CN106897678A - A kind of remote human face recognition methods of combination heartbeat signal, device and system - Google Patents
A kind of remote human face recognition methods of combination heartbeat signal, device and system Download PDFInfo
- Publication number
- CN106897678A CN106897678A CN201710069368.7A CN201710069368A CN106897678A CN 106897678 A CN106897678 A CN 106897678A CN 201710069368 A CN201710069368 A CN 201710069368A CN 106897678 A CN106897678 A CN 106897678A
- Authority
- CN
- China
- Prior art keywords
- face
- pedestrian
- image
- locus
- heartbeat signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Image Processing (AREA)
Abstract
The invention provides a kind of remote human face recognition methods of combination heartbeat signal, device and system, methods described comprises the steps:Image in acquisition monitoring region, pedestrian image is extracted from the image for collecting and pedestrian locus is determined;According to pedestrian locus, the locus of pedestrian's heartbeat signal is obtained using light vibration imaging technique;According to heartbeat signal and the natural place relation of face, the face locus of pedestrian is positioned;According to face locus, adjustment pickup area to human face region gathers clear face image.Present invention utilizes the locus of heartbeat signal, the locus of face, and the rapid locating human face region of energy are quickly calculated, capture high-resolution human face image, improve the discrimination and practicality of remote human face identifying system.
Description
Technical field
The invention belongs to living things feature recognition field, more particularly to a kind of remote human face identification side of combination heartbeat signal
Method, device and system.
Background technology
Recognition of face refers to a kind of technology that authentication is carried out using the invariant feature of face.Face is used as a kind of allusion quotation
The biological characteristic of type, keeps with the uniqueness different from other people and in the regular period constant stability.Compared to tradition
Identity identifying method (identity card/passport/key/password etc.), biological characteristic will not be lost, be difficult to forge and palm off, at any time may be used
With being preferable authentication carrier.
Recognition of face simultaneously is a kind of without invasive, convenience, friendly living things feature recognition mode.It is according to camera
Distance to user can be classified as closely recognition of face and remote human face identification technology.Closely recognition of face generally needs
User cooperates, and remote human face identification technology does not need the cooperation of object, and object will not also discovered, and is especially suitable for emphasis scene
Monitoring application, can be used for:
Access control:White list personage enters, automatic opening entrance.
Sensitive objects are recognized:Dangerous person (such as terrorist or criminal) enters monitor area, auto-alarm-signal keying device.
Region control:Personage's entrance does not obtain authority region, automatic alarm.
Logout:Automatic capture enters the facial image of some region of whole personages, is registered and is recorded.
Market analysis:The consumption track of crowd, analyzes its behavioral characteristic in record market.
Sum it up, remote human face identification technology suffers from wide application prospect in terms of business, security protection and national defence.But no
It is same as in closely recognition of face, camera can easily capture high-resolution facial image, remote human face identification needs
Monitoring extensive area, from large-range monitoring image, is only capable of obtaining the facial image of low resolution, but identification face biological characteristic
Need high-resolution facial image, it is impossible to meet the demand of follow-up recognition of face and authentication.If to meet face knowledge
The whole monitor area of other high-resolution requirement, then equipment cost cannot bear, and there is also the substantial amounts of wasting of resources.It is existing
Remote human face identification technology remote human face identification scene in, mobile pedestrian and less pedestrian image does not ensure that but
Output clear face image, can so cause certain influence to the accuracy for recognizing, how to obtain and can be used for reliable recognition
Clear face image is to realize the key issue of remote human face identification.Another aspect remote human face identifying system is needed to entering prison
People's quick response in control region, monitor area occurs multiple targets simultaneously, and target may be moved quickly, even across monitoring
Region, how quickly to position and catch human face region is also the problem that existing remote human face identification technology needs to solve.
The content of the invention
It is an object of the invention to provide a kind of remote human face recognition methods of combination heartbeat signal, device and it is
System, it is intended to solve in existing remote human face identification technology in remote human face identification scene it cannot be guaranteed that output clear face image
Problem, and can not quickly position and catch the problem of human face region.
The object of the present invention is achieved like this, a kind of remote human face recognition methods of combination heartbeat signal, for multiple spot
The remote human face identification of active vision system, it is characterised in that methods described comprises the following steps:
Image in acquisition monitoring region, pedestrian image is extracted from the image for collecting and pedestrian locus is determined;
According to pedestrian locus, the locus of pedestrian's heartbeat signal is obtained using light vibration imaging technique;
According to heartbeat signal and the natural place relation of face, the face locus of pedestrian is positioned;
According to face locus, adjustment pickup area to human face region gathers clear face image.
Preferably, it is described that pedestrian image is extracted from the image for collecting and pedestrian locus is determined, including following step
Suddenly:
Segment in sliding window is extracted using multi-scale sliding window mouthful method to collecting image, the direction ladder of segment in parsing sliding window
Degree histogram feature, pedestrian detection window is obtained using SVMs analysis directions histogram of gradients feature, extracts pedestrian detection
Pedestrian image in window, and obtain the locus of pedestrian.
Preferably, the use light vibration imaging technique obtains the locus of pedestrian's heartbeat signal, specifically includes following
Step:
According to pedestrian locus, using the structure light irradiation pedestrian after modulation, structure light is after reflection by position sensing
Detector array is received, and docks number vibration information for carrying out Fourier transformation output pedestrian of collecting mail, from the vibration information for obtaining
In extract the locus of heartbeat signal.
Preferably, the natural place relation according to heartbeat signal and face, positions the face locus of pedestrian, tool
Body is comprised the following steps:
The proportionate relationship of pedestrian and pedestrian image is obtained according to pedestrian image, proportionally relation and heartbeat signal and face
Natural place relation, the distance required for moving to face from heartbeat signal in image is calculated, so that with reference to heartbeat signal
Locus determine face locus.
Preferably, described according to face locus, adjustment pickup area to human face region gathers clear face image,
Specifically include following steps:
According to face locus, the face locus after 0.2~0.5s of Kalman prediction pedestrian, adjustment
Man face image acquiring area obtains the clear face image of pedestrian to after predicting face locus, and stores.
Preferably, also include with reference to the remote human face recognition methods of heartbeat signal:
According to facial image is collected, facial image is shear off, transmitted to recognition of face device, completed to pedestrian's face
Identification and authentication.
Another object of the present invention is to provide a kind of remote human face identifying device of combination heartbeat signal, for multiple spot master
The remote human face identification of dynamic vision system, it is characterised in that described device includes:
Pedestrian position collecting unit, the image in acquisition monitoring region, from the image for collecting extract pedestrian image and
Determine pedestrian locus;
Heartbeat signal perceives unit, according to pedestrian locus, pedestrian's heartbeat signal is obtained using light vibration imaging technique
Locus;
Face coordinate setting unit, according to heartbeat signal and the natural place relation of face, positions the face space of pedestrian
Position;
Man face image acquiring unit, according to face locus, adjustment pickup area to human face region gathers clear face
Image.
Preferably, the pedestrian position collecting unit includes more than one wide field's camera and pedestrian image treatment mould
Block, wide field's camera is used to detect and follow the trail of target pedestrian, and gathers pedestrian image, the pedestrian image processing module
Segment in sliding window is extracted using multi-scale sliding window mouthful method to collecting image, the histograms of oriented gradients of segment in parsing sliding window
Feature, pedestrian detection window is obtained using SVMs analysis directions histogram of gradients feature, extracts the row in pedestrian detection window
People's image, and obtain the locus of pedestrian.
