CN106599826A - Face 3D reconstruction method based on near-infrared light - Google Patents
Face 3D reconstruction method based on near-infrared light Download PDFInfo
- Publication number
- CN106599826A CN106599826A CN201611129277.XA CN201611129277A CN106599826A CN 106599826 A CN106599826 A CN 106599826A CN 201611129277 A CN201611129277 A CN 201611129277A CN 106599826 A CN106599826 A CN 106599826A
- Authority
- CN
- China
- Prior art keywords
- camera
- face
- matrixes
- infrared light
- method based
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- 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
-
- 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
-
- 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
Abstract
The invention relates to a 3D reconstruction method and particularly relates to a face 3D reconstruction method based on near-infrared light. The method comprises steps that S1, Q matrixes of inner and outer parameters of a camera and a distortion parameter are acquired; S2, according to the Q matrixes of the inner and outer parameters of the camera and the distortion parameter, an image acquired by the camera is corrected, and pole lines of left and right frames of the image after correction are on one horizontal line; S3, a parallax matrix disparity of left and right views is calculated; S4, three-dimensional coordinates are reconstructed through utilizing the Q matrixes of the inner and outer parameters and the parallax matrix disparity; S5, the acquired three-dimensional coordinates are stretched to be a set of characteristic vectors; and S6, an SVM classifier is utilized to train the face and non-face characteristic vectors, and in vivo detection is finally realized. Compared with the prior art, the method is advantaged in that 1, in vivo detection is realized; 2, the method can be utilized under the no-visible-light environment; 3, performance robustness is realized; and 4, a fast speed is realized.
Description
Technical field
The present invention relates to the method that 3D rebuilds, more particularly to a kind of method based near infrared light 3D face reconstruction.
Background technology
At present, biometrics identification technology has been widely used in the every aspect in daily life.Face is biological
Feature identification technique, because with facilitating easy-to-use, user friendly the advantages of contactless, achieves in recent years prominent flying suddenly
The development entered.These development have been embodied in each research field, including Face datection, and face characteristic is extracted, classifier design with
And hardware device manufacture etc..However, the living things feature recognition based on face is still faced with some tests on application, it is existing
Some technology of identification are complicated and discrimination is low.
The content of the invention
It is an object of the invention to avoid the deficiency existing for above-mentioned prior art, propose a kind of based near infrared light people
The method that face 3D rebuilds.
The invention provides a kind of method based near infrared light 3D face reconstruction, comprises the following steps:
S1:Obtain the Q matrixes and distortion parameter of the inside and outside ginseng of camera;
S2:According to the Q matrixes and distortion parameter of the inside and outside ginseng of camera, the picture collected to camera carries out school
Just, the polar curve of the left and right picture of picture is in the same horizontal line after correction;
S3:Calculate the parallax matrix disparity of left and right view;
S4:Three-dimensional coordinate is rebuild using the Q matrixes and parallax matrix disparity of inside and outside ginseng;
S5:The three-dimensional coordinate for obtaining is stretched as into stack features vector;
S6:Face and non-face characteristic vector are trained using SVM classifier, finally realize In vivo detection.
Further, the camera in step S1 adopts infrared camera.
Further, the Q matrixes and the acquisition methods of distortion parameter of the inside and outside ginseng of the step S1 mid-infrared camera
For Zhang Zhengyou standardizations.
Further, inside and outside ginseng in step S1
Further, the parallax matrix disparity in step S3 is obtained by solid matching method.
Further, it is using the method for internal reference, outer ginseng and parallax matrix reconstruction three-dimensional coordinate in step S4:Root
Three-dimensional coordinate is obtained according to Q matrixes and disparity:
The three-dimensional coordinate of reconstruction is
Compared with prior art, the application has advantages below:1. In vivo detection can be realized;2. without under visible light environment
Also can use;3. performance robust;4. speed is fast.
