CN106340139A - VTM (Virtual Teller Machine) capable of automatically recognizing human face - Google Patents
VTM (Virtual Teller Machine) capable of automatically recognizing human face Download PDFInfo
- Publication number
- CN106340139A CN106340139A CN201610715546.4A CN201610715546A CN106340139A CN 106340139 A CN106340139 A CN 106340139A CN 201610715546 A CN201610715546 A CN 201610715546A CN 106340139 A CN106340139 A CN 106340139A
- Authority
- CN
- China
- Prior art keywords
- face
- dimensional
- image
- information
- unit
- 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
- G07F—COIN-FREED OR LIKE APPARATUS
- G07F19/00—Complete banking systems; Coded card-freed arrangements adapted for dispensing or receiving monies or the like and posting such transactions to existing accounts, e.g. automatic teller machines
- G07F19/20—Automatic teller machines [ATMs]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Finance (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Accounting & Taxation (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Collating Specific Patterns (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses a VTM (Virtual Teller Machine) capable of automatically recognizing a human face. The VTM comprises a multi-human face picking device, a human face model establishing device, a human face model reverse modeling device and a client face database, wherein the multi-human face picking device is used for picking multiple human faces in images one by one; the a human face model establishing device is used for establishing corresponding three-dimensional array models for the human faces in the images; the human face model reverse modeling device is used for establishing simulation three-dimensional models from the three-dimensional array models in a reverse molding manner; the client face database is used for storing client face data information, comparing the simulation three-dimensional models after reverse molding with stored client face data information, and feeding back comparison results. When any person enters a view range of the VTM, human face information of the person can be automatically acquired by the VTM immediately, and can be compared through the database automatically. If the human face information is user data stored in the database, corresponding teller operation can be implemented according to grade classification of the client immediately, the operation that identity verification is carried out for the client for multiple times can be avoided, or further artificial verification of a background can be avoided. On the premise that the security is ensured, the simplest and most convenient financial service can be achieved.
Description
Technical field
The present invention relates to a kind of financial terminal, more particularly, to a kind of vtm machine of automatic identification face.
Background technology
From the virtual automatic teller machine of vtm(virtual teller machine) and stm(smart teller machine wisdom
Automatic teller machine) etc. after intelligent finance terminal emerges, embody due to operating manpower and lifting client can be saved in a large number, increasingly tend to
Replace traditional atm.Intelligent terminal can provide open an account, cancellation, transfer accounts, to the irrealizable function of the tradition atm institute such as public affairs, but
It is that these functions all must do harsh identity verification to user, be not as atm and just can be considered as authority confirmation means using card.
Intelligent finance terminal-pair user identification confirmation and license confirmation use multiple bio-identification side such as fingerprint, iris, face, vocal print
Method.Wherein recognition of face is belonging to the mandate identification of high authority, therefore very high to the technical requirements of recognition of face.And existing skill
It is all that the image first obtaining user in checking is changed into stereomodel to recognition of face in art, then by the new axonometric chart obtaining
As wanting to compare with the stereomodel that user is initially recorded.Although only processing speed is contrasted comparatively fast by mathematical model,
Backstage manually can not effectively be checked, and increased checking leak risk, and such as lawless person can be out-tricked checking by solid mask
Step etc..Simultaneously when multiple face in picture, the identifying system of prior art arises that interim card it is impossible to recognize
Destination object.The picture being arranged on the financial terminal crawl in roadside often comprises multiple faces, and after leading to extract, impact identity is tested
The accuracy of card.Authentication just will start in plug-in card or after carrying out other instructions, and safety is not enough.
Content of the invention
In order to solve above-mentioned technical problem, present invention aim at providing a kind of vtm machine of automatic identification face.
