CN105574525B - A kind of complex scene multi-modal biological characteristic image acquiring method and its device - Google Patents
A kind of complex scene multi-modal biological characteristic image acquiring method and its device Download PDFInfo
- Publication number
- CN105574525B CN105574525B CN201510967025.3A CN201510967025A CN105574525B CN 105574525 B CN105574525 B CN 105574525B CN 201510967025 A CN201510967025 A CN 201510967025A CN 105574525 B CN105574525 B CN 105574525B
- Authority
- CN
- China
- Prior art keywords
- default
- scene
- image
- imaging
- photographed scene
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000003384 imaging method Methods 0.000 claims abstract description 215
- 238000012545 processing Methods 0.000 claims abstract description 43
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 26
- 230000008447 perception Effects 0.000 claims abstract description 14
- 230000003287 optical effect Effects 0.000 claims description 61
- 238000003331 infrared imaging Methods 0.000 claims description 28
- 238000005286 illumination Methods 0.000 claims description 23
- 230000001815 facial effect Effects 0.000 claims description 11
- 238000011112 process operation Methods 0.000 claims description 9
- 238000002366 time-of-flight method Methods 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 7
- 230000005540 biological transmission Effects 0.000 claims description 2
- 230000005021 gait Effects 0.000 abstract description 22
- 238000005516 engineering process Methods 0.000 description 16
- 230000000007 visual effect Effects 0.000 description 9
- 238000004587 chromatography analysis Methods 0.000 description 6
- 230000004927 fusion Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000013441 quality evaluation Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000035479 physiological effects, processes and functions Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 210000003462 vein Anatomy 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 208000004350 Strabismus Diseases 0.000 description 1
- 230000001154 acute effect Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005538 encapsulation Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000010348 incorporation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 238000013517 stratification Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
- G06V10/14—Optical characteristics of the device performing the acquisition or on the illumination arrangements
- G06V10/147—Details of sensors, e.g. sensor lenses
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Vascular Medicine (AREA)
- Studio Devices (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of complex scene multi-modal biological characteristic image acquiring devices, it include: the polynary imaging unit of active, logic control platform, parallel data computing platform, wherein, logic control platform, it is connected with actively polynary imaging unit, parallel data computing platform is connected with logic control platform, and parallel data computing platform specifically includes multi-sensor data processing system, biological characteristic scene perception and modeling, multi-modal biological characteristic actively perceive system and advanced imaging algorithm processing system.In addition, the invention also discloses a kind of complex scene multi-modal biological characteristic image acquiring methods.A kind of complex scene multi-modal biological characteristic image acquiring method and its device disclosed by the invention, it can be to the biological characteristic progress multimodal recognition of multiple users under the conditions of real complex scene, the identification demand for meeting multi-user in complicated image scene is realized and is effectively obtained to a wide range of interior iris, face, gait multi-modal biological characteristic.
Description
Technical field
The present invention relates to the technical fields such as optics, computer vision and pattern-recognition, more particularly to a kind of complex scene
Multi-modal biological characteristic image acquiring method and its device.
Background technique
Currently, with the continuous development of human sciences' technology, biometrics identification technology is widely used.Biology
Feature identification refers to computer by obtaining and the physiology and behavioural characteristic of analysis human body, realize science that automatic identity identifies and
Technology.Common biological characteristic mode includes fingerprint, iris, face, palmmprint, hand shape, vein, person's handwriting, gait, voice etc..People
The biological information of body multiple modalities is distributed mainly on facial (face, iris) and hand (fingerprint, palmmprint, hand shape, vein).
Compared to hand biological characteristic, the face and iris feature of human body face have can apparently show, abundant information, acquisition it is non-contacting solely
Special advantage has irreplaceable important function in medium and long distance identification and intelligent video monitoring application scenarios, thus obtains
International academic community, industrial circle or even the highest attention of government department are arrived.In addition, gait biological characteristic has can feel at a distance
Know, information is unique and easy to secret capture, non-infringements property is high, is difficult to hide and the advantages that camouflage, also gradually by some research aircrafts
The attention of structure.
Bio-identification subject and technical field pass through the development of many years, substantially can be correct under conditions of carefully controlled
The user of identification height cooperation, still, in digitlization biological information acquisition process, if the physiology by inherent user
(such as illumination is blocked, distance for variation for variation (such as blink, strabismus, posture, expression, movement) and external environment (scene where i.e.)
Deng) influence, the performance of bio-identification will sharply decline, be unable to satisfy under complex scene environment to the need of the identification of user
It asks, therefore seriously constrains subject progress, Technique Popularizing and the industry development of bio-identification.
Currently, iris and face recognition technology are developed rapidly, and are imaged compared to iris and face, and gait biology is special
Sign is easier to perceive at a distance, generally can carry out data acquisition using camera ready-made in the market.Since gait biology is special
Sign influenced by imaging device visual angle it is very big, how to be laid out multiple-camera carry out multi-angle of view acquisition, Gait Recognition will be directly affected
Precision.
In addition, multi-modal biological characteristic identification target user's range, accuracy of identification and in terms of compare single mode
The biological identification technology of state has unique advantage, therefore the research of multimodal Biometrics device is also the overdevelopment of recent years
Point.
But existing multi-modal biological characteristic obtains equipment and is mainly directed towards controlled environment, range to imageable target and
Posture and image scene environmental condition all have more stringent limitation, meanwhile, existing multi-modal biological characteristic imaging is set
The standby information that can only once obtain a people, can not acquire the multi-modal biological characteristic of multiple targets simultaneously, be unable to satisfy complexity
The identification demand of user in image scene.Currently, the convenient and efficient of high quality, multi-modal biological characteristic data obtains
As the significant bottleneck that bio-identification subject and technical application are expanded and crossed over from controlled condition to complicated actual environment, serious shadow
Ring accuracy, robustness, the real-time even safety of living things feature recognition.
Therefore, there is an urgent need to develop a kind of technologies out at present, can be to real complex scene condition (as outdoor is a wide range of
And background is complicated, ambient lighting variation, target group are more and situations such as there is movement, attitudes vibration and mutually block) under it is multiple
The biological characteristic of user carries out multimodal recognition, meets the identification demand of multi-user in complicated image scene, realizes to big
Iris, face, gait multi-modal biological characteristic in range are effectively obtained.
Summary of the invention
In view of this, the object of the present invention is to provide a kind of complex scene multi-modal biological characteristic image acquiring method and its
Device, can be to real complex scene condition (as outdoor a wide range of and background is complicated, ambient lighting changes, target group is more
And there is movement, attitudes vibration and situations such as mutually block) under the biological characteristic of multiple users carry out multimodal recognition, meet multiple
The identification demand of multi-user in miscellaneous image scene is realized to a wide range of interior iris, face, gait multi-modal biological characteristic
It is effectively obtained, further increases the accuracy of identification, realize that remote, high-resolution, high speed multi-modal biological characteristic is known
Not, the product use feeling for being conducive to improve user, is of great practical significance.
