CN111583422B - Heuristic editing method and device for three-dimensional human body model - Google Patents
Heuristic editing method and device for three-dimensional human body model Download PDFInfo
- Publication number
- CN111583422B CN111583422B CN202010304897.2A CN202010304897A CN111583422B CN 111583422 B CN111583422 B CN 111583422B CN 202010304897 A CN202010304897 A CN 202010304897A CN 111583422 B CN111583422 B CN 111583422B
- Authority
- CN
- China
- Prior art keywords
- dimensional
- image
- editing
- human body
- body 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
- G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- 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
Abstract
The invention discloses a heuristic editing method and a device of a three-dimensional human body model, wherein the method comprises the following steps: acquiring a three-dimensional human body model to be edited; converting a three-dimensional human body model to be edited into a geometric image by using a curved surface parameterization method; and the processing geometric image trains the two-dimensional neural network, outputs a generated geometric image, and inverts the generated geometric image into three-dimensional information. The method only needs to select the vertex without defining the sketch plane, thereby getting rid of the limitation of visual angles and simplifying the editing steps.
Description
Technical Field
The invention relates to the technical field of computer vision and computer graphics, in particular to a heuristic editing method and device for a three-dimensional human body model.
Background
Editing of three-dimensional human models is an important issue in the field of computer graphics. However, the editing cost of the three-dimensional human body model is often very high, the traditional three-dimensional model editing method is complicated in steps, needs a professional model animator to complete, and is time-consuming and labor-consuming.
In recent years, the rapid development of deep neural networks provides a solution to many "from scratch" problems. However, three-dimensional neural networks are less mature than two-dimensional neural networks, and therefore, the three-dimensional problem cannot be solved by a neural network-based method as in the two-dimensional problem.
Disclosure of Invention
The present invention is directed to solving, at least to some extent, one of the technical problems in the related art.
It is therefore an object of the present invention to provide a heuristic editing method of a three-dimensional phantom which greatly simplifies the editing of the three-dimensional phantom.
Another object of the present invention is to provide a heuristic editing device for three-dimensional human body models.
In order to achieve the above object, an embodiment of the present invention provides a heuristic editing method for a three-dimensional human body model, including the following steps: s1, acquiring a three-dimensional human body model to be edited; s2, converting the three-dimensional human body model to be edited into a geometric image by using a curved surface parameterization method; and S3, processing the geometric image to train the two-dimensional neural network, outputting a generated geometric image, and inverting the generated geometric image into three-dimensional information.
The heuristic editing method of the three-dimensional human body model of the embodiment of the invention artificially provides a heuristic sketch by utilizing the curved surface parameterization and the deep neural network technology to generate the human body model with corresponding details, thereby greatly simplifying the editing of the three-dimensional human body model.
In addition, the heuristic editing method for the three-dimensional human body model according to the above embodiment of the present invention may further have the following additional technical features:
further, in an embodiment of the present invention, the step S3 includes: extracting line images of the geometric figures through an edge detection algorithm, and randomly generating mask images; performing mask calculation on the geometric image and the mask image to obtain a two-dimensional incomplete image; and training the two-dimensional neural network by using the two-dimensional incomplete image.
Further, in an embodiment of the present invention, the step S1 further obtains an editing area selected by the user.
Further, in an embodiment of the present invention, before obtaining the geometric image, one or more vertex chains that are desired to be deformed are selected in an editing area selected by a user, and a line image, that is, an editing sketch, is generated by using the vertex chains.
Further, in one embodiment of the present invention, a mask image is mapped according to the user-selected edit region.
In order to achieve the above object, another embodiment of the present invention provides a heuristic editing apparatus for a three-dimensional human body model, including: the acquisition module is used for acquiring a three-dimensional human body model to be edited; the conversion module is used for converting the three-dimensional human body model to be edited into a geometric image by using a curved surface parameterization method; and the training module is used for processing the geometric image to train the two-dimensional neural network, outputting a generated geometric image and inverting the generated geometric image into three-dimensional information.
The heuristic editing device for the three-dimensional human body model provided by the embodiment of the invention artificially provides a heuristic sketch by utilizing the curved surface parameterization and deep neural network technology to generate the human body model with corresponding details, thereby greatly simplifying the editing of the three-dimensional human body model.
In addition, the heuristic editing device for three-dimensional human body models according to the above embodiments of the present invention may further have the following additional technical features:
further, in one embodiment of the present invention, the training module comprises: the extraction unit is used for extracting the line image of the geometric figure through an edge detection algorithm and randomly generating a mask image; the calculation module is used for performing mask calculation on the geometric image and the mask image to obtain a two-dimensional incomplete image; and the training unit is used for training the two-dimensional neural network by utilizing the two-dimensional incomplete image.
Further, in an embodiment of the present invention, the obtaining module is further configured to obtain an editing area selected by a user.
Further, in an embodiment of the present invention, before obtaining the geometric image, one or more vertex chains that are desired to be deformed are selected in an editing area selected by a user, and a line image, that is, an editing sketch, is generated by using the vertex chains.
Further, in one embodiment of the present invention, a mask image is mapped according to the user-selected edit region.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a flow diagram of a heuristic editing method of a three-dimensional mannequin according to one embodiment of the present invention;
fig. 2 is a schematic structural diagram of a heuristic editing apparatus for three-dimensional human body models according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.
The following describes a heuristic editing method and apparatus for a three-dimensional human body model according to an embodiment of the present invention with reference to the drawings, and first, a heuristic editing method for a three-dimensional human body model according to an embodiment of the present invention will be described with reference to the drawings.
It should be noted that, the heuristic editing method refers to that an editor provides a "suggestion" for the editor. The results generated by the editor do not necessarily coincide exactly with the "suggestions", but are modified accordingly in the direction of the "suggestions". Namely, the editing process is greatly simplified under the condition of abandoning certain editing accuracy.
FIG. 1 is a flow chart of a heuristic editing method of a three-dimensional mannequin according to one embodiment of the present invention.
As shown in fig. 1, the heuristic editing method of the three-dimensional human body model comprises the following steps:
in step S1, a three-dimensional human body model to be edited is acquired.
Further, step S1 also obtains an editing area selected by the user, wherein the embodiment of the present invention may modify the three-dimensional geometry of the selected editing area according to a sketch input by the user.
In step S2, the three-dimensional human body model to be edited is converted into a geometric image using a surface parameterization method.
It is understood that current three-dimensional neural networks are less mature than two-dimensional neural networks. The three-dimensional information is converted into two-dimensional information which is easy to process by using a curved surface parameterization method, so that the neural network can process the information conveniently. The reversible transformation mode in turn allows information not to be lost too much before and after transformation.
Further, in an embodiment of the present invention, before obtaining the geometric image, one or more vertex chains that are desired to be deformed are selected in the editing area selected by the user, and the line image is generated by using the vertex chains.
It should be noted that, the conventional three-dimensional editing method often needs to select different sketch planes for editing at different viewing angles for many times, and the embodiment of the present invention only needs to automatically generate a line image (i.e., a sketch) by selecting a vertex chain (composed of a plurality of vertices connected in a certain order), and does not need to define a sketch plane, thereby getting rid of the limitation of the viewing angles and simplifying the editing steps.
In step S3, the geometric image is processed to train the two-dimensional neural network, a generated geometric image is output, and the generated geometric image is converted into three-dimensional information.
Specifically, a line image of a geometric figure is extracted through an edge detection algorithm, and a mask image is randomly generated; performing mask calculation on the geometric image and the mask image to obtain a two-dimensional incomplete image; and training the two-dimensional neural network by using the two-dimensional incomplete image. Inputting a three-dimensional human body model to be edited into the trained two-dimensional neural network, outputting and generating a geometric image, and inverting the two-dimensional information into three-dimensional information. Wherein the mask image is randomly mapped according to the editing area selected by the user.
It should be noted that the embodiment of the present invention only encodes the difference value of the three-dimensional coordinates of the human body model (containing the clothes details) relative to the human body model (the naked body). The human body model has overlarge numerical range difference on the xyz three axes, and the numerical range can be unified by only coding the difference, so that more detail information is reserved, and the training of a neural network is facilitated.
