CN108268839A - A kind of live body verification method and its system - Google Patents
A kind of live body verification method and its system Download PDFInfo
- Publication number
- CN108268839A CN108268839A CN201810009645.XA CN201810009645A CN108268839A CN 108268839 A CN108268839 A CN 108268839A CN 201810009645 A CN201810009645 A CN 201810009645A CN 108268839 A CN108268839 A CN 108268839A
- Authority
- CN
- China
- Prior art keywords
- infrared
- image
- module
- picture
- imaged
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000012795 verification Methods 0.000 title claims abstract description 27
- 238000002329 infrared spectrum Methods 0.000 claims abstract description 79
- 238000001429 visible spectrum Methods 0.000 claims abstract description 74
- 230000001815 facial effect Effects 0.000 claims abstract description 52
- 238000006243 chemical reaction Methods 0.000 claims abstract description 10
- 238000003384 imaging method Methods 0.000 claims description 31
- 238000001228 spectrum Methods 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 9
- 238000013507 mapping Methods 0.000 claims description 9
- 230000004888 barrier function Effects 0.000 claims description 7
- 241000208340 Araliaceae Species 0.000 claims description 5
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims description 5
- 235000003140 Panax quinquefolius Nutrition 0.000 claims description 5
- 235000008434 ginseng Nutrition 0.000 claims description 5
- 238000000701 chemical imaging Methods 0.000 claims description 3
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 238000001727 in vivo Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000031068 symbiosis, encompassing mutualism through parasitism Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
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
- G06V40/40—Spoof detection, e.g. liveness detection
- G06V40/45—Detection of the body part being alive
-
- 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
-
- 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
- G06V40/172—Classification, e.g. identification
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
A kind of live body verification method disclosed by the invention, includes the following steps:(1) it is imaged in picture in visible spectrum and obtains the location of facial image data;(2) it calculates the facial image and is imaged the position data in picture in infrared spectrum;(3) the facial image GLCM characteristic parameters are calculated, SVM vector machines is used whether can to judge for real human face.The invention also discloses live bodies to verify system, including visible spectrum image-forming module, obtains visible spectrum image location data module, conversion position data module, infrared structure light projection module, infrared spectrum image-forming module, GLCM characteristic parameters module, SVM vector machines.It is an advantage of the current invention that no matter counterfeiter uses photo, 3D masks that can be identified, and cooperated on one's own initiative without authenticatee when carrying out malicious attack to the authentication system based on face, so as to shorten the verification time, verification accuracy is improved.
Description
Technical field
The invention belongs to technical field of face recognition, specially a kind of live body verification method and its system.
Background technology
The authentication system for being currently based on face has been used widely, in the authentication system based on face,
It is the comparison between the human face photo based on current shooting and pre-stored human face photo, to carry out authentication, with
The popularization of authentication system based on face has derived the method for some malicious attack face authentications, for example, imitative
Emit people by handheld terminal store by counterfeit people's photo or the video prerecorded, come it is counterfeit I carry out authentication,
However, this counterfeiting measures in the prior art can be easily discernable, but work as and will be placed in by the photo of counterfeiter
When before the camera in this authentication system compared based on human face photo, this identity compared based on human face photo is tested
Card system can pass through subscriber authentication.In other words, malicious user can use the photo by counterfeiter and be attacked to carry out malice
It hits (that is, photo attack), this authentication system compared based on human face photo cannot resist photo attack.For above-mentioned photograph
Piece is attacked, and the method that currently available technology uses is, in the improved authentication system based on face, by checking face
Whether fine movement is had to effectively cope with above-mentioned photo attack.Further, user can be required to carry out required movement, so as to
Enhance the attack tolerant of the authentication system based on face.However, this method of the prior art needs authenticatee to coordinate
Initiatively do the malicious attack that some behaviors just can effectively prevent the authentication system based on face.
Invention content
During to solve to use the authentication system malicious attack carried out by counterfeiter's photo based on face in the prior art,
It is difficult to be identified and when identifying this malicious attack, the technological deficiency that authenticatee is needed to cooperate on one's own initiative, the present invention provides
A kind of live body verification method and its system, the purpose of realization for counterfeiter no matter using photo, 3D masks etc. to being based on face
Authentication system carry out malicious attack when, can all be identified, and cooperated on one's own initiative without authenticatee, so as to shorten
Verification time improves verification accuracy.