Preferably, the heartbeat signal perceives unit and includes:Light vibration image-forming module, for perceiving the small of heartbeat generation
Displacement, obtains the locus coordinate of target line people's heartbeat signal, and the light vibration image-forming module includes:Light source, position sensing
Detector array, signal processing circuit and microcontroller, the light source are used to produce the structure light irradiation pedestrian after modulation, described
Position-Sensitive Detector array is used to receive the echo-signal of structure light irradiation pedestrian, and the signal processing circuit is used for position
The signal that sensitive detector array is received is filtered noise reduction process, and the microcontroller is used for the docking collection of letters number to be carried out in Fu
Leaf transformation exports the vibration information of pedestrian, and the locus of heartbeat signal is extracted from the vibration information for obtaining.
Preferably, the face coordinate setting unit includes Face detection module, and the Face detection module is according to pedestrian
Image obtains the proportionate relationship of pedestrian and pedestrian image, proportionally the natural place relation of relation and heartbeat signal and face,
The distance required for moving to face from heartbeat signal in image is calculated, so that the locus with reference to heartbeat signal determines people
Face locus.
Preferably, the man face image acquiring unit include face prediction module, more than one narrow visual field camera and
The head matched with the narrow visual field camera, the face location prediction module is used to pass through Kalman prediction pedestrian
Face locus after 0.2~0.5s, the narrow visual field camera is arranged on the head of pairing, according to prediction face
Locus, by narrow visual field camera shooting angle and direction described in the cradle head control, alignment focuses on human face region, collection
Clear face image.
Preferably, the remote human face identifying device of the combination heartbeat signal also includes that facial image cuts module and face
Identifier, the facial image cuts module to be used to be sheared and be sent to the recognition of face to collecting facial image
Device, the recognition of face device is used to recognize face characteristic, completion pedestrian using convolutional neural networks to the facial image sheared
Recognition of face and authentication.
Dress is recognized another object of the present invention is to provide a kind of remote human face comprising the above-mentioned combination heartbeat signal of right
The system put.
From above-mentioned technical proposal as can be seen that obtaining the space of heartbeat signal using light vibration imaging technique in the present invention
Position, according to face locus and the natural place relation of heartbeat signal coordinate, the face space needed for effective acquisition system
Position, quick to guide head to adjust camera to face in-scope, calculating process is extremely simple, takes less, is almost not take up
Hardware resource, significantly saves and calculates consumption, improves systematic function, and solving in the prior art can not fast positioning and seizure soon
The problem of human face region.
Simultaneously the present invention in using by the way of multi-cam has complementary functions as remote human face identification scene in can not protect
The solution of card output clear face image problem, wide field's camera of low resolution is responsible for monitoring whole region, is carried out
The pedestrian detection in region, high-resolution narrow visual field camera is responsible for obtaining high definition facial image, and the face of pedestrian can be entered
The IMAQ of row high definition, solve the problems, such as in the prior art it cannot be guaranteed that output clear face image, so as to improve
Face identification rate.
Present invention utilizes the locus of heartbeat signal, the quick locus for calculating face manipulates head motion, makes
Obtaining narrow visual field camera can rapidly reach human face region, capture high-resolution human face image, improve remote human face identifying system
Discrimination and practicality.
Brief description of the drawings
Fig. 1 is the flow chart that the embodiment of the present invention provides the remote human face recognition methods for combining heartbeat signal;
Fig. 2 is the composition structure chart of the remote human face identifying device of combination heartbeat signal provided in an embodiment of the present invention;
Fig. 3 long-range heartbeat signal Cleaning Principle schematic diagrames provided in an embodiment of the present invention;
The dual camera pictorial diagram of Fig. 4 embodiment of the present invention embodiment offer devices.
