CN108875646A - A kind of real face image and identity card registration is dual compares authentication method and system - Google Patents
A kind of real face image and identity card registration is dual compares authentication method and system Download PDFInfo
- Publication number
- CN108875646A CN108875646A CN201810649673.8A CN201810649673A CN108875646A CN 108875646 A CN108875646 A CN 108875646A CN 201810649673 A CN201810649673 A CN 201810649673A CN 108875646 A CN108875646 A CN 108875646A
- Authority
- CN
- China
- Prior art keywords
- image
- face
- feature vector
- registration
- face image
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
Abstract
A kind of real face image provided by the embodiments of the present application and identity card registration is dual compares authentication method, including：Obtain the current face image and registration face-image of target body；Feature extraction is carried out to the current face image and the registration face-image respectively, generates corresponding texture feature vector；It is studied just using texture feature vector of the preset correction model to the registration face-image, generates amendment feature vector；The texture feature vector of the current face image and the diversity factor of the amendment feature vector are calculated, judges whether the value of the diversity factor is less than preset threshold；When the value of the diversity factor is less than preset threshold, determine that the current face image and the registration face-image match.By being modified during challenge to registration face-image, then matching verifying is carried out with current face image, reduce the diversity factor of registration face-image and current face image, to improve the speed that automatic face identifies.
Description
Technical field
This application involves image identification technical field more particularly to a kind of real face images and identity card to register dual ratio
To authentication method and system.
Background technique
In occasions such as hotel accommodations, airport security, real name bookings, identity card will be generally registered, someone holds in order to prevent
Other people identity card or personation identity card carries out identity camouflage, also to check whether people is consistent with card.
Currently, other than manually checking the real face image of holder can also be shot, and above mention from identity card
It takes registration face-image or remotely transfers registration face-image from public security data base manipulation ID card No., and then by true face
Portion's image carries out automatic face recognition with registration face-image and compares.
But, it is only just resurveyed in change certification, correction since the registration face-image of identity card is general, image is more
New frequency is very low, and hysteresis quality is obvious, changes as time goes by with body, the real face image of holder with step on
Obvious deviation has occurred in note face-image, this compares to automatic face identification and produces many obstacles：Such as when true
When face-image and registration face-image deviate larger, even if the testimony of a witness, which is consistent, can also occur false alarm；In another example some are looked into automatically
Check system extracts facial characteristics with matching verifying is carried out with complicated correction algorithm to reduce the mistake when testimony of a witness is consistent by surplus
Alarm rate, but it reduce recognition speeds, increase hardware and software cost, are unfavorable for quickly checking.
Summary of the invention
In view of this, the purpose of the application is to propose a kind of real face image and identity card registration is dual compares certification
Method and system, come solve in the prior art due to the registration face-image renewal frequency of identity card it is low caused by test in identity
Automatic face identification comparison speed is slow during card, and identification error rate is high, is unfavorable for the technical issues of quickly checking, to improve
The speed and accuracy of identity card registration and real human face image recognition matching process.
Based on above-mentioned purpose, in the one aspect of the application, proposes a kind of real face image and identity card registration is double
Authentication method is compared again, including：
Obtain the current face image and registration face-image of target body；
Feature extraction is carried out to the current face image and the registration face-image respectively, it is special to generate corresponding texture
Levy vector；
It is studied just using texture feature vector of the preset correction model to the registration face-image, generates amendment
Feature vector；
The texture feature vector of the current face image and the diversity factor of the amendment feature vector are calculated, described in judgement
Whether the value of diversity factor is less than preset threshold；
When the value of the diversity factor is less than preset threshold, the current face image and the registration face-image are determined
Match.
In some embodiments, described feature is carried out to the current face image and the registration face-image respectively to mention
It takes, generates corresponding texture feature vector, including：
Edge detection is carried out respectively to the current face image and the registration face-image, according to the number of closure edge
The current face image and the registration face-image are divided into multiple regions by amount, carry out texture to each region
Identification extracts human face characteristic value, generates the texture feature vector and the registration face-image of the current face image
Texture feature vector.
