Decimation method and apparatus for mesh model, electronic device, and storage medium

Abstract

Disclosed in the embodiments of the present disclosure are a decimation method and apparatus for a mesh model, an electronic device, and a storage medium. The method comprises: reconstructing an original mesh model as a first mesh model having sealing performance; performing decimation on the first mesh model; performing mesh unfolding on a second mesh model obtained by decimation, and determining the coordinates of each vertex of the second mesh model in an unfolded texture image; performing differentiable rendering on the second mesh model on the basis of the coordinates to obtain a target mesh model, wherein loss in a differentiable rendering process comprises first loss constructed on the basis of a rendered image of the currently rendered second mesh model under a preset viewing angle and a preset image, and second loss constructed on the basis of a currently rendered first texture image, wherein the preset image is an image of the original mesh model under the preset viewing angle after differentiable rendering. The method and the related apparatus provided by the embodiments of the present disclosure can ensure texture reducibility while ensuring the accurate topological structure of the mesh model subjected to decimation.

Description

Mesh model decimation method and device, electronic device and storage medium

[0001] This application claims priority to Chinese Patent Application No. 202410425297.X, filed on April 9, 2024, the disclosure of which is incorporated herein in its entirety as part of the present application. TECHNICAL FIELD

[0002] The embodiments of the present disclosure relate to a mesh model decimation method, device, electronic device and storage medium. BACKGROUND

[0003] The mesh model decimation method, for example, can reduce the complexity of the mesh model by removing redundant vertices, edges and faces in the model. With this decimation method, the decimated mesh model has problems such as poor topological accuracy in structure, and problems such as texture blurring and distortion in texture, i.e., both structural details and texture details need to be improved. SUMMARY

[0004] The embodiments of the present disclosure provide a mesh model decimation method, device, electronic device and storage medium, which can ensure the accuracy of the topological structure of the decimated mesh model while ensuring the restoration of the texture.

[0005] At least one embodiment of the present disclosure provides a mesh model decimation method, comprising:

[0006] reconstructing the original mesh model into a first mesh model with airtightness;

[0007] decimating the first mesh model to obtain a second mesh model;

[0008] unfolding the second mesh model to obtain a first texture image, and determining first texture coordinates of each vertex in the second mesh model in the first texture image;

[0009] differentially rendering the second mesh model based on the first texture coordinates to obtain a target mesh model;

[0010] In the differential rendering process, the loss constructed includes a first loss constructed based on a rendered image of the currently rendered second mesh model at a preset perspective and a preset image; a second loss constructed based on a first texture image of the currently rendered second mesh model; and the preset image is an image of the original mesh model at the preset perspective after differential rendering.

[0011] At least one embodiment of the present disclosure also provides a mesh model decimation device, comprising:

[0012] The reconstruction module is configured to reconstruct the original mesh model into a first mesh model with tightness;

[0013] The face reduction module is configured to reduce the faces of the first mesh model to obtain a second mesh model;

[0014] The unfolding module is configured to perform mesh unfolding on the second mesh model to obtain a first texture image, and determine first texture coordinates of each vertex in the second mesh model in the first texture image;

[0015] The rendering module is configured to perform differentiable rendering on the second mesh model based on the first texture coordinates to obtain a target mesh model;

[0016] In the differentiable rendering process, the loss is constructed based on a first loss constructed based on a rendered image of the currently rendered second mesh model at a preset view angle and a preset image, and a second loss constructed based on the first texture image of the currently rendered second mesh model; and the preset image is an image of the original mesh model at the preset view angle after differentiable rendering.

[0017] At least one embodiment of the present disclosure further provides an electronic device, which comprises:

[0018] One or more processors;

[0019] A storage device for storing one or more programs,

[0020] When the one or more programs are executed by the one or more processors, the one or more processors implement the face reduction method of the mesh model provided by at least one embodiment of the present disclosure.

[0021] At least one embodiment of the present disclosure further provides a storage medium containing computer executable instructions, which are used to execute the face reduction method of the mesh model provided by at least one embodiment of the present disclosure when executed by a computer processor. BRIEF DESCRIPTION OF DRAWINGS

[0022] The above and other features, advantages, and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, same or similar reference numerals can represent same or similar elements. It should be understood that the drawings are schematic, and the original and elements are not necessarily drawn according to the scale.

[0023] FIG. 1 is a flowchart of a face reduction method of a mesh model provided by at least one embodiment of the present disclosure;

[0024] FIG. 2 is a schematic diagram of loss construction in a differentiable rendering process in the face reduction method of the mesh model provided by at least one embodiment of the present disclosure;

[0025] FIG. 3 is a flow diagram of a mesh model surface reduction method according to at least one embodiment of the present disclosure;

[0026] FIG. 4 is a structural diagram of a mesh model surface reduction device according to at least one embodiment of the present disclosure;

[0027] FIG. 5 is a structural diagram of an electronic device according to at least one embodiment of the present disclosure. DETAILED DESCRIPTION

[0028] Embodiments of the present disclosure will be described in more detail with reference to the drawings. While certain embodiments of the present disclosure will be shown and described, it is to be understood that the present disclosure is not limited to the embodiments to be shown and described, but is intended to include all embodiments which will become apparent to those skilled in the art upon reading the present disclosure. It is to be understood that the drawings and the embodiments are merely for illustrative purposes and are not intended to limit the scope of the present disclosure.

[0029] It is to be understood that the various steps of the method embodiments of the present disclosure can be performed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit performing the steps shown. The scope of the present disclosure is not limited in this regard.

[0030] The term “comprising” and variations thereof as used in the present disclosure are open-ended, that is, “including but not limited to”. The term “based on” is “based, at least in part, on”. The term “one embodiment” means “at least one embodiment”; the term “another embodiment” means “at least one additional embodiment”; the term “some embodiments” means “at least some embodiments”. Related terms are defined in the following description.

[0031] It should be noted that the terms “first”, “second”, and the like in the present disclosure are used only to distinguish different devices, modules or units, and do not imply the order or interdependence of the functions performed by these devices, modules or units.

[0032] It should be noted that the terms “one”, “multiple”, “a number of” in the present disclosure are illustrative and not limiting, and those skilled in the art should understand that unless otherwise explicitly stated in the context, it should be understood as “one or more”.

[0033] The names of the messages or information exchanged between the devices in the embodiments of the present disclosure are only for illustrative purposes, and are not intended to limit the scope of the messages or information.

[0034] It can be understood that the data involved in the technical solutions of the present disclosure (including but not limited to the data itself, the acquisition or use of the data) should comply with the requirements of the corresponding laws, regulations and relevant provisions.

[0035] FIG. 1 is a flowchart of a mesh model surface reduction method according to at least one embodiment of the present disclosure. The embodiments of the present disclosure are applicable to the case of three-dimensional mesh model surface reduction. The method can be performed by a mesh model surface reduction device, which can be implemented in the form of software and / or hardware, and can be configured in an electronic device, such as a computer.

[0036] As shown in FIG. 1, the mesh model surface reduction method provided by the embodiments of the present disclosure can include:

[0037] S110, reconstructing the original mesh model into a first mesh model with watertight property.

[0038] In at least one embodiment of the present disclosure, the original mesh model can be considered as a three-dimensional mesh model with high face number and complex topology. The watertight property can mean that the mesh model is composed of closed surfaces, and can be understood as that the mesh model does not contain holes and has a clear internal definition.

[0039] In actual application process, due to the design factors of the original mesh model, the original mesh model may not have the watertight property. If the original mesh model is directly reduced in this case, the reduced mesh model is prone to have structural problems such as incorrect holes or edges.

[0040] In at least one embodiment of the present disclosure, the surface of the original mesh model can be reconstructed based on the surface estimation algorithm of the existing isosurface extraction technology to obtain a completely watertight first mesh model. The existing isosurface extraction technology can include but is not limited to the contour tracking method, the cubical method and the marching cubes method, etc.

[0041] In some optional implementation manners, reconstructing the original mesh model into a first mesh model with watertight property can include: obtaining a bounding volume surrounding the original mesh model, and performing cuboid division on the bounding volume; deleting the cuboids outside the original mesh model, and performing cuboid division again on the cuboids intersecting the surface of the original mesh model; in response to the cuboid set satisfying a preset condition, determining the first mesh model based on the cuboid set.

[0042] For example, the bounding volume can include a large cuboid surrounding the original mesh model. For example, the bounding volume can be divided into a plurality of small cuboids (for example, divided into a preset number of cuboids), and the divided cuboids can have three relative positional relationships with the original mesh model, i.e., "outside the original mesh model", "intersecting the surface of the original mesh model" and "inside the original mesh model".

[0043] For example, the cubes outside the original mesh model can be deleted to make the surface topology of the bounding volume close to the original mesh model. Moreover, the cubes intersecting with the surface of the original mesh model can be divided into smaller cubes. Furthermore, referring to the above steps of deleting and dividing the cubes, the smaller cubes can be deleted and divided according to the positional relationship between the smaller cubes and the original mesh model. The operations of deleting and dividing the cubes can gradually adjust the bounding volume to the shape of the original mesh model, that is, obtain the bounding volume whose surface gradually tends to be consistent with the original mesh model.

[0044] For example, the operations of deleting and dividing the cubes can be performed in a loop until the cube set composed of the cubes meets a preset condition, and the loop can be stopped. For example, the preset condition can include, but is not limited to, that the number of times of dividing the cubes in the current cube set reaches a preset number, that the shape loss constructed by the current cube set and the original mesh model is less than a preset threshold, and the like. When the preset condition is met, the division of the cubes can be stopped, and finally a cube set composed of multiple cubes can be obtained, and the cube set has a similar surface topology structure as the original mesh model. In the finally obtained cube set, multiple cubes of different sizes can be included, and each size of the cubes can include at least one. For example, the surface of the finally obtained cube set can be converted into a new mesh model based on the surface, and the first mesh model can be obtained.

[0045] In some optional implementations, by constructing multiple cubes in a three-dimensional space and judging the positional relationship between the cubes and the original mesh model, the connection relationship between the cubes can be estimated to obtain a surface topology structure similar to the shape of the original mesh model. Furthermore, the first mesh model can be generated according to the surface, and the surface reconstruction of the original mesh model can be implemented.

[0046] In at least one embodiment of the present disclosure, since the first mesh model has complete airtightness, the first mesh model can guarantee that the patches have a connection relationship, thereby avoiding the occurrence of incorrect holes and edges in the subsequent decimation process, and ensuring the correctness and accuracy of the mesh model in the topology structure after decimation.

[0047] S120, decimating the first mesh model to obtain a second mesh model.

[0048] In at least one embodiment of the present disclosure, the first mesh model after surface reconstruction generally has a higher number of patches. For example, the first mesh model can be decimated based on existing decimation algorithms, such as vertex merging, vertex deletion, edge collapse, face collapse, and mesh reconstruction, to obtain a mesh model with a target number of faces, i.e., to obtain a second mesh model. For example, the original mesh model can be automatically decimated based on a Quadric Error Metrics (QEM) algorithm by iteratively merging vertices in the original mesh model. The QEM algorithm can quickly and accurately decimate high-face-count complex meshes, effectively reducing the number of vertices while maintaining the overall shape and detailed features of the model, to obtain a more streamlined and efficient mesh model. By using the QEM algorithm, the performance and rendering efficiency of the original mesh model can be improved to meet the needs of various application fields for face number optimization.

[0049] In S130, the second mesh model is mesh unfolded to obtain a first texture image, and first texture coordinates of each vertex in the second mesh model in the first texture image are determined.

[0050] After decimation, the vertex positions of the second mesh model differ greatly from those of the original mesh model. Therefore, in an embodiment of the present disclosure, the second mesh model can be mesh unfolded based on existing mesh unfolding algorithms (such as mesh UV unfolding). For example, the two-dimensional image after mesh unfolding of the second mesh model can be referred to as a first texture image. The image coordinates of each vertex in the second mesh model in the first texture image can be obtained through mapping, which can be referred to as first texture coordinates. By determining the first texture coordinates of each vertex in the second mesh model, a foundation can be laid for subsequent differentiable rendering processes.

[0051] In S140, the second mesh model is differentially rendered based on the first texture coordinates to obtain a target mesh model.

[0052] For example, the loss constructed in the differentiable rendering process includes a first loss constructed based on a rendered image of the currently rendered second mesh model at a preset viewing angle and a preset image; and a second loss constructed based on a first texture image of the currently rendered second mesh model. For example, the preset image is an image of the original mesh model differentially rendered at a preset viewing angle.

[0053] In at least one embodiment of the present disclosure, the differentiable rendering process of the second mesh model can be implemented through an existing neural network structure. For example, the differentiable rendering of the mesh model can be represented based on the following formula: img = DiffRender(v, f, uv).

[0054] Wherein, img can represent a rendering image of the mesh model under a certain view angle; DiffRender(·) can represent a differentiable rendering process function; v, f, uv can represent a vertex, a face to which the vertex belongs, and a texture coordinate of the vertex in the mesh model, respectively.

[0055] Correspondingly, a preset image img of the original mesh model after differentiable rendering under a preset view angle can be obtained based on the vertex, the face to which the vertex belongs, and the texture coordinate of the vertex in the original mesh model. ref A rendering image img of the second mesh model under the preset view angle can be obtained based on the vertex, the face to which the vertex belongs, and the texture coordinate of the vertex of the second mesh model. src .

[0056] For example, the preset view angle can include at least one view angle, and a first loss can be constructed based on the rendering image img under each preset view angle and the preset image img under the corresponding view angle. src ref And, gradient backpropagation can be performed based on the first loss to adjust the first texture image and the texture coordinate, so that the second mesh model (i.e., the target mesh model) after rendering retains the accuracy and restoration of the texture, and is visually very similar to the original mesh model.

[0057] In addition, a second loss can also be constructed according to the first texture image of the currently rendered second mesh model. By constructing the second loss related to the first texture image, it can be ensured that the first texture image of the second mesh model has texture continuity, so that the visual effect of the target mesh model can be significantly improved.

[0058] For example, FIG. 2 is a schematic diagram of loss construction in a differentiable rendering process in a mesh model decimation method provided by at least one embodiment of the present disclosure.

[0059] Referring to FIG. 2, a first loss can be constructed based on the rendering image img and the preset image img src ref . The first loss can include at least one of the following: a color loss MSELoss determined according to the rendering image img and the preset image img src ref ; and a gradient loss GradientLoss determined according to the rendering image img and the preset image img src ref .

[0060] For example, the color loss can be determined by the formula MSELoss = (img src - img ref ) 2 ; and the gradient loss can be determined by the formula GradientLoss = Gradient(img​​​​src )Gradient(img ref ) is determined; wherein Gradient(·) can represent a gradient determination function. By constructing the color loss, the pixel distance between the rendered image img src and the preset image img ref can be measured; by optimizing the color loss, the color of the reduced mesh model can be made as close as possible to the original mesh model before reduction in the overall visual effect. By constructing the gradient loss, the difference between the rendered image img src and the preset image img ref on the image gradient can be measured; by optimizing the gradient loss, the reduced mesh model can be made as close as possible to the original mesh model before reduction in local details (including boundaries, textures, etc.).

[0061] Referring to FIG. 2 again, the second loss can include a texture smoothness loss TexLoss constructed according to the differential value of the first texture image tex src . The texture smoothness loss can be determined by the formula TexLoss = Diff(tex src , std); wherein Diff(·) can represent a differential function, and std can represent a preset coefficient. By constructing the texture smoothness loss, the smoothness of the first texture image can be measured; by optimizing the texture smoothness loss, the rendered image can be ensured to have visual continuity.

[0062] Referring to FIG. 2 again, the loss constructed in the differentiable rendering process can also include a third loss constructed based on the structure information of the currently rendered second mesh model; wherein the third loss can include a structure smoothness loss determined based on the position information of the points and faces of the currently rendered second mesh model.

[0063] By constructing the third loss related to the structure of the currently rendered second mesh model, the structure smoothness of the second mesh model can be ensured. For example, the structure smoothness loss can be composed of at least one of a Laplace loss and a norm loss; wherein the Laplace loss can be represented as LaplaceLoss(v src , f src ), and the norm loss can be represented as NormLoss(v src , f src ), and v src , f src can represent the position information of the vertices and the faces to which the vertices belong in the second mesh model, respectively. By constructing the structure smoothness loss, the smoothness of the second mesh model in structure can be measured; by optimizing the structure smoothness loss, the smoothness of the second mesh model in structure can be ensured, and the visual effect can be improved.

[0064] In at least one embodiment of the present disclosure, the total loss of the differentiable rendering process can include a weighted sum of the first loss, the second loss, and the third loss. The vertex coordinates and the texture information of the second mesh model are updated using the total loss until a preset stopping condition is met, such as a condition that the loss is less than a preset value, so as to obtain a target mesh model. By optimizing the vertex coordinates and the texture color values of the second mesh model based on the end-to-end gradient descent optimization algorithm of the differentiable rendering, the target mesh model obtained by rendering can be very similar to the original mesh model in vision.

[0065] The technical solution of at least one embodiment of the present disclosure provides a complete automatic face reduction link, which can include the steps of surface reconstruction, face reduction, mesh unfolding, texture coordinate determination of vertices, and differentiable rendering of the mesh model. By performing surface reconstruction, any original mesh model can be uniformly processed into a completely closed mesh model. Since the mesh model has complete closeness, it can ensure that the patches of the mesh model have a connection relationship, thereby avoiding the appearance of incorrect holes and edges in the subsequent face reduction process, and ensuring the correctness and accuracy of the mesh model in topology after face reduction. By performing mesh unfolding and texture coordinate determination of vertices on the mesh model after face reduction, a foundation can be laid for subsequent differentiable rendering. By constructing the first loss related to the rendered image in the differentiable rendering process, the similarity of the texture and structure of the mesh model after face reduction and the mesh model before face reduction can be ensured in vision; by constructing the second loss related to the texture image, the first texture image of the second mesh model can have texture continuity. Thus, the topology accuracy of the mesh model after face reduction can be ensured, and the texture restoration of the mesh model can also be ensured.

[0066] The embodiments of the present disclosure can be combined with the optional schemes of the mesh model face reduction method provided in the above embodiments. The mesh model face reduction method provided in the present embodiment optimizes the face reduction link. By mapping each vertex in the first mesh model back to each vertex in the original mesh model before model face reduction, the first mesh model can further fit the original mesh model, and the consistency of the first mesh model and the original mesh model can be improved. By deleting the invisible patches in the second mesh model after model face reduction, the interpenetration of patches during motion can be avoided to some extent, and the number of patches can be further reduced, thereby optimizing the face reduction result.

[0067] FIG. 3 is a flowchart of a mesh model face reduction method provided by at least one embodiment of the present disclosure. As shown in FIG. 3, the mesh model face reduction method provided by the present embodiment can include:

[0068] S310, reconstructing the original mesh model into a first mesh model having closeness.

[0069] S320, position adjustment is performed on each vertex in the first mesh model, so that the position-adjusted vertex has a mapping relationship with each vertex in the original mesh model.

[0070] In at least one embodiment of the present disclosure, the position adjustment of each vertex in the first mesh model can be achieved by mapping the vertex in each cube of the first mesh model back to the original mesh model based on the vertex mapping technology, so that the first mesh model can further fit the original mesh model in structure shape.

[0071] S330, the first mesh model after position adjustment is subtracted to obtain a second mesh model.

[0072] S340, a preset number of light rays are sampled, and based on the intersection of the light rays with the facets in the second mesh model, the facets to be deleted in the second mesh model are determined.

[0073] In at least one embodiment of the present disclosure, first, the brightness value of each facet in the second mesh model can be initially set to 0. Then, based on the point on any facet in the second mesh model and the normal vector corresponding to the facet, a ray is emitted in all directions of the upper hemisphere where the normal line is located, to sample a preset number of light rays. Finally, it can be detected whether the ray intersects with other facets in the second mesh model. If the light ray hits the environment, the brightness of the surface is increased; if it hits other facets, the surface brightness remains unchanged. In this way, the brightness information of each facet in the second mesh model can be obtained. If the brightness of a certain facet is closer to 0, it means that the probability of the facet being visible from all directions is smaller, and the facet can be considered as an external surface approaching to an invisible surface, i.e., as a facet to be deleted.

[0074] S350, the facet to be deleted is deleted to obtain a second mesh model after simplification.

[0075] In at least one embodiment of the present disclosure, the facet to be deleted in the second mesh model can be directly deleted, so that the second mesh model can be ensured not to have the phenomenon of internal surface penetrating into external surface when moving, and the number of facets can be further reduced to improve the quality of the result of facet reduction. Thus, the number of facets can be further reduced without affecting any visual effect.

[0076] S360, the second mesh model after simplification is meshed to obtain a first texture image, and first texture coordinates of each vertex in the second mesh model after simplification in the first texture image are determined.

[0077] S370, the second mesh model after simplification is differentially rendered based on the first texture coordinates to obtain a target mesh model.

[0078] For example, the loss constructed in the differentiable rendering process includes a first loss constructed based on a rendered image of the second mesh model at a preset view angle under current rendering and a preset image; and a second loss constructed based on the first texture image of the second mesh model under current rendering; and the preset image is an image of the original mesh model at the preset view angle after differentiable rendering.

[0079] The technical solution of at least one embodiment of the present disclosure optimizes the face reduction. By mapping each vertex in the first mesh model back to each vertex in the original mesh model before the model face reduction, the first mesh model can further fit the original mesh model, and the consistency of the first mesh model and the original mesh model is improved. By deleting the invisible patches in the second mesh model after the model face reduction, the interpenetration of the patches during the motion can be avoided to some extent, and the number of the patches can be further reduced, so as to optimize the face reduction result. The face reduction method of the mesh model provided in the present embodiment belongs to the same disclosure concept as the face reduction method of the mesh model provided in the above embodiment, and the technical details not described in detail in the present embodiment can be referred to the above embodiment, and the same technical features have the same beneficial effects in the present embodiment and the above embodiment.

[0080] FIG. 4 is a structural schematic diagram of a mesh model face reduction device provided by at least one embodiment of the present disclosure. The mesh model face reduction device provided by the present embodiment is suitable for the case of three-dimensional mesh model face reduction.

[0081] As shown in FIG. 4, the mesh model face reduction device provided by the present embodiment can include:

[0082] The reconstruction module 410 is configured to reconstruct the original mesh model into a first mesh model with airtightness;

[0083] The face reduction module 420 is configured to perform face reduction on the first mesh model to obtain a second mesh model;

[0084] The unfolding module 430 is configured to perform mesh unfolding on the second mesh model to obtain a first texture image, and determine first texture coordinates of each vertex in the second mesh model in the first texture image;

[0085] The rendering module 440 is configured to perform differentiable rendering on the second mesh model based on the first texture coordinates to obtain a target mesh model;

[0086] The loss constructed in the differentiable rendering process includes a first loss constructed based on a rendered image of the second mesh model at a preset view angle under current rendering and a preset image; and a second loss constructed based on the first texture image of the second mesh model under current rendering; and the preset image is an image of the original mesh model at the preset view angle after differentiable rendering.

[0087] In some optional implementation manners, the reconstruction module can be further configured to:

[0088] adjust positions of the vertices in the first mesh model so that the position-adjusted vertices have a mapping relationship with the vertices in the original mesh model.

[0089] In some optional implementation manners, the face reduction module can be further configured to:

[0090] sample a preset number of rays, and determine the face to be deleted in the second mesh model according to intersection of the rays with the faces in the second mesh model;

[0091] delete the face to be deleted to obtain the simplified second mesh model.

[0092] In some optional implementation manners, the reconstruction module can be configured to:

[0093] obtain a bounding volume enclosing the original mesh model, and perform cubic division on the bounding volume;

[0094] delete the cubes outside the original mesh model, and perform cubic division again on the cubes intersecting with the surface of the original mesh model;

[0095] determine the first mesh model based on the set of cubes in response to the set of cubes satisfying a preset condition.

[0096] In some optional implementation manners, the first loss includes at least one of the following:

[0097] a color loss determined according to the rendered image and a preset image;

[0098] a gradient loss determined according to the rendered image and the preset image.

[0099] In some optional implementation manners, the second loss includes a texture smoothness loss constructed according to a differential value of the first texture image.

[0100] In some optional implementation manners, the loss constructed in the differentiable rendering process further includes:

[0101] a third loss constructed based on structure information of the currently rendered second mesh model; the third loss includes a structure smoothness loss determined based on position information of points and faces of the currently rendered second mesh model.

[0102] The face reduction device for a mesh model provided in the embodiments of the present disclosure can perform the face reduction method for a mesh model provided in any of the embodiments of the present disclosure, and has the corresponding function modules and beneficial effects of the execution method.

[0103] It is noted that the units and modules included in the above apparatus are only divided according to functional logic, and are not limited to the above division, as long as the corresponding functions can be implemented; in addition, the specific names of the functional units are only for the convenience of mutual differentiation, and do not serve to limit the protection scope of the embodiments of the present disclosure.

[0104] Reference is made below to FIG. 5, which shows a structural schematic diagram of an electronic device (e.g., a terminal device or a server in FIG. 5) 500 suitable for implementing at least one embodiment of the present disclosure. The terminal device in the embodiments of the present disclosure can include, but is not limited to, a mobile terminal such as a mobile phone, a notebook computer, a digital broadcast receiver, a PDA (Personal Digital Assistant), a PAD (Tablet Personal Computer), a PMP (Portable Multimedia Player), a vehicle terminal (e.g., a vehicle navigation terminal), and the like, and a fixed terminal such as a digital TV, a desktop computer, and the like. The electronic device shown in FIG. 5 is only an example, and should not bring any limitation to the functions and use range of the embodiments of the present disclosure.

[0105] As shown in FIG. 5, the electronic device 500 can include a processing apparatus (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a Read-Only Memory (ROM) 502 or a program loaded from a storage apparatus 508 into a Random Access Memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing apparatus 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An Input / Output (I / O) interface 505 is also connected to the bus 504.

[0106] Generally, the following apparatuses can be connected to the I / O interface 505: an input apparatus 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, and the like; an output apparatus 507 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; a storage apparatus 508 including, for example, a magnetic tape, a hard disk, and the like; and a communication apparatus 509. The communication apparatus 509 can allow the electronic device 500 to communicate with other devices wirelessly or by wire to exchange data. Although FIG. 5 shows the electronic device 500 with various apparatuses, it should be understood that it is not required to implement or have all the shown apparatuses. More or fewer apparatuses can be alternatively implemented or had.

[0107] In particular, according to embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for executing the method shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from the network through the communication apparatus 509, or installed from the storage apparatus 508, or installed from the ROM 502. When the computer program is executed by the processing apparatus 501, the above-mentioned functions defined in the surface reduction method of the mesh model of the embodiments of the present disclosure are executed.

[0108] The electronic device provided by the embodiments of the present disclosure belongs to the same disclosure concept as the surface reduction method of the mesh model provided by the above-mentioned embodiments, and the technical details not described in detail in the present embodiment can be referred to the above-mentioned embodiments, and the present embodiment has the same beneficial effects as the above-mentioned embodiments.

[0109] The embodiments of the present disclosure provide a computer storage medium, which stores a computer program, and the program is executed by a processor to implement the surface reduction method of the mesh model provided by the above-mentioned embodiments.

[0110] It should be noted that the computer-readable medium described above can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium, for example, can be, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any suitable combination of the foregoing. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program used by or in connection with an instruction execution system, apparatus or device. In the present disclosure, the computer-readable signal medium can include a data signal propagated in baseband or propagated as a carrier wave in a propagated data signal, in which the computer-readable program code is contained. Such a propagated data signal can take any of a variety of forms, including, but not limited to, an electromagnetic signal, an optical signal, or any suitable combination of the foregoing. The computer-readable signal medium can also be any computer-readable medium that is not a storage medium and that can communicate, propagate or transport a program for use by or in connection with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including, but not limited to, wire, cable, RF (radio frequency), or the like, or any suitable combination of the foregoing.

[0111] In some embodiments, the client, server, or both can communicate using any current known or future developed network protocol, such as HTTP (Hyper Text Transfer Protocol), and can be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include local area networks ("LAN"), wide area networks ("WAN"), the Internet, and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any current known or future developed networks.

