CN114529648A - Model display method, device, apparatus, electronic device and storage medium - Google Patents

Abstract

The disclosure provides a model display method, equipment, a device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring octree information corresponding to each three-dimensional model in a target scene; wherein the octree information includes a plurality of nodes; the three-dimensional region represented by each node comprises a local model on the three-dimensional model; the fineness of the local model included by the node is positively correlated with the node depth of the node; the local model with high fineness comprises the number of vertexes, and the number of vertexes is larger than that of the local model with low fineness; determining area information of a visual area of a virtual camera based on current shooting parameter information of the virtual camera for shooting a three-dimensional model in the target scene; and determining a target local model included in at least one node positioned in the visualization region based on the octree information and the region information respectively corresponding to each three-dimensional model included in the target scene.

Description

Model display method, device, device, electronic device and storage medium

technical field

The present disclosure relates to the technical field of three-dimensional models, and in particular, to a model display method, device, device, electronic device, and storage medium.

Background technique

With the development of technology, more and more scenes need to render 3D models; for example, Augmented Reality (AR), Virtual Reality (VR), digital twin projects, etc. Among them, a model file corresponding to each 3D model can be generated, and the model file of each 3D model can be stored in the memory, so that the device uses the model file obtained from the memory to render and display the 3D model.

Generally, the 3D reconstruction algorithm can be used to generate the model file corresponding to the 3D model, and the model file generated by the 3D reconstruction algorithm has a large data capacity. When the size is small, loading the model files of each 3D model into the memory will cause devices with small memory capacity to be unable to render and display each 3D model, which brings inconvenience to the rendering and display of each 3D model.

SUMMARY OF THE INVENTION

In view of this, the present disclosure provides at least a model display method, device, apparatus, electronic device, and storage medium.

In a first aspect, the present disclosure provides a model display method, including:

Acquire octree information corresponding to each 3D model included in the target scene; wherein, the octree information includes multiple nodes; the 3D region represented by each node includes a local model on the 3D model; The fineness of the local model included in the node is positively correlated with the node depth of the node; the number of vertices included in the local model with high fineness is greater than the number of vertices included in the local model with low finesse;

Determine the area information of the visualization area of the virtual camera based on the current shooting parameter information of the virtual camera used for shooting the three-dimensional model in the target scene;

Based on the octree information and the region information respectively corresponding to each three-dimensional model included in the target scene, a target local model included in at least one node located in the visualization region is determined.

In the above method, the octree information corresponding to each three-dimensional model included in the target scene is obtained; wherein, the octree information includes a plurality of nodes, and the three-dimensional area represented by each node includes a local model on the three-dimensional model; The model is divided into multiple partial models, and each partial model corresponds to a node in the octree. After determining the area information of the visualized area of the virtual camera based on the current shooting parameter information of the virtual camera, based on the area information of the current visualized area of the virtual camera and the octree information of each 3D model, it is possible to more efficiently and accurately determine the location of the visual area. The target local model in the visualization area realizes the refined rendering display of each 3D model, and improves the display flexibility of the 3D model.

In a possible implementation manner, the obtaining octree information corresponding to each three-dimensional model included in the target scene includes:

Obtain model information of each 3D model included in the target scene;

Based on the model information corresponding to each of the three-dimensional models, octree information corresponding to the three-dimensional models is generated.

Here, by using the obtained model information of each 3D model included in the target scene, the octree information corresponding to the 3D model can be generated more accurately, which can provide data support for the subsequent determination of the target local model, and realize the detection of each 3D model in the target scene. Refinement display of 3D models.

In a possible implementation manner, generating the octree information corresponding to the three-dimensional model based on the model information corresponding to each of the three-dimensional models includes:

Based on the model information corresponding to each of the three-dimensional models, initial octree information corresponding to the three-dimensional model is generated; wherein, the three-dimensional region represented by each node in the initial octree information includes the corresponding three-dimensional model. The original local model; the original local model includes a plurality of vertices and a plurality of meshes formed by the connection relationship between the vertices;

For each node in the initial octree information, determine the vertex to be deleted in the original local model corresponding to the node, and determine the vertex to be deleted based on the multiple vertices included in the original local model except the vertex to be deleted vertex, generate the local model corresponding to the node;

Based on the local model information of the local model corresponding to each node, the octree information corresponding to the three-dimensional model is generated.

In the embodiment of the present disclosure, for each node in the generated initial octree information, the vertex to be deleted in the original local model corresponding to the node is determined, and the vertex to be deleted is determined based on the multiple vertices included in the original local model except the vertex to be deleted. For other vertices, the local model corresponding to the node is generated. The model accuracy of the local model after vertex deletion is lower than the original local model that was not deleted, which reduces the model accuracy of the model corresponding to the node. The local model information can more accurately generate the octree information corresponding to the 3D model, so that the octree information corresponding to the 3D model can be used to more accurately determine the local model of the target located in the visualization area, and the detailed 3D model can be realized. display.

In a possible implementation manner, generating initial octree information corresponding to the three-dimensional model based on the model information corresponding to each of the three-dimensional models, including:

Taking the three-dimensional model as the target model, based on the model information of the target model, according to the set division method, the three-dimensional area corresponding to the target model is divided into eight sub-three-dimensional areas; wherein, the three-dimensional area of the three-dimensional model The node corresponding to the area is a root node, and the node information of the root node includes the model information of the three-dimensional model; and each sub-three-dimensional area corresponds to a child node; the node information of each child node includes: located in the sub-three-dimensional corresponding to the node Model information or preset information of the original local model in the area;

When there is a child node that satisfies the preset condition, the original local model included in the child node that satisfies the preset condition is used as the target model, and the model information based on the target model is returned. The step of dividing the three-dimensional region corresponding to the target model into eight sub-three-dimensional regions, until the sub-nodes obtained after the division do not meet the preset conditions; wherein, the sub-nodes that meet the preset conditions include: nodes whose depth is less than the set depth threshold. Nodes, and/or, child nodes whose number of vertices in the model information included in the node information is greater than the set number threshold;

Based on the generated node information of the root node and the node information of each child node, initial octree information corresponding to the three-dimensional model is generated.

Here, by taking the 3D model as the target model, based on the model information of the target model, according to the set division method, the 3D region corresponding to the target model is divided into eight sub-3D regions; When the sub-nodes are divided into sub-nodes, the sub-nodes are divided until none of the sub-nodes obtained after division meet the preset conditions, and the node information of the root node and the node information of each sub-node are obtained. Furthermore, based on the node information of the root node and the node information of each child node, the initial octree information corresponding to the three-dimensional model is more accurately generated.

In a possible implementation manner, the determining of the vertex to be deleted in the original local model corresponding to the node includes:

Based on the model information of the original local model included in the node, determine the normal similarity between the normals corresponding to the multiple meshes indicated by the model information;

The vertex to be deleted in the original local model corresponding to the node is determined based on the normal similarity between the normals corresponding to the multiple meshes respectively.

Here, based on the model information of the original local model corresponding to the node, the normal similarity between the normals corresponding to the multiple grids indicated by the model information is determined; The normal similarity can easily determine the vertices to be deleted from the multiple vertices of the original local model; then based on the other vertices in the multiple vertices except the vertices to be deleted, the local model corresponding to the node is generated to realize the original local model. Simplify, improve the efficiency of 3D model simplification.

In a possible implementation manner, determining the vertex to be deleted in the original local model corresponding to the node based on the normal similarity between the normals corresponding to the multiple meshes, including:

Traversing the vertices of the original local model, and determining the target mesh including the traversed vertices for the traversed vertices;

Compare the normal similarity between the normals corresponding to the target grid and the normal similarity threshold corresponding to the node; wherein, the normal similarity threshold corresponding to the node corresponds to the node The node depth of is positively correlated;

In response to the normal similarity being greater than or equal to the normal similarity threshold, the traversed vertex is determined as a vertex to be deleted.

In this way, the boundary of vertex deletion is limited by a numerically controllable normal similarity threshold, so that the fineness of the simplification of the original local model is controllable, while the accuracy of the simplified local model is guaranteed, and the simplification of the model is improved at the same time. efficiency.

In a possible implementation manner, the target local model included in at least one node located in the visualization region is determined based on the octree information and the region information corresponding to each three-dimensional model included in the target scene. ,include:

determining at least one target three-dimensional model located in the visualization area based on the area information, the displayed pose and model size of each three-dimensional model included in the target scene;

Based on the octree information corresponding to the target three-dimensional model and the area information, a target local model included in at least one node in the visualization area is determined.

In the embodiment of the present disclosure, at least one target 3D model located in the visualization area is determined based on the area information, the displayed pose and model size of each 3D model included in the target scene; and then according to the octree information corresponding to the target 3D model and area information, the target local model included in at least one node in the visualization area can be more accurately determined, and the determination process of the target local model is relatively simple and efficient.

In a possible implementation manner, the determining, based on the octree information corresponding to the target three-dimensional model and the area information, determines the target local model included in at least one node in the visualization area, including:

Taking the root node in the octree information corresponding to the target three-dimensional model as the node to be processed, it is determined whether the three-dimensional area corresponding to the node to be processed is completely located in the visualization area indicated by the area information;

In the case that the three-dimensional area corresponding to the node to be processed is partially located in the visualization area, take each child node connected to the node to be processed as a node to be processed, and return to determining whether the three-dimensional area corresponding to the node to be processed is not The step of being completely located in the visualization area indicated by the area information, until the three-dimensional area corresponding to the node to be processed is completely located in the visualization area;

When the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, the local model included in the node to be processed is determined as the target local model included in the node located in the visualization area.

In the above manner, by judging whether the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, when the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, the local model included in the node to be processed is determined as the target local model, Improves the accuracy and efficiency of the target local model.

In a second aspect, the present disclosure provides a model display device, including: a central processing unit (CPU) and a graphics processing unit (GPU);

The CPU is configured to store octree information corresponding to each three-dimensional model included in the target scene, and after determining the target local model included in at least one node, transmit the model information of the target local model included in the node to the GPU; wherein, the target local model included in the at least one node is determined by using the model display method described in the first aspect or any implementation method;

The GPU is configured to store the received model information of the target local model included in the node in the internal memory; and by reading the model information in the internal memory, in the shooting picture of the virtual camera, The target local model is rendered and displayed.

Here, by using the model display method described in the first aspect or any one of the embodiments, after the target local model is more accurately determined, the model information of the target local model can be sent to the GPU, and there is no need to send the models of other local models to the GPU. information, reducing the amount of information transmission, alleviating the consumption of bandwidth resources when sending model information of other local models that do not need to be rendered to the GPU, and improving the utilization of bandwidth resources. At the same time, the model information of the target local model is loaded into the internal memory (memory), and there is no need to load the model information of other local models into the memory, which reduces memory overhead and improves memory utilization.

In a possible implementation manner, the CPU is further configured to send the updated model information of the target local model to the GPU after the target local model is updated;

The GPU is also used to delete the model information that does not match the updated target partial model in the internal memory; and, in the updated target partial model, the model of the unstored target partial model Information is stored in the internal memory.

In the above embodiment, after the target local model is updated, the model information of the updated target local model can be loaded into the memory, and the model information in the memory that does not match the updated target local model is deleted, so as to ensure that what is loaded in the memory is correct. Currently, the model information of the target local model needs to be rendered and displayed to avoid the waste of memory resources.

For descriptions of the effects of the following apparatuses, electronic devices, etc., reference may be made to the descriptions of the above-mentioned methods, which will not be repeated here.

In a third aspect, the present disclosure provides a model display device, comprising:

The acquisition module is used for acquiring octree information corresponding to each three-dimensional model included in the target scene; wherein, the octree information includes a plurality of nodes; the three-dimensional area represented by each node includes the information on the three-dimensional model. The fineness of the local model included in the node is positively correlated with the node depth of the node; the number of vertices included in the local model with high fineness is greater than the number of vertices included in the local model with low fineness;

a first determining module, configured to determine the area information of the visualized area of the virtual camera based on the current shooting parameter information of the virtual camera used for shooting the three-dimensional model in the target scene;

The second determination module is configured to determine, based on the octree information and the region information respectively corresponding to each three-dimensional model included in the target scene, a target local model included in at least one node located in the visualization region.

In a fourth aspect, the present disclosure provides an electronic device, comprising: a processor, a memory, and a bus, the memory stores machine-readable instructions executable by the processor, and when the electronic device runs, the processor communicates with the The memories communicate with each other through a bus, and when the machine-readable instructions are executed by the processor, the steps of the model presentation method according to the first aspect or any one of the implementation manners are executed.

In a fifth aspect, the present disclosure provides a computer-readable storage medium on which a computer program is stored, and when the computer program is run by a processor, executes the model described in the first aspect or any one of the embodiments above. Show the steps of the method.

In order to make the above-mentioned objects, features and advantages of the present disclosure more obvious and easy to understand, the preferred embodiments are exemplified below, and are described in detail as follows in conjunction with the accompanying drawings.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present disclosure more clearly, the accompanying drawings required in the embodiments will be briefly introduced below.

FIG. 1 shows a schematic flowchart of a model display method provided by an embodiment of the present disclosure;

2 shows a schematic diagram of a three-dimensional region, a sub-three-dimensional region, and an octree corresponding to a target model in a model display method provided by an embodiment of the present disclosure;

FIG. 3 shows a schematic diagram of an original local model in a model display method provided by an embodiment of the present disclosure;

FIG. 4 shows a schematic diagram of a sub-3D area and a visualization area in the model display method provided by an embodiment of the present disclosure;

5 shows a schematic diagram of another sub-three-dimensional area and a visualization area in the model display method provided by an embodiment of the present disclosure;

6 shows a schematic diagram of another sub-three-dimensional area and a visualization area in the model display method provided by an embodiment of the present disclosure;

FIG. 7 shows a schematic structural diagram of a model display device provided by an embodiment of the present disclosure.

FIG. 8 shows a schematic structural diagram of a model display apparatus provided by an embodiment of the present disclosure.

FIG. 9 shows a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure.

Detailed ways

To make the purposes, technical solutions, and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present disclosure.

For each 3D model, a 3D reconstruction algorithm can be used to generate a model file corresponding to the 3D model, and the model file of the 3D model generated by the 3D reconstruction algorithm has a large data capacity. When the memory capacity of the device for rendering display is small, loading the model files of each 3D model into the memory will cause the device with small memory capacity to be unable to render and display each 3D model, which brings inconvenience to the rendering and display of each 3D model.

Generally, for each 3D model, models of various precisions can be constructed, such as high-precision model, medium-precision model, low-precision model, etc.; among them, for the same model, the higher the accuracy of the model, the higher the accuracy required to construct the 3D model. More triangles. Furthermore, three-dimensional models with different precisions can be selected for rendering according to the distance between the three-dimensional model and the virtual camera used to photograph the three-dimensional model. For example, when the three-dimensional model is far away, a low-precision three-dimensional model can be selected for rendering display, and when the three-dimensional model is relatively close, a high-precision three-dimensional model can be selected for rendering display. However, by constructing 3D models with various precisions for rendering and display, the construction of 3D models is cumbersome, and when the 3D model is rendered and displayed, the model files of the 3D models with various precisions need to be loaded into the memory, increasing the size of the 3D model. memory overhead.

To alleviate the above problems, embodiments of the present disclosure provide a model display method, device, apparatus, electronic device, and storage medium.

It should be noted that like numerals and letters refer to like items in the following figures, so once an item is defined in one figure, it does not require further definition and explanation in subsequent figures.

In order to facilitate the understanding of the embodiments of the present disclosure, a model display method disclosed by the embodiments of the present disclosure is first introduced in detail. The execution body of the model display method provided by the embodiment of the present disclosure is generally a computer device with certain computing capabilities, and the computer device includes, for example, a terminal device or a server; the server may be, for example, a local server, a cloud server, or the like. The terminal device may be, for example, a mobile phone, a tablet, augmented reality (Augmented Reality, AR) glasses, a personal digital assistant (Personal Digital Assistant, PDA) and other devices. In some possible implementations, the model presentation method may be implemented by a processor invoking computer-readable instructions stored in a memory.

Referring to FIG. 1, which is a schematic flowchart of a model display method provided by an embodiment of the present disclosure, the method includes: S101-S103, wherein:

S101, obtain octree information corresponding to each three-dimensional model included in the target scene; wherein, the octree information includes a plurality of nodes; the three-dimensional region represented by each node includes a local model on the three-dimensional model; the local model included in the node The fineness of the node is positively correlated with the node depth of the node; the number of vertices included in the local model with high fineness is greater than the number of vertices included in the local model with low fineness;

S102, based on the current shooting parameter information of the virtual camera used for shooting the three-dimensional model in the target scene, determine the area information of the visualization area of the virtual camera;

S103: Determine a target local model included in at least one node located in the visualization region based on the octree information and region information respectively corresponding to each three-dimensional model included in the target scene.

In the above method, the octree information corresponding to each three-dimensional model included in the target scene is obtained; wherein, the octree information includes a plurality of nodes, and the three-dimensional area represented by each node includes a local model on the three-dimensional model; The model is divided into multiple partial models, and each partial model corresponds to a node in the octree. After determining the area information of the visual area of the virtual camera based on the current shooting parameter information of the virtual camera, based on the area information of the current visual area of the virtual camera and the octree information of each 3D model, it is possible to more efficiently and accurately determine the location of the visual area. The target local model in the visualization area realizes the refined rendering display of each 3D model, and improves the display flexibility of the 3D model.

S101-S103 will be specifically described below.

For S101:

The three-dimensional model may include models corresponding to real objects in the target scene. For example, the target scene may be a school scene, a shopping mall scene, an office scene, etc., and the three-dimensional model may include a teaching building included in the school scene, and models of real objects such as desks and chairs in the teaching building. Alternatively, the three-dimensional model may also include constructed virtual objects, for example, the three-dimensional model may include constructed virtual trainers, virtual animals, and the like.

The octree information corresponding to each three-dimensional model includes a plurality of nodes, each node corresponds to a three-dimensional area, and the three-dimensional area includes a local model on the three-dimensional model. The fineness of the local model included in the node is positively correlated with the node depth of the node, that is, the greater the node depth, the higher the fineness of the local model included in the three-dimensional region represented by the node. Among them, for the same local model, the number of vertices included in a local model with a high degree of detail is greater than the number of vertices included in a local model with a low degree of refinement; The triangular patch included in the model is composed of a plurality of vertices and the connection relationship between the vertices.

During implementation, a three-dimensional reconstruction algorithm may be used to generate a model file corresponding to each three-dimensional model in the target scene, and the three-dimensional model may be a high-precision model. After each three-dimensional model in the target scene is constructed, the octree information corresponding to the three-dimensional model can be determined for each three-dimensional model. Further, the octree information corresponding to each three-dimensional model can be obtained.

In a possible implementation manner, acquiring the octree information corresponding to each 3D model included in the target scene may include:

Step A1, obtaining model information of each three-dimensional model included in the target scene;

Step A2, based on the model information corresponding to each three-dimensional model, generate octree information corresponding to the three-dimensional model.

Generally, a three-dimensional model may include: a plurality of meshes; wherein any mesh of the multiple meshes has a common vertex with at least one other mesh. Multiple meshes are connected to each other to form a 3D model. where each mesh is a triangular patch.

During implementation, model information of each three-dimensional model included in the target scene may be acquired first, where the model information includes vertex coordinates of multiple vertices and information on connection edges between vertices.

Then, for each three-mode model, based on the model information corresponding to the three-dimensional model, the octree information corresponding to the three-dimensional model can be generated. For example, the 3D model can be used as the target model, and based on the model information of the target model, the target model can be divided into eight local models according to the set segmentation direction, and the model information of each local model can be obtained; then each local model can be used as the target. model, return to the step of dividing the target model into eight local models according to the set dividing direction, until the number of divisions is equal to the set number of times threshold, and/or, until the volume of the divided local models is less than or equal to the set volume threshold, etc. According to the model information of each partial model obtained after division, the octree information corresponding to the three-dimensional model is generated. Or, perform vertex deletion processing on at least one local model obtained after the division, and obtain model information of the deleted local model; information to generate the octree information corresponding to the 3D model.

Here, by using the obtained model information of each 3D model included in the target scene, the octree information corresponding to the 3D model can be generated more accurately, which can provide data support for the subsequent determination of the target local model, and realize the detection of each 3D model in the target scene. Refinement display of 3D models.

In a possible implementation manner, in step A2, based on the model information corresponding to each three-dimensional model, the octree information corresponding to the three-dimensional model is generated, which may include:

Step A21, based on the model information corresponding to each three-dimensional model, generate initial octree information corresponding to the three-dimensional model. Wherein, the three-dimensional region represented by each node in the initial octree information includes the original local model corresponding to the three-dimensional model; the original local model includes multiple vertices and multiple meshes formed by connection relationships between the vertices.

Step A22, for each node in the initial octree information, determine the vertex to be deleted in the original local model corresponding to the node, and based on the other vertices except the vertex to be deleted in the multiple vertices included in the original local model, generate the corresponding vertex of the node. the local model.

Step A23: Generate octree information corresponding to the three-dimensional model based on the local model information of the local model corresponding to each node.

In the embodiment of the present disclosure, for each node in the generated initial octree information, the vertex to be deleted in the original local model corresponding to the node is determined, and the vertex to be deleted is determined based on the multiple vertices included in the original local model except the vertex to be deleted. For other vertices, the local model corresponding to the node is generated. The model accuracy of the local model after vertex deletion is lower than the original local model that was not deleted, which reduces the model accuracy of the model corresponding to the node. The local model information can more accurately generate the octree information corresponding to the 3D model, so that the octree information corresponding to the 3D model can be used to more accurately determine the local model of the target located in the visualization area, and the detailed 3D model can be realized. display.

In step A21, after acquiring the model information of the three-dimensional model, initial octree information corresponding to the three-dimensional model may be generated based on the model information corresponding to the three-dimensional model; wherein the initial octree information includes a plurality of nodes, and each The three-dimensional region represented by each node includes an original local model corresponding to the three-dimensional model, and the original local model is a model with vertices not deleted. That is, the model accuracy of the original local model corresponding to each node in the initial octree information is the same.

For example, the three-dimensional model can be divided according to the set segmentation direction, and a plurality of original partial models and model information of the original partial models can be obtained. Then, according to the model information of the multiple original local models, the octree information corresponding to the three-dimensional model is generated.

In a possible implementation, in step A21, based on the model information corresponding to each three-dimensional model, the initial octree information corresponding to the three-dimensional model is generated, including:

Step A211, taking the three-dimensional model as the target model, and dividing the three-dimensional area corresponding to the target model into eight sub-three-dimensional areas based on the model information of the target model and according to the set division method.

Wherein, the node corresponding to the three-dimensional area of the three-dimensional model is the root node, and the node information of the root node includes the model information of the three-dimensional model; and each sub-three-dimensional area corresponds to a child node; the node information of each child node includes: Model information or preset information of the original local model in the area.

Step A212, when there is a child node that satisfies the preset condition, take the original local model included in the child node that satisfies the preset condition as the target model, return to the model information based on the target model, and divide the target model according to the set division method. The step of dividing the corresponding three-dimensional region into eight sub-three-dimensional regions, until none of the sub-nodes obtained after the division meet the preset conditions.

The child nodes satisfying the preset conditions include: child nodes whose node depth is less than a set depth threshold, and/or child nodes whose number of vertices in the model information included in the node information is greater than the set number threshold.

Step A213, based on the generated node information of the root node and the node information of each child node, generate initial octree information corresponding to the three-dimensional model.

The 3D model is taken as the target model, and based on the model information of the target model, the 3D region corresponding to the target model is divided into eight sub-3D regions according to the set division method. The three-dimensional area corresponding to the target model may be a three-dimensional area where a three-dimensional detection frame surrounding the target model is located. Alternatively, the three-dimensional region corresponding to the target model may also be the three-dimensional region where the contour of the target model is located.

The division manner may include multiple cutting directions, for example, may include a horizontal direction, a vertical direction, and a vertical direction, and any two of the horizontal direction, the vertical direction, and the vertical direction are perpendicular to each other. Wherein, the set cutting direction can be referred to as shown in FIG. 2 .

Referring to Figure 2, the three-dimensional model is used as the target model, and the three-dimensional area corresponding to the target model (that is, the three-dimensional area where the three-dimensional detection frame of the target model is located) is divided into 8 sub-three-dimensional areas along the multiple cutting directions indicated by the division method. area. Wherein, the node corresponding to the three-dimensional area of the three-dimensional model is the root node, and the node information of the root node includes the model information of the three-dimensional model, that is, node 1 in the octree on the right side of FIG. 2 is the root node. And each child three-dimensional area corresponds to a child node, namely child node 2, child node 3, ..., child node 9; the node information of each child node includes: the model information of the original local model located in the child three-dimensional area corresponding to the node, or preset information. When the original local model does not exist in the sub-three-dimensional region corresponding to the node, the node information of the node may be set as preset information, and the preset information may be, for example, 0 or the like.

After obtaining multiple sub-nodes, for each sub-node, determine whether the sub-node satisfies the preset condition. If the preset condition is met, the original local model included in the sub-node is used as the target model, and the model based on the target model is returned to the information, according to the set division method, the step of dividing the three-dimensional area corresponding to the target model into eight sub-three-dimensional areas. For example, if it is determined that the child node 4 in FIG. 2 satisfies the preset condition, the original local model included in the child node 4 is used as the target model, and the process returns to step A211. If it is determined that the child node 2 in FIG. 2 does not meet the preset condition, no subsequent processing is performed, that is, the child node 2 does not need to be segmented.

The preset condition may include that the node depth is less than a set depth threshold, and/or the number of vertices in the model information included in the node information is greater than the set number threshold. For example, the node depth of the child node is less than the set depth threshold, and/or the number of vertices in the model information included in the node information of the child node is greater than the set number threshold. The node depth of the child node is related to the layer number of the child node in the octree. For example, as shown in Figure 2, the node depth of the root node 1 can be 0, and the node depth of the child node 2 to the child node 9 can be 1. The node depth of child node 10 to child node 25 may be 2.

The depth threshold and the quantity threshold can be set according to the actual situation. For example, the depth threshold can be 4, 5, 6, etc., and the quantity threshold can be 5, 8, etc.

During implementation, if the node information of the child node includes preset information, it is determined that the child node does not meet the preset condition, and the child three-dimensional region corresponding to the child node will not be divided. If the node information of the sub-node includes the model information of the original local model located in the sub-3D region corresponding to the sub-node, it is judged whether the sub-node satisfies the preset condition, and if the sub-node satisfies the preset condition, The original local model included in the sub-node is used as the target model, and the three-dimensional area corresponding to the target model is divided until the sub-nodes obtained after division do not meet the preset conditions, and the root node, the node information of the root node, and multiple child nodes, node information for each child node.

Then, based on the node information of the root node and the node information of each child node, initial octree information corresponding to the three-dimensional model is generated. See the octree shown on the right side of Figure 2, where each circle in the octree represents a node.

Here, by taking the 3D model as the target model, based on the model information of the target model, according to the set division method, the 3D region corresponding to the target model is divided into eight sub-3D regions; When the sub-nodes are divided into sub-nodes, the sub-nodes are divided until none of the sub-nodes obtained after division meet the preset conditions, and the node information of the root node and the node information of each sub-node are obtained. Furthermore, based on the node information of the root node and the node information of each child node, the initial octree information corresponding to the three-dimensional model is more accurately generated.

In step A22, for each node in the initial octree information, the vertex to be deleted in the original local model corresponding to the node can be determined; The normal similarity between them is determined, and the vertex to be deleted is determined from the multiple vertices included in the original local model.

Then, a local model corresponding to the node is generated based on other vertices except the vertex to be deleted among the multiple vertices included in the original local model. For example, a triangulation algorithm can be used to triangulate other vertices except the vertex to be deleted among the vertices included in the original local model to generate a new mesh; according to the new mesh and other vertices, the The local model corresponding to the node.

During implementation, the triangulation algorithm includes, for example, the following steps: Step 1: Find the concave and convex vertices of a simple polygon. Step 2, sort the triangles according to the minimum value of the interior angle and remove the largest one to construct the triangle mesh; modify the vertex list of the polygon to recalculate the convexity of the polygon; repeat this process until the boundary vertex list is empty. Step 3, in the local scope, according to the maximum-minimum criterion, obtain the Delaunay triangulation, that is, obtain at least one mesh composed of other vertices except the vertices to be deleted.

In a possible implementation, in step A22, determining the vertex to be deleted in the original local model corresponding to the node, including:

Step A221: Based on the model information of the original local model included in the node, determine the normal similarity between the normals corresponding to the multiple meshes indicated by the model information.

Step A222: Determine the vertex to be deleted in the original local model corresponding to the node based on the normal similarity between the normals corresponding to the multiple meshes respectively.

During implementation, for each node in the initial octree information, based on the model information of the original local model included in the node, determine a plurality of meshes included in the original partial model, and determine the method of the surface corresponding to each mesh. For example, the normal vector of the face corresponding to the grid can be determined according to the coordinate information of the three vertices constituting the grid, and the normal vector can be used as the normal corresponding to the grid. Then, among the multiple meshes included in the original local model, the normal similarity between the normals of two meshes with the same vertex is determined. For example, the distance between the normals may be determined according to the normal vectors of the two grids; and the distance may be determined as the normal similarity between the normals of the two grids.

Then, based on the normal similarity between the normals corresponding to the multiple networks, the vertex to be deleted in the original local model corresponding to the node is determined. For example, the normal similarity between the normals corresponding to two grids with a common vertex and the normal similarity threshold corresponding to the node can be compared. If the normal similarity is greater than the normal similarity threshold corresponding to the node, the vertex shared between the two meshes is used as the vertex to be deleted. Among them, the normal similarity threshold corresponding to the node can be set according to the actual situation.

Here, based on the model information of the original local model corresponding to the node, the normal similarity between the normals corresponding to the multiple grids indicated by the model information is determined; The normal similarity can easily determine the vertices to be deleted from the multiple vertices of the original local model; then based on the other vertices in the multiple vertices except the vertices to be deleted, the local model corresponding to the node is generated to realize the original local model. Simplify, improve the efficiency of 3D model simplification.

In a possible implementation manner, in step A222, the vertex to be deleted in the original local model corresponding to the node is determined based on the normal similarity between the normals corresponding to the multiple grids, including:

Step 1: Traverse each vertex of the original local model, and determine a target mesh including the traversed vertex for the traversed vertex.

Step 2: Compare the normal similarity between the normals corresponding to the target grid and the normal similarity threshold corresponding to the node; wherein, the normal similarity threshold corresponding to the node is positively correlated with the node depth corresponding to the node .

Step 3: In response to the normal similarity being greater than or equal to the normal similarity threshold, determine the traversed vertex as the vertex to be deleted.

During implementation, the vertices that need to be traversed can be all vertices in the original local model, or some vertices. For example, only a certain side represented by the original local model is simplified, and the vertices in the side are regarded as vertices to be traversed. When traversing each vertex of the original local model, the target mesh corresponding to the vertex can be determined according to the position of the traversed vertex, that is, the target mesh including the vertex can be determined.

In the example of the original local model shown in FIG. 3 , the original local model includes 14 vertices, that is, a1 to a14 are traversed vertices, and also includes a mesh formed by the connection relationship between a1 to a14.

Assuming that the current traversed vertex is a4, it can be determined that all vertices that have a connection relationship with a4 include: a1, a2, a3, a5, a6, and the corresponding target meshes include: (a4, a1, a2), (a4, a2 , a6), (a4, a6, a5), (a4, a5, a3), (a4, a3, a1).

Assuming that the currently traversed vertex is a9, it can be determined that the vertices associated with a9 include: a5, a10, a13, a8, and the corresponding target meshes include: M1 (a9, a5, a10), M2 (a9, a10) , a13), M3 (a9, a13, a8), M4 (a9, a8, a5).

After the target grid is determined, the normal similarity between the normals corresponding to the target grid can be determined by calculating the distance between the normal vectors. The distance includes, for example: Euclidean distance, Manhattan distance, included angle cosine, At least one of the Hamming distance, the Chebyshev distance, etc., can be determined according to the actual situation.

After the normal similarity of the normals between the target grids is determined, the normal similarity between the normals corresponding to the target grids and the normal similarity threshold corresponding to the node can be compared; If the line similarity is greater than or equal to the normal similarity threshold, the traversed vertex is determined as the vertex to be deleted. Among them, the normal similarity threshold corresponding to the node is positively correlated with the node depth corresponding to the node. That is, the greater the node depth corresponding to the node, the greater the normal similarity threshold corresponding to the node.

Generally, for the same 3D model, the larger the normal similarity threshold is, the smaller the number of vertices to be deleted on the 3D model is determined, the smaller the simplification of the 3D model, and the greater the accuracy of the simplified 3D model. ; On the contrary, the smaller the normal similarity threshold, the more determined the number of vertices to be deleted on the 3D model, the greater the simplification of the 3D model, and the smaller the accuracy of the simplified 3D model. By setting different normal similarity thresholds for nodes with different node depths, the accuracy of the original local models corresponding to nodes with different node depths is different, and a refined display of each 3D model in the target scene is achieved.

For example, when the 3D model is closer to the virtual camera, it is determined to display the target local model corresponding to the node with a larger node depth on the 3D model. High target local model; when the 3D model is far from the virtual camera, the target local model corresponding to the node with a smaller node depth on the 3D model is determined to be displayed. The accuracy of the target local model is low, that is, the display in this case It is the target local model with lower accuracy, which realizes the refined display of the 3D model.

For example, in Figure 2, the depth of node 1 is 0, then the normal similarity threshold corresponding to node 1 can be, for example, 0.6; the depths of nodes 2 to 9 are all 1, then the normals corresponding to nodes 2 to 9 The similarity thresholds are the same, for example, may be 0.8; the depths of nodes 10 to 25 are all 2, then the normal similarity thresholds corresponding to nodes 10 to 25 are the same, for example, may be 0.9.

Exemplarily, continue to explain in conjunction with Fig. 3, for the vertex currently traversed to a9, the target network corresponding to the vertex a9 includes M1, M2, M3, M4, then M1 and M2, M1 and M3, M1 and M4 can be obtained, M2 and M3, M2 and M4, M3 and M4, a total of 6 target grid pairs corresponding to the normal similarity respectively, if M1 and M2, M1 and M3, M1 and M4, M2 and M3, M2 and M4, M3 and The normal similarity corresponding to M4 is greater than or equal to the normal similarity threshold corresponding to the node, which means that M1, M2, M3 and M4 can be merged into the same plane, and it will not cause any damage to the structure represented by the target local model. Influence outside the range is accepted, so the traversed vertex a9 can be used as the vertex to be deleted.

If among the above 6 target grid pairs, the normal similarity corresponding to any target grid pair is less than the normal similarity threshold corresponding to the node, it means that if a9 is deleted, M1, M2, M3 and M4 are merged into The same plane will have a greater impact on the structure represented by the three-dimensional model, so the vertex a9 cannot be deleted, that is, it cannot be used as the vertex to be deleted.

Traverse all vertices in the original local model, and take the vertices that can be deleted from the original local model as the vertices to be deleted.

In this way, the boundary of vertex deletion is limited by a numerically controllable normal similarity threshold, so that the fineness of the simplification of the original local model is controllable, while the accuracy of the simplified local model is guaranteed, and the simplification of the model is improved at the same time. efficiency.

In step A23, after obtaining the local model information of the local model corresponding to each node, the octree information corresponding to the three-dimensional model may be generated according to the local model information of the local model corresponding to each node.

For S102:

Here, the virtual camera may be a virtual camera included in the rendering engine for photographing the three-dimensional model. The current shooting parameter information of the virtual camera may include the current pose of the virtual camera, the viewing angle range of the virtual camera, the viewing distance of the virtual camera, and the like. Among them, the visible angle range and the visible distance can be set according to the actual situation. For example, the viewing angle range can be matched with the visual degree of human eyes, and the viewing distance can be 50 meters.

During implementation, the current visual area corresponding to the virtual camera and the area information of the visual area under the current pose of the virtual camera may be determined according to the visual angle range of the virtual camera and the visual distance of the virtual camera.

For S103:

During implementation, the three-dimensional model in the target scene located in the visualization area can be determined based on the area information of the visualization area of the virtual camera; then based on the octree information corresponding to the three-dimensional model located in the visualization area, at least The target local model contained in a node. For example, the three-dimensional region represented by each node included in the octree information may be compared with the visualization region to determine the target local model included in at least one node located in the visualization region.

In a possible implementation manner, the target local model included in at least one node located in the visualization area is determined based on the octree information and area information corresponding to each three-dimensional model included in the target scene, including:

Step B1: Determine at least one target three-dimensional model located in the visualization area based on the area information, the displayed pose and model size of each three-dimensional model included in the target scene.

Step B2, based on the octree information and the area information corresponding to the target three-dimensional model, determine the target local model included in at least one node in the visualization area.

During implementation, at least one target three-dimensional model located in the visualization area is determined based on the area information, the displayed pose and model size of each three-dimensional model included in the target scene. For example, according to the displayed pose and model size of each 3D model, determine the spatial area where the 3D model is located in the target scene; and then determine whether the spatial area corresponding to the 3D model overlaps with the visualization area. The 3D model belongs to the target 3D model; if it does not overlap, the 3D model does not belong to the target 3D model.

Then, based on the octree information and the area information corresponding to the target three-dimensional model, the target local model included in at least one node located in the visualization area is determined. For example, the three-dimensional region represented by each node included in the octree information can be compared with the current visualization region, and according to the comparison result, the target local model included in at least one node located in the current visualization region can be determined. For example, if the comparison result indicates that the current visualization area completely surrounds the three-dimensional area of the node, the local model included in the node belongs to the target local model.

In the embodiment of the present disclosure, at least one target 3D model located in the visualization area is determined based on the area information, the displayed pose and model size of each 3D model included in the target scene; and then according to the octree information corresponding to the target 3D model and area information, the target local model included in at least one node in the visualization area can be more accurately determined, and the determination process of the target local model is relatively simple and efficient.

In a possible implementation, in step B2, based on the octree information and the area information corresponding to the target three-dimensional model, determine the target local model included in at least one node in the visualization area, including:

Step B21, taking the root node in the octree information corresponding to the target three-dimensional model as the node to be processed, and determining whether the three-dimensional area corresponding to the node to be processed is completely located in the visualization area indicated by the area information;

Step B22, in the case that the three-dimensional area corresponding to the node to be processed is partially located in the visualization area, use each child node connected to the node to be processed as the node to be processed, and return to determine whether the three-dimensional area corresponding to the node to be processed is completely located in the area information indication The step of visualizing the area until the three-dimensional area corresponding to the node to be processed is completely located in the visual area;

Step B23, in the case that the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, determine the local model included in the node to be processed as the target local model included in the node located in the visualization area.

The root node in the octree information corresponding to the target three-dimensional model is used as the node to be processed, and it is determined whether the three-dimensional area corresponding to the node to be processed is completely located in the visualization area indicated by the area information. If the three-dimensional area corresponding to the node to be processed is partially located in the visualization area, then go to step B22; if the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, go to step B23; if the three-dimensional area corresponding to the node to be processed is not located in the visualization area at all, Then it is determined that the local model included in the node to be processed does not belong to the target local model.

During implementation, start traversing from the root node in the octree information corresponding to the target 3D model. If the 3D area corresponding to the root node is partially located in the visualization area, then traverse the child nodes connected to the root node, that is, each child node connected to the root node. The node is used as the node to be processed, and it is determined whether the three-dimensional area corresponding to the node to be processed is completely located in the visualization area.

When the three-dimensional area corresponding to the node to be processed is partially located in the visualization area, take each child node connected to the node to be processed as the node to be processed, and return to the step of determining whether the three-dimensional area corresponding to the node to be processed is completely located in the visualization area indicated by the area information , until the three-dimensional area corresponding to the node to be processed is completely located in the visualization area. Referring to FIG. 4 , it is a schematic diagram that the three-dimensional area corresponding to the node to be processed is partially located in the visualization area.

When the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, the local model included in the node to be processed is determined as the target local model. Referring to FIG. 5 , it is a schematic diagram that the three-dimensional area corresponding to the node to be processed is completely located in the visualization area.

When the three-dimensional area corresponding to the node is not located in the visualization area at all, it is determined that the local model included in the node does not belong to the target local model, and rendering display is not performed. Referring to Fig. 6, a schematic diagram showing that the three-dimensional area corresponding to the node is not located in the visualization area at all.

In the above method, by judging whether the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, when the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, the local model included in the node to be processed is determined as the target local model, Improves the accuracy and efficiency of the target local model.

Those skilled in the art can understand that in the above method of the specific implementation, the writing order of each step does not mean a strict execution order but constitutes any limitation on the implementation process, and the specific execution order of each step should be based on its function and possible Internal logic is determined.

Based on the same concept, an embodiment of the present disclosure also provides a model display device. Referring to FIG. 7 , a schematic diagram of the architecture of the model display device provided by the embodiment of the present disclosure includes a central processing unit (CPU) 701, A graphics processor (graphics processing unit, GPU) 702, specifically:

The CPU 701 is used to store the octree information corresponding to each three-dimensional model included in the target scene, and after determining the target local model included in at least one node, transmit the model information of the target local model included in the node. to the GPU; wherein, the target local model included in the at least one node is determined by using the model display method described in the above embodiment.

The GPU 702 is configured to store the received model information of the target local model included in the node into the internal memory; and by reading the model information in the internal memory, in the shooting picture of the virtual camera, the The target local model is rendered and displayed.

During implementation, after constructing each 3D model, according to the model information of each 3D model in the target scene, determine the octree information corresponding to each 3D model; and store the octree information corresponding to each 3D model in the CPU . And, the target local model to be rendered and displayed can be determined according to the above-mentioned implementation manner. Transfer the model information of the target local model to the connected GPU. For the determination process of the target local model, reference may be made to the specific descriptions of S101 to S103 above, which will not be repeated here.

The GPU stores the received model information of the target local model in the internal memory (memory), and by reading the model information in the memory, renders and displays the target local model in the shooting screen of the virtual camera.

Here, by using the model display method described in the first aspect or any one of the embodiments, after the target local model is more accurately determined, the model information of the target local model can be sent to the GPU, and there is no need to send the models of other local models to the GPU. information, reducing the amount of information transmission, alleviating the consumption of bandwidth resources when sending model information of other local models that do not need to be rendered to the GPU, and improving the utilization of bandwidth resources. At the same time, the model information of the target local model is loaded into the internal memory (memory), and there is no need to load the model information of other local models into the memory, which reduces memory overhead and improves memory utilization.

In an optional implementation manner, the CPU is further configured to send the updated model information of the target local model to the GPU after the target local model is updated;

The GPU is also used to delete the model information that does not match the updated target partial model in the internal memory; and, in the updated target partial model, the model of the unstored target partial model Information is stored in the internal memory.

When the current pose of the virtual camera changes, the current shooting parameter information of the virtual camera changes, and the visualization area corresponding to the virtual camera changes, so the determined local model of the target will change. The target local model can be re-determined using the processes of S101-S03 described above.

After determining the updated target local model from each 3D model in the target scene area based on the octree information and the updated current shooting parameter information corresponding to the virtual camera, the CPU can convert the updated target local model to the model Information is sent to the GPU. The GPU can delete the model information in the memory that does not match the updated target local model; and, in the updated target local model, the model information of the unstored target local model is stored in the memory, so that what is stored in the memory is required Rendering and displaying the model information of the target local model, alleviating the consumption of memory resources and improving the utilization of memory resources.

In the above embodiment, after the target local model is updated, the model information of the updated target local model can be loaded into the memory, and the model information in the memory that does not match the updated target local model is deleted, so as to ensure that what is loaded in the memory is correct. Currently, the model information of the target local model needs to be rendered and displayed to avoid the waste of memory resources.

Based on the same concept, an embodiment of the present disclosure also provides a model display device. Referring to FIG. 8 , a schematic diagram of the architecture of the model display device provided by the embodiment of the present disclosure includes an acquisition module 801 , a first determination module 802 , a first The second determination module 803, specifically:

The obtaining module 801 is configured to obtain octree information corresponding to each 3D model included in the target scene; wherein, the octree information includes a plurality of nodes; the 3D region represented by each node includes the 3D model The fineness of the local model included in the node is positively correlated with the node depth of the node; the number of vertices included in the local model with high fineness is greater than the number of vertices included in the local model with low fineness;

a first determination module 802, configured to determine the area information of the visualized area of the virtual camera based on the current shooting parameter information of the virtual camera used for shooting the three-dimensional model in the target scene;

The second determining module 803 is configured to determine, based on the octree information and the region information corresponding to each three-dimensional model included in the target scene, a target local model included in at least one node located in the visualization region.

In a possible implementation manner, the apparatus further includes: a first generation module 804, the first generation module 804 is configured to generate octree information corresponding to each three-dimensional model included in the target scene according to the following steps: :

Obtain model information of each 3D model included in the target scene;

Based on the model information corresponding to each of the three-dimensional models, octree information corresponding to the three-dimensional models is generated.

In a possible implementation manner, the first generation module 804, when generating the octree information corresponding to the three-dimensional model based on the model information corresponding to each of the three-dimensional models, is used for:

Based on the model information corresponding to each of the three-dimensional models, initial octree information corresponding to the three-dimensional model is generated; wherein, the three-dimensional region represented by each node in the initial octree information includes the corresponding three-dimensional model. The original local model; the original local model includes a plurality of vertices and a plurality of meshes formed by the connection relationship between the vertices;

For each node in the initial octree information, determine the vertex to be deleted in the original local model corresponding to the node, and determine the vertex to be deleted based on the multiple vertices included in the original local model except the vertex to be deleted vertex, generate the local model corresponding to the node;

Based on the local model information of the local model corresponding to each node, the octree information corresponding to the three-dimensional model is generated.

In a possible implementation manner, the device further includes: a second generation module 805, the second generation device generates an initial value corresponding to the three-dimensional model based on the model information corresponding to each of the three-dimensional models. When octree information is used, it is used to:

Taking the three-dimensional model as the target model, based on the model information of the target model, according to the set division method, the three-dimensional area corresponding to the target model is divided into eight sub-three-dimensional areas; wherein, the three-dimensional area of the three-dimensional model The node corresponding to the area is a root node, and the node information of the root node includes the model information of the three-dimensional model; and each sub-three-dimensional area corresponds to a child node; the node information of each child node includes: located in the sub-three-dimensional corresponding to the node Model information or preset information of the original local model in the area;

When there is a child node that satisfies the preset condition, the original local model included in the child node that satisfies the preset condition is used as the target model, and the model information based on the target model is returned. The step of dividing the three-dimensional region corresponding to the target model into eight sub-three-dimensional regions, until the sub-nodes obtained after the division do not meet the preset conditions; wherein, the sub-nodes that meet the preset conditions include: nodes whose depth is less than the set depth threshold. Nodes, and/or, child nodes whose number of vertices in the model information included in the node information is greater than the set number threshold;

Based on the generated node information of the root node and the node information of each child node, initial octree information corresponding to the three-dimensional model is generated.

In a possible implementation manner, the apparatus further includes: a third determination module 806, the third determination module 806 is configured to determine the vertex to be deleted in the original local model corresponding to the node:

Based on the model information of the original local model included in the node, determine the normal similarity between the normals corresponding to the multiple meshes indicated by the model information;

The vertex to be deleted in the original local model corresponding to the node is determined based on the normal similarity between the normals corresponding to the multiple meshes respectively.

In a possible implementation manner, the third determining module 806 determines, based on the normal similarity between the normals corresponding to the multiple grids, the to-be-deleted items in the original local model corresponding to the node. When vertices are used:

Traversing the vertices of the original local model, and determining the target mesh including the traversed vertices for the traversed vertices;

Compare the normal similarity between the normals corresponding to the target grid and the normal similarity threshold corresponding to the node; wherein, the normal similarity threshold corresponding to the node corresponds to the node The node depth of is positively correlated;

In response to the normal similarity being greater than or equal to the normal similarity threshold, the traversed vertex is determined as a vertex to be deleted.

In a possible implementation manner, the first determining module 802 determines at least one octree located in the visualization area based on the octree information and the area information corresponding to each three-dimensional model included in the target scene. When the target local model is included in the node, it is used to:

determining at least one target three-dimensional model located in the visualization area based on the area information, the displayed pose and model size of each three-dimensional model included in the target scene;

Based on the octree information corresponding to the target three-dimensional model and the area information, a target local model included in at least one node in the visualization area is determined.

In a possible implementation manner, the second determining module 803 determines the target included in at least one node in the visualization area based on the octree information corresponding to the target three-dimensional model and the area information. When a local model is used:

Taking the root node in the octree information corresponding to the target three-dimensional model as the node to be processed, it is determined whether the three-dimensional area corresponding to the node to be processed is completely located in the visualization area indicated by the area information;

In the case that the three-dimensional area corresponding to the node to be processed is partially located in the visualization area, take each child node connected to the node to be processed as a node to be processed, and return to determining whether the three-dimensional area corresponding to the node to be processed is not The step of being completely located in the visualization area indicated by the area information, until the three-dimensional area corresponding to the node to be processed is completely located in the visualization area;

When the three-dimensional area corresponding to the node to be processed is completely located in the visualization area, the local model included in the node to be processed is determined as the target local model included in the node located in the visualization area.

In some embodiments, the functions or templates included in the apparatus provided by the embodiments of the present disclosure may be used to execute the methods described in the above method embodiments. For specific implementation, reference may be made to the above method embodiments. For brevity, here No longer.

Based on the same technical concept, an embodiment of the present disclosure also provides an electronic device. Referring to FIG. 9 , a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure includes a processor 901 , a memory 902 , and a bus 903 . Among them, the memory 902 is used to store the execution instructions, including the memory 9021 and the external memory 9022; the memory 9021 here is also called the internal memory, and is used to temporarily store the operation data in the processor 901 and the data exchanged with the external memory 9022 such as the hard disk, The processor 901 exchanges data with the external memory 9022 through the memory 9021. When the electronic device 900 is running, the processor 901 communicates with the memory 902 through the bus 903, so that the processor 901 executes the following instructions:

Acquire octree information corresponding to each 3D model included in the target scene; wherein, the octree information includes multiple nodes; the 3D region represented by each node includes a local model on the 3D model; The fineness of the local model included in the node is positively correlated with the node depth of the node; the number of vertices included in the local model with high fineness is greater than the number of vertices included in the local model with low finesse;

Determine the area information of the visualization area of the virtual camera based on the current shooting parameter information of the virtual camera used for shooting the three-dimensional model in the target scene;

Based on the octree information and the region information respectively corresponding to each three-dimensional model included in the target scene, a target local model included in at least one node located in the visualization region is determined.

For the specific processing flow of the processor 901, reference may be made to the records of the foregoing method embodiments, which will not be repeated here.

In addition, an embodiment of the present disclosure further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, the steps of the model display method described in the above method embodiments are executed. . Wherein, the storage medium may be a volatile or non-volatile computer-readable storage medium.

Embodiments of the present disclosure further provide a computer program product, where the computer program product carries program codes, and the instructions included in the program codes can be used to execute the steps of the model display method described in the foregoing method embodiments. For details, please refer to the foregoing method. The embodiments are not repeated here.

Wherein, the above-mentioned computer program product can be specifically implemented by means of hardware, software or a combination thereof. In an optional embodiment, the computer program product is embodied as a computer storage medium, and in another optional embodiment, the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK), etc. Wait.

The present disclosure relates to the field of augmented reality. By acquiring image information of a target object in a real environment, various visual correlation algorithms are used to detect or identify the relevant features, states, and attributes of the target object, so as to obtain an image matching the specific application. AR effect that combines virtual and reality. Exemplarily, the target object may involve faces, limbs, gestures, movements, etc. related to the human body, or objects, markers, or sandboxes, display areas, or display items related to venues or venues. Vision-related algorithms may involve visual localization, SLAM, 3D reconstruction, image registration, background segmentation, object keypoint extraction and tracking, object pose or depth detection, etc. The specific application can not only involve interactive scenes such as navigation, navigation, explanation, reconstruction, and virtual effect overlay display related to real scenes or items, but also special effects processing related to people, such as makeup beautification, body beautification, special effects display, virtual Model display and other interactive scenarios. The relevant features, states and attributes of the target object can be detected or recognized through the convolutional neural network. The above convolutional neural network is a network model obtained by model training based on a deep learning framework.

Those skilled in the art can clearly understand that, for the convenience and brevity of description, for the specific working process of the system and device described above, reference may be made to the corresponding process in the foregoing method embodiments, which will not be repeated here. In the several embodiments provided by the present disclosure, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. The apparatus embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some communication interfaces, indirect coupling or communication connection of devices or units, which may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a processor-executable non-volatile computer-readable storage medium. Based on this understanding, the technical solutions of the present disclosure can be embodied in the form of software products in essence, or the parts that make contributions to the prior art or the parts of the technical solutions. The computer software products are stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present disclosure. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program codes.

The above are only specific embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any person skilled in the art who is familiar with the technical scope of the present disclosure can easily think of changes or substitutions, which should be covered within the scope of the present disclosure. within the scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims (13)

1. A method of model display, comprising:

acquiring octree information corresponding to each three-dimensional model included in a target scene; wherein the octree information comprises a plurality of nodes; each three-dimensional region characterized by a node comprises a local model on the three-dimensional model; the fineness of a local model included by the node is positively correlated with the node depth of the node; the local model with high fineness comprises the number of vertexes, and the number of vertexes is larger than that of the local model with low fineness;

determining area information of a visual area of a virtual camera based on current shooting parameter information of the virtual camera for shooting a three-dimensional model in the target scene;

and determining a target local model included in at least one node positioned in the visualization region based on the octree information and the region information respectively corresponding to each three-dimensional model included in the target scene.

2. The method according to claim 1, wherein the obtaining octree information corresponding to each three-dimensional model included in the target scene comprises:

obtaining model information of each three-dimensional model included in a target scene;

and generating octree information corresponding to the three-dimensional models based on the model information corresponding to each three-dimensional model.

3. The method of claim 2, wherein generating octree information corresponding to each of the three-dimensional models based on the model information corresponding to the three-dimensional model comprises:

generating initial octree information corresponding to the three-dimensional models based on the model information corresponding to each three-dimensional model; wherein, the three-dimensional region represented by each node in the initial octree information comprises an original local model corresponding to the three-dimensional model; the original local model comprises a plurality of vertexes and a plurality of meshes formed by the connection relations among the vertexes;

determining a vertex to be deleted in an original local model corresponding to each node in the initial octree information, and generating a local model corresponding to the node based on other vertexes except the vertex to be deleted in a plurality of vertexes included in the original local model;

and generating octree information corresponding to the three-dimensional model based on the local model information of the local model corresponding to each node.

4. The method of claim 3, wherein generating initial octree information corresponding to each of the three-dimensional models based on the model information corresponding to the three-dimensional model comprises:

taking the three-dimensional model as a target model, and dividing a three-dimensional area corresponding to the target model into eight sub three-dimensional areas according to a set dividing mode on the basis of model information of the target model; the node corresponding to the three-dimensional region of the three-dimensional model is a root node, and the node information of the root node comprises model information of the three-dimensional model; each sub three-dimensional area corresponds to one sub node; the node information of each child node includes: model information or preset information of the original local model located in the sub-three-dimensional region corresponding to the node;

when the child nodes meeting the preset conditions exist, the original local model included in the child nodes meeting the preset conditions is used as a target model, model information based on the target model is returned, and the three-dimensional region corresponding to the target model is divided into eight sub three-dimensional regions according to the set dividing mode until the divided child nodes do not meet the preset conditions; wherein, the child node satisfying the preset condition includes: the node depth is smaller than the child nodes of the set depth threshold value, and/or the number of vertexes in the model information included in the node information is larger than the child nodes of the set number threshold value;

and generating initial octree information corresponding to the three-dimensional model based on the generated node information of the root node and the node information of each child node.

5. The method according to claim 3 or 4, wherein the determining the vertex to be deleted in the original local model corresponding to the node comprises:

determining normal similarity between normals respectively corresponding to a plurality of grids indicated by the model information based on the model information of the original local model included by the node;

and determining the vertexes to be deleted in the original local model corresponding to the nodes based on the normal similarity among the normals respectively corresponding to the grids.

6. The method according to claim 5, wherein the determining vertices to be deleted in the original local model corresponding to the nodes based on the normal similarity between the normals corresponding to the meshes respectively comprises:

traversing each vertex of the original local model, and determining a target mesh comprising the traversed vertex aiming at the traversed vertex;

comparing the normal similarity between the normals corresponding to the target grid with a normal similarity threshold corresponding to the node; wherein the normal similarity threshold corresponding to the node is positively correlated with the node depth corresponding to the node;

and determining the traversed vertex as a vertex to be deleted in response to the normal similarity being greater than or equal to the normal similarity threshold.

7. The method according to any one of claims 1 to 6, wherein the determining a target local model included in at least one node located in the visualization region based on the octree information and the region information respectively corresponding to each three-dimensional model included in the target scene comprises:

determining at least one target three-dimensional model located within the visualization area based on the area information, the display pose and the model size of each three-dimensional model included in the target scene;

and determining a target local model included by at least one node in the visualization region based on the octree information corresponding to the target three-dimensional model and the region information.

8. The method according to claim 7, wherein the determining the target local model included in the at least one node in the visualization region based on the octree information corresponding to the target three-dimensional model and the region information comprises:

taking a root node in the octree information corresponding to the target three-dimensional model as a node to be processed, and determining whether a three-dimensional region corresponding to the node to be processed is completely located in a visual region indicated by the region information;

under the condition that the three-dimensional region corresponding to the node to be processed is partially located in the visualization region, taking each sub-node connected with the node to be processed as the node to be processed, and returning to the step of determining whether the three-dimensional region corresponding to the node to be processed is completely located in the visualization region indicated by the region information until the three-dimensional region corresponding to the node to be processed is completely located in the visualization region;

and under the condition that the three-dimensional region corresponding to the node to be processed is completely positioned in the visualization region, determining the local model included by the node to be processed as a target local model included by the node positioned in the visualization region.

9. A model display apparatus, comprising: a central processing unit CPU and a graphic processing unit GPU;

the CPU is used for storing octree information corresponding to each three-dimensional model included in a target scene, and transmitting model information of a target local model included by at least one node to the GPU after determining the target local model included by the node; wherein, the target local model included in the at least one node is determined by using the model display method of any one of claims 1 to 8;

the GPU is used for storing the received model information of the target local model included by the node into an internal memory; and rendering and displaying the target local model in a shooting picture of a virtual camera by reading the model information in the internal memory.

10. The apparatus of claim 9, wherein the CPU is further configured to send, after the target local model is updated, model information of the updated target local model to the GPU;

the GPU is also used for deleting the model information which is not matched with the updated target local model in the internal memory; and storing the model information of the target local model which is not stored in the updated target local model into the internal memory.

11. A model display apparatus, comprising:

the acquisition module is used for acquiring octree information corresponding to each three-dimensional model in a target scene; wherein the octree information comprises a plurality of nodes; each three-dimensional region characterized by a node comprises a local model on the three-dimensional model; the fineness of a local model included by the node is positively correlated with the node depth of the node; the local model with high fineness comprises the number of vertexes, and the number of vertexes is larger than that of the local model with low fineness;

the first determination module is used for determining the area information of a visual area of a virtual camera based on the current shooting parameter information of the virtual camera used for shooting the three-dimensional model in the target scene;

and the second determining module is used for determining a target local model included in at least one node located in the visualization region based on the octree information and the region information respectively corresponding to each three-dimensional model included in the target scene.

12. An electronic device, comprising: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory communicating via the bus when the electronic device is running, the machine-readable instructions when executed by the processor performing the steps of the model exhibition method according to any one of claims 1 to 8.

13. A computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of the model exhibition method according to one of the claims 1 to 8.

Images