CN115661354A - 3D intelligent fitting model system - Google Patents
3D intelligent fitting model system Download PDFInfo
- Publication number
- CN115661354A CN115661354A CN202211383588.4A CN202211383588A CN115661354A CN 115661354 A CN115661354 A CN 115661354A CN 202211383588 A CN202211383588 A CN 202211383588A CN 115661354 A CN115661354 A CN 115661354A
- Authority
- CN
- China
- Prior art keywords
- human body
- model
- dimensional
- fitting
- visual angle
- 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
- 230000036544 posture Effects 0.000 claims abstract description 68
- 230000000007 visual effect Effects 0.000 claims abstract description 65
- 230000009466 transformation Effects 0.000 claims abstract description 18
- 230000000694 effects Effects 0.000 claims abstract description 16
- 239000004744 fabric Substances 0.000 claims description 16
- 238000000034 method Methods 0.000 claims description 13
- 230000002452 interceptive effect Effects 0.000 claims description 9
- 238000013507 mapping Methods 0.000 claims description 7
- 238000001514 detection method Methods 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 238000004088 simulation Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 239000000725 suspension Substances 0.000 description 1
Landscapes
- Processing Or Creating Images (AREA)
Abstract
The invention relates to the technical field of virtual fitting, and discloses a 3D intelligent fitting model system which comprises a 3D fitting display module used for displaying virtual fitting effects of human body models with different visual angles and different postures and configured with a display visual angle posture model library, and a human body-clothing matching module used for generating clothing on the 3D human body model, wherein the 3D fitting display module parameterizes a 3D human body model of a fitting appointed style clothing generated by the human body-clothing matching module and a corresponding two-dimensional visual angle posture image template to a two-dimensional plane, carries out visual angle posture control transformation on the parameterized model in a two-dimensional space through a visual angle posture control point set, restores the model after the visual angle posture control transformation to a three-dimensional curved surface, obtains the 3D human body model of the fitting appointed style clothing under different visual angles and different postures, and obtains beneficial technical contribution of observing the virtual fitting effects of the three-dimensional human body models in multiple visual angles and multiple postures.
Description
Technical Field
The invention relates to the technical field of virtual fitting, in particular to a 3D intelligent fitting model system.
Background
The virtual fitting technology is a technical means which can help a user to change clothes and check the wearing effect without taking off the clothes. The virtual fitting system is based on three-dimensional human body reconstruction and three-dimensional dressing simulation, and the quality of the virtual fitting effect is directly influenced by the selection of a human body data acquisition mode, a human body model construction method, a garment library establishment mode, a fabric simulation method and the like. The human body model and the clothes are completely constructed in a 3D mode, 360-degree fitting effect display can be achieved, fitting effect is more realistic, and certain difference still exists between the fitting effect and real fitting effect due to the limitation of the prior art.
Disclosure of Invention
Technical problem to be solved
The invention provides a 3D intelligent fitting model system capable of observing the virtual fitting effect of a three-dimensional human body model in multiple visual angles and multiple postures.
(II) technical scheme
In order to achieve the purpose, the invention provides the following technical scheme:
A3D intelligent fitting model system comprises a 3D fitting display module and a human body-clothing matching module, wherein the 3D fitting display module is used for displaying virtual fitting effects of human body models with different visual angles and different postures and is provided with a display visual angle posture model library, the human body-clothing matching module is used for generating clothing on the 3D human body models, and the human body-clothing matching module and the 3D fitting display module perform data interaction with each other;
the 3D fitting display module parameterizes a 3D human body model of the fitting appointed style garment generated by the human body-garment matching module and a corresponding two-dimensional view angle posture image template to a two-dimensional plane, performs view angle posture control transformation on the parameterized model in a two-dimensional space through a view angle posture control point set, and restores the model after the view angle posture control transformation to a three-dimensional curved surface to obtain the 3D human body model of the fitting appointed style garment under different view angles and different postures.
Preferably, the specific implementation steps of the method for obtaining the 3D human body model for trying on the specific style of clothing at different viewing angles and different postures are as follows:
step1, selecting a two-dimensional visual angle posture image template from a display visual angle posture model library of a 3D fitting display module, and adjusting parameter information of the two-dimensional visual angle posture image template according to characteristic size data of human body characteristic points of a 3D human body model of the 3D human body model for fitting a specified style garment to obtain the two-dimensional visual angle posture image template corresponding to the 3D human body model for fitting the specified style garment;
step2, acquiring a vertex number set and a patch number set of a 3D human body model for trying on a specified style garment;
analyzing the two-dimensional view attitude image template, and extracting a contour line and a feature point set of a human body model in the two-dimensional view attitude image template;
performing triangular meshing on the two-dimensional visual angle posture image template to obtain a two-dimensional model which has the same vertex number set and surface patch number set as the 3D human body model trying on the specified style clothes;
step3, parameterizing a 3D human body model of the try-on appointed style garment into a two-dimensional plane disc through quasi-conformal mapping, and extracting a characteristic point set of the model;
parameterizing the two-dimensional model to a two-dimensional plane disc to serve as a basis for controlling deformation of the two-dimensional view attitude, and extracting a feature point set of the two-dimensional model;
step4, selecting a visual angle posture control point set from a feature point set of the 3D human body model; selecting a visual angle attitude control point set from a characteristic point set of the two-dimensional model; establishing a mapping relation between the visual angle posture control point set and the visual angle posture control point set;
selecting a visual angle posture control point set of the 3D human body model as a control point set of visual angle posture transformation, and using a visual angle posture control point set of the two-dimensional model as a control point set transformation result;
and establishing a two-dimensional model after the visual angle posture transformation with the minimum error with the two-dimensional model by using a quasi-conformal iterative algorithm, solving a cotangent Laplace matrix of the two-dimensional model, selecting a visual angle posture control point set as a known point for recovering the three-dimensional curved surface, and obtaining the 3D fitting models with different visual angles and different postures.
Preferably, the 3D intelligent fitting model system further comprises a 3D human body modeling and displaying module for constructing and displaying the 3D human body model and providing an interactive interface, and a clothing database for storing clothing images and cloth models.
Preferably, the bounding box of the 3D human body model is created according to the Spheres collision bounding box theory, a cloth model corresponding to the user-selected try-on clothes in the clothes database is called by applying a particle-spring model based on a physical modeling method and the Spheres bounding ball theory in the hierarchical bounding box, collision detection and response processing are performed on the cloth, the clothing of the user-specified style is generated by using the bounding box of the 3D human body model according to cloth initialization information, the cloth is wrapped around the 3D human body model, the physical characteristics of the clothing are simulated, and the 3D human body model of the try-on specified style clothing is obtained.
(III) advantageous technical effects
Compared with the prior art, the invention has the following beneficial technical effects:
according to the method, the 3D human body model of the try-on appointed style garment and the corresponding two-dimensional view angle posture image template are parameterized to the two-dimensional plane, the view angle posture control transformation is carried out on the parameterized model in the two-dimensional space through the view angle posture control point set, the model after the view angle posture control transformation is restored to the three-dimensional curved surface, the 3D human body model of the try-on appointed style garment under different view angles and different postures is obtained, and the beneficial technical contribution of the virtual try-on effect of the multi-view angle multi-posture observation three-dimensional human body model is obtained.
Detailed Description
A 3D intelligent fitting model system, comprising: 3D human body modeling and displaying module M for constructing and displaying 3D human body model and providing interactive interface hmd A clothing database Cd for storing clothing images and cloth models, a body-clothing matching module M for generating clothing on a 3D body model and providing an interactive interface bcm And a 3D fitting display module M for displaying the virtual fitting clothes effect of the human body models with different visual angles and different postures and providing an interactive interface fd The 3D fitting display module M fd Configured with display view angle attitude model library ML dwap ;
Human body-clothing matching module M bcm Respectively modeling with 3D human body and display module M hmd The clothing database Cd and the 3D fitting display module M fd Carrying out data interaction among the data;
wherein, the 3D human body modeling and display module M hmd The work flow of (2) comprises: through the provided interactive interface, the user ui inputs the characteristic dimension data of the human body characteristic points on the interface, the data is matched with each layer of section of the standard human body model after being preprocessed to obtain the value point information of each layer of the human body, and the surface of the human body is modeled by adopting a curved surface modeling methodFitting and splicing to generate a 3D human body model, then carrying out triangular gridding processing on generating points of the 3D human body model, calculating a normal vector of each point, and displaying the BMs-ui of the 3D human body model after adding illumination information of ambient light and refracted light;
human body-clothing matching module M bcm The work flow of (2) comprises: firstly, providing an interactive interface, calling a clothing image in a clothing database Cd to be displayed on the interactive interface and providing the clothing image to a user ui for selection, inputting the clothing to be tried by the user ui, and calling a 3D human body modeling and displaying module M hmd The method comprises the steps of generating a 3D human body model BMs-ui, obtaining body type and posture data of the 3D human body model BMs-ui, creating a bounding box of the 3D human body model BMs-ui according to a Spheres collision bounding box theory, calling a cloth model corresponding to a try-on garment selected by a user ui in a garment database Cd by applying a particle-spring model based on a physical modeling method and a Spheres bounding ball theory in a hierarchical bounding box, performing collision detection and response processing on the cloth, generating the garment of a style specified by the user ui by using the bounding box of the 3D human body model BMs-ui according to cloth initialization information, wrapping the cloth around the 3D human body model BMs-ui, simulating physical characteristics such as drape suspension of the garment, and obtaining the 3D human body model BMs-ui-ssc of the try-on garment of the specified style;
3D fitting display module M fd The work flow of (2) comprises: the user ui inputs the information of the characteristic points of his/her body on the interactive interface and displays the visual angle posture model library ML dwap Selecting a two-dimensional visual angle posture image template to obtain a 3D human body model BMs-ui-ssc-n of the try-on appointed style garment at different visual angles and different postures, and observing the virtual try-on effect of the 3D human body model BMs-ui-ssc of the try-on appointed style garment at different visual angles and different postures;
the specific implementation steps of the method for obtaining the 3D human body model BMs-ui-ssc-n of the try-on appointed style garment under different visual angles and postures are as follows:
step1, acquiring a two-dimensional visual angle posture image template IT of a 3D human body model BMs-ui-ssc 2Dpp-s The method comprises the following specific steps:
fitting display in 3DModule M fd Display view angle posture model library ML dwap In the method, a two-dimensional visual angle posture image template IT is selected 2Dpp ;
Adjusting a two-dimensional visual angle posture image template IT according to characteristic dimension data of human body characteristic points of a 3D human body model BMs-ui of the 3D human body model BMs-ui-ssc trying on a specified style garment 2Dpp Obtaining a two-dimensional view angle posture image template IT corresponding to a 3D human body model BMs-ui-ssc trying on a specified style garment 2Dpp-s ;
Step2, obtaining a two-dimensional model IT with the same vertex number set and surface patch number set as the 3D human body model BMs-ui-ssc for trying on the appointed style clothes 2Dpp-s-a The method comprises the following specific steps:
obtaining a vertex number set V of a 3D human body model BMs-ui-ssc for trying on a specified style garment BMs-ui-ssc-3D And the number set P of patches BMs-ui-ssc-3D ;
To two-dimensional visual angle attitude image template IT 2Dpp-s Analyzing and extracting a two-dimensional visual angle posture image template IT 2Dpp-s Contour line of middle human body model and characteristic point set S IT2Dpp-s ;
To two-dimensional visual angle attitude image template IT 2Dpp-s Performing triangular gridding to obtain a two-dimensional model IT with the same vertex number set and patch number set as the 3D human body model BMs-ui-ssc for trying on the specified style clothes 2Dpp-s-a ;
Step3, according to the quasi-conformal mapping, a 3D human body model BMs-ui-ssc and a two-dimensional model IT are processed 2Dpp-s-a Parameterizing to obtain a two-dimensional plane disc, which comprises the following specific steps:
parameterizing a 3D human body model BMs-ui-ssc for trying on a specified style garment into a two-dimensional plane disc through quasi-conformal mapping, and extracting a characteristic point set S of the parameterized three-dimensional human body model BMs-ui-ssc ;
Two-dimensional model IT 2Dpp-s-a Parameterizing to a two-dimensional plane disc as a basis for controlling deformation of two-dimensional view angle postures, and extracting a two-dimensional model IT 2Dpp-s-a Characteristic point set S of IT2Dpp-s-a ;
Step4, selecting a visual angle attitude control point set of each model in a two-dimensional plane, and carrying out visual angle attitude control transformation operation through the visual angle attitude control point set, wherein the method specifically comprises the following steps:
step4-1, in a feature point set S of the 3D human body model BMs-ui-ssc BMs-ui-ssc In-selection visual angle attitude control point set C SBMs-ui-ssc (ii) a In a two-dimensional model IT 2Dpp-s-a Characteristic point set S of IT2Dpp-s-a In-selection visual angle attitude control point set C SIT2Dpp-s-a ;
Step4-2, establishing a visual angle attitude control point set C SBMs-ui-ssc And view angle attitude control point set C SIT2Dpp-s-a The mapping relation f between BMs-IT ;
Step4-3, selecting a visual angle attitude control point set C of the 3D human body model BMs-ui-ssc SBMs-ui-ssc Two-dimensional model IT as a set of control points for view attitude transformation 2Dpp-s-a View angle attitude control point set C SIT2Dpp-s-a As a result of the control point set transformation;
step4-4, calculating a control point set C at the view angle attitude by using a quasi-conformal iterative algorithm SBMs-ui-ssc As the changes made by other points under the constraint condition, a set C of control points in the view angle attitude is obtained SBMs-ui-ssc Establishing a two-dimensional model IT as coordinates of all points under constraint conditions 2Dpp-s-a The two-dimensional model M-2D with the minimum error after the view angle posture transformation;
step5, obtaining a 3D fitting model BMs-ui-ssc-didaadp with different visual angles and different postures, and specifically comprising the following steps:
solving a cotangent Laplace matrix of the two-dimensional model M-2D, and selecting a visual angle attitude control point set C SBMs-ui-ssc And the known point is used as a known point for recovering the three-dimensional curved surface, the known point is used as a constraint condition, the information of the vertex of the three-dimensional model meeting the constraint condition is solved through a least square method, and the 3D trial model BMs-ui-ssc-didaadp with different visual angles and different postures is obtained.
Claims (4)
1. A3D intelligent fitting model system is characterized by comprising a 3D fitting display module and a human body-clothing matching module, wherein the 3D fitting display module is used for displaying virtual fitting effects of human body models with different visual angles and different postures and is provided with a display visual angle posture model library, the human body-clothing matching module is used for generating clothing on the 3D human body model, and the human body-clothing matching module and the 3D fitting display module perform mutual data interaction;
the three-dimensional (3D) fitting display module parameterizes a 3D human body model of the fitting appointed style clothes generated by the human body-clothes matching module and a corresponding two-dimensional view attitude image template to a two-dimensional plane, performs view attitude control transformation on the parameterized model in a two-dimensional space through a view attitude control point set, restores the model subjected to the view attitude control transformation to a three-dimensional curved surface, and obtains the 3D human body model of the fitting appointed style clothes at different views and different attitudes.
2. The system according to claim 1, wherein the 3D human body model for trying on a specific style of clothes at different viewing angles and different postures is obtained by the following steps:
step1, selecting a two-dimensional visual angle posture image template from a display visual angle posture model library of a 3D fitting display module, and adjusting parameter information of the two-dimensional visual angle posture image template according to characteristic size data of human body characteristic points of a 3D human body model of the 3D human body model for fitting a specified style garment to obtain the two-dimensional visual angle posture image template corresponding to the 3D human body model for fitting the specified style garment;
step2, acquiring a vertex number set and a patch number set of a 3D human body model for trying on a specified style garment;
analyzing the two-dimensional view attitude image template, and extracting a contour line and a feature point set of a human body model in the two-dimensional view attitude image template;
performing triangular meshing on the two-dimensional visual angle posture image template to obtain a two-dimensional model which has the same vertex number set and surface patch number set as the 3D human body model trying on the specified style clothes;
step3, parameterizing a 3D human body model of the try-on appointed style garment into a two-dimensional plane disc through quasi-conformal mapping, and extracting a characteristic point set of the parameterized human body model;
parameterizing the two-dimensional model to a two-dimensional plane disc to serve as a basis for controlling deformation of the two-dimensional view attitude, and extracting a feature point set of the two-dimensional model;
step4, selecting a visual angle posture control point set from a feature point set of the 3D human body model; selecting a visual angle attitude control point set from a feature point set of a two-dimensional model; establishing a mapping relation between a visual angle attitude control point set and a visual angle attitude control point set;
selecting a visual angle posture control point set of the 3D human body model as a control point set of visual angle posture transformation, and taking a visual angle posture control point set of the two-dimensional model as a control point set transformation result;
and establishing a two-dimensional model after the visual angle posture transformation with the minimum error with the two-dimensional model by using a quasi-conformal iterative algorithm, solving a cotangent Laplace matrix of the two-dimensional model, selecting a visual angle posture control point set as a known point for recovering the three-dimensional curved surface, and obtaining the 3D fitting models with different visual angles and different postures.
3. The 3D intelligent fitting model system according to claim 1, further comprising a 3D human modeling and display module for constructing and displaying the 3D human model and providing an interactive interface, a clothing database for storing clothing images and cloth models.
4. The 3D intelligent fitting model system according to claim 3, wherein a bounding box of the 3D human body model is created according to a Spheres collision bounding box theory, a cloth model corresponding to a user-selected fitting garment in a garment database is called by applying a particle-spring model based on a physical modeling method and a Spheres bounding box theory in a hierarchical bounding box, collision detection and response processing are performed on the cloth, a garment of a user-specified style is generated by using the bounding box of the 3D human body model according to cloth initialization information, the cloth is wrapped around the 3D human body model, physical characteristics of the garment are simulated, and the 3D human body model of the fitting specified style garment is obtained.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211383588.4A CN115661354A (en) | 2022-11-07 | 2022-11-07 | 3D intelligent fitting model system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211383588.4A CN115661354A (en) | 2022-11-07 | 2022-11-07 | 3D intelligent fitting model system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115661354A true CN115661354A (en) | 2023-01-31 |
Family
ID=85016034
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211383588.4A Pending CN115661354A (en) | 2022-11-07 | 2022-11-07 | 3D intelligent fitting model system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115661354A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016000304A1 (en) * | 2014-06-30 | 2016-01-07 | 京东方科技集团股份有限公司 | Virtual fitting method and virtual fitting system |
CN105354876A (en) * | 2015-10-20 | 2016-02-24 | 何家颖 | Mobile terminal based real-time 3D fitting method |
CN114119905A (en) * | 2020-08-27 | 2022-03-01 | 北京陌陌信息技术有限公司 | Virtual fitting method, system, equipment and storage medium |
CN114663199A (en) * | 2022-05-17 | 2022-06-24 | 武汉纺织大学 | Dynamic display real-time three-dimensional virtual fitting system and method |
-
2022
- 2022-11-07 CN CN202211383588.4A patent/CN115661354A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016000304A1 (en) * | 2014-06-30 | 2016-01-07 | 京东方科技集团股份有限公司 | Virtual fitting method and virtual fitting system |
CN105354876A (en) * | 2015-10-20 | 2016-02-24 | 何家颖 | Mobile terminal based real-time 3D fitting method |
CN114119905A (en) * | 2020-08-27 | 2022-03-01 | 北京陌陌信息技术有限公司 | Virtual fitting method, system, equipment and storage medium |
CN114663199A (en) * | 2022-05-17 | 2022-06-24 | 武汉纺织大学 | Dynamic display real-time three-dimensional virtual fitting system and method |
Non-Patent Citations (1)
Title |
---|
况立群: "《准共形映射的三维模型变形算法》" * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10922898B2 (en) | Resolving virtual apparel simulation errors | |
US10628666B2 (en) | Cloud server body scan data system | |
US10573091B2 (en) | Systems and methods to create a virtual object or avatar | |
US11640672B2 (en) | Method and system for wireless ultra-low footprint body scanning | |
US11961200B2 (en) | Method and computer program product for producing 3 dimensional model data of a garment | |
CN105354876B (en) | A kind of real-time volume fitting method based on mobile terminal | |
EP2686834B1 (en) | Improved virtual try on simulation service | |
CN114663199A (en) | Dynamic display real-time three-dimensional virtual fitting system and method | |
Magnenat-Thalmann | Modeling and simulating bodies and garments | |
KR20140077820A (en) | Method for fiiting virtual items using human body model and system for providing fitting service of virtual items | |
CN109242954A (en) | Multi-view angle three-dimensional human body reconstruction method based on template deformation | |
CN113610612B (en) | 3D virtual fitting method, system and storage medium | |
US20110148871A1 (en) | Apparatus and method for rendering woven fabric material based on type of weave | |
CN114450719A (en) | Human body model reconstruction method, reconstruction system and storage medium | |
JP2023515517A (en) | Fitting eyeglass frames including live fitting | |
CN110174940A (en) | Type of flight simulator unreal & real space real time integrating method | |
JP2013501284A (en) | Representation of complex and / or deformable objects and virtual fitting of wearable objects | |
EP3997670A1 (en) | Methods of estimating a bare body shape from a concealed scan of the body | |
CN108734772A (en) | High accuracy depth image acquisition methods based on Kinect fusion | |
US20170193677A1 (en) | Apparatus and method for reconstructing experience items | |
Alexiadis et al. | Fast and smooth 3d reconstruction using multiple rgb-depth sensors | |
CN107590858A (en) | Medical sample methods of exhibiting and computer equipment, storage medium based on AR technologies | |
WO2018182938A1 (en) | Method and system for wireless ultra-low footprint body scanning | |
CN112884556A (en) | Shop display method, system, equipment and medium based on mixed reality | |
CN115661354A (en) | 3D intelligent fitting model system |
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: 20230131 |
|
RJ01 | Rejection of invention patent application after publication |