For this purpose, the present invention provides a kind of complex scene multi-modal biological characteristic image acquiring methods, comprising steps of
Step 1: the default photographed scene for needing to shoot to user provides multispectral compensation illumination；
Step 2: the multiplex images of the acquisition default photographed scene, wherein the multiplex images of the default photographed scene
The infrared figure of iris and facial image, scene high-definition picture, scene high-dynamics image, multiple users including multiple users
Picture and the depth image of multiple users；Step 3: the multiplex images to the default photographed scene carry out preset data processing
Operation, obtains the multiplex images data of initialization；
Step 4: carrying out Sensing model according to the multiplex images data of the initialization to the default photographed scene, obtaining
The corresponding real structure model of place of the default photographed scene is obtained, and detection obtains presetting in the default photographed scene
Biological information；
Step 5: it is pre- to judge whether modeling information that the real structure model of place of the default photographed scene has meets
If whether the default biological information in modeling information condition and the default photographed scene meets default biological characteristic item
Part, when the modeling information that the real structure model of place of the default photographed scene has meets default modeling information condition, together
When the default biological information that Shi Suoshu presets in photographed scene meets default biological characteristic condition, according to the default shooting
The default biological characteristic in modeling information and the default photographed scene that the corresponding real structure model of place of scene has
Information executes default image forming process operation, obtains final multi-modal biological characteristic image and shown.
Wherein, in the first step, by from the multispectral light source module that multiple multispectral light sources form to described pre-
If photographed scene provides multispectral compensation illumination.
Wherein, in the second step, by forming optical field imaging module by multiple optical field imaging sensors to acquire
State the iris and facial image of multiple users in default photographed scene；
The default photographed scene is acquired by the high-resolution imaging module being made of multiple array image sensors
High-definition picture；
The default photographed scene is acquired by the high dynamic image-forming module being made of multiple Highly-dynamic image sensors
Scene high-dynamics image；
It is acquired by the infrared imaging module being made of multiple infrared image sensors multiple in the default photographed scene
The infrared image of user；
Multiple users in the default photographed scene are acquired by the Depth Imaging module being made of multiple depth cameras
Depth image.
Wherein, the depth camera is time-of-flight method TOF camera.
In addition, the present invention also provides a kind of complex scene multi-modal biological characteristic image acquiring devices, comprising:
Actively polynary imaging unit, the default photographed scene for needing to shoot to user provide multispectral compensation and shine
It is bright, and the multiplex images of the acquisition default photographed scene, wherein the multiplex images of the default photographed scene include multiple
The iris and facial image of user, scene high-definition picture, scene high-dynamics image, the infrared image of multiple users and more
Then the multiplex images are sent to logic control platform by the depth image of a user；
Logic control platform is connected with actively polynary imaging unit, and the image acquisition for being inputted according to user starts
Perhaps the starting or out of service of the polynary imaging unit of the corresponding control active of halt instruction, and it is actively polynary receiving
After the multiplex images for the default photographed scene that imaging unit is sent, it is transmitted to parallel data computing platform in real time, and sentence
Whether the modeling information that has of real structure model of place of the default photographed scene of breaking meets default modeling information condition, with
And whether the default biological information in the default photographed scene meets default biological characteristic condition, when the default shooting
The modeling information that the real structure model of place of scene has meets in default modeling information condition and the default photographed scene
Default biological information when meeting default biological characteristic condition, send the corresponding real structure field of the default photographed scene
The default biological information in modeling information and the default photographed scene that scape model has calculates flat to parallel data
Advanced imaging algorithm processing system in platform；
Parallel data computing platform specifically includes multi-sensor data processing system, biological characteristic scene perception and modeling
System, multi-modal biological characteristic actively perceive system and advanced imaging algorithm processing system；Wherein:
Multi-sensor data processing system is connected with logic control platform, for the more of the default photographed scene
First image carries out preset data processing operation, obtains the multiplex images data of initialization；
Biological characteristic scene perception and modeling, respectively with multi-sensor data processing system and logic control platform phase
Connection carries out Sensing model to the default photographed scene, described in acquisition for the multiplex images data according to the initialization
The default corresponding real structure model of place of photographed scene, and it is sent to logic control platform；
Multi-modal biological characteristic actively perceive system, respectively with multi-sensor data processing system and logic control platform phase
Connection for detecting the default biological information obtained in the default photographed scene, and is sent to logic control platform；
Advanced imaging algorithm processing system, is connected with logic control platform, for according to the default photographed scene pair
The default biological information in modeling information and the default photographed scene that the real structure model of place answered has, holds
The default image forming process operation of row, obtains final multi-modal biological characteristic image and is shown.
Wherein, the polynary imaging unit of the active includes multispectral light source module, optical field imaging module, high-resolution imaging
Module, high dynamic image-forming module, infrared imaging module and Depth Imaging module；
It is the multispectral light source module, optical field imaging module, high-resolution imaging module, high dynamic image-forming module, infrared
Image-forming module and Depth Imaging module are connected with the logic control platform respectively.
Wherein, including crust of the device, the front center position of described device shell are provided with main imaging lens；
It is provided with infrared imaging camera lens on the left of the main imaging lens, is provided with depth on the right side of the main imaging lens
Imaging lens；
The positive surrounding of described device shell is respectively uniformly distributed there are four optical field imaging lens group；
There are multiple multispectral light around the optical field imaging lens group, infrared imaging camera lens and Depth Imaging camera lens respectively
Source.
Wherein, the logic control platform includes on-site programmable gate array FPGA and Chip Microcomputer A RM logic control chip,
Wherein:
The image that on-site programmable gate array FPGA is used to be inputted according to user obtains beginning or halt instruction, to actively
Multispectral light source module, optical field imaging module, high-resolution imaging module, high dynamic image-forming module in polynary imaging unit,
Infrared imaging module and Depth Imaging module carry out drive control, realize corresponding control actively the starting of polynary imaging unit or
Person is out of service and control parameter adjusts and the transmission buffer of data；Chip Microcomputer A RM logic control chip is for being responsible for logic
Data dispatch cooperation between control platform and parallel data computing platform.
By the above technical solution provided by the invention as it can be seen that compared with prior art, the present invention provides a kind of complexity
Scene multi-modal biological characteristic image acquiring method and its device, can be to real complex scene condition (as outdoor is a wide range of
And background is complicated, ambient lighting variation, target group are more and situations such as there is movement, attitudes vibration and mutually block) under it is multiple
The biological characteristic of user carries out multimodal recognition, meets the identification demand of multi-user in complicated image scene, realizes to big
Iris, face, gait multi-modal biological characteristic in range are effectively obtained, and the accuracy of identification is further increased, and are realized
At a distance, the multi-modal biological characteristic identification of high-resolution, high speed, is conducive to the product use feeling for improving user, has great
Production practices meaning.
Detailed description of the invention
Fig. 1 is a kind of flow chart of complex scene multi-modal biological characteristic image acquiring method provided by the invention；
Fig. 2 is a kind of structural schematic diagram of complex scene multi-modal biological characteristic image acquiring device provided by the invention；
Fig. 3 is a kind of structure distribution forward sight of complex scene multi-modal biological characteristic image acquiring device provided by the invention
Schematic diagram；
Fig. 4 is a kind of light channel structure signal of complex scene multi-modal biological characteristic image acquiring device provided by the invention
Figure；
Fig. 5 is a kind of application scenarios signal of complex scene multi-modal biological characteristic image acquiring device provided by the invention
Figure.
Specific embodiment
In order to enable those skilled in the art to better understand the solution of the present invention, with reference to the accompanying drawing with embodiment to this
Invention is described in further detail.
Fig. 1 is a kind of flow chart of complex scene multi-modal biological characteristic image acquiring method provided by the invention；
Referring to Fig. 1, a kind of complex scene multi-modal biological characteristic image acquiring method provided by the invention, including following step
It is rapid:
Step S101: the default photographed scene for needing to shoot to user provides multispectral compensation illumination；
In the present invention, in step s101, user needs the default photographed scene shot to preset according to the needs of users
It is configured, can be any one scene of user's selection, such as hall doorway scene.
It in the present invention, in step s101, can be by the mostly light that is made of multiple multispectral light sources in specific implementation
Compose light source module and provide multispectral compensation to default photographed scene and illuminate, it is therefore intended that meet the high quality of round-the-clock scene at
As demand.
Step S102: the multiplex images of the acquisition default photographed scene, wherein the polynary figure of the default photographed scene
As include multiple users iris and facial image, scene high-definition picture, scene high-dynamics image, multiple users it is infrared
Image and the depth image (obtaining the depth information of user, such as three-dimensional feature information) of multiple users；
It in the present invention, in step s 102, can be by forming light by multiple optical field imaging sensors in specific implementation
Image-forming module acquires the iris and facial image of multiple users in the default photographed scene；
In step s 102, in specific implementation, can by the high-resolution that is made of multiple array image sensors at
As module acquires the scene high-definition picture of the default photographed scene；
In the present invention, it should be noted that in specific implementation, resolution that the scene high-definition picture specifically has
Rate size is acquired after being configured according to the needs of users to high-resolution imaging module and is obtained, for example, can be with 1280*
800 pixels can also be the image of the high-resolution of other users setting certainly.
In step s 102, in specific implementation, can by the high dynamic that is made of multiple Highly-dynamic image sensors at
As module acquires the scene high-dynamics image of the default photographed scene；
In the present invention, it should be noted that in specific implementation, resolution ratio that the scene high-dynamics image specifically has
Size is acquired after being configured according to the needs of users to high dynamic image-forming module and is obtained, for example, can be with 1280*720 picture
The dynamic image of element can also be the dynamic image of the high-resolution of other users setting certainly.
It in step s 102, can be by the infrared imaging mould that is made of multiple infrared image sensors in specific implementation
Block acquires the infrared image of multiple users in the default photographed scene；
In step s 102, it in specific implementation, can be adopted by the Depth Imaging module being made of multiple depth cameras
Collect the depth image of multiple users in the default photographed scene, the depth camera preferably selects the TOF (flight time
Method) video camera.
Step S103: preset data processing operation is carried out to the multiplex images of the default photographed scene, is initialized
Multiplex images data；
In step s 103, the preset data processing operation carried out to multiplex images includes: to carry out to multiplex images
Data prediction, coding and fusion, wherein pretreatment includes that image goes hot-tempered, normalization, the calibration of light field image and removes vignetting
Deng；Coding is to carry out data encoding to multiplex images using compressive sensing theory, reduces the data redundancy of more images, improves system
System performance；Fusion is will to obtain polynary image at each image-forming module to have complementary advantages, and the figure of multiple characteristic is provided in fusion
Picture.For the present invention, the redundancy of image is removed, the quality of image is higher and is conducive to further handle analysis.
Step S104: according to the multiplex images data of the initialization, carrying out Sensing model to the default photographed scene,
It obtains pre- in the corresponding real structure model of place of the default photographed scene, and the detection acquisition default photographed scene
If biological information (such as iris feature, face characteristic and gait feature information)；
For the present invention, it should be noted that multiplex images data are under the conditions of obvious to all the period of time, illumination variation
Complex scene is perceived, and the real structure model of place for covering time, illumination is finally inversed by, and is carried out background using depth data and is built
Mould carries out completing scene three-dimensional modeling using shape and texture, to reduce the variable factors such as illumination under complex scene, weather to life
The adverse effect of object characteristic imaging.
For the present invention, it should be noted that multiplex images data are used to detect the default biological information in scene,
It in specific implementation, is examined using target positioning unit, the region of interest characteristic of field in multi-modal biological characteristic actively perceive system
Extraction unit, multi-modal biological characteristic integrated unit and multi-modal biological characteristic quality evaluation unit are surveyed, multi-modal biology is passed through
Feature detection, feature extraction, fusion and quality evaluation technology obtain multiple target face, iris and gait biological information.
Step S105: whether the modeling information for judging that the real structure model of place of the default photographed scene has meets
Whether the default biological information in default modeling information condition and the default photographed scene meets default biological characteristic
Condition, when the modeling information that the real structure model of place of the default photographed scene has meets default modeling information condition,
When the default biological information in the default photographed scene meets default biological characteristic condition simultaneously, according to the default bat
The default biology taken the photograph in the modeling information and the default photographed scene that the corresponding real structure model of place of scene has is special
Reference breath executes default image forming process operation, obtains final multi-modal biological characteristic image and shown；
Otherwise, i.e., if the modeling information that the real structure model of place of the default photographed scene has do not meet it is default
Default biological information in modeling information condition and the default photographed scene does not meet default biological characteristic condition, then
The modeling information and default biological information being had according to real structure model of place are adjusted and are provided to default photographed scene
Multispectral compensation illumination parameter and the acquisition default photographed scene multiplex images imaging parameters, then return and execute
Step S101, until the modeling information that the real structure model of place of the default photographed scene has meets default modeling information
Until default biological information in condition and the default photographed scene meets default biological characteristic condition.
In the present invention, the modeling information that the real structure model of place has includes the light conditions of scene, weather
Situation, scene domain, the shape in scene and texture information；
For the present invention, default modeling information condition is primarily referred to as the attribute of scene, for example judges the illumination of photographed scene
Situation, weather condition, scene domain, the shape in scene and texture information etc..In the present invention, the default biological characteristic packet
Include face, iris, gait etc., default biological characteristic condition mainly include face, iris, gait image quality requirements, such as
Resolution ratio, brightness, clarity of image etc..
In the present invention, scene modeling information and multi-modal biological characteristic actively perceive information are transmitted to parallel data meter
The advanced imaging algorithm processing system in platform is calculated, depth chromatography type algorithm is recycled, goes to block algorithm, three-dimensional imaging algorithm,
Complete the subsequent processing of image；
For the present invention, the multi-modal biological characteristic image finally obtained include: the unobstructed iris 2D image of multiple target,
2D/3D image when face and gait short distance and it is remote when 2D image.(explanation that application scenarios Fig. 5 can be referred to)
It, can be according to attributes and multi-modal biologies such as illumination, weather, scene domains in scene modeling for the present invention
Characteristic image quality information, come adjust multispectral light source module lighting parameter and each image-forming module (optical field imaging module,
High-resolution imaging module, high dynamic image-forming module, infrared imaging module and Depth Imaging module) imaging parameters (differentiate
Rate, aperture, focal length, the depth of field, gain coefficient etc.).
It should be noted that for the present invention, the default image forming process operation can by depth chromatography type algorithm, go
Block algorithm, three-dimensional imaging algorithm carries out.Wherein, it using the digital technology of focusing in optical field imaging again, is obtained using single exposure
The field information taken realizes the multiple target in different focus face to scene objects by adjusting weight Focusing parameter in the depth direction
Chromatography type imaging.Using synthetic aperture imaging technology, to being blocked, target carries out reduction imaging.Target is rebuild using depth information
High accuracy three-dimensional shape.Target surface localized variation feature, application are analyzed using global illumination model based on 3D shape
Such as shape from shading algorithm restores local illumination model, and in conjunction with the surface texture information of scene image, final realization is more
The three-dimensional imaging of target.