Specific embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Embodiment one:
Referring to Fig. 1, Fig. 1 shows that the embodiment of the present invention provides the stream of the remote human face recognition methods for combining heartbeat signal
Journey, details are as follows for the method process:
In step S101, the image in acquisition monitoring region extracts pedestrian image and determination from the image for collecting
Pedestrian locus, its specific step includes in embodiments of the present invention:To collecting image using multi-scale sliding window mouthful
Method extracts segment in sliding window, the histograms of oriented gradients feature of segment in parsing sliding window, using SVMs analysis directions ladder
Degree histogram feature obtains pedestrian detection window, extracts the pedestrian image in pedestrian detection window, and obtain the locus of pedestrian;
In step s 102, according to pedestrian locus, the sky of pedestrian's heartbeat signal is obtained using light vibration imaging technique
Between position, its specific step includes in embodiments of the present invention:According to pedestrian locus, shone using the structure light after modulation
Pedestrian is penetrated, structure light is docked the collection of letters and number carry out Fourier transformation output after reflection by Position-Sensitive Detector array received
The vibration information of pedestrian, extracts the locus of heartbeat signal from the vibration information for obtaining;
In step s 103, according to heartbeat signal and the natural place relation of face, the face locus of pedestrian is positioned,
Its specific step includes in embodiments of the present invention:The proportionate relationship of pedestrian and pedestrian image is obtained according to pedestrian image, is pressed
According to the natural place relation of proportionate relationship and heartbeat signal and face, calculating needed for move to face from heartbeat signal in image
The distance wanted, so that the locus with reference to heartbeat signal determines face locus;
In step S104, according to face locus, adjustment pickup area to human face region gathers clear face figure
Picture, its specific step includes in embodiments of the present invention:According to face locus, by Kalman prediction pedestrian
Face locus after 0.2~0.5s, adjustment man face image acquiring area obtains the clear of pedestrian to after predicting face locus
Clear facial image, and store.
Used as another embodiment of the present invention, the remote human face recognition methods of the combination heartbeat signal also includes following step
Suddenly:
According to facial image is collected, facial image is shear off, transmitted to recognition of face device, completed to pedestrian's face
Identification and authentication.
Embodiment two:
Referring to Fig. 2, Fig. 2 shows the composition knot of the remote human face identifying device of the combination heartbeat signal that embodiment is provided
Structure, illustrate only the part related to the embodiment of the present invention.
This can be operate in software unit in each application system, hard with reference to the remote human face identifying device of heartbeat signal
Part unit or software and hardware unit.
The remote human face identifying device for combining heartbeat signal is perceived including pedestrian position collecting unit 201, heartbeat signal
Unit 202, face coordinate setting unit 203 and man face image acquiring unit 204.Wherein, the concrete function of each unit is as follows:
Pedestrian position collecting unit 201, the image in acquisition monitoring region, extracts pedestrian image from the image for collecting
With determine pedestrian locus, its specific device includes in embodiments of the present invention:Wide field's camera 2011 and pedestrian scheme
As processing module 2012, wide field's camera 2011 is used to detect and follow the trail of target pedestrian, and gathers pedestrian image, described
Pedestrian image processing module 2012 pairs collects image and extracts segment in sliding window using multi-scale sliding window mouthful method, in parsing sliding window
The histograms of oriented gradients feature of segment, pedestrian detection window is obtained using SVMs analysis directions histogram of gradients feature,
The pedestrian image in pedestrian detection window is extracted, and obtains the locus of pedestrian;
Heartbeat signal perceives unit 202, according to pedestrian locus, obtains pedestrian's heartbeat using light vibration imaging technique and believes
Number locus, its specific device includes in embodiments of the present invention:Light vibration image-forming module 2021, for perceiving heartbeat
The micro-displacement of generation, obtains the locus coordinate of target line people's heartbeat signal, referring to light vibration image-forming module described in Fig. 3
2021 include:Light source 301, Position-Sensitive Detector array 302, signal processing circuit 303 and microcontroller 304, referring to Fig. 3,
The light source 301 is used to produce the structure light irradiation pedestrian after modulation, and the Position-Sensitive Detector array 302 is used to receive to be tied
The echo-signal of structure light irradiation pedestrian, the signal processing circuit 303 is used for what location sensitive detector array 302 was received
Signal is filtered noise reduction process, and the microcontroller 304 is used for the docking collection of letters number carries out shaking for Fourier transformation output pedestrian
Dynamic information, and the locus of heartbeat signal is extracted from the vibration information for obtaining;
Face coordinate setting unit 203, according to heartbeat signal and the natural place relation of face, the face for positioning pedestrian is empty
Between position, its specific device includes in embodiments of the present invention:Face detection module 2031, the Face detection module 2031
The proportionate relationship of pedestrian and pedestrian image is obtained according to pedestrian image, proportionally the natural position of relation and heartbeat signal and face
Relation is put, the distance required for moving to face from heartbeat signal in image is calculated, so that with reference to the space bit of heartbeat signal
Put determination face locus.
