CN112686993B - Three-dimensional reconstruction method, apparatus and computer storage medium for three-dimensional object - Google Patents
Three-dimensional reconstruction method, apparatus and computer storage medium for three-dimensional object Download PDFInfo
- Publication number
- CN112686993B CN112686993B CN202110106935.8A CN202110106935A CN112686993B CN 112686993 B CN112686993 B CN 112686993B CN 202110106935 A CN202110106935 A CN 202110106935A CN 112686993 B CN112686993 B CN 112686993B
- Authority
- CN
- China
- Prior art keywords
- undirected graph
- matrix
- dimensional reconstruction
- point cloud
- translation
- 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
- 238000000034 method Methods 0.000 title claims abstract description 52
- 230000009466 transformation Effects 0.000 claims abstract description 52
- 239000011159 matrix material Substances 0.000 claims abstract description 51
- 230000004927 fusion Effects 0.000 claims abstract description 30
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 13
- 238000013519 translation Methods 0.000 claims description 35
- 238000009826 distribution Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 3
- 208000017604 Hodgkin disease Diseases 0.000 claims description 2
- 229940050561 matrix product Drugs 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 9
- 230000009977 dual effect Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000000844 transformation Methods 0.000 description 2
- XCWPUUGSGHNIDZ-UHFFFAOYSA-N Oxypertine Chemical compound C1=2C=C(OC)C(OC)=CC=2NC(C)=C1CCN(CC1)CCN1C1=CC=CC=C1 XCWPUUGSGHNIDZ-UHFFFAOYSA-N 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003631 expected effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Landscapes
- Image Processing (AREA)
- Processing Or Creating Images (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a three-dimensional reconstruction method, equipment and a computer storage medium for a three-dimensional object, wherein the method comprises the following steps: acquiring point cloud data of the surface of the three-dimensional object by using acquisition equipment, wherein the point cloud data comprises a plurality of pieces of point clouds; according to the overlapping relation between the point clouds, carrying out pairwise registration between the point clouds by using a point cloud registration method so as to realize local registration; constructing an undirected graph based on the pairwise registered point clouds; triangulating the undirected graph; determining a transformation matrix corresponding to each side of the undirected graph after triangulation; correcting the transformation matrix by using a discrete Hodgkin decomposition algorithm; and acting the corrected transformation matrix on the point cloud data to perform data fusion, thereby realizing three-dimensional reconstruction of global consistency. The invention can remarkably improve the overall continuity and accuracy of data fusion.
Description
Technical Field
The present invention relates generally to the field of three-dimensional reconstruction. More particularly, the present invention relates to a three-dimensional reconstruction method, apparatus and computer storage medium for three-dimensional objects.
Background
In a three-dimensional reconstruction scene in practical application, the problem of fusion of data acquired by a plurality of sensors and multiple views exists. Often the local data of each sensor is reasonable, but if it cannot be fused properly, the overall data obtained locally may be subject to large errors, even complete errors. Therefore, global consistency adjustment is carried out on the whole data with errors and deviations, and the method plays an important role in improving the three-dimensional reconstruction quality.
The traditional three-dimensional reconstruction method is to register according to the geometric features and the distribution features of the point cloud. In recent years, with the development of artificial intelligence, a method for deep learning for point cloud registration has also emerged. However, the two methods still solve the local registration problem, and global continuity is not considered. Therefore, how to achieve the overall consistency on the premise of good local registration becomes a problem to be solved.
Disclosure of Invention
The invention mainly solves the technical problem of how to realize overall consistency under the premise of good local registration of the three-dimensional reconstruction method in the prior art, and provides the three-dimensional reconstruction method, equipment and a computer storage medium for the three-dimensional object, which can reconstruct an overall continuous high-quality three-dimensional model based on geometric topology theory so as to obtain a fusion scheme with overall consistency and remarkably improve the overall continuity and accuracy of data fusion. In view of this, the present invention provides a reconstruction scheme in several aspects as follows.
In one aspect, the present invention provides a three-dimensional reconstruction method for a three-dimensional object, comprising: acquiring point cloud data of the surface of the three-dimensional object by using acquisition equipment, wherein the point cloud data comprises a plurality of pieces of point clouds; according to the overlapping relation between the point clouds, carrying out pairwise registration between the point clouds by using a point cloud registration method so as to realize local registration; constructing an undirected graph based on the pairwise registered point clouds; triangulating the undirected graph; determining a transformation matrix corresponding to each side of the undirected graph after triangulation; correcting the transformation matrix by using a discrete Hodgkin decomposition algorithm; and acting the corrected transformation matrix on the point cloud data to perform data fusion, thereby realizing three-dimensional reconstruction of global consistency.
In one embodiment, constructing an undirected graph based on the pairwise registered point clouds includes: and defining a node for each point cloud, and connecting two nodes corresponding to the two point clouds with the registration relationship to construct the undirected graph.
In one embodiment, the local registration includes translation and rotation, and the transformation matrix includes a translation matrix and a rotation matrix.
In one embodiment, determining the length of an edge in the undirected graph using the center point distance of the locally registered point cloud is further included.
In one embodiment, the acquisition device comprises one or more of a radar, a laser, or a depth camera.
In one embodiment, the point cloud registration method includes an iterative nearest point algorithm or a normal distribution transformation algorithm.
In one embodiment, determining a translation matrix and a rotation matrix corresponding to each edge of the undirected graph after the triangulation includes: for the original edge of the undirected graph, the corresponding translation matrix relates to a locally registered translation part between the point clouds corresponding to the two endpoints of the edge, and the corresponding rotation matrix relates to a locally registered rotation part between the point clouds corresponding to the two endpoints of the edge; for the newly added edge of the undirected graph after triangulation, the rotation matrix and the translation matrix are obtained according to the rotation matrix and the translation matrix of other edges in the triangle where the rotation matrix meets the requirement that the rotation matrix product on all three edges of the triangle is identical transformation, and the translation matrix meets the requirement that the sum of the translation matrices on all three edges of the triangle is zero.
In one embodiment, modifying the transformation matrix using the discrete hodgkin's decomposition algorithm comprises: the lie and translation matrices of the rotation matrix corresponding to each edge in the undirected graph are defined as discrete forms, denoted omega T ，ω R The method comprises the steps of carrying out a first treatment on the surface of the Omega is decomposed according to the discrete Hodgy decomposition algorithm T ，ω R Respectively decomposed into omega T ＝dτ T +δη T +h T And omega R ＝dτ R +δη R +h R And satisfy the following equations respectively:
dωT＝dδη T ，δw T ＝δdτ T
dω R ＝dδη R ，δw R ＝δdτ R
wherein omega T Representing a translation matrix, ω R Representing a rotation matrix dτ T And dτ R In the proper form, delta eta T And δη R In the form of closed form, h T And h R In a harmonized form; solving the above equation set to obtain τ T And τ R The method comprises the steps of carrying out a first treatment on the surface of the Applying a differential operator to said τ T And τ R To obtain omega T And omega R In the proper form dτ T And dτ R Wherein the proper form satisfies global consistency in the undirected graph; using the appropriate form dτ T And dτ R To correct the fusion result between the two corresponding point clouds.
In another aspect, the present invention provides an apparatus for three-dimensional reconstruction of a three-dimensional object, comprising: a processor; and a memory coupled to the processor, the memory having stored therein computer program code that, when executed by the processor, causes the apparatus to perform the method as described above and in various embodiments thereof.
In yet another aspect, the present invention provides a computer-readable storage medium having stored thereon computer-readable instructions for three-dimensional reconstruction of a three-dimensional object, which when executed by one or more processors, implement the method as described in the above and in various embodiments thereof.
From the foregoing description, those skilled in the art will appreciate that the present invention provides a method, apparatus, and computer storage medium for three-dimensional reconstruction of global consistency. Specifically, an undirected graph is constructed from registration relationships between point clouds and its corresponding triangulation is constructed. Then, the local registration transformation is defined as an edge form, the proper transformation is extracted through a Hodge decomposition algorithm, and the point cloud fusion transformation is integrally corrected on triangulation. And finally, updating the registration result according to the corrected point cloud fusion transformation, thereby realizing the integral adjustment of the point cloud fusion. Therefore, the invention provides a fusion scheme with global consistency, and the overall continuity and accuracy of data fusion are obviously improved.
By the scheme of the invention, the overall adjustment of registration can be performed on the data with larger error in the data fusion result. For example, on the basis of the local registration result of the point cloud, the registration data is corrected integrally, so that the problem of high-precision fusion between a plurality of sensors and multi-view acquired data is converted into the problem of geometric topology on an undirected graph. Based on the method, the scheme of the invention can effectively improve the three-dimensional reconstruction quality and provide a more effective way for point cloud fusion. Furthermore, by the global consistency three-dimensional reconstruction method, accurate three-dimensional reconstruction can be performed on complex real scenes. In some application scenarios, the obtained integral continuous three-dimensional reconstruction model can be applied to various fields of terrain exploration, automatic driving and the like, and corresponding solutions are provided from various aspects.
Drawings
FIG. 1 is a simplified flow chart of a three-dimensional reconstruction method for a three-dimensional object provided by the present invention;
FIG. 2 is a detailed flow chart of a three-dimensional reconstruction method for a three-dimensional object provided by the present invention;
FIG. 3 is an undirected graph corresponding to a point cloud set acquired by an acquisition device;
FIG. 4 is a schematic diagram of triangulation of undirected graph generation according to an embodiment of the present invention;
FIG. 5 is an exemplary schematic diagram of the vertices and their neighbors of triangulation of a portion of an undirected graph in accordance with an embodiment of the present invention;
FIG. 6 is an exemplary diagram of a comparison of before and after global coherency adjustment in accordance with an embodiment of the present invention;
FIG. 7 is a second exemplary diagram of a comparison of global coherency adjustment in accordance with an embodiment of the present invention; and
FIG. 8 is an exemplary diagram of a comparison of before and after global coherency adjustment in accordance with an embodiment of the present invention.
Detailed Description
In order to make the technical problems solved by the invention, the technical scheme adopted and the technical effects achieved clearer, the invention is further described in detail below with reference to the accompanying drawings and the embodiments. It is to be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention. It should be further noted that, for convenience of description, only some, but not all of the matters related to the present invention are shown in the accompanying drawings.
Fig. 1 is a simplified flow chart of a three-dimensional reconstruction method for a three-dimensional object provided by the present invention. As shown in fig. 1, the three-dimensional reconstruction method of the present invention starts at step 10. At this step, point cloud data of the three-dimensional object surface is acquired with an acquisition device, wherein the point cloud data comprises a plurality of pieces of point cloud. In one embodiment, the acquisition device may include one or more of a radar, a laser, or a depth camera.
At step 20, according to the overlapping relationship between the point clouds, the point clouds are registered in pairs by using a point cloud registration method, so as to realize local registration. In one embodiment, the point cloud registration method may include an iterative nearest point algorithm or a normal distribution transformation algorithm. In one embodiment, the local registration may include translation and rotation.
After pairwise registration, at step 30, an undirected graph is constructed based on the pairwise registered point clouds. In one embodiment, constructing the undirected graph based on the pairwise registered point clouds may include defining one node for each piece of point cloud, and connecting two nodes corresponding to the two pieces of point clouds having the registration relationship to construct the undirected graph. Next, at step 40, the undirected graph obtained above is triangulated, and at step 50, a transformation matrix corresponding to each edge of the undirected graph after triangulating is determined. In one embodiment, the transformation matrix herein may include a translation matrix and a rotation matrix.
At step 60, the transformation matrix is modified using a discrete hodgkin ("Hodge") decomposition algorithm; and at step 70, applying the modified transformation matrix to the point cloud data for data fusion, thereby achieving a three-dimensional reconstruction of global consistency.
Although not shown in fig. 1, in some embodiments, the lie and translation matrices of the rotation matrix corresponding to each edge in the undirected graph may be defined as discrete forms, and appropriate ones of the foregoing discrete forms may be obtained in accordance with the discrete hall and micro-sub-algorithms.
Based on the description of the invention, a person skilled in the art can understand that the invention extracts a proper form of global agreement in the form of triangulation generated by local registration of data through Hodge decomposition, so that new fusion contains richer global information, and the reconstructed object has integral continuous property.
Fig. 2 is a detailed flowchart of the three-dimensional reconstruction method for a three-dimensional object provided by the present invention. As shown in fig. 2, the method for reconstructing global consistency according to the embodiments of the present invention includes, at step 100, acquiring point cloud data of a three-dimensional object surface using an acquisition device. As previously mentioned, the acquisition device may include, but is not limited to, radar, laser, depth camera, and the like. Thus, the fused data of the present invention may be acquired by various types of depth camera devices including, for example, azureKinectDK, lidar, and structured light. Furthermore, the invention can complete fusion of the data acquired by the three devices and obtain the expected effect of overall consistency.
In step 200, according to the overlapping relationship between the point clouds, the point clouds are registered two by using a point cloud registration method, so as to obtain a reconstructed model with good local registration, and the local registration is completed. As previously described in connection with fig. 1, the local registration may include both translational and rotational portions. The point cloud registration method includes, but is not limited to, ICP (Iterative closest point, iterative closest point algorithm) and NDT (Normal Distributions Transform, normal distribution transformation algorithm) according to different application scenarios.
At step 300, a node may be defined for each point cloud, and two node wires corresponding to two point clouds with a registration relationship are connected, where a registration relationship exists between two point clouds corresponding to two nodes connected by each edge of the undirected graph, so as to complete the construction of the undirected graph. Here, the registration relationship may indicate that there is an overlapping region between the point clouds. For exemplary purposes only, FIG. 3 illustrates a schematic diagram of an undirected graph generated in accordance with an embodiment of the present invention.
At step 400, the lengths of edges in the undirected graph constructed in step 300 may be determined using the center point distances of the locally transformed point clouds obtained in step 200, and the undirected graph may be triangulated. In this step, the center point distance of the point cloud after the local transformation is defined as the length of the edge in the undirected graph. The process of constructing the triangulation of the undirected graph includes adding several new edges in the undirected graph until all the rings in the undirected graph are the boundaries of triangles or triangle areas. Similarly, for exemplary purposes only, FIG. 4 shows a schematic diagram of triangulation generated according to an embodiment of the present invention. In the embodiment of the invention, the data fusion condition with more complex topology can be processed by generating triangulation. Based on this, for the complex topology undirected graph, step 400 can complete triangulation, so that the point cloud fusion technique can be applied to the complex topology scene.
At step 500, a translational transformation and a rotational transformation corresponding to each edge of the undirected graph may be determined for different edges on the triangulation. In particular, for an edge of the undirected graph, its corresponding translational transformation is the partially registered translational portion between the two endpoint-corresponding point clouds of the edge, and its corresponding rotational transformation is the partially registered rotational portion between the two endpoint-corresponding point clouds of the edge; for newly added edges in triangulation, the corresponding rotation transformation form and translation transformation form are calculated according to the rotation transformation and translation transformation of other edges in the triangle where the rotation transformation form meets that the rotation transformation product on all three edges of the triangle is an identity transformation, and the translation transformation form meets that the sum of the translation transformations on all three edges of the triangle is zero.
At step 600, the translation transform and the rotation transform (i.e., the transform matrix of the present invention described previously) are modified using a discrete Hodge decomposition algorithm to implement a fusion transform, based on the determined translation transform form and rotation transform form. Therefore, the three-dimensional reconstruction method can realize the three-dimensional reconstruction of the global consistency of the three-dimensional object.
The invention is based on triangulation and two forms thereon, wherein the two forms are defined as lie algebra of rotation transformation and translation transformation, respectively, and the dual meshes of the triangulation are combined. For exemplary purposes only, FIG. 5 shows a schematic diagram of a portion of a triangulated dual mesh generated in accordance with an embodiment of the present invention.
Returning to flowchart 2, in one embodiment, step 600 may include steps 601 and 602 as follows:
at step 601, the lie and translation transforms of the rotation transform corresponding to each edge in the undirected graph are defined as discrete forms listing the equations of the Hodge decomposition algorithm, which can be decomposed into ω, respectively T ＝dτ T +δη T +h T And omega R ＝dτ R +δη R +h R And satisfy the equations:
dω T ＝dδη T ，δw T ＝δdτ T
dω R ＝dδη R ，δw R ＝δdτ R
wherein omega T Representing translation transformation ω R Representing rotation transformation dτ T And dτ R In the proper form, delta eta T And δη R In the form of closed form, h T And h R In a harmonic form.
At step 602, τ is found by solving the system of equations T And τ R I.e. the value at each vertex, and applying a differential operator to τ T And τ R Omega is obtained T And omega R In the proper form dτ T And dτ R And obtaining an appropriate form that any two connected points meet global consistency, and calculating the transformation between the two corresponding point clouds. Can be used forIt will be appreciated that the proper form found at step 600 of the present invention satisfies the consistent nature of either ring in the triangulation and undirected graph.
Next, at step 700, the fusion transformation modified by the above steps may be applied to all three-dimensional data (i.e., the aforementioned point cloud data) to complete the global modification of the data fusion.
Depending on the nature of the proper form, the final rotation transformation satisfies: the rotation transformation product on the three sides of each triangle is an identity transformation, and the final translation transformation satisfies: the sum of the translational transformations on the three sides of each triangle is zero.
As previously described, the modified data has globally consistent characteristics. At the same time, the locality between the data may no longer be optimal. In order to make the fusion compatible with the local optimization and the global agreement, the iteration of the local registration and the global optimization is performed on the updated point cloud data through the loops of the steps 200 to 700 when necessary. FIGS. 6-8 are exemplary diagrams of comparisons before and after global coherency adjustment in accordance with embodiments of the present invention; it can be seen from the figure that the reconstructed object can have overall continuous property through the global consistency processing of the invention.
In combination with the above description, those skilled in the art can understand that the method extracts a proper form of global agreement through the form of triangulation generated by local registration of data by the Hodge decomposition, so that the new fusion contains richer global information, and the reconstructed object can have integral continuous property. Meanwhile, compared with the traditional fusion method, the method is more reasonable in theory based on global fusion in geometric topology. Furthermore, by utilizing the global fusion algorithm provided by the invention and combining a traditional local registration method, data fusion can be further analyzed and processed.
Based on the above description in conjunction with the accompanying drawings, one skilled in the art will appreciate that embodiments of the present invention also provide an apparatus for three-dimensional reconstruction of a three-dimensional object, comprising: a processor; and a memory coupled to the processor, the memory having stored therein computer program code that, when executed by the processor, causes the apparatus to perform the method provided by any of the embodiments of the invention.
Embodiments of the present invention also provide a computer-readable storage medium having stored thereon computer-readable instructions for three-dimensional reconstruction of a three-dimensional object, which when executed by one or more processors, implement the method provided by any of the embodiments of the present invention.
The computer readable storage medium described above may be any suitable magnetic or magneto-optical storage medium, such as, for example, resistive Random Access Memory RRAM (Resistive Random Access Memory), dynamic Random Access Memory DRAM (Dynamic Random Access Memory), static Random Access Memory SRAM (Static Random-Access Memory), enhanced dynamic Random Access Memory EDRAM (Enhanced Dynamic Random Access Memory), high-Bandwidth Memory HBM (High-Bandwidth Memory), hybrid storage cube HMC (Hybrid Memory Cube), etc., or any other medium that may be used to store the desired information and that may be accessed by an application, a module, or both. Any such computer storage media may be part of, or accessible by, or connectable to, the device. Any of the applications or modules described herein may be implemented using computer-readable/executable instructions that may be stored or otherwise maintained by such computer-readable media.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments is modified or some or all of the technical features are replaced equivalently, so that the essence of the corresponding technical scheme does not deviate from the scope of the technical scheme of the embodiments of the present invention.
Claims (10)
1. A three-dimensional reconstruction method for a three-dimensional object, comprising:
acquiring point cloud data of the surface of the three-dimensional object by using acquisition equipment, wherein the point cloud data comprises a plurality of pieces of point clouds;
according to the overlapping relation between the point clouds, carrying out pairwise registration between the point clouds by using a point cloud registration method so as to realize local registration;
constructing an undirected graph based on the pairwise registered point clouds;
triangulating the undirected graph;
determining a transformation matrix corresponding to each side of the undirected graph after triangulation;
correcting the transformation matrix by using a discrete Hodgkin decomposition algorithm; and
and acting the corrected transformation matrix on the point cloud data to perform data fusion, thereby realizing three-dimensional reconstruction of global consistency.
2. The three-dimensional reconstruction method according to claim 1, wherein constructing an undirected graph based on the pairwise registered point clouds comprises:
and defining a node for each point cloud, and connecting two nodes corresponding to the two point clouds with the registration relationship to construct the undirected graph.
3. The three-dimensional reconstruction method of claim 1, wherein the local registration comprises translation and rotation and the transformation matrix comprises a translation matrix and a rotation matrix.
4. The three-dimensional reconstruction method according to claim 1, further comprising determining a length of an edge in the undirected graph using a center point distance of the locally registered point cloud.
5. The three-dimensional reconstruction method according to claim 1, wherein the acquisition device comprises one or more of a radar, a laser, or a depth camera.
6. The three-dimensional reconstruction method according to claim 1, wherein the point cloud registration method comprises an iterative nearest point algorithm or a normal distribution transformation algorithm.
7. A three-dimensional reconstruction method according to claim 3, wherein determining a translation matrix and a rotation matrix corresponding to each edge of the undirected graph after the triangulation comprises:
for the original edge of the undirected graph, the corresponding translation matrix relates to a locally registered translation part between the point clouds corresponding to the two endpoints of the edge, and the corresponding rotation matrix relates to a locally registered rotation part between the point clouds corresponding to the two endpoints of the edge;
for the newly added edge of the undirected graph after triangulation, the rotation matrix and the translation matrix are obtained according to the rotation matrix and the translation matrix of other edges in the triangle where the rotation matrix meets the requirement that the rotation matrix product on all three edges of the triangle is identical transformation, and the translation matrix meets the requirement that the sum of the translation matrices on all three edges of the triangle is zero.
8. The three-dimensional reconstruction method according to claim 7, wherein modifying the transformation matrix using the discrete hodgkin's decomposition algorithm comprises:
the lie and translation matrices of the rotation matrix corresponding to each edge in the undirected graph are defined as discrete forms, denoted omega T ，ω R ；
Omega is decomposed according to the discrete Hodgy decomposition algorithm T ，ω R Respectively decomposed into omega T ＝dτ T +δη T +h T And omega R ＝dτ R +δη R +h R And satisfy the following equations respectively:
dω T ＝dδη T ，δw T ＝δdτ T
dω R ＝dδη R ，δw R ＝δdτ R
wherein omega T Representing a translation matrix, ω R Representing a rotation matrix dτ T And dτ R In the proper form, delta eta T And δη R In the form of closed form, h T And h R For adjustingAnd form;
solving the above equation set to obtain τ T And τ R ；
Applying a differential operator to said τ T And τ R To obtain omega T And omega R In the proper form dτ T And dτ R Wherein the proper form satisfies global consistency in the undirected graph;
using the appropriate form dτ T And dτ R To correct the fusion result between the two corresponding point clouds.
9. An apparatus for three-dimensional reconstruction of a three-dimensional object, comprising:
a processor; and
a memory coupled to the processor, the memory having stored therein computer program code which, when executed by the processor, causes the apparatus to perform the method of any of claims 1-8.
10. A computer storage medium having stored thereon computer readable instructions for three-dimensional reconstruction of a three-dimensional object, which when executed by one or more processors, implement the method of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110106935.8A CN112686993B (en) | 2021-01-27 | 2021-01-27 | Three-dimensional reconstruction method, apparatus and computer storage medium for three-dimensional object |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110106935.8A CN112686993B (en) | 2021-01-27 | 2021-01-27 | Three-dimensional reconstruction method, apparatus and computer storage medium for three-dimensional object |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112686993A CN112686993A (en) | 2021-04-20 |
| CN112686993B true CN112686993B (en) | 2024-04-02 |
Family
ID=75459296
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110106935.8A Active CN112686993B (en) | 2021-01-27 | 2021-01-27 | Three-dimensional reconstruction method, apparatus and computer storage medium for three-dimensional object |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112686993B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113628307B (en) * | 2021-08-02 | 2023-07-28 | 聚好看科技股份有限公司 | Bone driving method and device for three-dimensional model |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105427317A (en) * | 2015-11-25 | 2016-03-23 | 武汉大学 | Method suitable for multi-view-angle automatic registration of ground laser point cloud data of multiple stations |
| CN106803267A (en) * | 2017-01-10 | 2017-06-06 | 西安电子科技大学 | Indoor scene three-dimensional rebuilding method based on Kinect |
| CN110211129A (en) * | 2019-05-17 | 2019-09-06 | 西安财经学院 | Low covering point cloud registration algorithm based on region segmentation |
| CN110889243A (en) * | 2019-12-20 | 2020-03-17 | 南京航空航天大学 | Aircraft fuel tank three-dimensional reconstruction method and detection method based on depth camera |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10489927B2 (en) * | 2016-07-18 | 2019-11-26 | King Abdullah University Of Science And Technology | System and method for three-dimensional image reconstruction using an absolute orientation sensor |
-
2021
- 2021-01-27 CN CN202110106935.8A patent/CN112686993B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105427317A (en) * | 2015-11-25 | 2016-03-23 | 武汉大学 | Method suitable for multi-view-angle automatic registration of ground laser point cloud data of multiple stations |
| CN106803267A (en) * | 2017-01-10 | 2017-06-06 | 西安电子科技大学 | Indoor scene three-dimensional rebuilding method based on Kinect |
| CN110211129A (en) * | 2019-05-17 | 2019-09-06 | 西安财经学院 | Low covering point cloud registration algorithm based on region segmentation |
| CN110889243A (en) * | 2019-12-20 | 2020-03-17 | 南京航空航天大学 | Aircraft fuel tank three-dimensional reconstruction method and detection method based on depth camera |
Non-Patent Citations (1)
| Title |
|---|
| 激光扫描立式罐倾斜测量新方法研究;陈伟;郝华东;李东升;陈贤雷;;中国测试;20150228(第02期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112686993A (en) | 2021-04-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zheng et al. | Fringe projection profilometry by conducting deep learning from its digital twin | |
| Berger et al. | A survey of surface reconstruction from point clouds | |
| Sweeney et al. | Large scale sfm with the distributed camera model | |
| CN112381879B (en) | Object posture estimation method, system and medium based on image and three-dimensional model | |
| CN112836618B (en) | Three-dimensional human body posture estimation method and computer readable storage medium | |
| Yoon et al. | Surface and normal ensembles for surface reconstruction | |
| CN109636912B (en) | Tetrahedral subdivision finite element interpolation method applied to three-dimensional sonar image reconstruction | |
| Cui et al. | High-precision 3D reconstruction for small-to-medium-sized objects utilizing line-structured light scanning: A review | |
| CN115761178A (en) | Multi-view three-dimensional reconstruction method based on implicit neural representation | |
| Yuan et al. | Feature preserving multiresolution subdivision and simplification of point clouds: A conformal geometric algebra approach | |
| CN112686993B (en) | Three-dimensional reconstruction method, apparatus and computer storage medium for three-dimensional object | |
| CN116958379A (en) | Image rendering method, device, electronic equipment, storage medium and program product | |
| Campos et al. | Splat-based surface reconstruction from defect-laden point sets | |
| Zhu et al. | Propagation strategies for stereo image matching based on the dynamic triangle constraint | |
| CN117132737B (en) | Three-dimensional building model construction method, system and equipment | |
| Chen et al. | Densefusion: Large-scale online dense pointcloud and dsm mapping for uavs | |
| De Araújo et al. | Adaptive polygonization of implicit surfaces | |
| Li et al. | Primitive-Based 3d building modeling, sensor simulation, and estimation | |
| Hyeon et al. | Automatic spatial template generation for realistic 3d modeling of large-scale indoor spaces | |
| Baudouin et al. | Lloyd’s energy minimization in the L p norm for quadrilateral surface mesh generation | |
| Oliveira et al. | Incremental scenario representations for autonomous driving using geometric polygonal primitives | |
| Zhang et al. | Fast Mesh Reconstruction from Single View Based on GCN and Topology Modification. | |
| Abdelfattah et al. | On Image to 3D Volume Construction for E-Commerce Applications | |
| Bærentzen et al. | Reconstruction of a Botanical Tree from a 3D Point Cloud | |
| CN113591369B (en) | Single-frame view three-dimensional model point cloud reconstruction method and storage device based on priori constraint |
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 |