CN107730587B - Rapid three-dimensional interactive modeling method based on pictures - Google Patents
Rapid three-dimensional interactive modeling method based on pictures Download PDFInfo
- Publication number
- CN107730587B CN107730587B CN201610656800.8A CN201610656800A CN107730587B CN 107730587 B CN107730587 B CN 107730587B CN 201610656800 A CN201610656800 A CN 201610656800A CN 107730587 B CN107730587 B CN 107730587B
- Authority
- CN
- China
- Prior art keywords
- dimensional
- contour
- modeling
- image
- technology
- 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 49
- 230000002452 interceptive effect Effects 0.000 title claims abstract description 28
- 238000013507 mapping Methods 0.000 claims abstract description 41
- 238000005516 engineering process Methods 0.000 claims abstract description 33
- 230000011218 segmentation Effects 0.000 claims abstract description 20
- 230000002093 peripheral effect Effects 0.000 claims abstract description 13
- 238000000605 extraction Methods 0.000 claims abstract description 6
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 4
- 230000014759 maintenance of location Effects 0.000 claims abstract description 4
- 230000000694 effects Effects 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 238000012423 maintenance Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 2
- 230000003993 interaction Effects 0.000 description 6
- 230000000007 visual effect Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 241000282994 Cervidae Species 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000010422 painting Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000010408 sweeping Methods 0.000 description 1
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
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Processing Or Creating Images (AREA)
- Image Generation (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a rapid three-dimensional interactive modeling method based on pictures, which comprises the following steps: utilizing an interactive contour extraction technology to outline the whole part of the colored drawing picture and extracting a peripheral contour line; the obtained peripheral contour line of the whole part is sketched by utilizing an interactive contour segmentation technology and a contour segmentation line between the parts is segmented; obtaining the peripheral contour lines of all parts of the colored drawing picture according to the contour dividing lines; establishing a two-dimensional outline frame of the three-dimensional model according to the two-dimensional outline of the characteristic region decomposition; generating a three-dimensional curved surface for the two-dimensional outline frame by utilizing a three-dimensional free-form surface modeling technology; performing texture mapping of the retention characteristics according to the generated three-dimensional curved surface model of each colored drawing part; and automatically splicing the geometric shapes of the three-dimensional curved surface models of the colored drawing parts after texture mapping and correcting the spliced textures to obtain a complete three-dimensional model.
Description
Technical Field
The invention relates to a three-dimensional modeling method, in particular to a rapid three-dimensional interactive modeling method based on pictures.
Background
Traditional three-dimensional modeling software such as 3dsmax, maya, blender and the like has strong modeling capability, a user can construct an arbitrarily complex three-dimensional model through some basic geometric units, and because the modeling engineer needs to have rich three-dimensional modeling experience and imagination, the software is usually used for constructing a complex geometric model and needs to be different from hours to days.
Another class is Sketch-based three-dimensional modeling (Sketch based modeling) techniques, such as:
1) multi-view based modeling
The modeling method based on multiple visual angles is not limited to drawing a sketch on a single plane, but adopts a modeling mode of rotating visual angles, and a user carries out three-dimensional modeling by drawing outlines at different visual angles [6,7,8 and 9 ]. Igarashi et al [6] designed interactive modeling System Teddy in 1999 is one of the most representative works. The system utilizes the hand-drawn contour to carry out free modeling, extracts a framework for a closed plane curve input by a user, calculates the distance from a vertex in a closed curve area to the framework, and generates a smooth three-dimensional model by offsetting to two sides of the framework. Fibermesh [9] designed by Nealen et al in 2007 uses a three-dimensional space curve input by a user as a constraint and uses a nonlinear optimization method to generate a smooth curved surface. The method has higher degree of freedom, however, frequent view angle switching increases the modeling difficulty, and the interaction mode is not intuitive when three-dimensional modeling is carried out by taking a single picture as a reference, so that a user is required to have a considerable modeling experience.
2) Single view based modeling
According to Hoffman et al study [10], it is also generally preferred that two-dimensional curves be used to represent the contour information of the three-dimensional model under this view. A single photo or drawing can only provide information of a single view angle, and in order to reduce the interaction degree of a user and improve the modeling efficiency, a modeling mode based on a single view has strong application requirements. However, the contour curve in the drawing not only contains the external contour of the three-dimensional object, but also contains rich detail characteristics, and the direct three-dimensional modeling of the contour curve has certain difficulty. And because the two-dimensional contour curve contains certain ambiguity, certain assumptions are often made to construct a three-dimensional model which has good visual effect and accords with general intuition. One common practice is to first construct an approximate three-dimensional model from the external contour information of the object, and then allow the user to further edit the model based thereon.
Such methods usually rely on the drawing sequence and different gestures to solve ambiguity problems, and achieve the purpose of modeling complex objects. Gigcolod et al [11] interactively edits generalized cylinders (generalized cylinders) based on a predefined representation to obtain complex objects in the drawing. Olsen [12] enables a user to construct different geometric shapes by outlining objects and using predefined gestures. Shtof et al [13] uses parameterized geometric primitives to fit two-dimensional contours to achieve modeling. Chen et al [14] model objects in photographs by a method of sweeping in conjunction with user-differentiated cues for tubular objects. Such methods allow a user to model objects in a picture through a single perspective and enhance the ability of the system to model through additional user cues. Bessmeltsev et al [15] propose a cartoon character modeling approach that incorporates a three-dimensional skeleton. The three-dimensional skeleton covers the depth information of all parts of the cartoon character, is used for guiding the segmentation of the outline, and finally utilizes the segmented outline to construct a rotator to obtain a three-dimensional model. In this way, cartoon characters with complex gestures can be modeled. Another class of methods proposes an effective three-dimensional model reconstruction method [16] by the perspective principle of sketching wireframes, however children's painting generally does not follow the perspective principle, and therefore this class of methods is not suitable for the problem herein. Therefore, how to generate a credible three-dimensional model on the premise of ensuring simple interaction according to a single image or drawing is a research problem with practical significance. Although the interactive modeling system introduced above can automatically generate a three-dimensional model, in the three-dimensional model modeling process, when a user manually outlines, deviation is easy to occur, so that the modeling effect is poor, a visual angle needs to be continuously rotated during modeling, modeling of a geometric body with a perspective angle is complex or cannot be performed at all, textures on the model need to be manually added by an artist, the smoothness of the manually added textures is poor, the textures on each part are not easy to be mutually and smoothly spliced, and for parts with large shielding, the existing system cannot be well rebuilt and correspond due to the fact that a large amount of information on images is lost when the corresponding three-dimensional model and the textures of the three-dimensional model are rebuilt, so that the overall effect after forming is relatively poor, the difficulty of operation is increased, and further, the use effect and the practicability are affected.
Disclosure of Invention
The invention aims to provide a rapid three-dimensional interactive modeling method based on pictures, which is designed for solving the defects of the technology and improving the using effect and the practical performance.
The invention discloses a rapid three-dimensional interactive modeling method based on pictures, which is characterized by comprising the following steps of:
A) utilizing an interactive contour extraction technology to outline the whole part of the colored drawing picture and extracting the peripheral contour line of the image to obtain the foreground and the background of the whole picture;
B) the obtained peripheral contour line of the whole part is sketched by utilizing an interactive contour segmentation technology, and contour segmentation lines between the parts are segmented to obtain the foreground and the background of the image in the whole picture;
C) obtaining two-dimensional contour lines of all parts of the colored drawing picture according to the contour dividing lines;
D) establishing a two-dimensional outline frame of the three-dimensional model according to the two-dimensional outline of the characteristic region decomposition;
E) generating a three-dimensional curved surface for the two-dimensional outline frame by utilizing a three-dimensional free-form surface modeling technology;
F) performing texture mapping of the retention characteristics according to the generated three-dimensional curved surface model of each colored drawing part;
G) and automatically splicing the geometric shapes of the three-dimensional curved surface models of the colored drawing parts after texture mapping and correcting the spliced textures to obtain a complete three-dimensional model.
