CN109671148A - A kind of human face three-dimensional model automatic generation method with dynamic expression movement - Google Patents
A kind of human face three-dimensional model automatic generation method with dynamic expression movement Download PDFInfo
- Publication number
- CN109671148A CN109671148A CN201910079081.1A CN201910079081A CN109671148A CN 109671148 A CN109671148 A CN 109671148A CN 201910079081 A CN201910079081 A CN 201910079081A CN 109671148 A CN109671148 A CN 109671148A
- Authority
- CN
- China
- Prior art keywords
- dimensional
- face model
- standard
- static
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
- G06T2207/30201—Face
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Processing Or Creating Images (AREA)
Abstract
The present invention provides a kind of human face three-dimensional model automatic generation method with dynamic expression movement, comprising the following steps: makes the standard three-dimensional faceform with movement expression, the three-dimensional vertices P in record standard three-dimensional face modelm, and mark the first key point O in standard three-dimensional faceformm;Static three-dimensional faceform is obtained using structural light three-dimensional scanning system, records the three-dimensional vertices P in static three-dimensional faceformn, and mark the second key point in static three-dimensional faceform;By by static three-dimensional faceform the second key point and standard three-dimensional faceform in the first key point OmIt is corresponding to be overlapped, using Principal Component Analysis by the three-dimensional vertices P in standard faces modelmIt is matched to the three-dimensional vertices P of static three-dimensional faceformnIn, realize that static three-dimensional faceform is merged with standard three-dimensional faceform.Human face three-dimensional model automatic generation method provided by the invention with dynamic expression movement, it is easy to use, producing efficiency is high.
Description
[technical field]
The present invention relates to technical field of computer vision more particularly to a kind of face three-dimensional moulds with dynamic expression movement
Type automatic generation method.
[background technique]
With the development of computer graphics techniques, three-dimensional face modeling and the three-dimensional face model with facial expressions and acts become
One research hotspot of field of Computer Graphics.Three-dimensional face modeling and the three-dimensional face model with facial expressions and acts gradually by
It promotes and applies to numerous areas such as virtual reality, production of film and TV, recognition of face, Entertainments, there is very strong application value.
The method that three-dimensional face model with facial expressions and acts generates at present are as follows: obtained by structured light three-dimensional imaging equipment
Static Human Face threedimensional model (referring in particular to CN106164979A) then recycles animation soft (such as U.S. Autodesk
The Maya software of company) it is acted manually to the binding of Static Human Face model, make the three-dimensional face with facial expressions and acts.But
Movement binding is carried out by Three-dimensional Animation Software, is appropriate only for the three-dimensional face production of single facial expressions and acts;In large batch of table
The three-dimensional face production of feelings movement, such as personalized face three-dimensional expression production (are embodied in same static three-dimensional model
Binding covers facial expressions and acts), the role switching in personalized three-dimensional game updates and (is embodied in different static models bindings
Identical facial expressions and acts) etc. application in, by animation soft carry out movement binding have considerable repetitive operation, make
Time length, low efficiency.
In consideration of it, it is really necessary to provide a kind of new human face three-dimensional model automatic generation method with dynamic expression movement
To overcome drawbacks described above.
[summary of the invention]
The object of the present invention is to provide a kind of easy to use, three-dimensional face model expression that producing efficiency is high sides of binding automatically
Method.
To achieve the goals above, the present invention provide it is a kind of with dynamic expression movement human face three-dimensional model automatically generate
Method, comprising the following steps:
S1 makes the standard three-dimensional faceform with movement expression, the three-dimensional top in record standard three-dimensional face model
Point Pm, and mark the first key point O in standard three-dimensional faceformm;
S2 obtains static three-dimensional faceform using structural light three-dimensional scanning system, records in static three-dimensional faceform
Three-dimensional vertices Pn, and mark the second key point in static three-dimensional faceform;
S3, by the way that first in the second key point and standard three-dimensional faceform in static three-dimensional faceform is crucial
Point OmIt is corresponding to be overlapped, using Principal Component Analysis by the three-dimensional vertices P in standard faces modelmIt is matched to static three-dimensional face mould
The three-dimensional vertices P of typenIn, realize that static three-dimensional faceform is merged with standard three-dimensional faceform.
In a preferred embodiment, the prenasale of the standard three-dimensional faceform is the first key point Om, described
The prenasale of static three-dimensional faceform is the second key point On。
In a preferred embodiment, by by the second key point in static three-dimensional faceform described in step S3
With the first key point O in standard three-dimensional faceformmIt is corresponding be overlapped the following steps are included:
S31, respectively with the first key point O of standard three-dimensional faceformmCentered on point, static three-dimensional faceform
Two key point OnCentered on point, take the three-dimensional vertices P of standard three-dimensional faceform Yu static three-dimensional faceform respectivelym、Pn;
S32, by principal component analysis system, respectively by the three-dimensional top of standard three-dimensional faceform and static three-dimensional faceform
Point Pm、PnIt is reduced to 2 D vertex data;In Two dimensional Distribution, the standard system of standard three-dimensional faceform is Om-XmYmZm, static state three
The standard system for tieing up faceform is On-XnYnZn
Then the standard system of static three-dimensional faceform is O by S33n-XnYnZnBecome by spin matrix R, transfer matrix T
It changes so that On-XnYnZnStandard system with standard three-dimensional faceform is Om-XmYmZmIt is overlapped.
In a preferred embodiment, described in step S32 in Two dimensional Distribution, the standard of standard three-dimensional faceform
System is Om-XmYmZm, the standard system of static three-dimensional faceform is On-XnYnZnDetailed process are as follows:
Standard three-dimensional faceform is to cross the first key point O in standard three-dimensional faceform front viewmLine of symmetry be Ym
Axis, excessively corresponding first key point OmAnd vertical YmThe straight line of axis is XmAxis is respectively perpendicular YmAxis, XmThe straight line of axis is ZmAxis is established
The standard system of standard three-dimensional faceform is Om-XmYmZm;
Static three-dimensional faceform is to cross the second key point O in static three-dimensional faceform's front viewnLine of symmetry be Yn
Axis crosses the second key point OnAnd vertical YnThe straight line of axis is XnAxis is respectively perpendicular YnAxis, XnThe straight line of axis is ZnAxis establishes standard
The standard system of three-dimensional face model is On-XnYnZn。
In a preferred embodiment, it is described using Principal Component Analysis by the three-dimensional vertices P in standard faces modelm
It is matched to the three-dimensional vertices P of static three-dimensional faceformnIn the following steps are included:
S34, static three-dimensional faceform is with the first key point OmFor the center of circle, rmFor the ball internal cutting of the radius of a ball, by static state three
Dimension faceform's is cut into the effective coverage that can cover face's main feature;
S35, each three-dimensional vertices P of standard three-dimensional faceformmAlong corresponding normal orientation, in static three-dimensional face mould
The three-dimensional vertices P of nearest static three-dimensional faceform is searched in the corresponding effective coverage that can cover face's main feature of typen;
S36, then by the three-dimensional vertices P of static three-dimensional faceformnIn three-dimensional vertices PmThrowing in corresponding normal orientation
Shadow is as the corresponding points in three-dimensional face model after fusion.
In a preferred embodiment, the standard three-dimensional faceform includes colorful three-dimensional model file and movement text
Part;The colorful three-dimensional model file includes the three-dimensional vertices P in standard faces modelmTriangle list, normal direction list, line
Reason and texture picture;The motion file includes the three-dimensional vertices P in standard faces modelmBinding relationship and movement with movement
The direction of motion.
In a preferred embodiment, static three-dimensional people is obtained using structural light three-dimensional scanning system described in step S2
Face model the following steps are included:
S21 orients the 2 D vertex in static three-dimensional faceform's image;
S22 orients three-dimensional vertices corresponding with the 2 D vertex in step S21 in static three-dimensional faceform's image
Pn;
S23 utilizes the three-dimensional vertices P in S2nThe shape similarity metric coefficient of weighted calculation static three-dimensional faceform, Yi Jicheng
As model parameter;
S24 using reconstructed coefficients and three-dimensional deformation model sub-unit synthesis three-dimensional face shape and is merged;
S25 extracts three-dimensional face texture using imaging model and facial symmetry.
Human face three-dimensional model automatic generation method provided by the invention with dynamic expression movement, by by standard three-dimensional
Three-dimensional vertices P in faceformmWith the three-dimensional vertices P in static three-dimensional faceformnFusion is corresponded, so that static state three
Tie up the three-dimensional vertices P in faceformnInherit the three-dimensional vertices P in standard three-dimensional faceformmColorful three-dimensional model file and
Motion file, and then three-dimensional face model dynamic expression automatically generates.People provided by the invention with dynamic expression movement
Face three-dimensional model automatic generation method, it is easy to use, producing efficiency is high.
[Detailed description of the invention]
Fig. 1 is the schematic diagram of the human face three-dimensional model automatic generation method provided by the invention with dynamic expression movement.
Fig. 2 is the schematic diagram of standard three-dimensional faceform provided by the invention.
Fig. 3 is the schematic diagram of static three-dimensional faceform provided by the invention.
Fig. 4 is standard three-dimensional faceform provided by the invention and the fused schematic diagram of static three-dimensional faceform.
Fig. 5 is standard three-dimensional faceform shown in Fig. 4 and the fused schematic diagram of static three-dimensional faceform.
[specific embodiment]
It is clear in order to be more clear the purpose of the present invention, technical solution and advantageous effects, below in conjunction with attached drawing and
Specific embodiment, the present invention will be described in further detail.It should be understood that specific implementation described in this specification
Mode is not intended to limit the present invention just for the sake of explaining the present invention.
Fig. 1 to Fig. 4 is please referred to, the present invention provides a kind of human face three-dimensional model side of automatically generating with dynamic expression movement
Method, comprising the following steps:
S1 makes the standard three-dimensional faceform with movement expression, the three-dimensional top in record standard three-dimensional face model
Point Pm, and mark the first key point O in standard three-dimensional faceformm;
Specifically, the prenasale of the standard three-dimensional faceform is the first key point O in step S1m。
The standard three-dimensional faceform includes colorful three-dimensional model file and motion file;The colorful three-dimensional model text
Part includes the three-dimensional vertices P in standard faces modelmTriangle list, normal direction list, texture and texture picture;The movement
File includes the three-dimensional vertices P in standard faces modelmWith the binding relationship of movement and the direction of motion of movement.
S2 obtains static three-dimensional faceform using structural light three-dimensional scanning system, records in static three-dimensional faceform
Three-dimensional vertices Pn, and mark the second key point O in static three-dimensional faceformn;
Specifically, the prenasale of the static three-dimensional faceform is the second key point O in step S2n。
Described in step S2 using structural light three-dimensional scanning system obtain static three-dimensional faceform the following steps are included:
S21 orients the 2 D vertex in static three-dimensional faceform's image;
S22 orients three-dimensional vertices corresponding with the 2 D vertex in step S21 in static three-dimensional faceform's image
Pn;
S23 utilizes the three-dimensional vertices P in S2nThe shape similarity metric coefficient of weighted calculation static three-dimensional faceform, Yi Jicheng
As model parameter;
S24 using reconstructed coefficients and three-dimensional deformation model sub-unit synthesis three-dimensional face shape and is merged;
S25 extracts three-dimensional face texture using imaging model and facial symmetry.
S3, by by the second key point O in static three-dimensional faceformnIt is closed with first in standard three-dimensional faceform
Key point OmIt is corresponding to be overlapped, using Principal Component Analysis by the three-dimensional vertices P in standard faces modelmIt is matched to static three-dimensional face
The three-dimensional vertices P of modelnIn, realize that static three-dimensional faceform is merged with standard three-dimensional faceform.
Specifically, by by the second key point O in static three-dimensional faceform described in step S3nWith standard three-dimensional people
The first key point O in face modelmIt is corresponding be overlapped the following steps are included:
S31, respectively with the first key point O of standard three-dimensional faceformmCentered on point, static three-dimensional faceform
Two key point OnCentered on point, take the three-dimensional vertices P of standard three-dimensional faceform Yu static three-dimensional faceform respectivelym、Pn;
S32, by principal component analysis system, respectively by the three-dimensional top of standard three-dimensional faceform and static three-dimensional faceform
Point Pm、PnIt is reduced to 2 D vertex data;In Two dimensional Distribution, the standard system of standard three-dimensional faceform is Om-XmYmZm, static state three
The standard system for tieing up faceform is On-XnYnZn
Then the standard system of static three-dimensional faceform is O by S33n-XnYnZnBecome by spin matrix R, transfer matrix T
It changes so that On-XnYnZnStandard system with standard three-dimensional faceform is Om-XmYmZmIt is overlapped.
Further, described in step S32 in Two dimensional Distribution, the standard system of standard three-dimensional faceform is Om-
XmYmZm, the standard system of static three-dimensional faceform is On-XnYnZnDetailed process are as follows:
Standard three-dimensional faceform is to cross the first key point O in standard three-dimensional faceform front viewmLine of symmetry be Ym
Axis, excessively corresponding first key point OmAnd vertical YmThe straight line of axis is XmAxis is respectively perpendicular YmAxis, XmThe straight line of axis is ZmAxis is established
The standard system of standard three-dimensional faceform is Om-XmYmZm;
Static three-dimensional faceform is to cross the second key point O in static three-dimensional faceform's front viewnLine of symmetry be Yn
Axis crosses the second key point OnAnd vertical YnThe straight line of axis is XnAxis is respectively perpendicular YnAxis, XnThe straight line of axis is ZnAxis establishes standard
The standard system of three-dimensional face model is On-XnYnZn。
Referring to figure 5., utilize Principal Component Analysis by the three-dimensional vertices P in standard faces model described in step S3m?
It is fitted on the three-dimensional vertices P of static three-dimensional faceformnIn the following steps are included:
S34, static three-dimensional faceform is with the first key point OmFor the center of circle, rmFor the ball internal cutting of the radius of a ball, by static state three
Dimension faceform's is cut into multiple effective coverages that can cover face's main feature;
S35, each three-dimensional vertices P of standard three-dimensional faceformmAlong corresponding normal orientation, in static three-dimensional face mould
The three-dimensional vertices P of nearest static three-dimensional faceform is searched in the corresponding effective coverage that can cover face's main feature of typen;
S36, then by the three-dimensional vertices P of static three-dimensional faceformnIn three-dimensional vertices PmThrowing in corresponding normal orientation
Shadow is as the corresponding points in three-dimensional face model after fusion.
Specifically, the three-dimensional vertices P of standard three-dimensional faceformmIncluding Pm1、Pm2、Pm3Deng static three-dimensional faceform's
Three-dimensional vertices PnIncluding Pn1、Pn2、Pn3Deng the three-dimensional vertices P of standard three-dimensional faceformm1、Pm2、Pm3Deng respectively with static three-dimensional
The three-dimensional vertices P of faceformn1、Pn2、Pn3Deng one-to-one correspondence, and then form the corresponding points in fused three-dimensional face model
Px1、Px2、Px3.Corresponding points P after fusion in three-dimensional face modelx1Inherit Pm1Colorful three-dimensional model file and motion file,
Inherit P again simultaneouslyn1Color texture and three-dimensional appearance, and then realize that dimension faceform dynamic expression automatically generates.
Human face three-dimensional model automatic generation method provided by the invention with dynamic expression movement, by by standard three-dimensional
Three-dimensional vertices P in faceformmWith the three-dimensional vertices P in static three-dimensional faceformnFusion is corresponded, so that static state three
Tie up the three-dimensional vertices P in faceformnInherit the three-dimensional vertices P in standard three-dimensional faceformmColorful three-dimensional model file and
Motion file, and then three-dimensional face model dynamic expression automatically generates.People provided by the invention with dynamic expression movement
Face three-dimensional model automatic generation method, it is easy to use, producing efficiency is high.
The present invention is not only in the description and the implementation described, therefore for the personnel of familiar field
Other advantage and modification is easily implemented, therefore in the essence without departing substantially from universal defined by claim and equivalency range
In the case where mind and range, the present invention is not limited to specific details, representative equipment and diagrams shown and described herein
Example.
Claims (7)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910079081.1A CN109671148B (en) | 2019-01-28 | 2019-01-28 | A method for automatically generating a 3D face model with dynamic facial expressions |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910079081.1A CN109671148B (en) | 2019-01-28 | 2019-01-28 | A method for automatically generating a 3D face model with dynamic facial expressions |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109671148A true CN109671148A (en) | 2019-04-23 |
CN109671148B CN109671148B (en) | 2025-03-07 |
Family
ID=66150860
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910079081.1A Active CN109671148B (en) | 2019-01-28 | 2019-01-28 | A method for automatically generating a 3D face model with dynamic facial expressions |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109671148B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110675475A (en) * | 2019-08-19 | 2020-01-10 | 腾讯科技(深圳)有限公司 | Face model generation method, device, equipment and storage medium |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101149840A (en) * | 2006-09-20 | 2008-03-26 | 清华大学 | A complex facial expression simulation system and its implementation method |
CN102298649A (en) * | 2011-10-09 | 2011-12-28 | 南京大学 | Space trajectory retrieval method of body movement data |
CN102376100A (en) * | 2010-08-20 | 2012-03-14 | 北京盛开互动科技有限公司 | Single-photo-based human face animating method |
CN103268623A (en) * | 2013-06-18 | 2013-08-28 | 西安电子科技大学 | A static facial expression synthesis method based on frequency domain analysis |
CN103854306A (en) * | 2012-12-07 | 2014-06-11 | 山东财经大学 | High-reality dynamic expression modeling method |
US9258550B1 (en) * | 2012-04-08 | 2016-02-09 | Sr2 Group, Llc | System and method for adaptively conformed imaging of work pieces having disparate configuration |
CN107209953A (en) * | 2014-12-23 | 2017-09-26 | 英特尔公司 | Sketch map for rendering 3D model incarnation is selected |
-
2019
- 2019-01-28 CN CN201910079081.1A patent/CN109671148B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101149840A (en) * | 2006-09-20 | 2008-03-26 | 清华大学 | A complex facial expression simulation system and its implementation method |
CN102376100A (en) * | 2010-08-20 | 2012-03-14 | 北京盛开互动科技有限公司 | Single-photo-based human face animating method |
CN102298649A (en) * | 2011-10-09 | 2011-12-28 | 南京大学 | Space trajectory retrieval method of body movement data |
US9258550B1 (en) * | 2012-04-08 | 2016-02-09 | Sr2 Group, Llc | System and method for adaptively conformed imaging of work pieces having disparate configuration |
CN103854306A (en) * | 2012-12-07 | 2014-06-11 | 山东财经大学 | High-reality dynamic expression modeling method |
CN103268623A (en) * | 2013-06-18 | 2013-08-28 | 西安电子科技大学 | A static facial expression synthesis method based on frequency domain analysis |
CN107209953A (en) * | 2014-12-23 | 2017-09-26 | 英特尔公司 | Sketch map for rendering 3D model incarnation is selected |
Non-Patent Citations (1)
Title |
---|
樊少菁: "数字图像视觉真实感若干问题的研究", 中国优秀硕士论文电子期刊网, 15 April 2016 (2016-04-15) * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110675475A (en) * | 2019-08-19 | 2020-01-10 | 腾讯科技(深圳)有限公司 | Face model generation method, device, equipment and storage medium |
CN110675475B (en) * | 2019-08-19 | 2024-02-20 | 腾讯科技(深圳)有限公司 | A face model generation method, device, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN109671148B (en) | 2025-03-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107316340B (en) | A fast face modeling method based on a single photo | |
US6054999A (en) | Method and apparatus for computer supported animation | |
CN111292427B (en) | Bone displacement information acquisition method, device, equipment and storage medium | |
CN109147048A (en) | A kind of three-dimensional grid method for reconstructing using individual cromogram | |
CN106709975B (en) | A kind of interactive three-dimensional facial expression animation edit methods, system and extended method | |
CN104599305B (en) | A kind of two three-dimensional animation producing methods combined | |
CN108876814A (en) | A method of generating posture stream picture | |
CN106710003A (en) | Three-dimensional photographing method and system based on OpenGL ES (Open Graphics Library for Embedded System) | |
EP0427723B1 (en) | Method and device for computerized animation | |
CN110889901B (en) | Large-scene sparse point cloud BA optimization method based on distributed system | |
CN101930618A (en) | Method for producing individual two-dimensional anime | |
CN110544314A (en) | Fusion method, system, medium and device of virtual reality and simulation model | |
Li et al. | Fast simulation of deformable characters with articulated skeletons in projective dynamics | |
Zhang et al. | The Application of Folk Art with Virtual Reality Technology in Visual Communication. | |
CN109671148A (en) | A kind of human face three-dimensional model automatic generation method with dynamic expression movement | |
CN114219001A (en) | A model fusion method and related device | |
Yuan et al. | The Fusion Method of Virtual Reality Technology and 3D Movie Animation Design. | |
Zhang et al. | Face animation making method based on facial motion capture | |
CN102855652B (en) | Method for redirecting and cartooning face expression on basis of radial basis function for geodesic distance | |
Min | Dynamic modeling of interactive scene in 3D animation teaching | |
Yang et al. | Application of augmented reality technology in smart cartoon character design and visual modeling | |
CN113554745B (en) | Three-dimensional face reconstruction method based on image | |
CN109829870B (en) | 3D Regenerated Kernel Space Function Image Synthesis Method | |
CN118967893B (en) | Human body reconstruction and real-time animation rendering system based on 3D Gaussian | |
Wang | Research on the production and application of virtual reality technology based on immersion and interaction in 3D animation scene |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |