CN115601502A - Human body model construction method, device, equipment and storage medium - Google Patents
Human body model construction method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN115601502A CN115601502A CN202211275952.5A CN202211275952A CN115601502A CN 115601502 A CN115601502 A CN 115601502A CN 202211275952 A CN202211275952 A CN 202211275952A CN 115601502 A CN115601502 A CN 115601502A
- Authority
- CN
- China
- Prior art keywords
- human body
- information
- body model
- dimensional
- image
- 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
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
Abstract
The embodiment of the invention provides a human body model construction method, a human body model construction device, human body model equipment and a storage medium, wherein the method comprises the following steps: obtaining an initial parameterized human body model based on the image to be processed; determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed; and determining a target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model. The human body model construction method, the human body model construction device, the human body model construction equipment and the storage medium solve the depth ambiguity problem in the prior art, and effectively improve the reconstruction precision of the parameterized human body model.
Description
Technical Field
The present invention relates to the field of image processing technologies, and in particular, to a method, an apparatus, a device, and a storage medium for constructing a human body model.
Background
Three-dimensional human body modeling techniques are diverse in variety, including reconstructing a three-dimensional digital human body model of a human body from a human body image, and have been widely applied to aspects such as human body motion migration, human body motion recognition, and control of virtual characters in scenes such as games.
At present, modeling can be performed by a parameterized human reconstruction method, and common parameterized human models include SCAPE (Shape Completion and Animation of pendant), SMPL (A Skinned Multi-Person Linear Model), SMPL-X (A Skinned Multi-Person Linear Model-eXpress), and the like.
With the popularization and development of new equipment and new technology, some new parameterized human body model reconstruction methods emerge in recent years, and the method takes a common parameterized human body model as a standard human body model and obtains other data such as two-dimensional human body joint point coordinates, human body contours, human body description parameters and the like by using a single RGB image so as to optimize the standard human body model, thereby improving the accuracy of reconstructing the human body model.
However, the human body model reconstructed by the above method often only uses two-dimensional monitoring signals, which is prone to depth ambiguity and other problems, and the reconstruction accuracy is still low.
Disclosure of Invention
The invention provides a human body model construction method, a human body model construction device, human body model equipment and a storage medium, which are used for overcoming the defect of low accuracy of a human body model in the prior art and achieving the purpose of improving the accuracy of the human body model.
The invention provides a method for constructing a human body model, which comprises the following steps:
obtaining an initial parameterized human body model based on the image to be processed;
determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed;
and determining a target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model.
According to the method for constructing the human body model, the method for determining the sparse three-dimensional attitude information and the dense three-dimensional geometric information based on the image to be processed comprises the following steps:
determining human body joint point information and human body part information based on the image to be processed;
determining the sparse three-dimensional pose information based on the human body joint point information, and determining the dense three-dimensional geometric information based on the human body component information.
According to the method for constructing a human body model provided by the invention, the determining the sparse three-dimensional posture information based on the human body joint point information and the dense three-dimensional geometric information based on the human body component information comprises the following steps:
respectively projecting the human body joint point information and the human body component information to a three-dimensional space based on the depth information in the image to be processed to obtain the sparse three-dimensional posture information and the dense three-dimensional geometric information.
According to the method for constructing the human body model, the step of determining the target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model comprises the following steps:
and optimizing the parameters of the initial parameterized human body model in an iterative mode based on the sparse three-dimensional attitude information and the dense three-dimensional geometric information to obtain the target human body model.
According to the construction method of the human body model provided by the invention, the image to be processed comprises an RGBD image.
The invention also provides a human body model constructing device, which comprises:
the initialization module is used for obtaining an initial parameterized human body model based on the image to be processed;
the determining module is used for determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed;
the determining module is further used for determining a target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model.
According to the apparatus for constructing a human body model provided by the present invention, the determining module is specifically configured to:
determining human body joint point information and human body part information based on the image to be processed;
and determining sparse three-dimensional posture information based on the human body joint point information, and determining dense three-dimensional geometric information based on the human body component information.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein when the processor executes the program, the processor realizes the human body model building method.
The invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method of constructing a mannequin as described in any of the above.
The invention also provides a computer program product comprising a computer program which, when executed by a processor, implements a method of constructing a mannequin as described in any one of the above.
According to the human body model construction method, the device, the equipment and the storage medium provided by the embodiment of the invention, the initial parameterized human body model is obtained through the image to be processed, on the basis, the sparse three-dimensional attitude information and the dense three-dimensional geometric information are obtained through the image to be processed, and the initial parameterized human body model is optimized based on the sparse three-dimensional attitude information and the dense three-dimensional geometric information, so that the problem of depth ambiguity in the initialized parameterized human body model is further solved, and the reconstruction precision of the parameterized human body model is effectively improved.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a method for constructing a human body model according to an embodiment of the present invention;
FIG. 2 is a second flowchart illustrating a method for constructing a human body model according to an embodiment of the present invention;
FIG. 3 is a schematic structural diagram of an apparatus for constructing a human body model according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In recent years, with the rapid development of computer technology, human body shape reconstruction methods have been widely used in animation, manufacturing, and medical fields, for example: the method based on human body type reconstruction can be used for virtual image design and garment digital design, and the three-dimensional human body model formed based on the method can be used for visually verifying products to be produced, so that the design efficiency is greatly improved. In this process, the human body shape reconstruction method plays a great role.
At present, human body model construction methods are roughly divided into two types, one is a non-parametric human body model construction method, and the other is a parametric human body model construction method. The construction of the unparameterized human body model generally needs special data acquisition equipment, such as a laser scanner, a depth camera and the like, for example, a commercial scanner, vironic, an object to be scanned is placed on a scanner platform, then 4 laser scanning probes move from top to bottom at a high speed to scan the whole object, the laser scanning probes obtain local point clouds under a single visual angle, and a three-dimensional grid can be directly reconstructed by matching software within a few seconds after the scanning is completed. The three-dimensional scanner can obtain a static three-dimensional object model quickly and accurately, but the laser scanner is expensive and large in size.
The construction method of the parameterized human body model well solves the problems that the acquisition equipment is high in cost and large in size, namely the human body model can be reconstructed through consumer-grade acquisition equipment, but depth information is not considered in the method, so that the problem of depth ambiguity is easily generated in the human body reconstruction process, and the accuracy of the reconstructed human body model is still low. For example, if a user has a knee bending motion at a small angle, the knee bending motion at the small angle cannot be reflected in the image due to the relationship between the image capturing angles, and thus the reconstructed human model cannot reflect the knee bending motion at the small angle, so that the accuracy of the reconstructed human model is low.
In order to solve the problem, the embodiment of the invention provides a method for constructing a human body model, which is characterized in that after an initial parameterized human body model is obtained by adopting an image to be processed, on the basis of the initial parameterized human body model, depth information contained in sparse three-dimensional attitude information and dense three-dimensional geometric information is fully utilized to further optimize the initial parameterized human body model, so that the reconstruction accuracy of a parameterized human body is effectively improved.
The human body model construction method provided in the embodiment of the present invention is described below with reference to fig. 1, where an execution subject of the method may be a terminal device, a server, or any device capable of reconstructing a human body model.
Fig. 1 is a schematic flow diagram of a method for constructing a human body model according to an embodiment of the present invention, and as shown in fig. 1, the method includes:
step 101: and obtaining an initial parameterized human body model based on the image to be processed.
The image to be processed includes a human body image including a human body, which is acquired by an image acquisition device or locally stored, where the human body image may be a whole body image of the human body or a half body image of the human body, and this is not particularly limited.
In this step, an initial parameterized human body model is obtained based on the image to be processed, and the shape of the human body in the standard human body model may be constrained and optimized by using key information in the image to be processed, such as human body contour coordinates, to adjust the standard human body model to obtain a new human body model, so that the new human body model approaches to the human body in the image to be processed. It should be noted that the standard human body model refers to a designed average parameterized human body model, and may include, for example, SCAPE, SMPL-X, and the like, and the specific type of the standard human body model is not specifically limited in the embodiment of the present invention.
Step 102: and determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed.
In this step, after the initial parameterized human body model is obtained, the three-dimensional attitude information and the three-dimensional geometric information of the current position of the human body are also required to be obtained so as to constrain and optimize the initial parameterized human body model, thereby obtaining a more accurate human body model.
The sparse three-dimensional posture information and the dense three-dimensional geometric information respectively represent human body three-dimensional posture information and human body three-dimensional geometric information. Illustratively, the sparse three-dimensional pose information may be body pose information characterized by a small number of key body three-dimensional pose features, such as: when a person stands on one foot, the three-dimensional coordinates of the knee and the foot of the other leg are a small amount of key human body three-dimensional posture characteristics. Similarly, the dense three-dimensional geometric information may be geometric information of the human body represented by a large number of geometric features of the human body, such as a curved arm, a tilted leg of the two langerhans, and a head or a neck deviated to one side, which are the geometric features of the human body describing the geometric information of the human body.
In addition, when the sparse three-dimensional attitude information and the dense three-dimensional geometric information are determined, the sparse three-dimensional attitude information and the dense three-dimensional geometric information can be obtained through a plurality of images to be processed which are obtained from different positions, and the sparse three-dimensional attitude information and the dense three-dimensional geometric information can also be directly obtained through the images to be processed. Specifically, the manner of acquiring the sparse three-dimensional pose information and the dense three-dimensional geometric information is not specifically limited herein.
Step 103: and determining a target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model.
After the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model are obtained in steps 101 and 102, the sparse three-dimensional attitude information and the dense three-dimensional geometric information can be used for constraining and optimizing the initial parameterized human body model so as to further determine the target human body model.
Specifically, constraint optimization is performed on the initial parameterized human body model through sparse three-dimensional attitude information and dense three-dimensional geometric information, simultaneous constraint optimization can be performed, separate constraint optimization can be performed, and the priority of the two constraint optimizations is not specifically limited.
According to the method for constructing the human body model, the initial parameterized human body model is obtained through the image to be processed, on the basis, sparse three-dimensional attitude information and dense three-dimensional geometric information containing depth information are obtained through the image to be processed, the initial parameterized human body model is jointly optimized based on the sparse three-dimensional attitude information and the dense three-dimensional geometric information, the problem of depth ambiguity in the initialized parameterized human body model is further solved, and therefore the reconstruction accuracy of the parameterized human body model is effectively improved.
Alternatively, the image to be processed in the above embodiment may be an RGBD image.
It should be understood that the RGBD image refers to two images, i.e., a common RGB image and a D image, and it should be noted that the joint information and the component segmentation information of the human body obtained by using the RGB image can fully use the camera projection principle, and determine sparse three-dimensional posture information and dense three-dimensional geometric information based on the D information, so that after the initial parameterized human body model is optimized based on the two pieces of information, a more accurate target human body model can be obtained. For example, when the user has a knee bending motion at a small angle, the knee bending motion can be reflected by the depth information, so that when the initial parameterized human model is optimized based on the depth information, the knee bending motion of the user can be reflected in the optimized target human model, and the accuracy of the target human model can be improved.
On the basis of any of the above embodiments, determining sparse three-dimensional pose information and dense three-dimensional geometric information based on an image to be processed may include:
determining human body joint point information and human body part information based on the image to be processed; and determining sparse three-dimensional posture information based on the human body joint point information, and determining dense three-dimensional geometric information based on the human body component information.
The human body joint points comprise knee joints, elbow joints, wrist joints and the like, and the human body parts comprise limbs, a trunk, a head and the like.
In this step, a mature two-dimensional human body posture estimation method opencast and a human body part segmentation method densepose can be used for effectively extracting human body joint point information and human body part information on an image to be processed, a two-dimensional human body posture estimation method and a human body part segmentation method with better effect and speed can be selected to replace the opencast and the densepose, and a specific method can be selected according to actual use requirements, wherein the two-dimensional human body posture estimation method is not specifically limited.