A kind of vtm machine of automatic identification face of the present invention is it is characterised in that include:
Plurality of human faces pick device, sets up device for picking up and being sequentially inputted to faceform faces multiple in image one by one；
In human body after vtm machine automatic image capturing；
Faceform sets up device, for the face in image is set up corresponding three-dimensional Array Model, is then input to face mould
The anti-molding apparatus of type；
The anti-molding apparatus of faceform, for moulding counter for three-dimensional array model as artificial stereo model；
Client's face database, for storing client's face data message, and by the artificial stereo model after counter moulding with store
Client's face data message compare, feed back comparison result.
Described plurality of human faces pick device, comprising:
Image collection module, for catching one section of image that current time starts；
Motion capture module, meets the dynamic image block of human body for catching from image；
Face extraction module, for extracting corresponding face information from image block.
Described motion capture module, after extracting all of image block, the left screening to image block size, will accord with after screening
The image block closing dimensional requirement is input to face extraction module.
Described plurality of human faces pick device, also includes time control module, for according to background environment real-time regulation image
Catch parameter；The definition of background environment is lower, and pull-in time is longer, sample frequency is lower；The definition of background environment is higher,
Pull-in time is shorter, sample frequency is higher.
Described faceform sets up device, comprising:
Two dimensional image acquiring unit, obtains the two dimensional image including face, and two dimensional image is sent to Face detection unit；
Face detection unit, to the extract facial feature in two dimensional image, sketches the contours of X-Y scheme according to face characteristic normal arrangement
Face scope in picture, reads two-dimensional coordinate to the pixel included by face scope, and two-dimensional coordinate is input to two-dimensional vector collection
Unit；
Two-dimensional vector collecting unit, according to the two-dimensional coordinate of pixel in the range of face, generates the two-dimensional vector of pixel, by all two
N dimensional vector n forms vector matrix, and vector matrix is delivered to deformation unit；
Three-dimensional face database, prestored conventional face three-dimensional information and characteristic information；
Deformation unit, vector matrix is contrasted with the three-dimensional information of storage in three-dimensional face database and characteristic information, from three-dimensional
Filter out the three-dimensional information closest with vector matrix feature and characteristic information in the information of human face data library storage, will screen
The information going out, as deformation parameter, carries out three-dimensional deformation to vector matrix and obtains deformation matrix；
Deformation constraint element, the deformation process that monitoring deformation unit is carried out to vector matrix, the arbitrary unit in constrained strain matrix
Element is not in illegal value；
Trivector signal generating unit, deformation matrix is converted to trivector group, to each trivector in trivector group
One pixel of distribution.
Described three-dimensional face database has prestored the three-dimensional information of many ethnic groups and many colours of skin and characteristic information；Described shape
Become unit in filter information, colour code and the immediate three-dimensional information of colour code and spy in prestored information in preferential screening face pixel
Reference ceases；The illegal value occurring in deformation matrix is constrained in the maximum legal value that can deviate by described deformation constraint element.
The anti-molding apparatus of described faceform, comprising:
Obtaining three-dimensional model unit, obtains three-dimensional face model data, obtains the original image generating three-dimensional face model simultaneously；
Original image is obtained with the COLOR COMPOSITION THROUGH DISTRIBUTION of face pixel；
Attached color element, carries out color assignment according to original image and COLOR COMPOSITION THROUGH DISTRIBUTION to three-dimensional face model；
Attitude dummy unit, carries out axle rotation to three-dimensional face model, angle effect corresponding with illumination in virtual rotary course, raw
Become attitude photometric data；
Counter mould unit, color correction is carried out according to attitude photometric data to the three-dimensional face model completing color assignment, generate with
Different rotary angle emulates faceform correspondingly；
Classification memory element, described classification memory element stores to emulation faceform according to different lighting angles.
Described attitude dummy unit, lighting simulation adopts single light source to irradiate or respectively to uniform light intense irradiation；Three-dimensional face
Respectively with each orthogonal axis as rotary shaft, rotation is not more than positive and negative 90 ° of angle to model；The described anglec of rotation is at -30 °
To between 30 °.
Described client's face database, comparison result be recorded client's face number for the artificial stereo model of non-customer
According in storehouse.
A kind of vtm machine of automatic identification face of the present invention, has an advantage in that, when anyone enters regarding of vtm machine
As scope, just its face information is automatically obtained by vtm machine at once, and pass through data base's automatic comparison.If in data base
The user data of storage, then can carry out corresponding sales counter operation according to the grade separation of this client at once, remove user from multiple
Carry out authentication, or connection backstage carries out further manual verification.Achieve the simplest in the case of ensureing safety
The most efficiently financial service.As the client of the non-storage entering video scope, then its face information all be recorded data
Storehouse, provides evidence in the application of follow-up security.
Brief description
Fig. 1 is the structural representation of automatic identification face vtm machine of the present invention.
Fig. 2 is plurality of human faces pick device structural representation of the present invention.
Fig. 3 is that faceform of the present invention sets up apparatus structure schematic diagram.
Fig. 4 is faceform of the present invention anti-molding apparatus structural representation.
Specific embodiment
According to Fig. 1, a kind of vtm machine of automatic identification face of the present invention, include plurality of human faces pick device,
Set up device for picking up and being sequentially inputted to faceform faces multiple in image one by one；Human body after vtm machine from
Dynamic startup image capturing；Faceform sets up device, for the face in image is set up corresponding three-dimensional Array Model, then defeated
Enter molding apparatus anti-to faceform；The anti-molding apparatus of faceform, for moulding counter for three-dimensional array model as artificial stereo model；Visitor
Family face database, for storing client's face data message, and the client by the artificial stereo model after counter moulding and storage
Human face data information compares, and feeds back comparison result.
Result after comparison is divided into " being client " and " non-customer ", and when result is " being client ", vtm machine will carry out this use
The corresponding financial operation of family place grade, for example, open an account, deposit, cancellation etc..Enter vtm machine video scope in user just to obtain at once
To its effective authentication, wait without user, improve Consumer's Experience sense in the case of safe enough.When result is " non-
During client ", this face information is all stored in data base, effective evidence can be provided for follow-up security work.
Plurality of human faces pick device according to Fig. 2, includes image collection module, for catching what current time started
One section of image；Motion capture module, meets the dynamic image block of human body for catching from image；Face extraction module, is used for
Corresponding face information is extracted from image block.Measure human body entrance video scope in vtm machine examination and need to carry out authentication
When, start and catch photographic head, catch photographic head and adopt wide-angle lens, the image close to 180 ° can be caught.Then from image
Extract all of image block successively, this image block extract according to the human body multidate information in human body dynamic base as foundation.Keep away
Exempt from the image of the movements such as typing cat and dog fallen leaves rather than human body information, waste memory space.Finally comprising the image of human body information
Extract corresponding face information in fast, face information is fed back to host computer and is for further processing.
In order to save memory space and transmission bandwidth, the described image display time is less than 1 second, and sample frequency is not less than
60hz.Using 60hz high sampling rate it is therefore intended that the image procossing of high computing is arranged in the financial terminal of front end, and need not
A large amount of computings are required to feed back to the server of far-end, save the bandwidth requirement of transmission.Within the sampling time of 1 second, can
To realize sampled images more than 60, the rate of change of each pixel is sufficient for motion capture module and completes dynamic detection.
Described motion capture module, after extracting all of image block, the left screening to image block size, will accord with after screening
The image block closing dimensional requirement is input to face extraction module, removes the image block being smaller in size than meansigma methodss 1/2nd simultaneously.
Avoid because heavy remaining data influence memory space and transmission bandwidth further, because sometimes using the user side of vtm or stm
Other retinue occurs, if only extracting the image block of large-size, easy error extraction is to non-targeted user's
Face information and eliminate the real user's face needing and extracting, increase error rate.And save again after meansigma methodss are set and be less than
The image block of meansigma methodss 1/2nd, can guarantee that the face near financial terminal is all effectively recognized completely, and by background
Face effectively reject, reduce formality image operation pressure.
