CN103116904A - Two-dimensional feature extraction system and two-dimensional feature extraction method of three-dimensional model - Google Patents
Two-dimensional feature extraction system and two-dimensional feature extraction method of three-dimensional model Download PDFInfo
- Publication number
- CN103116904A CN103116904A CN2012104073958A CN201210407395A CN103116904A CN 103116904 A CN103116904 A CN 103116904A CN 2012104073958 A CN2012104073958 A CN 2012104073958A CN 201210407395 A CN201210407395 A CN 201210407395A CN 103116904 A CN103116904 A CN 103116904A
- Authority
- CN
- China
- Prior art keywords
- dimensional
- dimensional model
- projection
- feature extraction
- 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.)
- Pending
Links
Images
Landscapes
- Processing Or Creating Images (AREA)
Abstract
Provided is a two-dimensional feature extraction system and a two-dimensional feature extraction method of a three-dimensional model. The two-dimensional feature extraction system and the two-dimensional feature extraction method of the three-dimensional model provide an effective aid decision making support in a searching process of the three-dimensional model based on vision. The two-dimensional feature extraction system is composed of four modules including a three-dimensional model modeling module, a visual angle vector calculation module, a three-dimensional model projection module and a three-dimensional model two-dimensional feature extraction module. The two-dimensional feature extraction method is achieved through four processes including modeling of the three-dimensional model, calculation of visual angle vectors, projection of the visual angle vector, and two-dimensional feature extraction of the two-dimensional feature extraction. The two-dimensional feature extraction system is enabled to have the advantages of being short in development period, good in maintainability and good in easy-modification performance. In addition, a user can obtain a two-dimensional projected image of the three-dimensional model by using the two-dimensional feature extraction system, and can obtain two-dimensional features of the three-dimensional model through the two-dimensional projected image, and effective system support is provided for a three-dimensional model searching system based on the vision.
Description
Technical field
The present invention relates to a kind of two dimensional character extraction system and extracting method of three-dimensional model, belong to the multimedia information retrieval technical field.
Background technology
Because the characteristic information that three-dimensional model comprises is very abundant, therefore produced a different feature description of multiple emphasis, as descriptive statistics, topological description, geometric description, vision description, local feature description and compound description etc.And in above-mentioned concentrated character description method, it is the cognition custom that meets human vision most that vision is described, and therefore also becomes one of focus of three-dimensional model search.
The concept that vision is described is first three-dimensional model to be converted to two-dimensional projection image, then utilize the two dimensional image process field proven technique the various features on these projecting planes are described.
Due to the initial gauges of three-dimensional model size, rotation towards and the position all can be different with the difference of model, therefore for effectively model being compared, it is important steps before feature extraction for routine based on the method for searching three-dimension model of many sightingpistons that all models are carried out standardization.
At present general has main member's analytic approach (PCA) and derivative algorithm thereof based on the standardized control method of rotation.Its cardinal principle is exactly to utilize a series of affined transformation to find the inherent coordinate system of object, and this coordinate system is adjusted to the position that overlaps with external unified coordinate system, normal vector principle component analysis (Normal PCA) finds the inherent coordinate system of object by the normal vector of the tri patch of analysis composition model, and continuously principal component analysis (CPCA) is considered as a continuous equation with model surface and carries out conversion, finally realizes same function.But the rotation standardized algorithm can only can reach very high standardization consistance to department pattern, also can guarantee to make all models of each class to be snapped to consistent direction by 100% ground without any a kind of algorithm gets on, this will make its inconsistent model difference more obverse with other models the time larger, finally causes retrieval to make mistakes.
The thought of not considering to rotate standardized control method is to find a kind of retrieval framework that need not to carry out model standardization.This framework is exactly the regular polygon projection scheme.The proper vector of extracting on different sightingpistons is merged as a whole characteristic set as the character representation of this three-dimensional model.Be exactly that the same feature set of three-dimensional model rotates at model and need to have certain stability when angle changes and can produce a requirement when using this framework.Effective ways of head it off are exactly to choose the Feature Descriptor that has unchanged view angle in a kind of certain angle to describe the viewing plane image.
Summary of the invention
Technology of the present invention is dealt with problems: overcome the deficiencies in the prior art, a kind of two dimensional character extraction system and extracting method of three-dimensional model are provided, the user needn't the Consideration of Three-dimensional model the rotation standardization issue, provide a kind of Man Machine Interface flexibly for the layman uses three-dimensional model searching system, strengthened practicality; Reduce system overhead, improved accuracy and the efficient of system.
Technical solution of the present invention: a kind of two dimensional character extraction system and extracting method of three-dimensional model comprise as shown in Figure 1:
The three-dimensional model MBM: the three-dimensional modeling data file to input carries out the modeling of three-dimensional model, adopts the regular dodecahedron modeling pattern.Then with three-dimensional model coordinate standard module, three-dimensional model is carried out the standardization pre-service of translation, rotation, calculate the pretreated three-dimensional model of described standardization with respect to the degree of depth on 20 surfaces of bounding box, rationally be positioned over three-dimensional model in regular dodecahedron;
Visual angle vector calculation module: after the three-dimensional modeling data file of inputting is carried out the regular dodecahedron modeling, utilize the coordinate on three summits of each face of regular dodecahedron, calculate the normal vector of this face, and should standardize by vector, as 20 two-dimensional projection's vectors of three-dimensional model;
The three-dimensional model projection module: 20 projection vectors according to calculating gained, respectively three-dimensional model is carried out two-dimensional projection on each projection vector, obtain 20 width two-dimensional projection image of three-dimensional model, generate and save as the projected image file;
Three-dimensional model two dimensional character extraction module: analyze the 20 width two-dimensional projection image that obtain, extract the SIFT Feature Descriptor of every width image.Metric space extreme point to every width two dimensional image detects, and then locates the key point in extreme point and the direction of key point is distributed, and generates at last the descriptor (SIFT Feature Descriptor) of key point.
Described three-dimensional model MBM implementation procedure is as follows:
The initialization three-dimensional model, three-dimensional model file to input is resolved, with three-dimensional model coordinate standard, three-dimensional model is carried out the standardization pre-service of translation, rotation, calculate the pretreated three-dimensional model of described standardization with respect to the degree of depth on 20 surfaces of bounding box, rationally be positioned over three-dimensional model in regular dodecahedron;
Described visual angle vector calculation module implementation procedure is as follows:
For the coordinate on three summits that utilize each face of regular dodecahedron, calculate the visual angle vector of this face, establish A
i(x
i1, y
i1, z
i1), B
i(x
i2, y
i2, z
i2), C
i(x
i3, y
i3, z
i3) be three apex coordinates of regular dodecahedron,
Be the visual angle vector of the face in regular dodecahedron, standardizing factor is made as k,
Computing method as follows:
Then right
Standardize, computing method are as follows:
With the standardization after
Visual angle vector as regular dodecahedron.
A kind of two dimensional character extracting method performing step of three-dimensional model is as follows:
(1) the three-dimensional modeling data file of input carried out the modeling of three-dimensional model, adopt the regular dodecahedron modeling pattern.Then with three-dimensional model coordinate standard module, three-dimensional model is carried out the standardization pre-service of translation, rotation, calculate the pretreated three-dimensional model of described standardization with respect to the degree of depth on 20 surfaces of bounding box, rationally be positioned over three-dimensional model in regular dodecahedron;
(2) the three-dimensional modeling data file of input is carried out the regular dodecahedron modeling after, utilize the coordinate on three summits of each face of regular dodecahedron, calculate the normal vector of this face, and should the vector standardization, as 20 two-dimensional projection's vectors of three-dimensional model
(3) according to 20 projection vectors that calculate gained, respectively three-dimensional model is carried out two-dimensional projection, obtain 20 width two-dimensional projection image of three-dimensional model, generate and save as the projected image file;
(4) analyze the 20 width two-dimensional projection image that obtain, extract the SIFT Feature Descriptor of every width image.Metric space extreme point to every width two dimensional image detects, and then locates the key point in extreme point and the direction of key point is distributed, and generates at last the descriptor (SIFT Feature Descriptor) of key point.
The present invention's advantage compared with prior art is:
(1) because the present invention has used the regular dodecahedron projection scheme, the rotation standardization issue of user during without the Consideration of Three-dimensional model projection just can get result two-dimensional projection image preferably;
(2) because the regular dodecahedron projection scheme in the present invention is compared with other projection scheme.Be best in efficient with above effect, can guarantee that the repetition rate of SIFT Feature Descriptor is stabilized in more than 80%;
(3) the present invention only needs the user to input the three-dimensional model file, then just can obtain the two dimensional character file of this three-dimensional model, and is simple to operate, user friendly.
Description of drawings
Fig. 1 is the system assumption diagram of system of the present invention;
Fig. 2 is the three-dimensional model MBM implementation procedure in system of the present invention;
Fig. 3 is the visual angle vector calculation module implementation procedure in system of the present invention.
Embodiment
The present invention is described in detail below in conjunction with accompanying drawing.
As shown in Figure 1, the two dimensional character extraction system of a kind of three-dimensional model of the present invention is made of three-dimensional model MBM, visual angle vector calculation module, three-dimensional model projection module and three-dimensional model two dimensional character extraction module.Three-dimensional model MBM, visual angle vector calculation module and three-dimensional model projection module belong to client modules, and the main function of client modules is that three-dimensional model is carried out projection, gathers and preserve two-dimensional projection's picture; The two dimensional character extraction module belongs to the server end module, and the main function of service end module is to provide the method for extracting the picture SIFT of two-dimensional projection feature, and two-dimensional projection's picture that client is obtained carries out the SIFT feature extraction, generates the SIFT tag file.
Whole implementation procedure is as follows:
(1) the three-dimensional modeling data file of input carried out the modeling of three-dimensional model, adopt the regular dodecahedron modeling pattern.Then with three-dimensional model coordinate standard module, three-dimensional model is carried out the standardization pre-service of translation, rotation, calculate the pretreated three-dimensional model of described standardization with respect to the degree of depth on 20 surfaces of bounding box, rationally be positioned over three-dimensional model in regular dodecahedron;
(2) the three-dimensional modeling data file of input is carried out the regular dodecahedron modeling after, utilize the coordinate on three summits of each face of regular dodecahedron, calculate the normal vector of this face, and should the vector standardization, as 20 two-dimensional projection's vectors of three-dimensional model;
(3) according to 20 projection vectors that calculate gained, respectively three-dimensional model is carried out two-dimensional projection, obtain 20 width two-dimensional projection image of three-dimensional model, generate and save as the projected image file;
(4) analyze the 20 width two-dimensional projection image that obtain, extract the SIFT Feature Descriptor of every width image.Metric space extreme point to every width two dimensional image detects, and then locates the key point in extreme point and the direction of key point is distributed, and generates at last the descriptor (SIFT Feature Descriptor) of key point.
The specific implementation process of above-mentioned each module is as follows:
1. three-dimensional model MBM
This Model Implement process is as shown in Figure 2:
(1) hardship is shielded a shortcoming or fault and is read the three-dimensional model file again, reads by row;
(2) successfully the explanation analysis is complete if do not read, execution in step (4);
(3) the three-dimensional model coordinate is standardized;
(4) three-dimensional model is carried out the regular dodecahedron modeling, calculate the pretreated three-dimensional model of described standardization with respect to the degree of depth on 20 surfaces of bounding box, rationally be positioned over three-dimensional model in regular dodecahedron;
(5) analysis is completed, and finishes.
2. visual angle vector calculation module
This Model Implement process is as shown in Figure 3:
(1) read the coordinate A of point of each face of regular dodecahedron in client
i(x
i1, y
i1, z
i1), B
i(x
i2, y
i2, z
i2), C
i(x
i3, y
i3, z
i3);
(2) according to the normal vector of this face of coordinate Calculation of each face.If
Be the visual angle vector of the face in regular dodecahedron, the computing method of visual angle vector are as follows:
(3) with the normal vector standardization, establishing standardizing factor is k, and normalization method is as follows:
With the standardization after
As the obverse two-dimensional projection's vector of regular dodecahedron;
(4) the visual angle vector calculation is completed, and finishes.
3. three-dimensional model projection module
(1) at two ten projection vectors of client according to the calculating gained, respectively three-dimensional model is carried out two-dimensional projection;
(2) obtain 20 width two-dimensional projection image of three-dimensional model.
4. three-dimensional model two dimensional character extraction module
(1) read 20 width two-dimensional projection image at server end;
(2) utilize the SIFT algorithm to extract the two dimensional character of three-dimensional model, metric space extreme point to every width two dimensional image detects, then locate the key point in extreme point and the direction of key point is distributed, generating at last the descriptor (SIFT Feature Descriptor) of key point.
One of ordinary skill in the art will appreciate that all or part of flow process that realizes in above-described embodiment method, to come the relevant hardware of instruction to complete by computer program, described program can be stored in a computer read/write memory medium, this program can comprise the flow process as the embodiment of above-mentioned each side method when carrying out.Wherein, described storage medium can be magnetic disc, CD, read-only store-memory body (Read-Only Memory, ROM) or random store-memory body (Random Access Memory, RAM) etc.
The part that the present invention does not describe in detail belongs to techniques well known.
Claims (3)
1. the two dimensional character extraction system of a three-dimensional model is characterized in that comprising:
The three-dimensional model MBM: the three-dimensional modeling data file to input carries out the modeling of three-dimensional model, adopt the regular dodecahedron modeling pattern, obtain three-dimensional model, again three-dimensional model is carried out the standardization pre-service of translation, rotation, calculate the pretreated three-dimensional model of described standardization with respect to the degree of depth on 20 surfaces of bounding box, three-dimensional model is positioned in regular dodecahedron;
Visual angle vector calculation module: to the three-dimensional model of input, utilize the coordinate on three summits of each face of regular dodecahedron, calculate the normal vector of this face, and with this normal vector standardization, as 20 two-dimensional projection's vectors of three-dimensional model;
Three-dimensional model projection module: 20 projection vectors that calculate gained according to visual angle vector calculation module, respectively three-dimensional model is carried out two-dimensional projection on each projection vector, obtain 20 width two-dimensional projection image of three-dimensional model, generate and save as the projected image file;
Three-dimensional model two dimensional character extraction module: analyze the 20 width two-dimensional projection image that obtain, extract the SIFT Feature Descriptor of every width image, namely the metric space extreme point of every width two dimensional image detected, then locate the key point in extreme point and the direction of described key point is distributed, generate at last the descriptor of key point, i.e. the SIFT Feature Descriptor.
2. the two dimensional character extraction system of a kind of three-dimensional model according to claim 1, it is characterized in that: described visual angle vector calculation module implementation procedure is as follows:
For the coordinate on three summits that utilize each face of regular dodecahedron, calculate the visual angle vector of this face,
If A
i(x
i1, y
i1, z
i1), B
i(x
i2, y
i2, z
i2), C
i(x
i3, y
i3, z
i3) be three apex coordinates of regular dodecahedron,
Be the visual angle vector of the face in regular dodecahedron, standardizing factor is made as k,
Computing method as follows:
Then right
Standardize, computing method are as follows:
With the standardization after
As the obverse two-dimensional projection's vector of regular dodecahedron.
3. the two dimensional character extracting method of a three-dimensional model is characterized in that step is as follows:
(1) the three-dimensional modeling data file of input carried out the modeling of three-dimensional model, adopt the regular dodecahedron modeling pattern, then with three-dimensional model coordinate standard module, three-dimensional model is carried out the standardization pre-service of translation, rotation, calculate the pretreated three-dimensional model of described standardization with respect to the degree of depth on 20 surfaces of bounding box, rationally be positioned over three-dimensional model in regular dodecahedron;
(2) the three-dimensional modeling data file of input is carried out the regular dodecahedron modeling after, utilize the coordinate on three summits of each face of regular dodecahedron, calculate the normal vector of this face, and should the vector standardization, as 20 two-dimensional projection's vectors of three-dimensional model;
(3) according to 20 projection vectors that calculate gained, respectively three-dimensional model is carried out two-dimensional projection, obtain 20 width two-dimensional projection image of three-dimensional model, generate and save as the projected image file;
(4) analyze the 20 width two-dimensional projection image that obtain, extract the SIFT Feature Descriptor of every width image, namely the metric space extreme point of every width two dimensional image detected, then locate the key point in extreme point and the direction of key point is distributed, generating at last the descriptor (SIFT Feature Descriptor) of key point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012104073958A CN103116904A (en) | 2012-10-23 | 2012-10-23 | Two-dimensional feature extraction system and two-dimensional feature extraction method of three-dimensional model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012104073958A CN103116904A (en) | 2012-10-23 | 2012-10-23 | Two-dimensional feature extraction system and two-dimensional feature extraction method of three-dimensional model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103116904A true CN103116904A (en) | 2013-05-22 |
Family
ID=48415268
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012104073958A Pending CN103116904A (en) | 2012-10-23 | 2012-10-23 | Two-dimensional feature extraction system and two-dimensional feature extraction method of three-dimensional model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103116904A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056551A (en) * | 2016-05-27 | 2016-10-26 | 西华大学 | Local similarity sample learning-based sparse de-noising method |
| CN107146287A (en) * | 2017-03-22 | 2017-09-08 | 西北大学 | Two-dimensional projection image to threedimensional model mapping method |
| CN107610105A (en) * | 2017-08-31 | 2018-01-19 | 沈阳东软医疗系统有限公司 | Position ROI method, apparatus, equipment and machinable medium |
| CN110115026A (en) * | 2016-12-19 | 2019-08-09 | 三星电子株式会社 | The method and system of 360 degree of contents is generated on rectangular projection in an electronic |
| CN110648396A (en) * | 2019-09-17 | 2020-01-03 | 西安万像电子科技有限公司 | Image processing method, device and system |
| WO2021012523A1 (en) * | 2019-07-19 | 2021-01-28 | 平安科技（深圳）有限公司 | Three-dimensional image search method and system, computer device, and storage medium |
| CN113744404A (en) * | 2021-07-21 | 2021-12-03 | 合肥泰瑞数创科技有限公司 | Three-dimensional model comparison processing method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101299218A (en) * | 2008-06-26 | 2008-11-05 | 覃征 | Method and device for searching three-dimensional model |
| CN101350016A (en) * | 2007-07-20 | 2009-01-21 | 富士通株式会社 | Device and method for searching three-dimensional model |
| CN102375831A (en) * | 2010-08-13 | 2012-03-14 | 富士通株式会社 | Three-dimensional model search device and method thereof and model base generation device and method thereof |
| CN102663087A (en) * | 2012-04-09 | 2012-09-12 | 北京邮电大学 | Three-dimensional model search method based on topology and visual feature |
-
2012
- 2012-10-23 CN CN2012104073958A patent/CN103116904A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101350016A (en) * | 2007-07-20 | 2009-01-21 | 富士通株式会社 | Device and method for searching three-dimensional model |
| CN101299218A (en) * | 2008-06-26 | 2008-11-05 | 覃征 | Method and device for searching three-dimensional model |
| CN102375831A (en) * | 2010-08-13 | 2012-03-14 | 富士通株式会社 | Three-dimensional model search device and method thereof and model base generation device and method thereof |
| CN102663087A (en) * | 2012-04-09 | 2012-09-12 | 北京邮电大学 | Three-dimensional model search method based on topology and visual feature |
Non-Patent Citations (2)
| Title |
|---|
| CHEN. C ET AL.: "A Stable Viewing Strategy for Rotation Normalization Free 3D Model Retrieval", 《2011 3RD INTERNATIONAL CONFERENCE ON ENVIRONMENTAL SCIENCE AND INFORMATION APPLICATION TECHNOLOGY ESIAT 2011》 * |
| 江秋鑫: "基于SIFT特征的图像相似性度量及其应用研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056551A (en) * | 2016-05-27 | 2016-10-26 | 西华大学 | Local similarity sample learning-based sparse de-noising method |
| CN110115026A (en) * | 2016-12-19 | 2019-08-09 | 三星电子株式会社 | The method and system of 360 degree of contents is generated on rectangular projection in an electronic |
| CN110115026B (en) * | 2016-12-19 | 2021-07-13 | 三星电子株式会社 | Method and system for generating 360-degree content on rectangular projection in electronic device |
| CN107146287A (en) * | 2017-03-22 | 2017-09-08 | 西北大学 | Two-dimensional projection image to threedimensional model mapping method |
| CN107146287B (en) * | 2017-03-22 | 2019-08-02 | 西北大学 | Two-dimensional projection image to threedimensional model mapping method |
| CN107610105A (en) * | 2017-08-31 | 2018-01-19 | 沈阳东软医疗系统有限公司 | Position ROI method, apparatus, equipment and machinable medium |
| CN107610105B (en) * | 2017-08-31 | 2020-12-11 | 东软医疗系统股份有限公司 | Method, device and equipment for positioning ROI and machine-readable storage medium |
| WO2021012523A1 (en) * | 2019-07-19 | 2021-01-28 | 平安科技（深圳）有限公司 | Three-dimensional image search method and system, computer device, and storage medium |
| CN110648396A (en) * | 2019-09-17 | 2020-01-03 | 西安万像电子科技有限公司 | Image processing method, device and system |
| CN113744404A (en) * | 2021-07-21 | 2021-12-03 | 合肥泰瑞数创科技有限公司 | Three-dimensional model comparison processing method and system |
| CN113744404B (en) * | 2021-07-21 | 2023-09-08 | 合肥泰瑞数创科技有限公司 | Comparison processing method and system of three-dimensional model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103116904A (en) | Two-dimensional feature extraction system and two-dimensional feature extraction method of three-dimensional model | |
| EP3096263B1 (en) | Human body orientation recognition method and system based on two-lens camera | |
| Shao et al. | An interactive approach to semantic modeling of indoor scenes with an rgbd camera | |
| Wang et al. | Efficient volume exploration using the gaussian mixture model | |
| Ngo et al. | Template-based monocular 3D shape recovery using laplacian meshes | |
| Kwon et al. | A geometric particle filter for template-based visual tracking | |
| Demir et al. | Skelneton 2019: Dataset and challenge on deep learning for geometric shape understanding | |
| Xu et al. | Back to reality: Weakly-supervised 3d object detection with shape-guided label enhancement | |
| CN108171133A (en) | A kind of dynamic gesture identification method of feature based covariance matrix | |
| Manni et al. | Snap2cad: 3D indoor environment reconstruction for AR/VR applications using a smartphone device | |
| CN103942832A (en) | Real-time indoor scene reconstruction method based on on-line structure analysis | |
| Wang et al. | Reconstructing action-conditioned human-object interactions using commonsense knowledge priors | |
| Yang et al. | Spatio-temporal reconstruction for 3D motion recovery | |
| CN103745209A (en) | Human face identification method and system | |
| CN111209811A (en) | Method and system for detecting eyeball attention position in real time | |
| Kao et al. | Toward 3d face reconstruction in perspective projection: Estimating 6dof face pose from monocular image | |
| Huang et al. | Automatic craniofacial registration based on radial curves | |
| CN104680135A (en) | Three-dimensional human face mark point detection method capable of resisting expression, posture and shielding changes | |
| Shi et al. | Geometric granularity aware pixel-to-mesh | |
| Lan et al. | The application of 3D morphable model (3DMM) for real-time visualization of acupoints on a smartphone | |
| Zhang et al. | 3D human body skeleton extraction from consecutive surfaces using a spatial–temporal consistency model | |
| Wang et al. | Swimmer’s posture recognition and correction method based on embedded depth image skeleton tracking | |
| Kazmi et al. | Efficient sketch‐based creation of detailed character models through data‐driven mesh deformations | |
| Fan et al. | A high-precision correction method in non-rigid 3D motion poses reconstruction | |
| Ding et al. | 3D face sparse reconstruction based on local linear fitting |
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 |
Application publication date: 20130522 |