WO2013000142A1 - 手机用户身份认证方法、云服务器以及网络系统 - Google Patents

手机用户身份认证方法、云服务器以及网络系统 Download PDF

Info

Publication number
WO2013000142A1
WO2013000142A1 PCT/CN2011/076623 CN2011076623W WO2013000142A1 WO 2013000142 A1 WO2013000142 A1 WO 2013000142A1 CN 2011076623 W CN2011076623 W CN 2011076623W WO 2013000142 A1 WO2013000142 A1 WO 2013000142A1
Authority
WO
WIPO (PCT)
Prior art keywords
face
sample
characteristic
user
sample image
Prior art date
Application number
PCT/CN2011/076623
Other languages
English (en)
French (fr)
Inventor
高东璇
Original Assignee
深圳市君盛惠创科技有限公司
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 深圳市君盛惠创科技有限公司 filed Critical 深圳市君盛惠创科技有限公司
Priority to PCT/CN2011/076623 priority Critical patent/WO2013000142A1/zh
Priority to US14/129,135 priority patent/US8861798B2/en
Priority to CN201180071166.7A priority patent/CN103814545B/zh
Publication of WO2013000142A1 publication Critical patent/WO2013000142A1/zh
Priority to US14/486,136 priority patent/US8989452B2/en
Priority to US14/486,112 priority patent/US8983145B2/en
Priority to US14/622,110 priority patent/US9537859B2/en
Priority to US15/355,268 priority patent/US9813909B2/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W12/00Security arrangements; Authentication; Protecting privacy or anonymity
    • H04W12/06Authentication
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/22Matching criteria, e.g. proximity measures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31User authentication
    • G06F21/32User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31User authentication
    • G06F21/40User authentication by quorum, i.e. whereby two or more security principals are required
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/44Program or device authentication
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/94Hardware or software architectures specially adapted for image or video understanding
    • G06V10/95Hardware or software architectures specially adapted for image or video understanding structured as a network, e.g. client-server architectures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16Human faces, e.g. facial parts, sketches or expressions
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161Detection; Localisation; Normalisation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168Feature extraction; Face representation
    • G06V40/171Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172Classification, e.g. identification
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00Network architectures or network communication protocols for network security
    • H04L63/08Network architectures or network communication protocols for network security for authentication of entities
    • H04L63/083Network architectures or network communication protocols for network security for authentication of entities using passwords
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00Network architectures or network communication protocols for network security
    • H04L63/08Network architectures or network communication protocols for network security for authentication of entities
    • H04L63/0861Network architectures or network communication protocols for network security for authentication of entities using biometrical features, e.g. fingerprint, retina-scan
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00Network architectures or network communication protocols for network security
    • H04L63/10Network architectures or network communication protocols for network security for controlling access to devices or network resources
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L2463/00Additional details relating to network architectures or network communication protocols for network security covered by H04L63/00
    • H04L2463/082Additional details relating to network architectures or network communication protocols for network security covered by H04L63/00 applying multi-factor authentication

Definitions

  • the present invention relates to the field of communications technologies, and in particular, to a mobile phone user identity authentication method, a cloud server, and a network system.
  • the embodiment of the invention provides a mobile phone user identity authentication method, a cloud server and a network system, which can bear the load of the identity authentication by the cloud server, improve the security of the mobile phone operating system, enhance the user experience, and improve the accuracy of the face verification.
  • An embodiment of the present invention provides a mobile phone user identity authentication method, where the mobile phone is connected to a cloud server through a communication network, and the cloud server stores a face sample image library corresponding to the user; the method includes:
  • the user enters a login account and password on the mobile phone
  • the login account and password are in error, the user is denied access to the mobile phone operating system; if the login account and password are correct, the login account and password are sent to the cloud server; the login account and password correspond to the cloud server. a face sample image library of the user stored therein;
  • the mobile phone camera acquires a face input image of the user, and sends the face input image to the cloud server;
  • the cloud server authenticates the user according to the login account and password and the face input image, and determines Whether to allow the user to enter the mobile phone operating system, specifically includes:
  • Step A The cloud server determines, according to the login account and the password, a face sample image library of the user corresponding to the login account and the password;
  • Step B obtaining a facial feature similarity value according to the face input image and the face sample image library; Step B includes:
  • Step C determining whether the facial feature similarity value is greater than a preset threshold, wherein the preset threshold is obtained according to a plurality of first characteristic distances between each facial sample image in the face sample image library;
  • Step D If the facial feature similarity value is not greater than the preset threshold, the user is allowed to enter the mobile phone operating system
  • Step E If the facial feature similarity value is greater than the preset threshold, calculating a first quantity and a second quantity, where the first quantity is a person corresponding to the first characteristic distance that is greater than the similarity value of the facial feature a number of face sample images in the face sample image library, the second number being a number of face sample images in the face sample image library corresponding to the first characteristic distance not greater than the face feature similarity value, and Determining whether the first quantity is greater than the second quantity;
  • Step F If the first quantity is less than the second quantity, rejecting the user to enter the mobile phone operating system; Step G. If the first quantity is not less than the second quantity, allowing the user to enter the mobile phone operating system.
  • An embodiment of the present invention provides a cloud server, including: a storage unit, configured to store a face sample image library of a user; a receiving unit, configured to receive a login account and a password from the user's mobile phone, and a face input image; And determining, according to the login account and the password, a user face sample image library stored in the storage unit corresponding to the login account and the password;
  • a face feature similarity value determining unit configured to obtain a face feature similarity value according to the face input image and the face sample image library;
  • the face feature similarity value determining unit includes a face region image acquiring unit, and a feature a value calculation unit and a characteristic distance calculation unit, wherein:
  • a face region image obtaining unit configured to obtain a face region image from the face input image by face detection
  • a feature value calculation unit configured to calculate each face sample image in the face sample image library a first characteristic value and a second characteristic value of the face region image
  • a characteristic distance calculation unit configured to calculate a characteristic value distance between a first characteristic value of each face sample image in the face sample image library and a second characteristic value of the face region image, to obtain a plurality of second Characteristic distance, and determining the facial feature similarity value according to the plurality of second characteristic distances;
  • a first determining unit configured to determine whether the facial feature similarity value is greater than a preset threshold, wherein the preset threshold is based on multiple first characteristics between each facial sample image in the face sample image library Obtained by distance;
  • a first enabling unit configured to allow the user to enter a mobile phone operating system when the facial feature similarity value is not greater than the preset threshold
  • a second determining unit configured to calculate a first quantity and a second quantity when the facial feature similarity value is greater than the preset threshold, where the first quantity is a first characteristic that is greater than the similarity value of the facial feature a number of face sample images in the corresponding face sample image library, the second number being a face sample image in the face sample image library corresponding to the first characteristic distance not greater than the face feature similarity value a number, and determining whether the first quantity is greater than the second quantity;
  • a rejecting unit configured to reject the user from entering a mobile phone operating system when the first quantity is less than the second quantity; and a first allowing unit, when the first quantity is not less than the second quantity, Allow the user to enter the mobile operating system.
  • An embodiment of the present invention provides a network system, including: a mobile phone and a cloud server, where the mobile phone is connected to the cloud server through a communication network;
  • the mobile phone is configured to receive a login account and a password input by the user, and determine whether the login account and the password are correct; if the login account and the password are in error, the user is denied to enter the mobile phone operating system; if the login account and the password are correct, the mobile phone
  • the login account and the password are sent to the cloud server; the login account and the password correspond to the face sample image library of the user stored in the cloud server; the face input image of the user is obtained, and the face input image is sent to
  • the cloud server is configured to store a face sample image library corresponding to the user; perform identity authentication on the user according to the login account and the password and the face input image, and determine whether the user is allowed to enter the mobile phone.
  • the operating system includes the following steps: Step A: determining, according to the login account and the password, a face sample image library of the user corresponding to the login account and the password; Step B. inputting the image according to the face and the face sample image library , obtaining a facial feature similarity value; wherein step B includes:
  • Step C determining whether the facial feature similarity value is greater than a preset threshold, wherein the preset threshold is obtained according to a plurality of first characteristic distances between each facial sample image in the face sample image library; Step D. If the facial feature similarity value is not greater than the preset threshold, the user is allowed to enter the mobile phone operating system;
  • Step E If the facial feature similarity value is greater than the preset threshold, calculating a first quantity and a second quantity, where the first quantity is a person corresponding to the first characteristic distance that is greater than the similarity value of the facial feature a number of face sample images in the face sample image library, the second number being a number of face sample images in the face sample image library corresponding to the first characteristic distance not greater than the face feature similarity value, and Determining whether the first quantity is greater than the second quantity;
  • Step F If the first quantity is less than the second quantity, rejecting the user to enter the mobile phone operating system; Step G. If the first quantity is not less than the second quantity, allowing the user to enter the mobile phone operating system.
  • the embodiment of the invention can bear the load of the identity authentication by the cloud server, improve the security of the mobile phone operating system, enhance the user experience, and improve the accuracy of the face verification.
  • Embodiment 1 is a flowchart of a method according to Embodiment 1 of the present invention.
  • FIG. 2 is a schematic structural diagram of a cloud server according to Embodiment 2 of the present invention.
  • FIG. 3 is a schematic structural diagram of a face feature similarity determining unit in a cloud server according to Embodiment 2 of the present invention
  • FIG. 4 is a schematic structural diagram of a network system according to Embodiment 3 of the present invention.
  • Cloud computing is an Internet-based computing approach in which shared hardware and software resources and information can be delivered to computers, mobile phones and other devices on demand.
  • Typical cloud computing providers often provide common network business applications that can be accessed through software such as browsers or other web services, while software and data are stored on the cloud server.
  • the embodiment of the invention is based on the cloud computing technology, and the identity authentication task of the mobile phone can be undertaken by the cloud server, thereby reducing the burden on the mobile phone, and introducing a high-cost service on the mobile phone to improve the user experience.
  • the user's mobile phone is connected to the cloud server through the communication network, and the cloud server stores the face sample image library corresponding to the user.
  • the cloud server may be managed by the telecom operator, and the user registers the face sample image when signing the contract. Go to the cloud server of the telecom operator.
  • the cloud server binds the user's mobile phone number, mobile phone operating system login account, and password to the face sample image library.
  • FIG. 1 A flowchart of a method for authenticating a mobile phone user identity according to Embodiment 1 of the present invention is shown in FIG. 1.
  • the method may include the following steps:
  • Step S101 The user inputs a login account and a password on the mobile phone;
  • Step S103 The mobile phone determines whether the login account and password are correct;
  • Step S105 If the login account and password are in error, the user is denied access to the mobile phone operating system, and an error occurs; Step S107. If the login account and password are correct, the login account and password are sent to the cloud server, and the login account and password correspond to the cloud. a face sample image library of the user stored in the server;
  • Step S109 The mobile phone starts the camera, acquires a face input image of the user, and sends the face input image to the cloud server;
  • Step S111 The cloud server authenticates the user according to the login account and the password and the face input image, and determines whether the user is allowed to enter the mobile phone operating system.
  • Step S111 specifically includes:
  • Step S111-2 The cloud server determines, according to the login account and the password, a face sample image library of the user corresponding to the login account and the password;
  • Step S111-4 According to the face input image and the face sample image library, the facial feature similarity value is obtained; the facial feature similarity value is the degree of similarity between the face input image and each face sample image, and the facial feature similarity value The smaller the more similar, the step S111-4 specifically includes:
  • Step S111-4-1 Obtaining a face region image from the face input image by face detection; the face detection method mainly performs face face color region comparison by using the face input image and the face sample image, and The face area image is taken out according to the ratio of the head shape.
  • Step S111-4-3 Calculating a first characteristic value of each face sample image in the face sample image library and a second characteristic value of the face region image;
  • Step S111-4-5 Calculating a feature value distance between a first characteristic value of each face sample image in the face sample image library and a second characteristic value of the face region image, to obtain a plurality of second characteristic distances And determining a facial feature similarity value according to the plurality of second characteristic distances; the facial feature similarity value is a degree of similarity between the facial input image and each facial sample image, and the similarity of the facial feature similarity value is similar.
  • the face feature similarity value may be a maximum of the plurality of second characteristic distances, or may be an average of the plurality of second characteristic distances.
  • Step S111-6 Determine whether the face feature similarity value is greater than a preset threshold, where the preset threshold is obtained according to multiple first characteristic distances between each face sample image in the face sample image library.
  • the preset threshold may be a maximum of the plurality of first characteristic distances, or may be an average of the plurality of first characteristic distances.
  • Step S111-8 If the similarity value of the face feature is not greater than the preset threshold, that is, the similarity of the face sample image in the user face image and the face sample image library meets the requirements, the user is allowed to enter the mobile phone. operating system;
  • Step S111-10 If the similarity value of the face feature is greater than the preset threshold, that is, the similarity between the face image of the user and the face sample image in the face sample image library fails to meet the requirements, how many individuals are counted separately
  • the first part of the face sample image The characteristic distance is greater than or less than the facial feature similarity value, that is, the first quantity and the second quantity are calculated, and the first quantity is a face in the face sample image library corresponding to the first characteristic distance greater than the facial feature similarity value. a number of sample images, the second number being a number of face sample images in the face sample image library corresponding to the first characteristic distance not greater than the face feature similarity value; and then determining the first quantity Whether it is greater than the second quantity;
  • Step S111-12 If the first quantity is less than the second quantity, the user is denied access to the mobile phone operating system; Step S111-14. If the first quantity is not less than the second quantity, the user is allowed to enter the mobile phone operating system.