Further, the three-dimensional free-form surface modeling technique includes a surface modeling method and a three-dimensional modeling method.
Further, the contour dividing line which is deviated when the contour dividing line is drawn at each part in the colored drawing picture is automatically judged and corrected according to the drawing direction and the tendency of the connecting points of the two ends of the deviated contour dividing line and the peripheral contour line, so that the contour of the target area is obtained.
Further, the specific steps of automatically judging and correcting the offset contour line during contour segmentation are as follows:
A) modeling a foreground and a background in the image by adopting a Gaussian mixture model, and predicting the probability of each pixel in the image belonging to the foreground or the background;
B) constructing a network according to the adjacency relation of the pixels;
C) according to the method, a pixel in an image is taken as a node in a network, and the nodes are a source point s and a sink point t respectively;
D) the weight of the edge in the network is obtained through the probability that the corresponding pixel point belongs to the foreground and the color difference between the adjacent pixel points;
F) and (5) completing network construction, and obtaining optimal contour segmentation by adopting a minimum segmentation algorithm.
Furthermore, the method also comprises the step of carrying out rapid three-dimensional interactive modeling on the colored drawing picture with the perspective angle of the planar sketch by adopting a Fibermesh technology of a three-dimensional free-form surface modeling technology as an algorithm for generating the grid.
Further, the Fibermesh technology of the three-dimensional free-form surface modeling technology as an algorithm for generating a grid is used for three-dimensional modeling operation of a colored drawing picture with perspective angles on a planar sketch, and comprises the following steps: and drawing a central line on the three-dimensional modeling object in the colored drawing picture, estimating a symmetrical plane of the three-dimensional model according to the central line, and combining the symmetrical plane with the two-dimensional outline frame to obtain a smooth three-dimensional initial model.
Further, the texture mapping comprises a parallel projection technology for automatically keeping high-frequency characteristic information of an image and a continuous mapping method for automatically keeping low-frequency background information and having small distortion.
Further, the high-frequency feature information parallel projection technology for automatically maintaining the image is to extract feature areas in the image by using the hierarchical information of the edges in the image, perform texture maintenance on each feature area, and directly adopt harmonic mapping for other areas not containing important features, and the specific mapping steps are as follows:
A) obtaining initial texture mapping by adopting harmonic mapping, wherein the two-dimensional texture pictures are uniformly distributed on the surface of the three-dimensional model at the moment, so that the texture mapping of the characteristic region is amplified;
B) and (3) subdividing the image area by utilizing a delaunay triangulation algorithm to obtain a proxy curved surface, and then deforming the initial proxy curved surface by adopting an as-rigid-as-capable algorithm of a mesh deformation algorithm to obtain texture mapping with maintained characteristics.
Further, the automatic splicing of the geometric shapes comprises a Poisson method for obtaining a Boolean operation of a watertight three-dimensional model and performing smooth processing on seams of all parts.
Further, the texture correction after splicing adopts Poisson processing to correct and restore the texture at the seam to obtain a reasonable effect.
The rapid three-dimensional interactive modeling method based on the pictures has the advantages that the two-dimensional cartoon is converted into the three-dimensional model through simple and few interactions, and the method has the biggest characteristic that the view angle does not need to be converted in the whole modeling process, and all modeling operations are completed in two dimensions; the conversion from the two-dimensional cartoon object to the three-dimensional model can be completed by the user without professional knowledge and experience related to three-dimensional modeling; the invention makes a significant contribution in this respect, in particular for the modeling of cartoon objects with a certain perspective.
In addition, the invention also emphasizes on optimizing a texture mapping algorithm, solves the problem brought by parallel projection, can obtain reasonable texture mapping at different visual angles, and simultaneously, specifies the corresponding relation of similar components through simple interaction to recover the texture of the shielding component.
Drawings
FIG. 1 is a block diagram showing the overall structure in the embodiment;
FIG. 2 is a schematic view of automatic judgment and correction of a parting line in the embodiment;
FIG. 3 is a schematic structural diagram of the parts of the cartoon object in the embodiment;
FIG. 4 is a schematic diagram of a three-dimensionally shaped mesh in an embodiment;
FIG. 5 is a diagram illustrating texture mapping of an image according to an embodiment.