To achieve these goals, the present invention provides following technical scheme:A kind of live body verification method provided by the invention,
Include the following steps:
(1) it is imaged in picture in visible spectrum and detects face, obtain the location of the facial image detected data;
(2) according to the facial image position data obtained in step (1), calculate the facial image infrared spectrum into
As the position data in picture;
(3) position data in picture is imaged in infrared spectrum according to the facial image that step (2) obtains, described
Detection face in infrared spectrum imaging picture position data area calculates the facial image GLCM features ginseng after detecting face
Number, is input to SVM vector machines by the facial image GLCM characteristic parameters being calculated and is compared, you can judge whether be true
Real face;
The SVM vector machines train to obtain through following method:1. prepare training sample set:The sample is included just
Sample and negative sample, the photo of real human face image taking is positive sample, and the picture of non-genuine facial image shooting is negative sample,
Sample training is carried out using the sorting algorithm study of SVM vector machines, obtains to distinguish the grader of positive sample and negative sample;
2. the positive sample 1. step is trained after is placed in a file, negative sample is placed in another file;And
All training samples are zoomed into same size;
3. extract the GLCM features of all positive samples;
4. extract the GLCM features of all negative samples;
5. respective label is assigned to the positive and negative samples respectively;
6. by the GLCM features of the positive negative sample, the label of the positive negative sample, it is all input to 1. point that step obtains
It is trained in class device;
7. after step 6. middle classifier training, the SVM vector machines are obtained.
Further, the step (1) includes being imaged mould to visible light image-forming module, infrared spectrum respectively first
Block carries out camera calibration, then is imaged in picture in visible spectrum and detects face;
The step (2) includes:1. calibrated visible light image-forming module, infrared spectrum image-forming module are placed in sky
Between same plane, while obtain the RGB image and infrared image of scaling board;
2. setting scaling board plane in the plane of world coordinate system Z=0, scaling board in visible spectrum imaging is acquired respectively
The infrared point set coordinate of scaling board characteristic point during RGB point sets coordinate, the infrared spectrum of characteristic point are imaged,
The RGB point sets coordinate is (Xrgb 1, Yrgb 1)、(Xrgb 2, Yrgb 2)…(Xrgb 3, Yrgb 3);
The infrared point set coordinate is (Xir 1, Yir 1)、(Xir 2, Yir 2)…(Xir 3, Yir 3);
3. according to the unitary linear mapping relation between the RGB point sets coordinate and infrared point set coordinate, obtain such as lower section
Journey group:
Formula one:
Formula two:
Using least square method, a is acquiredx、bx、ay、by,
4. a 3. obtained according to stepx、bx、ay、byAnd the formula one, formula two, it will be obtained from step (1)
Visible spectrum imaging picture described in the location of facial image data substitute into formula one, formula two, you can be somebody's turn to do
Facial image is imaged the position data in picture in infrared spectrum;
Infrared spectrum imaging picture is obtained using the infrared spectrum image-forming module that the camera calibration is crossed in the step (3)
Face, and detect face in infrared spectrum imaging picture position data area.
Further, the method that GLCM characteristic parameters are calculated in the step (3) includes:
1. extracting the facial image feature, according to the facial image character pixel extracted, calculating acquires the figure
The number of greyscale levels of picture, so as to obtain co-occurrence matrix, the co-occurrence matrix is square of number of greyscale levels being calculated;
2. the value in the co-occurrence matrix is converted to probability value, gray level co-occurrence matrixes are obtained;
3. the GLCM characteristic parameters include mean value Mean, variance Variance, contrast C ontrast, entropy Entropy,
Angular second moment ASM, correlation Correlation, the GLCM characteristic parameters are acquired respectively by following formula:
Formula three:
Formula four:
Formula five:
Formula six:
Formula seven:
Formula eight:
Wherein,Represent the gray probability value of all row coordinates of the gray level co-occurrence matrixes;
Represent the gray probability value of all row coordinates of the gray level co-occurrence matrixes;I represents the gray level co-occurrence matrixes row
Coordinate;J represents the gray level co-occurrence matrixes row coordinate;P (i, j) represents that certain a line coordinate in the gray level co-occurrence matrixes, row are sat
The gray probability value of certain point that mark determines.
Further, in the step (1) visible spectrum be imaged in picture detect after face first with it is pre-stored
Photo and/or the photo of reading compare, and if comparison result is inconsistent, directly terminate.
Further, when detecting face in infrared spectrum imaging picture position data area in the step (3),
Directly terminate if it can't detect face.
The invention also discloses the verification systems using the exploitation of above-mentioned verification method, including visible spectrum image-forming module, obtain
Take visible spectrum image location data module, conversion position data module, infrared structure light projection module, infrared spectrum imaging mould
Block, GLCM characteristic parameters module, SVM vector machines;
The visible spectrum image-forming module is imaged in picture in visible spectrum for shooting visible spectrum image and detects people
Face;
It is described to obtain visible spectrum image location data module for obtaining the location of the facial image detected number
According to;
The conversion position data module is used to be imaged the position data of facial image in picture according to visible spectrum, calculates
Go out the facial image and be imaged the position data in picture in infrared spectrum;
The infrared structure light projection module is used to emit the structure light of infrared spectrum;
The infrared spectrum image-forming module shoots infrared spectroscopic imaging for detecting face;
GLCM characteristic parameters module is used to calculate the GLCM features of facial image detected in infrared spectrum imaging picture
Parameter;
SVM vector machines are used for the GLCM characteristic parameters for calculating the GLCM characteristic parameters module and input, and input and sentence
Disconnected result.