Man face image acquiring unit 204, according to face locus, adjustment pickup area to human face region, collection is clear
Facial image, its specific device includes in embodiments of the present invention:Face location prediction module 2041, narrow visual field camera
2042 and with the narrow visual field camera match head 2043, the face location prediction module 2041 be used for pass through Kalman
Face locus after 0.2~0.5s of filter forecasting pedestrian, the narrow visual field camera 2042 is arranged on the cloud of pairing
On platform 2043, according to prediction face locus, the shooting angle of narrow visual field camera 2042 is controlled by the head 2043
Degree and direction, alignment focus on human face region, gather clear face image.
Further, the remote human face identifying device of the combination heartbeat signal also includes that facial image cuts module 205
With recognition of face device 206, the facial image cuts module 205 to be used to be sheared and be sent to institute to collecting facial image
Recognition of face device 206 is stated, the recognition of face device 206 is used to recognize people using convolutional neural networks to the facial image sheared
Face feature, completes pedestrian's recognition of face and authentication.
In the present embodiment, the recognition of face device 206 is including but not limited to using identification software of increasing income
SeetaFaceEngine completes recognition of face, and it uses one 9 layers of convolutional neural networks (CNN) to recognize face characteristic,
97.1% precision is reached on internal authority recognition of face open test collection (Labeled Faces in the Wild, LFW).
In embodiments of the present invention, present apparatus operation principle is:First using wide field's camera to monitor area pedestrian
Detected and followed the trail of, then by light vibration imaging technique detecting pedestrian heartbeat signal, and obtained the space of its heartbeat signal
Position, according to heartbeat signal and the natural place relation of face, by simple computation, obtains the locus of pedestrian's face, according to
This regulation narrow visual field camera, makes its quick arrival human face region, clear face image of the collection with enough resolution ratio.Need
What is illustrated is the distance relation that heartbeat signal refers to heart to vertical direction human face center with the natural place relation of face, body
The human body face center of 170cm high is normally at 30~40cm of heart vertical direction or so.
The remote human face identifying device of the combination heartbeat signal that the present embodiment is provided can be used in foregoing corresponding combination
The remote human face recognition methods of heartbeat signal, details referring to above-mentioned combination heartbeat signal remote human face recognition methods embodiment one
Associated description, no longer Ao Shu herein.
Used as another preferred embodiment of the invention, Fig. 4 illustrates the dual camera pictorial diagram of the present embodiment device, configuration
1 wide field's camera and 1 narrow visual field camera.Using dual camera as solution, the wide field of low resolution
Camera is responsible for monitoring whole region, carries out the pedestrian detection in region.High-resolution narrow visual field camera is responsible for obtaining high definition
Facial image.After device obtains the face locus of current pedestrian, head regulation is rotated, and narrow visual field camera is adjusted rapidly
To face in-scope.The present invention can also be preferably configured multiple wide or narrow visual field cameras according to scene demand.
It will be appreciated by those skilled in the art that the unit included by said apparatus is simply logically divided
, but above-mentioned division is not limited to, as long as corresponding function can be realized;In addition, the specific name of each functional unit
Title also simply facilitates mutually differentiation, the protection domain being not intended to limit the invention.