Furthermore, the sparse three-dimensional attitude information can be determined based on the determined human body joint point information, and the dense three-dimensional geometric information can be determined based on the human body component information.
Illustratively, when the sparse three-dimensional posture information is determined based on the human body joint point information and the dense three-dimensional geometric information is determined based on the human body component information, the human body joint point information and the human body component information are projected to a three-dimensional space respectively by combining the depth information of the image to be processed, and the sparse three-dimensional posture information and the dense three-dimensional geometric information are obtained.
In this embodiment, by acquiring depth information in an image to be processed, two-dimensional human body joint point information and the human body component information are projected to a three-dimensional space, so that the acquired sparse three-dimensional posture information and dense three-dimensional geometric information are more accurate, and a human body model optimized and reconstructed based on the sparse three-dimensional posture information and the dense three-dimensional geometric information is more accurate.
On the basis of any one of the above embodiments, determining a target human body model corresponding to a human body in an image to be processed based on sparse three-dimensional pose information, dense three-dimensional geometric information, and an initial parameterized human body model includes:
and optimizing the parameters of the initial parameterized human body model in an iterative mode based on the sparse three-dimensional attitude information and the dense three-dimensional geometric information to obtain the target human body model.
In the step, the sparse three-dimensional posture information and the dense three-dimensional geometric information are used as reference information to perform constraint optimization on inaccurate positions in the initial parameterized human body model, in specific implementation, the reference information of the sparse three-dimensional posture information and the dense three-dimensional geometric information can be used for constraining three-dimensional bone joints and the human body surface of the initial parameterized human body model, and parameters of the human body model are optimized in an iteration mode, so that the optimized human body model is closer to a real human body, and the iteration times are not specifically limited.
In this embodiment, all parameters in the initial parameterized model are optimized by using the acquired sparse three-dimensional attitude information and dense three-dimensional geometric information as reference information, so that model parameters for constructing the initial parameterized human body model are more accurate, and a more accurate human body model is constructed.
Fig. 2 is a second schematic diagram of a human body model constructing method according to an embodiment of the present invention, as shown in fig. 2, first, an initial parameterized human body model is obtained based on RGB images, that is, the human body shape of a designed average parameterized human body model is constrained and optimized, where the human body models may be human body models such as SCAPE, SMPL-X, and the like, then, 2D human body joint point and human body part information on the RGB images is obtained by a two-dimensional human body posture estimating method openposition and a human body part segmenting method segend, and on this basis, the D image is combined to promote the 2D human body joint point and human body part information to a three-dimensional space to obtain sparse three-dimensional posture information and dense three-dimensional geometric information, and the sparse three-dimensional posture information and dense three-dimensional geometric information are used as references to optimize the parameterized human body model to obtain a final parameterized human body model, that is a target human body model.
Further, on the basis of obtaining the parameterized human body model based on the RGB image, the sparse three-dimensional attitude information and the dense three-dimensional geometric information are further obtained through the RGBD image, on one hand, auxiliary optimization information of the parameterized human body model is added, on the other hand, the problem of depth ambiguity commonly existing in the prior art is also avoided, and therefore the reconstruction precision of the parameterized human body model is effectively improved. In addition, compared with a construction method of a non-parametric human body model, the method has no strict requirement on acquisition equipment, and the human body model constructed by the method can be directly driven and used in design software such as animation, clothes and the like, so that the method can be popularized and applied to numerous consumption scenes.
Therefore, the human body model construction method provided by the invention has the advantages that the human body model reconstruction precision, speed, equipment availability and model drivability are considered, and the method has stronger scene applicability.
The following describes the human body model construction device provided in the embodiment of the present invention, and the following human body model construction device and the above described human body model construction method may be referred to in correspondence with each other.
