装置及终端设备
CN106886698B (zh) * 2016-06-08 2021-01-29 创新先进技术有限公司 一种认证方法及装置
CN107786487B (zh) * 2016-08-24 2021-02-02 腾讯科技（深圳）有限公司 一种信息认证处理方法、***以及相关设备
CN106372591B (zh) * 2016-08-30 2019-05-07 湖南强视信息科技有限公司 一种面向无人监考的防止虚拟摄像头作弊***
CN107992728B (zh) * 2016-10-27 2022-05-20 腾讯科技（深圳）有限公司 人脸验证方法及装置
CN106992968B (zh) * 2017-03-03 2020-05-19 浙江智贝信息科技有限公司 一种基于客户端的人脸持续认证方法
CN106713368B (zh) * 2017-03-08 2019-09-27 四川大学 一种身份验证方法及装置
CN106937039B (zh) * 2017-04-26 2020-08-11 安徽龙运智能科技有限公司 一种基于双摄像头的成像方法、移动终端及存储介质
CN107403165B (zh) * 2017-08-01 2021-12-24 杰创智能科技股份有限公司 智能人脸识别***的数据管理架构及使用方法
CN108647576A (zh) * 2018-04-10 2018-10-12 阳光暖果（北京）科技发展有限公司 一种基于视频序列人脸识别的活体身份认证方法
CN109408651B (zh) * 2018-09-21 2020-09-29 神思电子技术股份有限公司 一种基于人脸脸势识别的人脸检索方法
CN111476189B (zh) * 2020-04-14 2023-10-13 北京爱笔科技有限公司 一种身份识别方法和相关装置
CN113810667B (zh) * 2021-11-19 2022-02-08 成都旺小宝科技有限公司 基于视频回溯模式的渠道风控***

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050044143A1 (en) * 2003-08-19 2005-02-24 Logitech Europe S.A. Instant messenger presence and identity management
CN101114909A (zh) * 2007-08-17 2008-01-30 搜图科技（南京）有限公司 全自动视频身份认证***及方法
CN101216887A (zh) * 2008-01-04 2008-07-09 浙江大学 照片人脸与活体人脸的计算机自动鉴别方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1403997A (zh) * 2001-09-07 2003-03-19 昆明利普机器视觉工程有限公司 数字视频人脸自动识别***
CN1284111C (zh) * 2003-12-15 2006-11-08 中国科学院自动化研究所 移动计算环境下分布式的人脸检测与识别方法
KR101242390B1 (ko) * 2011-12-29 2013-03-12 인텔 코오퍼레이션 사용자를 인증하기 위한 방법, 장치, 및 컴퓨터 판독 가능한 기록 매체

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050044143A1 (en) * 2003-08-19 2005-02-24 Logitech Europe S.A. Instant messenger presence and identity management
CN101114909A (zh) * 2007-08-17 2008-01-30 搜图科技（南京）有限公司 全自动视频身份认证***及方法
CN101216887A (zh) * 2008-01-04 2008-07-09 浙江大学 照片人脸与活体人脸的计算机自动鉴别方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108639012A (zh) * 2018-05-14 2018-10-12 冼汉生 基于物联网的供电***启动方法、装置及计算机存储介质
CN116152936A (zh) * 2023-02-17 2023-05-23 深圳市永腾翼科技有限公司 一种带交互式活体检测的人脸身份认证***及其方法

Also Published As

Publication number Publication date
CN103384234B (zh) 2016-09-28
US20140341445A1 (en) 2014-11-20
BR112014026846A2 (pt) 2017-06-27
US9524441B2 (en) 2016-12-20
BR112014026846B1 (pt) 2021-09-21
CN103384234A (zh) 2013-11-06

Similar Documents

Publication Publication Date Title
WO2013163915A1 (zh) 人脸身份认证方法和***、计算机存储介质
US10135818B2 (en) User biological feature authentication method and system
EP3005202B1 (en) System and method for biometric authentication with device attestation
WO2022019693A1 (ko) 사용자 비대면 인증 vc(verifiable credential)를 이용한 did(decentralized identifier) 기반의 사용자 비대면 인증 방법 및 이를 이용한 인증 지원 서버
TW201629851A (zh) 一種影像識別系統及方法
WO2013100699A1 (en) Method, apparatus, and computer-readable recording medium for authenticating a user
US20120086785A1 (en) Apparatus for identifying protecting, requesting, assisting and managing information
US20160371533A1 (en) Method for recognizing operation body's characteristic information, electronic apparatus, safety apparatus and palm print recognizing device
EP3001343B1 (en) System and method of enhanced identity recognition incorporating random actions
WO2022166090A1 (zh) 自助打印方法、装置、设备、存储介质及程序产品
CN101976356A (zh) 网吧实名制人脸识别方法及识别***
CN102054202A (zh) 一种基于指纹的考生信息管理***及其控制方法
WO2018058833A1 (zh) 验证方法、验证装置、终端和服务器
WO2012018173A2 (ko) 무인대출 처리방법
WO2021060670A1 (ko) 보안 카드를 이용한 사용자 인증 장치 및 사용자 인증 방법
KR20130133676A (ko) 카메라를 통한 얼굴인식을 이용한 사용자 인증 방법 및 장치
TWI338866B (en) Face recognition system, security system comprising same and method for operating same
WO2018179723A1 (ja) 顔認証処理装置、顔認証処理方法及び顔認証処理システム
KR101751175B1 (ko) 온라인 신분증 확인, 안면 인증 및 음성 인증을 활용한 생체정보 등록 방법 및 시스템
WO2022010022A1 (ko) 비대면 본인인증 고객확인 장치 및 그 방법
TWI621074B (zh) 巡邏簽到系統及其方法
CN111611569A (zh) 一种人脸声纹复核终端及其身份认证方法
WO2023074919A1 (ko) 탈중앙화 분산 데이터 저장 방식을 이용한 무인 주차 시스템 및 이를 운영하는 방법
CN115688071B (zh) 一种防止智能手表信息篡改的处理方法及***
CN112115454B (zh) 单点登录方法、第一服务器及电子设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13784272

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

REG Reference to national code

Ref country code: BR

Ref legal event code: B01A

Ref document number: 112014026846

Country of ref document: BR

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 09/04/15)

122 Ep: pct application non-entry in european phase

Ref document number: 13784272

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 112014026846

Country of ref document: BR

Kind code of ref document: A2

Effective date: 20141027