In some embodiments, described feature is carried out to the current face image and the registration face-image respectively to mention
It takes, including：Edge is carried out to the current face image and the registration face-image using canny edge detection operator respectively
Detection is extracted the image-region surrounded by closed edge, is specifically included：
Convolution is made with Gauss mask respectively to the current face image and the registration face-image, works as front to described
Portion's image and the registration face-image are smoothed；
Each of described current face image and the registration face-image after calculating smoothing processing using Sobel operator
The gradient of pixel；
Retain the maximum of gradient intensity on each pixel of the current face image and the registration face-image,
Delete other values；
Set the maximum of gradient intensity on each pixel of the current face image and the registration face-image
The threshold value upper bound and threshold value lower bound, by the maximum of gradient intensity be greater than the threshold value upper bound pixel be confirmed as boundary, will
The maximum of gradient intensity is greater than the threshold value lower bound and is confirmed as weak boundary less than the pixel in the threshold value upper bound, and gradient is strong
The pixel that the maximum of degree is less than the threshold value lower bound is confirmed as non-boundary；
The weak boundary being connected with the boundary is confirmed into boundary, other weak boundaries are confirmed as non-boundary.
In some embodiments, the preset correction model includes：
The parameter group (α 1, α 2, α 3 ... α n) that the texture feature vector of the registration face-image is modified.
In some embodiments, it is described using preset correction model to it is described registration face-image textural characteristics
Vector is studied just, and amendment feature vector is generated, including：
Using preset correction model (α 1, α 2, α 3 ... α n) to it is described registration face-image textural characteristics to
Amount (D1, D2, D3 ... Dn) is modified, and generates amendment feature vector (D1* (1+ α 1), D2* (1+ α 2), D3* (1+ α 3) ...
Dn*(1+αn))。
In some embodiments, calculated using following formula the texture feature vector of the current face image with it is described
The diversity factor for correcting feature vector, judges whether the value of the diversity factor is less than preset threshold：
Wherein, the texture feature vector of current face image is (R1, R2, R3 ... Rn), and I is diversity factor.
It in some embodiments, further include using following formula to institute when the value of the diversity factor is less than preset threshold
Correction model is stated to be updated：
Wherein α itIndicate the updated corrected parameter of epicycle, α it-1Indicate that epicycle updates pervious corrected parameter, i value
Range is 1 to n.
In some embodiments, the corrected parameter is the corrected parameter for meeting preset threshold range, the preset threshold
Range is (0.5,1.5).
In some embodiments, the parameter group in the correction model has number weight parameter group (δ 1, δ 2, δ 3 ... δ
N), when the texture feature vector (D1, D2, D3 ... Dn) to the registration face-image is modified, the amendment of generation is special
Levying vector is (D1* (1+ δ 1* α 1), D2* (1+ δ 2* α 2), D3* (1+ δ 3* α 3) ... Dn* (1+ δ n* α n)), wherein number power
The value range of δ n in weight parameter group (δ 1, δ 2, δ 3 ... δ n) is 0.5-1.5, and the value of δ n is gradually increased with times of revision.
Based on above-mentioned purpose, in further aspect of the application, proposes a kind of real face image and identity card is registered
Dual comparison Verification System, including：
Image collection module, for obtaining the current face image and registration face-image of target body；
Texture feature vector generation module, for being carried out respectively to the current face image and the registration face-image
Feature extraction generates corresponding texture feature vector；
Texture feature vector correction module, for the line using preset correction model to the registration face-image
It manages feature vector education just, generates amendment feature vector；
Diversity factor determining module, for calculate the current face image texture feature vector and the amendment feature to
The diversity factor of amount, judges whether the value of the diversity factor is less than preset threshold；
As a result output module, for determining the current face image when the value of the diversity factor is less than preset threshold
Match with the registration face-image.
Real face image provided by the embodiments of the present application and identity card registration it is dual compare authentication method and system, every time
Verifying is first modified registration face-image with correction model, then carries out matching verifying with current face image, and pass through
In the case where assert that the testimony of a witness is consistent during challenge, correction model is carried out using the real face image currently acquired
It updates, so that the diversity factor of registration face-image and current face image is reduced, to improve the speed of automatic face identification
Degree.
Detailed description of the invention
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the application's is other
Feature, objects and advantages will become more apparent upon：
Fig. 1 is that the real face image of the embodiment of the present application one and identity card register the dual process for comparing authentication method
Figure；
Fig. 2 is that the real face image of the embodiment of the present application two and identity card register the dual process for comparing authentication method
Figure；
Fig. 3 is that the real face image of the embodiment of the present application three and identity card register the dual structure for comparing Verification System
Figure；
Fig. 4 is to register the dual Verification System that compares using the real face image and identity card of the embodiment of the present application four to carry out
The schematic diagram of challenge.
Specific embodiment
The application is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to
Convenient for description, part relevant to related invention is illustrated only in attached drawing.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
As one embodiment of the application, as shown in Figure 1, being the real face image and identity of the embodiment of the present application one
Card registers the flow chart of dual comparison authentication method.It can be seen from the figure that real face image provided in this embodiment and body
Part card registers dual comparison authentication method, includes the following steps：
S101：Obtain the current face image and registration face-image of target body.
In the present embodiment, the current face image that target body can be acquired by image capture device, for example, working as mesh
Human body is marked in occasions such as hotel accommodations, airport security, real name bookings, the current of the target body can be shot by camera
Face-image, due to needing to show the certificates such as identity card in occasions such as hotel accommodations, airport security, real name bookings, can be with
Registration face-image is extracted from identity card or remotely transfers stepping on for target body from public security data base manipulation ID card No.
Remember face-image.In the present embodiment, the target body is the human body for carrying out challenge, such as carries out identity to human body A
Examination, then human body A is target body.It should be noted that examples detailed above is intended to illustratively to how obtaining target person
The current face image and registration face-image of body are illustrated, and are understood not to the limit to technical scheme
It is fixed.
S102：Feature extraction is carried out to the current face image and the registration face-image respectively, is generated corresponding
Texture feature vector.
In the present embodiment, it in the current face image and registration face-image for getting target body, needs to acquisition
Current face image and registration face-image to target body carry out feature extraction, to generate corresponding texture feature vector.
Specifically, it can use edge detection algorithm, above-mentioned face-image be divided by multiple regions according to the quantity of closure edge, so
Texture recognition is carried out to each region again afterwards, extracts the characteristic value of human face, the texture for generating the current face image is special
Levy the texture feature vector that face-image is registered described in vector sum.It, will according to the quantity of closure edge using edge detection algorithm
The specific method that above-mentioned face-image is divided into multiple regions see below in embodiment two, which is not described herein again.The present embodiment
In, face identification method in the prior art can be used by carrying out texture recognition to each region, identify corresponding face device
Official, then in the characteristic value for extracting corresponding human face.Such as glabella away from, eye spacing, eye is wide, eye is long, nose length, nasal base breadth,
The characteristic value for the numeralizations such as lip is wide, lip is thick, and it is special according to the texture that the characteristic value of these human faces generates current face image
Levy the texture feature vector of vector sum registration face-image.In the present embodiment, the textural characteristics of the current face image to
Amount can be denoted as (R1, R2, R3 ... Rn), and the texture feature vector for registering face-image can be denoted as (D1, D2, D3 ...
Dn)。
S103：It is studied just using texture feature vector of the preset correction model to the registration face-image, it is raw
At amendment feature vector.
In the present embodiment, since registration face-image is collected target person when target body handles identity card
The face-image of body, and the face-image does not update, and with growth and development of human body, and makeup etc. other because
Element, leads to register face-image that there are biggish differences with current face image, is unfavorable for challenge, therefore, can be to stepping on
The texture feature vector of note face-image is modified, and generates amendment feature vector, in order to challenge.Specifically, described
Preset correction model may include the parameter group (α being modified to the texture feature vector of the registration face-image
1, α 2, α 3 ... α n), the amendment feature vector of generation is (D1* (1+ α 1), D2* (1+ α 2), D3* (1+ α 3) ... Dn* (1+ α
n)).It will be described below, the corrected parameter group used here is on the basis of an initial assignment, every time in challenge
In the case where assert that the testimony of a witness is consistent, correction model is updated using the real face image currently acquired it is obtained, because
This is modified registration face-image by all previous updated correction model, can make up because of registration face-image itself not
Update and caused by deviation.