[0112] The computer-readable medium described above can be included in the electronic device; or exist separately from the electronic device, and not be assembled into the electronic device.

[0113] The computer-readable medium described above carries one or more programs, when the one or more programs are executed by the electronic device, cause the electronic device to:

[0114] reducing the first mesh model to obtain a second mesh model; performing mesh unfolding on the second mesh model to obtain a first texture image, and determining first texture coordinates of each vertex in the second mesh model in the first texture image; performing differentiable rendering on the second mesh model based on the first texture coordinates to obtain a target mesh model; wherein the loss constructed in the process of differentiable rendering comprises a first loss constructed based on a rendered image of the current rendered second mesh model under a preset perspective and a preset image; a second loss constructed based on the first texture image of the current rendered second mesh model; wherein the preset image is an image of the original mesh model under the preset perspective after differentiable rendering.

[0115] Computer program code for carrying out operations of the present disclosure can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0116] The flow diagrams and the block diagrams in the drawings are illustrations of architectures, functionalities, and operations of possible implementations of systems, methods, and computer program products according to various embodiments of present disclosure. In this regard, each block in the flow diagrams and the block diagrams can represent a module, a procedure, or a part of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in some cases, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flow diagrams, and combinations thereof, can be implemented by a dedicated hardware-based system that carries out specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0117] The units described in the embodiments of the present disclosure can be implemented in the form of software, or can be implemented in the form of hardware. In some cases, the names of the units and modules do not constitute a limitation on the units and modules themselves.

[0118] The functions described above herein can be performed at least in part by one or more hardware logic components. For example, and without limitation, exemplary types of hardware logic components that can be used include Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Parts (ASSPs), System on Chips (SOCs), Complex Programmable Logic Devices (CPLDs), etc.

[0119] In the context of the present disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable storage media can include, without limitation, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media can include one or more lines of electrical connections, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0120] According to one or more embodiments of the present disclosure, a mesh model reduction method is provided, which includes:

[0121] reconstructing the original mesh model into a first mesh model with tightness;

[0122] reducing the first mesh model to obtain a second mesh model;

[0123] performing mesh unfolding on the second mesh model to obtain a first texture image, and determining first texture coordinates of each vertex in the second mesh model in the first texture image;

[0124] performing differentiable rendering on the second mesh model based on the first texture coordinates to obtain a target mesh model;

[0125] The loss constructed in the differentiable rendering process includes a first loss constructed based on a rendered image of the second mesh model at a preset view angle and a preset image, and a second loss constructed based on a first texture image of the second mesh model at the current rendering.

[0126] In some optional implementations, before the first mesh model is decimated, the mesh model decimation method further includes:

[0127] The positions of the vertices in the first mesh model are adjusted so that the adjusted vertices have a mapping relationship with the vertices in the original mesh model.

[0128] In some optional implementations, after the second mesh model is obtained, the mesh model decimation method further includes:

[0129] A preset number of rays are sampled, and based on the intersection of the rays and the facets in the second mesh model, the facets to be deleted in the second mesh model are determined;

[0130] The facets to be deleted are deleted to obtain the second mesh model after simplification.

[0131] In some optional implementations, reconstructing the original mesh model into the first mesh model with closedness includes:

[0132] An enclosing volume enclosing the original mesh model is obtained, and the enclosing volume is cubed;

[0133] The cubes outside the original mesh model are deleted, and the cubes intersecting the surface of the original mesh model are cubed again;

[0134] In response to the cube set satisfying a preset condition, the first mesh model is determined based on the cube set.

[0135] In some optional implementations, the first loss includes at least one of:

[0136] a color loss determined according to the rendered image and the preset image;

[0137] a gradient loss determined according to the rendered image and the preset image.

[0138] In some optional implementations, the second loss includes a texture smoothness loss constructed according to the differential value of the first texture image.

[0139] In some optional implementations, the loss constructed in the differentiable rendering process further includes:

[0140] a third loss constructed based on structure information of the second mesh model in the current rendering; wherein the third loss comprises a structure smoothness loss determined based on position information of points and faces of the second mesh model in the current rendering.

[0141] According to one or more embodiments of the present disclosure, a mesh model face reduction device is provided, which comprises:

[0142] a reconstruction module configured to reconstruct an original mesh model into a first mesh model with closeness;

[0143] a face reduction module configured to reduce faces of the first mesh model to obtain a second mesh model;

[0144] an unfolding module configured to perform mesh unfolding on the second mesh model to obtain a first texture image, and determine first texture coordinates of each vertex in the second mesh model in the first texture image;

[0145] a rendering module configured to perform differentiable rendering on the second mesh model based on the first texture coordinates to obtain a target mesh model;

[0146] wherein losses constructed in the process of differentiable rendering comprise a first loss constructed based on a rendering image of the second mesh model in the current rendering at a preset view angle and a preset image; and a second loss constructed based on a first texture image of the second mesh model in the current rendering; wherein the preset image is an image of the original mesh model at the preset view angle after differentiable rendering.

[0147] The above description is merely preferred embodiments of the present disclosure and a description of principles of applied technologies. It should be understood by those skilled in the art that the disclosed scope of the present disclosure is not limited to the technical solutions formed by the specific combinations of the above technical features, and also covers other technical solutions formed by any combinations of the above technical features or equivalent features without departing from the above disclosed concepts. For example, the above features can be replaced with technical features disclosed in the present disclosure (but not limited to) having similar functions to form technical solutions.

[0148] Further, while operations are depicted in a particular order, this should not be understood as requiring these operations to be performed in the particular order shown or in sequential order, as some other operations can be performed in parallel or concurrently. Additionally, the various elements depicted can also be implemented in a different order than that which is depicted. Furthermore, no element or component is intended to be dependent on another, unless the element or component is deemed absolutely necessary due to a specific context. Also, specific functional units described in the above specification are meant only to be functional equivalents representing means-plus-function, unless explicitly stated otherwise. Although the above disclosure discusses various exemplary embodiments with reference to the attached drawings, the present disclosure is not limited to the embodiments and the constructive language recited in the specification is not intended to limit the scope of the present disclosure, but rather to what is required by the patent law. Therefore, the scope of the present disclosure should be determined by the following claims.

[0149] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

1. A method for reducing the surface area of ​​a mesh model, comprising: Reconstructing the original grid model into a first grid model with airtightness; Performing face reduction on the first mesh model to obtain a second mesh model; Expanding the second mesh model to obtain a first texture image, and determining first texture coordinates of each vertex in the second mesh model in the first texture image; as well as Performing differentiable rendering on the second mesh model based on the first texture coordinates to obtain a target mesh model; Among them, the loss constructed in the differentiable rendering process includes a first loss constructed based on the rendered image of the currently rendered second mesh model at a preset perspective and a preset image; a second loss constructed based on the first texture image of the currently rendered second mesh model; wherein, the preset image is the image of the original mesh model at the preset perspective after differentiable rendering.

2. The method according to claim 1, wherein Before reducing the first mesh model, the method further includes: The positions of the vertices in the first mesh model are adjusted so that the vertices after the positions are adjusted have a mapping relationship with the vertices in the original mesh model.

3. The method according to claim 1 or 2, wherein: After obtaining the second grid model, the method further includes: Sampling a preset number of rays, and determining to-be-deleted facets in the second mesh model based on intersections between the rays and facets in the second mesh model; and The to-be-deleted facets are deleted to obtain a streamlined second mesh model.

4. The method according to any one of claims 1 to 3, wherein The reconstructing the original grid model into a first grid model having airtightness includes: Obtaining a bounding volume surrounding the original mesh model, and dividing the bounding volume into cubes; Deleting cubes outside the original grid model and re-dividing the cubes intersecting with the surface of the original grid model; and In response to the cube set satisfying a preset condition, a first grid model is determined based on the cube set.

5. The method according to any one of claims 1 to 4, wherein: The first loss includes at least one of the following: a color loss determined based on the rendered image and the preset image; A gradient loss is determined based on the rendered image and the preset image.

6. The method according to any one of claims 1 to 5, wherein: The second loss includes: a texture smoothness loss constructed according to a differential value of the first texture image.

7. The method according to any one of claims 1 to 6, wherein: The losses constructed during the differentiable rendering process also include: A third loss is constructed based on structural information of the currently rendered second mesh model; wherein the third loss includes a structural smoothness loss determined based on position information of points and surfaces of the currently rendered second mesh model.

8. A surface reduction device for a mesh model, comprising: A reconstruction module is configured to reconstruct the original grid model into a first grid model with airtightness; a face reduction module configured to reduce faces of the first mesh model to obtain a second mesh model; an unfolding module configured to perform mesh unfolding on the second mesh model to obtain a first texture image, and determine a first texture coordinate of each vertex in the second mesh model in the first texture image; as well as a rendering module configured to perform differentiable rendering on the second mesh model based on the first texture coordinates to obtain a target mesh model; Among them, the loss constructed in the differentiable rendering process includes a first loss constructed based on the rendered image of the currently rendered second mesh model at a preset perspective and a preset image; a second loss constructed based on the first texture image of the currently rendered second mesh model; wherein, the preset image is the image of the original mesh model at the preset perspective after differentiable rendering.

9. An electronic device comprising: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the mesh model surface reduction method according to any one of claims 1 to 7.

10. A storage medium containing computer-executable instructions, wherein: When executed by a computer processor, the computer executable instructions are used to perform the mesh model surface reduction method according to any one of claims 1 to 7.

Images