Description of the drawings
With reference to the accompanying drawings and detailed description the present invention is further detailed explanation.
Fig. 1 is the schematic flow sheet of the embodiment of the present invention.
Specific embodiment
Below in conjunction with accompanying drawing, technical scheme is further described, but the present invention is not limited to these realities
Apply example.
With reference to accompanying drawing 1, the invention provides a kind of method based near infrared light 3D face reconstruction, comprises the following steps:
S1:Obtain the Q matrixes and distortion parameter of the inside and outside ginseng of camera;
S2:According to the Q matrixes and distortion parameter of the inside and outside ginseng of camera, the picture collected to camera carries out school
Just, the polar curve of the left and right picture of picture is in the same horizontal line after correction;
S3:Calculate the parallax matrix disparity of left and right view;
S4:Three-dimensional coordinate is rebuild using the Q matrixes and parallax matrix disparity of inside and outside ginseng;
S5:The three-dimensional coordinate for obtaining is stretched as into stack features vector;
S6:Face and non-face characteristic vector are trained using SVM classifier, finally realize In vivo detection.
Preferably, the camera in step S1 adopts infrared camera.Traditional camera is in the environment of without visible ray
Unusable, the infrared camera adopted in the present embodiment one equally can be used in the environment of without visible ray.
Further, the Q matrixes and the acquisition methods of distortion parameter of the inside and outside ginseng of the step S1 mid-infrared camera
For Zhang Zhengyou standardizations.Wherein, inside and outside ginsengHerein "
Positive friend's standardization " refers to that Zhang Zhengyou teaches the tessellated video camera in monoplane proposed in 1998 also known as " Zhang Shi standardizations "
Scaling method.Zhang Shi standardizations are widely used as tool box or packaged function.The original text that Zhang Shi is demarcated is " A
Flexible New Technique forCamera Calibration”.The method being previously mentioned in this text, is that camera calibration is carried
Convenience has been supplied, and with very high precision.Special demarcation thing can not be needed from this demarcation, it is only necessary to a printing
Gridiron pattern out.
Further, the parallax matrix disparity in step S3 is obtained by solid matching method.It is described here
Volume solid matching method belongs to prior art, and simple introduction is only done here.Here solid matching method adopts block
Matching algorithms.
It is using the method for internal reference, outer ginseng and parallax matrix reconstruction three-dimensional coordinate in step S4:According to Q matrixes and
Disparity obtains three-dimensional coordinate:
The three-dimensional coordinate of reconstruction is
SVM classifier is, by Nonlinear Mapping p, sample space to be mapped to a higher-dimension or even infinite dimensional spy
In levying space so that the problem of Nonlinear separability is converted into the linear separability in feature space in original sample space
Problem.Briefly, peacekeeping linearisation is exactly risen.Dimension is risen, exactly sample is done to higher dimensional space and is mapped, generally this meeting
Increase the complexity for calculating, or even " dimension disaster " can be caused, thus people seldom make inquiries.But ask as classification, recurrence etc.
For topic, it is likely that low-dimensional sample space cannot linear process sample set, in high-dimensional feature space but can pass through one
Individual linear hyperplane realizes linear partition (or recurrence).General liter dimension can all bring the complication of calculating, and SVM methods are dexterously
Solve this difficult problem:Using the expansion theorem of kernel function, avoid the need for knowing the explicit expression of Nonlinear Mapping;Due to being
Linear learning machine is set up in high-dimensional feature space, so compared with linear model, not only hardly increase the complexity of calculating,
And avoid to a certain extent " dimension disaster ".Everything will give the credit to the expansion of kernel function and computational theory.In this reality
In applying example, SVM classifier is trained to face and non-face characteristic vector, finally realizes In vivo detection, performance robust, speed
Degree is fast.
Specific embodiment described herein is only explanation for example spiritual to the present invention.Technology neck belonging to of the invention
The technical staff in domain can be made various modifications to described specific embodiment or supplement or replaced using similar mode
Generation, but without departing from the spiritual of the present invention or surmount scope defined in appended claims.
Claims (6)
1. a kind of method based near infrared light 3D face reconstruction, it is characterised in that comprise the following steps:
S1:Obtain the Q matrixes and distortion parameter of the inside and outside ginseng of camera;
S2:According to the Q matrixes and distortion parameter of the inside and outside ginseng of camera, the picture that camera is collected is corrected, school
The polar curve of the left and right picture of just rear picture is in the same horizontal line;
S3:Calculate the parallax matrix disparity of left and right view;
S4:Three-dimensional coordinate is rebuild using the Q matrixes and parallax matrix disparity of inside and outside ginseng;
S5:The three-dimensional coordinate for obtaining is stretched as into stack features vector;
S6:Face and non-face characteristic vector are trained using SVM classifier, finally realize In vivo detection.
2. the method based near infrared light 3D face reconstruction according to claim 1, it is characterised in that in step S1
Camera adopt infrared camera.
3. the method based near infrared light 3D face reconstruction according to claim 2, it is characterised in that in step S1
The Q matrixes of the inside and outside ginseng of infrared camera and the acquisition methods of distortion parameter are Zhang Zhengyou standardizations.
4. the method based near infrared light 3D face reconstruction according to claim 3, it is characterised in that in step S1
Inside and outside ginseng
5. the method based near infrared light 3D face reconstruction according to claim 1, it is characterised in that in step S3
Parallax matrix disparity obtained by solid matching method.
6. the method based near infrared light 3D face reconstruction according to claim 1, it is characterised in that in step S4
It is using the method for internal reference, outer ginseng and parallax matrix reconstruction three-dimensional coordinate:Three-dimensional seat is obtained according to Q matrixes and disparity
Mark:
The three-dimensional coordinate of reconstruction is
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611129277.XA CN106599826A (en) | 2016-12-09 | 2016-12-09 | Face 3D reconstruction method based on near-infrared light |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611129277.XA CN106599826A (en) | 2016-12-09 | 2016-12-09 | Face 3D reconstruction method based on near-infrared light |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106599826A true CN106599826A (en) | 2017-04-26 |
Family
ID=58597958
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611129277.XA Pending CN106599826A (en) | 2016-12-09 | 2016-12-09 | Face 3D reconstruction method based on near-infrared light |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106599826A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107590473A (en) * | 2017-09-19 | 2018-01-16 | 杭州登虹科技有限公司 | A kind of human face in-vivo detection method, medium and relevant apparatus |
CN108334817A (en) * | 2018-01-16 | 2018-07-27 | 深圳前海华夏智信数据科技有限公司 | Living body faces detection method and system based on three mesh |
CN109241832A (en) * | 2018-07-26 | 2019-01-18 | 维沃移动通信有限公司 | A kind of method and terminal device of face In vivo detection |
CN109299677A (en) * | 2018-09-07 | 2019-02-01 | 西安知微传感技术有限公司 | A kind of recognition of face living body judgment method and system |
CN111191556A (en) * | 2019-12-25 | 2020-05-22 | 杭州宇泛智能科技有限公司 | Face recognition method and device and electronic equipment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104346829A (en) * | 2013-07-29 | 2015-02-11 | 中国农业机械化科学研究院 | Three-dimensional color reconstruction system and method based on PMD (photonic mixer device) cameras and photographing head |
CN105203044A (en) * | 2015-05-27 | 2015-12-30 | 珠海真幻科技有限公司 | Method and system for stereoscopic vision three-dimensional measurement taking computing laser speckles as texture |
CN103065289B (en) * | 2013-01-22 | 2016-04-06 | 清华大学 | Based on four lens camera front face method for reconstructing of binocular stereo vision |
CN105740779A (en) * | 2016-01-25 | 2016-07-06 | 北京天诚盛业科技有限公司 | Method and device for human face in-vivo detection |
-
2016
- 2016-12-09 CN CN201611129277.XA patent/CN106599826A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103065289B (en) * | 2013-01-22 | 2016-04-06 | 清华大学 | Based on four lens camera front face method for reconstructing of binocular stereo vision |
CN104346829A (en) * | 2013-07-29 | 2015-02-11 | 中国农业机械化科学研究院 | Three-dimensional color reconstruction system and method based on PMD (photonic mixer device) cameras and photographing head |
CN105203044A (en) * | 2015-05-27 | 2015-12-30 | 珠海真幻科技有限公司 | Method and system for stereoscopic vision three-dimensional measurement taking computing laser speckles as texture |
CN105740779A (en) * | 2016-01-25 | 2016-07-06 | 北京天诚盛业科技有限公司 | Method and device for human face in-vivo detection |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107590473A (en) * | 2017-09-19 | 2018-01-16 | 杭州登虹科技有限公司 | A kind of human face in-vivo detection method, medium and relevant apparatus |
CN108334817A (en) * | 2018-01-16 | 2018-07-27 | 深圳前海华夏智信数据科技有限公司 | Living body faces detection method and system based on three mesh |
CN109241832A (en) * | 2018-07-26 | 2019-01-18 | 维沃移动通信有限公司 | A kind of method and terminal device of face In vivo detection |
CN109241832B (en) * | 2018-07-26 | 2021-03-30 | 维沃移动通信有限公司 | Face living body detection method and terminal equipment |
CN109299677A (en) * | 2018-09-07 | 2019-02-01 | 西安知微传感技术有限公司 | A kind of recognition of face living body judgment method and system |
CN111191556A (en) * | 2019-12-25 | 2020-05-22 | 杭州宇泛智能科技有限公司 | Face recognition method and device and electronic equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106599826A (en) | Face 3D reconstruction method based on near-infrared light | |
CA2934514C (en) | System and method for identifying faces in unconstrained media | |
CN106067190B (en) | A kind of generation of fast face threedimensional model and transform method based on single image | |
US10198623B2 (en) | Three-dimensional facial recognition method and system | |
Hassner et al. | Effective face frontalization in unconstrained images | |
CN110348330B (en) | Face pose virtual view generation method based on VAE-ACGAN | |
CN103530599B (en) | The detection method and system of a kind of real human face and picture face | |
CN105447441B (en) | Face authentication method and device | |
CN110110629A (en) | Personal information detection method and system towards indoor environmental condition control | |
CN104298995B (en) | Three-dimensional face identifying device and method based on three-dimensional point cloud | |
CN104809638A (en) | Virtual glasses trying method and system based on mobile terminal | |
CN103325120A (en) | Rapid self-adaption binocular vision stereo matching method capable of supporting weight | |
CN103971137A (en) | Three-dimensional dynamic facial expression recognition method based on structural sparse feature study | |
CN105243376A (en) | Living body detection method and device | |
CN103745209B (en) | A kind of face identification method and system | |
Song et al. | Describing trajectory of surface patch for human action recognition on RGB and depth videos | |
Smith et al. | Facial shape-from-shading and recognition using principal geodesic analysis and robust statistics | |
WO2020135125A1 (en) | Living body detection method and device | |
CN112257641A (en) | Face recognition living body detection method | |
CN112750531A (en) | Automatic inspection system, method, equipment and medium for traditional Chinese medicine | |
Li et al. | Head pose classification based on line portrait | |
Chen et al. | 2D facial landmark model design by combining key points and inserted points | |
CN109859306A (en) | A method of extracting manikin in the slave photo based on machine learning | |
CN108924542A (en) | Based on conspicuousness and sparsity without reference three-dimensional video quality evaluation method | |
CN104636727A (en) | Face recognition method applicable to multiple expressions and multiple gestures |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170426 |
|
RJ01 | Rejection of invention patent application after publication |