  • the following is a detailed description of how the embodiment of the present invention extracts facial image features, determines a first characteristic value of each face sample image in the face sample image library, a second characteristic value of the face region image, and a face sample. a first characteristic distance between two pairs of face sample images in the image library, a second characteristic distance of each face sample image in the face sample image library and the face region image, and a preset threshold and a face feature worth it.
  • the face sample image 3 ⁇ 4 _ is a two-dimensional 64 ⁇ 64 gray image representing the abscissa pixel and _ representing the ordinate pixel.
  • the face sample image is superimposed and superimposed on the face position, and the mean value of all the images is overlapped.
  • the size of the covariance matrix C is 4096X4096, and it is very difficult to solve the eigenvalues and eigenvectors directly.
  • each face sample image can be projected to the feature space ft composed of l, 2 , L, and Zhang, and the previous feature value can be selected as the feature space, because the feature space
  • the dimension of the original face sample image is lower than that of the original face sample image. Therefore, after each face sample image is projected onto the feature space formed by u, u 2 , L , the face sample image dimension is greatly reduced. Thereby achieving a reduction in dimensionality and The purpose of extracting features.
  • the embodiment of the present invention proposes to obtain a block image of the face sample.
  • the method of eigenvectors In view of the three distinctive features of the human face: eyes, nose and mouth, and they are respectively in the upper, middle and lower parts of the face, according to these three salient features, the face image is divided into three independent Sub-block - the upper part includes: the eye, the middle includes the nose, and the lower part includes the mouth.
  • the upper sub-images of all face sample image libraries constitute the upper sub-block image library, also the middle and lower parts
  • the sub-image constitutes a middle sub-block image library and a lower sub-block image library. In the process of feature extraction, they will be treated as three separate sub-block image libraries.
  • the embodiment of the present invention proposes an algorithm that can increase the number of samples without sampling, thereby improving the accuracy of feature extraction.
  • the method specifically includes:
  • the feature space of X is the direct sum of the feature space of the first sample x e and the feature space of the second sample x 0. Therefore, the first feature space u e and the feature extraction algorithm can be respectively obtained for x e and x 0 respectively.
  • the second feature space u o is then selected from the first feature space u e and the second feature space u o to select a feature vector with high recognition accuracy and large difference to form the feature space u.
  • the method of the embodiment of the present invention is described in conjunction with the segmented face sample image library.
  • the feature space of each sample X/" and X ( 1, 2, L, ⁇ ) for the middle sub-block image library and the lower sub-block image library
  • each sample X; 1 , X/" and X 1 characteristic values ⁇ ", ? And 7 to obtain an average value, and obtain a first characteristic value ⁇ of each face sample C in the face sample image library ;
  • the face region image is also processed correspondingly, that is, the face region image is segmented, the corresponding characteristic values of each block are respectively calculated, the sum is averaged, and finally the second feature of the face region image is obtained.
  • the value is ⁇ .
  • Embodiments of the present invention provide a method for calculating a characteristic distance—calculating a plurality of first characteristic distances between face sample images according to a first characteristic value of each face sample image in the face sample image library. Specifically include:
  • a plurality of first characteristic distances between the two face sample images are calculated, a total of a first characteristic distance.
  • a preset threshold is obtained according to a first characteristic distance between each face sample image in the face sample image library, and the preset threshold may be ⁇ ( ⁇ - ⁇ / maximum of the first characteristic distances, It can also be ⁇ ( ⁇ - ⁇ /2 the average of the first characteristic distances.
  • the facial feature similarity value is determined according to the second characteristic distance, and the facial feature similarity value may be a maximum value of the second characteristic distances, or may be an average of the second characteristic distances.
  • Step S 111-14 of the embodiment of the present invention further includes: if the first quantity is not less than the second quantity, updating the face sample image library by using the face input image; the update strategy may be an alternative The oldest face sample image, or the face sample image that has the largest difference from the face input image.
  • the first characteristic distance of the face sample image library in the cloud server may be recalculated, and a new preset threshold is determined according to the first characteristic distance, and the new preset threshold is replaced by the preset Set the threshold.
  • the mobile phone user identity authentication method in the embodiment of the present invention can bear the load of the identity authentication by the cloud server, improve the security of the mobile phone operating system, enhance the user experience, and improve the accuracy of the face verification.
  • the embodiment of the present invention further provides a cloud server 100, as shown in FIG. 2, including:
  • the storage unit 200 is configured to store a face sample image library of the user
  • the receiving unit 201 is configured to receive a login account and a password from the user's mobile phone, and a face input image.
  • the determining unit 203 is configured to determine, according to the login account and the password, that the login account and the password are stored in the storage unit 200.
  • User's face sample image library
  • the face feature similarity value determining unit 205 is configured to obtain a face feature similarity value according to the face input image and the face sample image library; as shown in FIG. 3, the face feature similarity determining unit 205 includes The face region image acquiring unit 205-2, the characteristic value calculating unit 205-4, and the characteristic distance calculating unit 205-6, wherein:
  • a face area image obtaining unit 205-2 configured to obtain a face area image from the face input image by face detection
  • the feature value calculation unit 205-4 is configured to calculate a first characteristic value of each face sample image in the face sample image library and a second characteristic value of the face region image;
  • the characteristic distance calculation unit 205-6 is configured to calculate a characteristic value distance between a first characteristic value of each face sample image in the face sample image library and a second characteristic value of the face region image, to obtain a plurality of a second characteristic distance, and determining the facial feature similarity value according to the plurality of second characteristic distances;
  • the first determining unit 207 is configured to determine whether the facial feature similarity value is greater than a preset threshold, where the preset threshold is according to multiple firsts between each facial sample image in the face sample image library. Characteristic distance;
  • the first permission unit 209 is configured to allow the user to enter the mobile phone operating system when the facial feature similarity value is not greater than the preset threshold;
  • the second determining unit 211 is configured to: when the facial feature similarity value is greater than the preset threshold, calculate a first quantity and a second quantity, where the first quantity is a first value that is greater than the similarity value of the facial feature a number of face sample images in the face sample image library corresponding to the feature distance, the second number being a face sample image in the face sample image library corresponding to the first characteristic distance not greater than the face feature similarity value And determining, by the number of the first quantity, whether the first quantity is greater than the second quantity; the rejecting unit 213, configured to reject the user from entering the mobile phone operating system when the first quantity is less than the second quantity;
  • the second permitting unit 215 is configured to allow the user to enter the mobile operating system when the first quantity is not less than the second quantity.
  • the cloud server may further include: a first updating unit 217, configured to perform, by using the face input image, the face sample image library when the first quantity is not less than the second quantity Update.
  • a first updating unit 217 configured to perform, by using the face input image, the face sample image library when the first quantity is not less than the second quantity Update.
  • the cloud server may further include: a second update unit 219, configured to recalculate a first characteristic distance of the face sample image library in the cloud server, and determine a new pre-determination according to the first characteristic distance A threshold is set, and the new preset threshold is replaced by the preset threshold.
  • a second update unit 219 configured to recalculate a first characteristic distance of the face sample image library in the cloud server, and determine a new pre-determination according to the first characteristic distance A threshold is set, and the new preset threshold is replaced by the preset threshold.
  • the characteristic value calculation unit 205-4 includes:
  • a first feature vector calculation unit 205-49 configured to respectively determine an orthogonal normalized feature vector of the first sample covariance matrix and an orthogonal normalized feature vector of the second sample covariance matrix;
  • a first projection calculation unit 205-411 a first feature space composed of orthogonal normalized feature vectors of the first sample covariance matrix, and an orthogonal normalization feature of the second sample covariance matrix a second feature space composed of a vector, determining a projection of the first sample and the second sample in the first feature space and the second feature space, respectively;
  • the first characteristic value determining unit 205-413 configured to The projections of the first sample and the second sample in the first feature space and the second feature space determine characteristic values of X, ", XTM and X t b ; according to characteristics of X, ", ⁇ and a value determining a first characteristic value of the face sample image x t ;
  • a second dividing unit 205-415 configured to divide the face region image into three sub-images;
  • a second generating unit 205-417 configured to respectively generate a corresponding dual sample for the three sub-images
  • a second decomposing unit 205-419 configured to: according to the dual samples corresponding to the three sub-images, the three sub-images Decomposed into a first sample and a second sample, respectively;
  • a second covariance matrix construction unit 205-421 configured to respectively construct a covariance matrix for the first sample and the second sample of the three sub-images
  • a second feature vector calculating unit 205-423 configured to respectively determine an orthogonal normalized feature vector of the first sample covariance matrix and an orthogonal normalized feature vector of the second sample covariance matrix;
  • a second projection calculation unit 205-425 configured to use a feature space composed of orthogonal normalized feature vectors of the first sample covariance matrix, and an orthogonal normalized feature vector of the second sample covariance matrix a feature space, determining a projection of the first sample and the second sample in a feature space;
  • a second characteristic value determining unit 205-427 configured to determine a characteristic value of the three sub-images according to the projection of the first sample and the second sample in a feature space; according to characteristic values of the three sub-images Determining a second characteristic value of the face region image.
  • the embodiments of the present invention provide that the load of the identity authentication can be assumed by the cloud server, improve the security of the mobile phone operating system, enhance the user experience, and improve the accuracy of the face verification.
  • the embodiment of the present invention further provides a network system, including a mobile phone and a cloud server, where the mobile phone is connected to the cloud server through a communication network;
  • the mobile phone is configured to receive a login account and a password input by the user, and determine whether the login account and the password are correct; if the login account and the password are in error, the user is denied to enter the mobile phone operating system; if the login account and the password are correct, the mobile phone
  • the login account and the password are sent to the cloud server; the login account and the password correspond to the face sample image library of the user stored in the cloud server; the face input image of the user is obtained, and the face input image is sent to
  • the cloud server is configured to store a face sample image library corresponding to the user; perform identity authentication on the user according to the login account and the password and the face input image, and determine whether the user is allowed to enter the mobile phone.
  • the operating system includes the following steps: Step A: determining, according to the login account and the password, a face sample image library of the user corresponding to the login account and the password; Step B. inputting the image according to the face and the face sample image library , obtaining a facial feature similarity value; wherein step B includes: Bl. obtaining a face region image from the face input image by face detection;
  • Step C determining whether the facial feature similarity value is greater than a preset threshold, wherein the preset threshold is obtained according to a plurality of first characteristic distances between each facial sample image in the face sample image library;
  • Step D If the facial feature similarity value is not greater than the preset threshold, the user is allowed to enter the mobile phone operating system
  • Step E If the facial feature similarity value is greater than the preset threshold, calculating a first quantity and a second quantity, where the first quantity is a person corresponding to the first characteristic distance that is greater than the similarity value of the facial feature a number of face sample images in the face sample image library, the second number being a number of face sample images in the face sample image library corresponding to the first characteristic distance not greater than the face feature similarity value, and Determining whether the first quantity is greater than the second quantity;
  • Step F If the first quantity is less than the second quantity, rejecting the user to enter the mobile phone operating system; Step G. If the first quantity is not less than the second quantity, allowing the user to enter the mobile phone operating system.
  • the specific structure of the cloud server may be as described in Embodiment 2.
  • the embodiment of the invention can bear the load of the identity authentication by the cloud server, improve the security of the mobile phone operating system, enhance the user experience, and improve the accuracy of the face verification.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read- Only Memory ROM), or a random access memory (RAM).

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Computer Hardware Design (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Multimedia (AREA)
  • Software Systems (AREA)
  • Computing Systems (AREA)
  • Human Computer Interaction (AREA)
  • Biomedical Technology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Collating Specific Patterns (AREA)
  • Image Analysis (AREA)
  • Processing Or Creating Images (AREA)

Abstract

本发明实施例提供一种手机用户身份认证方法、云服务器以及网络系统,所述手机通过通信网络与云服务器连接,所述云服务器中存储用户对应的人脸样本图像库;该方法包括:云服务器根据登录账号和密码,确定所述登录账号和密码对应的用户的人脸样本图像库;根据登录账号和密码以及所述人脸输入图像,对所述用户进行身份认证,判断是否允许所述用户进入手机操作系统,能够将身份认证的负载由云服务器承担,提高手机操作系统的安全性、增强用户体验、提高人脸验证的精确度。

Description

手机用户身份认证方法、 云服务器以及网络系统 技术领域
本发明涉及通信技术领域,尤其涉及一种手机用户身份认证方法、云服务器以及网络系 统。
背景技术
随着手机特别是智能手机的广泛流行,手机操作系统的安全性也变得越来越重要。目前, 大部分智能手机仅仅把用户登录账号和密码作为身份认证的手段。 但这种方法的安全性不 高, 一旦登录账号和密码被他人盗窃, 手机操作系统上所有数据就暴露无遗。
将人体生物特征特别是人脸作为安全认证的技术发展迅速。但是人脸验证的运算复杂度 较高, 而手机的计算资源一般比较有限, 难以支撑运算量大的人脸验证。 另外, 在已有人脸 验证系统中, 人脸验证算法比较粗糙, 误判的机率很高。
发明内容
本发明实施例提供一种手机用户身份认证方法、云服务器以及网络系统, 能够将身份认 证的负载由云服务器承担, 提高手机操作系统的安全性、增强用户体验、提高人脸验证的精 确度。
本发明实施例提供一种手机用户身份认证方法, 所述手机通过通信网络与云服务器连 接, 所述云服务器中存储用户对应的人脸样本图像库; 该方法包括:
用户在手机上输入登录账号和密码;
判断登录账号和密码是否正确;
如果登录账号和密码出错,则拒绝所述用户进入手机操作系统;若登录账号和密码正确, 则将所述登录账号和密码发送到所述云服务器;所述登录账号和密码对应所述云服务器中存 储的用户的人脸样本图像库;
手机摄像头获取用户的人脸输入图像, 将所述人脸输入图像发送到所述云服务器; 所述云服务器根据登录账号和密码以及所述人脸输入图像, 对所述用户进行身份认证, 判断是否允许所述用户进入手机操作系统, 具体包括:
步骤 A. 所述云服务器根据登录账号和密码,确定所述登录账号和密码对应的用户的人 脸样本图像库;
步骤 B. 根据所述人脸输入图像与所述人脸样本图像库, 得到人脸特征相似值; 其中步骤 B包括:
B1. 通过人脸检测, 从所述人脸输入图像中获得人脸区域图像;
B2. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值;
B3. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值之间的特性值距离,得到多个第二特性距离,并根据所述多个第二特性距离确 定所述人脸特征相似值;
步骤 C. 判断所述人脸特征相似值是否大于预设阈值,其中所述预设阈值是根据所述人 脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;
步骤 D. 若所述人脸特征相似值不大于所述预设阈值, 则允许所述用户进入手机操作系 统;
步骤 E. 若所述人脸特征相似值大于所述预设阈值, 则计算第一数量和第二数量, 所述 第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人脸样本图 像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的人脸样本图像 库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量;
步骤 F. 若所述第一数量小于所述第二数量, 则拒绝所述用户进入手机操作系统; 步骤 G. 若所述第一数量不小于所述第二数量, 则允许所述用户进入手机操作系统。 本发明实施例提供一种云服务器, 包括: 存储单元, 用于存储用户的人脸样本图像库; 接收单元, 用于接收来自用户手机的登录账号和密码, 以及人脸输入图像; 确定单元,用于根据所述登录账号和密码,确定所述登录账号和密码对应的存储在所述 存储单元的用户人脸样本图像库;
人脸特征相似值确定单元,用于根据所述人脸输入图像与所述人脸样本图像库,得到人 脸特征相似值;该人脸特征相似值确定单元包括人脸区域图像获取单元、特性值计算单元和 特性距离计算单元, 其中:
人脸区域图像获取单元,用于通过人脸检测,从所述人脸输入图像中获得人脸区域图像; 特性值计算单元,用于计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所 述人脸区域图像的第二特性值;
特性距离计算单元,用于计算所述人脸样本图像库中每个人脸样本图像的第一特性值与 所述人脸区域图像的第二特性值之间的特性值距离,得到多个第二特性距离,并根据所述多 个第二特性距离确定所述人脸特征相似值; 第一判断单元,用于判断所述人脸特征相似值是否大于预设阈值,其中所述预设阈值是 根据所述人脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;
第一允许单元,用于当所述人脸特征相似值不大于所述预设阈值时,则允许所述用户进 入手机操作系统;
第二判断单元,用于当所述人脸特征相似值大于所述预设阈值时,计算第一数量和第二 数量,所述第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人 脸样本图像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的人脸 样本图像库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量;
拒绝单元,用于当所述第一数量小于所述第二数量时,拒绝所述用户进入手机操作系统; 第一允许单元,用于当所述第一数量不小于所述第二数量时,允许所述用户进入手机操 作系统。
本发明实施例提供一种网络系统, 包括: 手机和云服务器, 所述手机通过通信网络与所 述云服务器连接; 其中,
所述手机用于接收用户输入的登录账号和密码,判断登录账号和密码是否正确;如果登 录账号和密码出错, 则拒绝所述用户进入手机操作系统; 若登录账号和密码正确, 则将所述 登录账号和密码发送到所述云服务器;所述登录账号和密码对应所述云服务器中存储的用户 的人脸样本图像库; 获取用户的人脸输入图像, 将所述人脸输入图像发送到所述云服务器; 所述云服务器用于存储用户对应的人脸样本图像库;根据登录账号和密码以及所述人脸 输入图像,对所述用户进行身份认证,判断是否允许所述用户进入手机操作系统,具体包括: 步骤 A.根据登录账号和密码,确定所述登录账号和密码对应的用户的人脸样本图像库; 步骤 B. 根据所述人脸输入图像与所述人脸样本图像库, 得到人脸特征相似值; 其中步骤 B包括:
B1. 通过人脸检测, 从所述人脸输入图像中获得人脸区域图像;
B2. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值;
B3. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值之间的特性值距离,得到多个第二特性距离,并根据所述多个第二特性距离确 定所述人脸特征相似值;
步骤 C. 判断所述人脸特征相似值是否大于预设阈值,其中所述预设阈值是根据所述人 脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的; 步骤 D. 若所述人脸特征相似值不大于所述预设阈值, 则允许所述用户进入手机操作系 统;
步骤 E. 若所述人脸特征相似值大于所述预设阈值, 则计算第一数量和第二数量, 所述 第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人脸样本图 像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的人脸样本图像 库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量;
步骤 F. 若所述第一数量小于所述第二数量, 则拒绝所述用户进入手机操作系统; 步骤 G. 若所述第一数量不小于所述第二数量, 则允许所述用户进入手机操作系统。 本发明实施例能够将身份认证的负载由云服务器承担,提高手机操作系统的安全性、增 强用户体验、 提高人脸验证的精确度。
附图说明
图 1为本发明实施例一提供的方法的流程图;
图 2为本发明实施例二提供的云服务器的结构示意图;
图 3为本发明实施例二提供的云服务器中的人脸特征相似值确定单元的结构示意图; 图 4为本发明实施例三提供的网络系统的结构示意图。
具体实施方式
为使本发明技术方案的优点更加清楚, 下面结合附图和实施例对本发明作详细说明。 云计算 (Cloud computing) 是一种基于互联网的计算方式, 通过这种方式, 共享的软 硬件资源和信息可以按需提供给计算机、手机和其他设备。典型的云计算提供商往往提供通 用的网络业务应用,可以通过浏览器等软件或者其他 Web服务来访问,而软件和数据都存储 在云服务器上。本发明实施例基于云计算技术,可以将手机的身份认证任务由云服务器承担, 从而减轻手机的负担, 也能够在手机上引入开销较高的服务, 提高用户体验。
在本发明实施例中,用户手机通过通信网络与云服务器连接,云服务器中存储用户对应 的人脸样本图像库,云服务器可以是由电信运营商管理,用户在签约时将人脸样本图像注册 到该电信运营商的云服务器。云服务器将用户的手机号码、手机操作系统登录账户和密码等 信息同人脸样本图像库绑定。
实施例一
本发明实施例一提供的手机用户身份认证方法流程图如图 1所示,该方法可以包括以下 步骤:
步骤 S101. 用户在手机上输入登录账号和密码; 步骤 S103. 手机判断登录账号和密码是否正确;
步骤 S105. 如果登录账号和密码出错, 则拒绝所述用户进入手机操作系统, 提示出错; 步骤 S107. 若登录账号和密码正确, 将该登录账号和密码发送到云服务器, 登录账号 和密码对应云服务器中存储的用户的人脸样本图像库;
步骤 S109. 手机启动摄像头, 获取用户的人脸输入图像, 并将该人脸输入图像发送到 云服务器;
步骤 S111. 云服务器根据登录账号和密码以及人脸输入图像, 对用户进行身份认证, 判断是否允许所述用户进入手机操作系统; 步骤 S111具体包括:
步骤 S111-2. 云服务器根据登录账号和密码, 确定所述登录账号和密码对应的用户的 人脸样本图像库;
步骤 S111-4. 根据人脸输入图像与人脸样本图像库, 得到人脸特征相似值; 该人脸特 征相似值即人脸输入图像与每个人脸样本图像之相似程度, 人脸特征相似值越小越相似; 步骤 S111-4具体包括:
步骤 S111-4-1. 通过人脸检测, 从人脸输入图像中获得人脸区域图像; 该人脸检测方 法主要是通过将人脸输入图像与人脸样本图像进行人脸肤色区域对比, 并按照头形的比例, 取出该人脸区域图像。
步骤 S111-4-3. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人 脸区域图像的第二特性值;
步骤 S111-4-5. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人 脸区域图像的第二特性值的特征值距离,得到多个第二特性距离,并根据所述多个第二特性 距离确定人脸特征相似值;该人脸特征相似值即人脸输入图像与每个人脸样本图像之相似程 度, 人脸特征相似值越小越相似。人脸特征相似值可以是多个第二特性距离中的最大值, 或 者可以是多个第二特性距离的平均值。
步骤 S111-6. 判断所述人脸特征相似值是否大于预设阈值, 其中所述预设阈值是根据 所述人脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;预设阈值可以是 多个第一特性距离中的最大值, 或者可以是多个第一特性距离的平均值。
步骤 S111-8.若所述人脸特征相似值不大于所述预设阈值, 即用户人脸图像与人脸样本 图像库中的人脸样本图像相似度符合要求, 则允许所述用户进入手机操作系统;
步骤 S111-10. 若所述人脸特征相似值大于所述预设阈值, 即用户人脸图像与人脸样 本图像库中的人脸样本图像相似度未能符合要求,则分别统计有多少个人脸样本图像的第一 特性距离大于或小于人脸特征相似值, 即计算第一数量和第二数量,所述第一数量为大于所 述人脸特征相似值的第一特性距离对应的人脸样本图像库中人脸样本图像的个数,所述第二 数量为不大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人脸样本图像 的个数; 然后, 判断所述第一数量是否大于所述第二数量;
步骤 S111-12. 若第一数量小于所述第二数量, 则拒绝用户进入手机操作系统; 步骤 S111-14. 若第一数量不小于所述第二数量, 则允许用户进入手机操作系统。 下文进一步详细具体地说明本发明实施例是如何提取人脸图像特征,确定人脸样本图像 库中每个人脸样本图像的第一特性值、人脸区域图像的第二特性值, 以及人脸样本图像库中 人脸样本图像两两之间的第一特性距离,人脸样本图像库中每个人脸样本图像与所述人脸区 域图像的第二特性距离, 以及预设阈值和人脸特征相似值的。
以人脸样本图像 7)为例, 人脸样本图像 ¾ _ )为二维 64X64灰度图像, 表示横 坐标像素, _ 表示纵坐标像素。 人脸样本图像库由 幅人脸样本图像构成, 则可以用 {X;I =1,2,L ,Μ}表示人脸样本图像库。将 幅人脸样本图像按人脸位置重合叠加, 求所 有图像重叠后的均值, 其均值为 = 1£Χ;, 每幅人脸样本图像 与均值 的差为: = X,- ( = l,L ,Μ) 构造协方差矩阵: C = A^ 其中 A = [ ,^,L 为差值向量的线性组合。 对于 64X64人脸图像, 协方差矩阵 C 的大小为 4096X4096, 直接对它求解特征值和特征向量是很困难的。 根据奇异值分解定理, 通过求解 A的特征值和特征向量来获得 C = A ^的特征值和特征向量。 设 ( = 1,2,L ,r)为矩阵 的 r个非零特征值, v,为 对应于 的特征向量, 则
C = AAT的正交归一特征向量 i =-=AVi, 将样本协方差对应的特征值按大小排列:
11>A2≥L ≥ΛΓ 。 设其对应的特征向量为 ;, 这样每幅人脸样本图像都可以投影到由 l, 2,L , 张成的特征空间 ft 具体运用时可以选取前面 个特征值作为特征空间, 因为 该特征空间的维数比原人脸样本图像的维数低, 所以将每幅人脸样本图像投影到由 u,,u2,L , 张成的特征空间 之后, 人脸样本图像维数也大大降低, 从而达到降低维数和 提取特征的目的。 选取的原则按照特征值所占的能量比例来确定, 通常取 "=95%~99%之间。 为了提高特征提取的效率和精度,本发明实施例提出了对人脸样本图像分块求取特征向 量的方法。 鉴于人脸具有三个显著特征: 眼睛、 鼻子和嘴巴, 而且它们分别处在人脸的上中 下三块, 根据这三个显著特征把人脸图像分为三个独立的子块——上部包括: 眼睛, 中部包 括鼻子, 下部包括嘴巴。
经过分块, 一幅人脸样本图像就变成了三个子图像, 那么每幅人脸样本图像 X;可以表 示成 =[ " Χ X'f ( = 1,2,L , ). 原来的一个人脸样本图像库变成三个相互独立的子块图像库, 即 Χ,", X/"和 X. ( = 1,2,L ,M)。 如果 为 行 列的矩阵, 则 X,"为的 A行 列的矩阵, X™为 ¾行 列的矩阵, X 为 ¾行 列的矩阵, 其中 A +P2 + = P . 所有人脸样本图像库的上部子图像构成上部子块图像库,同样中部和下部子图像就构成 了中部子块图像库和下部子块图像库。在特征提取的过程中,它们将会被当作是三个独立的 子块图像库。
鉴于人脸样本图像库的样本有限,本发明实施例提出如下算法,可以在不采样的情况下 增加样本数量, 从而提高特征提取的精度。 该方法具体包括:
1.对人脸样本图像 X (mx«的矩阵)生成其对偶样本 X',其中 X' =ΧΓ, 为《x«的 矩阵, 它的反对角元素为 1, 其他元素为 0, 即有:
X(l,l) X(l,2) L X(l,n) X\l,n) X\l,n-l) L X'(l,l)
X M M O M , X' M M O M
X(m,l) X(m,2) L X(m,n) X m,n) X\m,n-1) L X\m,l) 其中矩阵 具有对称性, 即 y = }^; 以及正交性, 即7}^ = }^ =/ (/表示单位矩 阵)。
将 J分解为第一样本 Xe = (X + X ') / 2和第二样本 X。 =(X_X' /2,则对偶样本 X ' 均值、 协方差矩阵 C'之间的关系为:
Y' = XY , C' = YTCY 第一样本 x„均值、 协方差矩阵 α之间的关系为: X =X(I + Y)/2 C = (I + YYC(I + Y)/4 第二样本 xn均值、 协方差矩阵 cn之间的关系为:
X。 = X(J— Y、I2 C0 =(I-YY C(I-Y)/4 通过理论推导, 可以得到: 第一样本:^的特征空间和第二样本 ^的特征空间互相正 交, 而且 X的特征空间是第一样本 xe的特征空间和第二样本 x0的特征空间的直接和。 因此, 可以分别对 xe和 x0分别根据特征提取算法得到第一特征空间 ue和第二特征空 间 uo, 然后从第一特征空间 ue和第二特征空间 uo中挑选出识别精度高且差异大的特征向 量构成特征空间 u。
3.将 作为特征变换矩阵, 通过 =^提取特征。
结合分块后的人脸样本图像库对本发明实施例的方法进行说明。 以上部子块图像库为 例, 对上部子块图像库的每个样本 C," G' = 1,2,L ,Μ)分别生成每个样本的对偶样本
X. 1 ( =1,2,L ,Μ),其中 X,"' = Χ Υ , 为 βχβ的矩阵, 它的反对角元素为 1, 其他
Figure imgf000010_0001
1)
将 X,"分解为第一样本 = (X- + X- ') 12和第二样本 X = (X - X," ') / 2。 分别 对 和 X 分别根据上述特征提取算法, 得到第一特征空间 Uu i e和第二特征空间 Uu i 0, 然后从第一特征空间 Uu ie和第二特征空间 Uu i 0中挑选出识别精度高且差异大的特征向量 构造特征空间 将 作为特征变换矩阵,通过 "= " 提取 X,"在特征空间 ";的 投影, 即 Κ"。 以上述同样的方法对中部子块图像库和下部子块图像库每个样本 C,m和 X- ( =1,2,L ,M)进行特征提取, 记中部子块图像库和下部子块图像库每个样本 X/"和 X. ( = 1,2,L ,Μ)的在各自特征空间的投影为 ^和 。 假设 ^"为 维向量, 对上部子块图像库中每个样本 X," ( = 1,2,L ,Μ)的特征矩阵
Figure imgf000011_0001
对中部子块图像库和下部子块图像库每个样本 X/"和 X ( =1,2,L ,Μ)的特征空间 和
Figure imgf000011_0002
对上部子块图像库、 中部子块图像库和下部子块图像库每个样本 X;1, X/"和 X 1的特 性值 η", ? 和 7 求平均值, 得到人脸样本图像库中每个人脸样本 C; 的第一特性值 η =
(Τ +Ti m+Ti b) /3. ( = 1,2,L ,Μ) 以上描述的是针对人脸样本图像库的处理。根据上述同样的方法对人脸区域图像也作相 应处理, 即对人脸区域图像进行分块, 分别计算每块相应的特性值, 求和取平均值, 最后得 到人脸区域图像的第二特性值 Τ。
本发明实施例提出一种计算特性距离的方法——根据人脸样本图像库中每个人脸样本 图像的第一特性值, 计算人脸样本图像之间的多个第一特性距离。 具体包括:
对人脸样本图像 和:^ (i, 7=1,2,L ,Μ, 且 =/), 这两个人脸样本图像之间的 第一特性距离为 D Xi, X. )= Τ ΊΫ .计算两两人脸样本图像之间的多个第一特性距 离, 一共有 个第一特性距离。
然后, 根据人脸样本图像库中每个人脸样本图像之间的 个第一特性距离求得 预设阈值, 该预设阈值可以是 Μ(Λί-υ/ 个第一特性距离中的最大值, 也可以是 Λί(Λί-υ/2 个第一特性距离的平均值。
同样地,根据人脸区域图像的第二特性值 和人脸样本图像库中每个人脸样本图像的第 一特性值, 可以求得多个第二特性距离 XX,, X ) = >/(Γ. - Γ)2 ( = 1, 2,L , ) , 一共有 个第二特性距离。 然后, 再根据 个第二特性距离确定人脸特征相似值, 所述人脸特征相 似值可以是 个第二特性距离中的最大值, 也可以是 个第二特性距离的平均值。
也就是说, 计算所述人脸样本图像库中每个人脸样本图像的第一特性值的步骤包括: 将人脸样本图像 X,分成三个子图像, 即 X,", X™和 Xt b (i = 1, 2,L ,Μ); 对 X,", X™和 Xt b分别生成对偶样本; 根据所述对偶样本, 将 X,", xr和 X 1分别分解为第一样本 X = (x; + X ') / 2和 第二样本 二 ^"— ^" ') / ; 分别对所述第一样本和第二样本构造协方差矩阵;
分别确定所述第一样本协方差矩阵的正交归一特征向量和所述第二样本协方差矩阵的 正交归一特征向量;
根据所述第一样本协方差矩阵的正交归一特征向量组成的第一特征空间, 以及所述第 二样本协方差矩阵的正交归一特征向量组成的第二特征空间,确定所述第一样本和所述第二 样本分别在所述第一特征空间和第二特征空间的投影;
根据所述第一样本和所述第二样本在所述第一特征空间和第二特征空间的投影确定 χ , χ;"和 ^的特性值; 根据 X,", X™和 X 的特性值确定所述人脸样本图像 X,的第一特性值; 计算所述人脸区域图像的第二特性值包括的步骤:
将所述人脸区域图像分成三个子图像;
对所述三个子图像分别生成对应的对偶样本;
根据所述三个子图像对应的对偶样本,将所述三个子图像分别分解为第一样本和第二样 本;
分别对所述三个子图像的第一样本和第二样本构造协方差矩阵;
分别确定所述第一样本协方差矩阵的正交归一特征向量和所述第二样本协方差矩阵的 正交归一特征向量;
根据所述第一样本协方差矩阵的正交归一特征向量组成的特征空间, 以及所述第二样 本协方差矩阵的正交归一特征向量组成的特征空间,确定所述第一样本和所述第二样本在特 征空间的投影;
根据所所述第一样本和所述第二样本在特征空间的投影确定所述三个子图像的特性值; 根据所述三个子图像的特性值确定所述人脸区域图像的第二特性值。
本发明实施例的步骤 S 111-14还包括:若第一数量不小于所述第二数量,则利用所述人 脸输入图像对所述人脸样本图像库进行更新; 更新的策略可以是替代最久远的人脸样本图 像, 或者替代与所述人脸输入图像差异最大的人脸样本图像。 另外, 还可以重新计算所述云 服务器中的人脸样本图像库的第一特性距离, 并根据所述第一特性距离确定新的预设阈值, 将所述新的预设阈值替代所述预设阈值。 从而实现人脸样本图库的动态更新。
本发明实施例的手机用户身份认证方法, 能够将身份认证的负载由云服务器承担,提高 手机操作系统的安全性、 增强用户体验、 提高人脸验证的精确度。
实施例二
本发明实施例还提供一种云服务器 100, 如图 2所示, 包括:
存储单元 200, 用于存储用户的人脸样本图像库;
接收单元 201, 用于接收来自用户手机的登录账号和密码, 以及人脸输入图像; 确定单元 203, 用于根据所述登录账号和密码, 确定所述登录账号和密码对应的存储在 存储单元 200用户的人脸样本图像库;
人脸特征相似值确定单元 205, 用于根据所述人脸输入图像与所述人脸样本图像库, 得 到人脸特征相似值;如图 3所示,该人脸特征相似值确定单元 205包括人脸区域图像获取单 元 205-2、 特性值计算单元 205-4和特性距离计算单元 205-6, 其中:
人脸区域图像获取单元 205-2, 用于通过人脸检测, 从所述人脸输入图像中获得人脸区 域图像;
特性值计算单元 205-4,用于计算所述人脸样本图像库中每个人脸样本图像的第一特性 值与所述人脸区域图像的第二特性值;
特性距离计算单元 205-6,用于计算所述人脸样本图像库中每个人脸样本图像的第一特 性值与所述人脸区域图像的第二特性值之间的特性值距离,得到多个第二特性距离,并根据 所述多个第二特性距离确定所述人脸特征相似值;
第一判断单元 207, 用于判断所述人脸特征相似值是否大于预设阈值, 其中所述预设阈 值是根据所述人脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;
第一允许单元 209, 用于当所述人脸特征相似值不大于所述预设阈值时, 则允许所述用 户进入手机操作系统; 第二判断单元 211, 用于当所述人脸特征相似值大于所述预设阈值时, 计算第一数量和 第二数量,所述第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库 中人脸样本图像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的 人脸样本图像库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量; 拒绝单元 213, 用于当所述第一数量小于所述第二数量时, 拒绝所述用户进入手机操作 系统;
第二允许单元 215, 用于当所述第一数量不小于所述第二数量时, 允许所述用户进入手 机操作系统。
可选的, 该云服务器还可以包括: 第一更新单元 217, 用于当所述第一数量不小于所述 第二数量时, 利用所述人脸输入图像对所述人脸样本图像库进行更新。
可选的, 该云服务器还可以包括: 第二更新单元 219, 用于重新计算所述云服务器中的 人脸样本图像库的第一特性距离,并根据所述第一特性距离确定新的预设阈值,将所述新的 预设阈值替代所述预设阈值。
所述特性值计算单元 205-4包括:
第一划分单元 205-41, 用于将人脸样本图像 分成三个子图像, 即 X; "和 ( = 1, 2,L , ) ; 第一生成单元 205-43, 用于对 X,", X; "和 Xt b分别生成对偶样本; 第一分解单元 205-45, 用于根据所述对偶样本, 将 Χ,", X/"和 分别分解为第一 样本 X = (X + X ') 1 2和第二样本 X;1。 = (X - ') ; 第一协方差矩阵构造单元 205-47, 用于分别对所述第一样本和第二样本构造协方差矩 阵;
第一特征向量计算单元 205-49, 用于分别确定所述第一样本协方差矩阵的正交归一特 征向量和所述第二样本协方差矩阵的正交归一特征向量;
第一投影计算单元 205-411, 用于根据所述第一样本协方差矩阵的正交归一特征向量 组成的第一特征空间,以及所述第二样本协方差矩阵的正交归一特征向量组成的第二特征空 间, 确定所述第一样本和所述第二样本分别在所述第一特征空间和第二特征空间的投影; 第一特性值确定单元 205-413,用于根据所述第一样本和所述第二样本在所述第一特征 空间和第二特征空间的投影确定 X,", X™和 Xt b的特性值; 根据 X,", Χ 和 的特性 值确定所述人脸样本图像 xt的第一特性值; 第二划分单元 205-415, 用于将所述人脸区域图像分成三个子图像;
第二生成单元 205-417, 用于对所述三个子图像分别生成对应的对偶样本; 第二分解单元 205-419, 用于根据所述三个子图像对应的对偶样本, 将所述三个子图像 分别分解为第一样本和第二样本;
第二协方差矩阵构造单元 205-421, 用于分别对所述三个子图像的第一样本和第二样 本构造协方差矩阵;
第二特征向量计算单元 205-423,用于分别确定所述第一样本协方差矩阵的正交归一特 征向量和所述第二样本协方差矩阵的正交归一特征向量;
第二投影计算单元 205-425, 用于根据所述第一样本协方差矩阵的正交归一特征向量 组成的特征空间, 以及所述第二样本协方差矩阵的正交归一特征向量组成的特征空间,确定 所述第一样本和所述第二样本在特征空间的投影;
第二特性值确定单元 205-427,用于根据所所述第一样本和所述第二样本在特征空间的 投影确定所述三个子图像的特性值;根据所述三个子图像的特性值确定所述人脸区域图像的 第二特性值。
本发明实施例提供能够将身份认证的负载由云服务器承担, 提高手机操作系统的安全 性、 增强用户体验、 提高人脸验证的精确度。
实施例三
本发明实施例还提供一种网络系统,包括手机和云服务器,所述手机通过通信网络与所 述云服务器连接; 其中,
所述手机用于接收用户输入的登录账号和密码,判断登录账号和密码是否正确;如果登 录账号和密码出错, 则拒绝所述用户进入手机操作系统; 若登录账号和密码正确, 则将所述 登录账号和密码发送到所述云服务器;所述登录账号和密码对应所述云服务器中存储的用户 的人脸样本图像库; 获取用户的人脸输入图像, 将所述人脸输入图像发送到所述云服务器; 所述云服务器用于存储用户对应的人脸样本图像库;根据登录账号和密码以及所述人脸 输入图像,对所述用户进行身份认证,判断是否允许所述用户进入手机操作系统,具体包括: 步骤 A.根据登录账号和密码,确定所述登录账号和密码对应的用户的人脸样本图像库; 步骤 B. 根据所述人脸输入图像与所述人脸样本图像库, 得到人脸特征相似值; 其中步骤 B包括: Bl. 通过人脸检测, 从所述人脸输入图像中获得人脸区域图像;
B2. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值;
B3. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值之间的特性值距离,得到多个第二特性距离,并根据所述多个第二特性距离确 定所述人脸特征相似值;
步骤 C. 判断所述人脸特征相似值是否大于预设阈值,其中所述预设阈值是根据所述人 脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;
步骤 D. 若所述人脸特征相似值不大于所述预设阈值, 则允许所述用户进入手机操作系 统;
步骤 E. 若所述人脸特征相似值大于所述预设阈值, 则计算第一数量和第二数量, 所述 第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人脸样本图 像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的人脸样本图像 库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量;
步骤 F. 若所述第一数量小于所述第二数量, 则拒绝所述用户进入手机操作系统; 步骤 G. 若所述第一数量不小于所述第二数量, 则允许所述用户进入手机操作系统。 其中, 云服务器的具体结构可以如实施例二所描述。
本发明实施例能够将身份认证的负载由云服务器承担,提高手机操作系统的安全性、增 强用户体验、 提高人脸验证的精确度。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计 算机程序来指令相关的硬件来完成,所述的程序可存储于一计算机可读取存储介质中,该程 序在执行时,可包括如上述各方法的实施例的流程。其中,所述的存储介质可为磁碟、光盘、 只读存储记忆体(Read- Only Memory ROM)或随机存储记忆体(Random Access Memory, RAM) 等。
以上所述, 仅为本发明的具体实施方式, 但本发明的保护范围并不局限于此, 任何熟悉 本技术领城的技术人员在本发明揭露的技术范围内,可轻易想到的变化或替换,都应涵盖在 本发明的保护范围之内。 因此, 本发明的保护范围应该以权利要求的保护范围为准。