Detailed Description
The present invention is not limited to the above-mentioned preferred embodiments, and any other products in various forms can be obtained by anyone in the light of the present invention, but any changes in the shape or structure thereof, which have the same or similar technical solutions as those of the present application, fall within the protection scope of the present invention.
As shown in fig. 1, the rapid three-dimensional interactive modeling method based on pictures described in this embodiment adopts the following steps of importing a colored drawing picture in a deer form or a fox form in the system:
A) utilizing an interactive contour extraction technology to outline the whole part of the colored drawing picture and extracting the peripheral contour line of the image to obtain the foreground and the background of the whole picture;
B) the obtained peripheral contour line of the whole part is sketched by utilizing an interactive contour segmentation technology, and contour segmentation lines between the parts are segmented to obtain the foreground and the background of the image in the whole picture;
C) obtaining two-dimensional contour lines of all parts of the colored drawing picture according to the contour dividing lines;
D) establishing a two-dimensional outline frame of the three-dimensional model according to the two-dimensional outline of the characteristic region decomposition;
E) generating a three-dimensional curved surface for the two-dimensional outline frame by utilizing a three-dimensional free-form surface modeling technology;
F) performing texture mapping of the retention characteristics according to the generated three-dimensional curved surface model of each colored drawing part;
G) and automatically splicing the geometric shapes of the three-dimensional curved surface models of the colored drawing parts after texture mapping and correcting the spliced textures to obtain a complete three-dimensional model.
The interactive contour extraction technology and the interactive contour segmentation technology both adopt the grabcut algorithm of the interactive image segmentation algorithm to obtain the whole picture and the foreground and the background in the picture.
Extracting the foreground and the background of the sketched colored drawing picture to obtain the outline of the two-dimensional target; after the curve of the foreground is drawn in the peripheral contour extraction process, the peripheral contour curve of the background is drawn, so that the contour of the two-dimensional object can be accurately extracted during rapid three-dimensional modeling, and the purpose of starting and ending through modeling is achieved.
As shown in fig. 2 and fig. 3, in this embodiment, when the contour dividing lines are drawn at each part of the colored drawing picture, the offset contour dividing lines are automatically determined and corrected according to the trend of the dividing lines, so as to obtain the contour of the target area, and after the overall contour of the cartoon object is obtained, the overall contour needs to be further divided to obtain the contour of each component. In order to model the hind legs of the fawn, the hind legs of the fawn need to be segmented, and the intention of the user is automatically judged according to the tangential directions of the two ends of the segmentation line to obtain the contour of the hind legs, wherein the contour line of the offset is automatically judged and corrected during the contour segmentation, and the specific steps are as follows:
A) modeling a foreground and a background in the image by adopting a Gaussian mixture model, and predicting the probability of each pixel in the image belonging to the foreground or the background;
B) constructing a network according to the adjacency relation of the pixels;
C) according to the method, a pixel in an image is taken as a node in a network, and the nodes are a source point s and a sink point t respectively;
D) the weight of the edge in the network is obtained through the probability that the corresponding pixel point belongs to the foreground and the color difference between the adjacent pixel points;
F) and (5) completing network construction, and obtaining optimal contour segmentation by adopting a minimum segmentation algorithm.
The method and the device prevent the user from influencing the overall model effect due to deviation when the user uses the contour dividing line to outline, further improve the service performance, enable the three-dimension on the picture to be carried out more quickly, and improve the working efficiency.
As shown in fig. 4, the three-dimensional free-form surface modeling technique according to the present embodiment includes a surface modeling method and a three-dimensional modeling method; and obtaining the side-looking three-dimensional modeling in the image.
In the embodiment, the rapid three-dimensional interactive modeling of the colored drawing picture with the planar sketch and the perspective angle adopts the Fibermesh technology of the three-dimensional free-form surface modeling technology as an algorithm for generating the grid; the Fibermesh technology of the three-dimensional free-form surface modeling technology is used as an algorithm for generating a grid, and the three-dimensional modeling operation steps of the colored drawing picture with perspective angles on a planar sketch are as follows: and drawing a central line on the three-dimensional modeling object in the colored drawing picture, estimating a symmetrical plane of the three-dimensional model according to the central line, and combining the symmetrical plane with the two-dimensional outline frame to obtain a smooth three-dimensional initial model.
According to the three-dimensional modeling algorithm of the Fibermesh technology, firstly, a symmetrical plane of the three-dimensional model is presumed according to a midline of a three-dimensional cartoon object drawn by a user, and an initial three-dimensional model is obtained by combining two-dimensional contour lines. Then the user can adjust the angle of the model, the system can feed the adjusted result back to the user in real time, and the modeling time of a single part is not more than half a minute; the cartoon shape with a certain perspective angle is subjected to three-dimensional modeling by a fibermesh technology through the following equation, and the specific solved equation is as follows:
1)
in this equation, c is the curvature of the curved surface, ciIs the target curvature of the ith vertex, ci' is the current curvature of the ith vertex, and L is the Laplacian on a curved surface. The effect of the first formula is to smooth the curvature of the curved surface.
2)
In this equation, viIs the target coordinate of the ith vertex, vi"is the current coordinate of the ith vertex, C is the constrained region (i.e., vertices inside the region remain unchanged), and L is the Laplace operator on the surface; deltaiIs represented as follows:
δi＝Ai·ci·niwherein A isiVoronoi region for the ith vertex, ciIs the curvature of the ith vertex, niIs the normal vector of the ith vertex.
Solving the equation in an iterative manner in the equations 1) and 2) to obtain a smooth curved surface.
As shown in fig. 5, in this embodiment, the texture mapping includes a high-frequency feature information parallel projection technology for automatically maintaining an image and a continuous mapping method for automatically maintaining low-frequency background information with less distortion, where the high-frequency feature information parallel projection technology for automatically maintaining an image is to extract feature regions in an image by using hierarchical information of edges in the image, perform texture maintenance on each feature region, and directly adopt harmonic mapping for other regions not including important features, and the specific mapping steps are as follows:
A) obtaining initial texture mapping by adopting harmonic mapping, wherein the two-dimensional texture pictures are uniformly distributed on the surface of the three-dimensional model at the moment, so that the texture mapping of the characteristic region is amplified;
B) utilizing a delaunay triangulation algorithm to subdivide an image area to obtain a proxy curved surface, and then adopting an as-rigid-as-posab le algorithm of a mesh deformation algorithm to deform an initial proxy curved surface (figure 5a) to obtain texture mapping (figure 5c) keeping characteristics;
wherein, the step A) is to distribute the two-dimensional texture image on the curved surface uniformly through a harmonic mapping, when the specific implementation steps of the method are mainly implemented by solving a series of differential equations, the specific equations are as follows:
and u isiTexture coordinate u as the ith vertexi' is the target texture coordinate of the ith vertex, L is the Laplace operator on the curved surface, and C is the boundary region (fixing the texture coordinates of the boundary region, solving the texture coordinates inside so that the two-dimensional texture image can be uniformly distributed on the surface of the three-dimensional model)
Wherein, B) a method of image deformation, wherein an as-rigid-applicable algorithm is adopted to deform the proxy curved surface to obtain the final texture mapping for keeping the characteristics.
The as-rigid-applicable algorithm of the mesh deformation algorithm is an algorithm of surface deformation, and the algorithm can deform a surface under the condition of keeping a local rigid structure of the surface as much as possible. The energy function optimized by the algorithm is defined as follows:
in the above formula, S 'is a deformed curved surface, E (S') is the overall rigidity energy of the deformed curved surface, CiIs a neighborhood before the vertex i is deformed, CiBy deforming to obtain Ci＇，wiAnd wijAre corresponding weights, PiIs the coordinate before the deformation of the ith vertex, PiObtaining P by deformationi", N (i) is a neighborhood of vertex i, RiIs a rotation matrix.
The deformation of the local rigid structure can be obtained by minimizing the energy function, so that the two algorithms are combined to solve the texture mapping problem of local feature maintenance;
the texture mapping implemented in the series of steps can ensure that the texture of any part of the object of the three-dimensional model cannot be distorted, and a better texture mapping structure is provided to obtain a more perfect three-dimensional model.
In this embodiment, the automatic splicing of the geometric shapes includes a boolean operation to obtain a watertight three-dimensional model and a poisson method to perform smoothing processing on seams of each part, in this step, we first obtain a watertight three-dimensional model by using the boolean operation, and further, in order to maintain smoothness of seams of the model, we perform smoothing processing on seams of the parts by using the poisson method, so that the three-dimensional model parts of each part are spliced to obtain a complete three-dimensional model.
In this embodiment, the texture at the seam is corrected and restored by poisson for the texture correction after the splicing to obtain a reasonable effect, and in order to avoid that the connection of the textures is cut when the mesh models obtained by respectively expanding each part are finally fused together, that is, the whole model is subjected to texture mapping which is continuous with respect to the original image slices, thereby integrally ensuring the continuity of the texture of each part.
Meanwhile, the surface modeling technology in the invention is divided into two parts: specifically, the side view modeling and the curved surface modeling with a certain perspective angle are adopted, and the curved surface can be constructed by using a fibermesh algorithm through a contour line for a side view part in the whole image. And with a certain perspective angle, a symmetrical plane needs to be estimated according to the combination of a central line and an interaction technology, and simultaneously, a target shape is obtained by utilizing contour line deformation.
The invention can transmit the space geometric information of the object by combining simple stroke combination and drawing and the cognition of people on the space object. Therefore, the simple and effective expression mode has wide application in professional design and daily drawing.
Claims (4)
1. A rapid three-dimensional interactive modeling method based on pictures is characterized by comprising the following steps:
A) utilizing an interactive contour extraction technology to outline the whole part of the colored drawing picture and extracting the peripheral contour line of the image to obtain the foreground and the background of the whole picture;
B) the obtained peripheral contour line of the whole part is sketched by utilizing an interactive contour segmentation technology, and contour segmentation lines between the parts are segmented to obtain the foreground and the background of the image in the whole picture;
C) obtaining two-dimensional contour lines of all parts of the colored drawing picture according to the contour dividing lines;
D) establishing a two-dimensional outline frame of the three-dimensional model according to the two-dimensional outline of the characteristic region decomposition;
E) generating a three-dimensional curved surface for the two-dimensional outline frame by utilizing a three-dimensional free-form surface modeling technology;
F) performing texture mapping of the retention characteristics according to the generated three-dimensional curved surface model of each colored drawing part;
G) according to the three-dimensional curved surface models of the colored drawing parts after texture mapping, automatic splicing of geometric shapes and texture correction after splicing are carried out to obtain a complete three-dimensional model;
the automatic splicing of the geometric shapes comprises a Poisson method for obtaining Boolean operation of a watertight three-dimensional model and performing smooth processing on seams of all parts;
the texture correction after splicing adopts Poisson treatment to correct and restore the texture at the seam to obtain a reasonable effect;
the contour dividing line which is deviated when the contour dividing line is drawn at each part in the colored drawing picture is automatically judged and corrected according to the drawing direction and the tendency of the connecting points of the two ends of the deviated contour dividing line and the peripheral contour line, so that the contour of the target area is obtained;
the specific steps of self-judging and correcting the deviated contour line during contour segmentation are as follows:
A) modeling a foreground and a background in the image by adopting a Gaussian mixture model, and predicting the probability of each pixel in the image belonging to the foreground or the background;
B) constructing a network according to the adjacency relation of the pixels;
C) according to the method, a pixel in an image is taken as a node in a network, and the nodes are a source point s and a sink point t respectively;
D) the weight of the edge in the network is obtained through the probability that the corresponding pixel point belongs to the foreground and the color difference between the adjacent pixel points;
F) completing network construction, and obtaining optimal contour segmentation by adopting a minimum segmentation algorithm;
the texture mapping comprises a high-frequency characteristic information parallel projection technology for automatically keeping an image and a continuous mapping method for automatically changing low-frequency background information and having small distortion;
the high-frequency feature information parallel projection technology for automatically maintaining the image extracts feature areas in the image by utilizing the level information of the edge in the image, carries out texture maintenance on each feature area, and directly adopts harmonic mapping for other areas without important features, wherein the specific mapping steps are as follows:
A) obtaining initial texture mapping by adopting harmonic mapping, wherein the two-dimensional texture pictures are uniformly distributed on the surface of the three-dimensional model at the moment, so that the texture mapping of the characteristic region is amplified;
B) and (3) subdividing the image area by using a delaunay triangulation algorithm to obtain a proxy curved surface, and then deforming the initial proxy curved surface by using an as-rigid-as-capable algorithm of a mesh deformation algorithm to obtain texture mapping keeping characteristics.
2. The picture-based rapid three-dimensional interactive modeling method according to claim 1, characterized in that: the three-dimensional free-form surface modeling technology comprises a surface modeling method and a three-dimensional modeling method.
3. The picture-based rapid three-dimensional interactive modeling method according to claim 1, characterized in that: the method also comprises the step of carrying out rapid three-dimensional interactive modeling on the colored drawing picture with the perspective angle on the plane sketch by adopting a Fibermesh technology of a three-dimensional free-form surface modeling technology as an algorithm for generating the grid.
4. The picture-based rapid three-dimensional interactive modeling method according to claim 3, characterized in that: the Fibermesh technology of the three-dimensional free-form surface modeling technology is used as an algorithm for generating a grid, and the three-dimensional modeling operation steps of the colored drawing picture with perspective angles of a planar sketch are as follows: and drawing a central line on the three-dimensional modeling object in the colored drawing picture, estimating a symmetrical plane of the three-dimensional model according to the central line, and combining the symmetrical plane with the two-dimensional outline frame to obtain a smooth three-dimensional initial model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610656800.8A CN107730587B (en) | 2016-08-11 | 2016-08-11 | Rapid three-dimensional interactive modeling method based on pictures |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610656800.8A CN107730587B (en) | 2016-08-11 | 2016-08-11 | Rapid three-dimensional interactive modeling method based on pictures |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107730587A CN107730587A (en) | 2018-02-23 |
| CN107730587B true CN107730587B (en) | 2021-08-13 |
Family
ID=61199619
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610656800.8A Active CN107730587B (en) | 2016-08-11 | 2016-08-11 | Rapid three-dimensional interactive modeling method based on pictures |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107730587B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108665530B (en) * | 2018-04-25 | 2021-12-14 | 厦门大学 | Three-dimensional modeling implementation method based on single picture |
| CN108898653A (en) * | 2018-06-20 | 2018-11-27 | 广州晖恒广告策划有限公司 | A kind of method for designing three-dimensional cartoon toys based on picture |
| CN109671153B (en) * | 2018-12-05 | 2022-12-30 | 安徽慧软科技有限公司 | Three-dimensional reconstruction method for contour line of complex organ |
| CN111709347A (en) * | 2020-06-10 | 2020-09-25 | 金波 | Hand-drawing modeling interaction system |
| CN112215959A (en) * | 2020-10-19 | 2021-01-12 | 华能阜新风力发电有限责任公司 | Three-dimensional model mapping system using picture cutting |
| CN112508784A (en) * | 2020-11-20 | 2021-03-16 | 中南大学 | Panoramic image method of planar object contour model based on image stitching |
| CN112799517B (en) * | 2021-02-23 | 2022-08-16 | 中国科学院深圳先进技术研究院 | Plant modeling method based on gesture interaction and plant modeling device and equipment thereof |
| CN116824055B (en) * | 2023-06-30 | 2024-06-25 | 深圳市魔数智擎人工智能有限公司 | Picture-based rapid three-dimensional interactive modeling method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101814196A (en) * | 2010-03-09 | 2010-08-25 | 浙江大学 | Method for designing three-dimensional cartoon toys based on pictures |
| CN103559719A (en) * | 2013-11-20 | 2014-02-05 | 电子科技大学 | Interactive graph cutting method |
| CN104574429A (en) * | 2015-02-06 | 2015-04-29 | 北京明兰网络科技有限公司 | Automatic selection method for intersection hot spots in panorama roaming |
-
2016
- 2016-08-11 CN CN201610656800.8A patent/CN107730587B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101814196A (en) * | 2010-03-09 | 2010-08-25 | 浙江大学 | Method for designing three-dimensional cartoon toys based on pictures |
| CN103559719A (en) * | 2013-11-20 | 2014-02-05 | 电子科技大学 | Interactive graph cutting method |
| CN104574429A (en) * | 2015-02-06 | 2015-04-29 | 北京明兰网络科技有限公司 | Automatic selection method for intersection hot spots in panorama roaming |
Non-Patent Citations (4)
| Title |
|---|
| DEFORMING A MAP INTO A HARMONIC MAP;DEANE YANG;《arXiv.org》;20001231;全文 * |
| 交互式虚拟人图像分割系统的设计与实现;邹海滨;《中国优秀硕士学位论文全文数据库 信息科技辑》;20140315;摘要，正文第2.2.4.1-2.2.4.2、2.5节 * |
| 基于调和映射的约束纹理映射方法;郭延文 等;《计算机辅助设计与图形学学报》;20050731;第17卷(第7期);全文 * |
| 基于调和映射的纹理映射方法;吴亚东 等;《计算机工程与应用》;20031031;第39卷(第17期);摘要，正文第2-4节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107730587A (en) | 2018-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107730587B (en) | Rapid three-dimensional interactive modeling method based on pictures | |
| He et al. | Arch++: Animation-ready clothed human reconstruction revisited | |
| CN109003325B (en) | Three-dimensional reconstruction method, medium, device and computing equipment | |
| CN102592275B (en) | Virtual viewpoint rendering method | |
| CN103473806B (en) | A kind of clothes 3 D model construction method based on single image | |
| Li et al. | Robust single-view geometry and motion reconstruction | |
| CN108038906B (en) | Three-dimensional quadrilateral mesh model reconstruction method based on image | |
| CN104616286B (en) | Quick semi-automatic multi views depth restorative procedure | |
| US20150178988A1 (en) | Method and a system for generating a realistic 3d reconstruction model for an object or being | |
| Sýkora et al. | TexToons: practical texture mapping for hand-drawn cartoon animations | |
| Zeng et al. | Region-based bas-relief generation from a single image | |
| CN112767531B (en) | Mobile-end-oriented human body model face area modeling method for virtual fitting | |
| WO2017185730A1 (en) | Method for generating hexahedral mesh based on polycube of closed form | |
| CN101866497A (en) | Binocular stereo vision based intelligent three-dimensional human face rebuilding method and system | |
| WO2011079421A1 (en) | Method for global paremeterization and quadrilateral gridding of point cloud data | |
| CN110197462A (en) | A kind of facial image beautifies in real time and texture synthesis method | |
| CN101303772A (en) | Method for modeling non-linear three-dimensional human face based on single sheet image | |
| US20230086928A1 (en) | Light field reconstruction method and apparatus of a dynamic scene | |
| CN104751508B (en) | The full-automatic of new view is quickly generated and complementing method in the making of 3D three-dimensional films | |
| CN107369204A (en) | A kind of method for recovering the basic three-dimensional structure of scene from single width photo based on deep learning | |
| CN114419297A (en) | 3D target camouflage generation method based on background style migration | |
| CN106296706B (en) | A kind of depth calculation method for reconstructing for combining global modeling and non-local filtering | |
| CN110648394B (en) | Three-dimensional human body modeling method based on OpenGL and deep learning | |
| CN116894914A (en) | Plane clothing visual three-dimensional method and system based on image processing | |
| CN104091318B (en) | A kind of synthetic method of Chinese Sign Language video transition frame |
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 |