An embodiment of the present invention further provides a human body model building apparatus, and fig. 3 is a schematic structural diagram of the human body model building apparatus provided by the present invention, as shown in fig. 3, the apparatus includes:
an initialization module 310, configured to obtain an initial parameterized human body model based on an image to be processed;
a determining module 320, configured to determine sparse three-dimensional pose information and dense three-dimensional geometric information based on the image to be processed;
the determining module 320 is further configured to determine a target human body model corresponding to a human body in the image to be processed based on the sparse three-dimensional pose information, the dense three-dimensional geometric information, and the initial parameterized human body model.
According to the human body model construction device provided by the embodiment of the invention, the initial parameterized human body model is obtained through the image to be processed, and on the basis, the sparse three-dimensional attitude information and the dense three-dimensional geometric information containing the depth information are obtained through the image to be processed, so that the initial parameterized human body model is optimized, the problem of depth ambiguity in the initialized parameterized human body model is further solved, and the reconstruction precision of the parameterized human body model is effectively improved.
Optionally, the determining module 320 is specifically configured to:
determining human body joint point information and human body part information based on the image to be processed;
and determining sparse three-dimensional posture information based on the human body joint point information, and determining dense three-dimensional geometric information based on the human body component information.
Optionally, the determining module 320 is specifically configured to:
and respectively projecting the human body joint point information and the human body component information to a three-dimensional space based on the depth information to obtain sparse three-dimensional posture information and dense three-dimensional geometric information.
Optionally, the determining module 320 includes:
and the adjusting unit is used for adjusting the parameters of the initial parameterized human body model based on the sparse three-dimensional attitude information and the dense three-dimensional geometric information to obtain the target human body model.
Based on any of the above embodiments, the image to be processed comprises an RGBD image.
The apparatus of this embodiment may be configured to perform the method of any embodiment in the foregoing electronic device side method embodiment, and specific implementation processes and technical effects thereof are similar to those in the electronic device side method embodiment, and specific reference may be made to detailed descriptions in the electronic device side method embodiment, which are not described herein again.
Fig. 4 illustrates a physical structure diagram of an electronic device, which may include, as shown in fig. 4: a processor (processor) 410, a communication Interface 420, a memory (memory) 430 and a communication bus 440, wherein the processor 410, the communication Interface 420 and the memory 430 are communicated with each other via the communication bus 440. The processor 410 may invoke logic instructions in the memory 430 to perform a method of construction of a mannequin, the method comprising: obtaining an initial parameterized human body model based on an image to be processed; determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed; and determining a target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model.
In addition, the logic instructions in the memory 430 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, and various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product, the computer program product comprising a computer program, the computer program being stored on a non-transitory computer-readable storage medium, wherein when the computer program is executed by a processor, the computer is capable of executing the method for constructing a human body model provided by the above methods, the method comprising: obtaining an initial parameterized human body model based on an image to be processed; determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed; and determining a target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model.
In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, the computer program being implemented by a processor to perform the method for constructing a human body model provided by the above methods, the method comprising: obtaining an initial parameterized human body model based on an image to be processed; determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed; and determining a target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for constructing a human body model is characterized by comprising the following steps:
obtaining an initial parameterized human body model based on the image to be processed;
determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed;
and determining a target human body model corresponding to the human body in the image to be processed based on the sparse three-dimensional attitude information, the dense three-dimensional geometric information and the initial parameterized human body model.
2. The human body model construction method according to claim 1, wherein the determining sparse three-dimensional pose information and dense three-dimensional geometric information based on the image to be processed comprises:
determining human body joint point information and human body part information based on the image to be processed;
determining the sparse three-dimensional pose information based on the human body joint point information, and determining the dense three-dimensional geometric information based on the human body component information.
3. The method of constructing a human body model according to claim 2, wherein the determining the sparse three-dimensional pose information based on the human body joint point information and the dense three-dimensional geometry information based on the human body part information comprises:
and respectively projecting the human body joint point information and the human body part information to a three-dimensional space based on the depth information in the image to be processed to obtain the sparse three-dimensional posture information and the dense three-dimensional geometric information.