Claims

权利 要 求
1. 一种手机用户身份认证方法,其特征在于,所述手机通过通信网络与云服务器连接, 所述云服务器中存储用户对应的人脸样本图像库; 该方法包括:
用户在手机上输入登录账号和密码;
判断登录账号和密码是否正确;
如果登录账号和密码出错,则拒绝所述用户进入手机操作系统;若登录账号和密码正确, 则将所述登录账号和密码发送到所述云服务器;所述登录账号和密码对应所述云服务器中存 储的用户的人脸样本图像库;
手机摄像头获取用户的人脸输入图像, 将所述人脸输入图像发送到所述云服务器; 所述云服务器根据登录账号和密码以及所述人脸输入图像, 对所述用户进行身份认证, 判断是否允许所述用户进入手机操作系统, 具体包括:
步骤 A. 所述云服务器根据登录账号和密码,确定所述登录账号和密码对应的用户的人 脸样本图像库;
步骤 B. 根据所述人脸输入图像与所述人脸样本图像库, 得到人脸特征相似值; 其中步骤 B包括:
B1. 通过人脸检测, 从所述人脸输入图像中获得人脸区域图像;
B2. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值;
B3. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值之间的特性值距离,得到多个第二特性距离,并根据所述多个第二特性距离确 定所述人脸特征相似值;
步骤 C. 判断所述人脸特征相似值是否大于预设阈值,其中所述预设阈值是根据所述人 脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;
步骤 D. 若所述人脸特征相似值不大于所述预设阈值, 则允许所述用户进入手机操作系 统;
步骤 E. 若所述人脸特征相似值大于所述预设阈值, 则计算第一数量和第二数量, 所述 第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人脸样本图 像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的人脸样本图像 库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量;
步骤 F. 若所述第一数量小于所述第二数量, 则拒绝所述用户进入手机操作系统; 步骤 G. 若所述第一数量不小于所述第二数量, 则允许所述用户进入手机操作系统。 2. 根据权利要求 1所述的手机用户身份认证方法,其特征在于,所述步骤 B2中的计算 所述人脸样本图像库中每个人脸样本图像的第一特性值包括:
将人脸样本图像 X;分成三个子图像, 即 X,", X/"和 X ( = 1,
2,L , ) ; 对 X,", X™和 Xt b分别生成对偶样本; 根据所述对偶样本, 将 X,", xr和 X 1分别分解为第一样本 ;e = (x; + X ') / 2和 第二样本 "。 = ( " _:^" ') / 2 ; 分别对所述第一样本和第二样本构造协方差矩阵;
分别确定所述第一样本协方差矩阵的正交归一特征向量和所述第二样本协方差矩阵的 正交归一特征向量;
根据所述第一样本协方差矩阵的正交归一特征向量组成的第一特征空间, 以及所述第 二样本协方差矩阵的正交归一特征向量组成的第二特征空间,确定所述第一样本和所述第二 样本分别在所述第一特征空间和第二特征空间的投影;
根据所述第一样本和所述第二样本在所述第一特征空间和第二特征空间的投影确定 χ , X/"和 的特性值; 根据 X,", X™和 X 的特性值确定所述人脸样本图像 X的第一特性值;
所述步骤 B2中计算所述人脸区域图像的第二特性值包括:
将所述人脸区域图像分成三个子图像;
对所述三个子图像分别生成对应的对偶样本;
根据所述三个子图像对应的对偶样本,将所述三个子图像分别分解为第一样本和第二样 本;
分别对所述三个子图像的第一样本和第二样本构造协方差矩阵;
分别确定所述第一样本协方差矩阵的正交归一特征向量和所述第二样本协方差矩阵的 正交归一特征向量;
根据所述第一样本协方差矩阵的正交归一特征向量组成的特征空间, 以及所述第二样 本协方差矩阵的正交归一特征向量组成的特征空间,确定所述第一样本和所述第二样本在特 征空间的投影;
根据所所述第一样本和所述第二样本在特征空间的投影确定所述三个子图像的特性值; 根据所述三个子图像的特性值确定所述人脸区域图像的第二特性值。
3. 根据权利要求 1所述的手机用户身份认证方法, 其特征在于, 其中所述人脸特征相 似值是所述多个第二特性距离中的最大值, 或者是所述多个第二特性距离的平均值。
4. 根据权利要求 1所述的手机用户身份认证方法, 其特征在于, 其中所述预设阈值是 所述第一特性距离中的最大值, 或者是所述多个第一特性距离的平均值。
5. 根据权利要求 1所述的手机用户身份认证方法, 其特征在于, 所述步骤 G还包括: 若所述第一数量不小于所述第二数量,则利用所述人脸输入图像对所述人脸样本图像库 进行更新。
6. 根据权利要求 5所述的手机用户身份认证方法, 其特征在于, 还包括:
重新计算所述云服务器中的人脸样本图像库的第一特性距离,并根据所述第一特性距离 确定新的预设阈值, 将所述新的预设阈值替代所述预设阈值。
7. —种云服务器, 其特征在于, 包括:
存储单元, 用于存储用户的人脸样本图像库;
接收单元, 用于接收来自用户手机的登录账号和密码, 以及人脸输入图像; 确定单元,用于根据所述登录账号和密码,确定所述登录账号和密码对应的存储在所述 存储单元的用户人脸样本图像库;
人脸特征相似值确定单元,用于根据所述人脸输入图像与所述人脸样本图像库,得到人 脸特征相似值;该人脸特征相似值确定单元包括人脸区域图像获取单元、特性值计算单元和 特性距离计算单元, 其中:
人脸区域图像获取单元,用于通过人脸检测,从所述人脸输入图像中获得人脸区域图像; 特性值计算单元,用于计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所 述人脸区域图像的第二特性值;
特性距离计算单元,用于计算所述人脸样本图像库中每个人脸样本图像的第一特性值与 所述人脸区域图像的第二特性值之间的特性值距离,得到多个第二特性距离,并根据所述多 个第二特性距离确定所述人脸特征相似值;
第一判断单元,用于判断所述人脸特征相似值是否大于预设阈值,其中所述预设阈值是 根据所述人脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;
第一允许单元,用于当所述人脸特征相似值不大于所述预设阈值时,则允许所述用户进 入手机操作系统;
第二判断单元,用于当所述人脸特征相似值大于所述预设阈值时,计算第一数量和第二 数量,所述第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人 脸样本图像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的人脸 样本图像库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量;
拒绝单元,用于当所述第一数量小于所述第二数量时,拒绝所述用户进入手机操作系统; 第一允许单元,用于当所述第一数量不小于所述第二数量时,允许所述用户进入手机操 作系统。
8. 根据权利要求 7所述的云服务器, 其特征在于, 所述特性值计算单元包括: 第一划分单元, 用于将人脸样本图像 X;分成三个子图像, 即 X,", X/"和
X " ( = 1, 2,L , ) ; 第一生成单元, 用于对 X,", X/"和 分别生成对偶样本; 第一分解单元, 用于根据所述对偶样本, 将 X,", X/"和 分别分解为第一样本 Xi u e = (X + X ') / 2和第二样本 X;1。 = (X 2; 第一协方差矩阵构造单元, 用于分别对所述第一样本和第二样本构造协方差矩阵; 第一特征向量计算单元,用于分别确定所述第一样本协方差矩阵的正交归一特征向量和 所述第二样本协方差矩阵的正交归一特征向量;
第一投影计算单元, 用于根据所述第一样本协方差矩阵的正交归一特征向量组成的第 一特征空间, 以及所述第二样本协方差矩阵的正交归一特征向量组成的第二特征空间,确定 所述第一样本和所述第二样本分别在所述第一特征空间和第二特征空间的投影;
第一特性值确定单元,用于根据所述第一样本和所述第二样本在所述第一特征空间和第 二特征空间的投影确定 X,", Χ 和 Xt b的特性值; 根据 X,", Χ 和 X 的特性值确定所 述人脸样本图像 X;的第一特性值; 第二划分单元, 用于将所述人脸区域图像分成三个子图像;
第二生成单元, 用于对所述三个子图像分别生成对应的对偶样本;
第二分解单元,用于根据所述三个子图像对应的对偶样本,将所述三个子图像分别分解 为第一样本和第二样本;
第二协方差矩阵构造单元, 用于分别对所述三个子图像的第一样本和第二样本构造协 方差矩阵; 第二特征向量计算单元,用于分别确定所述第一样本协方差矩阵的正交归一特征向量和 所述第二样本协方差矩阵的正交归一特征向量;
第二投影计算单元, 用于根据所述第一样本协方差矩阵的正交归一特征向量组成的特 征空间, 以及所述第二样本协方差矩阵的正交归一特征向量组成的特征空间,确定所述第一 样本和所述第二样本在特征空间的投影;
第二特性值确定单元,用于根据所所述第一样本和所述第二样本在特征空间的投影确定 所述三个子图像的特性值;根据所述三个子图像的特性值确定所述人脸区域图像的第二特性 值。
9. 根据权利要求 7所述的云服务器, 其特征在于, 还包括:
第一更新单元,用于当所述第一数量不小于所述第二数量时,利用所述人脸输入图像对 所述人脸样本图像库进行更新。
10. 根据权利要求 9所述的云服务器, 其特征在于, 还包括:
第二更新单元,用于重新计算所述云服务器中的人脸样本图像库的第一特性距离,并根 据所述第一特性距离确定新的预设阈值, 将所述新的预设阈值替代所述预设阈值。
11. 一种网络系统, 其特征在于, 包括手机和云服务器, 所述手机通过通信网络与所述 云服务器连接; 其中,
所述手机用于接收用户输入的登录账号和密码,判断登录账号和密码是否正确;如果登 录账号和密码出错, 则拒绝所述用户进入手机操作系统; 若登录账号和密码正确, 则将所述 登录账号和密码发送到所述云服务器;所述登录账号和密码对应所述云服务器中存储的用户 的人脸样本图像库; 获取用户的人脸输入图像, 将所述人脸输入图像发送到所述云服务器; 所述云服务器用于存储用户对应的人脸样本图像库;根据登录账号和密码以及所述人脸 输入图像,对所述用户进行身份认证,判断是否允许所述用户进入手机操作系统,具体包括: 步骤 A.根据登录账号和密码,确定所述登录账号和密码对应的用户的人脸样本图像库; 步骤 B. 根据所述人脸输入图像与所述人脸样本图像库, 得到人脸特征相似值; 其中步骤 B包括:
B1. 通过人脸检测, 从所述人脸输入图像中获得人脸区域图像;
B2. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值;
B3. 计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所述人脸区域图像 的第二特性值之间的特性值距离,得到多个第二特性距离,并根据所述多个第二特性距离确 定所述人脸特征相似值;
步骤 C. 判断所述人脸特征相似值是否大于预设阈值,其中所述预设阈值是根据所述人 脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;
步骤 D. 若所述人脸特征相似值不大于所述预设阈值, 则允许所述用户进入手机操作系 统;
步骤 E. 若所述人脸特征相似值大于所述预设阈值, 则计算第一数量和第二数量, 所述 第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人脸样本图 像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的人脸样本图像 库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量;
步骤 F. 若所述第一数量小于所述第二数量, 则拒绝所述用户进入手机操作系统; 步骤 G. 若所述第一数量不小于所述第二数量, 则允许所述用户进入手机操作系统。
12. 根据权利要求 11所述的网络系统, 其特征在于, 其特征在于, 所述云服务器包括: 存储单元, 用于存储用户的人脸样本图像库;
接收单元, 用于接收来自用户手机的登录账号和密码, 以及人脸输入图像; 确定单元,用于根据所述登录账号和密码,确定所述登录账号和密码对应的存储在所述 存储单元的用户人脸样本图像库;
人脸特征相似值确定单元,用于根据所述人脸输入图像与所述人脸样本图像库,得到人 脸特征相似值;该人脸特征相似值确定单元包括人脸区域图像获取单元、特性值计算单元和 特性距离计算单元, 其中:
人脸区域图像获取单元,用于通过人脸检测,从所述人脸输入图像中获得人脸区域图像; 特性值计算单元,用于计算所述人脸样本图像库中每个人脸样本图像的第一特性值与所 述人脸区域图像的第二特性值;
特性距离计算单元,用于计算所述人脸样本图像库中每个人脸样本图像的第一特性值与 所述人脸区域图像的第二特性值之间的特性值距离,得到多个第二特性距离,并根据所述多 个第二特性距离确定所述人脸特征相似值;
第一判断单元,用于判断所述人脸特征相似值是否大于预设阈值,其中所述预设阈值是 根据所述人脸样本图像库中每个人脸样本图像之间的多个第一特性距离得到的;
第一允许单元,用于当所述人脸特征相似值不大于所述预设阈值时,则允许所述用户进 入手机操作系统;
第二判断单元,用于当所述人脸特征相似值大于所述预设阈值时,计算第一数量和第二 数量,所述第一数量为大于所述人脸特征相似值的第一特性距离对应的人脸样本图像库中人 脸样本图像的个数,所述第二数量为不大于所述人脸特征相似值的第一特性距离对应的人脸 样本图像库中人脸样本图像的个数, 并判断所述第一数量是否大于所述第二数量;
拒绝单元,用于当所述第一数量小于所述第二数量时,拒绝所述用户进入手机操作系统; 第一允许单元,用于当所述第一数量不小于所述第二数量时,允许所述用户进入手机操 作系统。
PCT/CN2011/076623 2011-06-30 2011-06-30 手机用户身份认证方法、云服务器以及网络系统 WO2013000142A1 (zh)

Priority Applications (7)

Application Number Priority Date Filing Date Title
PCT/CN2011/076623 WO2013000142A1 (zh) 2011-06-30 2011-06-30 手机用户身份认证方法、云服务器以及网络系统
US14/129,135 US8861798B2 (en) 2011-06-30 2011-06-30 Method for authenticating identity of handset user
CN201180071166.7A CN103814545B (zh) 2011-06-30 2011-06-30 手机用户身份认证方法
US14/486,136 US8989452B2 (en) 2011-06-30 2014-09-15 Method for authenticating identity of handset user
US14/486,112 US8983145B2 (en) 2011-06-30 2014-09-15 Method for authenticating identity of handset user
US14/622,110 US9537859B2 (en) 2011-06-30 2015-02-13 Method for authenticating identity of handset user in a cloud-computing environment
US15/355,268 US9813909B2 (en) 2011-06-30 2016-11-18 Cloud server for authenticating the identity of a handset user

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2011/076623 WO2013000142A1 (zh) 2011-06-30 2011-06-30 手机用户身份认证方法、云服务器以及网络系统

Related Child Applications (4)

Application Number Title Priority Date Filing Date
US14/129,135 A-371-Of-International US8861798B2 (en) 2011-06-30 2011-06-30 Method for authenticating identity of handset user
US14/486,136 Continuation US8989452B2 (en) 2011-06-30 2014-09-15 Method for authenticating identity of handset user
US14/486,136 Continuation-In-Part US8989452B2 (en) 2011-06-30 2014-09-15 Method for authenticating identity of handset user
US14/486,112 Continuation US8983145B2 (en) 2011-06-30 2014-09-15 Method for authenticating identity of handset user

Publications (1)

Publication Number Publication Date
WO2013000142A1 true WO2013000142A1 (zh) 2013-01-03

Family

ID=47423382

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/076623 WO2013000142A1 (zh) 2011-06-30 2011-06-30 手机用户身份认证方法、云服务器以及网络系统

Country Status (3)

Country Link
US (5) US8861798B2 (zh)
CN (1) CN103814545B (zh)
WO (1) WO2013000142A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015103970A1 (en) * 2014-01-09 2015-07-16 Tencent Technology (Shenzhen) Company Limited Method, apparatus and system for authenticating user
CN105279416A (zh) * 2015-10-27 2016-01-27 上海川织金融信息服务有限公司 基于多生物特征结合设备指纹的身份识别方法及其系统
CN110012018A (zh) * 2019-04-11 2019-07-12 国网山东省电力公司 一种工业网络安全系统

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150067786A1 (en) * 2013-09-04 2015-03-05 Michael Stephen Fiske Visual image authentication and transaction authorization using non-determinism
JP6292295B2 (ja) 2014-04-07 2018-03-14 日本電気株式会社 連携システム、装置、方法、および記録媒体
CN105577378A (zh) * 2014-10-14 2016-05-11 阿里巴巴集团控股有限公司 身份认证方法、装置和系统
GB2551688B (en) * 2016-03-18 2021-12-22 Vzinternet Ltd Methods, user devices, access control equipments, computer software, computer program products and systems for facilitating authentication or access control
CN107360119A (zh) * 2016-05-09 2017-11-17 中兴通讯股份有限公司 一种云桌面登陆验证方法、云桌面控制系统及客户端
CN106022043A (zh) * 2016-05-23 2016-10-12 南京甄视智能科技有限公司 一种生物特征识别业务处理平台的系统架构及实现方法
CN106603562A (zh) * 2016-12-30 2017-04-26 山东中架工人信息技术股份有限公司 一种rim云数据验证身份系统及方法
CN107093066B (zh) * 2017-03-22 2020-09-01 阿里巴巴集团控股有限公司 业务实现方法和装置
CN106951762A (zh) * 2017-03-29 2017-07-14 中控智慧科技股份有限公司 一种身份识别方法及系统
CN106980838A (zh) * 2017-03-29 2017-07-25 中控智慧科技股份有限公司 一种身份识别方法及系统
CN110166412A (zh) * 2018-02-14 2019-08-23 上海硅孚信息科技有限公司 基于区块链的智能身份生物认证凭证管理方法及系统
WO2019222345A1 (en) * 2018-05-18 2019-11-21 South 8 Technologies, Inc. Electrochemical cell cap
CN109492555A (zh) * 2018-10-26 2019-03-19 平安科技(深圳)有限公司 新生身份认证方法、电子装置及计算机可读存储介质
CN110210307B (zh) * 2019-04-30 2023-11-28 中国银联股份有限公司 人脸样本库部署方法、基于人脸识别业务处理方法及装置
US10860705B1 (en) 2019-05-16 2020-12-08 Capital One Services, Llc Augmented reality generated human challenge
CN110677537B (zh) * 2019-09-29 2021-05-11 维沃移动通信有限公司 便签信息显示方法、便签信息发送方法及电子设备
CN113259136B (zh) * 2021-07-07 2021-11-16 浙江宇视科技有限公司 特征识别的多客户端协同认证方法、装置、设备及介质
CN114915439A (zh) * 2021-10-27 2022-08-16 杭州拼便宜网络科技有限公司 电商平台身份验证的方法、装置、电子设备及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1629875A (zh) * 2003-12-15 2005-06-22 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
CN1972186A (zh) * 2005-11-24 2007-05-30 中国科学院自动化研究所 一种移动式身份认证系统及其认证方法
CN101132557A (zh) * 2007-09-28 2008-02-27 赵颜 一种可为手机用户提供设备和数据安全服务的方法
CN101393598A (zh) * 2007-09-21 2009-03-25 希姆通信息技术(上海)有限公司 利用手机摄像头对人脸识别来决定允许开机/解锁的方法
CN101557551A (zh) * 2009-05-11 2009-10-14 成都市华为赛门铁克科技有限公司 一种移动终端访问云服务的方法、装置和通信系统
WO2010022185A1 (en) * 2008-08-19 2010-02-25 Digimarc Corporation Methods and systems for content processing

Family Cites Families (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR19980703120A (ko) 1995-03-20 1998-10-15 조안나 티. 라우 이미지 식별 시스템 및 방법
US7114080B2 (en) * 2000-12-14 2006-09-26 Matsushita Electric Industrial Co., Ltd. Architecture for secure remote access and transmission using a generalized password scheme with biometric features
JP3873793B2 (ja) 2002-03-29 2007-01-24 日本電気株式会社 顔メタデータ生成方法および顔メタデータ生成装置
JP2003346149A (ja) * 2002-05-24 2003-12-05 Omron Corp 顔照合装置および生体情報照合装置
JP4397212B2 (ja) * 2003-02-05 2010-01-13 富士フイルム株式会社 本人認証装置
KR100571813B1 (ko) 2003-07-28 2006-04-17 삼성전자주식회사 로컬 선형변환함수를 이용한 특징벡터 추출방법과 이를이용한 영상인식방법 및 장치
US20050210267A1 (en) * 2004-03-18 2005-09-22 Jun Sugano User authentication method and system, information terminal device and service providing server, subject identification method and system, correspondence confirmation method and system, object confirmation method and system, and program products for them
CN101174948A (zh) * 2006-11-02 2008-05-07 上海银晨智能识别科技有限公司 具有人脸认证的网络登陆系统及登陆方法
US8558663B2 (en) * 2007-11-30 2013-10-15 Bank Of America Corporation Integration of facial recognition into cross channel authentication
CN101216884B (zh) * 2007-12-29 2012-04-18 北京中星微电子有限公司 一种人脸认证的方法及系统
US20100138914A1 (en) * 2008-12-01 2010-06-03 Research In Motion Limited System and method of providing biometric quick launch
JP4636190B2 (ja) * 2009-03-13 2011-02-23 オムロン株式会社 顔照合装置、電子機器、顔照合装置の制御方法、および顔照合装置制御プログラム
JP5230501B2 (ja) * 2009-03-26 2013-07-10 富士フイルム株式会社 認証装置及び認証方法
US8194938B2 (en) 2009-06-02 2012-06-05 George Mason Intellectual Properties, Inc. Face authentication using recognition-by-parts, boosting, and transduction
US8621583B2 (en) * 2010-05-14 2013-12-31 Microsoft Corporation Sensor-based authentication to a computer network-based service
US9030292B2 (en) * 2011-04-18 2015-05-12 Telmate, Llc Interactive audio/video system and device for use in a secure facility

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1629875A (zh) * 2003-12-15 2005-06-22 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
CN1972186A (zh) * 2005-11-24 2007-05-30 中国科学院自动化研究所 一种移动式身份认证系统及其认证方法
CN101393598A (zh) * 2007-09-21 2009-03-25 希姆通信息技术(上海)有限公司 利用手机摄像头对人脸识别来决定允许开机/解锁的方法
CN101132557A (zh) * 2007-09-28 2008-02-27 赵颜 一种可为手机用户提供设备和数据安全服务的方法
WO2010022185A1 (en) * 2008-08-19 2010-02-25 Digimarc Corporation Methods and systems for content processing
CN101557551A (zh) * 2009-05-11 2009-10-14 成都市华为赛门铁克科技有限公司 一种移动终端访问云服务的方法、装置和通信系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015103970A1 (en) * 2014-01-09 2015-07-16 Tencent Technology (Shenzhen) Company Limited Method, apparatus and system for authenticating user
CN105279416A (zh) * 2015-10-27 2016-01-27 上海川织金融信息服务有限公司 基于多生物特征结合设备指纹的身份识别方法及其系统
CN105279416B (zh) * 2015-10-27 2018-05-29 上海川织金融信息服务有限公司 基于多生物特征结合设备指纹的身份识别方法及其系统
CN110012018A (zh) * 2019-04-11 2019-07-12 国网山东省电力公司 一种工业网络安全系统

Also Published As

Publication number Publication date
US8983145B2 (en) 2015-03-17
US20150047005A1 (en) 2015-02-12
US20140250516A1 (en) 2014-09-04
US8861798B2 (en) 2014-10-14
US20160080371A1 (en) 2016-03-17
CN103814545A (zh) 2014-05-21
US20150047006A1 (en) 2015-02-12
US9537859B2 (en) 2017-01-03
CN103814545B (zh) 2016-10-26
US9813909B2 (en) 2017-11-07
US20170070885A1 (en) 2017-03-09
US8989452B2 (en) 2015-03-24

Similar Documents

Publication Publication Date Title
WO2013000142A1 (zh) 手机用户身份认证方法、云服务器以及网络系统
US20120110341A1 (en) Mobile Device Transaction Using Multi-Factor Authentication
KR20190136349A (ko) 3d 영상 기반의 얼굴 인증 방법 및 장치
US11288530B1 (en) Systems and methods for liveness-verified identity authentication
JP2006338092A (ja) パタン照合方法、パタン照合システム及びパタン照合プログラム
US11663306B2 (en) System and method for confirming a person's identity
CN103646201A (zh) 一种人脸组合身份验证方法
US20210194874A1 (en) Privacy-Preserving Biometric Authentication
US9690972B1 (en) Method and apparatus for fingerprint encoding, identification and authentication
WO2023071812A1 (zh) 用于多方安全计算系统的生物特征提取方法及设备
KR20210004319A (ko) 안면 인식을 사용하여 인증을 수행하는 방법, 장치 및 시스템
WO2018059127A1 (zh) 一种安全验证的方法及装置
CN107368722A (zh) 生物图像的验证方法、计算机可读存储介质、移动终端
Kwon et al. CCTV-based multi-factor authentication system
US20130198836A1 (en) Facial Recognition Streamlined Login
Alotaibi et al. Enhancing OAuth services security by an authentication service with face recognition
CN110651268B (zh) 认证用户的方法和电子设备
WO2019245437A1 (en) Method and electronic device for authenticating a user
KR101740574B1 (ko) 확인자의 자필 서명을 이용한 개인 인증 방법 및 이를 활용한 핀테크 시스템
Pakulova et al. Multi-path multimodal authentication system for remote information system
KR101988365B1 (ko) 얼굴 인식 시스템 및 그 방법
CN107292146B (zh) 用户特征向量选取方法及系统
Zakharov et al. Embedding Non-Distortive Cancelable Face Template Generation
TWI850800B (zh) 用於多方安全計算系統的生物特徵提取方法及設備、電腦儲存媒介、電腦程式產品
US20210336951A1 (en) Fusion template for user authentication and vault for storing and using the same

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: 11868828

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 14129135

Country of ref document: US

NENP Non-entry into the national phase

Ref country code: DE

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 15/05/2014)

122 Ep: pct application non-entry in european phase

Ref document number: 11868828

Country of ref document: EP

Kind code of ref document: A1