The above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvement, replacement and change can also be made.These improve, replace and change and are not departing from the right by attaching
It is required that and its on the premise of the scope of equivalent technologies, also should be regarded as protection scope of the present invention.
Claims (13)
1. a kind of remote human face recognition methods of combination heartbeat signal, for the remote human face identification of multiple spot active vision system,
Characterized in that, methods described comprises the following steps:
Image in acquisition monitoring region, pedestrian image is extracted from the image for collecting and pedestrian locus is determined;
According to pedestrian locus, the locus of pedestrian's heartbeat signal is obtained using light vibration imaging technique;
According to heartbeat signal and the natural place relation of face, the face locus of pedestrian is positioned;
According to face locus, adjustment pickup area to human face region gathers clear face image.
2. the as claimed in claim 1 remote human face recognition methods for combining heartbeat signal, it is characterised in that described from collecting
Image in extract pedestrian image and determine pedestrian locus, comprise the following steps:
Segment in sliding window is extracted using multi-scale sliding window mouthful method to collecting image, the direction gradient of segment is straight in parsing sliding window
Square figure feature, pedestrian detection window is obtained using SVMs analysis directions histogram of gradients feature, is extracted in pedestrian detection window
Pedestrian image, and obtain the locus of pedestrian.
3. the remote human face recognition methods for combining heartbeat signal as claimed in claim 1, it is characterised in that the use light shakes
Dynamic imaging technique obtains the locus of pedestrian's heartbeat signal, specifically includes following steps:
According to pedestrian locus, using the structure light irradiation pedestrian after modulation, structure light is detected by position sensing after reflection
Device array received, and number vibration information for carrying out Fourier transformation output pedestrian of collecting mail is docked, carried from the vibration information for obtaining
Take out the locus of heartbeat signal.
4. the remote human face recognition methods of the combination heartbeat signal described in claim 1, it is characterised in that described to be believed according to heartbeat
Natural place relation number with face, positions the face locus of pedestrian, specifically includes following steps:
The proportionate relationship of pedestrian and pedestrian image is obtained according to pedestrian image, proportionally the day of relation and heartbeat signal and face
Right position relationship, calculates the distance required for moving to face from heartbeat signal in image, so that with reference to the sky of heartbeat signal
Between position determine face locus.
5. the remote human face recognition methods of the combination heartbeat signal described in claim 1, it is characterised in that described empty according to face
Between position, to human face region, collection clear face image specifically includes following steps to adjustment pickup area:
According to face locus, the face locus after 0.2~0.5s of Kalman prediction pedestrian adjusts face
Image acquisition areas obtain the clear face image of pedestrian to after predicting face locus, and store.
6. the remote human face recognition methods of the combination heartbeat signal of any one of Claims 1 to 5, it is characterised in that methods described
Also include:
According to facial image is collected, facial image is shear off, transmitted to recognition of face device, completed to pedestrian's recognition of face
And authentication.
7. a kind of remote human face identifying device of combination heartbeat signal, for the remote human face identification of multiple spot active vision system,
Characterized in that, described device includes:
Pedestrian position collecting unit, the image in acquisition monitoring region extracts pedestrian image and determination from the image for collecting
Pedestrian locus;
Heartbeat signal perceives unit, according to pedestrian locus, the sky of pedestrian's heartbeat signal is obtained using light vibration imaging technique
Between position;
Face coordinate setting unit, according to heartbeat signal and the natural place relation of face, positions the face locus of pedestrian;
Man face image acquiring unit, according to face locus, adjustment pickup area to human face region gathers clear face figure
Picture.
8. the remote human face identifying device of heartbeat signal is combined as claimed in claim 7, it is characterised in that the pedestrian position
Collecting unit include more than one wide field's camera and pedestrian image processing module, wide field's camera be used for detect and
Target pedestrian is followed the trail of, and gathers pedestrian image, the pedestrian image processing module uses multi-scale sliding window to collecting image
Mouth method extracts segment in sliding window, the histograms of oriented gradients feature of segment in parsing sliding window, using SVMs analysis directions
Histogram of gradients feature obtains pedestrian detection window, extracts the pedestrian image in pedestrian detection window, and obtain the locus of pedestrian.
