A cloud-based 3D effect rendering method and system

Abstract

The present invention discloses a cloud-based 3D effect rendering method and system, and relates to the technical field of cloud-based 3D effect rendering. The present invention adopts a computing power dynamic allocation algorithm to dynamically generate computing resources that match the current 3D rendering request of a user terminal, and redistributes computing resources after each 3D rendering request of multiple user terminals, so that computing resources are used more evenly, and the computing resources of a cloud server are used to the maximum extent. The user terminal does not need high-performance local computing equipment, and only needs a basic network connection to perform high-quality 3D rendering. The rendering speed is greatly improved through the dynamic allocation of computing resources and multi-threaded parallel computing of the cloud server; the user terminal does not need to purchase expensive hardware equipment, and adopts optimized data transmission and dynamic adjustment strategies to ensure that stable rendering effects can be obtained under different network conditions. Through bandwidth adaptive technology, it is ensured that a smooth 1080p rendering effect can still be provided under a 10Mbps network.

Description

Technical Field

[0001] The present invention relates to the technical field of cloud 3D effect rendering, and in particular to a cloud 3D effect rendering method and system. Background Art

[0002] Currently, 3D rendering mainly relies on local computing resources such as CPU and GPU. This method not only requires high-performance hardware support, but also requires a lot of storage space and memory.

[0003] Therefore, for users with limited device performance, the efficiency and effect of 3D rendering are often unsatisfactory. With the development of cloud computing technology, it has become possible to use cloud resources for 3D rendering. However, existing cloud 3D rendering technology has problems such as high latency, high cost, and unstable rendering quality. These problems limit the widespread application of cloud 3D rendering technology.

[0004] The Chinese invention patent application with publication number CN118485755A discloses a method, device, medium and product for cloud rendering. The method can be applied to the application logic node, rendering node and streaming node of the distributed cloud rendering system. The method applied to the application logic node includes: determining the object to be rendered and capturing multiple snapshot frame data of the object to be rendered; calculating the number of frames of the multiple snapshot frame data; determining the required number of rendering nodes according to the number of frames, wherein the rendering node is located in the distributed cloud rendering system; sending the multiple snapshot frame data to at least one rendering node frame by frame. Therefore, through the distributed cloud rendering system, the effective integration and utilization of rendering resources can be achieved under the premise of ensuring the rendering of the entire frame.

[0005] The above cloud rendering method does not consider the data interaction between the user end and the cloud server, and does not dynamically adjust the computing power resources of the cloud server under the multi-rendering requirements of multiple user ends, and the optimization effect of computing power resources is poor. Summary of the invention

[0006] In order to solve the technical problems of data interaction between the user end and the cloud server and dynamic adjustment of the computing resources of the cloud server, the present invention provides a cloud 3D effect rendering method and system. The following technical solutions are adopted:

[0007] A cloud-based 3D effect rendering method comprises the following steps:

[0008] Step 1, resource request and allocation: The client sends the 3D rendering request content to the cloud server;

[0009] The cloud server uses a computing power dynamic allocation algorithm to generate computing resources that match the 3D rendering request content of the user end according to the 3D rendering request content and the current computing resource usage data;

[0010] Step 2: Data transmission and preprocessing: The user uploads the 3D model and material data to be rendered corresponding to the 3D rendering request content, and the cloud server receives the data and preprocesses it;

[0011] Step 3, rendering task execution: the cloud server uses the computing resources allocated in step 1 to perform 3D rendering;

[0012] Step 4, result transmission and display: After rendering is completed, the cloud server compresses the rendering result and transmits the compressed rendering result data back to the client, which decompresses it and displays the rendering effect.

[0013] By adopting the above technical solution, the client sends a 3D rendering request to the cloud server.

[0014] The cloud server dynamically allocates appropriate computing resources based on the request content and current resource usage. The computing resources include the number of CPU cores (e.g., 8-64 cores) and GPU memory (e.g., 8-32GB).

[0015] The dynamic allocation algorithm of computing power can dynamically generate computing resources that match the current 3D rendering request of the user end, so that computing resources can be reallocated after each 3D rendering request of multiple users, so that computing resources are used more evenly and the computing resources of the cloud server are used to the maximum extent. The user end does not need high-performance local computing equipment, but only needs a basic network connection to perform high-quality 3D rendering. The dynamic allocation of computing resources of the cloud server and multi-threaded parallel computing can greatly improve the rendering speed. For example, under the configuration of 32-core CPU and 16GB GPU memory, the rendering time of a typical 3D scene is reduced to less than 1 minute;

[0016] The user does not need to purchase expensive hardware equipment, but only needs to pay for the use of cloud server computing resources on demand. Optimized data transmission and dynamic adjustment strategies are adopted to ensure stable rendering effects under different network conditions. Through bandwidth adaptive technology, smooth 1080p rendering effects can still be provided under 10Mbps network.

[0017] Optionally, in step 1, the specific method of using the computing power dynamic allocation algorithm to generate computing resources that match the user-side 3D rendering request content is: first use the following formula to calculate the computing power requirement of the user-side 3D rendering request content:

[0018] ;

[0019] in The computing power required for the i-th rendering request, The priority weight of the i-th rendering request, is the resource requirement of the i-th rendering request, where the resource requirement refers to the number of CPU cores and GPU memory;

[0020] Then use the following formula to calculate the total dynamic computing power redundancy:

[0021] ;

[0022] in is the total dynamic computing power redundancy, is the dynamic redundancy coefficient, ranging from 10% to 30%. is the total available computing power of the cloud server at present;

[0023] Finally, the following formula is used to generate computing resources that match the 3D rendering request content of the user end:

[0024] ;

[0025] in is the computing power matching the i-th rendering request, j is an index variable, an integer from 0 to n, Used to traverse all rendering requests. Indicates the computing power required for the j-th rendering request;

[0026] It is the total available computing power of the cloud server.

[0027] .

[0028] Optional, The value of is re-prioritized based on the currently running rendering tasks and the rendering tasks of the current rendering request. The priority re-prioritization considers the following four variables:

[0029] Task urgency score, resource requirement score, user importance score and estimated completion time score; calculated using the following formula :

[0030] ;

[0031] in is the priority score of the ith task, is the task urgency score of the i-th task, is the resource requirement score of the ith task, is the user importance score of the ith task, is the estimated completion time score of the i-th task;

[0032] The priority weight change value of all rendering tasks is recalculated based on the priority score of the tasks added to the rendering task of the current rendering request:

[0033] ;

[0034] in is the priority weight change value of the i-th task, is the weight change coefficient, is the original priority score of the i-th task before the rendering task of the current rendering request is added;

[0035] The priority weight value calculation formula of the rendering task of the current rendering request is:

[0036] .

[0037] By adopting the above technical solution, every time a new 3D rendering request is input to the cloud server, the priority weight values ​​of all rendering tasks need to be recalculated, thereby realizing dynamic adjustment of computing resources, so that each 3D rendering task is performed under the optimal computing resource allocation, and improving the utilization rate of computing resources;

[0038] First, let's define some factors that affect the priority of rendering tasks:

[0039] Task urgency: How long does the task need to be completed?

[0040] Resource requirements: The amount of computing resources required for the task.

[0041] Client Importance: Tasks for different clients may have different priorities.

[0042] Estimated completion time: The estimated completion time based on task requirements and current resource usage.

[0043] By determining the above factors, the task priority score can be quantified, thereby achieving dynamic adjustment of the priority weight value.

[0044] Optionally, the cloud server receives the data and performs preprocessing including data format conversion and compression.

[0045] By adopting the above technical solution, the user uploads the 3D model and material data to be rendered, and the data size is generally between 100MB and 2GB. The cloud server receives the data and performs preprocessing, including data format conversion and compression, to reduce the amount of data transmission. The compression ratio can reach 3:1.

[0046] Optionally, in step 3, the cloud server first provides a rendering thumbnail, which is generated based on the 3D rendering request content, and transmits the rendering thumbnail and the rendering resolution re-selection item back to the user end. After the user end personnel observes the rendering thumbnail, they finally determine the rendering resolution in the rendering resolution re-selection item. The user end transmits the determined rendering resolution to the cloud server, and the cloud server uses the computing resources allocated in step 1 to perform 3D rendering according to the determined rendering resolution.

[0047] By adopting the above technical solution, before actually running the rendering task, a rendering thumbnail matching the initial rendering parameters, such as rendering resolution, can be generated according to the rendering parameters initially set, and the rendering thumbnail can be sent back to the user end for confirmation of the final rendering parameters, so as to avoid a large gap between the final rendering result obtained by the user end and the expected result. When sending back the rendering thumbnail, a rendering resolution re-selection option can also be provided. If the user end is not satisfied with the effect after viewing the rendering thumbnail, the rendering resolution requirement can be increased in the rendering resolution re-selection option.

[0048] Optionally, in step 4, according to the format of the rendering result, use Python and Pillow library to select a suitable compression algorithm from JPEG, PNG, WebP, and HEVC. If JPEG is used to compress the image, the compression ratio is adjusted by selecting a quality factor. If PNG is used to compress the image, the compression ratio is adjusted by selecting different compression levels. If WebP is used to compress the image, the compression ratio is adjusted by selecting a quality factor. If HEVC is used to compress the video, the compression ratio is adjusted by adjusting the quantization parameter.

[0049] Optionally, the user uses Python and the Pillow library to decompress the file and restore the original rendering result.

[0050] By adopting the above technical solution, after rendering is completed, the cloud server compresses the rendering results, and transmits the compressed rendering results back to the user end through the network. The network bandwidth requirement is 10Mbps to 100Mbps. The user end decompresses and displays the rendering effect, reducing the network bandwidth requirement during the transmission process.

[0051] A cloud-based 3D effect rendering system is used to implement a cloud-based 3D effect rendering method. The cloud-based 3D effect rendering system includes a user terminal, a network transmission module and a cloud server. The user terminal is connected to the cloud server through the network transmission module. The cloud server is provided with a data analysis chip. The data analysis chip is used to run a computing power dynamic allocation algorithm to generate computing resources that match the 3D rendering request content of the user terminal.

[0052] Optionally, the cloud server and the user end are both provided with encryption and decryption modules, and the encryption and decryption modules provide the same encryption protocol and algorithm for the user end and the cloud server.

[0053] Optionally, a computing power warning module is also included, and the computing power warning module is connected to the cloud server in communication. When the total available computing power of the cloud server is determined, An alarm will be issued when it is less than 15% of the total computing power.

[0054] By adopting the above technical solution, the computing power warning module can calculate the total available computing power of the cloud server. When it is less than 15% of the total computing power, an alarm will be issued to prevent the cloud server from running continuously under high load.

[0055] In summary, the present invention includes at least one of the following beneficial technical effects:

[0056] The present invention can provide a cloud-based 3D effect rendering method and system, which adopts a computing power dynamic allocation algorithm to dynamically generate computing resources that match the current 3D rendering request of the user end, and redistribute computing resources after each 3D rendering request of multiple user ends, so that computing resources are used more evenly and the computing resources of the cloud server are used to the maximum extent. The user end does not need a high-performance local computing device, but only needs a basic network connection to perform high-quality 3D rendering. The rendering speed is greatly improved through the dynamic allocation of computing resources of the cloud server and multi-threaded parallel computing.

[0057] Users do not need to purchase expensive hardware equipment, but only need to pay on demand to use cloud server computing resources.

[0058] Optimized data transmission and dynamic adjustment strategies are used to ensure stable rendering effects under different network conditions. Bandwidth adaptive technology is used to ensure smooth 1080p rendering effects under 10Mbps network. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flow chart of a cloud 3D effect rendering method of the present invention;

[0060] Figure 2 It is a schematic diagram of the component connection principle of a cloud-based 3D effect rendering system of the present invention.

[0061] Explanation of the accompanying drawings: 100, user end; 101, cloud server; 1011, data analysis chip; 102, encryption and decryption module; 103, network transmission module; 104, computing power warning module. DETAILED DESCRIPTION

[0062] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0063] The embodiments of the present invention disclose a cloud-based 3D effect rendering method and system.

[0064] Reference Figure 1 and Figure 2 , Embodiment 1, a cloud 3D effect rendering method, comprising the following steps:

[0065] Step 1, resource request and allocation: the client 100 sends a 3D rendering request content to the cloud server 101;

[0066] The cloud server 101 generates computing resources that match the 3D rendering request content of the client 100 using a computing power dynamic allocation algorithm based on the 3D rendering request content and current computing resource usage data;

[0067] Step 2, data transmission and preprocessing: the client 100 uploads the 3D model and material data to be rendered corresponding to the 3D rendering request content, and the cloud server 101 receives the data and preprocesses it;

[0068] Step 3, rendering task execution: the cloud server 101 performs 3D rendering using the computing resources allocated in step 1;

[0069] Step 4, result transmission and display: After the rendering is completed, the cloud server 101 compresses the rendering result, and transmits the compressed rendering result data back to the client 100, which decompresses it and displays the rendering effect.

[0070] The client 100 sends a 3D rendering request to the cloud server 101 .

[0071] The cloud server 101 dynamically allocates appropriate computing resources according to the request content and current resource usage, and the computing resources include the number of CPU cores (e.g., 8-64 cores) and GPU memory (e.g., 8-32 GB);

[0072] The computing power dynamic allocation algorithm can dynamically generate computing resources that match the current 3D rendering request of the user terminal 100, so that the computing resources can be reallocated after each 3D rendering request of multiple user terminals 100, so that the computing resources are used more evenly and the computing resources of the cloud server 101 are used to the maximum extent. The user terminal 100 does not need high-performance local computing equipment, but only needs a basic network connection to perform high-quality 3D rendering. The dynamic allocation of computing resources and multi-threaded parallel computing of the cloud server 101 greatly improve the rendering speed. For example, under the configuration of 32-core CPU and 16GB GPU memory, the rendering time of a typical 3D scene is reduced to less than 1 minute;

[0073] The client 100 does not need to purchase expensive hardware equipment, but only needs to pay on demand to use the computing resources of the cloud server 101. Optimized data transmission and dynamic adjustment strategies are adopted to ensure stable rendering effects under different network conditions. Bandwidth adaptive technology is used to ensure smooth 1080p rendering effects under 10Mbps network.

[0074] In Example 2, the specific method of generating computing resources that match the 3D rendering request content of the user end by the computing power dynamic allocation algorithm in step 1 is: firstly use the following formula to calculate the computing power requirement of the 3D rendering request content of the user end:

[0075] ;

[0076] in The computing power required for the i-th rendering request, The priority weight of the i-th rendering request, is the resource requirement of the i-th rendering request, where the resource requirement refers to the number of CPU cores and GPU memory;

[0077] Then use the following formula to calculate the total dynamic computing power redundancy:

[0078] ;

[0079] in is the total dynamic computing power redundancy, is the dynamic redundancy coefficient, ranging from 10% to 30%. is the total available computing power of the cloud server 101 at present;

[0080] Finally, the following formula is used to generate computing resources that match the 3D rendering request content of the client 100:

[0081] ;

[0082] in is the computing power matching the i-th rendering request, j is an index variable, an integer from 0 to n, Used to traverse all rendering requests. Indicates the computing power required for the j-th rendering request;

[0083] is the total allocatable computing power of the cloud server 101 at present;

[0084] .

[0085] Embodiment 3, The value of is re-prioritized based on the currently running rendering tasks and the rendering tasks of the current rendering request. The priority re-prioritization considers the following four variables:

[0086] Task urgency score, resource requirement score, user importance score and estimated completion time score; calculated using the following formula :

[0087] ;

[0088] in is the priority score of the ith task, is the task urgency score of the i-th task, is the resource requirement score of the ith task, is the user importance score of the ith task, is the estimated completion time score of the i-th task;

[0089] The priority weight change value of all rendering tasks is recalculated based on the priority score of the tasks added to the rendering task of the current rendering request:

[0090] ;

[0091] in is the priority weight change value of the i-th task, is the weight change coefficient, is the original priority score of the i-th task before the rendering task of the current rendering request is added;

[0092] The priority weight value calculation formula of the rendering task of the current rendering request is:

[0093] .

[0094] Whenever a new 3D rendering request is input to the cloud server 101, the priority weight values ​​of all rendering tasks need to be recalculated, thereby achieving dynamic adjustment of computing resources, so that each 3D rendering task is performed under the optimal computing resource allocation, and improving the utilization rate of computing resources;

[0095] First, let's define some factors that affect the priority of rendering tasks:

[0096] Task urgency: How long does the task need to be completed?

[0097] Resource requirements: The amount of computing resources required for the task.

[0098] Client Importance: Tasks for different clients may have different priorities.

[0099] Estimated completion time: The estimated completion time based on task requirements and current resource usage.

[0100] By determining the above factors, the task priority score can be quantified, thereby achieving dynamic adjustment of the priority weight value.

[0101] In Embodiment 4, the cloud server 101 receives data and performs pre-processing including data format conversion and compression.

[0102] The client 100 uploads the 3D model and material data to be rendered, and the data size is generally between 100MB and 2GB. The cloud server 101 receives the data and performs preprocessing, including data format conversion and compression, to reduce the amount of data transmission. The compression ratio can reach 3:1.

[0103] In Example 5, in step 3, the cloud server 101 first provides a rendering thumbnail, which is generated based on the 3D rendering request content, and transmits the rendering thumbnail and the rendering resolution re-selection item back to the user terminal 100. After the user terminal 100 observes the rendering thumbnail, he finally determines the rendering resolution in the rendering resolution re-selection item. The user terminal 100 transmits the determined rendering resolution to the cloud server 101, and the cloud server 101 uses the computing resources allocated in step 1 to perform 3D rendering according to the determined rendering resolution.

[0104] Before actually running the rendering task, a rendering thumbnail that matches the initial rendering parameters can be generated according to the rendering parameters initially set, such as the rendering resolution, and the rendering thumbnail can be sent back to the user terminal 100 for confirmation of the final rendering parameters to avoid a large discrepancy between the final rendering result obtained by the user terminal 100 and the expected result. When sending back the rendering thumbnail, a rendering resolution re-selection option can also be provided. If the user at the user terminal 100 is not satisfied with the effect after viewing the rendering thumbnail, the rendering resolution requirement can be increased in the rendering resolution re-selection option.

[0105] To generate a rendering thumbnail that corresponds to the rendering parameters set before rendering, you need to add a pre-rendering stage to the rendering process, which is specifically used to generate thumbnails, including the following steps:

[0106] 1. Define thumbnail generation parameters:

[0107] Thumbnail Size: Determines the size of the thumbnail, which is usually much smaller than the final rendered image.

[0108] Rendering Settings: Select or simplify rendering settings, such as lowering the resolution, reducing lighting and shadow calculations, etc.

[0109] 2. Create a pre-rendering script

[0110] Write a script to set rendering parameters and generate thumbnails. Here is a simplified example flow:

[0111] a. Rendering engine preparation

[0112] Prepare the rendering environment according to the rendering engine used (such as Blender, Maya, 3dsMax, etc.).

[0113] b. Set rendering parameters

[0114] Adjust rendering parameters to suit thumbnail generation.

[0115] Here is the relevant code example:

[0116] # Assuming using Blender's Python API

[0117] import bpy

[0118] # Set the rendering resolution

[0119] bpy.context.scene.render.resolution_x = 256 # thumbnail width

[0120] bpy.context.scene.render.resolution_y = 256 # thumbnail height

[0121] # Simplify rendering settings

[0122] bpy.context.scene.render.use_shadows = False

[0123] bpy.context.scene.render.use_antialiasing = False

[0124] # Other necessary simplification settings.

[0125] Perform pre-rendering to generate thumbnails.

[0126] # Set the output path

[0127] bpy.context.scene.render.filepath = ' / path / to / thumbnail.png'

[0128] # Start rendering (this will generate thumbnails)

[0129] bpy.ops.render.render(write_still=True);

[0130] In Example 6, in step 4, according to the format of the rendering result, use Python and Pillow library to select a suitable compression algorithm from JPEG, PNG, WebP, and HEVC. If JPEG is used to compress the image, the compression ratio is adjusted by selecting a quality factor. If PNG is used to compress the image, the compression ratio is adjusted by selecting different compression levels. If WebP is used to compress the image, the compression ratio is adjusted by selecting a quality factor. If HEVC is used to compress the video, the compression ratio is adjusted by adjusting the quantization parameter.

[0131] In embodiment 7, the client 100 uses Python and Pillow library to decompress the file and restore the original rendering result.

[0132] After the rendering is completed, the cloud server 101 compresses the rendering result, and transmits the compressed rendering result back to the client 100 via the network, with a network bandwidth requirement of 10 Mbps to 100 Mbps. The client decompresses and displays the rendering effect, which reduces the network bandwidth requirement during the transmission process.

[0133] Embodiment 8, a cloud-based 3D effect rendering system, is used to implement a cloud-based 3D effect rendering method. The cloud-based 3D effect rendering system includes a user terminal 100, a network transmission module 103 and a cloud server 101. The user terminal 100 is connected to the cloud server 101 through the network transmission module 103. The cloud server 101 is provided with a data analysis chip 1011. The data analysis chip 1011 is used to run a computing power dynamic allocation algorithm to generate computing resources that match the 3D rendering request content of the user terminal 100.

[0134] In Example 9, both the cloud server 101 and the client 100 are provided with an encryption and decryption module 102 , and the encryption and decryption module 102 provides the same encryption protocol and algorithm for the client 100 and the cloud server 101 .

[0135] Embodiment 10 further includes a computing power warning module 104, which is connected to the cloud server 101 in communication with the cloud server 101. When the total available computing power of the cloud server 101 is determined, An alarm will be issued when it is less than 15% of the total computing power.

[0136] The computing power warning module 104 can calculate the total available computing power of the cloud server 101. When it is less than 15% of the total computing power, an alarm is issued to prevent the cloud server 101 from continuously running under a high load state;

[0137] Example 11, cloud-based 3D rendering task based on shoe design:

[0138] The client 100 uploads the 3D model and material data of the shoe design, the data size is 300MB, and the rendering resolution is set to 1080p. The cloud server 101 allocates 16 core CPU and 8GB GPU memory for rendering, and the rendering time is 30 seconds. The rendering results are transmitted back to the client via a 50Mbps network, and the client displays the real-time rendering effect, allowing the designer to instantly view and adjust the design details.

[0139] Example 12: Cloud-based 3D rendering task based on clothing style design:

[0140] The user uploads the 3D model and material data of the clothing style design, the data size is 600MB, and the rendering resolution is set to 4K. The cloud server allocates 32-core CPU and 16GB GPU memory for rendering, and the rendering time is 1 minute. The rendering results are transmitted back to the user end through the 100Mbps network for design display and modification, which facilitates designers to quickly iterate and optimize design solutions.

[0141] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A cloud-based 3D effect rendering method, characterized in that: The following steps are involved: Step 1, resource request and allocation: the client (100) sends a 3D rendering request content to the cloud server (101); The cloud server (101) generates computing resources matching the 3D rendering request content of the user terminal (100) by using a computing power dynamic allocation algorithm according to the 3D rendering request content and current computing resource usage data; Step 2, data transmission and preprocessing: the user end (100) uploads the 3D model and material data to be rendered corresponding to the 3D rendering request content, and the cloud server (101) receives the data and preprocesses it; Step 3, rendering task execution: the cloud server (101) performs 3D rendering using the computing resources allocated in step 1; Step 4, result transmission and display: After the rendering is completed, the cloud server (101) compresses the rendering result, and transmits the compressed rendering result data back to the user terminal (100), and the user terminal (100) decompresses and displays the rendering effect; The specific method of the computing power dynamic allocation algorithm in step 1 to generate computing resources that match the user-side 3D rendering request content is: first use the following formula to calculate the computing power requirement of the user-side 3D rendering request content: ； in The computing power required for the i-th rendering request, The priority weight of the i-th rendering request, is the resource requirement of the i-th rendering request, where the resource requirement refers to the number of CPU cores and GPU memory; Then use the following formula to calculate the total dynamic computing power redundancy: ； in is the total dynamic computing power redundancy, is the dynamic redundancy coefficient, ranging from 10% to 30%. is the total available computing power of the cloud server (101); Finally, the following formula is used to generate the computing resources that match the 3D rendering request content of the user end (100): ； in is the computing power matching the i-th rendering request, j is an index variable, an integer from (0) to n, Used to traverse all rendering requests. Indicates the computing power required for the j-th rendering request; is the total available computing power of the cloud server (101); ； The value of is re-prioritized based on the currently running rendering tasks and the rendering tasks of the current rendering request. The priority re-prioritization considers the following four variables: Task urgency score, resource requirement score, user importance score and estimated completion time score; Calculate using the following formula : ； in is the priority score of the ith task, is the task urgency score of the i-th task, is the resource requirement score of the ith task, is the user importance score of the ith task, is the estimated completion time score of the i-th task; The priority weight change value of all rendering tasks is recalculated based on the priority score of the tasks added to the rendering task of the current rendering request: ； in is the priority weight change value of the i-th task, is the weight change coefficient, is the original priority score of the i-th task before the rendering task of the current rendering request is added; The priority weight value calculation formula of the rendering task of the current rendering request is: 。 2. The cloud-based 3D effect rendering method according to claim 1, characterized in that: The cloud server (101) receives the data and performs pre-processing including data format conversion and compression.

3. The cloud-based 3D effect rendering method according to claim 2, characterized in that: In step 3, the cloud server (101) first provides a rendering thumbnail, which is generated based on the 3D rendering request content, and transmits the rendering thumbnail and the rendering resolution re-selection item back to the user end (100). After the user end (100) observes the rendering thumbnail, he finally determines the rendering resolution in the rendering resolution re-selection item. The user end (100) transmits the determined rendering resolution to the cloud server (101), and the cloud server (101) uses the computing resources allocated in step 1) to perform 3D rendering according to the determined rendering resolution.

4. The cloud-based 3D effect rendering method according to claim 3, characterized in that: In step 4, according to the format of the rendering result, use Python and Pillow library to select the appropriate compression algorithm from JPEG, PNG, WebP, and HEVC. If JPEG is used to compress the image, the compression ratio is adjusted by selecting the quality factor. If PNG is used to compress the image, the compression ratio is adjusted by selecting different compression levels. If WebP is used to compress the image, the compression ratio is adjusted by selecting the quality factor. If HEVC is used to compress the video, the compression ratio is adjusted by adjusting the quantization parameter.

5. The cloud-based 3D effect rendering method according to claim 4, characterized in that: The user end (100) uses Python and the Pillow library to decompress the file and restore the original rendering result.

6. A cloud-based 3D effect rendering system, characterized in that: A cloud-based 3D effect rendering method for implementing claim 5, wherein the cloud-based 3D effect rendering system comprises a user terminal (100), a network transmission module (103) and a cloud server (101), wherein the user terminal (100) is connected to the cloud server (101) for communication via the network transmission module (103), and the cloud server (101) is provided with a data analysis chip (1011), wherein the data analysis chip (1011) is used to run a computing power dynamic allocation algorithm to generate computing resources that match the 3D rendering request content of the user terminal (100).

7. The cloud-based 3D effect rendering system according to claim 6, characterized in that: The cloud server (101) and the user terminal (100) are both provided with an encryption and decryption module (102), and the encryption and decryption module (102) provides the same encryption protocol and algorithm for the user terminal (100) and the cloud server (101).

8. The cloud-based 3D effect rendering system according to claim 7, characterized in that: The computing power warning module (104) is also included. The computing power warning module (104) is in communication with the cloud server (101). When determining that the total available computing power of the cloud server (101) is An alarm will be issued when it is less than 15% of the total computing power.

Images