Therefore, based on above technical solution it is found that a kind of complex scene multi-modal biological characteristic figure provided by the invention
It, can be to real complex scene condition (as outdoor a wide range of and background is complicated, ambient lighting variation, target as acquisition methods
Crowd is more and situations such as there is movement, attitudes vibration and mutually block) under the biological characteristic of multiple users carry out multimodal recognition,
The identification demand for meeting multi-user in complicated image scene is realized to a wide range of interior iris, face, the multi-modal life of gait
Object feature is effectively obtained, and the accuracy of identification is further increased, and realizes remote, high-resolution, high speed multi-modal biology
Feature identification is conducive to the product use feeling for improving user, is of great practical significance.
Referring to fig. 2, in specific implementation, in order to run a kind of complex scene multi-modal biological characteristic of aforementioned present invention offer
Image acquiring method, the present invention also provides a kind of complex scene multi-modal biological characteristic image acquiring device, which includes:
Actively polynary imaging unit 201, the default photographed scene for needing to shoot to user provide multispectral compensation
Illumination, and the multiplex images of the acquisition default photographed scene, wherein the multiplex images of the default photographed scene include more
The iris and facial image of a user, scene high-definition picture, scene high-dynamics image, multiple users infrared image and
The depth image (obtaining the depth information of user, such as three-dimensional feature information) of multiple users, then sends out the multiplex images
Give logic control platform 202；
Logic control platform 202 is connected with actively polynary imaging unit 201, and the image for being inputted according to user obtains
The beginning perhaps corresponding starting or out of service for controlling actively polynary imaging unit of halt instruction is taken away, and is receiving master
After the multiplex images for moving the default photographed scene that polynary imaging unit 201 is sent, it is flat that it is transmitted to parallel data calculating in real time
Platform 203, and judge whether the modeling information that the real structure model of place of the default photographed scene has meets default build
Whether the default biological information in mould information condition and the default photographed scene meets default biological characteristic condition,
When the modeling information that the real structure model of place of the default photographed scene has meets default modeling information condition and described
When default biological information in default photographed scene meets default biological characteristic condition, the default photographed scene pair is sent
The default biological information in modeling information and the default photographed scene that the real structure model of place answered has to
Advanced imaging algorithm processing system in parallel data computing platform；
It should be noted that for logic control platform 202, if by judgement, the default photographed scene it is true
The modeling information that structure model of place has does not meet default in default modeling information condition and the default photographed scene
Biological information does not meet default biological characteristic condition, then the modeling information that is had according to real structure model of place and pre-
If biological information, the parameter and the acquisition default bat that the multispectral compensation provided to default photographed scene illuminates are adjusted
The imaging parameters for taking the photograph scene multiplex images adjust parameter and acquisition that the multispectral compensation provided to default photographed scene illuminates
Then the imaging parameters of the default photographed scene multiplex images return to operation actively polynary imaging unit 201, until described pre-
If the modeling information that the real structure model of place of photographed scene has meets default modeling information condition and the default bat
Until the default biological information taken the photograph in scene meets default biological characteristic condition；
Parallel data computing platform 203 specifically includes multi-sensor data processing system, biological characteristic scene perception and builds
Modular system, multi-modal biological characteristic actively perceive system and advanced imaging algorithm processing system；Wherein:
Multi-sensor data processing system 2031 is connected with logic control platform, for the default photographed scene
Multiplex images carry out preset data processing operation, obtain the multiplex images data of initialization；
Biological characteristic scene perception and modeling 2032 are flat with multi-sensor data processing system and logic control respectively
Platform is connected, and for the multiplex images data according to the initialization, carries out Sensing model to the default photographed scene, obtains
The corresponding real structure model of place of the default photographed scene, and it is sent to logic control platform；
Multi-modal biological characteristic actively perceive system 2033, it is flat with multi-sensor data processing system and logic control respectively
Platform is connected, and for detecting the default biological information obtained in the default photographed scene, and it is flat to be sent to logic control
Platform；
Advanced imaging algorithm processing system 2034, is connected with logic control platform 202, for according to the default shooting
The default biological characteristic in modeling information and the default photographed scene that the corresponding real structure model of place of scene has
Information executes default image forming process operation, and obtaining final multi-modal biological characteristic image and being shown (can pass through calculating
The data terminals platform such as machine is shown).
For the present invention, it should be noted that special by biological characteristic scene perception and modeling and multi-modal biology
Sign actively perceive system to actual physical scene carry out Sensing model, biological information is detected, for actively it is polynary at
It is motivated as unit provides feedback compensation, to realize the feedback closed loop robust control of system entirety.
In the present invention, it should be noted that user needs the default photographed scene shot to preset according to the needs of users
It is configured, can be any one scene of user's selection, such as hall doorway scene.
It in the present invention, can be by from the multispectral light source module that multiple multispectral light sources form to default photographed scene
Multispectral compensation illumination is provided.
In the present invention, in specific implementation, the polynary imaging unit 201 of active is comprised the following modules:
Multispectral light source module 2011, including multiple multispectral light sources, are connected with logic control platform 202, for
The default photographed scene that user needs to shoot provides multispectral compensation illumination；Purpose is to meet the high quality of round-the-clock scene
Imaging demand；
Optical field imaging module 2012, including multiple optical field imaging sensors, are connected with logic control platform 202, are used for
Acquire the iris and facial image of multiple users in the default photographed scene；
High-resolution imaging module 2013, including multiple array image sensors, are connected with logic control platform 202,
For acquiring the scene high-definition picture of the default photographed scene, therefore the big view of the default photographed scene may be implemented
The high-resolution imaging of field；
High dynamic image-forming module 2014, including multiple Highly-dynamic image sensors, are connected with logic control platform 202,
For acquiring the scene high-dynamics image of the default photographed scene；
In specific implementation, the high dynamic image-forming module 2014 can use two high dynamics with different factors
Imaging sensor obtains the high-dynamics image of the big visual field of the default photographed scene.
Infrared imaging module 2015, including multiple infrared image sensors, are connected with logic control platform 202, are used for
Acquire the infrared image of multiple users in the default photographed scene；The infrared imaging module 2015 can be convenient will be described pre-
If the background in photographed scene is split with people and interpersonal image；
Depth Imaging module 2016, including multiple depth cameras, are connected, for acquiring with logic control platform 202
The depth image of multiple users in the default photographed scene, therefore, the Depth Imaging module 2016 can be by described
Target (people etc.) in default photographed scene is imaged in the depth direction, obtains depth information of scene, convenient for described pre-
If the target (people etc.) in photographed scene perceives and establishes the three-dimensional information of scene.
In the present invention, in specific implementation, the depth camera preferably selects TOF (time-of-flight method) video camera.
In the present invention, in specific implementation, the optical field imaging module 2012 specifically uses microlens array optical field imaging
Mode acquires the iris and face figure of multiple users of high angle and high spatial resolution using integrated antenna package technology
Picture.
In specific implementation, the optical field imaging module 2012 can be by the sub- lens group of optical field imaging, high reflection mirror microscope group, half
Anti- pellicle mirror, field stitching lens group (including the first field stitching camera lens and second field stitching camera lens) and high-resolution light field
Sensor (the first high-resolution light field sensor and the second high-resolution light field sensor) group is constituted, wherein the sub- camera lens of optical field imaging
Using structure design of looking in the distance, receive multiple visual angle incident beams of scene；High reflection mirror mirror is used for multiple visual angles of scene are anti-
Light is penetrated to be deviated；Incident beam is divided into transmitted light beam and the reflected beams by half-reflecting half mirror, and transmitted light beam is through the first visual field
Splicing camera lens is incident on the first high-resolution light field sensor, is used for near-infrared iris imaging by the first high-resolution light field sensor,
Reflected light, which is incident on, is incident on the second high-resolution light field sensor through the second field stitching camera lens, is sensed by the second high-resolution light field
Device is imaged for visible light face.
In the present invention, in specific implementation, the logic control platform 202 may include FPGA (field-programmable gate array
Column) and ARM (single-chip microcontroller) logic control chip, wherein the image that FPGA is used to be inputted according to user, which obtains, to be started or stops
Instruction, in actively polynary imaging unit 201 multispectral light source module 2011, optical field imaging module 2012, high-resolution at
As module 2013, high dynamic image-forming module 2014, infrared imaging module 2015 and Depth Imaging module 2016 carry out driving control
System realizes the starting of the polynary imaging unit of corresponding control active or out of service and control parameter adjusts and the biography of data
Defeated caching；The data dispatch cooperation that ARM is used to be responsible between logic control platform and parallel data computing platform.
The parallel data computing platform includes combining the high-performance calculation processing system constituted by FPGA, DSP and GPU,
Multi-sensor data processing system, biological characteristic scene perception and modeling, multi-modal are further divided into according to function realization
Biological characteristic actively perceive system and advanced imaging algorithm processing system.
The multi-sensor data processing system, is connected with logic control platform, for the default photographed scene
Multiplex images carry out preset data processing operation, obtain the multiplex images data of initialization；In specific implementation, the polynary biography
Feel data processing system to be used to carry out processing, including data prediction list to the data that actively polynary imaging system obtains
Member, coding unit and integrated unit.
The biological characteristic scene perception and modeling are flat with multi-sensor data processing system and logic control respectively
Platform is connected, and for the multiplex images data according to the initialization, carries out Sensing model to the default photographed scene, obtains
The corresponding real structure model of place of the default photographed scene, and it is sent to logic control platform；In specific implementation, the life
Object feature scene perception with modeling be used for it is obvious to all the period of time, illumination variation under the conditions of complex scene perceive, instead
The real structure model of place of time, illumination are covered in performance, while providing accurate feedback operation ginseng for actively polynary imaging system
Number.
The multi-modal biological characteristic actively perceive system, it is flat with multi-sensor data processing system and logic control respectively
Platform is connected, and for detecting the default biological information obtained in the default photographed scene, and it is flat to be sent to logic control
Platform；In specific implementation, the multi-modal biological characteristic actively perceive system be used for multiple target multi-modal biological characteristic information into
Row actively perceive, including target positioning unit, the region of interest characteristic of field Detection and Extraction unit, multi-modal biology in complex scene
Fusion Features unit and multi-modal biological characteristic quality evaluation unit, according to the polynary imaging system of actively perceive information guiding active
Targetedly imaging parameters are carried out to configure.
As described above, the advanced imaging algorithm processing system, is connected with logic control platform 202, for according to institute
It states pre- in the modeling information and the default photographed scene that the corresponding real structure model of place of default photographed scene has
If biological information, default image forming process operation is executed, final multi-modal biological characteristic image is obtained and shown (can
To be shown by data terminals platforms such as computers).In specific implementation, the advanced imaging algorithm processing system is for realizing more mesh
Mark multi-modal biological characteristic depth chromatography type imaging and penetration three-dimensional imaging, aggrandizement apparatus Depth of field range it is same
When, device is improved to the imaging capability of shelter target.
It should be noted that for the present invention, the default image forming process operation can by depth chromatography type algorithm, go
Block algorithm, three-dimensional imaging algorithm carries out.Wherein, it using the digital technology of focusing in optical field imaging again, is obtained using single exposure
The field information taken realizes the multiple target in different focus face to scene objects by adjusting weight Focusing parameter in the depth direction
Chromatography type imaging.Using synthetic aperture imaging technology, to being blocked, target carries out reduction imaging.Target is rebuild using depth information
High accuracy three-dimensional shape.Target surface localized variation feature, application are analyzed using global illumination model based on 3D shape
Such as shape from shading algorithm restores local illumination model, and in conjunction with the surface texture information of scene image, final realization is more
The three-dimensional imaging of target.
Referring to Fig. 3, Fig. 3 is active in a kind of complex scene multi-modal biological characteristic image acquiring device provided by the invention
The structure distribution front-view schematic diagram of polynary imaging unit illustrates multispectral light source module and polynary image-forming module (including light field
Image-forming module 2012, high-resolution imaging module 2013, high dynamic image-forming module 2014, infrared imaging module 2015 and depth
Image-forming module 2016) mechanical distribution situation.
Shown in Figure 3, a kind of complex scene multi-modal biological characteristic image acquiring device provided by the invention includes dress
Shell 300 is set, the front center position of described device shell is provided with main imaging lens 303, and the main imaging lens 303 are used for
For high-resolution imaging module 2013 (high-resolution imaging module) and high dynamic image-forming module 2014 (high dynamic image-forming module)
Incident light camera lens access is provided.
It should be noted that the high-resolution imaging and high dynamic imaging are shared by main imaging lens for the present invention
The 303 same main imaging optical paths provided, for main imaging optical path using structure design of looking in the distance, aperture is big, can satisfy to the big model of scene
It surrounds as needs；It is (a kind of high that light through main imaging optical path through one group of half-reflecting half mirror is incident on array image sensor respectively
Resolution image sensor) and Highly-dynamic image sensor group, it is respectively used to high-resolution imaging and high dynamic range imaging；Institute
High dynamic sensor group is stated to be made of two high dynamic sensors with different factors.
In the present invention, the left side of the main imaging lens 303 is provided with infrared imaging camera lens 304, as infrared imaging
The incident light camera lens access of module 2015；The right side of the main imaging lens 303 is provided with Depth Imaging camera lens 305 (specially
TOF Depth Imaging camera lens), the incident light camera lens access as Depth Imaging module 2016；
The positive surrounding of described device shell is respectively uniformly distributed there are four optical field imaging lens group 306, and each group by 4
Optical field imaging camera lens camera lens composition, as the incident light camera lens access of optical field imaging module 2012, per upper and lower two light fields at
As sub- camera lens constitutes an optical field imaging modular unit.It should be noted that the optical field imaging camera lens that optical field imaging module has
Group 306 is not limited to as shown in Figure 3 four, can also be no less than two any number of.
In the present invention, referring to Fig. 3, the optical field imaging lens group 306, infrared imaging camera lens 304 and Depth Imaging mirror
There are multiple multispectral light sources 302 around first 305 respectively, multispectral light source module is formed by the multiple multispectral light source 302
2011.Namely illustrate, the multispectral light source module 2011 is distributed in the optical field imaging lens group 306, red in array
Around outer imaging lens 304 and Depth Imaging camera lens 305.
Fig. 4 is the polynary imaging system light channel structure schematic diagram of active provided by the invention, illustrates optical field imaging module
2012, high-resolution imaging module 2013, high dynamic image-forming module 2014, infrared imaging module 2015 and Depth Imaging module
2016 optical path is constituted.
The four of image-forming module are distributed in array wherein 2011 be multispectral light source module referring to Fig. 3, Fig. 4 together
Week drive control is carried out by logic control platform 202, is polynary image-forming module (including optical field imaging module 2012, high score
Resolution image-forming module 2013, high dynamic image-forming module 2014, infrared imaging module 2015 and Depth Imaging module 2016) it provides
Multispectral compensation illumination.
Main imaging optical path 402 includes main imaging lens 303, using structure design of looking in the distance, is located at actively polynary imaging unit
The object space light beam of large-scale default photographed scene is assembled in 201 center, and the light after convergence is anti-through the first half
Semi-transparent prism 403 divides for transmitted light beam and the reflected beams, wherein transmitted light beam is incident on a kind of array image sensor (high score
Resolution imaging sensor) 404 and generate the high-definition picture of big visual field, array image sensor 404 is using integrated circuit
Encapsulation technology processes to obtain；The reflected beams are projected to the second half-reflection and half-transmission prism 405, the output of the second half-reflection and half-transmission prism 405
Transmitted light beam and the reflected beams project two high-dynamics image sensing units 406, each high-dynamics image sensing unit respectively
406 include an imaging lens and a Highly-dynamic image sensor, and wherein imaging lens are for assembling incident ray
After project on corresponding sensor, the difference of two high dynamics sensing imaging units is that respective high-dynamics image senses
Device has different factors, can get the different same view field image of exposure, to realize that high dynamic is imaged.
Infrared imaging module 2015 is located at the side of main imaging lens 303, as shown in Figure 4, red for carrying out to scene
Outer imaging is made of infrared imaging camera lens and infrared imaging sensor, and incident light projects it after the convergence of infrared imaging camera lens
On the infrared imaging sensor of rear end, infrared imaging process is completed.
Depth Imaging module 2016 is located at the side of main imaging lens 303, as shown in Figure 4, deep for carrying out to scene
Degree imaging, obtains depth information of scene；The Depth Imaging module 2016 is by Depth Imaging camera lens and Depth Imaging sensor group
At, on the Depth Imaging sensor that incident light projects its rear end after the convergence of Depth Imaging camera lens, completion Depth Imaging process.
Optical field imaging module 2012 is located at the surrounding of main imaging optical path 402, and every four optical field imaging camera lenses are one group, such as schemes
Shown in 3, upper and lower two are an optical field imaging unit 410, as shown in figure 4, the optical field imaging unit 410 includes two light
Field 414, two, the half-reflection and half-transmission prism of the second reflecting mirror 413, one of the first reflecting mirror 412, two of imaging lens 411, two view
Sensing unit 418 is imaged in field splicing 415, iris imaging sensing units 416 of camera lens and a face；
Wherein, the dead astern inclination of each optical field imaging camera lens 411 is respectively arranged with first reflecting mirror
412, the longitudinally middle position of two first reflecting mirrors 412 is obliquely installed there are two the second reflecting mirror 413, and described two the
The dead astern inclination of two-mirror 413 is respectively arranged with a half-reflection and half-transmission prism 414；
The angle of 414 left end of half-reflection and half-transmission prism and horizontal plane is acute angle；The half-reflection and half-transmission prism 414 just under
Side and directly to the right are respectively arranged with the field stitching camera lens 415, wherein just positioned at the half-reflection and half-transmission prism 414
The field stitching camera lens 415 of lower section is connected with a face imaging sensing unit 418, is located at the half-reflection and half-transmission prism 414
The field stitching camera lens 415 of directly to the right be connected with an iris imaging sensing unit 416.
Referring to fig. 4, for optical field imaging module 2012, every four optical field imaging camera lenses 411 are one group, each light field at
As lens group 306 includes two optical field imaging units 410, as shown in Figure 4, the top of the infrared imaging module 2015 and depth
The lower section of degree image-forming module 2016 is respectively provided with the optical field imaging unit 410, and (practical includes two, in order to which Fig. 4 is shown
It is convenient, only draw an optical field imaging unit 410 and illustrate), incident light is distinguished after the convergence of upper and lower two optical field imaging camera lenses 411
It projects on corresponding first reflecting mirror 412, optical field imaging camera lens 411 is using structure design of looking in the distance, to receive photographed scene
Multi-angle of view incident beam, the first reflecting mirror 412 can reflect incident ray；The reflection light of first reflecting mirror 412 is again
It is projected on half-reflection and half-transmission prism 414 after the reflection of the second reflecting mirror of another set 412；The light after half-reflection and half-transmission prism 414
Line is divided into transmitted light beam and the reflected beams, wherein transmitted light beam is incident on field stitching camera lens 415, and upper and lower two optical paths are obtained
The light beam arrived carries out visual field merging, then projects the high-resolution iris imaging sensing unit 416 of its rear end, high-resolution rainbow
Film imaging sensing unit is placed in the high-resolution processed using integrated antenna package technology thereafter by microlens array group and tightly
Iris optical field imaging sensor composition, and the reflected beams are incident on field stitching camera lens 415, the light that upper and lower two optical paths are obtained
Shu Jinhang visual field merges, then projects the high-resolution face imaging sensing unit 418 of the microlens array form of its rear end,
High-resolution human face imaging sensing unit is placed in use integrated antenna package technology thereafter by microlens array group and tightly and processes
High-resolution human face optical field imaging sensor composition, that is to say, that the high score that optical field imaging module 2012 includes by it
Sensing unit is imaged in resolution iris imaging sensing unit and high-resolution human face, can be finally completed the rainbow of the big visual field of multi-angle of view
Film, face optical field imaging.
Referring to Fig. 5, a kind of complex scene multi-modal biological characteristic image acquiring device for being provided based on aforementioned present invention
Application scenarios are as shown in the figure.Iris, face and the step of the available multiple target of device (such as user 2000) provided by the invention
State image information realizes that the biological information of " multi-modal, multiple target, various dimensions " effectively obtains.
Device 1000 provided by the present invention have iris biological characteristic effectively perceive region 1001, and have face and
Gait biological characteristic effectively perceive region 1002；In specific implementation, after testing, wherein iris biological characteristic effectively perceive area
The range in domain 1001 is about away from device 1000 apart from the regional scope for 1m (L1) to 5m (L2)；Face and gait biological characteristic
Effectively perceive region 1002 is the regional scope for being 5m (L2) -15m (L3) away from 1000 distance of device.
In addition, device 100 provided by the invention can obtain multiple target in iris biological characteristic effectively perceive region 1001
Iris 2D image, and can be in the immediate area in face and gait biological characteristic effectively perceive region 1002 (away from device
1000 distance be 5m-8m within the scope of) obtain multiple target face and gait 3D rendering, compared with far region (away from 1000 distance of device
For within the scope of 8m-15m) obtain multiple target face and gait 2D image.It is application with living things feature recognition for the present invention
Background, towards real complex scene condition (as outdoor a wide range of and background is complicated, ambient lighting variation, target group is more and deposits
In movement, attitudes vibration and situations such as mutually block), by the technological incorporation in the fields such as photoelectric technology, vision computing technique with
Innovation carries out effectively detection to a wide range of interior iris, face, gait multi-modal biological characteristic and obtains, and stratification acquisition is more
The multidimensional image information of mode biological characteristic, before there is wide application in fields such as bio-identification, video monitoring, intelligent security guards
Scape.
It is also possible to actively be felt by biological characteristic scene perception and modeling and multi-modal biological characteristic
Know that system carries out Sensing model to actual physical scene, carries out detection positioning to target, is provided instead for actively polynary imaging system
Feedback correction, completes that outdoor a wide range of and background is complicated, the target biometric perception with scene under the conditions of illumination variation etc. round the clock,
The robust control imaging for realizing more scene multi-modal biological characteristic information, using imaging end and algorithm end cooperate with effectively control come
Ensure that the multivariate data of high quality obtains.
In addition, the present invention breaches the office that traditional biological characteristic-acquisition device Depth of field is limited and target object is single
Limit calculates imaging technique using optical field imaging etc., realizes and select the multiple target biological characteristic in scene different depth
Property imaging, not only expand the Depth of field of scene, also can selectively determine imaging plane, raising image quality enhances simultaneously
The flexibility of instrument and equipment；It is with depth information for generation crowd's occlusion issue, the present invention between multiple target in complex environment
Priori knowledge is identified, divided and is rejected to the shelter of scene objects using relevant calculation imaging technique, is realized to more
The more fully fine perception of biological information under crowd's target environment.
In conclusion compared with prior art, the present invention provides a kind of complex scene multi-modal biological characteristic images
Acquisition methods and its device, can be to real complex scene condition (as outdoor a wide range of and background is complicated, ambient lighting becomes
Change, target group is more and situations such as there is movement, attitudes vibration and mutually block) under the biological characteristic of multiple users carry out multimode
State identifies, meets the identification demand of multi-user in complicated image scene, realizes to a wide range of interior iris, face, gait
Multi-modal biological characteristic is effectively obtained, and the accuracy of identification is further increased, and is realized remote, high-resolution, high speed more
Mode living things feature recognition is conducive to the product use feeling for improving user, is of great practical significance.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (8)
1. a kind of complex scene multi-modal biological characteristic image acquiring method, which is characterized in that comprising steps of
Step 1: the default photographed scene for needing to shoot to user provides multispectral compensation illumination；
Step 2: the multiplex images of the acquisition default photographed scene, wherein the multiplex images of the default photographed scene include
The iris and facial image of multiple users, scene high-definition picture, scene high-dynamics image, multiple users infrared image with
And the depth image of multiple users；
Step 3: the multiplex images to the default photographed scene carry out preset data processing operation, the polynary of initialization is obtained
Image data；
Step 4: carrying out Sensing model according to the multiplex images data of the initialization to the default photographed scene, obtaining institute
The corresponding real structure model of place of default photographed scene is stated, and detection obtains the default biology in the default photographed scene
Characteristic information；
Step 5: judging whether the modeling information that the real structure model of place of the default photographed scene has meets default build
Whether the default biological information in mould information condition and the default photographed scene meets default biological characteristic condition,
When the modeling information that the real structure model of place of the default photographed scene has meets default modeling information condition, while institute
When stating the default biological information in default photographed scene and meeting default biological characteristic condition, according to the default photographed scene
The default biological information in modeling information and the default photographed scene that corresponding real structure model of place has,
Default image forming process operation is executed, final multi-modal biological characteristic image is obtained and is shown.
2. the method as described in claim 1, which is characterized in that in the first step, by by multiple multispectral light source groups
At multispectral light source module provide multispectral compensation to the default photographed scene and illuminate.
3. the method as described in claim 1, which is characterized in that in the second step, by being sensed by multiple optical field imagings
Device forms optical field imaging module to acquire the iris and facial image of multiple users in the default photographed scene；
The scene of the default photographed scene is acquired by the high-resolution imaging module being made of multiple array image sensors
High-resolution；
The scene of the default photographed scene is acquired by the high dynamic image-forming module being made of multiple Highly-dynamic image sensors
High-dynamics image；
Multiple users in the default photographed scene are acquired by the infrared imaging module being made of multiple infrared image sensors
Infrared image；
The depth of multiple users in the default photographed scene is acquired by the Depth Imaging module being made of multiple depth cameras
Spend image.
4. method as claimed in claim 3, which is characterized in that the depth camera is time-of-flight method TOF camera.
5. a kind of complex scene multi-modal biological characteristic image acquiring device characterized by comprising
Actively polynary imaging unit, the default photographed scene for needing to shoot to user provide multispectral compensation illumination, with
And the multiplex images of the acquisition default photographed scene, wherein the multiplex images of the default photographed scene include multiple users
Iris and facial image, scene high-resolution, scene high-dynamics image, the infrared image of multiple users and multiple users
Then the multiplex images are sent to logic control platform by depth image；
Logic control platform is connected with actively polynary imaging unit, the image for being inputted according to user obtain start or
Halt instruction, the starting or out of service of the polynary imaging unit of corresponding control active, and receiving actively polynary imaging
After the multiplex images for the default photographed scene that unit is sent, it is transmitted to parallel data computing platform in real time, and judge institute
Whether the modeling information that stating the real structure model of place of default photographed scene has meets default modeling information condition, Yi Jisuo
State whether the default biological information in default photographed scene meets default biological characteristic condition, when the default photographed scene
The modeling information that has of real structure model of place meet it is pre- in default modeling information condition and the default photographed scene
If biological information meets default biological characteristic condition, the corresponding real structure scene mould of the default photographed scene is sent
The default biological information in modeling information and the default photographed scene that type has is in parallel data computing platform
Advanced imaging algorithm processing system；
Parallel data computing platform, specifically include multi-sensor data processing system, biological characteristic scene perception and modeling,
Multi-modal biological characteristic actively perceive system and advanced imaging algorithm processing system；Wherein:
Multi-sensor data processing system is connected with logic control platform, for the polynary figure to the default photographed scene
As carrying out preset data processing operation, the multiplex images data of initialization are obtained；
Biological characteristic scene perception and modeling are connected with multi-sensor data processing system and logic control platform respectively
It connects, for the multiplex images data according to the initialization, Sensing model is carried out to the default photographed scene, is obtained described pre-
If the corresponding real structure model of place of photographed scene, and it is sent to logic control platform；
Multi-modal biological characteristic actively perceive system, is connected with multi-sensor data processing system and logic control platform respectively
It connects, for detecting the default biological information obtained in the default photographed scene, and is sent to logic control platform；
Advanced imaging algorithm processing system, is connected with logic control platform, for corresponding according to the default photographed scene
The default biological information in modeling information and the default photographed scene that real structure model of place has executes pre-
If image forming process operation, obtains final multi-modal biological characteristic image and shown.
6. device as claimed in claim 5, which is characterized in that the polynary imaging unit of active includes multispectral light source mould
Block, optical field imaging module, high-resolution imaging module, high dynamic image-forming module, infrared imaging module and Depth Imaging module；
The multispectral light source module, optical field imaging module, high-resolution imaging module, high dynamic image-forming module, infrared imaging
Module and Depth Imaging module are connected with the logic control platform respectively.
7. device as claimed in claim 5, which is characterized in that including crust of the device (300), in the front of described device shell
Heart position is provided with main imaging lens (303)；
It is provided with infrared imaging camera lens (304) on the left of the main imaging lens (303), the right side of the main imaging lens (303)
Side is provided with Depth Imaging camera lens (305)；
The positive surrounding of described device shell is respectively uniformly distributed there are four optical field imaging lens group (306)；
Have respectively around the optical field imaging lens group (306), infrared imaging camera lens (304) and Depth Imaging camera lens (305)
Multiple multispectral light sources (302).
8. device as claimed in claim 5, which is characterized in that the logic control platform includes field programmable gate array
FPGA and Chip Microcomputer A RM logic control chip, in which:
The image that on-site programmable gate array FPGA is used to be inputted according to user obtains beginning or halt instruction, to actively polynary
It is multispectral light source module, optical field imaging module, high-resolution imaging module, high dynamic image-forming module in imaging unit, infrared
Image-forming module and Depth Imaging module carry out drive control, realize the starting of the polynary imaging unit of corresponding control active or stop
The only transmission buffer of operation and control parameter adjusting and data；Chip Microcomputer A RM logic control chip is for being responsible for logic control
Data dispatch cooperation between platform and parallel data computing platform.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510967025.3A CN105574525B (en) | 2015-12-18 | 2015-12-18 | A kind of complex scene multi-modal biological characteristic image acquiring method and its device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510967025.3A CN105574525B (en) | 2015-12-18 | 2015-12-18 | A kind of complex scene multi-modal biological characteristic image acquiring method and its device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105574525A CN105574525A (en) | 2016-05-11 |
| CN105574525B true CN105574525B (en) | 2019-04-26 |
Family
ID=55884632
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510967025.3A Active CN105574525B (en) | 2015-12-18 | 2015-12-18 | A kind of complex scene multi-modal biological characteristic image acquiring method and its device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105574525B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106504751A (en) * | 2016-08-01 | 2017-03-15 | 深圳奥比中光科技有限公司 | Self adaptation lip reading exchange method and interactive device |
| JP7018022B2 (en) | 2016-11-29 | 2022-02-09 | ヌヴォトンテクノロジージャパン株式会社 | Solid-state image sensor and range-finding image sensor |
| CN107016339A (en) * | 2017-03-01 | 2017-08-04 | 泰州市创新电子有限公司 | Multizone collaboration identity corrector with information feedback function |
| CN107590461B (en) * | 2017-09-12 | 2021-04-02 | Oppo广东移动通信有限公司 | Face recognition method and related product |
| CN111465937B (en) | 2017-12-08 | 2024-02-02 | 上海科技大学 | Face detection and recognition method employing light field camera system |
| CN108108693B (en) * | 2017-12-20 | 2019-02-19 | 深圳市安博臣实业有限公司 | Intelligent identification monitoring device and recognition methods based on 3D high definition VR panorama |
| CN108197566B (en) * | 2017-12-29 | 2022-03-25 | 成都三零凯天通信实业有限公司 | Monitoring video behavior detection method based on multi-path neural network |
| CN108478207B (en) * | 2018-01-09 | 2021-08-06 | 南京航空航天大学 | Multi-modal brain physiological monitoring system |
| CN108509868B (en) * | 2018-03-12 | 2020-08-04 | 杭州软库科技有限公司 | Face recognition system and method based on light field camera |
| CN108710859A (en) * | 2018-05-23 | 2018-10-26 | 曜科智能科技（上海）有限公司 | Method for detecting human face and equipment, face identification method and equipment and storage medium |
| GB202011144D0 (en) | 2019-08-28 | 2020-09-02 | ams Sensors Germany GmbH | Systems for characterizing ambient illumination |
| CN110610160B (en) * | 2019-09-16 | 2022-04-19 | 长江师范学院 | Multi-biological-feature fusion identity authentication device |
| CN110781856B (en) * | 2019-11-04 | 2023-12-19 | 浙江大华技术股份有限公司 | Heterogeneous face recognition model training method, face recognition method and related device |
| CN111191644B (en) * | 2020-04-10 | 2020-10-20 | 支付宝(杭州)信息技术有限公司 | Identity recognition method, system and device |
| CN112507781B (en) * | 2020-10-21 | 2023-11-21 | 天津中科智能识别产业技术研究院有限公司 | Multi-dimensional multi-mode group biological feature recognition system and method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102800126A (en) * | 2012-07-04 | 2012-11-28 | 浙江大学 | Method for recovering real-time three-dimensional body posture based on multimodal fusion |
| CN102810154A (en) * | 2011-06-02 | 2012-12-05 | 国民技术股份有限公司 | Method and system for biological characteristic acquisition and fusion based on trusted module |
| CN103745242A (en) * | 2014-01-30 | 2014-04-23 | 中国科学院自动化研究所 | Cross-equipment biometric feature recognition method |
| CN104246796A (en) * | 2012-04-13 | 2014-12-24 | 高通股份有限公司 | Object recognition using multi-modal matching scheme |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130250087A1 (en) * | 2012-03-23 | 2013-09-26 | Peter A. Smith | Pre-processor imaging system and method for remotely capturing iris images |
-
2015
- 2015-12-18 CN CN201510967025.3A patent/CN105574525B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102810154A (en) * | 2011-06-02 | 2012-12-05 | 国民技术股份有限公司 | Method and system for biological characteristic acquisition and fusion based on trusted module |
| CN104246796A (en) * | 2012-04-13 | 2014-12-24 | 高通股份有限公司 | Object recognition using multi-modal matching scheme |
| CN102800126A (en) * | 2012-07-04 | 2012-11-28 | 浙江大学 | Method for recovering real-time three-dimensional body posture based on multimodal fusion |
| CN103745242A (en) * | 2014-01-30 | 2014-04-23 | 中国科学院自动化研究所 | Cross-equipment biometric feature recognition method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105574525A (en) | 2016-05-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105574525B (en) | A kind of complex scene multi-modal biological characteristic image acquiring method and its device | |
| Bai et al. | Smart guiding glasses for visually impaired people in indoor environment | |
| US11693242B2 (en) | Head-mounted display for virtual and mixed reality with inside-out positional, user body and environment tracking | |
| CN107635129B (en) | Three-dimensional trinocular camera device and depth fusion method | |
| CN112507781B (en) | Multi-dimensional multi-mode group biological feature recognition system and method | |
| JP6974873B2 (en) | Devices and methods for retrieving depth information from the scene | |
| KR101874494B1 (en) | Apparatus and method for calculating 3 dimensional position of feature points | |
| CN104036488B (en) | Binocular vision-based human body posture and action research method | |
| US9432593B2 (en) | Target object information acquisition method and electronic device | |
| Zhu et al. | Panoramic virtual stereo vision of cooperative mobile robots for localizing 3d moving objects | |
| US10674139B2 (en) | Methods and systems for human action recognition using 3D integral imaging | |
| CN105554385A (en) | Remote multimode biometric recognition method and system thereof | |
| CN102855471B (en) | Remote iris intelligent imaging device and method | |
| CN108234984A (en) | Binocular depth camera system and depth image generation method | |
| US20200210733A1 (en) | Enhanced video-based driver monitoring using phase detect sensors | |
| CN105357515A (en) | Color and depth imaging method and device based on structured light and light-field imaging | |
| CN107004275A (en) | For determining that at least one of 3D in absolute space ratio of material object reconstructs the method and system of the space coordinate of part | |
| KR102441171B1 (en) | Apparatus and Method for Monitoring User based on Multi-View Face Image | |
| CN116194866A (en) | Alignment of images from separate cameras using 6DOF pose information | |
| KR101444538B1 (en) | 3d face recognition system and method for face recognition of thterof | |
| KR20130107981A (en) | Device and method for tracking sight line | |
| CN108415875A (en) | The method of Depth Imaging mobile terminal and face recognition application | |
| GB2544263A (en) | Systems and methods for imaging three-dimensional objects | |
| US11882354B2 (en) | System for acquisiting iris image for enlarging iris acquisition range | |
| KR20120002723A (en) | Device and method for recognizing person by using 3 dimensional image information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190107 Address after: Unit 1005, Block MSD-G1, Teda, 57 Second Avenue, Jinshi Economic and Technological Development Zone, 300457 Applicant after: TIANJIN ZHONGKE HONGXING TECHNOLOGY CO.,LTD. Address before: 300457 No. 57, Second Avenue, Economic and Technological Development Zone, Binhai New Area, Tianjin Applicant before: TIANJIN ZHONGKE INTELLIGENT IDENTIFICATION INDUSTRY TECHNOLOGY RESEARCH INSTITUTE Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A method and device for obtaining multimodal biological feature images in complex scenes Granted publication date: 20190426 Pledgee: Bohai Bank Co.,Ltd. Tianjin Branch Pledgor: TIANJIN ZHONGKE HONGXING TECHNOLOGY CO.,LTD. Registration number: Y2024120000040 |