CN101593365A - A kind of method of adjustment of universal three-dimensional human face model - Google Patents
A kind of method of adjustment of universal three-dimensional human face model Download PDFInfo
- Publication number
- CN101593365A CN101593365A CNA2009100596644A CN200910059664A CN101593365A CN 101593365 A CN101593365 A CN 101593365A CN A2009100596644 A CNA2009100596644 A CN A2009100596644A CN 200910059664 A CN200910059664 A CN 200910059664A CN 101593365 A CN101593365 A CN 101593365A
- Authority
- CN
- China
- Prior art keywords
- model
- face
- dimensional
- human face
- universal
- 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
Images
Landscapes
- Processing Or Creating Images (AREA)
- Image Processing (AREA)
Abstract
A kind of method of adjustment of universal three-dimensional human face model belongs to the computer image processing technology field, is mainly concerned with the three-dimensional facial reconstruction technology in the three-dimensional face recognition technology individual in the living things feature recognition.The present invention is based on universal three-dimensional human face model, utilize the two-dimension human face image of two quadratures, adjust three-dimensional face model, finally obtain three-dimensional specific faceform by the three-dimensional coordinate of estimating human face characteristic point in two quadrature facial images.Active shape model, Procrustes are analyzed in the present invention and the symmetry architectural feature of people's face is applied in the adjustment process of universal three-dimensional human face model.Estimate the face feature point with active shape model, adjust three-dimensional face model by the symmetry architectural feature of Procrustes analysis and people's face then.With general method based on image comparatively speaking, the method among the present invention has not only improved the accuracy that model is adjusted, and speed also increases to a certain extent, has very strong versatility in the three-dimensional facial reconstruction based on image.
Description
Technical field
The invention belongs to the computer image processing technology field, be mainly concerned with the three-dimensional facial reconstruction technology in the three-dimensional face recognition technology individual in the living things feature recognition.
Background technology
Recognition of face, can be finished identification, verification process, thereby be subjected to researchist's extensive concern because of it does not have infringement for the person as one of biological characteristic authentication technology under the situation that the authentication object is not discovered.The business-like face identification system of part has also appearred at present.But when the imaging condition was uncertain, as different illumination conditions, different attitudes, the performance of face identification system may be plummeted to 60%～70%, and under many practical application conditions, the performance of each recognition system does not often reach 50%.Therefore, eliminate different light and attitude for the influence of recognition of face, become the subject matter that needs to be resolved hurrily, at the influence of illumination and attitude, early stage researchist is from the solution of starting with of the method for two dimension, as many view approach.Though these class methods can reduce the influence of illumination and attitude to a certain extent, but consider the influence of illumination and attitude for recognition of face, mainly be relevant with people's face inherent attributes such as the 3D shape of people's face, illumination reflectivity, therefore starting with from three-dimensional method is the basic method that solves illumination and attitude problem, and three-dimensional face is identified in the focus that has become research both at home and abroad at present.Three-dimensional face identification mainly comprises the reconstruction and the identification of specific people's face 3D shape.Wherein difficult point is the reconstruction of specific people's face 3D shape, the reconstruction of at present popular people's face 3D shape is divided into based on the three-dimensional reconstruction of 3 d discrete point data with based on three-dimensional reconstruction two classes of image according to the source of data, and the concrete people who how universal three-dimensional human face model is adjusted in the scene automatically is a key issue of the reconstruction of specific people's face 3D shape on the face.
Present normally used three-dimensional face model adjusting method has:
(1) based on the three-dimensional model method of adjustment of 3 d discrete point data.The method is obtained face's 3 d discrete point data with spatial digitizer earlier, handle face's 3 d discrete point data then obtaining the three-dimensional data of face's key feature point, and then adjust universal three-dimensional human face model with face's key feature point three-dimensional data of obtaining.Though this method precision height need expensive 3-D scanning equipment, and calculated amount is very big, is not suitable for practical application.See document for details: A-Nasser Ansari, " 3DFace Mesh Modeling from Range Images for 3D Face Recognition " Proceeding of IEEEICIP, pp.509-512,2007.
(2) based on the three-dimensional model method of adjustment of image.According to the difference of the number of the two dimensional image of using in the algorithm, can be divided into three-dimensional model adjustment based on the model method of adjustment of image based on single width, several and sequence image.The main thought of these class methods is by handling the three-dimensional data that two-dimension human face image obtains facial image, adjusting the general three-dimensional model by the three-dimensional data of obtaining then.The method has the little advantage of operand, and has taken into account cost and precision, as long as typical CCD camera just can satisfy the needs of application.See document for details: R.Lengagne, P.Fua, and O.Monga, " 3D Face Modeling fromStereo and differential Constraints ", and Proc.ICAFGR, April 1998, pp.148-153.
Summary of the invention
The invention provides a kind of universal three-dimensional human face model method of adjustment, this method is based on universal three-dimensional human face model, utilize the two-dimension human face image of two quadratures, adjust three-dimensional face model by the three-dimensional coordinate of estimating human face characteristic point in two quadrature facial images, finally obtain three-dimensional specific faceform.
In order to describe content of the present invention easily, at first some terms are defined.
Definition 1: universal three-dimensional human face model.Easy in order to calculate, the universal three-dimensional human face model of Shi Yonging is the Candide_3 model in the present invention.The Candide_3 model is a kind of parameterized model, and it is made up of 113 summits and 168 tri patchs.Each summit P
iThree coordinate figure x by its correspondence
i, y
i, z
iExpression, i.e. P
i=(x
i, y
i, z
i)
T, 1≤i≤113, the three-dimensional data serial connection on all 113 summits forms matrix Q and represents whole model, i.e. Q=(P
1, P
2... P
i... P
113).
Definition 2: facial contour.Facial contour comprises the profile of people's face shape of face, describes the profile of shapes such as eyes, nose, face.People's face of a people can be aged in people's face in life, but profile roughly can change hardly, and the profile of different people is different.
Definition 3: human face characteristic point.Human face characteristic point is meant some discrete points on the peripheral profile that can characterize people's face shape and face shape.
Definition 4: model points.Model points refers to 113 summits forming the Candide_3 model.
Definition 5: svd.M * n rank matrix A can be write as the form of A=USV ', and U is m rank orthogonal matrix, and V is n rank orthogonal matrix, S=diag (σ
1, σ
2..., σ
r), σ
i＞0 (i=1 ..., r), r=rank (A) is respectively the singular vector of A among U and the V, and the diagonal matrix that S is made up of the singular value of A.Svd and eigenvalue problem are closely connected, the quadrature unit character vector of AA ' is formed U, eigenwert is formed S ' S, the quadrature unit character vector of AA ' is formed V, eigenwert (identical with AA ') is formed S ' S, and svd provides some information about A, and for example the number of non-zero singular value (exponent number of S) is identical with the order of A, in case order r determines that the preceding r row of U have constituted the orthogonal basis in the column vector space of A so.
Definition 6: active shape model.Active shape model is the deformable model of a statistics, can the accurate localization human face characteristic point.
Technical solution of the present invention is as follows:
A kind of faceform's method of adjustment of universal three-dimensional human face model as shown in Figure 1, comprises the steps:
Step 1: two quadrature facial images gathering are carried out respectively detecting based on people's face of complexion model.
At first utilize the connective cut zone of skin pixel, use each zone of ellipse fitting then, judge according to the ratio of ellipse long and short shaft whether it is human face region, as shown in Figure 2.
Step 2: people's face key feature point.
To the human face region of detected two quadrature facial images in the step 1, adopt active shape model to carry out the location of people's face key feature point, as shown in Figure 3, specifically comprise step by step following:
Step 2-1: every in facial image database facial image is carried out the manual demarcation of unique point, obtain the unique point vector (i.e. shape sample) of every facial image, the unique point vector of face images is formed a characteristic point matrix (being the shape sample set).
Step 2-2: the alignment of shape sample set.With align all shape samples in the shape sample set of Procrustes method.
Step 2-3: the foundation of shape.Shape sample after all alignment is done principal component analysis (PCA), thereby shape is set up in the variation that obtains shape.
Step 2-4: the foundation of local gray level model.For each point on the shape is set up the local gray level derivative model.
Step 2-5: the active shape model of forming based on the local gray level model of the shape of step 2-3 gained and step 2-4 gained, carry out the face characteristic point location to the human face region of step 1 gained.
Step 3: the obtaining of human face characteristic point three-dimensional coordinate.
Face characteristic point coordinate by the front face image of step 2-5 gained is X
f=[x
f, y
f]
T, the corresponding human face characteristic point coordinate of Side Face Image is X
p=[y
p, z
p]
TBecause front input picture and side input picture are orthogonal images, y is arranged
f=y
pSo the three-dimensional coordinate that obtains this human face characteristic point is P
i=[x
f, y
f, z
p]
T
Step 4: universal three-dimensional human face model is carried out overall situation adjustment, as shown in Figure 4.
At first carry out overall situation adjustment at front face image.By rotation, translation, three kinds of conversion of convergent-divergent, make on the universal three-dimensional human face model with the projection of the corresponding summit of human face characteristic point on the x-y plane and input facial image in the human face characteristic point estimated match.The three-dimensional coordinate data of N human face characteristic point is formed the matrix P of N * 3
I, form the matrix P of N * 3 on the universal three-dimensional human face model with the three-dimensional coordinate data on the corresponding N of a human face characteristic point summit
MSpecifically comprise step by step following:
Step 4-1: centralization P
IAnd P
MTo P
IAnd P
MEach row of matrix are averaged, then with P
IAnd P
MIn each element deduct the mean value of this element column, obtain centralization matrix P '
IAnd P '
M
Step 4-2: normalization P '
IAnd P '
M, obtain normalization matrix P "
IAnd P "
M
Step 4-3: order
Matrix A is carried out svd, obtain A=USV
TForm, wherein U is n rank orthogonal matrix, V is n rank orthogonal matrix, the diagonal matrix that S is made up of the singular value of A.
Step 4-4: calculate rotation matrix R=VU
T
Step 4-5: calculate zoom factor S=∑ S * ‖ P "
I‖/‖ P "
M‖, wherein ‖ ‖ representing matrix is asked modular arithmetic.
Step 4-6: calculating translation vector T=E (P "
I)-SE (P "
M) R.
Step 4-7: employing zoom factor S, rotation matrix R and translation vector T adjust all summits on the universal three-dimensional human face model, obtain overall adjusted universal three-dimensional human face model summit matrix P, and it is P=SP that the concrete overall situation is adjusted formula
MR+T.
Carry out overall situation adjustment at Side Face Image then.By rotation, translation, three kinds of conversion of convergent-divergent, make on the universal three-dimensional human face model to match with importing the human face characteristic point of estimating in the facial image with the projection of the corresponding summit of human face characteristic point on the y-z plane, concrete grammar such as step 4-1 are to step 4-7.
Step 5: universal three-dimensional human face model is carried out part adjustment.
In order to make universal three-dimensional human face model, need carry out part adjustment to overall situation adjustment model later more near three-dimensional specific faceform.The part adjustment of model is the adjustment to critical organ in people's face, comprises the adjustment of critical organ in the adjustment of critical organ in the front face image and the Side Face Image.
At first critical organ in the front face image is adjusted, is specifically comprised step by step following:
Step 5-1: the part of eyes is adjusted.The adjusted universal three-dimensional human face model of the overall situation is rotated to standard front face, then with model projection to the x-y plane.If the left eye angular vertex p of left eye eyeball on the universal three-dimensional human face model
1Projection coordinate on the x-y plane is (x
1, y
1), the summit p at the right eye angle of right eye eyeball on the model
2Projection coordinate on the x-y plane is (x
2, y
2).By the structural attitude of people's face eyes, the y coordinate at the left eye angle of left eye eyeball should equate with the y coordinate on the summit at the right eye angle of right eye eyeball, therefore can be with p
1, p
2The coordinate of point is adjusted into (x
1, (y
1+ y
2)/2) and (x
2, (y
1+ y
2)/2).According to the method described above, the coordinate on other eyes summit on the adjustment model.
Step 5-2: the adjustment of face and nose.By the symmetry of people's face as can be known, in universal three-dimensional human face model, the projection on the x-y plane is with x=(x with face and the corresponding model vertices of nose unique point
1+ x
2)/2 are axis of symmetry, wherein x
1, x
2Be respectively the horizontal ordinate on summit at the left eye angle of left eye eyeball on the three-dimensional face model, the horizontal ordinate on the summit at the right eye angle of right eye eyeball on the model.If any one face summit m on the model
1Projection coordinate on the x-y plane is (x
3, y
3), symmetrical with it summit m on model
2Projection coordinate on the x-y plane is (x
4, y
4), because m
1And m
2With x=(x
1+ x
2)/2 are symmetry, so summit m
1And m
2Coordinate should be adjusted into (x
1+ x
2-x
4, (y
3+ y
4)/2) and (x
4, (y
3+ y
4)/2).According to the geometric relationship between projection summit and the subpoint, we can adjust and m
1And m
2The coordinate of corresponding model vertices.According to the method described above, the coordinate on other faces and nose summit on the adjustment model.
Need to prove:
1, the bold race capital in the formula all represents it is matrix or vector in this patent, other be scalar.
2, step 2 is with reference to classical active shape model algorithm.
3, the Procrustes analytical approach among the step 2-1 can be with reference to the paper Procrustes Analysis of Amy Ross.
Innovation part of the present invention is:
Active shape model, Procrustes are analyzed in the present invention and the symmetry architectural feature of people's face is applied in the adjustment process of universal three-dimensional human face model.Estimate the face feature point with active shape model, adjust three-dimensional face model by the symmetry architectural feature of Procrustes analysis and people's face then.With general method based on image comparatively speaking, the method among the present invention has not only improved the accuracy that model is adjusted, and speed also increases to a certain extent, has very strong versatility in the three-dimensional facial reconstruction based on image.
Description of drawings
Fig. 1 is the schematic flow sheet of general three-dimensional people face wire-frame model method of adjustment of the present invention.
Fig. 2 is people's face testing process synoptic diagram.
Fig. 3 is a human face characteristic point positioning flow synoptic diagram.
Fig. 4 is that the overall situation of general three-dimensional people face wire-frame model is adjusted schematic flow sheet.
Embodiment
Adopt method of the present invention, at first, carry out the programming of system software then with the C language in the emulation of the enterprising line algorithm of Matlab platform.By a large amount of experiment showed, with traditional method comparatively speaking, specific faceform is more accurate for the resulting three-dimensional of this algorithm, and speed also increases.
In sum, with the method among the present invention, can adjust automatically three-dimensional face model fast and accurately, thereby obtain three-dimensional specific faceform.
Claims (1)
1, a kind of faceform's method of adjustment of universal three-dimensional human face model comprises the steps:
Step 1: two quadrature facial images gathering are carried out respectively detecting based on people's face of complexion model;
At first utilize the connective cut zone of skin pixel, use each zone of ellipse fitting then, judge according to the ratio of ellipse long and short shaft whether it is human face region;
Step 2: people's face key feature point;
To the human face region of detected two quadrature facial images in the step 1, adopt active shape model to carry out the location of people's face key feature point, specifically comprise step by step following:
Step 2-1: every in facial image database facial image is carried out the manual demarcation of unique point, obtain the unique point vector of every facial image, i.e. a shape sample; The unique point vector of face images is formed a characteristic point matrix, i.e. shape sample set;
Step 2-2: the alignment of shape sample set; With align all shape samples in the shape sample set of Procrustes method;
Step 2-3: the foundation of shape; Shape sample after all alignment is done principal component analysis (PCA), thereby shape is set up in the variation that obtains shape;
Step 2-4: the foundation of local gray level model; For each point on the shape is set up the local gray level derivative model;
Step 2-5: the active shape model of forming based on the local gray level model of the shape of step 2-3 gained and step 2-4 gained, carry out the face characteristic point location to the human face region of step 1 gained;
Step 3: the obtaining of human face characteristic point three-dimensional coordinate;
Face characteristic point coordinate by the front face image of step 2-5 gained is X
f=[x
f, y
f]
T, the corresponding human face characteristic point coordinate of Side Face Image is X
p=[y
p, z
p]
TBecause front input picture and side input picture are orthogonal images, y is arranged
f=y
pSo the three-dimensional coordinate that obtains this human face characteristic point is P
i=[x
f, y
f, z
p]
T
Step 4: universal three-dimensional human face model is carried out overall situation adjustment;
At first carry out overall situation adjustment at front face image; By rotation, translation, three kinds of conversion of convergent-divergent, make on the universal three-dimensional human face model with the projection of the corresponding summit of human face characteristic point on the x-y plane and input facial image in the human face characteristic point estimated match; The three-dimensional coordinate data of N human face characteristic point is formed the matrix P of N * 3
I, form the matrix P of N * 3 on the universal three-dimensional human face model with the three-dimensional coordinate data on the corresponding N of a human face characteristic point summit
MSpecifically comprise step by step following:
Step 4-1: centralization P
IAnd P
MTo P
IAnd P
MEach row of matrix are averaged, then with P
IAnd P
MIn each element deduct the mean value of this element column, obtain centralization matrix P '
IAnd P '
M
Step 4-2: normalization P '
IAnd P '
M, obtain normalization matrix P "
IAnd P "
M
Step 4-3: make A=P "
I TP "
M, matrix A is carried out svd, obtain A=USV
TForm, wherein U is n rank orthogonal matrix, V is n rank orthogonal matrix, the diagonal matrix that S is made up of the singular value of A;
Step 4-4: calculate rotation matrix R=VU
T
Step 4-5: calculating zoom factor S=∑ S * || P "
I||/|| P "
M||, wherein || || representing matrix is asked modular arithmetic;
Step 4-6: calculating translation vector T=E (P "
I)-SE (P "
M) R;
Step 4-7: employing zoom factor S, rotation matrix R and translation vector T adjust all summits on the universal three-dimensional human face model, obtain overall adjusted universal three-dimensional human face model summit matrix P, and it is P=SP that the concrete overall situation is adjusted formula
MR+T;
Carry out overall situation adjustment at Side Face Image then.By rotation, translation, three kinds of conversion of convergent-divergent, make on the universal three-dimensional human face model to match with importing the human face characteristic point of estimating in the facial image with the projection of the corresponding summit of human face characteristic point on the y-z plane, concrete grammar such as step 4-1 are to step 4-7;
Step 5: universal three-dimensional human face model is carried out part adjustment;
In order to make universal three-dimensional human face model, need carry out part adjustment to overall situation adjustment model later more near three-dimensional specific faceform; The part adjustment of model is the adjustment to critical organ in people's face, comprises the adjustment of critical organ in the adjustment of critical organ in the front face image and the Side Face Image;
At first critical organ in the front face image is adjusted, is specifically comprised step by step following:
Step 5-1: the part of eyes is adjusted; The adjusted universal three-dimensional human face model of the overall situation is rotated to standard front face, then with model projection to the x-y plane; If the left eye angular vertex p of left eye eyeball on the universal three-dimensional human face model
1Projection coordinate on the x-y plane is (x
1, y
1), the summit p at the right eye angle of right eye eyeball on the model
2Projection coordinate on the x-y plane is (x
2, y
2); By the structural attitude of people's face eyes, the y coordinate at the left eye angle of left eye eyeball should equate with the y coordinate on the summit at the right eye angle of right eye eyeball, therefore can be with p
1, p
2The coordinate of point is adjusted into (x
1, (y
1+ y
2)/2) and (x
2, (y
1+ y
2)/2); According to the method described above, the coordinate on other eyes summit on the adjustment model;
Step 5-2: the adjustment of face and nose; By the symmetry of people's face as can be known, in universal three-dimensional human face model, the projection on the x-y plane is with x=(x with face and the corresponding model vertices of nose unique point
1+ x
2)/2 are axis of symmetry, wherein x
1, x
2Be respectively the horizontal ordinate on summit at the left eye angle of left eye eyeball on the three-dimensional face model, the horizontal ordinate on the summit at the right eye angle of right eye eyeball on the model; If any one face summit m on the model
1Projection coordinate on the x-y plane is (x
3, y
3), symmetrical with it summit m on model
2Projection coordinate on the x-y plane is (x
4, y
4), because m
1And m
2With x=(x
1+ x
2)/2 are symmetry, so summit m
1And m
2Coordinate should be adjusted into (x
1+ x
2-x
4, (y
3+ y
4)/2) and (x
4, (y
3+ y
4)/2); According to the geometric relationship between projection summit and the subpoint, we can adjust and m
1And m
2The coordinate of corresponding model vertices; According to the method described above, the coordinate on other faces and nose summit on the adjustment model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2009100596644A CN101593365A (en) | 2009-06-19 | 2009-06-19 | A kind of method of adjustment of universal three-dimensional human face model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2009100596644A CN101593365A (en) | 2009-06-19 | 2009-06-19 | A kind of method of adjustment of universal three-dimensional human face model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101593365A true CN101593365A (en) | 2009-12-02 |
Family
ID=41408006
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2009100596644A Pending CN101593365A (en) | 2009-06-19 | 2009-06-19 | A kind of method of adjustment of universal three-dimensional human face model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101593365A (en) |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819627A (en) * | 2010-03-30 | 2010-09-01 | 北京蓝色星河软件技术发展有限公司 | Method for identifying abnormal behavior of people based on space conversion |
| CN102663820A (en) * | 2012-04-28 | 2012-09-12 | 清华大学 | Three-dimensional head model reconstruction method |
| CN103116902A (en) * | 2011-11-16 | 2013-05-22 | 华为软件技术有限公司 | Three-dimensional virtual human head image generation method, and method and device of human head image motion tracking |
| WO2015054991A1 (en) * | 2013-10-18 | 2015-04-23 | 华为技术有限公司 | Method and apparatus for positioning characteristic point |
| CN104574432A (en) * | 2015-02-15 | 2015-04-29 | 四川川大智胜软件股份有限公司 | Three-dimensional face reconstruction method and three-dimensional face reconstruction system for automatic multi-view-angle face auto-shooting image |
| CN104599284A (en) * | 2015-02-15 | 2015-05-06 | 四川川大智胜软件股份有限公司 | Three-dimensional facial reconstruction method based on multi-view cellphone selfie pictures |
| CN104833373A (en) * | 2015-04-21 | 2015-08-12 | 西安理工大学 | Procrustes-based geomagnetic matching navigation method |
| CN104978764A (en) * | 2014-04-10 | 2015-10-14 | 华为技术有限公司 | Three-dimensional face mesh model processing method and three-dimensional face mesh model processing equipment |
| CN105701763A (en) * | 2015-12-30 | 2016-06-22 | 青岛海信移动通信技术股份有限公司 | Method and device for adjusting face image |
| CN105763829A (en) * | 2014-12-18 | 2016-07-13 | 联想(北京)有限公司 | Image processing method and electronic device |
| CN106462738A (en) * | 2014-05-20 | 2017-02-22 | 埃西勒国际通用光学公司 | Method for constructing a model of the face of a person, method and device for posture analysis using such a model |
| CN104077804B (en) * | 2014-06-09 | 2017-03-01 | 广州嘉崎智能科技有限公司 | A kind of method based on multi-frame video picture construction three-dimensional face model |
| WO2017202191A1 (en) * | 2016-05-27 | 2017-11-30 | 广州帕克西软件开发有限公司 | Facial data measurement method and system |
| CN108154550A (en) * | 2017-11-29 | 2018-06-12 | 深圳奥比中光科技有限公司 | Face real-time three-dimensional method for reconstructing based on RGBD cameras |
| CN108463823A (en) * | 2016-11-24 | 2018-08-28 | 华为技术有限公司 | A kind of method for reconstructing, device and the terminal of user's Hair model |
| CN109064472A (en) * | 2017-03-28 | 2018-12-21 | 合肥工业大学 | A kind of approximating method and device of the three-dimensional space model fit Plane of vertebrae |
| CN109087340A (en) * | 2018-06-04 | 2018-12-25 | 成都通甲优博科技有限责任公司 | A kind of face three-dimensional rebuilding method and system comprising dimensional information |
| CN109118579A (en) * | 2018-08-03 | 2019-01-01 | 北京微播视界科技有限公司 | The method, apparatus of dynamic generation human face three-dimensional model, electronic equipment |
| CN109151540A (en) * | 2017-06-28 | 2019-01-04 | 武汉斗鱼网络科技有限公司 | The interaction processing method and device of video image |
| CN109389682A (en) * | 2017-08-09 | 2019-02-26 | 上海影子智能科技有限公司 | A kind of three-dimensional face model automatic adjusting method |
| CN109859305A (en) * | 2018-12-13 | 2019-06-07 | 中科天网（广东）科技有限公司 | Three-dimensional face modeling, recognition methods and device based on multi-angle two-dimension human face |
| CN110363175A (en) * | 2019-07-23 | 2019-10-22 | 厦门美图之家科技有限公司 | Image processing method, device and electronic equipment |
| CN111324250A (en) * | 2020-01-22 | 2020-06-23 | 腾讯科技（深圳）有限公司 | Three-dimensional image adjusting method, device and equipment and readable storage medium |
| CN111582143A (en) * | 2020-05-06 | 2020-08-25 | 郑州工程技术学院 | Student classroom attendance method and device based on image recognition and storage medium |
| CN112232274A (en) * | 2020-11-03 | 2021-01-15 | 支付宝(杭州)信息技术有限公司 | Depth image model training method and device |
| CN112464699A (en) * | 2019-09-06 | 2021-03-09 | 富士通株式会社 | Image normalization method, system and readable medium for face analysis |
| CN113554745A (en) * | 2021-07-15 | 2021-10-26 | 电子科技大学 | Three-dimensional face reconstruction method based on image |
| CN117082188A (en) * | 2023-10-12 | 2023-11-17 | 广东工业大学 | Consistency video generation method and related device based on Pruk analysis |
| CN117238020A (en) * | 2023-11-10 | 2023-12-15 | 杭州启源视觉科技有限公司 | Face recognition method, device and computer equipment |
-
2009
- 2009-06-19 CN CNA2009100596644A patent/CN101593365A/en active Pending
Cited By (41)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819627A (en) * | 2010-03-30 | 2010-09-01 | 北京蓝色星河软件技术发展有限公司 | Method for identifying abnormal behavior of people based on space conversion |
| CN103116902A (en) * | 2011-11-16 | 2013-05-22 | 华为软件技术有限公司 | Three-dimensional virtual human head image generation method, and method and device of human head image motion tracking |
| CN102663820A (en) * | 2012-04-28 | 2012-09-12 | 清华大学 | Three-dimensional head model reconstruction method |
| CN102663820B (en) * | 2012-04-28 | 2014-10-22 | 清华大学 | Three-dimensional head model reconstruction method |
| WO2015054991A1 (en) * | 2013-10-18 | 2015-04-23 | 华为技术有限公司 | Method and apparatus for positioning characteristic point |
| CN104978764A (en) * | 2014-04-10 | 2015-10-14 | 华为技术有限公司 | Three-dimensional face mesh model processing method and three-dimensional face mesh model processing equipment |
| CN104978764B (en) * | 2014-04-10 | 2017-11-17 | 华为技术有限公司 | 3 d human face mesh model processing method and equipment |
| CN106462738A (en) * | 2014-05-20 | 2017-02-22 | 埃西勒国际通用光学公司 | Method for constructing a model of the face of a person, method and device for posture analysis using such a model |
| US10380411B2 (en) | 2014-05-20 | 2019-08-13 | Essilor International | Method for constructing a model of the face of a person, method and device for posture analysis using such a model |
| CN106462738B (en) * | 2014-05-20 | 2020-10-09 | 依视路国际公司 | Method for constructing a model of a person's face, method and apparatus for analyzing a pose using such a model |
| CN104077804B (en) * | 2014-06-09 | 2017-03-01 | 广州嘉崎智能科技有限公司 | A kind of method based on multi-frame video picture construction three-dimensional face model |
| CN105763829A (en) * | 2014-12-18 | 2016-07-13 | 联想(北京)有限公司 | Image processing method and electronic device |
| CN104574432B (en) * | 2015-02-15 | 2017-05-24 | 四川川大智胜软件股份有限公司 | Three-dimensional face reconstruction method and three-dimensional face reconstruction system for automatic multi-view-angle face auto-shooting image |
| CN104599284B (en) * | 2015-02-15 | 2017-06-13 | 四川川大智胜软件股份有限公司 | Three-dimensional facial reconstruction method based on various visual angles mobile phone auto heterodyne image |
| CN104599284A (en) * | 2015-02-15 | 2015-05-06 | 四川川大智胜软件股份有限公司 | Three-dimensional facial reconstruction method based on multi-view cellphone selfie pictures |
| CN104574432A (en) * | 2015-02-15 | 2015-04-29 | 四川川大智胜软件股份有限公司 | Three-dimensional face reconstruction method and three-dimensional face reconstruction system for automatic multi-view-angle face auto-shooting image |
| CN104833373A (en) * | 2015-04-21 | 2015-08-12 | 西安理工大学 | Procrustes-based geomagnetic matching navigation method |
| CN105701763B (en) * | 2015-12-30 | 2019-05-21 | 青岛海信移动通信技术股份有限公司 | A kind of method and device adjusting facial image |
| CN105701763A (en) * | 2015-12-30 | 2016-06-22 | 青岛海信移动通信技术股份有限公司 | Method and device for adjusting face image |
| WO2017202191A1 (en) * | 2016-05-27 | 2017-11-30 | 广州帕克西软件开发有限公司 | Facial data measurement method and system |
| CN108463823A (en) * | 2016-11-24 | 2018-08-28 | 华为技术有限公司 | A kind of method for reconstructing, device and the terminal of user's Hair model |
| CN109064472B (en) * | 2017-03-28 | 2020-09-04 | 合肥工业大学 | Fitting method and device for fitting plane of three-dimensional space model of vertebra |
| CN109064472A (en) * | 2017-03-28 | 2018-12-21 | 合肥工业大学 | A kind of approximating method and device of the three-dimensional space model fit Plane of vertebrae |
| CN109151540B (en) * | 2017-06-28 | 2021-11-09 | 武汉斗鱼网络科技有限公司 | Interactive processing method and device for video image |
| CN109151540A (en) * | 2017-06-28 | 2019-01-04 | 武汉斗鱼网络科技有限公司 | The interaction processing method and device of video image |
| CN109389682A (en) * | 2017-08-09 | 2019-02-26 | 上海影子智能科技有限公司 | A kind of three-dimensional face model automatic adjusting method |
| CN108154550B (en) * | 2017-11-29 | 2021-07-06 | 奥比中光科技集团股份有限公司 | RGBD camera-based real-time three-dimensional face reconstruction method |
| CN108154550A (en) * | 2017-11-29 | 2018-06-12 | 深圳奥比中光科技有限公司 | Face real-time three-dimensional method for reconstructing based on RGBD cameras |
| CN109087340A (en) * | 2018-06-04 | 2018-12-25 | 成都通甲优博科技有限责任公司 | A kind of face three-dimensional rebuilding method and system comprising dimensional information |
| CN109118579A (en) * | 2018-08-03 | 2019-01-01 | 北京微播视界科技有限公司 | The method, apparatus of dynamic generation human face three-dimensional model, electronic equipment |
| CN109859305A (en) * | 2018-12-13 | 2019-06-07 | 中科天网（广东）科技有限公司 | Three-dimensional face modeling, recognition methods and device based on multi-angle two-dimension human face |
| CN110363175A (en) * | 2019-07-23 | 2019-10-22 | 厦门美图之家科技有限公司 | Image processing method, device and electronic equipment |
| CN112464699A (en) * | 2019-09-06 | 2021-03-09 | 富士通株式会社 | Image normalization method, system and readable medium for face analysis |
| CN111324250A (en) * | 2020-01-22 | 2020-06-23 | 腾讯科技（深圳）有限公司 | Three-dimensional image adjusting method, device and equipment and readable storage medium |
| CN111582143A (en) * | 2020-05-06 | 2020-08-25 | 郑州工程技术学院 | Student classroom attendance method and device based on image recognition and storage medium |
| CN112232274A (en) * | 2020-11-03 | 2021-01-15 | 支付宝(杭州)信息技术有限公司 | Depth image model training method and device |
| CN113554745A (en) * | 2021-07-15 | 2021-10-26 | 电子科技大学 | Three-dimensional face reconstruction method based on image |
| CN117082188A (en) * | 2023-10-12 | 2023-11-17 | 广东工业大学 | Consistency video generation method and related device based on Pruk analysis |
| CN117082188B (en) * | 2023-10-12 | 2024-01-30 | 广东工业大学 | Consistency video generation method and related device based on Pruk analysis |
| CN117238020A (en) * | 2023-11-10 | 2023-12-15 | 杭州启源视觉科技有限公司 | Face recognition method, device and computer equipment |
| CN117238020B (en) * | 2023-11-10 | 2024-04-26 | 杭州启源视觉科技有限公司 | Face recognition method, device and computer equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101593365A (en) | A kind of method of adjustment of universal three-dimensional human face model | |
| Jiang et al. | Wearable vision assistance system based on binocular sensors for visually impaired users | |
| CN105069746B (en) | Video real-time face replacement method and its system based on local affine invariant and color transfer technology | |
| US7512255B2 (en) | Multi-modal face recognition | |
| CN106951840A (en) | A kind of facial feature points detection method | |
| US20020097906A1 (en) | Method and apparatus for collating object | |
| CN101561710A (en) | Man-machine interaction method based on estimation of human face posture | |
| JP6207210B2 (en) | Information processing apparatus and method | |
| CN108537191A (en) | A kind of three-dimensional face identification method based on structure light video camera head | |
| WO2007062478A1 (en) | Visual tracking of eye glasses in visual head and eye tracking systems | |
| CN107229920B (en) | Behavior identification method based on integration depth typical time warping and related correction | |
| CN101650777A (en) | Corresponding three-dimensional face recognition method based on dense point | |
| CN101996416A (en) | 3D face capturing method and equipment | |
| CN103093237B (en) | A kind of method for detecting human face of structure based model | |
| Wang et al. | Dynamic human body reconstruction and motion tracking with low-cost depth cameras | |
| CN113920571A (en) | Micro-expression identification method and device based on multi-motion feature fusion | |
| CN106204564A (en) | A kind of laser photocentre extracting method | |
| Lee et al. | Anlysis methods of the variation of facial size and shape based on 3D face scan images | |
| JP6118976B2 (en) | Head posture estimation apparatus, head posture estimation method, and program for causing computer to execute head posture estimation method | |
| Pentland | Parts: Structured Descriptions of Shape. | |
| CN106406507A (en) | Image processing method and electronic equipment | |
| CN101393639A (en) | Bidirectional constraint method for matching two-dimensional object point with image point | |
| CN106485667B (en) | A kind of cranium face restored method | |
| Ye et al. | Active contour image segmentation method for training talents of computer graphics and image processing technology | |
| López et al. | Shape model and Hermite features for the segmentation of the cerebellum in fetal ultrasound |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20091202 |