9. the remote human face identifying device of heartbeat signal is combined as claimed in claim 7, it is characterised in that the heartbeat signal
Perceiving unit includes:Light vibration image-forming module, the micro-displacement for perceiving heartbeat generation, obtains target line people's heartbeat signal
Locus coordinate, the light vibration image-forming module includes:Light source, Position-Sensitive Detector array, signal processing circuit and micro-
Controller, the light source is used to produce the structure light irradiation pedestrian after modulation, and the Position-Sensitive Detector array is used to receive
The echo-signal of structure light irradiation pedestrian, the signal processing circuit is used for the signal that location sensitive detector array is received
Noise reduction process is filtered, the microcontroller is used to dock number vibration information for carrying out Fourier transformation output pedestrian of collecting mail,
And the locus of heartbeat signal is extracted from the vibration information for obtaining.
10. the remote human face identifying device of heartbeat signal is combined as claimed in claim 7, it is characterised in that the face is sat
Demarcating bit location includes Face detection module, and the Face detection module obtains the ratio of pedestrian and pedestrian image according to pedestrian image
Example relation, the proportionally natural place relation of relation and heartbeat signal and face is calculated in image from heartbeat signal movement
Distance to required for face, so that the locus with reference to heartbeat signal determines face locus.
The 11. remote human face identifying devices for combining heartbeat signal as claimed in claim 7, it is characterised in that the face figure
As collecting unit includes face prediction module, more than one narrow visual field camera and the cloud matched with the narrow visual field camera
Platform, the face location prediction module is used for by the face locus after 0.2~0.5s of Kalman prediction pedestrian, institute
State narrow visual field camera to be arranged on the head of pairing, according to prediction face locus, by the cradle head control institute
Narrow visual field camera shooting angle and direction are stated, alignment focuses on human face region, gathers clear face image.
The 12. remote human face identifying devices for combining heartbeat signal as claimed in claim 7, it is characterised in that device also includes
Facial image cuts module and recognition of face device, and the facial image cuts module to be used to be sheared to collecting facial image
And the recognition of face device is sent to, the recognition of face device is used to know the facial image sheared using convolutional neural networks
Other face characteristic, completes pedestrian's recognition of face and authentication.
A kind of 13. systems of the remote human face identifying device of the combination heartbeat signal comprising any one of claim 7~12.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710069368.7A CN106897678B (en) | 2017-02-08 | 2017-02-08 | A kind of remote human face recognition methods of combination heartbeat signal, device and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710069368.7A CN106897678B (en) | 2017-02-08 | 2017-02-08 | A kind of remote human face recognition methods of combination heartbeat signal, device and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106897678A true CN106897678A (en) | 2017-06-27 |
CN106897678B CN106897678B (en) | 2019-11-12 |
Family
ID=59199059
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710069368.7A Active CN106897678B (en) | 2017-02-08 | 2017-02-08 | A kind of remote human face recognition methods of combination heartbeat signal, device and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106897678B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107633244A (en) * | 2017-11-13 | 2018-01-26 | 姜茂清 | Human body direct picture harvester, the image capturing system comprising it and its application |
CN109448026A (en) * | 2018-11-16 | 2019-03-08 | 南京甄视智能科技有限公司 | Passenger flow statistical method and system based on head and shoulder detection |
CN109754602A (en) * | 2019-01-15 | 2019-05-14 | 珠海格力电器股份有限公司 | The method and apparatus of the anti-erroneous judgement of pedestrian running red light |