Further, which further includes mapping block, and the visible spectrum image-forming module includes visible spectrum imaging mark
Cover half block, infrared spectrum image-forming module include infrared spectrum and are imaged demarcating module;
The visible spectrum imaging demarcating module is used to light spectrum image-forming module be carried out camera calibration;
The infrared spectrum imaging demarcating module is used to infrared spectrum image-forming module carrying out camera calibration;
The mapping block is used to obtain the RGB point set coordinates of scaling board characteristic point in visible spectrum image respectively,
The infrared point set coordinate of scaling board characteristic point, RGB point sets coordinate, infrared point further according to the scaling board in infrared spectrum imaging
Unary linear relation between collection coordinate is obtained the facial image detected in visible spectrum image-forming module and is imaged in infrared spectrum
Position data in picture.
Further, the visible spectrum image-forming module includes obtaining photo module in advance, the advance acquisition photo mould
Block is used for the first photo with pre-stored photo and/or reading after visible spectrum is imaged in picture and detects face and compares,
If comparison result is inconsistent, result is exported.
Further, the infrared structure light projection module includes RF transmitter, and the infrared ray includes infrared dissipate
Spot structure light, infrared stripes structure light.
Further, the visible spectrum image-forming module is visible spectrum cameras, it is seen that inside light spectrum image-forming video camera
It is equipped with outside the first cmos image sensor, camera lens and infrared barrier filter is installed.
Further, the infrared spectrum image-forming module is red-light spectrum video camera, inside the infrared spectrum video camera
Second cmos image sensor is installed, visible ray barrier filter is installed outside camera lens.
Further, which further includes microprocessor, and microprocessor is visible with visible spectrum image-forming module, acquisition respectively
Spectrum picture position data module, conversion position data module, infrared structure light projection module, infrared spectrum image-forming module,
GLCM characteristic parameters module, the connection of SVM vector machines.
Further, the visible spectrum image-forming module and the same central shaft of infrared spectrum image-forming module coexistence.
Further, the advance acquisition photo module includes photo memory, certificate reading device.
The present invention is using above-mentioned technical proposal, including following advantageous effect:No matter counterfeiter uses photo, 3D masks etc. pair
When authentication system based on face carries out malicious attack, can all it be identified, and cooperated on one's own initiative without authenticatee,
So as to shorten the verification time, verification accuracy is improved.
Description of the drawings
Fig. 1 is live body verification method flow chart of the present invention in embodiment one;
Fig. 2 verifies system block diagram for live body of the present invention in embodiment one;
Fig. 3 is live body verification method flow chart of the present invention in embodiment two;
Fig. 4 verifies system block diagram for live body of the present invention in embodiment two;
Fig. 5 is visible spectrum image-forming module and the signal of infrared spectrum image-forming module another kind example structure in embodiment two
Figure;
In figure, 1, visible spectrum cameras;2nd, red-light spectrum video camera;3rd, the first cmos image sensor;4th, infrared resistance
Every filter;5th, the second cmos image sensor;6th, visible ray barrier filter.
Specific embodiment
The present invention is described in further detail below by specific embodiment and with reference to attached drawing.
Embodiment one:A kind of live body verification method with reference to shown in Fig. 1, includes the following steps:
(1) it is imaged in picture in visible spectrum and detects face, obtain the location of the facial image detected data;Example
Such as coordinate data of the facial image in entire shooting picture,
(2) according to the facial image position data obtained in step (1), calculate the facial image infrared spectrum into
As the position data in picture;
(3) position data in picture is imaged in infrared spectrum according to the facial image that step (2) obtains, described
Detection face in infrared spectrum imaging picture position data area calculates the facial image GLCM features ginseng after detecting face
Number, is input to SVM vector machines by the facial image GLCM characteristic parameters being calculated and is compared, you can judge whether be true
Real face;
The SVM vector machines train to obtain through following method:1. prepare training sample set:The sample is included just
Sample and negative sample, the photo of real human face image taking is positive sample, and the picture of non-genuine facial image shooting is negative sample,
Sample training is carried out using the sorting algorithm study of SVM vector machines, obtains to distinguish the grader of positive sample and negative sample;Instruction
It should be infinite number of to practice sample, and training sample should cover the various situations that may occur in actual application.It is real
In the application process of border, training sample can not possibly be infinitely more, according to the complexity of In vivo detection, should choose 3,000 to 5,000 positive samples
This, 3,000 to 5,000 negative samples;
2. the positive sample 1. step is trained after is placed in a file, negative sample is placed in another file;And
All training samples are zoomed into same size;
3. extract the GLCM features of all positive samples;
4. extract the GLCM features of all negative samples;
5. respective label is assigned to the positive and negative samples respectively;
6. by the GLCM features of the positive negative sample, the label of the positive negative sample, it is all input to 1. point that step obtains
It is trained in class device;
7. after step 6. middle classifier training, the SVM vector machines are obtained.
Further, in order to remove influence of the other parameter for In vivo detection, the step (1) includes distinguishing first
Camera calibration is carried out, then be imaged in picture and examine in visible spectrum to visible light image-forming module, infrared spectrum image-forming module
Survey face;
The step (2) includes:1. calibrated visible light image-forming module, infrared spectrum image-forming module are placed in sky
Between same plane, while obtain the RGB image and infrared image of scaling board;
2. setting scaling board plane in the plane of world coordinate system Z=0, scaling board in visible spectrum imaging is acquired respectively
The infrared point set coordinate of scaling board characteristic point during RGB point sets coordinate, the infrared spectrum of characteristic point are imaged,
The RGB point sets coordinate is (Xrgb 1, Yrgb 1)、(Xrgb 2, Yrgb 2)…(Xrgb 3, Yrgb 3);
The infrared point set coordinate is (Xir 1, Yir 1)、(Xir 2, Yir 2)…(Xir 3, Yir 3);
3. according to the unitary linear mapping relation between the RGB point sets coordinate and infrared point set coordinate, obtain such as lower section
Journey group:
Formula one:
Formula two:
Using least square method, a is acquiredx、bx、ay、by,
4. a 3. obtained according to stepx、bx、ay、byAnd the formula one, formula two, it will be obtained from step (1)
Visible spectrum imaging picture described in the location of facial image data substitute into formula one, formula two, you can be somebody's turn to do
Facial image is imaged the position data in picture in infrared spectrum;
Infrared spectrum imaging picture is obtained using the infrared spectrum image-forming module that the camera calibration is crossed in the step (3)
Face, and detect face in infrared spectrum imaging picture position data area.
The method that GLCM characteristic parameters are calculated in the step (3) includes 1. extracting the facial image feature, according to carrying
The facial image character pixel taken out calculates the number of greyscale levels for acquiring the image, so as to obtain co-occurrence matrix, the symbiosis
Matrix is square of number of greyscale levels being calculated;
2. the value in the co-occurrence matrix is converted to probability value, gray level co-occurrence matrixes are obtained;
For example, obtained co-occurrence matrix is as shown in the table:
Grayvalue transition in above-mentioned co-occurrence matrix is probability value, and method is that own in each element value divided by matrix
Element and (in upper table element and for 18), it is as shown in the table to finally obtain gray level co-occurrence matrixes:
3. the GLCM characteristic parameters include mean value Mean, variance Variance, contrast C ontrast, entropy Entropy,
Angular second moment ASM, correlation Correlation, the GLCM characteristic parameters are acquired respectively by following formula:
Formula three:
Formula four:
Formula five:
Formula six:
Formula seven:
Formula eight:
Wherein,Represent the gray probability value of all row coordinates of the gray level co-occurrence matrixes;
Represent the gray probability value of all row coordinates of the gray level co-occurrence matrixes;I represents the gray level co-occurrence matrixes row
Coordinate;J represents the gray level co-occurrence matrixes row coordinate;P (i, j) represents that certain a line coordinate in the gray level co-occurrence matrixes, row are sat
The gray probability value of certain point that mark determines.
GLCM characteristic parameters 14 parameters in total, in In vivo detection is carried out, it is only necessary to can be realized using 6 parameters high-precision
The detection of degree, the present invention are repeated as many times parallel test, detect accurate rate of accuracy reached to more than 98%.
The verification system developed using above-mentioned verification method with reference to shown in Fig. 2, including visible spectrum image-forming module, is obtained
Take visible spectrum image location data module, conversion position data module, infrared structure light projection module, infrared spectrum imaging mould
Block, GLCM characteristic parameters module, SVM vector machines;
The visible spectrum image-forming module is imaged in picture in visible spectrum for shooting visible spectrum image and detects people
Face;
It is described to obtain visible spectrum image location data module for obtaining the location of the facial image detected number
According to;
The conversion position data module is used to be imaged the position data of facial image in picture according to visible spectrum, calculates
Go out the facial image and be imaged the position data in picture in infrared spectrum;
The infrared structure light projection module is used to emit the structure light of infrared spectrum;
The infrared spectrum image-forming module shoots infrared spectroscopic imaging for detecting face;
GLCM characteristic parameters module is used to calculate the GLCM features of facial image detected in infrared spectrum imaging picture
Parameter;
SVM vector machines are used for the GLCM characteristic parameters for calculating the GLCM characteristic parameters module and input, and export and sentence
Disconnected result.
The system further includes mapping block, and the visible spectrum image-forming module includes visible spectrum and is imaged demarcating module, red
External spectrum image-forming module includes infrared spectrum and is imaged demarcating module;
The visible spectrum imaging demarcating module is used to light spectrum image-forming module be carried out camera calibration;
The infrared spectrum imaging demarcating module is used to infrared spectrum image-forming module carrying out camera calibration;
The mapping block is used to obtain the RGB point set coordinates of scaling board characteristic point in visible spectrum image respectively,
The infrared point set coordinate of scaling board characteristic point, RGB point sets coordinate, infrared point further according to the scaling board in infrared spectrum imaging
Unary linear relation between collection coordinate is obtained the facial image detected in visible spectrum image-forming module and is imaged in infrared spectrum
Position data in picture.
Embodiment two:Other than one disclosure of embodiment, in order to further make detection method more quick,
With reference to shown in Fig. 3, in the step (1) visible spectrum be imaged in picture detect after face first with pre-stored photo
And/or the photo read compares, and if comparison result is inconsistent, directly terminates.
When face can also be detected in infrared spectrum imaging picture position data area in the step (3),
Directly terminate if it can't detect face.
By above-mentioned steps, then can not enter follow-up comparison process, direct output error as a result, save compare when
Between.
The system further includes microprocessor, microprocessor respectively with visible spectrum image-forming module, obtain visible spectrum image
Position data module, conversion position data module, infrared structure light projection module, infrared spectrum image-forming module, GLCM features ginseng
Digital-to-analogue block, the connection of SVM vector machines.The message obtained from modules is carried out integration processing by microprocessor, and classification will belong to each
The information of module is distributed transmission again.
Other than system disclosed in embodiment one, with reference to shown in Fig. 4, using described in the verification system of above method exploitation
Visible spectrum image-forming module includes obtaining photo module in advance, and the advance acquisition photo module is used to be imaged picture in visible spectrum
The first photo with pre-stored photo and/or reading compares after face is detected in face, if comparison result is inconsistent, exports
As a result.The advance acquisition photo module includes photo memory, certificate reading device.It can get user's by internet
Photo, can also be obtained by certificate reading device the photo on the certificates such as identity card, social security card of user to current authentication
User compares.
More specifically, the infrared structure light projection module used in the present invention includes RF transmitter, the infrared ray
Including infrared pattern light, infrared stripes structure light.In order to ensure visible spectrum image-forming module and infrared spectrum image-forming module
When obtaining verifier's facial image, the range of shooting is consistent with angle, and the visible spectrum image-forming module is imaged with infrared spectrum
The same central shaft of module coexistence, with reference to shown in Fig. 5, the visible spectrum image-forming module is visible spectrum cameras 1, it is seen that spectrum
It is equipped with inside imaging camera machine outside the first cmos image sensor 3, camera lens and infrared barrier filter 4 is installed;It is described
Infrared spectrum image-forming module is red-light spectrum video camera 2, and the second cmos image biography is equipped with inside the infrared spectrum video camera
Visible ray barrier filter 6 is installed outside sensor 5, camera lens.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, that is made any repaiies
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of live body verification method, which is characterized in that this method comprises the following steps:
(1) it is imaged in picture in visible spectrum and detects face, obtain the location of the facial image detected data;
(2) it according to the facial image position data obtained in step (1), calculates the facial image and is imaged picture in infrared spectrum
Position data in face;
(3) position data in picture is imaged in infrared spectrum according to the facial image that step (2) obtains, described infrared
Detection face, the facial image GLCM characteristic parameters are calculated after detecting face in light spectrum image-forming picture position data area,
The facial image GLCM characteristic parameters being calculated are input to SVM vector machines to be compared, you can judge whether be true
Face;
The SVM vector machines train to obtain through following method:1. prepare training sample set:The sample includes positive sample
And negative sample, the photo of real human face image taking is positive sample, and the picture of non-genuine facial image shooting is negative sample, is utilized
The sorting algorithm study of SVM vector machines carries out sample training, obtains to distinguish the grader of positive sample and negative sample;
2. the positive sample 1. step is trained after is placed in a file, negative sample is placed in another file;And by institute
There is training sample to zoom to same size;
3. extract the GLCM features of all positive samples;
4. extract the GLCM features of all negative samples;
5. respective label is assigned to the positive and negative samples respectively;
6. by the GLCM features of the positive negative sample, the label of the positive negative sample, it is all input to 1. grader that step obtains
In be trained;
7. after step 6. middle classifier training, the SVM vector machines are obtained.
2. live body verification method according to claim 1, which is characterized in that it is right respectively first that the step (1) includes
Visible light image-forming module, infrared spectrum image-forming module carry out camera calibration, then are imaged in picture and detect in visible spectrum
Face;
The step (2) includes:1. by calibrated visible light image-forming module, that infrared spectrum image-forming module is placed in space is same
One plane, while obtain the RGB image and infrared image of scaling board;
2. setting scaling board plane in the plane of world coordinate system Z=0, scaling board feature in visible spectrum imaging is acquired respectively
The infrared point set coordinate of scaling board characteristic point during RGB point sets coordinate, the infrared spectrum of point are imaged,
The RGB point sets coordinate is (Xrgb 1, Yrgb 1)、(Xrgb 2, Yrgb 2)…(Xrgb 3, Yrgb 3);
The infrared point set coordinate is (Xir 1, Yir 1)、(Xir 2, Yir 2)…(Xir 3, Yir 3);
3. according to the unitary linear mapping relation between the RGB point sets coordinate and infrared point set coordinate, equation below group is obtained:
Formula one:
Formula two:
Using least square method, a is acquiredx、bx、ay、by,
4. a 3. obtained according to stepx、bx、ay、byAnd the formula one, formula two, it will be obtained from step (1)
The location of facial image data described in light spectrum image-forming picture is substituted into formula one, formula two, you can obtains the face figure
As the position data being imaged in infrared spectrum in picture;
Infrared spectrum imaging picture is obtained using the infrared spectrum image-forming module that the camera calibration is crossed in the step (3),
And detect face in infrared spectrum imaging picture position data area.
3. live body verification method according to claim 1, which is characterized in that GLCM features ginseng is calculated in the step (3)
Several methods include:
1. extracting the facial image feature, according to the facial image character pixel extracted, calculating acquires the image
Number of greyscale levels, so as to obtain co-occurrence matrix, the co-occurrence matrix is square of number of greyscale levels being calculated;
2. the value in the co-occurrence matrix is converted to probability value, gray level co-occurrence matrixes are obtained;
3. the GLCM characteristic parameters include mean value Mean, variance Variance, contrast C ontrast, entropy Entropy, angle two
Rank square ASM, correlation Correlation, the GLCM characteristic parameters are acquired respectively by following formula:
Formula three:
Formula four:
Formula five:
Formula six:
Formula seven:
Formula eight:
Wherein,Represent the gray probability value of all row coordinates of the gray level co-occurrence matrixes;
Represent the gray probability value of all row coordinates of the gray level co-occurrence matrixes;I represents the gray level co-occurrence matrixes row coordinate;
J represents the gray level co-occurrence matrixes row coordinate;P (i, j) represents that certain a line coordinate, row coordinate determine in the gray level co-occurrence matrixes
Certain point gray probability value.
4. live body verification method according to claim 1, which is characterized in that be imaged in the step (1) in visible spectrum
The first photo with pre-stored photo and/or reading compares after face is detected in picture, straight if comparison result is inconsistent
Binding beam.
5. live body verification method according to claim 1, which is characterized in that in the infrared spectrum in the step (3)
When being imaged detection face in picture position data area, directly terminate if it can't detect face.
6. a kind of live body verifies system, which is characterized in that the system includes visible spectrum image-forming module, obtains visible spectrum image
Position data module, conversion position data module, infrared structure light projection module, infrared spectrum image-forming module, GLCM features ginseng
Digital-to-analogue block, SVM vector machines;
The visible spectrum image-forming module is imaged in picture in visible spectrum for shooting visible spectrum image and detects face;
It is described to obtain visible spectrum image location data module for obtaining the location of the facial image detected data;
The conversion position data module is used to be imaged the position data of facial image in picture according to visible spectrum, calculates institute
It states facial image and is imaged the position data in picture in infrared spectrum;
The infrared structure light projection module is used to emit the structure light of infrared spectrum;
The infrared spectrum image-forming module shoots infrared spectroscopic imaging for detecting face;
GLCM characteristic parameters module is used to calculate the GLCM characteristic parameters of facial image detected in infrared spectrum imaging picture;
SVM vector machines are used for the GLCM characteristic parameters for calculating the GLCM characteristic parameters module and input, and export judgement knot
Fruit.
7. live body according to claim 6 verifies system, which is characterized in that the system further includes mapping block, it is described can
See that light spectrum image-forming module includes visible spectrum and is imaged demarcating module, infrared spectrum image-forming module includes infrared spectrum and is imaged calibration mold
Block;
The visible spectrum imaging demarcating module is used to light spectrum image-forming module be carried out camera calibration;
The infrared spectrum imaging demarcating module is used to infrared spectrum image-forming module carrying out camera calibration;
The mapping block is infrared for obtaining the RGB point set coordinates of scaling board characteristic point in visible spectrum image respectively
The infrared point set coordinate of scaling board characteristic point in light spectrum image-forming, RGB point sets coordinate, infrared point set further according to the scaling board are sat
Unary linear relation between mark is obtained the facial image detected in visible spectrum image-forming module and is imaged picture in infrared spectrum
In position data.
8. live body according to claim 6 verifies system, which is characterized in that the visible spectrum image-forming module includes advance
Obtain photo module, the advance acquisition photo module be used for visible spectrum be imaged in picture detect after face first in advance
The photo of storage and/or the photo of reading compare, and if comparison result is inconsistent, export result.
9. live body verification system according to claim 6, which is characterized in that the infrared structure light projection module includes infrared
Line transmitter, the infrared ray include infrared pattern light, infrared stripes structure light.
10. live body verification system according to claim 6, which is characterized in that the visible spectrum image-forming module is visible ray
Compose video camera, it is seen that be equipped with inside light spectrum image-forming video camera outside the first cmos image sensor, camera lens be equipped with it is red
Outer barrier filter;
The infrared spectrum image-forming module is red-light spectrum video camera, and the 2nd CMOS is equipped with inside the infrared spectrum video camera
Visible ray barrier filter is installed outside imaging sensor, camera lens;
The system further includes microprocessor, microprocessor respectively with visible spectrum image-forming module, obtain visible spectrum picture position
Data module, conversion position data module, infrared structure light projection module, infrared spectrum image-forming module, GLCM characteristic parameter moulds
Block, the connection of SVM vector machines;
The visible spectrum image-forming module and the same central shaft of infrared spectrum image-forming module coexistence;
The advance acquisition photo module includes photo memory, certificate reading device.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810009645.XA CN108268839A (en) | 2018-01-05 | 2018-01-05 | A kind of live body verification method and its system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810009645.XA CN108268839A (en) | 2018-01-05 | 2018-01-05 | A kind of live body verification method and its system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108268839A true CN108268839A (en) | 2018-07-10 |
Family
ID=62773427
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810009645.XA Pending CN108268839A (en) | 2018-01-05 | 2018-01-05 | A kind of live body verification method and its system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108268839A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110490044A (en) * | 2019-06-14 | 2019-11-22 | 杭州海康威视数字技术股份有限公司 | Face modelling apparatus and human face model building |
CN110909617A (en) * | 2019-10-28 | 2020-03-24 | 广州多益网络股份有限公司 | Living body face detection method and device based on binocular vision |
WO2020073993A1 (en) * | 2018-10-12 | 2020-04-16 | 杭州海康威视数字技术股份有限公司 | Face anti-spoof detection method, device and multi-view camera |
CN111879724A (en) * | 2020-08-05 | 2020-11-03 | 中国工程物理研究院流体物理研究所 | Human skin mask identification method and system based on near infrared spectrum imaging |
US11443550B2 (en) * | 2020-06-05 | 2022-09-13 | Jilin Qs Spectrum Data Technology Co. Ltd | Face recognition monitoring system based on spectrum and multi-band fusion and recognition method using same |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101793562A (en) * | 2010-01-29 | 2010-08-04 | 中山大学 | Face detection and tracking algorithm of infrared thermal image sequence |
CN102542281A (en) * | 2010-12-27 | 2012-07-04 | 北京北科慧识科技股份有限公司 | Non-contact biometric feature identification method and system |
CN102622589A (en) * | 2012-03-13 | 2012-08-01 | 辉路科技(北京)有限公司 | Multispectral face detection method based on graphics processing unit (GPU) |
CN103106401A (en) * | 2013-02-06 | 2013-05-15 | 北京中科虹霸科技有限公司 | Mobile terminal iris recognition device with human-computer interaction mechanism and method |
WO2013131407A1 (en) * | 2012-03-08 | 2013-09-12 | 无锡中科奥森科技有限公司 | Double verification face anti-counterfeiting method and device |
CN103605958A (en) * | 2013-11-12 | 2014-02-26 | 北京工业大学 | Living body human face detection method based on gray scale symbiosis matrixes and wavelet analysis |
CN104978077A (en) * | 2014-04-08 | 2015-10-14 | 联想(北京)有限公司 | Interaction method and interaction system |
CN106530484A (en) * | 2016-11-17 | 2017-03-22 | 深圳怡化电脑股份有限公司 | Method and device for counterfeit detection of banknote |
CN107292285A (en) * | 2017-07-14 | 2017-10-24 | 广东欧珀移动通信有限公司 | Living iris detection method and Related product |
-
2018
- 2018-01-05 CN CN201810009645.XA patent/CN108268839A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101793562A (en) * | 2010-01-29 | 2010-08-04 | 中山大学 | Face detection and tracking algorithm of infrared thermal image sequence |
CN102542281A (en) * | 2010-12-27 | 2012-07-04 | 北京北科慧识科技股份有限公司 | Non-contact biometric feature identification method and system |
WO2013131407A1 (en) * | 2012-03-08 | 2013-09-12 | 无锡中科奥森科技有限公司 | Double verification face anti-counterfeiting method and device |
CN102622589A (en) * | 2012-03-13 | 2012-08-01 | 辉路科技(北京)有限公司 | Multispectral face detection method based on graphics processing unit (GPU) |
CN103106401A (en) * | 2013-02-06 | 2013-05-15 | 北京中科虹霸科技有限公司 | Mobile terminal iris recognition device with human-computer interaction mechanism and method |
CN103605958A (en) * | 2013-11-12 | 2014-02-26 | 北京工业大学 | Living body human face detection method based on gray scale symbiosis matrixes and wavelet analysis |
CN104978077A (en) * | 2014-04-08 | 2015-10-14 | 联想(北京)有限公司 | Interaction method and interaction system |
CN106530484A (en) * | 2016-11-17 | 2017-03-22 | 深圳怡化电脑股份有限公司 | Method and device for counterfeit detection of banknote |
CN107292285A (en) * | 2017-07-14 | 2017-10-24 | 广东欧珀移动通信有限公司 | Living iris detection method and Related product |
Non-Patent Citations (2)
Title |
---|
BENJAMI´N HERNA´NDEZ ET AL.: "Visual learning of texture descriptors for facial expression recognition in thermal imagery", COMPUTER VISION AND IMAGE UNDERSTANDING, 4 January 2007 (2007-01-04), pages 258 - 269 * |
丁明跃 等著: "《地质勘查图像分析与综合》", 北京:中国铁道出版社, pages: 107 - 109 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020073993A1 (en) * | 2018-10-12 | 2020-04-16 | 杭州海康威视数字技术股份有限公司 | Face anti-spoof detection method, device and multi-view camera |
US11869255B2 (en) | 2018-10-12 | 2024-01-09 | Hangzhou Hikvision Digital Technology Co., Ltd. | Anti-counterfeiting face detection method, device and multi-lens camera |
CN110490044A (en) * | 2019-06-14 | 2019-11-22 | 杭州海康威视数字技术股份有限公司 | Face modelling apparatus and human face model building |
CN110909617A (en) * | 2019-10-28 | 2020-03-24 | 广州多益网络股份有限公司 | Living body face detection method and device based on binocular vision |
CN110909617B (en) * | 2019-10-28 | 2022-03-25 | 广州多益网络股份有限公司 | Living body face detection method and device based on binocular vision |
US11443550B2 (en) * | 2020-06-05 | 2022-09-13 | Jilin Qs Spectrum Data Technology Co. Ltd | Face recognition monitoring system based on spectrum and multi-band fusion and recognition method using same |
CN111879724A (en) * | 2020-08-05 | 2020-11-03 | 中国工程物理研究院流体物理研究所 | Human skin mask identification method and system based on near infrared spectrum imaging |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108268839A (en) | A kind of live body verification method and its system | |
CN103530599B (en) | The detection method and system of a kind of real human face and picture face | |
US10650261B2 (en) | System and method for identifying re-photographed images | |
CN107451510B (en) | Living body detection method and living body detection system | |
CN105740780B (en) | Method and device for detecting living human face | |
CN106529414A (en) | Method for realizing result authentication through image comparison | |
RU2431190C2 (en) | Facial prominence recognition method and device | |
KR20170006355A (en) | Method of motion vector and feature vector based fake face detection and apparatus for the same | |
JP3855025B2 (en) | Personal authentication device | |
CN106886216A (en) | Robot automatic tracking method and system based on RGBD Face datections | |
CN107240112B (en) | Individual X corner extraction method in complex scene | |
CN108734235A (en) | A kind of personal identification method and system for electronic prescription | |
CN111222380B (en) | Living body detection method and device and recognition model training method thereof | |
WO2020195732A1 (en) | Image processing device, image processing method, and recording medium in which program is stored | |
CN111104833A (en) | Method and apparatus for in vivo examination, storage medium, and electronic device | |
CN111666835A (en) | Face living body detection method and device | |
CN109492509A (en) | Personal identification method, device, computer-readable medium and system | |
Nikitin et al. | Face anti-spoofing with joint spoofing medium detection and eye blinking analysis | |
Song et al. | Face spoofing detection by fusing binocular depth and spatial pyramid coding micro-texture features | |
CN113642639A (en) | Living body detection method, living body detection device, living body detection apparatus, and storage medium | |
WO2020204150A1 (en) | System and method for adaptively constructing a three-dimensional facial model based on two or more inputs of a two-dimensional facial image | |
Samal | Minimum resolution for human face detection and identification | |
CN112183357B (en) | Multi-scale living body detection method and system based on deep learning | |
CN111126283B (en) | Rapid living body detection method and system for automatically filtering fuzzy human face | |
CN111191519B (en) | Living body detection method for user access of mobile power supply device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |