CN109727350A - A kind of Door-access control method and device based on recognition of face - Google Patents
A kind of Door-access control method and device based on recognition of face Download PDFInfo
- Publication number
- CN109727350A CN109727350A CN201811535230.2A CN201811535230A CN109727350A CN 109727350 A CN109727350 A CN 109727350A CN 201811535230 A CN201811535230 A CN 201811535230A CN 109727350 A CN109727350 A CN 109727350A
- Authority
- CN
- China
- Prior art keywords
- facial image
- user
- identified
- features
- face
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Landscapes
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a kind of Door-access control method and devices based on recognition of face, this method comprises: the facial image of acquisition user, obtains facial image to be identified;Facial image to be identified is identified using the human face recognition model based on deep learning, is judged in facial image to be identified with the presence or absence of five features;If so, extracting the five features vector of facial image to be identified, and five features vector is matched with the five features vector of multiple facial image samples in member database;Matching result is obtained, if matching result instruction user is member, access control is opened;If matching result indicates that user is non-member, gate inhibition's closure is kept, and exports member registration link.Technical solution provided in an embodiment of the present invention is able to solve the problem of access control low efficiency in the prior art.
Description
[technical field]
The present invention relates to technical field of face recognition more particularly to a kind of access control methods and dress based on recognition of face
It sets.
[background technique]
The places such as existing self-service gymnasium, bank, self-help shopping store, are typically provided with access control system, currently, user one
As need brush member card just to can enter.However this entity member card is when in use, and old member is needed to carry, new user is then
Need registered members that can just handle member card.Access control, inefficiency are realized with member card.
Therefore, how to solve access control low efficiency in the prior art becomes a technical problem to be solved urgently.
[summary of the invention]
In view of this, the embodiment of the invention provides a kind of Door-access control method and device based on recognition of face, to
Solve the problems, such as access control low efficiency in the prior art.
To achieve the goals above, according to an aspect of the invention, there is provided a kind of gate inhibition's control based on recognition of face
Method processed, which comprises the facial image for acquiring user obtains facial image to be identified;Using based on deep learning
Human face recognition model identifies the facial image to be identified, judges in the facial image to be identified with the presence or absence of face
Feature;If so, extract the five features vector of the facial image to be identified, and by the five features vector and member's number
It is matched according to the five features vector of multiple facial image samples in library;Matching result is obtained, if the matching result
Indicate that the user is member, access control is opened;If the matching result indicates that the user is non-member, gate inhibition is kept
Closure, and export member registration link.
Further, described that the facial image to be identified is known using the face recognition algorithms based on deep learning
Not, the method for judging to whether there is five features in the facial image to be identified, comprising:
The five features point in the facial image to be identified is identified using active appearance models;Using principal component analysis point
The feature vector of the five features point is not extracted;Using clustering algorithm to the principal component analysis treated feature vector into
Row classification, obtains face data;By the face data train classification models, so that the disaggregated model can judge institute
It states in facial image to be identified with the presence or absence of five features.
Further, identified described using active appearance models five features point in the facial image to be identified it
Before, method further include: acquire multiple facial images of different user;Five features point is demarcated on the multiple facial image,
Obtain the location feature point of face in each facial image;The location feature point of face in each facial image is carried out pair
Together;It is established using the location feature point of face in the multiple facial image after alignment and constitutes face model, the face
Model includes phantom eye, nose model, mouth model, eyebrow model and ear model;The face model of foundation is combined to obtain
Obtain the active appearance models.
Further, the face by multiple facial image samples in the five features vector and member database
Feature vector is matched, comprising: calculates the five features vector and the multiple face of the facial image to be identified one by one
The similarity value of the five features vector of image pattern, wherein each facial image sample is associated with a member;Judgement meter
Whether at least one reaches threshold value to the multiple similarity values calculated;If calculated multiple similarity values are at least
There is one to reach the threshold value, then the corresponding facial image sample of maximum similarity value is confirmed as to the face figure of the user
Picture;Output is used to indicate the instruction information that the user is member;If calculated all similarity values are not up to described
Threshold value, then be confirmed as that it fails to match, and output is used to indicate the instruction information that the user is non-member.
Further, if indicating that the user is non-member in the matching result, gate inhibition's closure is kept, and defeated
Out after member registration link, the method also includes: the registration information of the user is obtained, the registration information includes described
The facial image and personal information of user;The address name in the personal information is extracted, using the address name as user
Label;By the registration information and the user tag associated storage into the member database.
Further, after the registration information for obtaining the user, the method also includes: obtain the user
The bank card account number of input and the cell-phone number bound with the bank card account number;The identifying code for bank card authentication is sent to institute
State cell-phone number;Obtain the check code of user's input;Judge whether the check code is consistent with the identifying code, if one
It causes, confirmation authentication passes through, and confirms that the registration information of the user is real information.
It further, is member when the matching result is the user described, after access control is opened, the side
Method further include: customer manager's propelling user information of Xiang Suoshu user, the user information include user basic information, Yong Huhui
Member grade, the pending business of user.