, with body contour line as border, the data outside border is all with transparent figure layer process for described image block.Transparent figure layer
Data processing be simplest processing mode, the pixel outside body contour line is not the data needed for certification, therefore may be used
To reduce heavy remainder as far as possible according to the impact to image procossing.
Described plurality of human faces pick device, also includes time controller, for being caught according to background environment real-time regulation image
Catch parameter；The definition of background environment is lower, and pull-in time is longer, sample frequency is lower；The definition of background environment is higher, catches
Time of catching is shorter, sample frequency is higher.Definition according to background environment arranges pull-in time and sample rate, Ke Yibao come inverse ratio
Only at daytime or night, the total number that can automatically adjust its sampled images is consistent card, the pressure of such image operation
Will not increase because of background environment difference, also will not waste computing power because background environment is good.Specifically, such as standard
Under background environment, the setting sampling time is 0.5 second, and sample frequency is 100hz, obtains 50 continuous images in once sampling
The image of composition.When environmental background definition is relatively low, motion capture module is not allowed disposable, because the change of pixel
Less, easily catch error.Therefore the sampling time can be adjusted to 0.8 second, sample frequency is 62.5hz, and this is equally once
The image of 50 continuous image compositions is obtained in sampling.As a same reason, when environmental background definition is preferable, dynamically catch
Catch module and can effectively catch out human body quickly dynamically, so there is no need to long pull-in time.Time control module is automatic
To be adjusted in the sampling time 0.2 second, sample frequency be 250hz it is achieved that according to environmental background real-time adjustment sampling parameter, and not shadow
Ring the accuracy rate of face extraction.
Faceform according to Fig. 3 sets up device, includes two dimensional image acquiring unit, Face detection unit, and two
N dimensional vector n collecting unit, three-dimensional face database, deformation unit, deformation constraint element, trivector signal generating unit.Two dimensional image
The face picture that acquiring unit will go in image pickup scope is stored with two-dimensional format, includes the color of each pixel in two dimensional image
Mark, in case deformation unit screens the colour of skin and ethnic group information according to colour code from three-dimensional face database.Two dimensional image acquiring unit
Two dimensional image is delivered to after Face detection unit, first two dimensional image is put into relief area, wait just clear after the completion of whole modeling process
Buffer empty is it is ensured that arbitrary process occurs can again transferring original image repair data during error in data, especially in three-dimensional shaped
During change and constrained strain, error in data the most easily occurs.
Face detection unit receives after two dimensional image to the extract facial feature in image, for example, retrieve one group of pixel
By olive shape continuous lines add central circular dark color colour code pixel form, and separated by a distance after equally retrieve other one
Organize akin pixel groups, then judge that this two groups of pixels form eye feature.Same principle, can examine according to the pixel form of the composition
Rope is to the face necessary feature such as nose, mouth, forehead and ear.According to the normal arrangement of face characteristic, for example pass through eye with
The normal arrangement of mouth, the scope that continuous lines that surround each face characteristic, to meet facial contour are covered is judged as face
Scope.After sketching the contours of face scope, two-dimensional coordinate is read to the pixel included by face scope, two-dimensional coordinate is input to two dimension
Vector collecting unit.When reading two-dimensional coordinate, can sample or all read two-dimensional coordinate according to system processing power.
Two-dimensional vector collecting unit, generates corresponding two-dimensional vector to each two-dimensional coordinate, forms two-dimensional vector simultaneously
Vector matrix.Matrix operationss, according to the soft or hard environment being suitable for General System, do pretreatment to three-dimensional operation.Deformation unit is by vector
Three-dimensional information in matrix and three-dimensional face database and characteristic information contrast, sieve from the information of three-dimensional face database storage
Select the three-dimensional information closest with vector matrix feature and characteristic information.Wherein three-dimensional face database prestored normal in a large number
Rule face three-dimensional information and characteristic information, especially includes three-dimensional information and the characteristic information of many ethnic groups and many colours of skin.Such as Europe
Descendants people is more prominent than the nose feature of Asian, the more low information classification of colour code.Facilitate deformation unit preferentially to screen, subtract
Few operation time, deformation unit can preferentially screen colour code and colour code in prestored information in face pixel and connect the most in filter information
Near three-dimensional information and characteristic information.Deformation unit after the information of having screened using this information as deformation parameter, deformation is joined
Several three-dimensional deformation is carried out to vector matrix, obtain deformation matrix.
Deformation matrix is converted to trivector group by last trivector signal generating unit, distributes one to each trivector
Pixel.Without colour code, the set of all new distribution pixels is the required faceform setting up to newly assigned pixel.?
To after faceform, vtm machine can carry out subsequent verification operations in this, as person identifier.
During deformation, in order to avoid illegal value in deformation matrix, deformation constraint element can monitor in real time shape
Each of become the operation result of element and corresponding parameter.After deformation constraint element is judged as legitimate value, deformation unit ability
Operation values are given new matrix element.There is various feature structure in view of different ethnic group all ages and classes and different sexes,
Illegal value is only constrained in the maximum legal value that can deviate by deformation constraint element, and such as illegal value judgment threshold can be set to swear
The each of amount deviates the 30% of deformation parameter respective value to component.
Deformation constraint both allow for each individuality and some positions unexpected deviation from the norm faceform occurred, also reasonable
In the range of limit its deviation situation.Make that faceform more conforms to modeling object is really three-dimensional situation, and the model obtaining is more
There is the uniqueness of identity, have significant auxiliaring effect to financial certification authority.
The anti-molding apparatus of faceform according to Fig. 4, include obtaining three-dimensional model unit, attached color element, attitude void
Quasi-simple unit and counter mould unit.The anti-molding apparatus of face first pass through obtaining three-dimensional model unit and set up acquisition device from faceform
Three-dimensional face model data, and obtain the original image generating this three-dimensional face model.Conventional according to face to original image
The regularity of distribution, scans face scope in this image.Then to the pixel collection colour code in the range of face, each colour code is assigned
Give coordinate figure, obtain the COLOR COMPOSITION THROUGH DISTRIBUTION data of face pixel.
By original image and COLOR COMPOSITION THROUGH DISTRIBUTION data input to attached color element, attached color element is when color assignment with three-dimensional
On the basis of faceform offsets relative to 3 orthogonal axis 0, according to the pixel two-dimensional coordinate of original image and COLOR COMPOSITION THROUGH DISTRIBUTION, will
The colour code of each pixel projects to the corresponding two-dimensional coordinate of three-dimensional face model and fastens.
Attitude dummy unit obtains the three-dimensional face model of not attached color, and three-dimensional face model is selected respectively, relatively
The rotation of each orthogonal axis can rotate a circle or control in positive and negative 90 °.Because under true environment, rotate over
90 ° can obtain back side image, do not meet authentication requesting.In order to reduce operand, the anglec of rotation of each orthogonal axis controls
It is optimal between positive and negative 30 °.Reach high emulation reduction, also reduce hardware computation resource to greatest extent.Attitude dummy unit is adopted
Irradiated with single light source or respectively to uniform light intense irradiation, the illumination projection that three-dimensional face model is simulated, the different rings of emulation
Light conditions under border.For the angle change in each direction, brightness on each pixel of three-dimensional face model can occur
Trickle respective change.The change of brightness in angle change and each pixel is corresponded by attitude dummy unit, after corresponding to
Data composition attitude photometric data.
Anti- unit of moulding obtains, from attached color element, the three-dimensional face model completing color assignment respectively, obtains from attitude dummy unit
Take attitude photometric data.The brightness flop occurring in the angle change in each direction is mapped on three-dimensional face model,
Specifically each of three-dimensional face model pixel all to revise this according to the brightness flop in attitude photometric data correspondence position
The color assignment of pixel.After completing the correction of color assignment, counter mould unit generate emulate correspondingly with different rotary angle
Faceform.
For quick obtaining with reduce data screening number of times, the anti-molding apparatus of faceform of the present invention, also include point
Class memory element.Described classification memory element according to different lighting angles to prevent faceform carry out classification storage.?
When needing to transfer human face data and compare with user's real time imaging, simply enter user's current light parameter, you can fast
Speed classification draws under this illumination condition, this user difference attitude corresponding emulation faceform.No matter user is when checking,
In image, whether facial angle tilts, and does not affect checking accuracy, improves in financial services industry the efficiency and accurately of checking
Degree.
For a person skilled in the art, can technical scheme as described above and design, make other each
Plant corresponding change and deformation, and all these change and deformation all should belong to the protection model of the claims in the present invention
Within enclosing.
Claims (9)
1. a kind of vtm machine of automatic identification face is it is characterised in that include:
Plurality of human faces pick device, sets up device for picking up and being sequentially inputted to faceform faces multiple in image one by one；
In human body after vtm machine automatic image capturing；
Faceform sets up device, for the face in image is set up corresponding three-dimensional Array Model, is then input to face mould
The anti-molding apparatus of type；
The anti-molding apparatus of faceform, for moulding counter for three-dimensional array model as artificial stereo model；
Client's face database, for storing client's face data message, and by the artificial stereo model after counter moulding with store
Client's face data message compare, feed back comparison result.
2. vtm machine according to claim 1 is it is characterised in that described plurality of human faces pick device, comprising:
Image collection module, for catching one section of image that current time starts；
Motion capture module, meets the dynamic image block of human body for catching from image；
Face extraction module, for extracting corresponding face information from image block.
3. vtm machine according to claim 2, it is characterised in that described motion capture module, is extracting all of image
After block, the left screening to image block size, the image block meeting dimensional requirement after screening is input to face extraction module.
4. vtm machine according to claim 2, it is characterised in that described plurality of human faces pick device, also includes time control
Module, for according to background environment real-time regulation image capturing parameter；The definition of background environment is lower, pull-in time is longer,
Sample frequency is lower；The definition of background environment is higher, and pull-in time is shorter, sample frequency is higher.
5. vtm machine according to claim 1 is it is characterised in that described faceform sets up device, comprising:
Two dimensional image acquiring unit, obtains the two dimensional image including face, and two dimensional image is sent to Face detection unit；
Face detection unit, to the extract facial feature in two dimensional image, sketches the contours of X-Y scheme according to face characteristic normal arrangement
Face scope in picture, reads two-dimensional coordinate to the pixel included by face scope, and two-dimensional coordinate is input to two-dimensional vector collection
Unit；
Two-dimensional vector collecting unit, according to the two-dimensional coordinate of pixel in the range of face, generates the two-dimensional vector of pixel, by all two
N dimensional vector n forms vector matrix, and vector matrix is delivered to deformation unit；
Three-dimensional face database, prestored conventional face three-dimensional information and characteristic information；
Deformation unit, vector matrix is contrasted with the three-dimensional information of storage in three-dimensional face database and characteristic information, from three-dimensional
Filter out the three-dimensional information closest with vector matrix feature and characteristic information in the information of human face data library storage, will screen
The information going out, as deformation parameter, carries out three-dimensional deformation to vector matrix and obtains deformation matrix；
Deformation constraint element, the deformation process that monitoring deformation unit is carried out to vector matrix, the arbitrary unit in constrained strain matrix
Element is not in illegal value；
Trivector signal generating unit, deformation matrix is converted to trivector group, to each trivector in trivector group
One pixel of distribution.
6. vtm machine according to claim 5 it is characterised in that described three-dimensional face database has prestored many ethnic groups and
The three-dimensional information of many colours of skin and characteristic information；Described deformation unit, in filter information, preferentially screens colour code in face pixel
Immediate three-dimensional information and characteristic information with colour code in prestored information；Described deformation constraint element will occur in deformation matrix
Illegal value constrain in the maximum legal value that can deviate.
7. vtm machine according to claim 1 is it is characterised in that the anti-molding apparatus of described faceform, comprising:
Obtaining three-dimensional model unit, obtains three-dimensional face model data, obtains the original image generating three-dimensional face model simultaneously；
Original image is obtained with the COLOR COMPOSITION THROUGH DISTRIBUTION of face pixel；
Attached color element, carries out color assignment according to original image and COLOR COMPOSITION THROUGH DISTRIBUTION to three-dimensional face model；
Attitude dummy unit, carries out axle rotation to three-dimensional face model, angle effect corresponding with illumination in virtual rotary course, raw
Become attitude photometric data；
Counter mould unit, color correction is carried out according to attitude photometric data to the three-dimensional face model completing color assignment, generate with
Different rotary angle emulates faceform correspondingly；
Classification memory element, described classification memory element stores to emulation faceform according to different lighting angles.
8. it is characterised in that described attitude dummy unit, lighting simulation adopts monochromatic light to vtm machine according to claim 7
Source is irradiated or respectively to uniform light intense irradiation；Respectively with each orthogonal axis as rotary shaft, rotation is not more than three-dimensional face model
Positive and negative 90 ° of angle；The described anglec of rotation is between -30 ° to 30 °.
9. vtm machine according to claim 1, it is characterised in that described client's face database, comparison result is non-
The artificial stereo model of client recorded in client's face database.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610715546.4A CN106340139A (en) | 2016-08-25 | 2016-08-25 | VTM (Virtual Teller Machine) capable of automatically recognizing human face |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610715546.4A CN106340139A (en) | 2016-08-25 | 2016-08-25 | VTM (Virtual Teller Machine) capable of automatically recognizing human face |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106340139A true CN106340139A (en) | 2017-01-18 |
Family
ID=57825271
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610715546.4A Pending CN106340139A (en) | 2016-08-25 | 2016-08-25 | VTM (Virtual Teller Machine) capable of automatically recognizing human face |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106340139A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107179827A (en) * | 2017-04-01 | 2017-09-19 | 深圳怡化电脑股份有限公司 | The intelligent interactive method and system of a kind of finance device |
| CN112382015A (en) * | 2020-11-11 | 2021-02-19 | 华南麟睿科技(衡阳)有限公司 | Intelligent article storing and taking method based on face recognition technology |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060078172A1 (en) * | 2004-06-03 | 2006-04-13 | Arizona Board Of Regents, A Body Corporate Of The State Of Arizona | 3D face authentication and recognition based on bilateral symmetry analysis |
| JP2008269534A (en) * | 2007-04-25 | 2008-11-06 | Sony Corp | Biometric authentication system, authentication server, terminal device, and biometric authentication method |
| CN104376249A (en) * | 2014-11-28 | 2015-02-25 | 苏州福丰科技有限公司 | Automatic teller system and processing method based on three-dimensional face recognition |
| CN104469252A (en) * | 2014-11-20 | 2015-03-25 | 深圳怡化电脑股份有限公司 | Facial image extraction achieving method, device and system for VTM |
| CN104599367A (en) * | 2014-12-31 | 2015-05-06 | 苏州福丰科技有限公司 | Multi-user parallel access control recognition method based on three-dimensional face image recognition |
| CN104732210A (en) * | 2015-03-17 | 2015-06-24 | 深圳超多维光电子有限公司 | Target human face tracking method and electronic equipment |
| CN104820829A (en) * | 2015-05-08 | 2015-08-05 | 常州大学 | Novel face recognition algorithm based on depth data and application thereof in intelligent permission assignment system |
-
2016
- 2016-08-25 CN CN201610715546.4A patent/CN106340139A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060078172A1 (en) * | 2004-06-03 | 2006-04-13 | Arizona Board Of Regents, A Body Corporate Of The State Of Arizona | 3D face authentication and recognition based on bilateral symmetry analysis |
| JP2008269534A (en) * | 2007-04-25 | 2008-11-06 | Sony Corp | Biometric authentication system, authentication server, terminal device, and biometric authentication method |
| CN104469252A (en) * | 2014-11-20 | 2015-03-25 | 深圳怡化电脑股份有限公司 | Facial image extraction achieving method, device and system for VTM |
| CN104376249A (en) * | 2014-11-28 | 2015-02-25 | 苏州福丰科技有限公司 | Automatic teller system and processing method based on three-dimensional face recognition |
| CN104599367A (en) * | 2014-12-31 | 2015-05-06 | 苏州福丰科技有限公司 | Multi-user parallel access control recognition method based on three-dimensional face image recognition |
| CN104732210A (en) * | 2015-03-17 | 2015-06-24 | 深圳超多维光电子有限公司 | Target human face tracking method and electronic equipment |
| CN104820829A (en) * | 2015-05-08 | 2015-08-05 | 常州大学 | Novel face recognition algorithm based on depth data and application thereof in intelligent permission assignment system |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107179827A (en) * | 2017-04-01 | 2017-09-19 | 深圳怡化电脑股份有限公司 | The intelligent interactive method and system of a kind of finance device |
| CN112382015A (en) * | 2020-11-11 | 2021-02-19 | 华南麟睿科技(衡阳)有限公司 | Intelligent article storing and taking method based on face recognition technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106874871B (en) | Living body face double-camera identification method and identification device | |
| WO2019128507A1 (en) | Image processing method and apparatus, storage medium and electronic device | |
| EP1413972B1 (en) | Examination of images of persons | |
| CN108717524B (en) | Gesture recognition system based on double-camera mobile phone and artificial intelligence system | |
| CN108470169A (en) | Face identification system and method | |
| CN108447159B (en) | Face image acquisition method and device and entrance and exit management system | |
| CN106529414A (en) | Method for realizing result authentication through image comparison | |
| CN204155293U (en) | A kind of demo plant based on recognition of face and verification system | |
| CN110163078A (en) | The service system of biopsy method, device and application biopsy method | |
| CN108268864A (en) | Face identification method, system, electronic equipment and computer program product | |
| CN109815960A (en) | Reproduction image-recognizing method, device, equipment and medium based on deep learning | |
| CN107111704A (en) | System and method based on biological recognition system fraud detection in iris | |
| CN105022999A (en) | Man code company real-time acquisition system | |
| US9449217B1 (en) | Image authentication | |
| CN105141842A (en) | Tamper-proof license camera system and method | |
| CN111967319B (en) | Living body detection method, device, equipment and storage medium based on infrared and visible light | |
| WO2022222575A1 (en) | Method and system for target recognition | |
| CN109598210A (en) | A kind of image processing method and device | |
| CN208351494U (en) | Face identification system | |
| CN111191521A (en) | Face living body detection method and device, computer equipment and storage medium | |
| CN204143555U (en) | The Certificate of House Property printing terminal of identification self-aided terminal and correspondence | |
| CN104217503A (en) | Self-service terminal identity identification method and corresponding house property certificate printing method | |
| CN106340139A (en) | VTM (Virtual Teller Machine) capable of automatically recognizing human face | |
| CN104217504A (en) | Identity recognition self-service terminal and corresponding certificate of house property printing terminal | |
| CN112037235B (en) | Injury picture automatic auditing method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170118 |