By reference to one exemplary embodiment, the purpose of the present invention and function and the side for realizing these purposes and function
Method will be illustrated.However, the present invention is not limited to one exemplary embodiment disclosed below；Can by multi-form Lai
It is realized.The essence of specification is only to aid in the detail of the various equivalent modifications Integrated Understanding present invention.
Mobile terminal iris recognition identification authentication device according to the present invention and method are in order to solve traditional iris image
Imaging device and identifying system are difficult to miniaturization with the problem of mobile it is proposed that a kind of reality is exercisable, easy-to-use
Mobile terminal iris identification device and method.The present invention has designed and developed and can be used for mobile terminal, user-friendly a new generation
Iris identification device and method, specifically, propose a kind of mobile terminal rainbow with active vision, tactile and sense of hearing interaction mechanism
Film identifying device and method.
Fig. 1 shows and is known according to the mobile terminal iris with active vision, tactile and sense of hearing interaction mechanism of the present invention
The block diagram of system of other device 100.This device 100 includes human-computer interaction module 110, multispectral iris image optically detecting mould
Block 120, iris image analysis processing module 130, feedback control module 140 and supply module 150.
Specifically, described human-computer interaction module 110 be used for completing vision in iris image acquiring processing procedure, tactile and
Multiple human-computer interaction function such as sense of hearing, so that user is configured to device 100 and is obtained the functions such as identification information.Described many
Spectrum iris image optically detecting module 120 is used for realizing multispectral, auto-focusing iris image acquiring.Special iris figure
As analysis and processing module 130 is used for the iris image collecting being analyzed process, and by result or instruction (as focused
Control instruction, picture quality result etc.) it is delivered to feedback control module 140.Feedback control module 140 is used for dividing iris image
The image analysis processing result that analysis processing module 130 obtains feeds back to multispectral iris image optically detecting module 120, and according to
Rule set in advance carries out Intelligent adjustment to the imaging parameters of optically detecting module 120；Meanwhile, feedback control module 140 is gone back
Image analysis processing result for obtaining image processing module 130 feeds back to human-computer interaction module 110, and according to setting in advance
Fixed rule output is instructed and is fed back to use to user.Described supply module 150 is used for equipment is powered, example
To realize such as by rechargeable battery.The equipment AC-less power being located in mobile terminal iris identification device 100 is powered
In the case of, this supply module is powered for device 100.When equipment insertion AC power, this supply module can be filled
Electricity.The detailed description of modules is described below with regard to accompanying drawing.
The each functions of modules of system describes in detail
Fig. 2 a further illustrates the composition frame chart of the human-computer interaction module 110 shown in Fig. 1.This human-computer interaction module
110 include possessing the display screen 111 of multi-point touch function, possess the microphone 112 of voice collecting function, have voice output work(
The loudspeaker 113 of energy, the positioning indicator 114 for instruction state, range finder module 115 and input/output interface 116.Man-machine friendship
Mutually module 110 is for completing multiple people such as the visual interactive in iris image acquiring processing procedure, haptic interaction and sense of hearing interaction
Machine interactive function, thus instructing user to use iris authentication system, improves user-friendliness.Specifically, this human-computer interaction module
110 mainly implement function such as：
A) active vision feedback：By display screen, the information of the user of Real-time Collection is shown with video mode
Come, thus the video that user can see according to oneself quickly adjusts to optimal acquisition position；
B) image processing and analyzing result shows：By display screen, the image matter that image processing algorithm analysis can be obtained
Amount judged result (as clear, unintelligible, please away from) etc. information show in real time and utilize different colors or verbal feedback extremely
User, thus user can perceive current acquisition state in real time；
C) control parameter input：By display screen 111, user can be configured it is also possible to right to the parameter of system
Acquisition process process is fed back and is interacted.For example, by touch-screen, user can be by way of clicking on area-of-interest
Specify the position needing focusing.Image processing algorithm can also obtain iris center by this area-of-interest of user input
One rough estimate, realizes the function of iris image coarse positioning.Again for example, during registration, operating personnel (or user
Oneself) can be selected by the image that touch-screen obtains to registration algorithm, such as click on screen representation and accept this image, streak
Screen representation abandons this image.Again for example, by touch-screen, system can be controlled, for example, can be arranged by touch-screen
Device parameter, input oneself ID card No. and password, equipment output result is confirmed etc..Preferably, described display
Screen possesses multi-point touch function to realize more preferable user's manipulation.
D) phonetic entry：By built-in microphone 112, active human-computer interaction module 110 can receive user's input
Voice command, and execute corresponding operation.