To achieve the goals above, according to an aspect of the invention, there is provided a kind of gate inhibition's control based on recognition of face
Device processed, described device include: acquisition unit, for acquiring the facial image of user, obtain facial image to be identified;Identification is single
Member, for being identified to the facial image to be identified using the human face recognition model based on deep learning, judgement it is described to
It identifies and whether there is five features in facial image;Matching unit, for when there are face spies in the facial image to be identified
When sign, the five features vector of the facial image to be identified is extracted, and will be in the five features vector and member database
The five features vectors of multiple facial image samples matched;First execution unit, for obtaining matching result, if institute
It states matching result and indicates that the user is member, access control is opened;Second execution unit, if referred to for the matching result
Show that the user is non-member, keep gate inhibition's closure, and exports member registration link.
To achieve the goals above, according to an aspect of the invention, there is provided a kind of computer non-volatile memories are situated between
Matter, the storage medium include the program of storage, wherein equipment where controlling the storage medium in described program operation is held
Access control method based on recognition of face described in the above-mentioned any one of row.
To achieve the goals above, according to an aspect of the invention, there is provided a kind of server, including memory and place
Device is managed, the memory is used to control the execution of program instruction, institute for storing the information including program instruction, the processor
It states when program instruction is loaded and executed by processor and realizes the access control side based on recognition of face described in above-mentioned any one
The step of method.
In the present solution, identifying the user of disengaging gate inhibition by face recognition technology, it is no longer necessary to user shows member card,
After recognition of face matching, access control system is automatically opened for the user for being confirmed as member, promotes user experience;It is not to being confirmed as
The user of member pushes member registration link, can quick user bound, solve asking for prior art access control low efficiency
Topic.
[Detailed description of the invention]
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this field
For those of ordinary skill, without any creative labor, it can also be obtained according to these attached drawings other attached
Figure.
Fig. 1 is a kind of flow chart of access control method based on recognition of face according to an embodiment of the present invention;
Fig. 2 is a kind of schematic diagram of access control device based on recognition of face according to an embodiment of the present invention;
Fig. 3 is a kind of schematic diagram of server according to an embodiment of the present invention.
[specific embodiment]
For a better understanding of the technical solution of the present invention, being retouched in detail to the embodiment of the present invention with reference to the accompanying drawing
It states.
It will be appreciated that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Base
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts it is all its
Its embodiment, shall fall within the protection scope of the present invention.
The term used in embodiments of the present invention is only to be not intended to be limiting merely for for the purpose of describing particular embodiments
The present invention.In the embodiment of the present invention and the "an" of singular used in the attached claims, " described " and "the"
It is also intended to including most forms, unless the context clearly indicates other meaning.
It should be appreciated that term "and/or" used herein is only a kind of incidence relation for describing affiliated partner, indicate
There may be three kinds of relationships, for example, A and/or B, can indicate: individualism A, exist simultaneously A and B, individualism B these three
Situation.In addition, character "/" herein, typicallys represent the relationship that forward-backward correlation object is a kind of "or".
It will be appreciated that though terminal may be described using term first, second, third, etc. in embodiments of the present invention,
But these terminals should not necessarily be limited by these terms.These terms are only used to for terminal being distinguished from each other out.For example, not departing from the present invention
In the case where scope of embodiments, the first diagnostic result can also be referred to as second opinion as a result, similarly, second opinion result
The first diagnostic result can be referred to as.
Depending on context, word as used in this " if " can be construed to " ... when " or " when ...
When " or " in response to determination " or " in response to detection ".Similarly, depend on context, phrase " if it is determined that " or " if detection
(condition or event of statement) " can be construed to " when determining " or " in response to determination " or " when the detection (condition of statement
Or event) when " or " in response to detection (condition or event of statement) ".
Fig. 1 is a kind of flow chart of access control method based on recognition of face according to an embodiment of the present invention, such as Fig. 1 institute
Show, this method comprises:
Step S101 acquires the facial image of user, obtains facial image to be identified.
Step S102 identifies facial image to be identified using the human face recognition model based on deep learning, judges
It whether there is five features in facial image to be identified.Wherein, face characteristic includes eyes, eyebrow, nose, mouth, ear.
Step S103, if so, extract the five features vector of facial image to be identified, and by five features vector and meeting
The five features vector of multiple facial image samples in member's database is matched.
Step S104 obtains matching result, if matching result instruction user is member, access control is opened.
Step S105 keeps gate inhibition's closure, and export member registration chain if matching result instruction user is non-member
It connects.
In the present solution, identifying the user of disengaging gate inhibition by face recognition technology, it is no longer necessary to user shows member card,
After recognition of face matching, access control system is automatically opened for the user for being confirmed as member, promotes user experience;It is not to being confirmed as
The user of member pushes member registration link, can quick user bound, solve asking for prior art access control low efficiency
Topic.
Optionally, facial image to be identified is identified using the human face recognition model based on deep learning, judge to
It identifies and whether there is five features in facial image, comprising:
The five features point in facial image to be identified is identified using active appearance models;
Extract the feature vector of five features point respectively using principal component analysis;
Using clustering algorithm, to principal component analysis, treated that feature vector is classified, and obtains face data;In this reality
It applies in mode, clustering algorithm can be Kmeans method.
By face data train classification models, so that disaggregated model can judge whether deposit in facial image to be identified
In five features.
It is to be appreciated that enabling to matching result more accurate reliable, one come match cognization with human face five-sense-organ image
A little higher places of business secret degree such as bank, insurance company, can effective guarantee member's equity, avoid some intermediaries or
Other non-members enter the region that member enjoys equity.
Optionally, before identifying the five features point in facial image to be identified using active appearance models, method is also
Include:
Acquire multiple facial images of different user.
The characteristic point that face are demarcated on multiple facial images, obtains the location feature point of face in each facial image.
For example, manually demarcating multiple location feature points of eyes, eyebrow, nose, mouth, ear on facial image, wherein eyes
Multiple location feature point difference branch is in positions such as upper lower eyelid boundary, iris boundary, pupil boundaries.Multiple positioning of nose are special
Branch is in positions such as the wing of nose, nose, nasal septum, the nasions respectively for sign point, and multiple location feature points difference branch of mouth is in upper lip
The positions such as contour line, lower lip contour line.
The location feature point of face in each facial image is aligned.Specifically, to the location feature of calibration face
Facial image after point carries out general formula analysis, wherein the center of gravity for calculating separately multiple facial images, by the weight of multiple facial images
Point is aligned, then the difference of rotation angle is calculated by the position of default corresponding points, and then rotating each facial image makes
The corresponding location feature point obtained on facial image is aligned.
It is established using the location feature point of face in multiple facial images after alignment and constitutes face model, face model
Including phantom eye, nose model, mouth model, eyebrow model and ear model.Specifically, by the eyes in each facial image after alignment
Multiple location feature points be connected to form shape vector;Eyes are extracted from the shape vector of eyes using principal component analysis
Feature vector reduces data dimension simultaneously, it will be appreciated that ground can using principal component analysis dimensionality reduction by extracting main feature vector
Reduce computational burden.Then, a mould is constructed and trained to principal component analysis treated feature vector using Kmeans method
Type.Similarly method building nose model, mouth model, eyebrow model and ear model.
The face model of foundation is combined to obtain active appearance models.Active appearance models can be detected wait know
Face image in others' face image and the feature vector for extracting face
Optionally, by the five features vector of multiple facial image samples in five features vector and member database into
Row matching, comprising: calculate the five features vector of facial image to be identified and the five features of multiple facial image samples one by one
The similarity value of vector, wherein each facial image sample is associated with a member;Judge calculated multiple similarity values
Whether at least one reaches threshold value;If at least one reaches threshold value to calculated multiple similarity values, by maximum phase
The facial image of user is confirmed as like the corresponding facial image sample of angle value;Output is used to indicate the instruction that user is member and believes
Breath;If calculated all similarity values are not up to threshold value, it is confirmed as that it fails to match, it is non-that output, which is used to indicate user,
The instruction information of member.It is to be appreciated that passing through the phase for calculating facial image and the feature vector of facial image sample to be identified
Like degree, to confirm whether user is member, the precision and reliability of user's identification can be improved.
Optionally, if being non-member in matching result instruction user, gate inhibition's closure is kept, and export member registration link
Later, method further include: obtain the registration information of user, registration information includes the facial image and personal information of user;It extracts
Address name in personal information, using address name as user tag;By registration information and the extremely meeting of user tag associated storage
In member's database.It is to be appreciated that by automatically extracting address name, and using address name as the registration of label storage user
Information, it is no longer necessary to manually handle registered members, save time, raising efficiency.
Optionally, obtain user registration information after, method further include: obtain user input bank card account number and
With the cell-phone number of bank card account number binding;The identifying code for bank card authentication is sent to cell-phone number;Obtain the school of user's input
Test code;It is whether consistent with identifying code to judge check code, if unanimously, confirmation authentication passes through, and confirms the registration letter of user
Breath is real information.Whether the registration information by bank card authentication verification user is true, ensures the authenticity of user information.
It optionally, is being member when matching result is user, after access control is opened, method further include: to user's
Customer manager's propelling user information, user information include user basic information, user's membership grade, the pending business of user.To
It allows users to quickly handle required business, such as handles social security, purchase insurance etc..If membership grade be it is advanced, accordingly
Customer manager personal VIP service can be provided, promote customer experience.
The embodiment of the invention provides a kind of access control device based on recognition of face, the device is for executing above-mentioned base
In the access control method of recognition of face, as shown in Fig. 2, the device includes: acquisition unit 10, recognition unit 20, matching unit
30, the first execution unit 40 and the second execution unit 50.
Acquisition unit 10 obtains facial image to be identified for acquiring the facial image of user.
Recognition unit 20, for being known using the human face recognition model based on deep learning to facial image to be identified
Not, judge in facial image to be identified with the presence or absence of five features.Wherein, face characteristic include eyes, eyebrow, nose, mouth,
Ear.
Matching unit 30, for when, there are when five features, extracting facial image to be identified in facial image to be identified
Five features vector, and by the five features vector of multiple facial image samples in five features vector and member database into
Row matching.
First execution unit 40, for obtaining matching result, if matching result instruction user is member, access control is opened
It opens.
Second execution unit 50 keeps gate inhibition's closure, and export meeting if being non-member for matching result instruction user
Member's login link.
In the present solution, identifying the user of disengaging gate inhibition by face recognition technology, it is no longer necessary to user shows member card,
After recognition of face matching, access control system is automatically opened for the user for being confirmed as member, promotes user experience;It is not to being confirmed as
The user of member pushes member registration link, can quick user bound, solve asking for prior art access control low efficiency
Topic.
Optionally, recognition unit 20 includes identification subelement, extracts subelement, classification subelement, training subelement.
Subelement is identified, for identifying the five features point in facial image to be identified using active appearance models;
Subelement is extracted, for extracting the feature vector of five features point respectively using principal component analysis;
Optionally, classification subelement is also performed the steps of when program instruction is loaded and executed by processor, for using
Treated that feature vector is classified to principal component analysis for clustering algorithm, obtains face data;In the present embodiment, it clusters
Algorithm can be Kmeans method.
Training subelement, for by face data train classification models so that disaggregated model can judge it is to be identified
It whether there is five features in facial image.
It is to be appreciated that enabling to matching result more accurate reliable, one come match cognization with human face five-sense-organ image
A little higher places of business secret degree such as bank, insurance company, can effective guarantee member's equity, avoid some intermediaries or
Other non-members enter the region that member enjoys equity.
Optionally, recognition unit 20 further includes acquisition subelement, calibration subelement, alignment subelement, building subelement, knot
Zygote unit:
Subelement is acquired, for acquiring multiple facial images of different user.
Subelement is demarcated, for demarcating the characteristic point of face on multiple facial images, is obtained five in each facial image
The location feature point of official.For example, manually demarcating multiple positioning spy of eyes, eyebrow, nose, mouth, ear on facial image
Levy point, wherein multiple location feature points difference branch of eyes is in positions such as upper lower eyelid boundary, iris boundary, pupil boundaries.
Multiple location feature points difference branch of nose is in the positions such as the wing of nose, nose, nasal septum, the nasion, multiple location features of mouth
Point difference branch is in positions such as upper lip contour line, lower lip contour lines.
It is aligned subelement, for the location feature point of face in each facial image to be aligned.Specifically, to calibration
Facial image after the location feature point of face carries out general formula analysis, wherein the center of gravity for calculating separately multiple facial images, it will be more
The emphasis of a facial image is aligned, then the difference of rotation angle is calculated by the position of default corresponding points, is then rotated
Each facial image is aligned the corresponding location feature point on facial image.
Subelement is constructed, establishes composition five for the location feature point using face in multiple facial images after alignment
Official's model, face model include phantom eye, nose model, mouth model, eyebrow model and ear model.It specifically, will be each of after alignment
Multiple location feature points of eyes in facial image are connected to form shape vector;Using principal component analysis from the shape of eyes
The feature vector that eyes are extracted in vector reduces data dimension simultaneously, it will be appreciated that ground is used by extracting main feature vector
PCA dimensionality reduction can reduce computational burden.Then, principal component analysis treated feature vector is constructed simultaneously using Kmeans method
Training phantom eye.Similarly method building nose model, mouth model, eyebrow model and ear model.
In conjunction with subelement, active appearance models are obtained for the face model of foundation to be combined.Active appearance mould
Type can detect the face image in facial image to be identified and extract the feature vector of face
Optionally, matching unit 30 includes computation subunit, judgment sub-unit, the first confirmation subelement and the second confirmation
Unit.
Computation subunit, for calculating the five features vector and multiple facial image samples of facial image to be identified one by one
Five features vector similarity value, wherein each facial image sample is associated with a member.Judgment sub-unit is used
In judging whether at least one reaches threshold value to calculated multiple similarity values;First confirmation subelement, if for calculating
At least one reaches threshold value to multiple similarity values out, then is confirmed as using by the corresponding facial image sample of maximum similarity value
The facial image at family;Output is used to indicate the instruction information that user is member;Second confirmation subelement, if for calculated
All similarity values are not up to threshold value, then are confirmed as that it fails to match, and output is used to indicate the instruction information that user is non-member.
It is to be appreciated that being come by the similarity for the feature vector for calculating facial image and facial image sample to be identified
Confirm whether user is member, the precision and reliability of user's identification can be improved.
Optionally, device further includes first acquisition unit, extraction unit and storage unit.
First acquisition unit, for obtaining the registration information of user, registration information includes facial image and the individual of user
Information;Extraction unit, for extracting the address name in personal information, using address name as user tag;Storage unit is used
In by registration information and user tag associated storage into member database.It is to be appreciated that by automatically extracting address name,
And using address name as the registration information of label storage user, it is no longer necessary to manually handle registered members, save the time, be promoted
Efficiency.
Optionally, device further includes second acquisition unit, transmission unit, third acquiring unit, judging unit.
Second acquisition unit, the cell-phone number for obtaining the bank card account number of user's input and with bank card account number binding;
Transmission unit, for sending the identifying code authenticated for bank card to cell-phone number;Third acquiring unit, for obtaining user's input
Check code;Judging unit, it is whether consistent with identifying code for judging check code, if unanimously, confirmation authentication passes through, and
The registration information for confirming user is real information.Whether the registration information by bank card authentication verification user is true, ensures and uses
The authenticity of family information.
Optionally, device further include: push unit, for customer manager's propelling user information to user, user information
Including user basic information, user's membership grade, the pending business of user.It enables a user to quickly handle required business,
Such as social security is handled, purchase insurance etc..If membership grade be it is advanced, corresponding customer manager can provide personal VIP service,
Promote customer experience.
The embodiment of the invention provides a kind of computer non-volatile memory medium, storage medium includes the program of storage,
Wherein, when program is run, equipment where control storage medium executes following steps:
The facial image for acquiring user, obtains facial image to be identified;Using the human face recognition model based on deep learning
Facial image to be identified is identified, is judged in facial image to be identified with the presence or absence of five features;If so, extracting wait know
The five features vector of others' face image, and by five of multiple facial image samples in five features vector and member database
Official's feature vector matches;Matching result is obtained, if matching result instruction user is member, access control is opened;If
Matching result indicates that user is non-member, keeps gate inhibition's closure, and exports member registration link.
Optionally, when program is run, equipment where control storage medium also executes following steps: using active appearance mould
Type identifies the five features point in facial image to be identified;Extracted respectively using principal component analysis the feature of five features point to
Amount;Using clustering algorithm, to principal component analysis, treated that feature vector is classified, and obtains face data;Pass through face data
Train classification models, so that disaggregated model can judge in facial image to be identified with the presence or absence of five features.
Optionally, when program is run, equipment where control storage medium also executes following steps: calculating one by one to be identified
The similarity value of the feature vector of facial image and the feature vector of multiple facial image samples, wherein each facial image sample
This is associated with a member;Judge whether at least one reaches threshold value to calculated multiple similarity values;If calculated
Multiple similarity values at least one reach threshold value, then the corresponding facial image sample of maximum similarity value is confirmed as user
Facial image;Output is used to indicate the instruction information that user is member;If calculated all similarity values are not up to
Threshold value, then be confirmed as that it fails to match, and output is used to indicate the instruction information that user is non-member.It is to be appreciated that passing through calculating
The similarity of the feature vector of facial image to be identified and facial image sample can be improved to confirm whether user is member
The precision and reliability of user's identification.
Optionally, when program is run, equipment where control storage medium also executes following steps: obtaining the registration of user
Information, registration information include the facial image and personal information of user;The address name in personal information is extracted, by address name
As user tag;By registration information and user tag associated storage into member database.It is to be appreciated that by mentioning automatically
Address name is taken, and using address name as the registration information of label storage user, it is no longer necessary to manually handle registered members, save
It makes an appointment, raising efficiency.
Optionally, when program is run, equipment where control storage medium also executes following steps: obtaining user's input
Bank card account number and the cell-phone number bound with bank card account number;The identifying code for bank card authentication is sent to cell-phone number;It obtains
The check code of user's input;It is whether consistent with identifying code to judge check code, if unanimously, confirmation authentication passes through, and confirms
The registration information of user is real information.Whether the registration information by bank card authentication verification user is true, ensures user's letter
The authenticity of breath.
The embodiment of the invention provides a kind of server 100, including memory 101 and processor 102, memory 101 is used
In the information that storage includes program instruction 103, processor 102 is used to control the execution of program instruction 103, and program instruction is processed
Device is loaded and is performed the steps of when executing
The facial image for acquiring user, obtains facial image to be identified;Using the human face recognition model based on deep learning
Facial image to be identified is identified, is judged in facial image to be identified with the presence or absence of five features;If so, extracting wait know
The five features vector of others' face image, and by five of multiple facial image samples in five features vector and member database
Official's feature vector matches;Matching result is obtained, if matching result instruction user is member, access control is opened;If
Matching result indicates that user is non-member, keeps gate inhibition's closure, and exports member registration link.
Optionally, it also performs the steps of when program instruction is loaded and executed by processor and is known using active appearance models
Five features point in facial image not to be identified;Extract the feature vector of five features point respectively using principal component analysis;It adopts
With clustering algorithm, to principal component analysis, treated that feature vector is classified, and obtains face data;Pass through the training of face data
Disaggregated model, so that disaggregated model can judge in facial image to be identified with the presence or absence of five features.
Optionally, it is also performed the steps of when program instruction is loaded and executed by processor and calculates face to be identified one by one
The similarity value of the feature vector of the feature vector of image and multiple facial image samples, wherein each facial image sample with
One member is associated;Judge whether at least one reaches threshold value to calculated multiple similarity values;If calculated more
At least one reaches threshold value to a similarity value, then the corresponding facial image sample of maximum similarity value is confirmed as to the people of user
Face image;Output is used to indicate the instruction information that user is member;If calculated all similarity values are not up to threshold value,
Then it is confirmed as that it fails to match, output is used to indicate the instruction information that user is non-member.It is to be appreciated that be identified by calculating
User's knowledge can be improved to confirm whether user is member in the similarity of the feature vector of facial image and facial image sample
Other precision and reliability.
Optionally, the registration information for obtaining user is also performed the steps of when program instruction is loaded and executed by processor,
Registration information includes the facial image and personal information of user;Extract personal information in address name, using address name as
User tag;By registration information and user tag associated storage into member database.It is to be appreciated that by automatically extracting use
Family name, and using address name as the registration information of label storage user, it is no longer necessary to registered members are manually handled, when saving
Between, raising efficiency.
Optionally, the bank for obtaining user's input is also performed the steps of when program instruction is loaded and executed by processor
Card account and the cell-phone number bound with bank card account number;The identifying code for bank card authentication is sent to cell-phone number;Obtain user
The check code of input;It is whether consistent with identifying code to judge check code, if unanimously, confirmation authentication passes through, and confirms user
Registration information be real information.Whether the registration information by bank card authentication verification user is true, ensures user information
Authenticity.
It should be noted that terminal involved in the embodiment of the present invention can include but is not limited to personal computer
(Personal Computer, PC), personal digital assistant (Personal Digital Assistant, PDA), wireless handheld
Equipment, tablet computer (Tablet Computer), mobile phone, MP3 player, MP4 player etc..
It is understood that using the application program (nativeApp) that can be mounted in terminal, or can also be
One web page program (webApp) of the browser in terminal, the embodiment of the present invention is to this without limiting.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In several embodiments provided by the present invention, it should be understood that disclosed system, device and method can be with
It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the division of unit,
Only a kind of logical function partition, there may be another division manner in actual implementation, for example, multiple units or components can be with
In conjunction with or be desirably integrated into another system, or some features can be ignored or not executed.Another point, it is shown or discussed
Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING of device or unit or
Communication connection can be electrical property, mechanical or other forms.
Unit may or may not be physically separated as illustrated by the separation member, shown as a unit
Component may or may not be physical unit, it can and it is in one place, or may be distributed over multiple networks
On unit.It can some or all of the units may be selected to achieve the purpose of the solution of this embodiment according to the actual needs.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of hardware adds SFU software functional unit.
The above-mentioned integrated unit being realized in the form of SFU software functional unit can store and computer-readable deposit at one
In storage media.Above-mentioned SFU software functional unit is stored in a storage medium, including some instructions are used so that a computer
It is each that device (can be personal computer, server or network equipment etc.) or processor (Processor) execute the present invention
The part steps of embodiment method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (Read-Only
Memory, ROM), random access memory (Random Access Memory, RAM), magnetic or disk etc. is various to deposit
Store up the medium of program code.
The above is merely preferred embodiments of the present invention, be not intended to limit the invention, it is all in spirit of the invention and
Within principle, any modification, equivalent substitution, improvement and etc. done be should be included within the scope of the present invention.
Claims (10)
1. a kind of access control method based on recognition of face, which is characterized in that the described method includes:
The facial image for acquiring user, obtains facial image to be identified;
The facial image to be identified is identified using the face recognition algorithms based on deep learning, is judged described to be identified
It whether there is five features in facial image;
If so, extract the five features vector of the facial image to be identified, and by the five features vector and member's number
It is matched according to the five features vector of multiple facial image samples in library;
Matching result is obtained, if the matching result indicates that the user is member, access control is opened;
If the matching result indicates that the user is non-member, gate inhibition's closure is kept, and exports member registration link.
2. the method according to claim 1, wherein described use the face recognition algorithms pair based on deep learning
The facial image to be identified is identified that the method for judging to whether there is five features in the facial image to be identified is wrapped
It includes:
The five features point in the facial image to be identified is identified using active appearance models;
Extract the feature vector of the five features point respectively using principal component analysis;
Using clustering algorithm, to the principal component analysis, treated that feature vector is classified, and obtains face data;
By the face data train classification models, so that the disaggregated model can judge the facial image to be identified
In whether there is five features.
3. according to the method described in claim 2, it is characterized in that, described described to be identified using active appearance models identification
Before five features point in facial image, method further include:
Acquire multiple facial images of different user;
Five features point is demarcated on the multiple facial image, obtains the location feature point of face in each facial image;
The location feature point of face in each facial image is aligned;
It is established using the location feature point of face in the multiple facial image after alignment and constitutes face model, the face
Model includes phantom eye, nose model, mouth model, eyebrow model and ear model;
The face model of foundation is combined to obtain the active appearance models.
4. the method according to claim 1, wherein described will be in the five features vector and member database
The five features vectors of multiple facial image samples matched, comprising:
The five features vector of the facial image to be identified and the five features of the multiple facial image sample are calculated one by one
The similarity value of vector, wherein each facial image sample is associated with a member;
Judge whether at least one reaches threshold value to calculated multiple similarity values;
If at least one reaches the threshold value to calculated multiple similarity values, and maximum similarity value is corresponding
Facial image sample is confirmed as the facial image of the user;Output is used to indicate the instruction information that the user is member;
If calculated all similarity values are not up to the threshold value, it is confirmed as that it fails to match, output is used to indicate institute
State the instruction information that user is non-member.
5. if the method according to claim 1, wherein indicate that the user is in the matching result
Non-member, keep gate inhibition closure, and export member registration link after, the method also includes:
The registration information of the user is obtained, the registration information includes the facial image and personal information of the user;
The address name in the personal information is extracted, using the address name as user tag;
By the registration information and the user tag associated storage into the member database.
6. according to the method described in claim 5, it is characterized in that, after the registration information for obtaining the user, institute
State method further include:
Obtain the bank card account number of user's input and the cell-phone number with bank card account number binding;
The identifying code for bank card authentication is sent to the cell-phone number;
Obtain the check code of user's input;
It is whether consistent with the identifying code to judge the check code, if unanimously, confirmation authentication passes through, and confirms the use
The registration information at family is real information.
7. method according to any one of claims 1 to 6, which is characterized in that described when the matching result is described
User is member, after access control is opened, the method also includes:
To customer manager's propelling user information of the user, the user information includes user basic information, user member etc.
Grade, the pending business of user.
8. a kind of access control device based on recognition of face, which is characterized in that described device includes:
Acquisition unit obtains facial image to be identified for acquiring the facial image of user;
Recognition unit, for being identified using the human face recognition model based on deep learning to the facial image to be identified,
Judge in the facial image to be identified with the presence or absence of five features;
Matching unit, for when there are when five features, extracting the facial image to be identified in the facial image to be identified
Five features vector, and by the five features of multiple facial image samples in the five features vector and member database
Vector is matched;
First execution unit, for obtaining matching result, if the matching result indicates that the user is member, access control
It opens;
Second execution unit keeps gate inhibition's closure, and export if indicating that the user is non-member for the matching result
Member registration link.
9. a kind of storage medium, which is characterized in that the storage medium includes the program of storage, wherein run in described program
When control the storage medium where equipment perform claim require any one of 1 to 7 described in gate inhibition's control based on recognition of face
Method processed.
10. a kind of server, including memory and processor, the memory is for storing the information including program instruction, institute
Processor is stated for controlling the execution of program instruction, it is characterised in that: described program instruction is real when being loaded and executed by processor
The step of showing the access control method described in claim 1 to 7 any one based on recognition of face.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811535230.2A CN109727350A (en) | 2018-12-14 | 2018-12-14 | A kind of Door-access control method and device based on recognition of face |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811535230.2A CN109727350A (en) | 2018-12-14 | 2018-12-14 | A kind of Door-access control method and device based on recognition of face |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109727350A true CN109727350A (en) | 2019-05-07 |
Family
ID=66297572
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811535230.2A Pending CN109727350A (en) | 2018-12-14 | 2018-12-14 | A kind of Door-access control method and device based on recognition of face |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109727350A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110866469A (en) * | 2019-10-30 | 2020-03-06 | 腾讯科技(深圳)有限公司 | Human face facial features recognition method, device, equipment and medium |
CN111028390A (en) * | 2019-11-29 | 2020-04-17 | 浙江威欧希科技股份有限公司 | Intelligent lock and face recognition optimization method applied to intelligent lock |
CN111325132A (en) * | 2020-02-17 | 2020-06-23 | 深圳龙安电力科技有限公司 | Intelligent monitoring system |
CN111651742A (en) * | 2020-04-29 | 2020-09-11 | 华为技术有限公司 | Method, electronic equipment and system for verifying user identity |
CN111798603A (en) * | 2020-09-08 | 2020-10-20 | 腾讯科技(深圳)有限公司 | Access control method and device, computer equipment and storage medium |
CN111932754A (en) * | 2019-08-19 | 2020-11-13 | 北京戴纳实验科技有限公司 | Laboratory entrance guard verification system and verification method |
CN112084904A (en) * | 2020-08-26 | 2020-12-15 | 武汉普利商用机器有限公司 | Face searching method, device and storage medium |
CN112669509A (en) * | 2020-12-11 | 2021-04-16 | 深圳市航天华拓科技有限公司 | Access control management method, system, electronic device and storage medium |
CN113591782A (en) * | 2021-08-12 | 2021-11-02 | 北京惠朗时代科技有限公司 | Training-based face recognition intelligent safety box application method and system |
CN110866469B (en) * | 2019-10-30 | 2024-07-16 | 腾讯科技(深圳)有限公司 | Facial five sense organs identification method, device, equipment and medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104899579A (en) * | 2015-06-29 | 2015-09-09 | 小米科技有限责任公司 | Face recognition method and face recognition device |
CN206348809U (en) * | 2016-08-26 | 2017-07-21 | 浙江维融电子科技股份有限公司 | A kind of face identification system for bank VIP personnel's business handling |
CN107679546A (en) * | 2017-08-17 | 2018-02-09 | 平安科技(深圳)有限公司 | Face image data acquisition method, device, terminal device and storage medium |
CN107978044A (en) * | 2017-11-29 | 2018-05-01 | 南京甄视智能科技有限公司 | Based on face recognition technology and the high-precision identifying system of RFID technique and recognition methods |
CN108364422A (en) * | 2018-02-24 | 2018-08-03 | 广州逗号智能零售有限公司 | Self-service method and device |
CN207833591U (en) * | 2017-12-29 | 2018-09-07 | 深圳正品创想科技有限公司 | A kind of access control system and unmanned shop |
-
2018
- 2018-12-14 CN CN201811535230.2A patent/CN109727350A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104899579A (en) * | 2015-06-29 | 2015-09-09 | 小米科技有限责任公司 | Face recognition method and face recognition device |
CN206348809U (en) * | 2016-08-26 | 2017-07-21 | 浙江维融电子科技股份有限公司 | A kind of face identification system for bank VIP personnel's business handling |
CN107679546A (en) * | 2017-08-17 | 2018-02-09 | 平安科技(深圳)有限公司 | Face image data acquisition method, device, terminal device and storage medium |
CN107978044A (en) * | 2017-11-29 | 2018-05-01 | 南京甄视智能科技有限公司 | Based on face recognition technology and the high-precision identifying system of RFID technique and recognition methods |
CN207833591U (en) * | 2017-12-29 | 2018-09-07 | 深圳正品创想科技有限公司 | A kind of access control system and unmanned shop |
CN108364422A (en) * | 2018-02-24 | 2018-08-03 | 广州逗号智能零售有限公司 | Self-service method and device |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111932754A (en) * | 2019-08-19 | 2020-11-13 | 北京戴纳实验科技有限公司 | Laboratory entrance guard verification system and verification method |
CN110866469A (en) * | 2019-10-30 | 2020-03-06 | 腾讯科技(深圳)有限公司 | Human face facial features recognition method, device, equipment and medium |
CN110866469B (en) * | 2019-10-30 | 2024-07-16 | 腾讯科技(深圳)有限公司 | Facial five sense organs identification method, device, equipment and medium |
CN111028390A (en) * | 2019-11-29 | 2020-04-17 | 浙江威欧希科技股份有限公司 | Intelligent lock and face recognition optimization method applied to intelligent lock |
CN111028390B (en) * | 2019-11-29 | 2021-10-22 | 浙江威欧希科技股份有限公司 | Intelligent lock and face recognition optimization method applied to intelligent lock |
CN111325132A (en) * | 2020-02-17 | 2020-06-23 | 深圳龙安电力科技有限公司 | Intelligent monitoring system |
CN111651742A (en) * | 2020-04-29 | 2020-09-11 | 华为技术有限公司 | Method, electronic equipment and system for verifying user identity |
CN112084904A (en) * | 2020-08-26 | 2020-12-15 | 武汉普利商用机器有限公司 | Face searching method, device and storage medium |
CN111798603A (en) * | 2020-09-08 | 2020-10-20 | 腾讯科技(深圳)有限公司 | Access control method and device, computer equipment and storage medium |
CN112669509A (en) * | 2020-12-11 | 2021-04-16 | 深圳市航天华拓科技有限公司 | Access control management method, system, electronic device and storage medium |
CN113591782A (en) * | 2021-08-12 | 2021-11-02 | 北京惠朗时代科技有限公司 | Training-based face recognition intelligent safety box application method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109727350A (en) | A kind of Door-access control method and device based on recognition of face | |
WO2019120115A1 (en) | Facial recognition method, apparatus, and computer apparatus | |
CN108009521B (en) | Face image matching method, device, terminal and storage medium | |
CN107742100B (en) | A kind of examinee's auth method and terminal device | |
CN105069622B (en) | A kind of face recognition payment system and method for facing moving terminal | |
WO2020155939A1 (en) | Image recognition method and device, storage medium and processor | |
WO2019200781A1 (en) | Receipt recognition method and device, and storage medium | |
WO2019090769A1 (en) | Human face shape recognition method and apparatus, and intelligent terminal | |
CN112328999B (en) | Double-recording quality inspection method and device, server and storage medium | |
CN106295482B (en) | A kind of update method and device of face database | |
CN108197557A (en) | Testimony of a witness consistency check method, terminal device and computer readable storage medium | |
CN107093066A (en) | Service implementation method and device | |
CN112364803B (en) | Training method, terminal, equipment and storage medium for living body identification auxiliary network | |
JP5785667B1 (en) | Person identification system | |
CN111639584A (en) | Risk identification method and device based on multiple classifiers and computer equipment | |
CN107103218A (en) | A kind of service implementation method and device | |
WO2017217314A1 (en) | Response device, response system, response method, and recording medium | |
CN108549848A (en) | Method and apparatus for output information | |
JP2020526835A (en) | Devices and methods that dynamically identify a user's account for posting images | |
WO2019178753A1 (en) | Payment method, device and system | |
CN107491757A (en) | A kind of antifraud system and control method based on living body characteristics | |
CN110415113A (en) | Finance data processing method, device, server and readable storage medium storing program for executing | |
CN109145786A (en) | A kind of image identification method, device, equipment, medium and product | |
CN109670420A (en) | Store the control method and device, storage terminal, electronic equipment, medium of terminal | |
CN113591603A (en) | Certificate verification method and device, electronic equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190507 |
|
RJ01 | Rejection of invention patent application after publication |