CN108492422A - Face characteristic online recognition equipment - Google Patents
Face characteristic online recognition equipment Download PDFInfo
- Publication number
- CN108492422A CN108492422A CN201810272808.3A CN201810272808A CN108492422A CN 108492422 A CN108492422 A CN 108492422A CN 201810272808 A CN201810272808 A CN 201810272808A CN 108492422 A CN108492422 A CN 108492422A
- Authority
- CN
- China
- Prior art keywords
- unit
- matrix
- data
- image
- obtains
- 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
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/00174—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys
- G07C9/00563—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys using personal physical data of the operator, e.g. finger prints, retinal images, voicepatterns
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Collating Specific Patterns (AREA)
Abstract
In order to improve the recognition accuracy to real human body characteristic parameter, improve the identification safety of gate inhibition's unit, the present invention provides a kind of face characteristic online recognition equipment, recognition of face for gate inhibition's unit with identification sensor, including local, long-range two parts are identified respectively and comprehensive descision, the prompt of None- identified is provided for the false medium being identified, the prompt of identification success or failure is provided for the true medium that can be identified, the influence of the drift generated in temperature-rise period to gate inhibition's unit identification sensor of itself by way of heating in identification process, which is not given, to be eliminated, but it energetically pays attention to and has carried out the comparison that square root calculates this experience calculation in the calculating of testing result twice, to be omitted in analysis and calculating, square root when calculating v is eliminated in calculating.Through experiment, 40% or so is improved for the recognition accuracy of false body to be identified.
Description
Technical field
The invention belongs to security protections and technical field of image processing, and in particular to a kind of face characteristic online recognition equipment.
Background technology
Living things feature recognition is a kind of new identity identifying technology.In actual life, everyone has with other people not
Same unique biological characteristic.With the development of computer technology, people can extract the biological information of itself, than
Such as face, face refer to vein, iris, vocal print.The technology that this physical trait by people carries out identification is referred to as giving birth to
Object feature identification technique.
Face recognition technology is a kind of biometrics identification technology, with the advantages such as convenient, fast, accurate, in recent years
Obtain the development advanced by leaps and bounds.The input terminal input of face identification system is usually a face for containing identity to be detected
The facial image of several known identities in image and face database, and its output is then that a series of human face similarity degrees obtain
Point, with the identity of this face for showing identification.At present, face recognition technology is widely used to criminal investigation and case detection, department of banking
The fields such as system, customs inspection, the civil affairs department, work and rest attendance.However, with the continuous extension of face recognition technology application range,
Some safety problems also occur therewith, and criminal cheats face identification system using the human face photo forged, to legal
User causes heavy economic losses.Therefore, the judgement of the source authenticity of facial image is particularly important, here it is live bodies
Detection.
Invention content
In view of the above analysis knowledge of gate inhibition's unit is improved in order to improve the recognition accuracy to real human body characteristic parameter
Other safety, the present invention provides a kind of face characteristic online recognition equipment, for gate inhibition's unit with identification sensor
Recognition of face, including:
First facial image obtains and storage unit, for by identification sensor the first facial image of acquisition, described the
One facial image includes first part and second part, the data volume of wherein first part be the data volume of second part at most
One third;
First local analytic unit carries out the first local analytics for the first part to the first facial image, obtains
To the first local analysis result;
First remote analysis unit carries out the first remote analysis for the second part to the first facial image, obtains
To the first remote analysis result;
First judging unit, for when the first local analysis result meets and is less than predetermined the First Eigenvalue, leading to
Face the second facial image of acquisition is detected again after crossing gate inhibition's unit heating, while being detected finger to be identified and being sensed with identification
The polar coordinates angle theta of the identification plane of devicemn∈ [0,1], second facial image include Part III and Part IV, wherein
The data volume of Part III is at most the 30% of the data volume of Part IV;Otherwise prompt None- identified is without obtaining the second face
Image;
Second local analytics unit carries out the second local analytics for the Part III to the second facial image, obtains
To the second local analytics result;
Second remote analysis unit carries out the second remote analysis for the Part IV to the second facial image, obtains
To the second remote analysis result;According to the first local analysis result, the second local analytics result, the first remote analysis result and the
Two remote analysis are as a result, determine the face recognition result of gate inhibition's unit.
Further, the described first local analytic unit includes:
First neighborhood determination unit is used for centered on the image geometry center of first part, preset length is radius
Data are as pending first part's image data in neighborhood;
First pretreatment unit is obtained for carrying out binaryzation and noise reduction process to pending first part's image data
To data set O;
Encryption unit obtains data set O ' for the data set O to be carried out symmetry encryption;
The First Eigenvalue acquiring unit, for data set O ' and preset reference human face data M to be handled as follows:To pre-
If carrying out binary conversion treatment with reference to human face data M obtains M ';The diagonal matrix of data set O ' is calculated;According to the diagonal matrix
Exponent number intercepts the intermediary matrix P of same exponent number since the value of first, the upper left corners the data set M ';Calculate intermediary matrix P with
The characteristic value K1 for the matrix that data set O ' multiplication crosses obtain.
Further, first remote analysis unit includes:
First transmission unit is transferred to remote server for the second part to first facial image;
First hash value determination unit, for carrying out Hash fortune to the data of the second part received in remote server
Calculation obtains the first hash value corresponding with the second part;
Second pretreatment unit, for M pairs of matrix data K corresponding to the second part and preset reference human face data
The matrix S answered is multiplied, and one that wherein exponent number is smaller in the two matrixes obtains matrix K with diagonal matrix polishing after multiplication ';
Third pretreatment unit, for centered on the gray scale barycenter of the image of the second part, the preset length
For radius, the image data matrix L of second part is obtained, matrix L is projected to obtain matrix L ', by matrix L ' and matrix K '
Multiplication cross is carried out, wherein terminates smaller one in the two matrixes with diagonal matrix polishing, matrix Q is obtained after multiplication cross;
Second Eigenvalue acquiring unit is used for the characteristic value F of calculating matrix Q.
Further, the second local analytics unit includes:
Second neighborhood determination unit is used for centered on the image geometry center of Part III, preset length is radius
Data are as pending Part III image data in neighborhood;
Third pretreatment unit is obtained for carrying out binaryzation and noise reduction process to pending Part III image data
To data set R;
Second encryption unit, for the data set R to be carried out symmetry encryption by key of the average value of data set O,
Obtain data set R ';
Third feature value acquiring unit, for data set R ' and preset reference human face data M to be handled as follows:To pre-
If carrying out binary conversion treatment with reference to human face data M obtains M ';The diagonal matrix of data set R ' is calculated;According to the diagonal matrix
Exponent number intercepts the intermediary matrix T of same exponent number since the value of first, the upper left corners the data set M ';Calculate intermediary matrix T with
The characteristic value K2 for the matrix that data set R ' multiplication crosses obtain.
Further, second remote analysis unit includes:
Second transmission unit is transferred to remote server for the Part IV to second facial image;
Fourth feature value acquiring unit, for setting
Wherein, kmnIndicate the gray value of the image pixel (m, n) of Part IV;
Greyscale transformation Tr () is carried out to the image of Part IV and obtains θ 'mn:
R=2 ..., N, N are the natural number more than 2;
Wherein
Wherein θcFor Boundary Recognition threshold value, is determined by face Boundary Recognition empirical value, then calculated as follows again:
Transformation coefficient k 'mn=(K-1) θmn
Image boundary is extracted, the image boundary matrix extracted is
Edges=[k 'mn]
Calculate the characteristic value E of the image boundary matrix;
Second hash value determination unit, for carrying out Hash fortune to the data of the Part IV received in remote server
Calculation obtains the second hash value corresponding with the Part IV;
Third hash value determination unit obtains third hash value for carrying out Hash operations to preset reference human face data.
Further, second remote analysis unit further includes recognition result determination unit, for determining the gate inhibition
The face recognition result of unit, including:
Similarity calculated, for the first hash value to be added to the second hash value and carried out consistency Hash operations,
Operation result and third hash value are calculated into degree a similar to each other using MinHash algorithms;
Recognition result determination unit, for determiningWhether preset knowledge is less than
Other threshold value, when less than when prompt identify successfully, otherwise prompt recognition failures.
Technical scheme of the present invention has the following advantages:
The face characteristic online recognition equipment of the present invention can be based on long-range and factor calculating is locally identified respectively, know
The influence for the drift that gate inhibition's unit identification sensor of itself is generated in temperature-rise period by way of heating during not
Do not give and eliminate, but energetically pay attention to and carried out in the calculating of testing result twice square root calculate this once
The comparison of calculation is tested, to be omitted in analysis and calculating (square root when calculating v is eliminated in calculating)
Through experiment, 40% or so is improved for the recognition accuracy of false body to be identified.
Description of the drawings
Fig. 1 shows the composition frame chart of the apparatus according to the invention.
Specific implementation mode
As shown in Figure 1, face characteristic online recognition equipment according to a preferred embodiment of the invention, for having identification to pass
The recognition of face of gate inhibition's unit of sensor, including:
First facial image obtains and storage unit, for by identification sensor the first facial image of acquisition, described the
One facial image includes first part and second part, the data volume of wherein first part be the data volume of second part at most
One third;
First local analytic unit carries out the first local analytics for the first part to the first facial image, obtains
To the first local analysis result;
First remote analysis unit carries out the first remote analysis for the second part to the first facial image, obtains
To the first remote analysis result;
First judging unit, for when the first local analysis result meets and is less than predetermined the First Eigenvalue, leading to
Face the second facial image of acquisition is detected again after crossing gate inhibition's unit heating, while being detected finger to be identified and being sensed with identification
The polar coordinates angle theta of the identification plane of devicemn∈ [0,1], second facial image include Part III and Part IV, wherein
The data volume of Part III is at most the 30% of the data volume of Part IV;Otherwise prompt None- identified is without obtaining the second face
Image;
Second local analytics unit carries out the second local analytics for the Part III to the second facial image, obtains
To the second local analytics result;
Second remote analysis unit carries out the second remote analysis for the Part IV to the second facial image, obtains
To the second remote analysis result;According to the first local analysis result, the second local analytics result, the first remote analysis result and the
Two remote analysis are as a result, determine the face recognition result of gate inhibition's unit.
Further, the described first local analytic unit includes:
First neighborhood determination unit is used for centered on the image geometry center of first part, preset length is radius
Data are as pending first part's image data in neighborhood;
First pretreatment unit is obtained for carrying out binaryzation and noise reduction process to pending first part's image data
To data set O;
Encryption unit obtains data set O ' for the data set O to be carried out symmetry encryption;
The First Eigenvalue acquiring unit, for data set O ' and preset reference human face data M to be handled as follows:To pre-
If carrying out binary conversion treatment with reference to human face data M obtains M ';The diagonal matrix of data set O ' is calculated;According to the diagonal matrix
Exponent number intercepts the intermediary matrix P of same exponent number since the value of first, the upper left corners the data set M ';Calculate intermediary matrix P with
The characteristic value K1 for the matrix that data set O ' multiplication crosses obtain.
Further, first remote analysis unit includes:
First transmission unit is transferred to remote server for the second part to first facial image;
First hash value determination unit, for carrying out Hash fortune to the data of the second part received in remote server
Calculation obtains the first hash value corresponding with the second part;
Second pretreatment unit, for M pairs of matrix data K corresponding to the second part and preset reference human face data
The matrix S answered is multiplied, and one that wherein exponent number is smaller in the two matrixes obtains matrix K with diagonal matrix polishing after multiplication ';
Third pretreatment unit, for centered on the gray scale barycenter of the image of the second part, the preset length
For radius, the image data matrix L of second part is obtained, matrix L is projected to obtain matrix L ', by matrix L ' and matrix K '
Multiplication cross is carried out, wherein terminates smaller one in the two matrixes with diagonal matrix polishing, matrix Q is obtained after multiplication cross;
Second Eigenvalue acquiring unit is used for the characteristic value F of calculating matrix Q.
Further, the second local analytics unit includes:
Second neighborhood determination unit is used for centered on the image geometry center of Part III, preset length is radius
Data are as pending Part III image data in neighborhood;
Third pretreatment unit is obtained for carrying out binaryzation and noise reduction process to pending Part III image data
To data set R;
Second encryption unit, for the data set R to be carried out symmetry encryption by key of the average value of data set O,
Obtain data set R ';
Third feature value acquiring unit, for data set R ' and preset reference human face data M to be handled as follows:To pre-
If carrying out binary conversion treatment with reference to human face data M obtains M ';The diagonal matrix of data set R ' is calculated;According to the diagonal matrix
Exponent number intercepts the intermediary matrix T of same exponent number since the value of first, the upper left corners the data set M ';Calculate intermediary matrix T with
The characteristic value K2 for the matrix that data set R ' multiplication crosses obtain.
Further, second remote analysis unit includes:
Second transmission unit is transferred to remote server for the Part IV to second facial image;
Fourth feature value acquiring unit, for setting
Wherein, kmnIndicate the gray value of the image pixel (m, n) of Part IV;
Greyscale transformation Tr () is carried out to the image of Part IV and obtains θ 'mn:
R=2 ..., N, N are the natural number more than 2;
Wherein
Wherein θcFor Boundary Recognition threshold value, is determined by face Boundary Recognition empirical value, then calculated as follows again:
Transformation coefficient k 'mn=(K-1) θmn
Image boundary is extracted, the image boundary matrix extracted is
Edges=[k 'mn]
Calculate the characteristic value E of the image boundary matrix;
Second hash value determination unit, for carrying out Hash fortune to the data of the Part IV received in remote server
Calculation obtains the second hash value corresponding with the Part IV;
Third hash value determination unit obtains third hash value for carrying out Hash operations to preset reference human face data.
Further, second remote analysis unit further includes recognition result determination unit, for determining the gate inhibition
The face recognition result of unit, including:
Similarity calculated, for the first hash value to be added to the second hash value and carried out consistency Hash operations,
Operation result and third hash value are calculated into degree a similar to each other using MinHash algorithms;
Recognition result determination unit, for determiningWhether it is less than preset
Recognition threshold, when less than when prompt identify successfully, otherwise prompt recognition failures.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
All any modification, equivalent and improvement etc., should all be included in the protection scope of the present invention made by within refreshing and principle.
Claims (6)
1. a kind of face characteristic online recognition equipment, for the recognition of face of gate inhibition's unit with identification sensor, including:
First facial image obtains and storage unit, described the first for obtaining the first facial image by identification sensor
Face image includes first part and second part, and the data volume of wherein first part is at most three points of the data volume of second part
One of;
First local analytic unit carries out the first local analytics for the first part to the first facial image, obtains the
One local analytics result;
First remote analysis unit obtains for carrying out the first remote analysis to the second part of the first facial image
One remote analysis result;
First judging unit, for when the first local analysis result meets and is less than predetermined the First Eigenvalue, passing through institute
It states and detects face again after the heating of gate inhibition's unit and obtain the second facial image, while detecting finger to be identified and identification sensor
Identify the polar coordinates angle theta of planemn∈ [0,1], second facial image include Part III and Part IV, wherein third
Partial data volume is at most the 30% of the data volume of Part IV;Otherwise prompt None- identified is without obtaining the second face figure
Picture;
Second local analytics unit obtains for carrying out the second local analytics to the Part III of the second facial image
Two local analytics results;
Second remote analysis unit obtains for carrying out the second remote analysis to the Part IV of the second facial image
Two remote analysis results;It is remote according to the first local analysis result, the second local analytics result, the first remote analysis result and second
Journey analysis result determines the face recognition result of gate inhibition's unit.
2. the apparatus according to claim 1, which is characterized in that the first local analytic unit includes:
First neighborhood determination unit, for centered on the image geometry center of first part, preset length for radius neighborhood
Interior data are as pending first part's image data;
First pretreatment unit is counted for carrying out binaryzation and noise reduction process to pending first part's image data
According to collection O;
Encryption unit obtains data set O ' for the data set O to be carried out symmetry encryption;
The First Eigenvalue acquiring unit, for data set O ' and preset reference human face data M to be handled as follows:To default ginseng
It examines human face data M progress binary conversion treatment and obtains M ';The diagonal matrix of data set O ' is calculated;According to the rank of the diagonal matrix
Number, intercepts the intermediary matrix P of same exponent number since the value of first, the upper left corners the data set M ';Calculate intermediary matrix P and number
According to the characteristic value K1 for the matrix that collection O ' multiplication crosses obtain.
3. the apparatus of claim 2, which is characterized in that first remote analysis unit includes:
First transmission unit is transferred to remote server for the second part to first facial image;
First hash value determination unit is obtained for carrying out Hash operations to the data of the second part received in remote server
To the first hash value corresponding with the second part;
Second pretreatment unit, it is corresponding with preset reference human face data M for matrix data K corresponding to the second part
Matrix S is multiplied, and one that wherein exponent number is smaller in the two matrixes obtains matrix K with diagonal matrix polishing after multiplication ';
Third pretreatment unit is used for centered on the gray scale barycenter of the image of the second part, the preset length is half
Diameter obtains the image data matrix L of second part, is projected to obtain matrix L to matrix L ', by matrix L ' and matrix K ' carry out
Multiplication cross wherein terminates smaller one with diagonal matrix polishing, matrix Q is obtained after multiplication cross in the two matrixes;
Second Eigenvalue acquiring unit is used for the characteristic value F of calculating matrix Q.
4. device according to claim 3, which is characterized in that the second local analytics unit includes:
Second neighborhood determination unit, for centered on the image geometry center of Part III, preset length for radius neighborhood
Interior data are as pending Part III image data;
Third pretreatment unit is counted for carrying out binaryzation and noise reduction process to pending Part III image data
According to collection R;
Second encryption unit is obtained for the data set R to be carried out symmetry encryption by key of the average value of data set O
Data set R ';
Third feature value acquiring unit, for data set R ' and preset reference human face data M to be handled as follows:To default ginseng
It examines human face data M progress binary conversion treatment and obtains M ';The diagonal matrix of data set R ' is calculated;According to the rank of the diagonal matrix
Number, intercepts the intermediary matrix T of same exponent number since the value of first, the upper left corners the data set M ';Calculate intermediary matrix T and number
According to the characteristic value K2 for the matrix that collection R ' multiplication crosses obtain.
5. device according to claim 4, which is characterized in that second remote analysis unit includes:
Second transmission unit is transferred to remote server for the Part IV to second facial image;
Fourth feature value acquiring unit, for setting
Wherein, kmnIndicate the gray value of the image pixel (m, n) of Part IV;
Greyscale transformation Tr () is carried out to the image of Part IV and obtains θ 'mn:
R=2 ..., N, N are the natural number more than 2;
Wherein
Wherein θcFor Boundary Recognition threshold value, is determined by face Boundary Recognition empirical value, then calculated as follows again:
Transformation coefficient k 'mn=(K-1) θmn
Image boundary is extracted, the image boundary matrix extracted is
Edges=[k 'mn]
Calculate the characteristic value E of the image boundary matrix;
Second hash value determination unit is obtained for carrying out Hash operations to the data of the Part IV received in remote server
To the second hash value corresponding with the Part IV;
Third hash value determination unit obtains third hash value for carrying out Hash operations to preset reference human face data.
6. device according to claim 5, which is characterized in that second remote analysis unit further includes that recognition result is true
Order member, the face recognition result for determining gate inhibition's unit, including:
Similarity calculated will be transported for the first hash value to be added to the second hash value and carried out consistency Hash operations
It calculates result and calculates degree a similar to each other using MinHash algorithms with third hash value;
Recognition result determination unit, for determiningWhether preset identification threshold is less than
Value, when less than when prompt identify successfully, otherwise prompt recognition failures.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810272808.3A CN108492422A (en) | 2018-03-29 | 2018-03-29 | Face characteristic online recognition equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810272808.3A CN108492422A (en) | 2018-03-29 | 2018-03-29 | Face characteristic online recognition equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108492422A true CN108492422A (en) | 2018-09-04 |
Family
ID=63317044
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810272808.3A Pending CN108492422A (en) | 2018-03-29 | 2018-03-29 | Face characteristic online recognition equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108492422A (en) |
-
2018
- 2018-03-29 CN CN201810272808.3A patent/CN108492422A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
El-Abed et al. | Evaluation of biometric systems | |
Zheng et al. | A unified distance measure scheme for orientation coding in identification | |
CN102156887A (en) | Human face recognition method based on local feature learning | |
CN112052731B (en) | Intelligent portrait identification card punching attendance system and method | |
EP3057037A1 (en) | Biometric information registration apparatus and biometric information registration method | |
WO2017075913A1 (en) | Mouse behaviors based authentication method | |
US20150178544A1 (en) | System for estimating gender from fingerprints | |
Akbar et al. | Palm vein biometric identification system using local derivative pattern | |
Shawkat et al. | The new hand geometry system and automatic identification | |
CN108491814A (en) | Recognition of face monitoring device | |
CN113591921B (en) | Image recognition method and device, electronic equipment and storage medium | |
CN105184236A (en) | Robot-based face identification system | |
Su et al. | Evaluation of rarity of fingerprints in forensics | |
Harb et al. | Palm print recognition | |
CN108492421A (en) | Low-power consumption face identification method | |
CN104615985B (en) | A kind of recognition methods of human face similarity degree | |
CN108492422A (en) | Face characteristic online recognition equipment | |
CN108416882A (en) | Portable fingerprint identification device | |
CN108492420A (en) | Safety-protection system based on fingerprint recognition | |
CN111401348B (en) | Living body detection method and system for target object | |
CN111428670B (en) | Face detection method, face detection device, storage medium and equipment | |
Balazia et al. | How unique is a face: An investigative study | |
Hemanth et al. | Improving Accuracy of Face Detection in ID Proofs using CNN and Comparing with DLNN | |
Amin et al. | RELIABLE PERSON IDENTIFICATION USING A NOVEL MULTIBIOMETRIC IMAGE SENSOR FUSION ARCHITECTURE | |
CN108416329A (en) | Remote fingerprint recognition methods |
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: 20180904 |
|
RJ01 | Rejection of invention patent application after publication |