The heuristic editing method of the three-dimensional human body model of the present invention is further described below with reference to specific examples.
This specific example converts three-dimensional information into a 512 × 512 × 3 two-dimensional image. The 3 image channels are used to store information in the xyz dimension, respectively. The following steps are divided into a training step and a using step, as the training and the using of the neural network are involved.
Training step 1: the three-dimensional coordinate difference of the phantom (containing the clothing details) in the data set with respect to the phantom (the naked body) is converted into a 512 × 512 × 3 two-dimensional image (hereinafter, referred to as "geometric image") by using a surface parameterization method.
Training step 2: extracting 512 multiplied by 1 line images of the geometric images by using an edge detection algorithm;
training step 3: randomly generating a 512 × 512 × 1 mask image;
training step 4: a two-dimensional neural network is trained, which inputs a 512 × 512 × 3 two-dimensional incomplete image (obtained by masking the geometric image with a mask image) and a 512 × 512 × 1 line image, and outputs a 512 × 512 × 3 generated geometric image. During training, except for L1 loss constraint between the original geometric image and the generated geometric image, the confrontation training technology is also utilized, the original geometric image is used as a positive sample, the generated geometric image is used as a negative sample, and the diversity of generated results is improved.
Using the steps of 1: selecting a three-dimensional human body area needing editing;
using the step 2: selecting one or more vertex chains (consisting of a plurality of vertexes connected in a certain sequence) to be deformed in the editing area;
using the step 3: by utilizing the previously defined surface parameterization mapping, the embodiment of the invention can automatically map the human body model to be edited into a 512 x 3 geometric image, automatically map the editing area using the step 1 into a 512 x 1 mask image, automatically map the vertex chain using the step 2 into a 512 x 1 line image, input the information into a trained neural network and output the information to generate the geometric image;
using the step 4: the embodiment of the invention can automatically invert the geometric image generated by the network into three-dimensional information, thereby obtaining the generated three-dimensional human body model.
According to the heuristic editing method of the three-dimensional human body model provided by the embodiment of the invention, three-dimensional information input by a user is converted into two-dimensional information which is easy to process by a curved surface parameterization method, and the two-dimensional information is converted into three-dimensional information to be displayed after being processed by a two-dimensional deep neural network, so that the complicated three-dimensional editing operation is simplified, the graph is irrelevant to the visual angle, the limitation of the number of lines is avoided, and the wrinkle details of clothes can be generated.
Next, a heuristic editing apparatus for a three-dimensional human body model proposed according to an embodiment of the present invention is described with reference to the drawings.
Fig. 2 is a schematic structural diagram of a heuristic editing apparatus for three-dimensional human body models according to an embodiment of the present invention.
As shown in fig. 2, the apparatus 10 includes: an acquisition module 100, a conversion module 200, and a training module 300.
The obtaining module 100 is configured to obtain a three-dimensional human body model to be edited. The conversion module 200 is used to convert the three-dimensional human body model to be edited into a geometric image by using a surface parameterization method. The training module 300 is configured to process the geometric image to train the two-dimensional neural network, output a generated geometric image, and invert the generated geometric image into three-dimensional information.
Further, in one embodiment of the present invention, the training module 300 comprises: the extraction unit is used for extracting the line image of the geometric figure through an edge detection algorithm and randomly generating a mask image; the calculation module is used for performing mask calculation on the geometric image and the mask image to obtain a two-dimensional incomplete image; and the training unit is used for training the two-dimensional neural network by utilizing the two-dimensional incomplete image.
Further, in an embodiment of the present invention, the obtaining module is further configured to obtain the editing area selected by the user.
Further, in an embodiment of the present invention, before obtaining the geometric image, one or more vertex chains that are desired to be deformed are selected in the editing area selected by the user, and a line image is generated by using the vertex chains (i.e., an editing sketch).
Further, in one embodiment of the present invention, the mask image is mapped according to the edit region selected by the user.
According to the heuristic editing device of the three-dimensional human body model provided by the embodiment of the invention, three-dimensional information input by a user is converted into two-dimensional information which is easy to process by a curved surface parameterization method, and the two-dimensional information is converted into three-dimensional information to be displayed after being processed by a two-dimensional deep neural network, so that the complicated three-dimensional editing operation is simplified, the graph is irrelevant to the visual angle, the limitation of the number of lines is avoided, and the wrinkle details of clothes can be generated.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Although embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are exemplary and not to be construed as limiting the present invention, and that changes, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (8)
1. A heuristic editing method of a three-dimensional human body model is characterized by comprising the following steps:
s1, acquiring a three-dimensional human body model to be edited;
s2, converting the three-dimensional human body model to be edited into a geometric image by using a curved surface parameterization method; and
s3, processing the geometric image to train the two-dimensional neural network, outputting a generated geometric image, inverting the generated geometric image into three-dimensional information,
the step S3 includes:
extracting a line image of the geometric image through an edge detection algorithm, and randomly generating a mask image;
performing mask calculation on the geometric image and the mask image to obtain a two-dimensional incomplete image;
and training the two-dimensional neural network by using the two-dimensional incomplete image.
2. The heuristic editing method of a three-dimensional human model according to claim 1, wherein the step S1 further obtains an editing area selected by a user.
3. A heuristic editing method of a three-dimensional human model as claimed in claim 1, characterized in that before the geometric image is obtained, one or more vertex chains that are desired to be deformed are selected in an editing area selected by a user, and a line image is generated using the vertex chains.
4. A heuristic editing method of a three-dimensional human model according to claim 2, characterized in that a mask image is mapped according to the editing area selected by the user.
5. A heuristic editing apparatus of a three-dimensional human body model, comprising:
the acquisition module is used for acquiring a three-dimensional human body model to be edited;
the conversion module is used for converting the three-dimensional human body model to be edited into a geometric image by using a curved surface parameterization method; and
a training module for processing the geometric image to train the two-dimensional neural network, outputting a generated geometric image, and inverting the generated geometric image into three-dimensional information,
wherein the training module comprises:
the extraction unit is used for extracting the line image of the geometric image through an edge detection algorithm and randomly generating a mask image;
the calculation module is used for performing mask calculation on the geometric image and the mask image to obtain a two-dimensional incomplete image;
and the training unit is used for training the two-dimensional neural network by using the two-dimensional incomplete image.
6. The apparatus for heuristically editing of a three-dimensional human model of claim 5, wherein the obtaining module is further configured to obtain a user-selected editing area.
7. The heuristic editing device of claim 5, wherein before the geometric image is obtained, one or more vertex chains to be deformed are selected in the editing area selected by the user, and the line image is generated by using the vertex chains.
8. The apparatus for heuristically editing of a three-dimensional human model of claim 6, wherein a mask image is mapped according to the user-selected editing area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010304897.2A CN111583422B (en) | 2020-04-17 | 2020-04-17 | Heuristic editing method and device for three-dimensional human body model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010304897.2A CN111583422B (en) | 2020-04-17 | 2020-04-17 | Heuristic editing method and device for three-dimensional human body model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111583422A CN111583422A (en) | 2020-08-25 |
| CN111583422B true CN111583422B (en) | 2023-03-28 |
Family
ID=72113088
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010304897.2A Active CN111583422B (en) | 2020-04-17 | 2020-04-17 | Heuristic editing method and device for three-dimensional human body model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111583422B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102158706A (en) * | 2011-05-19 | 2011-08-17 | 大连民族学院 | Image compressing storage method on the basis of compressing vertex chain code |
| CN109903396A (en) * | 2019-03-20 | 2019-06-18 | 洛阳中科信息产业研究院(中科院计算技术研究所洛阳分所) | A kind of tooth three-dimensional model automatic division method based on surface parameterization |
| CN109902616A (en) * | 2019-02-25 | 2019-06-18 | 清华大学 | Face three-dimensional feature point detecting method and system based on deep learning |
| CN110490235A (en) * | 2019-07-23 | 2019-11-22 | 武汉大学 | A kind of Vehicle Object view prediction towards 2D image and threedimensional model restoration methods and device |
| CN110660123A (en) * | 2018-06-29 | 2020-01-07 | 清华大学 | Three-dimensional CT image reconstruction method and device based on neural network and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8781557B2 (en) * | 1999-08-11 | 2014-07-15 | Osteoplastics, Llc | Producing a three dimensional model of an implant |
-
2020
- 2020-04-17 CN CN202010304897.2A patent/CN111583422B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102158706A (en) * | 2011-05-19 | 2011-08-17 | 大连民族学院 | Image compressing storage method on the basis of compressing vertex chain code |
| CN110660123A (en) * | 2018-06-29 | 2020-01-07 | 清华大学 | Three-dimensional CT image reconstruction method and device based on neural network and storage medium |
| CN109902616A (en) * | 2019-02-25 | 2019-06-18 | 清华大学 | Face three-dimensional feature point detecting method and system based on deep learning |
| CN109903396A (en) * | 2019-03-20 | 2019-06-18 | 洛阳中科信息产业研究院(中科院计算技术研究所洛阳分所) | A kind of tooth three-dimensional model automatic division method based on surface parameterization |
| CN110490235A (en) * | 2019-07-23 | 2019-11-22 | 武汉大学 | A kind of Vehicle Object view prediction towards 2D image and threedimensional model restoration methods and device |
Non-Patent Citations (2)
| Title |
|---|
| 三维点云模型的参数化方法;王丽辉等;《信息与电脑(理论版)》;20130515(第05期);全文 * |
| 曲面参数化定义;鱼然;《https://blog.csdn.net/qq_29698809/article/details/55107040》;20170214;第一页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111583422A (en) | 2020-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7011146B2 (en) | Image processing device, image processing method, image processing program, and teacher data generation method | |
| Zhou et al. | Topology repair of solid models using skeletons | |
| CN104156997B (en) | Quick volume data skeleton extraction method based on rendering | |
| CN113066171B (en) | Face image generation method based on three-dimensional face deformation model | |
| Zhang et al. | Critical regularizations for neural surface reconstruction in the wild | |
| CN111382618A (en) | Illumination detection method, device, equipment and storage medium for face image | |
| CN112734887A (en) | Face mixing-deformation generation method and device based on deep learning | |
| CN112966390B (en) | Method and apparatus for garment processing based on dual three-dimensional distance fields | |
| CN116993926B (en) | Single-view human body three-dimensional reconstruction method | |
| CN111583422B (en) | Heuristic editing method and device for three-dimensional human body model | |
| US7388584B2 (en) | Method and program for determining insides and outsides of boundaries | |
| CN116385667B (en) | Reconstruction method of three-dimensional model, training method and device of texture reconstruction model | |
| CN111476899A (en) | Three-dimensional reconstruction method for dense texture coordinates of human hand based on single-viewpoint RGB camera | |
| CN115496864B (en) | Model construction method, model reconstruction device, electronic equipment and storage medium | |
| CN116310753A (en) | Vectorized skeleton extraction method and system for outdoor scene point cloud data | |
| CN113435273B (en) | Data augmentation method, data augmentation device, electronic device, and storage medium | |
| JP2739447B2 (en) | 3D image generator capable of expressing wrinkles | |
| CN111611997B (en) | Cartoon customized image motion video generation method based on human body action migration | |
| Brett et al. | A Method of 3D Surface Correspondence for Automated Landmark Generation. | |
| JPH07175921A (en) | Density converter | |
| Knyaz et al. | Machine learning for approximating unknown face | |
| Yang et al. | HiLo: Detailed and Robust 3D Clothed Human Reconstruction with High-and Low-Frequency Information of Parametric Models | |
| CN117456099A (en) | Method for generating character rendering model from a group of action sequences based on key frames | |
| CN116363329B (en) | Three-dimensional image generation method and system based on CGAN and LeNet-5 | |
| KR20010073916A (en) | Real time picture mapping device and methode |
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 |