4. The method for constructing a human body model according to any one of claims 1-3, wherein the determining a target human body model corresponding to a human body in the image to be processed based on the sparse three-dimensional pose information, the dense three-dimensional geometric information and the initial parameterized human body model comprises:
and optimizing the parameters of the initial parameterized human body model in an iterative mode based on the sparse three-dimensional attitude information and the dense three-dimensional geometric information to obtain the target human body model.
5. A method of constructing a human model according to any of claims 1-3, wherein the image to be processed comprises an RGBD image.
6. An apparatus for constructing a human body model, comprising:
the initialization module is used for obtaining an initial parameterized human body model based on the image to be processed;
the determining module is used for determining sparse three-dimensional attitude information and dense three-dimensional geometric information based on the image to be processed;
the determining module is further configured to determine a target human body model corresponding to a human body in the image to be processed based on the sparse three-dimensional pose information, the dense three-dimensional geometric information, and the initial parameterized human body model.
7. The human model building apparatus according to claim 6, wherein the determining module is specifically configured to:
determining human body joint point information and human body part information based on the image to be processed;
and determining sparse three-dimensional posture information based on the human body joint point information, and determining dense three-dimensional geometric information based on the human body component information.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of constructing a mannequin according to any one of claims 1 to 5 when executing the program.
9. A non-transitory computer-readable storage medium on which a computer program is stored, the computer program, when being executed by a processor, implementing the method for constructing a mannequin according to any one of claims 1 to 5.
10. A computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the method of mannequin construction according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211275952.5A CN115601502A (en) | 2022-10-18 | 2022-10-18 | Human body model construction method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211275952.5A CN115601502A (en) | 2022-10-18 | 2022-10-18 | Human body model construction method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115601502A true CN115601502A (en) | 2023-01-13 |
Family
ID=84849307
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211275952.5A Pending CN115601502A (en) | 2022-10-18 | 2022-10-18 | Human body model construction method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115601502A (en) |
-
2022
- 2022-10-18 CN CN202211275952.5A patent/CN115601502A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110889890B (en) | Image processing method and device, processor, electronic equipment and storage medium | |
| Stoll et al. | Fast articulated motion tracking using a sums of gaussians body model | |
| Feng et al. | Avatar reshaping and automatic rigging using a deformable model | |
| CN110168608B (en) | System for acquiring 3-dimensional digital representations of physical objects | |
| Li et al. | Global correspondence optimization for non‐rigid registration of depth scans | |
| KR101911133B1 (en) | Avatar construction using depth camera | |
| KR101608253B1 (en) | Image-based multi-view 3d face generation | |
| Popa et al. | Globally consistent space‐time reconstruction | |
| CN113366491B (en) | Eyeball tracking method, device and storage medium | |
| CN113924600A (en) | Real-time body animation based on single image | |
| KR20090092473A (en) | 3D Face Modeling Method based on 3D Morphable Shape Model | |
| Chen et al. | Realtime reconstruction of an animating human body from a single depth camera | |
| CN113012293A (en) | Stone carving model construction method, device, equipment and storage medium | |
| US11922593B2 (en) | Methods of estimating a bare body shape from a concealed scan of the body | |
| Fan et al. | Convex hull aided registration method (CHARM) | |
| Yookwan et al. | Multimodal fusion of deeply inferred point clouds for 3D scene reconstruction using cross-entropy ICP | |
| Hartl et al. | Rapid reconstruction of small objects on mobile phones | |
| KR102655987B1 (en) | Apparatus and method for generating 3d avatar | |
| CN115496864B (en) | Model construction method, model reconstruction device, electronic equipment and storage medium | |
| Ju et al. | Individualising Human Animation Models. | |
| Jang et al. | 3D human modeling from a single depth image dealing with self-occlusion | |
| Starck et al. | Animated statues | |
| CN115601502A (en) | Human body model construction method, device, equipment and storage medium | |
| Villa-Uriol et al. | Automatic creation of three-dimensional avatars | |
| Ohtake et al. | A composite approach to meshing scattered data |
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 |