E) audio feedback：By having the loudspeaker 113 of voice output function, system can be by feedback command, image
Reason analysis result, identification authentication result etc. feed back to user in the way of voice, improve the practical degree of convenience of user.For example,
Can by image processing algorithm or range finder module analysis obtain customer location information (as " please away from a bit ", " please be close
A bit ", " please keep " etc.) feed back to user, with auditory tone cues user's adjustment position.
F) indicate function：Described active human-computer interaction module 110 can also preferably comprise the state for instruction state
Indicator 114, this positioning indicator 114 can be realized by sending the indicator lamp of particular color light, by instruction
Lamp can by image analysis processing result, identification authentication result etc. with indicator lamp in the way of show.For example, the instruction of green
Lamp can represent and is identified by, and the indicator lamp of redness can represent recognition failures.
G) distance measurement function：Described active human-computer interaction module 110 can realize distance measurement function further, works as real-time measurement
Front user and the distance of iris optical imaging device, thus obtain optimal iris imaging effect.As shown in Figure 2 a, range finding mould
Block 115 is made up of hardware range finder module 1151 and algorithm range finder module 1152.Wherein, hardware range finder module 1151 is red by launching
Outer light or laser measure apart from d1 to user and mobile terminal iris identification device 100；Algorithm range finder module
1152 pass through to calculate the iris radius of collected iris image, by the distance of radius size estimated service life person and device
d2.The estimated distance value of hardware range finder module 1151 and algorithm range finder module 1152 acquisition merges after filtering afterwards, is used
The estimated distance D of person.Fusion formula is as follows：
D=0.5* (d1+d2) (1)
Fig. 2 b shows the calculating of estimated distance and uses flow chart.It comprises the following steps that：
Step S201：By hardware range finder module 1151 measure user and mobile terminal iris identification device 100 away from
Step S202：User is obtained with device apart from d2 by algorithm range finder module 1152；
Step S203：D1 and d2 obtaining in step S201 and step S202 is carried out fusion and obtains estimated distance D；
Step S204：Processed according to distance value point situation：If current distance (i.e. estimated distance D) is not mobile whole
In effective acquisition range of end iris identification device, (effective acquisition range is designated as [D_min, D_max], wherein D_min, D_max
Determine when determining in collection camera lens and CMOS.For different optical lens parameters, this parameter is probably change),
Then in step S205, point out user's adjustment position by human-computer interaction module 110, such as away from, close, move left and right, then
System carries out Resurvey, then the result collecting is entered row distance again through hardware range finder module and algorithm range finder module estimates
Calculate, repeat above step S201-S204；If current distance is in effective acquisition range of mobile terminal iris identification device
Interior, by voice or screen instruction prompting user " current location is suitable, please keep a period of time ", will currently estimate simultaneously
It is transferred to automatic focusing module apart from D, provide initial focusing range for automatic focusing module.
Step S206：Run auto-focusing algorithm, after collecting the image that a width is focused, to image in step S207
Carry out Treatment Analysis, and transmit recognition result to human-computer interaction module 110 in step S208, feed back to user.
Fig. 3 a further illustrates the composition frame of the described multispectral iris image optically detecting module 120 shown in Fig. 1
Figure.As shown in Figure 3 a, described multispectral iris image optically detecting module 120 is by optical imagery module 121 and multispectral illumination
Unit 122 forms.Wherein, optical imagery module 121 is by special image acquiring sensor 211, special iris capturing optics
Camera lens 212, zoom motor 213 form, for realizing multispectral, auto-focusing iris image acquiring.This module both can be adopted
It is used as front-facing camera and be placed on the front end of mobile device it is also possible to be placed on after mobile device as post-positioned pick-up head
End.IMAQ distance is 5-30cm, more than effective depth of field 10cm.
Multispectral lighting unit 122 is arranged in a combination by the illuminating lamp of multiple different-wavebands, is used for sending some strength
Light.Under the illumination of multispectral lighting unit 122, imaging modules 121 can obtain multispectral iris image.Multispectral photograph
Bright system includes main light source 1221 and secondary light source 1222 again.Main light source 1221 is indispensable lighting unit, for example with wavelength
Scope in the near-infrared illumination LED of 750-950nm, for gathering near infrared iris image；Secondary light source 1222 is optional
Lighting unit, can be preferably other wavelength illumination lamps such as blue and green light, gold-tinted, visible ray, for gathering multispectral iris figure
Picture.The multispectral iris image being gathered can be used for focusing position and judges the In vivo detection with iris image.Lighting unit 122
Locus arrangement method according to actual conditions flexible configuration, can on camera lens side, both sides are it is also possible in mobile lens
Offside.Fig. 7-9 shows multispectral lighting unit in mobile computing device (such as mobile phone, panel computer, notebook computer)
Optional location layout mode during enforcement, in being embodied as, the layout of multispectral lighting unit for example can select in figure to justify
Fork () combination of any one or more in the position that marks, that is, in the frame surrounding of mobile computing device.
Optical imagery module 121 by image acquiring sensor 211 (such as CMOS, CCD), iris capturing optical lens 212,
Motor 213 and multispectral optical filter 214 form.This module 121 can gather the iris image of user's right and left eyes simultaneously.Have two
Plant implementation, one kind is to gather user using the imaging modules of single high-resolution (more than 5,000,000 pixels) simultaneously
Iris of both eyes image, another kind is the iris figure gathering user's right and left eyes using the imaging modules of two low resolution respectively
Picture.When using single high-resolution imaging module, the image acquiring sensor resolution ratio being adopted is not less than 3,000,000 pixels.
Fig. 7, Fig. 8, Fig. 9 respectively illustrate optional position when optical imagery module is implemented in mobile phone, panel computer, notebook computer
Layout type, can be assemblied in before equipment it is also possible to be assemblied in equipment back side.In figure, empty circles (zero) have marked employing
Location layout mode during single imaging modules, solid circles (●) have marked using location layout side during two imaging modules
Preferably, when using two imaging modules, the distance between two imaging modules are between 6-7cm.
Preferably, the multispectral optical filter 214 that this optical imagery module 121 comprises is to make multispectral lighting unit
The optical filter that the light of the light source included in 122 passes through can be a piece of optical filter comprising multiple spectrum composition it is also possible to
It is made up of the optical filter of multi-disc single spectrum.
Preferably, this module has autozoom function, and wherein zoom function can be carried by image acquiring sensor
Automatic focusing system complete it is also possible to by special iris image Focus value analysis method, motor 213 is controlled and
Realize zoom function.
Especially, the present invention adopts a kind of auto-focusing control algolithm of iteration, first, collects image, and passes through rainbow
Film image definition algorithm, obtain change zoom lens focal length after obtain a series of images definition description, when
When to picture rich in detail, send a zoom lens registration signal, and think that focal length now is optimum focal length.
Fig. 3 b shows the implementation method of the autozoom function of optical imagery module 121.Specifically, as shown in Figure 3 b,
Described given auto-focusing control system focus process is as follows：
Step S301：Input focusing position.By residing multi-point touch control display screen, manual putting touches (or mouse to user
Click on) area-of-interest (i.e. focusing position).Described area-of-interest is to be with screen touch point (or mouse clicking point)
The heart, the square with H as the length of side.Especially, the present invention supports 2 points of focal modes, and that is, user can click on the iris in screen
The region of right and left eyes, realizes the entirety focusing to eyes imaging region.
Step S302：The range information obtaining currently used person is measured according to above-mentioned distance-measurement module, and root
Obtain the scope [P of auto-focusing according to range informationmin,Pmax].As previously described, because having had learned that the distance of currently used person
Information, in theory, if this range information is accurate, can uniquely determine that (i.e. optical lens arrives a focusing position
The distance of sensor).But because this distance is estimated to obtain, it is inaccurate, has certain error, therefore, focusing position
It is also uncertain, but within the scope of one, i.e. [Pmin,Pmax])；
Step S303：In the range of auto-focusing, gather the image of multiple focusings.Herein, a focusing corresponding
Individual iris capturing optical lens 212 arrives the distance of image acquiring sensor 211.Motor control iris capturing optical lens 212 from
PmaxP is moved to towards image acquiring sensor 211 with certain step-lengthminIt is assumed that step-length is A, then the position moved is [Pmin,
Pmin+A,Pmin+2A,…,Pmax], and piece image is gathered on each position.
Step S304：By sharpness computation, obtain optimal focusing position：Calculate area-of-interest at each position
Focus value, if the definition of each point is Fmin,Fmin+A,Fmin+2A,Fmin+3A,…,Fmax, and find highest definition Fbest
Corresponding position, that is, optimal focusing position Pbest.Especially, when user inputs the focusing position of eyes, need a left side
The definition of right eye area-of-interest carries out merging definition F obtaining final area-of-interestroi, concrete fusion method is：
Wherein, FleftAnd FrightDefinition for right and left eyes area-of-interest.
Step S305：Obtain optimal focusing position PbestAfterwards, iris capturing optical lens 212 and image are adjusted by motor
The relative position of collection sensor 211, imaging modules are adjusted to this optimal focusing position Pbest.
Step S306：Optimal focusing according to obtaining updates the focusing range in next round focus process.Obtain current
Definition highest focusing corresponding focusing position Pbest, and update minimax focusing model during position adjustment next time
Enclose, and step-length is adjusted to A '=0.2A.I.e.
As it was previously stated, this method is the method for an iteration, after each iteration, focusing range is gradually reduced.
Step S307：Repeat step S303 to step S306, until in front and back adjacent move twice after the F obtainingbestDifference
Different less than certain threshold value, such as 5%.
Especially, in order to improve the processing speed of system, iterations herein and its threshold value of convergence can be according to realities
Border application scenarios are different and flexibly arrange, the minimum number of times of iteration is 1, the most very much not should exceed 3 times.
Especially, in order to improve adaptivity under different photoenvironments for the system, described definition of iris image is calculated
Method, on the basis of conventional iris image definition algorithm, has merged the saturation infromation of image, specifically, if image is full
With spend height, then to image definition calculated value increase a penalty term, such as 10% punishment；If the saturation of image is spent
Low, then a gain term, such as 10% gain are increased to image definition calculated value.
Fig. 4 a shows the composition frame chart of described special iris image analysis processing module 130.This module is by special
Iris image processing module 131, image processor 132 and memory 133 form.Iris image analysis processing module 130 is used for
The iris image collecting is analyzed process, such as Quality estimation, auto-focusing calculate, feature extraction and than reciprocity function.
Image processor and memory can run iris image Treatment Analysis and process.Both can be directly using shifting in image processing process
Moving platform (include smart mobile phone, notebook computer, panel computer) in existing computing unit (such as notebook computer
Arm processor in Intel processor or smart mobile phone) realizing the function of graphics processing unit it is also possible to separately configure specially
Carrying out above-mentioned image procossing, these different modes broadly fall into the present invention's to the graphic process unit (such as ARM, DSP etc.) of door
Spirit and essential scope.
Fig. 4 b shows the workflow of the iris image analysis processing module 130 in Fig. 4 a.As shown in Figure 4 b, described
The process that iris image processing module 131 is carried out mainly includes iris image quality analysis, iris preprocessing, iris image
In vivo detection, iris image feature coding and iris image feature templates compare five parts.
In step S401, iris image is gathered by multispectral iris image optically detecting module 120.
In step S402, by iris image quality analysis module, the iris image of acquisition is analyzed processing, and
Audio visual interaction feedback system feedback is passed through to user according to the Quality estimation result that obtains of analysis, be used for instructing user's positioning or
Person is used for iris image acquiring unit reasonably being configured and controlling.Specifically, when iris image quality analysis module is sentenced
Break iris image off quality when, can by audio visual interaction feedback system feedback to user's Resurvey instruction,
To instruct user's Resurvey iris, till the quality of iris image is reasonable.
Then in step S403, up-to-standard iris image is pre-processed.First sub-steps are that iris is slightly fixed
Position, in this step, Primary Location iris Position Approximate in the picture, if can't detect suitable iris image, need
Return to step S401 prompting user's Resurvey iris image.User carrys out adjustment position according to prompting, and eye alignment is shot
Camera.If the position of iris is detected, carrying out the second sub-steps and iris segmentation is carried out to image, obtaining iris
Exradius, and the big mere suggestion user's adjustment position according to obtained exradius, such as too big, then prompting user away from
From a little further, if too little, point out user near a bit.Meanwhile, the position in the picture according to eyes, carries to user
Show, if too inclined, point out the suitable shift position of user.Additionally, pretreatment module also includes positioning eyelid, eyelashes of image etc.
Occlusion area.After segmentation obtains iris effective coverage, under the 3rd sub-steps, iris image normalization is carried out to image, that is,
Based on iris inside and outside circle positioning result, carry out rectangular co-ordinate to polar conversion, by the original iris image under rectangular co-ordinate
Normalize to the rectangular area of a fixation under polar coordinates.
Then, in step S404, In vivo detection is carried out to iris image.Run iris image In vivo detection algorithm, judge to work as
Before the iris image that collects whether from the eyes of live body, rather than from counterfeit, such as glass eyeball, print iris and shine
Piece, iris video, U.S. pupil etc..Come from the iris image of counterfeit when present image is detected, then enter step S405
Prompting or report to the police, thus avoiding system to be subject to the attack of counterfeit it is ensured that the security of iris authentication system.Especially, institute
The iris image In vivo detection algorithm stated is analyzed based on multispectral image, and multispectral image comprises visible images, closely red
Outer image and blue light images or green light image (or optical imagery of its all band), processing method used is texture analysis
And algorithm for pattern recognition, texture analysis method including but not limited to local binary patterns feature, Gabor textural characteristics etc., pattern
Recognition methods includes but is not limited to SVMs, Adaboost, neutral net scheduling algorithm.
If judging, presented iris image is derived from live body, enters step S406 and carries out iris feature extraction and coding.Returning
One changes extraction iris texture characteristic code (as traditional binary Gabor phase code, sequencing condition code etc.) on image, is worked as
The iris feature template of front image.
Then, carry out iris feature template matching in step S 407.By prestoring in this feature code and registry
Registration feature code in iris feature template database carries out the comparison clearing of Hamming distance (Hamming Distance), and
Identification authentication result is drawn according to alignment score, thus verifying to the identity of currently used person.If authentication failed, enter
To step S405, checking end or warning；If being proved to be successful, method terminates.
Especially, described iris image compares and can complete in described mobile terminal iris identification device, also may be used
To complete in remote server, described remote server includes the data processing server at individual data center, also may be used
To be analogous to store virtual server beyond the clouds.When comparing in the machine, registration feature should be previously stored in the machine
Memory cell in or be downloaded in advance the memory cell of the machine by network.When comparing in remote server,
The iris image collecting completes iris image quality analysis, iris preprocessing, monitored in vivo and feature coding in the machine
Process, only transmits the iris feature being drawn into code to remote server, after server end completes to compare, by recognition result
Or follow-up result is transmitted to the machine.
Especially, for the security of guarantee information processing system, under carrying out remote server mode of operation, described
Feature masterplate also need to check hardware identity (ID) information of described mobile terminal device, such as the state in smart mobile phone application
Border IMSI (IMSI), hard disk serial number of panel computer or notebook computer etc. can uniquely indicate this movement
The information of terminal.
Especially, image recognition processes support Real time identification pattern and non real-time recognition mode.Under Real time identification pattern,
System is processed, encoded and is compared to every piece image, is identified result.Under non-real-time mode, system gathers first
Iris image in the range of certain time, then therefrom selects 1-3 up-to-standard image, and is identified comparing.
User uses the process of apparatus of the present invention
User is known using the mobile terminal iris with active vision, tactile and sense of hearing interaction mechanism of the present invention
The implementation that other device 100 carries out authentication is as follows：
A) user is brought into operation by clicking on the iris capturing processing system software in startup mobile terminal；
B) under conditions of radiation of visible light, system starts to gather iris image.In the process, need user suitable
Adjusting position so that iris optical imaging device can collect eyes or the image of at least eyes；
C) user clicks on the area-of-interest needing to focus on；
D) system passes through range finder module or image analysis algorithm obtains the distance range of currently used person and device；
E), after obtaining area-of-interest position and distance range, device starts near-infrared illuminator, and starts to gather one
The near-infrared iris image of the different focal position of group.Herein, different focal positions are by the configuration to iris optical imagery module
F) image acquisition and processing module carries out Treatment Analysis to the image sequence collecting, and calculates each image region of interest
The mass fraction in domain, and optimal focal position is obtained based on mass fraction；
G) optical imagery module is configured to this best focus position, the mark that display image is aligned on the display screen simultaneously
(as the check mark of green).The mark of described image be aligned can also be realized by other instruction equipments on mobile terminal, if
Send the indicator lamp of particular color light；
H) optical imagery module gathers several near-infrared iris images near this focal position；
I) image analysis processing module is processed based on the image that h step collects, and obtains mass fraction highest three
Width image, carries out feature extraction and coding to it；
J) feature coding of three width optimized images described in and the condition code of registration in advance are compared, and obtain authentication
Result.Herein, the iris feature code of registration in advance can be stored in locally it is also possible to be stored on the server of far-end.If
It is the former, comparison process locally completes in mobile terminal；If the latter, comparison process completes in the server of far-end.
K) comparison result is processed.If comparing unsuccessfully, this time ID authentication request of refusal user, and answer specific
With middle startup warning device；If compared successfully, by using this time ID authentication request of person, and it is allowed to carry out subsequently
Especially, above-mentioned using method can adopt forward type, it would however also be possible to employ rear-mounted, that is, harvester is arranged on hand
Machine front portion or back.
Especially, the click described in above-mentioned steps may refer to mouse click it is also possible to refer to finger or hand-written style of writing
Touch.Especially, when optical pickup collects binocular images, described area-of-interest refers to iris and the lesser ring of Merkel of eyes
Domain；When only optical pickup only collects eyes, eyes that described area-of-interest refers to currently to be collected
Iris and pupil position.