CN110321756A (en) * | 2018-03-28 | 2019-10-11 | 北京语智科技有限公司 | In vivo detection system and method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1877599A (en) * | 2006-06-29 | 2006-12-13 | 南京大学 | Face setting method based on structured light |
CN101216885A (en) * | 2008-01-04 | 2008-07-09 | 中山大学 | Passerby face detection and tracing algorithm based on video |
CN103761514A (en) * | 2014-01-26 | 2014-04-30 | 公安部第三研究所 | System and method for achieving face recognition based on wide-angle gun camera and multiple dome cameras |
US20150148687A1 (en) * | 2013-11-22 | 2015-05-28 | Samsung Electronics Co., Ltd. | Method and apparatus for measuring heart rate |
-
2017
- 2017-02-08 CN CN201710069368.7A patent/CN106897678B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1877599A (en) * | 2006-06-29 | 2006-12-13 | 南京大学 | Face setting method based on structured light |
CN101216885A (en) * | 2008-01-04 | 2008-07-09 | 中山大学 | Passerby face detection and tracing algorithm based on video |
US20150148687A1 (en) * | 2013-11-22 | 2015-05-28 | Samsung Electronics Co., Ltd. | Method and apparatus for measuring heart rate |
CN103761514A (en) * | 2014-01-26 | 2014-04-30 | 公安部第三研究所 | System and method for achieving face recognition based on wide-angle gun camera and multiple dome cameras |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107633244A (en) * | 2017-11-13 | 2018-01-26 | 姜茂清 | Human body direct picture harvester, the image capturing system comprising it and its application |
CN110321756A (en) * | 2018-03-28 | 2019-10-11 | 北京语智科技有限公司 | In vivo detection system and method |
CN109448026A (en) * | 2018-11-16 | 2019-03-08 | 南京甄视智能科技有限公司 | Passenger flow statistical method and system based on head and shoulder detection |
CN109754602A (en) * | 2019-01-15 | 2019-05-14 | 珠海格力电器股份有限公司 | The method and apparatus of the anti-erroneous judgement of pedestrian running red light |
Also Published As
Publication number | Publication date |
---|---|
CN106897678B (en) | 2019-11-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10672140B2 (en) | Video monitoring method and video monitoring system | |
US9875392B2 (en) | System and method for face capture and matching | |
US10303172B2 (en) | Unmanned aerial vehicle having automatic tracking function and method of controlling the same | |
CN103106393B (en) | A kind of embedded human face identification intelligent identity authorization system based on robot platform | |
US7574021B2 (en) | Iris recognition for a secure facility | |
CN102473238B (en) | Method and system for image analysis | |
CN106897678A (en) | A kind of remote human face recognition methods of combination heartbeat signal, device and system | |
US20160321498A1 (en) | Warning a user about adverse behaviors of others within an environment based on a 3d captured image stream | |
CN108073577A (en) | A kind of alarm method and system based on recognition of face | |
CN105279496B (en) | A kind of method and apparatus of recognition of face | |
CN104378582A (en) | Intelligent video analysis system and method based on PTZ video camera cruising | |
KR102544391B1 (en) | A method of automatically detecting event of object using wearable device and a management server operating the same | |
Bashir et al. | Eagle-eyes: A system for iris recognition at a distance | |
Mian | Realtime face detection and tracking using a single pan, tilt, zoom camera | |
Yoon et al. | Tracking System for mobile user Based on CCTV | |
CN108596057A (en) | A kind of Information Security Management System based on recognition of face | |
Agarwal et al. | Impact of super-resolution and human identification in drone surveillance | |
CN110026982A (en) | Robot servo system | |
CN109344776A (en) | Data processing method | |
Bashir et al. | Video surveillance for biometrics: long-range multi-biometric system | |
Li et al. | A non-cooperative long-range biometric system for maritime surveillance | |
CN110991316B (en) | Method for automatically acquiring shape and identity information applied to open environment | |
CN111277745A (en) | Target person tracking method and device, electronic equipment and readable storage medium | |
KR20090053295A (en) | Method and system for monitoring video | |
Wei et al. | Biometrics: applications, challenges and the future |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |