A method of avatar rendering

Abstract

The application discloses a virtual image rendering method, and belongs to the technical field of computers. The method comprises the following steps: receiving a virtual image rendering instruction, wherein the virtual image rendering instruction comprises a game character selected by a user; controlling a user terminal to collect point cloud data and color image data of the user; wherein the point cloud data and the color image data are collected under the same visual angle; generating a characteristic three-dimensional image according to a preset characteristic type based on the point cloud data and the color image data; performing splicing processing on the characteristic three-dimensional image and a three-dimensional virtual image corresponding to the game character, and performing rendering on the spliced three-dimensional virtual image to obtain a target three-dimensional virtual image. The application aims to solve the technical problem that the existing game characters cannot be customized individually.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a virtual image rendering method. BACKGROUND

[0002] With the popularization of intelligent devices and the promotion of various applications, users' demand for personalized interaction using 3D images in related application scenarios is increasing.

[0003] At present, in the game application scenario, the 3D images used are all pre-defined and designed by game developers and stored in the game program. In the prior art, the 3D images in the game application scenario are obtained by 3D modeling, and the game character image and style are unified, which cannot meet the personalized needs of game players.

[0004] The above content is only used to assist in understanding the technical solutions of the present application and does not represent the acknowledgement of the above content as prior art. SUMMARY

[0005] The main purpose of the present application is to provide a virtual image rendering method, which aims to solve the technical problem that the existing game characters cannot be personalized.

[0006] To achieve the above purpose, the present application provides a virtual image rendering method, comprising the following steps:

[0007] Receiving a virtual image rendering instruction, the virtual image rendering instruction comprising a game character selected by a user;

[0008] Controlling the user terminal to collect point cloud data and color image data of the user; wherein the point cloud data and the color image data are collected under the same perspective;

[0009] Based on the point cloud data and the color image data, generating a feature three-dimensional image according to a preset feature type;

[0010] Splicing the feature three-dimensional image with a three-dimensional virtual image corresponding to the game character, and rendering the spliced three-dimensional virtual image to obtain a target three-dimensional virtual image.

[0011] Optionally, the step of generating a feature three-dimensional image based on the point cloud data and the color image data according to a preset feature type comprises:

[0012] Based on the preset feature type, extracting feature point cloud data corresponding to the preset feature type from the point cloud data;

[0013] Based on a pre-set mapping relationship, extracting feature points corresponding to the feature point cloud data from the color image data;

[0014] The feature color image data corresponding to the cloud data; wherein, the mapping relationship includes the established mapping relationship between the pixel coordinates of the color image data and the point cloud spatial coordinates of the point cloud data;

[0015] After converting the feature point cloud data into a feature 3D model and the feature color data into texture parameters, the feature 3D model and the texture parameters are spliced ​​together to generate a feature 3D image.

[0016] Optionally, the preset feature type is extracted from the point cloud data.

[0017] Before the steps for obtaining the feature point cloud data corresponding to the feature type, the following steps are also included:

[0018] The point cloud data is subjected to noise reduction processing to obtain noise-reduced point cloud data;

[0019] The step of extracting feature point cloud data corresponding to the preset feature type from the point cloud data based on the preset feature type includes: extracting the preset feature from the denoised point cloud data based on the preset feature type.

[0020] Feature point cloud data corresponding to the type.

[0021] Optionally, the point cloud data is subjected to noise reduction processing to obtain noise-reduced point cloud data.

[0022] The steps include:

[0023] Traverse the point cloud data and divide the point cloud data into multiple point cloud windows;

[0024] For each point cloud window, clustering processing is performed on the point cloud window based on the mean clustering algorithm;

[0025] Based on the clustering results, the point cloud data in the point cloud window is subjected to noise reduction processing to obtain the noise-reduced point cloud data.

[0026] Optionally, the point cloud data in the point cloud window is processed according to the results of the clustering process.

[0027] The steps for noise reduction processing to obtain denoised point cloud data include:

[0028] If the number of cluster centers corresponding to the point cloud window is less than or equal to a preset threshold, then all point clouds in the point cloud window are determined to be valid point clouds.

[0029] If the data of the cluster center corresponding to the point cloud window is greater than the preset threshold, then the center of the point cloud window is determined based on the cluster center corresponding to the point cloud 0 window;

[0030] obtaining a minimum circumscribed frame of the point cloud window, and determining a point cloud inside the point cloud window and outside the minimum circumscribed frame as a noise point cloud; wherein a center of the minimum circumscribed frame is a center of the point cloud window;

[0031] Alternatively, a reference vector is constructed according to a center of the point cloud window and a center of a neighboring point cloud window of the point cloud window;

[0032] For each point cloud in the point cloud window, a target vector is constructed according to the point cloud and the center of the point cloud window;

[0033] If an included angle between the reference vector and the target vector is less than a preset included angle threshold, the point cloud is determined as a valid point cloud;

[0034] If the included angle between the reference vector and the target vector is greater than or equal to the preset included angle threshold, the point cloud is determined as a noise point cloud;

[0035] The noise point cloud is removed to obtain denoised point cloud data.

[0036] Optionally, the step of determining the center of the point cloud window based on the clustering center corresponding to the point cloud window comprises:

[0037] A clustering polygon is constructed based on coordinate information of the clustering center corresponding to the point cloud window; wherein each vertex of the clustering polygon is the clustering center corresponding to the point cloud window;

[0038] A centroid of the clustering polygon is calculated as the center of the point cloud window.

[0039] Optionally, the step of traversing the point cloud data and dividing the point cloud data into a plurality of point cloud windows comprises:

[0040] The point cloud data is sampled by an FPS algorithm to obtain N key points;

[0041] The point cloud data is divided into a plurality of point cloud windows by searching and grouping neighboring points around each key point by a K-neighbor algorithm.

[0042] Optionally, the preset feature type comprises at least one of the following: hairstyle, accessory.

[0043] Optionally, the step of rendering the spliced three-dimensional virtual image to obtain a target three-dimensional virtual image comprises:

[0044] The spliced three-dimensional virtual image is configured with attributes;

[0045] Render the three-dimensional virtual image after attribute configuration to obtain a target three-dimensional virtual image.

[0046] The application provides a virtual image rendering method, compared with the prior art in which a game character cannot be customized, the application receives a virtual image rendering instruction, the virtual image rendering instruction includes a game character selected by a user; a user terminal is controlled to collect point cloud data and color image data of the user; wherein the point cloud data and the color image data are collected under the same angle of view; a feature three-dimensional image is generated according to a preset feature type based on the point cloud data and the color image data; the feature three-dimensional image is spliced with a three-dimensional virtual image corresponding to the game character, and the three-dimensional virtual image after splicing is rendered to obtain a target three-dimensional virtual image. Therefore, in the application, a feature three-dimensional image corresponding to the user can be generated according to the preset feature type, the feature three-dimensional image is spliced with a three-dimensional virtual image corresponding to the game character selected by the user, and a three-dimensional virtual image corresponding to the user can be obtained, and the individual customization of the game character is realized. BRIEF DESCRIPTION OF DRAWINGS

[0047] The accompanying drawings incorporated in and forming a part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application.

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced here, and obviously, other drawings can be obtained by those skilled in the art without creative labor.

[0049] Figure 1 The flowchart of an embodiment of the virtual image rendering method of the present application.

[0050] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0051] It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0052] The application provides a virtual image rendering method, referring to Figure 1 , Figure 1 The flowchart of the first embodiment of the virtual image rendering method of the present application.

[0053] In the embodiment, the virtual image rendering method comprises the following steps:

[0054] Step S10, receiving a virtual image rendering instruction, the virtual image rendering instruction including a game character selected by the user.

[0055] Step S20, controlling the user terminal to collect point cloud data and color image data of the user; wherein the point cloud data and the color image data are collected under the same perspective.

[0056] Step S30, generating a feature three-dimensional image according to a preset feature type based on the point cloud data and the color image data.

[0057] Step S40, splicing the feature three-dimensional image with a three-dimensional virtual image corresponding to the game character, and rendering the spliced three-dimensional virtual image to obtain a target three-dimensional virtual image.

[0058] Compared with the prior art in which a game character cannot be customized, in the embodiment, a feature three-dimensional image corresponding to the user can be generated according to a preset feature type, the feature three-dimensional image is spliced with a three-dimensional virtual image corresponding to the game character selected by the user, and a three-dimensional virtual image corresponding to the user is obtained, thereby realizing the customization of the game character. It should be noted that in the embodiment, only the feature three-dimensional image corresponding to the user needs to be generated, the feature three-dimensional image is only a part of the three-dimensional virtual image corresponding to the user, and the three-dimensional virtual image corresponding to the user is not directly generated, thereby reducing the consumption of the computing resources of the server and enabling the three-dimensional virtual image corresponding to the user to be obtained more quickly.

[0059] The specific steps are as follows:

[0060] Step S10, receiving a virtual image rendering instruction, the virtual image rendering instruction including a game character selected by the user.

[0061] It should be noted that in the embodiment, the virtual image rendering method is applied to the server. The application scenario of the virtual image rendering method is a game application scenario.

[0062] As an example, the permission of the virtual image rendering can be whether the interface permission of accessing the data is opened by the game application developer or operator according to the situation of the game application developer or operator. If the interface permission of accessing the data is opened for free, the user can directly initiate the rendering request of the virtual image. If the interface permission of accessing the data is opened for a fee, the user initiates the rendering request of the virtual image only after completing the payment. That is, the server needs to judge whether the user has paid the relevant fees or other restrictive conditions according to the received user information (ID information or account information of the user), and thereby judge whether the user meets the condition of granting the rendering request of the virtual image.

[0063] As an example, the server is provided with a three-dimensional virtual image repository storing three-dimensional virtual images corresponding to a plurality of game characters. Therefore, the virtual image rendering instruction sent by the user terminal to the server only needs to include the game character selected by the user, and the game character selected by the user occupies a smaller amount of data compared to the three-dimensional virtual image corresponding to it, thereby reducing the consumption of communication resources between the user terminal and the server.

[0064] As an example, the server and the user terminal communicate data through USB protocol, LVDS protocol, MIPI protocol, etc.

[0065] Step S20, control the user terminal to collect the point cloud data and the color image data of the user; wherein the point cloud data and the color image data are collected under the same angle of view.

[0066] As an example, the user terminal is fixedly provided with a laser radar camera and a camera, and there is a fixed transformation relationship between the coordinate system of the laser radar camera and the coordinate system of the camera. Among them, the laser radar camera is used to obtain the point cloud data of the user, and the camera is used to obtain the color image data of the user.

[0067] As an example, the user terminal can also use a three-dimensional scanning device (such as a depth camera) to take three-dimensional photos of the user from multiple angles to obtain the point cloud data of the user.

[0068] As an example, when the user terminal collects the point cloud data and the color image data of the user, the user can be stationary or active, and the shooting place can be anywhere, including but not limited to the office of the game developer, the home of the game designer, the home of the game user, or the game operating place.

[0069] Step S30, generating a feature three-dimensional image according to a preset feature type based on the point cloud data and the color image data.

[0070] As an example, the step of generating a feature three-dimensional image according to a preset feature type based on the point cloud data and the color image data includes:

[0071] Step S31, extracting feature point cloud data corresponding to the preset feature type from the point cloud data based on the preset feature type.

[0072] Among them, the above-mentioned preset feature type includes at least one of the following: hairstyle, accessory. It should be noted that in this embodiment, the accessory includes but is not limited to: glasses, hats, earring, scarves, watches, bracelets, etc.

[0073] As an example, each point cloud in the point cloud data belongs to a different type. On this basis, the feature point cloud data corresponding to the preset feature type can be extracted from the point cloud data according to the belonging type of the point cloud.

[0074] In step S32, the feature color image data corresponding to the feature point cloud data is extracted from the color image data based on a preset mapping relationship. The mapping relationship includes a mapping relationship between the pixel coordinates of the color image data and the point cloud space coordinates of the point cloud data.

[0075] As an example, the mapping relationship between the pixel coordinates of the color image data and the point cloud space coordinates of the point cloud data is established as follows:

[0076] 1) Extract the same position points from the color image data and the point cloud data, and solve the projection transformation matrix according to the calibration model;

[0077] 2) Solve the external conversion parameters of the laser radar unit coordinate system and the camera unit coordinate system according to the projection transformation matrix parameters, complete the conversion of the two coordinate systems, and obtain the mapping relationship between the pixel coordinates of the color image data and the point cloud space coordinates of the point cloud data.

[0078] In step S33, after the feature point cloud data is converted into a feature three-dimensional model and the feature color data is converted into a texture parameter, the feature three-dimensional model and the texture parameter are spliced to generate a feature three-dimensional image.

[0079] As an example, the step of converting the feature point cloud data into a feature three-dimensional model includes:

[0080] Point cloud body construction is performed using surface fitting (such as B-spline surface fitting), a spatial geometric modeling of the preset feature is constructed by a surface modeling method, a three-dimensional rectangular coordinate system is established by taking a feature point of the point cloud body as a coordinate origin, other feature values of the point cloud body are corresponded to the three-dimensional rectangular coordinate system, the point cloud body is divided into a plurality of surface patches, and a feature three-dimensional model with a geometric feature curve is obtained after curve fitting (such as B-spline surface fitting).

[0081] As an example, the texture parameter includes but is not limited to contour, color, material, etc.

[0082] As an example, the step of splicing the feature three-dimensional model and the texture parameter to generate a feature three-dimensional image includes:

[0083] Based on the mapping relationship between the feature point cloud data and the feature color data, the texture parameter is surface-fitted with the feature three-dimensional model to generate a feature three-dimensional image.

[0084] In an embodiment of the present application, before the step of extracting the feature point cloud data corresponding to the preset feature type from the point cloud data based on the preset feature type, the method further comprises:

[0085] The point cloud data is denoised to obtain denoised point cloud data.

[0086] Based on the above, the step of extracting the feature point cloud data corresponding to the preset feature type from the point cloud data based on the preset feature type comprises:

[0087] The feature point cloud data corresponding to the preset feature type is extracted from the denoised point cloud data based on the preset feature type.

[0088] In an example, the step of denoising the point cloud data to obtain denoised point cloud data comprises:

[0089] Step A1, traversing the point cloud data, and dividing the point cloud data into a plurality of point cloud windows.

[0090] Specifically, the step of traversing the point cloud data and dividing the point cloud data into a plurality of point cloud windows comprises:

[0091] The point cloud data is sampled by an FPS algorithm to obtain N key points;

[0092] The neighboring points around each key point are found and grouped by a K-neighbor algorithm, and the point cloud data is divided into a plurality of point cloud windows.

[0093] Step A2, for each point cloud window, based on a mean clustering algorithm, the point cloud window is clustered;

[0094] As an example, based on the mean clustering algorithm, the point cloud window is clustered, comprising:

[0095] 1) initializing K clustering centers, and the K clustering centers are U1, U2, …, Uk respectively;

[0096] 2) all point clouds in the point cloud window are assigned to the nearest cluster set according to the principle of minimum distance, wherein the distance is calculated using Euclidean distance;

[0097] 3) the mean value of the spatial coordinates of all point clouds in each cluster set is taken as a new cluster center;

[0098] 4) repeating steps 1) to 3) until the cluster center no longer changes;

[0099] 5) end, obtaining the k cluster centers corresponding to the point cloud window.

[0100] Step A3, according to the result of the clustering processing, the point cloud data in the point cloud window is denoised to obtain denoised point cloud data.

[0101] As an example, the step of denoising the point cloud data in the point cloud window according to the result of the clustering processing to obtain denoised point cloud data includes:

[0102] If the number of cluster centers corresponding to the point cloud window is less than or equal to a preset threshold, it is determined that the point cloud in the point cloud window is all valid point cloud;

[0103] If the data of the cluster center corresponding to the point cloud window is greater than a preset threshold, the center of the point cloud window is determined based on the cluster center corresponding to the point cloud window;

[0104] The minimum inscribed frame of the point cloud window is obtained, the point cloud inside the point cloud window and outside the minimum inscribed frame is determined as noise point cloud, and the point cloud inside the minimum inscribed frame is determined as valid point cloud; wherein the center of the minimum inscribed frame is the center of the point cloud window;

[0105] Alternatively, a reference vector is constructed according to the center of the point cloud window and the center of its adjacent point cloud window;

[0106] For each point cloud in the point cloud window, a target vector is constructed according to the point cloud and the center of the point cloud window;

[0107] If the included angle between the reference vector and the target vector is less than a preset included angle threshold, the point cloud is determined as valid point cloud;

[0108] If the included angle between the reference vector and the target vector is greater than or equal to a preset included angle threshold, the point cloud is determined as noise point cloud;

[0109] The noise point cloud is removed to obtain denoised point cloud data.

[0110] It should be noted that the step of determining the center of the point cloud window based on the cluster center corresponding to the point cloud window includes:

[0111] A cluster polygon is constructed based on the coordinate information of each cluster center corresponding to the point cloud window; wherein each vertex of the cluster polygon is the cluster center corresponding to the point cloud window;

[0112] The centroid of the cluster polygon is calculated as the center of the point cloud window.

[0113] It should be noted that the reference vector is constructed according to the center of the point cloud window and the center of the adjacent point cloud window, that is, the reference vector is calculated through the coordinate information of the center of the point cloud window and the coordinate information of the center of the adjacent point cloud window.

[0114] It should be noted that the target vector is constructed according to the point cloud and the center of the point cloud window, that is, the target vector is calculated through the coordinate information of the point cloud and the coordinate information of the center of the point cloud window.

[0115] It should be noted that in the embodiment, the preset threshold and the threshold angle threshold can be set according to actual application, which is not limited here.

[0116] As an example, the random noise removal and outlier rejection of the point cloud data can also be realized through the bilateral filtering algorithm, and then the least square method is used to repair the damage in the point cloud body.

[0117] Step S40, splicing the feature three-dimensional image and the three-dimensional virtual image corresponding to the game character, and rendering the spliced three-dimensional virtual image to obtain a target three-dimensional virtual image.

[0118] As an example, the process of splicing the feature three-dimensional image and the three-dimensional virtual image corresponding to the game character includes:

[0119] Obtaining depth data of the feature three-dimensional image and depth data of the three-dimensional virtual image corresponding to the game character;

[0120] Fusing the depth data of the feature three-dimensional image and the depth data of the three-dimensional virtual image corresponding to the game character according to the positional relationship between the feature three-dimensional image and the three-dimensional virtual image corresponding to the game character, to obtain fused depth data;

[0121] Converting the fused depth data into the spliced three-dimensional virtual image.

[0122] Wherein, the specific implementation process of converting the fused depth data into the spliced three-dimensional virtual image is the same as the specific implementation process of converting the feature point cloud data into the feature three-dimensional model, which will not be repeated here.

[0123] As an example, the step of rendering the spliced three-dimensional virtual image to obtain a target three-dimensional virtual image includes:

[0124] Attribute configuration is performed on the spliced three-dimensional virtual image;

[0125] The attribute configured three-dimensional virtual image is rendered to obtain a target three-dimensional virtual image.

[0126] The attributes used in the attribute configuration of the spliced three-dimensional virtual image are pre-designed basic data, which can give the game character the state required by the user, so that the game character is the creation type expected by the user, and has certain functions in the application to perform certain tasks. Since the game character is a character developed by the game developer, its characteristics are not completely satisfactory to the user, so the user can make secondary adjustment to the generated game character (spliced three-dimensional virtual image) according to his own expectations, so that the game character has a virtual-real combined visual effect and meets the user's personalized expectations.

[0127] For example, the attributes used in the attribute configuration of the spliced three-dimensional virtual image can be pre-designed function parameters, which include but are not limited to: behavior parameters, expression parameters, tracking parameters, skill release parameters, etc. For example, in a network game, each original character will contain behavior parameters, tracking parameters, etc. The character model is configured with multiple fusion deformations, each of which can correspond to a polygon mesh or point cloud or any other representation of a geometric three-dimensional surface suitable for a surface such as a body movement or / with a facial expression, so that the original character has activity ability; At the same time, the tracking parameters, i.e. the weights of the fusion deformations, are configured, which are weighted combined to generate body movements or / with facial expressions, so that the original character can better realize certain functions as required by the game plot setting in advance.

[0128] For example, the attributes used in the attribute configuration of the spliced three-dimensional virtual image can also be pre-designed logic parameters, which include but are not limited to: sequencer, selector, circulator and randomizer, etc. The function parameters and condition parameters are combined according to a certain logic flow, so as to give the game character a soul or spirit function, and make it lifelike. For example, in a network game, the game character is movable, and can have expressions and actions, and even can release skills, but not just a simple appearance effect. The game character can be similar to a real person, and even can have a more cool effect, which cannot be separated from the support of logic parameters. Therefore, the game character is more attractive and fascinating, so the user player is more interested in playing the game application.

[0129] In addition, it should be noted that the target three-dimensional virtual image can also be adjusted according to the actual needs of the user. For example, in a network game, if the user feels that the current target three-dimensional virtual image does not conform to or is not liked, all details and large outlines such as hairstyle, hair color, face shape, skin color, even fatness, chest circumference, etc. can be changed to shape according to the user's own aesthetic view to obtain an image effect close to the user's expectation, so as to obtain the above-mentioned final virtual image.

[0130] It should be noted that, in the present document, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0131] The embodiment serial numbers of the present application are only for description, and do not represent the advantages or disadvantages of the embodiments.

[0132] From the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be realized by means of software and necessary general hardware platforms, and of course can also be realized by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a ROM / RAM, a magnetic disk, or an optical disk) and includes a plurality of instructions for causing a terminal device (which can be a mobile phone, a computer, a server, a device, or a network device) to execute the methods described in the various embodiments of the present application.

[0133] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation using the content of the present application specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method of avatar rendering, the method comprising: The virtual image rendering method comprises the following steps: receiving a virtual image rendering instruction, the virtual image rendering instruction comprising a game character selected by a user; controlling a user terminal to collect point cloud data and color image data of the user, wherein the point cloud data and the color image data are collected under the same perspective; sampling the point cloud data through an FPS algorithm to obtain N key points; finding and grouping neighboring points around each key point through a K-neighbor algorithm to divide the point cloud data into N point cloud windows; for each point cloud window, performing clustering processing on the point cloud window based on a mean clustering algorithm; if the number of cluster centers corresponding to the point cloud window is less than or equal to a preset threshold, determining that all the point clouds in the point cloud window are valid point clouds; if the number of cluster centers corresponding to the point cloud window is greater than the preset threshold, determining a center of the point cloud window based on the cluster centers corresponding to the point cloud window; obtaining a minimum inscribed frame of the point cloud window, and determining point clouds outside the minimum inscribed frame as noise point clouds, wherein the center of the minimum inscribed frame is the center of the point cloud window; performing removal processing on the noise point clouds to obtain denoised point cloud data; extracting feature point cloud data corresponding to a preset feature type from the denoised point cloud data based on the preset feature type; extracting feature color image data corresponding to the feature point cloud data from the color image data based on a preset mapping relationship, the mapping relationship comprising a mapping relationship between pixel coordinates of the color image data and point cloud space coordinates of the point cloud data; performing splicing processing on the feature three-dimensional model and the texture parameters after converting the feature point cloud data into a feature three-dimensional model and converting the feature color data into texture parameters, to generate a feature three-dimensional image; performing splicing processing on the feature three-dimensional image and a three-dimensional virtual image corresponding to the game character, and rendering the spliced three-dimensional virtual image to obtain a target three-dimensional virtual image.

2. The avatar rendering method of claim 1, wherein, The step of determining the center of the point cloud window based on the cluster centers corresponding to the point cloud window comprises: constructing a cluster polygon based on coordinate information of the cluster centers corresponding to the point cloud window, wherein each vertex of the cluster polygon is a cluster center corresponding to the point cloud window; calculating the centroid of the cluster polygon as the center of the point cloud window.

3. The avatar rendering method of claim 1, wherein, The preset feature type comprises at least one of the following: hairstyle, accessory.

4. The avatar rendering method of claim 1, wherein, The step of rendering the spliced three-dimensional virtual image to obtain a target three-dimensional virtual image comprises: performing attribute configuration on the spliced three-dimensional virtual image; rendering the three-dimensional virtual image after attribute configuration to obtain a target three-dimensional virtual image.

Images