Grid denoising method and device, three-dimensional scanning system and storage medium

Abstract

The invention relates to a grid denoising method and device, a three-dimensional scanning system and a storage medium, and the method comprises the steps: obtaining all grid patches in a preset neighborhood of a current grid patch in to-be-processed grid data, and obtaining a patch set; performing curved surface fitting based on the patch set to obtain a target curved surface; determining a denoising normal vector of the current grid patch according to the target curved surface, and updating the vertex position of the current grid patch according to the denoising normal vector; and for each grid patch in the to-be-processed grid data, executing the above steps in sequence. According to the method, the curved surface fitting is carried out based on the collected surface patch set, and a more accurate normal vector can be obtained, so that the accurate correction of the vertex of the current grid surface patch is realized at a relatively high speed, and the denoising efficiency and the denoising scale in the grid denoising process are considered.

Description

Technical Field

[0001] The present application relates to the field of three-dimensional scanning, and in particular to a grid denoising method, device, three-dimensional scanning system and storage medium. Background Art

[0002] Mesh data is an effective way to represent point cloud data after 3D scanning, which can more intuitively display the 3D shape of an object. At present, mesh data has been widely used in many 3D application scenarios, such as 3D printing, virtual reality (VR), augmented reality (AR), etc.

[0003] During the 3D scanning process, the mesh data obtained by scanning often contains noise due to the influence of various factors such as equipment accuracy, environmental interference, and surface characteristics of the object. These noises appear as isolated points, burrs, and uneven areas on the mesh model, which will seriously affect the accuracy and reliability of the data. Through denoising, these unnecessary noise points can be removed, making the mesh model smoother and more accurate, thereby improving data quality.

[0004] At present, a two-step denoising framework is usually used, using anisotropic operators such as bilateral filter operators and guided filter operators to filter the patch normal vectors. This method cannot remove large-scale noise such as knife marks. In addition, artificial intelligence-based denoising methods usually extract features from the mesh topology and input them into the neural network to obtain the denoised patch normal vectors, but this method is limited by the complex network structure and has low processing efficiency.

[0005] There is no effective solution to the problem that the efficiency and denoising scale of grid data denoising cannot be balanced in related technologies. Summary of the invention

[0006] In this embodiment, a grid denoising method, device, three-dimensional scanning system and storage medium are provided to solve the problem in the related art that the efficiency and denoising scale of grid data denoising cannot be taken into account at the same time.

[0007] In a first aspect, a grid denoising method is provided in this embodiment, comprising:

[0008] Obtain all mesh patches within a preset neighborhood of a current mesh patch in the mesh data to be processed, and obtain a patch set;

[0009] Performing surface fitting based on the facet set to obtain a target surface;

[0010] Determine a denoised normal vector of the current mesh surface according to the target surface, and update the vertex position of the current mesh surface according to the denoised normal vector;

[0011] For each mesh face in the mesh data to be processed, the above steps are performed in sequence.

[0012] In some of the embodiments, the preset neighborhood is a ring neighborhood having at least 2 rings.

[0013] In some embodiments, performing surface fitting based on the patch set to obtain a target surface includes:

[0014] Establishing a local coordinate system according to the centroid coordinates of each mesh face in the face set;

[0015] In the local coordinate system, polynomial surface fitting is performed based on the centroid coordinates of each mesh patch in the patch set to obtain a target surface.

[0016] In some of the embodiments, a local coordinate system is established according to the centroid coordinates of each mesh face in the face set, including:

[0017] Obtaining the centroid coordinates of each mesh face in the face set to form a centroid set;

[0018] The coordinates of the center point in the center of gravity set are taken as the origin of the coordinate system, and the directions of the coordinate axes are determined based on the center of gravity set to establish a local coordinate system.

[0019] In some embodiments, determining the denoised normal vector of the current mesh patch according to the target surface includes:

[0020] Determine the coordinates corresponding to the center point of the centroid coordinates of each mesh face in the face set on the target surface as the target coordinates;

[0021] The normal vector at the target coordinate is used as the denoised normal vector of the current mesh face.

[0022] In some embodiments, updating the vertex coordinates of the current mesh patch according to the denoised normal vector includes:

[0023] The vertex coordinates of the current mesh face are determined according to the denoised normal vector of the current mesh face, the number of facets in a ring neighborhood of the current mesh face, and the center of gravity of the current mesh face.

[0024] In some embodiments, the method further comprises:

[0025] After updating the vertex coordinates of all mesh facets in the mesh data to be processed, denoised mesh data is obtained.

[0026] In a second aspect, a grid denoising device is provided in this embodiment, including: an acquisition module, a fitting module, an update module and an iteration module; wherein:

[0027] The acquisition module is used to acquire all mesh facets within a preset neighborhood of a current mesh facet in the mesh data to be processed, and obtain a facet set;

[0028] The fitting module is used to perform surface fitting based on the face set to obtain a target surface;

[0029] The updating module is used to determine the denoised normal vector of the current mesh surface according to the target surface, and update the vertex position of the current mesh surface according to the denoised normal vector;

[0030] The iteration module is used to sequentially execute the above steps for each mesh face in the mesh data to be processed.

[0031] In a third aspect, a three-dimensional scanning system is provided in this embodiment, including: a scanning device and a server; the scanning device and the server are connected;

[0032] The scanning device is used to perform three-dimensional scanning on the scanning object to obtain grid data to be processed;

[0033] The server is used to execute the grid denoising method described in the first aspect above.

[0034] In a fourth aspect, in this embodiment, a storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the grid denoising method described in the first aspect is implemented.

[0035] Compared with the related art, the present embodiment provides a mesh denoising method, device, three-dimensional scanning system and storage medium. The mesh denoising method obtains all mesh patches within a preset neighborhood of the current mesh patch in the mesh data to be processed to obtain a patch set; performs surface fitting based on the patch set to obtain a target surface; determines the denoised normal vector of the current mesh patch according to the target surface, and updates the vertex position of the current mesh patch according to the denoised normal vector; and performs the above steps in sequence for each mesh patch in the mesh data to be processed. It performs surface fitting based on the collected patch set, and can obtain a more accurate normal vector, so that the vertices of the current mesh patch can be accurately corrected at a faster speed. Compared with the related art, it takes into account both the denoising efficiency and the denoising scale in the mesh denoising process.

[0036] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0038] Figure 1 is a hardware structure block diagram of a terminal of the grid denoising method of this embodiment;

[0039] Figure 2 is a flow chart of the grid denoising method of this embodiment;

[0040] Figure 3 is a neighborhood diagram of a vertex in this embodiment;

[0041] Figure 4 is a schematic diagram of a patch neighborhood in this embodiment;

[0042] Figure 5 is a flow chart of a mesh denoising method of some of the embodiments;

[0043] Figure 6 is a structural block diagram of the grid denoising device of this embodiment;

[0044] Figure 7 Schematic diagram of the structure of the three-dimensional scanning system of this embodiment. DETAILED DESCRIPTION

[0045] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments.

[0046] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the general meaning understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the", "these" and the like in this application do not represent quantitative restrictions, and they can be singular or plural. The terms "include", "comprise", "have" and any variants thereof involved in this application are intended to cover non-exclusive inclusions; for example, a process, method and system, product or device comprising a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether directly or indirectly. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships. For example, "A and / or B" can mean: A exists alone, A and B exist at the same time, and B exists alone. Usually, the character " / " indicates that the objects associated with each other are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0047] The method embodiment provided in this embodiment can be executed in a terminal, a computer or a similar computing device. For example, running on a terminal, Figure 1 is a hardware structure block diagram of a terminal of the grid denoising method of this embodiment. Figure 1 As shown, the terminal may include one or more ( Figure 1 Only one is shown in the figure) processor 102 and memory 104 for storing data, wherein processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA. The above terminal may also include a transmission device 106 and an input and output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is for illustration only and does not limit the structure of the above terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0048] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the grid denoising method in the present embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0049] The transmission device 106 is used to receive or send data via a network. The above network includes a wireless network provided by the communication provider of the terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0050] In this embodiment, a grid denoising method is provided. Figure 2 is a flow chart of the grid denoising method of this embodiment. Figure 2 As shown, the process includes the following steps:

[0051] Step S210, obtaining all mesh patches within a preset neighborhood of a current mesh patch in the mesh data to be processed, and obtaining a patch set.

[0052] The mesh data to be processed is mesh data obtained based on three-dimensional scanning technology. The mesh data may be composed of a series of interconnected triangles, quadrilaterals or other polygons. The preset neighborhood of the current mesh face may be a neighborhood of a preset number of rings of the vertices of the current mesh face, or a neighborhood of a preset number of rings of the current mesh face, or may be all faces within a preset radius of the current mesh face. The preset number of rings may be 1 or more. Figure 3 is a neighborhood diagram of a vertex in this embodiment, such as Figure 3 As shown in the figure, the area surrounded by the polyline 301 is the first ring neighborhood of point M; the area surrounded by the polyline 302 is the second ring neighborhood of point M. By analogy, the k-ring neighborhood of point M can be defined as follows: the condition is that the shortest path to the point M passes through k vertices; the area surrounded by all points that meet the condition is the k-ring neighborhood of point M.

[0053] Accordingly, Figure 4 is a schematic diagram of a patch neighborhood in this embodiment, such as Figure 4 As shown, the area surrounded by the broken line 401 is the 1-ring neighborhood of the center face patch; the area surrounded by the broken line 402 is the 2-ring neighborhood of the center face patch.

[0054] In this step, the face set composed of all mesh faces can be obtained by searching the neighborhood of the preset number of rings of the current mesh face; or the face set can be obtained by searching the neighborhood of the preset number of rings of the vertices of the current mesh face. Each mesh face in the mesh data to be processed can be taken as the current mesh face f (f∈F) in turn, and all mesh faces within the preset neighborhood of the current mesh face can be searched to obtain the face set N corresponding to the current mesh face. f (including f).

[0055] Step S220, performing surface fitting based on the patch set to obtain a target surface.

[0056] The target surface is constructed by performing surface fitting on the patch set obtained in step S210. Specifically, the surface fitting method may include quadratic polynomial surface fitting, cubic polynomial surface fitting, quartic polynomial surface fitting, or even higher-order polynomial surface fitting. Among them, the higher the degree, the stronger the expression ability of the surface, but the calculation time is also more. Therefore, the degree of polynomial surface fitting can be set according to the needs of the actual application scenario, balancing the surface denoising and the calculation time. In this step, the purpose of performing polynomial surface fitting is to use the surface to realize the denoising of the grid data, thereby realizing the description of the surface of the scanned object.

[0057] In particular, in order to improve the fitting effect and make the target surface obtained after fitting closer to the actual surface condition of the scanned object, mesh patches within a larger neighborhood range may be selected in step S210 to form a patch set. For example, a 2-ring neighborhood, a 3-ring neighborhood, or even a more-ring neighborhood of the current mesh patch may be used as the above-mentioned preset neighborhood, and all mesh patches therein may be obtained to form a patch set.

[0058] Step S230, determining the denoised normal vector of the current mesh face according to the target surface, and updating the vertex positions of the current mesh face according to the denoised normal vector.

[0059] After fitting the target surface, the denoised normal vector of the current mesh patch can be determined based on the target surface. Specifically, the coordinates corresponding to the current mesh patch can be determined on the target surface, and the normal vector at the coordinates can be used as the denoised normal vector of the current mesh patch. Afterwards, the vertex coordinates of the current mesh patch are updated based on the denoised normal vector.

[0060] Step S240, for each mesh face in the mesh data to be processed, the above steps are sequentially performed. The above steps S210 to S230 are sequentially performed for each mesh face in the mesh data to be processed, and finally the denoised mesh data is obtained.

[0061] In the existing related technologies, a two-step denoising framework is usually adopted, using anisotropic operators, such as bilateral filter operators, guided filter operators, etc. to filter the patch normal vectors. This method cannot remove large-scale noise such as knife marks. In addition, artificial intelligence-based denoising methods usually extract features from the mesh topology and input them into the neural network to obtain the denoised patch normal vectors, but this method is limited by the complex network structure and has low processing efficiency. Therefore, in the existing related technologies, it is impossible to achieve a balance between the efficiency of mesh data denoising and the denoising scale.

[0062] In this embodiment, by introducing surface fitting into mesh data denoising, polynomial surface fitting is performed based on a facet set, the fitted surface is used to perform normal vector denoising, and then the mesh facet vertices are updated based on the denoised normal vectors, and finally the denoised mesh data is obtained. Therefore, this embodiment can achieve a mesh data denoising that takes into account both denoising scale and denoising efficiency.

[0063] In the above steps S210 to S240, all mesh patches within the preset neighborhood of the current mesh patch in the mesh data to be processed are obtained to obtain a patch set; surface fitting is performed based on the patch set to obtain a target surface; the denoised normal vector of the current mesh patch is determined according to the target surface, and the vertex position of the current mesh patch is updated according to the denoised normal vector; the above steps are performed in sequence for each mesh patch in the mesh data to be processed. The surface fitting is performed based on the collected patch set, and a more accurate normal vector can be obtained, so that the vertices of the current mesh patch can be accurately corrected at a faster speed. Compared with related technologies, the denoising efficiency and denoising scale in the mesh denoising process are taken into account.

[0064] Among them, in one embodiment, the preset neighborhood is a ring neighborhood with at least 2 rings. In the related art, mesh patches in a one-ring neighborhood of vertices / facets are often used for denoising, and the denoising accuracy is low. In this embodiment, a more accurate denoising normal can be obtained by fitting a polynomial surface using patches in a larger neighborhood, so that large-scale noise such as surface scratches can be dealt with and the denoising scale can be improved. In addition, the denoising method combined with surface fitting can improve the denoising efficiency while ensuring the denoising scale.

[0065] Additionally, in one embodiment, performing surface fitting based on the patch set to obtain a target surface may include:

[0066] A local coordinate system is established according to the centroid coordinates of each mesh face in the face set; in the local coordinate system, polynomial surface fitting is performed based on the centroid coordinates of each mesh face in the face set to obtain a target surface.

[0067] The barycentric coordinates of each mesh face in the face set are extracted to form a barycentric set, and a local coordinate system is constructed based on the barycentric set. Then, in the local coordinate system, polynomial surface fitting is performed using the barycentric coordinates of each barycentric set to obtain a target surface.

[0068] Next, take fitting a quadratic polynomial surface as an example, assuming the surface equation is:

[0069] h(u,v)=au 2 +buv+cv 2 +du+ev+f

[0070] List the linear equations as:

[0071]

[0072] The above linear equations can be simplified to:

[0073] AX=h

[0074] Among them, (u i , v i ,h i ) represents the coordinates of the i-th point on the target surface; (u i , v i ,h i )∈C, i=1,2,…,n. C is the centroid set. a, b, c, d, e, f refer to the fitting coefficients used to describe the fitting surface. Solve using the linear least squares method, that is, solve A T AX=A T h, and the target surface is obtained. In addition, a more accurate surface can be fitted using the moving least squares method (MLS) or the robust moving least squares method (RMLS).

[0075] In this embodiment, a local coordinate system is established based on the patch set, and surface fitting is performed in the local coordinate system, so that an accurate target surface can be obtained, thereby improving the denoising accuracy.

[0076] In one embodiment, a local coordinate system is established according to the centroid coordinates of each mesh face in the face set, which may specifically include:

[0077] The centroid coordinates of each mesh face in the face set are obtained to form a centroid set; the coordinates of the center point in the centroid set are used as the origin of the coordinate system, and the direction of the coordinate axis is determined based on the centroid set to establish a local coordinate system.

[0078] Let the face set be N f From N f Get the centroid of each mesh patch, form a centroid set C, and calculate the center point p of C:

[0079]

[0080] Establish a local coordinate system (u axis , v axis ,h axis ). Among them, the directions of the three axes of the local coordinate system can be calculated by fitting the plane where C is located or using the normal tensor voting method. In this way, the establishment of the local coordinate system is realized.

[0081] Additionally, in one embodiment, determining the denoised normal vector of the current mesh patch according to the target surface may specifically include:

[0082] The coordinates corresponding to the center point of the centroid coordinates of each mesh face in the face set on the target surface are determined as the target coordinates; the normal vector at the target coordinates is used as the denoising normal vector of the current mesh face.

[0083] The coordinates of the center point of the above centroid set on the target surface in the local coordinate system (u p ,v p ,h p ) This point can also be expressed as h(u p , v p ). Calculate h(u p , v p ) at the normal direction n, we get:

[0084] n=h axis -du axis -ev axis

[0085] The calculated n is used as the denoised normal vector of the current mesh face f.

[0086] In this embodiment, the normal vector at the coordinate corresponding to the current mesh patch position on the target surface obtained by fitting is used as the denoised normal vector of the current mesh patch, which can improve the accuracy of denoising the normal vector of the current mesh patch.

[0087] In one embodiment, updating the vertex coordinates of the current mesh patch according to the denoised normal vector may include:

[0088] The vertex coordinates of the current mesh patch are determined according to the denoised normal vector of the current mesh patch, the number of patches in a ring neighborhood of the current mesh patch, and the center of gravity of the current mesh patch.

[0089] The denoised patch normals can be used to iteratively update the vertex coordinates of the patch. For example, for each vertex v in the current mesh patch, i ∈V, the corresponding face set in the neighborhood of a ring is F V (i), the position of the vertex after each iteration is:

[0090]

[0091] Among them, |F V (i)| represents F V (i) The number of patches, n′ k Yes F V The denoised normal of the kth face in (i), c k Indicates the centroid of the kth face. Based on the above process, each mesh face in the mesh data to be processed is used as the current mesh face in turn, and the vertices of the current mesh face are updated after obtaining the denoised normal vector. Finally, after several iterations, the denoised mesh data is obtained.

[0092] In one embodiment, the mesh denoising method further includes: obtaining denoised mesh data after updating the vertex coordinates of all mesh facets in the mesh data to be processed.

[0093] Figure 5 is a flowchart of a grid denoising method of some embodiments, such as Figure 5 As shown, the grid denoising method includes the following steps:

[0094] Step S501, for the current mesh face, search all mesh faces within the preset neighborhood to obtain a face set; wherein, for the current mesh face f∈F, search all mesh faces within the corresponding preset neighborhood to obtain the corresponding face set Nf (including f); in particular, a ring neighborhood of 2 or more rings can be selected as the preset neighborhood.

[0095] Step S502, obtaining the centroid coordinates of each facet from the facet set Nf, forming a centroid set C, and calculating the center point p of C.

[0096] Step S503, establish a local coordinate system with the center point of the centroid set as the origin, and calculate the directions of the three coordinate axes of the local coordinate system by fitting the plane where the centroid set is located or using the normal tensor voting method.

[0097] Step S504: In the local coordinate system, a polynomial surface is fitted using the points in the centroid set to obtain a target surface.

[0098] Step S505: Calculate the corresponding position h(u p ,v p) as the denoised normal vector of the current mesh face f.

[0099] Step S506: using the denoised normal vector, update the vertex positions of the current mesh patch.

[0100] Repeat the above steps S501 to S506 to achieve denoising of all mesh patches in the processed mesh data. This embodiment uses the patch centroids in a larger neighborhood to fit a polynomial surface, and obtains a more accurate denoising vector based on the fitted surface, thereby improving the denoising efficiency while ensuring the denoising scale.

[0101] In this embodiment, a grid denoising device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. The terms "module", "unit", "subunit", etc. used below can implement a combination of software and / or hardware of predetermined functions. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0102] Figure 6 is a structural block diagram of the grid denoising device 60 of this embodiment, as shown in Figure 6 As shown, the grid denoising device 60 includes: an acquisition module 62, a fitting module 64, an update module 66 and an iteration module 68; wherein:

[0103] The acquisition module 62 is used to acquire all mesh patches within a preset neighborhood of the current mesh patch in the mesh data to be processed, and obtain a patch set; the fitting module 64 is used to perform surface fitting based on the patch set to obtain a target surface; the update module 66 is used to determine the denoised normal vector of the current mesh patch according to the target surface, and update the vertex position of the current mesh patch according to the denoised normal vector; the iteration module 68 is used to execute the above steps in sequence for each mesh patch in the mesh data to be processed.

[0104] It should be noted that the above modules can be functional modules or program modules, and can be implemented by software or hardware. For modules implemented by hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0105] In this embodiment, a three-dimensional scanning system is also provided. Figure 7 is a schematic diagram of the structure of the three-dimensional scanning system 70 of this embodiment, as shown in Figure 7As shown, the three-dimensional scanning system 70 includes: a scanning device 72 and a server 74; the scanning device 72 and the server 74 are connected; the scanning device 72 is used to perform three-dimensional scanning on the scanned object to obtain the grid data to be processed; the server 74 is used to execute the grid denoising method provided in the above embodiment. Among them, the scanning device 72 can be a three-dimensional scanner, and the server 74 can be an electronic device with the ability to process three-dimensional scanning data.

[0106] The above-mentioned three-dimensional scanning system performs surface fitting based on the collected facet set, and can obtain more accurate normal vectors, thereby realizing accurate correction of the vertices of the current mesh facet at a faster speed. Compared with related technologies, it takes into account the denoising efficiency and denoising scale in the mesh denoising process.

[0107] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and will not be repeated in this embodiment.

[0108] In addition, in combination with the grid denoising method provided in the above embodiments, a storage medium may be provided in this embodiment to implement the method. The storage medium stores a computer program; when the computer program is executed by a processor, any grid denoising method in the above embodiments is implemented.

[0109] It should be understood that the specific embodiments described herein are only used to explain the application, rather than to limit it. Based on the embodiments provided in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the protection scope of this application.

[0110] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0111] Obviously, the drawings are only some examples or embodiments of the present application. For ordinary technicians in the field, the present application can also be applied to other similar situations based on these drawings without creative work. In addition, it is understandable that although the work done in this development process may be complicated and lengthy, for ordinary technicians in the field, certain changes in design, manufacturing or production based on the technical content disclosed in this application are only conventional technical means and should not be regarded as insufficient content disclosed in this application.

[0112] The term "embodiment" in this application refers to a specific feature, structure or characteristic described in conjunction with the embodiment that can be included in at least one embodiment of the present application. The appearance of this phrase in various locations in the specification does not necessarily mean the same embodiment, nor does it mean that it is mutually exclusive with other embodiments and is independent or optional. It is clearly or implicitly understood by those of ordinary skill in the art that the embodiments described in this application can be combined with other embodiments without conflict.

[0113] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of patent protection. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the scope of protection of the present application. Therefore, the scope of protection of the present application shall be subject to the attached claims.

Claims

1. A grid denoising method, characterized in that: include: Obtain all mesh patches within a preset neighborhood of a current mesh patch in the mesh data to be processed, and obtain a patch set; Performing surface fitting based on the facet set to obtain a target surface; Determine a denoised normal vector of the current mesh surface according to the target surface, and update the vertex position of the current mesh surface according to the denoised normal vector; For each mesh face in the mesh data to be processed, the above steps are performed in sequence.

2. The grid denoising method according to claim 1, characterized in that: The preset neighborhood is a ring neighborhood with at least 2 rings.

3. The grid denoising method according to claim 1, characterized in that: Performing surface fitting based on the patch set to obtain a target surface includes: Establishing a local coordinate system according to the centroid coordinates of each mesh face in the face set; In the local coordinate system, polynomial surface fitting is performed based on the centroid coordinates of each mesh patch in the patch set to obtain a target surface.

4. The grid denoising method according to claim 3, characterized in that: According to the centroid coordinates of each mesh face in the face set, a local coordinate system is established, including: Obtaining the centroid coordinates of each mesh face in the face set to form a centroid set; The coordinates of the center point in the center of gravity set are taken as the origin of the coordinate system, and the directions of the coordinate axes are determined based on the center of gravity set to establish a local coordinate system.

5. The grid denoising method according to claim 3 or claim 4, characterized in that: Determining a denoising normal vector of the current mesh surface according to the target surface includes: Determine the coordinates corresponding to the center point of the centroid coordinates of each mesh face in the face set on the target surface as the target coordinates; The normal vector at the target coordinate is used as the denoised normal vector of the current mesh face.

6. The grid denoising method according to claim 1, characterized in that: Updating the vertex coordinates of the current mesh patch according to the denoised normal vector includes: The vertex coordinates of the current mesh face are determined according to the denoised normal vector of the current mesh face, the number of facets in a ring neighborhood of the current mesh face, and the center of gravity of the current mesh face.

7. The grid denoising method according to claim 1, characterized in that: The method further comprises: After updating the vertex coordinates of all mesh facets in the mesh data to be processed, denoised mesh data is obtained.

8. A grid denoising device, characterized in that: include: Acquisition module, fitting module, update module and iteration module; among them: The acquisition module is used to acquire all mesh facets within a preset neighborhood of a current mesh facet in the mesh data to be processed, and obtain a facet set; The fitting module is used to perform surface fitting based on the face set to obtain a target surface; The updating module is used to determine the denoised normal vector of the current mesh surface according to the target surface, and update the vertex position of the current mesh surface according to the denoised normal vector; The iteration module is used to sequentially execute the above steps for each mesh face in the mesh data to be processed.

9. A three-dimensional scanning system, characterized in that: include: A scanning device and a server; the scanning device and the server are connected; The scanning device is used to perform three-dimensional scanning on the scanning object to obtain grid data to be processed; The server is used to execute the grid denoising method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the mesh denoising method according to any one of claims 1 to 7 are implemented.

Images