Three-dimensional rendering file batch splitting rendering layer processing system

Abstract

The invention discloses a three-dimensional rendering file batch splitting rendering layer processing system, and relates to the technical field of three-dimensional rendering, and the system comprises a data input module, a content intelligent analysis module, a rendering layer classification module, a real-time rendering preview module and a closed-loop management module; the data input module is used for receiving three-dimensional rendering files input by a user in batches and distributing unique identifiers to the three-dimensional rendering files, and by conducting intelligent analysis on elements in the three-dimensional rendering files, feature information, such as geometrical shapes, textures and spatial positions, of each element can be accurately captured; according to the method, the feature information of the rendering layer is obtained, scientific and reasonable rendering layer classification is carried out based on the feature information, the problem that traditional manual classification is tedious and prone to making mistakes is solved, the processing efficiency is greatly improved, a user does not need to carry out tedious manual operation, the system can automatically complete rendering layer classification only by simply inputting files, and the user experience is improved. And a lot of time and energy are saved for the user.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional rendering, and in particular to a three-dimensional rendering file batch splitting rendering layer processing system. Background Art

[0002] With the rapid development of 3D graphics technology, 3D rendering is increasingly used in various fields, such as 3D animation production, game development, and virtual reality. In these applications, 3D rendering files usually contain a large number of elements, such as characters, objects, backgrounds, and special effects. These elements need to be classified and processed according to different rendering layers to achieve more refined rendering effects and higher rendering efficiency. However, traditional 3D rendering file processing methods often rely on manual rendering layer classification, which is not only time-consuming and labor-intensive, but also prone to errors and cannot meet the needs of large-scale and efficient 3D rendering.

[0003] There are many shortcomings in the traditional 3D rendering file processing method. First, manual rendering layer classification requires a lot of time and effort, and the classification results often rely on the classifier's experience and subjective judgment, resulting in inconsistency and inaccuracy in the classification results. Secondly, the traditional 3D rendering method lacks real-time preview and adjustment functions during the rendering process. Users cannot discover and correct problems in time during the rendering process, resulting in reduced rendering efficiency and quality. In addition, when processing large-scale and complex 3D rendering files, the traditional 3D rendering method often has problems of slow rendering speed and low rendering quality, which cannot meet the needs of modern 3D graphics technology.

[0004] Therefore, a 3D rendering file batch splitting rendering layer processing system is developed. This innovative 3D rendering file processing method not only improves rendering efficiency and quality, but also reduces the cost and risk of manual intervention. It has broad application prospects and market value. Summary of the invention

[0005] The purpose of the present invention is to make up for the shortcomings of the prior art and provide a three-dimensional rendering file batch splitting rendering layer processing system, which can realize the automatic processing of three-dimensional rendering files through data input, content intelligent analysis, rendering layer classification, real-time rendering preview and interactive editing and output modules. The system can intelligently analyze the file content, classify the rendering layers based on element characteristics and semantic relationships, and provide real-time rendering preview function, supporting user interactive editing. The invention significantly improves the efficiency and flexibility of three-dimensional rendering, enables users to easily adjust the rendering layer order and parameters, and optimizes the rendering effect.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a three-dimensional rendering file batch split rendering layer processing system, the system comprising: a data input module, a content intelligent analysis module, a rendering layer classification module, a real-time rendering preview module and a closed-loop management module; The data input module receives the 3D rendering files input by the user in batches, assigns a unique identifier to the files, records basic information, preliminarily checks the integrity of the files, determines the file format based on the extension name and file header information, and passes the checked files to the content intelligent analysis module; The content intelligent analysis module: after receiving the 3D rendering file, it parses it and decomposes the file into different element data structures, respectively extracting the geometric shape, face, and clothing texture features of the human elements, the geometric shape, surface texture, and spatial position features of the object elements, the spatial layout, color tone, and light and shadow distribution features of the background elements, and the visual features of the special effect elements. The feature vector of each element is extracted through the element feature extraction formula. Suppose the element set in the 3D rendering file is , for each element , whose eigenvector The calculation formula is: ,in Is an element No. The original eigenvalues, is the weight of the corresponding feature, and these feature information are stored and associated with the element unique identifier. Based on the extracted features, the semantic association between elements is calculated through the semantic understanding formula. Let the element and , the semantic correlation between them The calculation formula is: ,in is from Different semantic dimensions evaluate elements and The function value of the relation, It is The weights of the semantic dimensions are calculated by using the evaluation functions of different semantic dimensions. , determine the interaction between characters and objects and their types, the logical relationship between objects, the role of background and special effects in the scene, and assign semantic labels to each element according to the value of the semantic understanding formula; The rendering layer classification module: creates a rule library based on multi-dimensional classification rules summarized from the research and analysis of three-dimensional rendering projects, classifies objects used for character interaction and decorative background objects into different rendering layers from a functional perspective, classifies elements with similar color tones and texture types into different layers from a visual feature perspective, and classifies elements related to specific characters into corresponding rendering layers from a semantic relationship perspective, obtains element features and semantic information extracted by the content intelligent analysis module, matches them with the rules in the rule library, searches for the most suitable rule for each element, and processes fuzzy and multiple matching situations according to pre-set priority rules. After the classification is completed, a detailed rendering layer classification list is generated, and the rendering layer information to which each element belongs is recorded; The real-time rendering preview module receives classified 3D rendering file information, initializes each rendering layer, allocates memory and GPU resources according to element information, renders characters and object elements by combining ray tracing algorithm and rasterization technology, and uses special algorithms to process large-area light and shadow effects and dynamic effects when processing background and special effect elements. The module has multi-thread processing capabilities, improves rendering speed through multi-core CPU and GPU parallel computing, and uses thread synchronization mechanism to ensure data security. It obtains rendering results in real time during rendering and displays them on the preview interface. When displaying, it is adjusted according to user settings and system default parameters, and supports users to zoom, rotate, and translate the preview interface. The interactive editing and output module provides users with rich interactive editing functions based on real-time rendering preview, allowing users to directly drag and drop in the preview interface to modify the rendering layer order, adjust the parameter slider to change the parameters of transparency and color saturation, and allow users to finely edit individual elements. The editing operation is reflected in real time in the preview interface. After the user completes the editing and confirms it, the processed 3D rendering file is output. The file retains the edited rendering layer information, can be directly used in 3D animation production, game development, virtual reality fields, and has good compatibility with other software and platforms.

[0007] Furthermore, the content intelligent analysis module analyzes the interactive relationship between people and objects, determines whether they are within a certain distance range based on their spatial position information, and further analyzes the person's posture and movement information and the object's attribute information within the distance range.

[0008] Furthermore, when analyzing the relationship between multiple objects in the content intelligent analysis module, the spatial position, shape and function information of the objects are compared to determine whether they belong to the same group. For objects with occlusion relationships, it is determined who occludes whom based on their depth information and rendering order. At the same time, the functional correlation between objects is analyzed.

[0009] Furthermore, the content intelligent analysis module understands the relationship between the background and other elements, analyzes how the background creates an atmosphere for the entire scene, and judges the role of special effects in the scene based on information about the type, location, and intensity of the special effects. Through semantic understanding, a semantic label is given to each element, and additional attribute data is stored in the feature data structure of the element, which together with other feature information constitutes a complete element description.

[0010] Furthermore, the rendering layer classification module calculates the probability that the element belongs to each rendering layer through the classification decision formula, determines the rendering layers into which the element is divided, and assumes that the rendering layer set is ,element Belongs to the rendering layer Probability The calculation formula is: ,in Is an element With render layers In the The matching degree on the classification factors is It is The weight of the classification factor, weight The elements are obtained by cross-validation optimization on a large number of annotated 3D rendering files. Probability of being classified Maximum render layer .

[0011] Furthermore, the setting of priorities in fuzzy and multiple matching situations in the rendering layer classification module is based on the priority of element type and function: for objects with fuzzy features, priority is set according to their main functions; priority is based on the importance of semantic relationships: an element has a potential interactive relationship with multiple other elements, and priority is given to semantic relationships related to the main characters and core plots. For situations where the semantic relationships between background elements and character elements are fuzzy, semantic relationships related to characters are given priority; priority is based on the clarity and importance of user operations: user operations involve multiple elements and rendering layers, and specific elements and rendering layers that are closer to the operation area are selected as priority elements and rendering layers. For operations that affect multiple elements, priority is set according to the clarity and importance of the operations, and fine-tuning operations on individual elements have higher priority than general adjustment operations on the overall scene.

[0012] Furthermore, the real-time rendering preview module determines whether the light intersects with the object in the scene by using the light-object intersection judgment formula of the distance field. Assume that the object in the scene is represented by the distance field function Indicates that is a point in three-dimensional space, Indicate point The shortest distance to the surface of an object. The distance inside the object is negative and outside is positive. The parametric equation of the light is: ,in It is the starting point of light. is the direction unit vector of the light, It is a parameter that defines a new function , the condition for the light to intersect the object is There is a solution, through Taylor expansion solution : ,in , , yes The gradient of t is solved by Newton iteration method, given a threshold , given a threshold , when the iteration value , no t that satisfies the conditions is found, and it is considered that the light does not intersect the object.

[0013] Furthermore, the rendering engine submodule in the real-time rendering preview module uses the light reflection direction formula to calculate the direction of the light after it is reflected on the surface of the object. Suppose the incident light direction vector is , the normal vector of the object surface is , the reflected light direction vector for: .

[0014] Furthermore, the real-time rendering preview module converts the three-dimensional model into a two-dimensional image through rasterization technology, and projects the object into a two-dimensional screen space according to its position and projection relationship in the three-dimensional space. For characters and object elements, the position and size of each polygonal surface on the screen are calculated, and the perspective effect and occlusion relationship are considered. The object corresponding to each pixel is determined through the depth buffer to solve the occlusion problem between objects. When processing texture mapping, the texture image is accurately mapped to the surface of the object according to the texture coordinates of the object surface, so that the object presents a real material appearance.

[0015] Compared with the prior art, the system for processing batch splitting rendering layers of 3D rendering files has the following beneficial effects: 1. The present invention can accurately capture the characteristic information of each element, such as geometric shape, texture, and spatial position, by performing intelligent analysis on the elements in the three-dimensional rendering file, and perform scientific and reasonable classification of rendering layers based on these characteristic information. This function not only avoids the tedious and error-prone problems of traditional manual classification, but also greatly improves the processing efficiency. Users no longer need to perform tedious manual operations, but only need to simply input the file, and the system can automatically complete the classification of the rendering layer, saving users a lot of time and energy.

[0016] 2. Through the real-time preview interface of the present invention, the user can intuitively understand the current rendering status and quickly adjust and optimize according to the automatically classified rendering layers. This function greatly enhances the user's operation convenience and enables the user to process 3D rendering files more efficiently. At the same time, the real-time preview function also provides rich interactive editing functions. The user can directly drag and drop in the preview interface to modify the rendering layer order and adjust parameters, making the rendering process more flexible and controllable. These features further enhance the practicality and user experience of the system, making the system have significant advantages in the field of 3D rendering.

[0017] Other advantages, objectives and features of the present invention will be set forth in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 A data flow diagram of a three-dimensional rendering file batch split rendering layer processing system; Figure 2 The present invention is a flowchart of a system for batch splitting rendering layers of 3D rendering files. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0021] Application in 3D animation production Preliminary preparation: The animation production team uses the system of the present invention to batch import multiple 3D rendering files containing characters, scenes, and special effect elements into the data input module. The system assigns a unique identifier to each file and records basic information, and checks the file integrity and format. For example, a 3D rendering file package containing an ancient castle scene, a protagonist knight, magic special effects, and various props is imported, totaling 50 files.

[0022] Intelligent content analysis: The intelligent content analysis module parses the file and extracts element features. For example, for the protagonist knight, it extracts its geometric shape (such as the proportion of each part of the body), facial features (expressions, facial features), clothing texture (armor texture details), the spatial layout of the castle background (the position relationship between towers and city walls), color tone (ancient and mysterious dark tones), light and shadow distribution (sunlight angle and shadow area), and the visual features of magic effects (light color, flashing frequency). It calculates the semantic correlation between elements. Suppose the element set in the 3D rendering file is , for each element , whose eigenvector The calculation formula is: ,in Is an element No. The original eigenvalues, is the weight of the corresponding feature, such as the "running" action element of the knight , which may have original eigenvalues ​​of "movement speed" and "limb swing amplitude" , , through the pre-set weights , Calculate the eigenvector , let element and , the semantic correlation between them The calculation formula is: ,in is from Different semantic dimensions evaluate elements and The function value of the relation, It is For example, when judging the semantic relevance between the knight and the castle gate, the semantic dimensions of "spatial position relationship" and "interaction possibility" are used to determine the semantic relevance between the knight and the castle gate. Evaluate and obtain , combined with the corresponding weights Calculate semantic relevance , and assign a semantic label to each element according to the value of the semantic understanding formula.

[0023] Rendering layer classification: The rendering layer classification module classifies elements according to classification rules. The interactive objects of the character knight and his weapons and equipment are divided into one rendering layer. The castle background and decorative objects are classified into multiple rendering layers according to the similarity of color tone and texture type. The elements related to the knight (such as the knight's mount and the elves following him) are divided into the corresponding character-related rendering layers. For fuzzy matching situations, such as a mysterious object in the castle that looks like both a decoration and an interactive mechanism, the probability of the element belonging to each rendering layer is calculated through the classification decision formula. Let the rendering layer set be ,element Belongs to the rendering layer The probability is: ,in Is an element With render layers In the The matching degree on the classification factors is It is The weight of the classification factor, weight The formula is obtained by cross-validation optimization on a large number of annotated 3D rendering files. The rendering layers into which the mysterious objects are divided are determined according to this formula, and a rendering layer classification list is generated.

[0024] Real-time rendering preview and editing: The real-time rendering preview module starts rendering, allocates memory and GPU resources, and renders elements by combining ray tracing and rasterization technology. For example, when rendering a knight, ray tracing calculates the reflection effect of light on the armor, and rasterization technology accurately handles its projection and texture mapping on the screen. In ray tracing, the distance field ray-object intersection judgment formula is used to determine whether the ray intersects with the objects in the scene. Assume that the objects in the scene are represented by the distance field function. Indicates that is a point in three-dimensional space, Indicate point The shortest distance to the surface of the object, the parametric equation of the light is: ,in It is the starting point of light. is the direction unit vector of the light, It is a parameter, defining a new function , the condition for the light to intersect the object is There is a solution, through Taylor expansion solution ,in , yes The gradient of is solved by Newton's iteration method , given a threshold , when the iteration value No results were found that met the conditions. , assuming that light does not intersect with objects, animators can perform interactive editing in the preview interface, adjust the brightness of knight armor to show the internal structure, change the color saturation of magic effects to enhance the visual impact, check scene details through zooming, rotating, and panning operations, and see the editing effects in real time.

[0025] Output finished product: After editing, the processed 3D rendering file is output and directly used for 3D animation production, ensuring good rendering effect of each element in the animation, good compatibility with other animation production software and platforms, and can be smoothly integrated into the final animation work.

[0026] According to the above embodiments, in the 3D animation production scenario, the system of the present invention demonstrates the ability to efficiently process 3D rendering files, from the orderly management of batch files in the early preparation stage, to the in-depth analysis of element features and semantic associations by the content intelligent analysis module, to the precise classification by the rendering layer classification module, and the real-time rendering preview module uses advanced technology to achieve fast and high-quality rendering preview and flexible editing, and finally outputs files suitable for animation production while ensuring compatibility. The entire process is closely connected, effectively improving the efficiency and quality of 3D rendering file processing during the animation production process, providing strong technical support for the animation production team, and helping to create more exquisite and vivid animation works. Example

[0027] Application in game development Resource import and analysis: The game development team batch inputs 3D rendering files of game scenes, character models, and props and special effects into the system. For example, a file collection containing multiple game level scenes, different character types, weapon props, and environmental effects, a total of 80 files.

[0028] The system processes files, and the intelligent content analysis module parses the files to extract the character's different postures and movements (such as standing, running, and attacking postures), clothing texture changes (textures under different equipment), the geometric shapes of object elements (props have different shapes), surface textures (weapon surface wear textures), spatial positions (the placement of props in the scene), spatial layout of background elements (level terrain layout), color tone (tone atmosphere of different scenes), light and shadow distribution (light and shadow changes between day and night), and the visual characteristics of special effect elements (skill release special effects).

[0029] Calculate the semantic association of elements. Suppose the element set in the 3D rendering file is , for each element , whose eigenvector The calculation formula is: ,in Is an element No. The original eigenvalues, is the weight of the corresponding feature, for example, for the character "attack" action element , which may have original feature values ​​of "attack speed" and "attack range" , through the pre-set weights , The calculated eigenvector of , let element and , the semantic correlation between them The calculation formula is: ,in is from Different semantic dimensions evaluate elements and The function value of the relation, It is The weight of each semantic dimension. For example, when judging the semantic relevance between a character and a weapon, the semantic dimensions of "equipment compatibility" and "usage frequency" are used to determine the semantic relevance between a character and a weapon. Evaluate and obtain , combined with the corresponding weights Calculate semantic relevance , and assign a semantic label to each element according to the value of the semantic understanding formula.

[0030] Rendering layer classification and optimization: The rendering layer classification module classifies according to rules. Characters and their interactive props are grouped into one layer. Buildings and ground background elements in the scene are grouped into different rendering layers according to color and texture similarity. Elements related to specific characters (such as character-exclusive skill effects) are divided into corresponding layers. Fuzzy and multiple matches are handled. For example, a wooden box that can be used as a scene decoration and can be destroyed by a character is used. The probability of an element belonging to each rendering layer is calculated through the classification decision formula. Suppose the rendering layer set is ,element Belongs to the rendering layer The probability calculation formula is: ,in Is an element With render layers In the The matching degree on the classification factors is It is The weight of the classification factor, weight The cross-validation optimization was performed on a large number of annotated 3D rendering files. , calculate its difference with different rendering layers (such as decoration layer and interaction layer ) in the "Function" "Visual Features" classification factor The matching degree on , , combined with the weight , determine the rendering layers into which the wooden boxes are divided, and generate a rendering layer classification list.

[0031] Real-time preview and adjustment: The real-time rendering preview module renders the scene, multi-threaded processing improves speed, and uses ray tracing and rasterization technology to present realistic images. When the player character moves in the scene, ray tracing calculates the interaction between light and objects, and rasterization ensures accurate display of the picture. In ray tracing, the distance field ray-object intersection judgment formula is used to determine whether the light intersects with the objects in the scene. Assume that the objects in the scene are represented by the distance field function Indicates that is a point in three-dimensional space, Indicate point The shortest distance to the surface of the object, the parametric equation of the light is: ,in It is the starting point of light. is the direction unit vector of the light, It is a parameter, defining a new function , the condition for the light to intersect the object is There is a solution, through Taylor expansion solution ,in , yes The gradient of is solved by Newton's iteration method , , given a threshold , when the iteration value No results were found that met the conditions. , assuming that the light does not intersect the object, when the light is reflected on the surface of the object, let the direction vector of the incident light be , the normal vector of the object surface is , the reflected light direction vector for For example, when light shines on a metal shield in the game scene, the direction of the reflected light is calculated based on the direction of the incident light and the normal direction of the shield surface to present realistic light and shadow effects. Game developers can directly drag and drop to modify the order of rendering layers in the preview interface, such as ensuring the display priority of important character rendering layers in advance, adjusting the transparency of special effects to avoid blocking key information, and fine-tuning individual elements (such as adjusting weapon model details) to view the effects in real time. When adjusting the transparency of special effects, you can adjust the transparency parameters according to the transparency parameters. ( ) for the color value of the special effect element Make adjustments, new color values , thereby changing the display effect of the special effects in the picture.

[0032] Integration and publishing: After editing is completed, the 3D rendering file is output and directly used in the game development engine and integrated with the game program code. The rendering effects of various elements in the game are excellent and compatible with different game platforms (PC, console, mobile), ensuring that players have a high-quality visual experience. For example, on different platforms, by adapting to different resolutions and graphics APIs, the game screen can be displayed correctly and run smoothly on various devices.

[0033] According to the above embodiments, the system of the present invention plays an important role in the field of game development. First, it comprehensively extracts the characteristics of various elements and accurately determines the semantic relationship during resource import and analysis. Then, it rationally plans the element layering in the rendering layer classification link, handles fuzzy situations to ensure accurate classification, and uses ray tracing and rasterization technology in the real-time rendering preview stage, combined with relevant formulas to achieve realistic picture presentation and efficient interactive editing. Finally, the output file can be perfectly integrated into the game development process and adapt to the requirements of different platforms. The system significantly improves the processing level of three-dimensional rendering resources in game development, helps to optimize the visual effects of the game, enhances the immersion and playability of the game, brings many conveniences to game developers, and promotes the improvement of game quality.

[0034] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A 3D rendering file batch splitting rendering layer processing system, characterized in that: The system includes: a data input module, a content intelligent analysis module, a rendering layer classification module, a real-time rendering preview module and a closed-loop management module; The data input module receives the 3D rendering files input by the user in batches, assigns a unique identifier to the files, records basic information, preliminarily checks the integrity of the files, determines the file format based on the extension name and file header information, and passes the checked files to the content intelligent analysis module; The content intelligent analysis module: after receiving the 3D rendering file, it parses it and decomposes the file into different element data structures, respectively extracting the geometric shape, face, and clothing texture features of the human elements, the geometric shape, surface texture, and spatial position features of the object elements, the spatial layout, color tone, and light and shadow distribution features of the background elements, and the visual features of the special effect elements. The feature vector of each element is extracted through the element feature extraction formula. Suppose the element set in the 3D rendering file is , for each element , whose eigenvector The calculation formula is: ,in Is an element No. The original eigenvalues, is the weight of the corresponding feature, and these feature information are stored and associated with the element unique identifier. Based on the extracted features, the semantic association between elements is calculated through the semantic understanding formula. Let the element and , the semantic correlation between them The calculation formula is: ,in is from Different semantic dimensions evaluate elements and The function value of the relation, It is The weights of the semantic dimensions are calculated by using the evaluation functions of different semantic dimensions. , determine the interaction between characters and objects and their types, the logical relationship between objects, the role of background and special effects in the scene, and assign semantic labels to each element according to the value of the semantic understanding formula; The rendering layer classification module: creates a rule library based on multi-dimensional classification rules summarized from the research and analysis of three-dimensional rendering projects, classifies objects used for character interaction and decorative background objects into different rendering layers from a functional perspective, classifies elements with similar color tones and texture types into different layers from a visual feature perspective, and classifies elements related to specific characters into corresponding rendering layers from a semantic relationship perspective, obtains element features and semantic information extracted by the content intelligent analysis module, matches them with the rules in the rule library, searches for the most suitable rule for each element, and processes fuzzy and multiple matching situations according to pre-set priority rules. After the classification is completed, a detailed rendering layer classification list is generated, and the rendering layer information to which each element belongs is recorded; The real-time rendering preview module receives classified 3D rendering file information, initializes each rendering layer, allocates memory and GPU resources according to element information, renders characters and object elements by combining ray tracing algorithm and rasterization technology, and uses special algorithms to process large-area light and shadow effects and dynamic effects when processing background and special effect elements. The module has multi-thread processing capabilities, improves rendering speed through multi-core CPU and GPU parallel computing, and uses thread synchronization mechanism to ensure data security. It obtains rendering results in real time during rendering and displays them on the preview interface. When displaying, it is adjusted according to user settings and system default parameters, and supports users to zoom, rotate, and translate the preview interface. The interactive editing and output module provides users with rich interactive editing functions based on real-time rendering preview, allowing users to directly drag and drop in the preview interface to modify the rendering layer order, adjust the parameter slider to change the parameters of transparency and color saturation, and allow users to finely edit individual elements. The editing operation is reflected in real time in the preview interface. After the user completes the editing and confirms it, the processed 3D rendering file is output. The file retains the edited rendering layer information, can be directly used in 3D animation production, game development, virtual reality fields, and has good compatibility with other software and platforms.

2. A 3D rendering file batch splitting rendering layer processing system according to claim 1, characterized in that: The content intelligent analysis module analyzes the interactive relationship between characters and objects, determines whether they are within a certain distance range based on their spatial position information, and further analyzes the character's posture and movement information and the object's attribute information within the distance range.

3. A 3D rendering file batch splitting rendering layer processing system according to claim 1, characterized in that: When analyzing the relationship between multiple objects in the content intelligent analysis module, the spatial position, shape and function information of the objects are compared to determine whether they belong to the same group. For objects with occlusion relationships, it is determined who occludes whom based on their depth information and rendering order. At the same time, the functional correlation between the objects is analyzed.

4. A 3D rendering file batch splitting rendering layer processing system according to claim 1, characterized in that: The content intelligent analysis module understands the relationship between the background and other elements, analyzes how the background creates an atmosphere for the entire scene, and judges the role of special effects in the scene based on information about the type, location, and intensity of the special effects. Through semantic understanding, a semantic label is given to each element, and additional attribute data is stored in the feature data structure of the element, which together with other feature information constitutes a complete element description.

5. The system for processing batch split rendering layers of 3D rendering files according to claim 1, characterized in that: The rendering layer classification module calculates the probability that the element belongs to each rendering layer through the classification decision formula, determines the rendering layers divided by the element, and sets the rendering layer set as ,element Belongs to the rendering layer Probability The calculation formula is: ,in Is an element With render layers In the The matching degree on the classification factors is It is The weight of the classification factor, weight The elements are obtained by cross-validation optimization on a large number of annotated 3D rendering files. Probability of being classified Maximum render layer .

6. A 3D rendering file batch splitting rendering layer processing system according to claim 1, characterized in that: The setting of priorities in fuzzy and multiple matching situations in the rendering layer classification module is based on the priority of element type and function: for objects with fuzzy features, the priority is set according to their main functions; the priority is based on the importance of semantic relationships: an element has a potential interactive relationship with multiple other elements, and the semantic relationship related to the main character and the core plot is given priority. For the situation where the semantic relationship between background elements and character elements is fuzzy, the semantic relationship related to the character is given priority; the priority is based on the clarity and importance of user operations: user operations involve multiple elements and rendering layers, and specific elements and rendering layers that are closer to the operation area are selected as priority elements and rendering layers. For operations that affect multiple elements, the priority is set according to the clarity and importance of the operations, and fine-tuning operations on individual elements have a higher priority than general adjustment operations on the overall scene.

7. The system for processing batch split rendering layers of 3D rendering files according to claim 1, characterized in that: In the real-time rendering preview module, the distance field light-object intersection judgment formula is used to determine whether the light intersects with the object in the scene. Assume that the object in the scene is represented by the distance field function Indicates that is a point in three-dimensional space, Indicate point The shortest distance to the surface of an object. The distance inside the object is negative and outside is positive. The parametric equation of the light is: ,in It is the starting point of light. is the direction unit vector of the light, It is a parameter that defines a new function , the condition for the light to intersect the object is There is a solution, through Taylor expansion solution : ,in , , yes The gradient of t is solved by Newton iteration method, given a threshold , given a threshold , when the iteration value , no t that satisfies the conditions is found, and it is considered that the light does not intersect the object.

8. The system for processing batch split rendering layers of 3D rendering files according to claim 1, characterized in that: The rendering engine submodule in the real-time rendering preview module uses the light reflection direction formula to calculate the direction of the light after reflection on the object surface. Suppose the incident light direction vector is , the normal vector of the object surface is , the reflected light direction vector for: .

9. The system for processing batch split rendering layers of 3D rendering files according to claim 1, characterized in that: In the real-time rendering preview module, the three-dimensional model is converted into a two-dimensional image through rasterization technology, and the object is projected into the two-dimensional screen space according to its position and projection relationship in the three-dimensional space. For characters and object elements, the position and size of each polygonal surface on the screen are calculated, and the object corresponding to each pixel is determined through the depth buffer considering the perspective effect and occlusion relationship, so as to solve the occlusion problem between objects. When processing texture mapping, the texture image is accurately mapped to the surface of the object according to the texture coordinates of the surface of the object, so that the object presents a real material appearance.

Images