S104：The texture feature vector of the current face image and the diversity factor of the amendment feature vector are calculated, is sentenced
Whether the value for the diversity factor of breaking is less than preset threshold.
In the present embodiment, can be calculated by the following formula current face image texture feature vector and the amendment
The diversity factor of feature vector：
Wherein, I is the texture feature vector of current face image and the diversity factor of the amendment feature vector.Pass through judgement
Whether the value of I is less than preset threshold to judge whether revised registration face-image matches with current face image.
S105：When the value of the diversity factor is less than preset threshold, the current face image and the registration face are determined
Portion's image matches.
In the present embodiment, as the diversity factor I of the texture feature vector of current face image and the amendment feature vector
Value when being less than preset threshold, determine that current face image and the registration face-image match, and then described in can determining
The challenge of target body passes through, i.e., the current face image and registration face-image of the described target body coincide.
It further include using following formula to the correction model also, when the value of the diversity factor is less than preset threshold
It is updated：
Wherein α itIndicate the updated corrected parameter of epicycle, α it-1Indicate that epicycle updates pervious corrected parameter, i value
Range is 1 to n.
Real face image and the identity card registration of the embodiment of the present application are dual to compare authentication method, by challenge
Registration face-image is modified in the process, then carries out matching verifying with current face image, reduces registration face-image
With the diversity factor of current face image, thus improve automatic face identification speed.
If Fig. 2 is that the real face image of the embodiment of the present application two and identity card register the dual process for comparing authentication method
Figure.It in the present embodiment, can be using canny edge detection operator to the current face image and the registration face-image
Edge detection is carried out respectively, is extracted the image-region surrounded by closed edge, is specifically included：
S201：Convolution is made with Gauss mask respectively to the current face image and the registration face-image, to described
Current face image and the registration face-image are smoothed.
S202：The current face image and the registration face-image after calculating smoothing processing using Sobel operator
Each pixel gradient.
S203：Retain the pole of gradient intensity on each pixel of the current face image and the registration face-image
Big value, deletes other values；
S204：Set the pole of gradient intensity on each pixel of the current face image and the registration face-image
The pixel that the maximum of gradient intensity is greater than the threshold value upper bound is confirmed as side by the threshold value upper bound being worth greatly and threshold value lower bound
The maximum of gradient intensity is greater than the threshold value lower bound and is confirmed as weak boundary less than the pixel in the threshold value upper bound by boundary, will
The pixel that the maximum of gradient intensity is less than the threshold value lower bound is confirmed as non-boundary；
S205：The weak boundary being connected with the boundary is confirmed into boundary, other weak boundaries are confirmed as non-boundary.
After the current face image and the registration face-image are divided into multiple regions in the manner described above, with
One embodiment is identical to be, carries out texture recognition to each region, extracts human face characteristic value, described in generation
The texture feature vector of the texture feature vector of current face image and the registration face-image.
In turn, the texture feature vector of registration face-image is modified during challenge, it is special generates amendment
Levy vector；Matching verifying is carried out with current face image again.When the value of diversity factor is less than preset threshold, also to the amendment mould
Type is updated, and details are not described herein.
In the present embodiment, when using the present embodiment real face image and identity card registration dual compare authentication method
After carrying out multiple challenge, due to all special to the texture of the registration face-image of target body during each challenge
Sign vector is corrected, the texture feature vector of the registration face-image repeatedly obtained after amendment and former registration face-image
Texture feature vector accumulation generates biggish deviation, therefore, be to the texture to the registration face-image in correction model
The value range for the corrected parameter that feature vector is modified carries out necessary limitation.So correcting ginseng described in the present embodiment
Number is the corrected parameter for meeting preset threshold range, and the preset threshold range is (0.5,1.5).If repaired after certain update
Positive parameter exceeds the preset threshold range, then its value is set as lowest threshold 0.5 when the corrected parameter is less than lowest threshold,
Its value is set as highest threshold value 1.5 when the corrected parameter is greater than highest threshold value.
Equally, in the present embodiment, the parameter group in the correction model has number weight parameter group (δ 1, δ 2, δ
3 ... δ n), when the texture feature vector (D1, D2, D3 ... Dn) to the registration face-image is modified, generation
Correcting feature vector is (D1* (1+ δ 1* α 1), D2* (1+ δ 2* α 2), D3* (1+ δ 3* α 3) ... Dn* (1+ δ n* α n)), wherein
The value range of δ n in number weight parameter group (δ 1, δ 2, δ 3 ... δ n) is 0.5-1.5, the value of δ n with times of revision by
It is cumulative big.That is, the amplitude being modified to the texture feature vector of the registration face-image of target body is with amendment
The increase of number and be gradually increased.This is because the deviation of registration face-image and real face image is the day with the time
The product moon is tired and gradually widens, and therefore, increasing and (meaning that distance last enrollment time is more long) with certification number is then repaired
It is necessary to incrementally increase for positive amplitude.On the contrary, amendment amplitude should then be reduced for certification initial after registration, to avoid
By initial target body revised several times registration face-image texture feature vector just with original registration face
The texture feature vector of portion's image deviates excessive.
As shown in figure 3, being that the real face image of the embodiment of the present application three and identity card registration dual compare Verification System
Structure chart.It can be seen from the figure that the real face image of the present embodiment and identity card registration is dual compares Verification System, packet
It includes：
Image collection module 301, for obtaining the current face image and registration face-image of target body.Specifically,
Described image obtains module 301 can acquire the current face image of target body, such as camera by image capture device,
Registration face-image can also be extracted from identity card or remotely transfers target person from public security data base manipulation ID card No.
The registration face-image of body.
Texture feature vector generation module 302, for distinguishing the current face image and the registration face-image
Feature extraction is carried out, corresponding texture feature vector is generated.
Texture feature vector correction module 303, for utilizing preset correction model to the registration face-image
Texture feature vector education just, generate amendment feature vector.
Diversity factor determining module 304, the texture feature vector and the amendment for calculating the current face image are special
The diversity factor for levying vector, judges whether the value of the diversity factor is less than preset threshold.
As a result output module 305, for determining the current face figure when the value of the diversity factor is less than preset threshold
Picture and the registration face-image match.
This implementation can obtain the technical effect similar with above method embodiment, and which is not described herein again.
As shown in figure 4, being to register dual to compare certification using the real face image and identity card of the embodiment of the present application four
The schematic diagram of system progress challenge.It can be seen from the figure that when real face image and body using the embodiment of the present application
When part card registers dual comparison Verification System progress challenge, the registration face of target body is obtained by image collection module first
Portion's image and current face image, the registration face-image are that the face-image on the identity card with the target body is corresponding
Face-image, described image obtains module can directly extract the registration face-image from the identity card of the target body,
Registration face-image can also be remotely transferred from public security data Cooley according to the ID card No. of the target body.Described image obtains
Modulus block can use the current face image that image capture device (such as video camera) acquires the target body, obtain mesh
After the registration face-image and current face image of marking human body, by the registration face-image and current face figure of the target body
As being sent to texture feature vector generation module, the texture feature vector generation module is according to the registration face figure of target body
Picture and current face image generate the texture feature vector of registration face-image and the texture feature vector of current face image, will
The texture feature vector of the current face image is sent directly to diversity factor determining module, by the line of the registration face-image
Reason feature vector is sent to texture feature vector correction module, by texture feature vector correction module to the registration face-image
Texture feature vector be modified generation amendment feature vector after be sent to diversity factor determining module, specifically, generation is worked as
The texture feature vector of front face image can be (R1, R2, R3 ... Rn), generation registration face-image textural characteristics to
Amount is (D1, D2, D3 ... Dn), and the amendment feature vector of generation is (D1* (1+ α 1), D2* (1+ α 2), D3* (1+ α 3) ...
Dn* (1+ α n)), then by diversity factor determining module according to the following formula determine current face image texture feature vector with repair
The diversity factor of positive feature vector：
Wherein, I is the texture feature vector of current face image and the diversity factor of the amendment feature vector.Work as diversity factor
After the texture feature vector of current face image and the diversity factor I of the amendment feature vector is calculated in determining module, I is sent out
It send to result output module, can store the threshold range of variant degree I in the result output module, the result exports mould
Block matches diversity factor I with threshold range, judges whether the value of the diversity factor is less than preset threshold, and works as the difference
When the value of degree is less than preset threshold, determine that the current face image and the registration face-image match, i.e., the described target
The challenge of human body passes through.
In conclusion real face image provided by the embodiments of the present application and identity card registration it is dual compare authentication method and
System, verifying is first modified registration face-image with correction model every time, then carries out matching verifying with current face image,
And by during challenge assert the testimony of a witness be consistent in the case where, using the real face image currently acquired to amendment
Model is updated, so that the diversity factor of registration face-image and current face image is reduced, to improve automatic face
The speed of identification.
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.Those skilled in the art
Member is it should be appreciated that invention scope involved in the application, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from foregoing invention design, it is carried out by above-mentioned technical characteristic or its equivalent feature
Any combination and the other technical solutions formed.Such as features described above has similar function with (but being not limited to) disclosed herein
Can technical characteristic replaced mutually and the technical solution that is formed.
Claims (10)
1. a kind of real face image and identity card registration dual compare authentication method, which is characterized in that including：
Obtain the current face image and registration face-image of target body；
Feature extraction is carried out respectively to the current face image and the registration face-image, generate corresponding textural characteristics to
Amount；
It is studied just using texture feature vector of the preset correction model to the registration face-image, generates amendment feature
Vector；
The texture feature vector of the current face image and the diversity factor of the amendment feature vector are calculated, judges the difference
Whether the value of degree is less than preset threshold；
When the value of the diversity factor is less than preset threshold, the current face image and the registration face-image phase are determined
Match.
2. the method according to claim 1, wherein described to the current face image and registration face
Image carries out feature extraction respectively, generates corresponding texture feature vector, including：
Edge detection is carried out respectively to the current face image and the registration face-image, it will according to the quantity of closure edge
The current face image and the registration face-image are divided into multiple regions, carry out texture knowledge to each region
Not, human face characteristic value is extracted, the texture feature vector and the registration face-image of the current face image are generated
Texture feature vector.
3. according to the method described in claim 2, it is characterized in that, the preset correction model includes：
The parameter group (α 1, α 2, α 3 ... α n) that the texture feature vector of the registration face-image is modified.
4. according to the method described in claim 3, it is characterized in that, described to the current face image and registration face
Image carries out feature extraction respectively, including：Using canny edge detection operator to the current face image and the registration face
Portion's image carries out edge detection respectively, extracts the image-region surrounded by closed edge, specifically includes：
Convolution is made with Gauss mask respectively to the current face image and the registration face-image, to the current face figure
Picture and the registration face-image are smoothed；
Each pixel of the current face image and the registration face-image after calculating smoothing processing using Sobel operator
The gradient of point；
The maximum for retaining gradient intensity on each pixel of the current face image and the registration face-image, is deleted
Other values；
Set the threshold of the maximum of gradient intensity on each pixel of the current face image and the registration face-image
It is worth the upper bound and threshold value lower bound, the pixel that the maximum of gradient intensity is greater than the threshold value upper bound is confirmed as boundary, by gradient
The maximum of intensity is greater than the threshold value lower bound and is confirmed as weak boundary less than the pixel in the threshold value upper bound, by gradient intensity
The pixel that maximum is less than the threshold value lower bound is confirmed as non-boundary；
The weak boundary being connected with the boundary is confirmed into boundary, other weak boundaries are confirmed as non-boundary.
5. according to the method described in claim 3, it is characterized in that, described utilize preset correction model to the registration
The texture feature vector education of face-image just, generates amendment feature vector, including：
Using preset correction model (α 1, α 2, α 3 ... α n) to the texture feature vector of the registration face-image
(D1, D2, D3 ... Dn) is modified, and generates amendment feature vector (D1* (1+ α 1), D2* (1+ α 2), D3* (1+ α 3) ...
Dn*(1+αn))。
6. according to the method described in claim 5, it is characterized in that, calculating the current face image using following formula
The diversity factor of texture feature vector and the amendment feature vector, judges whether the value of the diversity factor is less than preset threshold：
Wherein, the texture feature vector of current face image is (R1, R2, R3 ... Rn), and I is diversity factor.
7. according to the method described in claim 6, it is characterized in that, also being wrapped when the value of the diversity factor is less than preset threshold
It includes and the correction model is updated using following formula：
Wherein α itIndicate the updated corrected parameter of epicycle, α it-1Indicate that epicycle updates pervious corrected parameter, i value range is
1 to n.
8. the method according to the description of claim 7 is characterized in that the corrected parameter is the amendment for meeting preset threshold range
Parameter, the preset threshold range are (0.5,1.5).
9. according to the method described in claim 8, it is characterized in that, the parameter group in the correction model is joined with number weight
Array (δ 1, δ 2, δ 3 ... δ n) is repaired in the texture feature vector (D1, D2, D3 ... Dn) to the registration face-image
Timing, the amendment feature vector of generation are (D1* (1+ δ 1* α 1), D2* (1+ δ 2* α 2), D3* (1+ δ 3* α 3) ... Dn* (1+ δ n*
α n)), wherein the value range of the δ n in number weight parameter group (δ 1, δ 2, δ 3 ... δ n) is 0.5-1.5, the value of δ n with
Times of revision is gradually increased.
10. a kind of real face image and identity card registration dual compare Verification System, which is characterized in that including：
Image collection module, for obtaining the current face image and registration face-image of target body；
Texture feature vector generation module, for carrying out feature respectively to the current face image and the registration face-image
It extracts, generates corresponding texture feature vector；
Texture feature vector correction module, for special using texture of the preset correction model to the registration face-image
It levies vector education just, generates amendment feature vector；
Diversity factor determining module, for calculate the current face image texture feature vector and the amendment feature vector
Diversity factor, judges whether the value of the diversity factor is less than preset threshold；
As a result output module, for determining the current face image and institute when the value of the diversity factor is less than preset threshold
Registration face-image is stated to match.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810649673.8A CN108875646B (en) | 2018-06-22 | 2018-06-22 | Method and system for double comparison and authentication of real face image and identity card registration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810649673.8A CN108875646B (en) | 2018-06-22 | 2018-06-22 | Method and system for double comparison and authentication of real face image and identity card registration |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108875646A true CN108875646A (en) | 2018-11-23 |
| CN108875646B CN108875646B (en) | 2022-09-27 |
Family
ID=64340367
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810649673.8A Active CN108875646B (en) | 2018-06-22 | 2018-06-22 | Method and system for double comparison and authentication of real face image and identity card registration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108875646B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110084098A (en) * | 2019-03-14 | 2019-08-02 | 杭州笔声智能科技有限公司 | A kind of paper corrects method and system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040062423A1 (en) * | 2002-09-27 | 2004-04-01 | Miwako Doi | Personal authentication apparatus and personal authentication method |
| CN104021397A (en) * | 2014-06-13 | 2014-09-03 | 中国民航信息网络股份有限公司 | Face identifying and comparing method and device |
| CN105139007A (en) * | 2015-09-30 | 2015-12-09 | 小米科技有限责任公司 | Positioning method and apparatus of face feature point |
| CN105956578A (en) * | 2016-05-23 | 2016-09-21 | 深圳华中科技大学研究院 | Face verification method based on identity document information |
| CN106203294A (en) * | 2016-06-30 | 2016-12-07 | 广东微模式软件股份有限公司 | The testimony of a witness unification auth method analyzed based on face character |
| CN106973192A (en) * | 2015-10-13 | 2017-07-21 | 柯尼卡美能达株式会社 | Image processing apparatus and image processing method |
| CN107358174A (en) * | 2017-06-23 | 2017-11-17 | 浙江大学 | A kind of hand-held authentication idses system based on image procossing |
| CN107944395A (en) * | 2017-11-27 | 2018-04-20 | 浙江大学 | A kind of method and system based on neutral net verification testimony of a witness unification |
-
2018
- 2018-06-22 CN CN201810649673.8A patent/CN108875646B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040062423A1 (en) * | 2002-09-27 | 2004-04-01 | Miwako Doi | Personal authentication apparatus and personal authentication method |
| CN104021397A (en) * | 2014-06-13 | 2014-09-03 | 中国民航信息网络股份有限公司 | Face identifying and comparing method and device |
| CN105139007A (en) * | 2015-09-30 | 2015-12-09 | 小米科技有限责任公司 | Positioning method and apparatus of face feature point |
| CN106973192A (en) * | 2015-10-13 | 2017-07-21 | 柯尼卡美能达株式会社 | Image processing apparatus and image processing method |
| CN105956578A (en) * | 2016-05-23 | 2016-09-21 | 深圳华中科技大学研究院 | Face verification method based on identity document information |
| CN106203294A (en) * | 2016-06-30 | 2016-12-07 | 广东微模式软件股份有限公司 | The testimony of a witness unification auth method analyzed based on face character |
| CN107358174A (en) * | 2017-06-23 | 2017-11-17 | 浙江大学 | A kind of hand-held authentication idses system based on image procossing |
| CN107944395A (en) * | 2017-11-27 | 2018-04-20 | 浙江大学 | A kind of method and system based on neutral net verification testimony of a witness unification |
Non-Patent Citations (1)
| Title |
|---|
| 尚政国 等: "基于Contourlet 特征修正的纹理图像识别算法", 《计算机工程》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110084098A (en) * | 2019-03-14 | 2019-08-02 | 杭州笔声智能科技有限公司 | A kind of paper corrects method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108875646B (en) | 2022-09-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108416336B (en) | A kind of method and system of intelligence community recognition of face | |
| CN105956578B (en) | A kind of face verification method of identity-based certificate information | |
| CN104751108B (en) | Facial image identification device and facial image recognition method | |
| WO2021036436A1 (en) | Facial recognition method and apparatus | |
| CN105022999B (en) | A kind of adjoint real-time acquisition system of people's code | |
| CN102542275B (en) | Automatic identification method for identification photo background and system thereof | |
| CN106815914A (en) | A kind of gate control system and unlocking method based on face recognition technology | |
| CN106529414A (en) | Method for realizing result authentication through image comparison | |
| CN105354558B (en) | Humanface image matching method | |
| CN103473564B (en) | A kind of obverse face detection method based on sensitizing range | |
| WO2016084072A1 (en) | Anti-spoofing system and methods useful in conjunction therewith | |
| JP2017502366A (en) | Iris biometric recognition module and access control assembly | |
| CN112487921B (en) | Face image preprocessing method and system for living body detection | |
| US20210089763A1 (en) | Animal identification based on unique nose patterns | |
| CN108335099A (en) | Method, apparatus, mobile terminal and the storage medium of mobile payment | |
| CN109145745A (en) | A kind of face identification method under circumstance of occlusion | |
| CN108629262A (en) | Iris identification method and related device | |
| CN108765014A (en) | A kind of intelligent advertisement put-on method based on access control system | |
| CN110503760A (en) | Access control method and access control system | |
| JP2004318248A (en) | Iris authentication system, iris authentication method, and iris authentication program | |
| CN106845449A (en) | A kind of image processing apparatus, method and face identification system | |
| CN109492555A (en) | Newborn identity identifying method, electronic device and computer readable storage medium | |
| CN107944395A (en) | A kind of method and system based on neutral net verification testimony of a witness unification | |
| CN107256388A (en) | A kind of method and device for obtaining front face image | |
| CN206849035U (en) | A kind of image processing apparatus and face identification system |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20220907 Address after: 266000 room 230, science and Technology Museum, EMU Town, west end of Chunyang Road, Jihongtan street, Chengyang District, Qingdao, Shandong Province Applicant after: QINGDAO CIVIL AVIATION CARES Co.,Ltd. Address before: Room b8406-03, 4th floor, 818 Huayuan Road, Xiangcheng District, Suzhou City, Jiangsu Province Applicant before: SUZHOU QIXIAN INTELLIGENT TECHNOLOGY Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |