Automatic imaging method and system based on large model

Abstract

The invention relates to the technical field of automatic imaging, in particular to an automatic imaging method and system based on a large model. The method comprises the following steps: acquiring and analyzing imaging information, and determining an imaging keyword; analyzing the imaging keyword, and determining a picture element, a picture logic and a scene; and according to the picture logic, designing a plurality of picture modules, and according to the scene, combining and / or adjusting the plurality of picture modules to generate a picture. The advanced image recognition and natural language processing technology is used for automatically extracting and analyzing the imaging keywords from the imaging information, recognition errors are reduced, the accuracy of the imaging keywords is ensured, and therefore a more reliable basis is provided for subsequent picture design. A large model is utilized to deeply analyze imaging keywords, and picture elements, picture logic and scenes can be more accurately determined. The analysis is not only based on the keywords, but also takes context information into consideration, so that the rationality of picture design is improved.

Description

Technical Field

[0001] The present application relates to the field of automated imaging technology, and in particular to an automated imaging method and system based on a large model. Background Art

[0002] With the rapid development of science and technology, automated imaging technology has been widely used in various fields. In the fields of image processing, film and television production, virtual reality, etc., automated imaging technology can improve production efficiency, reduce costs, and bring users a richer visual experience.

[0003] However, due to the limitations of the design of traditional automated imaging methods, automated imaging technology often fails to achieve the ideal imaging effect; the imaging effect is affected by factors such as the accuracy of imaging keywords and the rationality of the picture logic, resulting in unsatisfactory picture effects. Existing technologies are difficult to cope with imaging tasks in different scenarios and with different needs; in practical applications, customized designs are often required according to specific scenarios and needs, and existing technologies are difficult to meet such flexibility requirements. Summary of the invention

[0004] The present application provides an automated imaging method and system based on a large model to solve the above-mentioned problems.

[0005] In a first aspect, the present application provides an automated imaging method based on a large model, the method comprising: Acquire and analyze imaging information and determine imaging keywords; Analyze the imaging keywords to determine picture elements, picture logic and scenes; According to the picture logic, several picture modules are designed and combined and / or adjusted according to the scene to generate a picture.

[0006] Through this solution, by using advanced image recognition and natural language processing technology, imaging keywords can be automatically extracted and analyzed from imaging information, reducing recognition errors. This helps to ensure the accuracy of imaging keywords, thereby providing a more reliable basis for subsequent screen design. Using large models to conduct in-depth analysis of imaging keywords can more accurately determine screen elements (such as object type, form, etc.), screen logic (such as the relationship between objects, actions, etc.), and scenes (such as background, atmosphere, etc.). This analysis is not only based on the keywords themselves, but also takes into account contextual information, thereby improving the rationality of screen design. Automated design of screen modules and combining and adjusting them according to scene requirements can ensure that the generated screen is more visually coordinated and natural. The system can automatically handle complex situations such as occlusion, motion trends, special effects, etc. according to pre-set rules and algorithms to generate high-quality imaging screens.

[0007] Optionally, analyzing the imaging keywords to determine picture elements, picture logic and scenes includes: Analyzing the imaging keywords to determine keyword semantic information; Determining screen elements according to the keyword semantic information; Filtering imaging keywords according to the keyword semantic information to obtain non-object keywords; Analyzing the non-object keywords to determine the behavior performance of the screen elements; Determine the screen logic based on the behavior performance; Analyze the behavior and predict possible scenarios; The scene corresponding to the imaging keyword is determined by combining the picture elements, the behavior performance and the possible scene.

[0008] Through this solution, through deep learning models and natural language processing technology, the extraction of imaging keywords is more accurate, errors are reduced, and the accuracy of keywords is improved. Using image recognition technology, it is possible to more accurately identify elements in the picture, such as people, objects, and environments, thereby providing a reliable data foundation for subsequent picture design and scene construction. By analyzing keyword semantic information, the system can more reasonably construct the picture logic, ensure that the relationship between elements in the picture conforms to the logic of the real world, and avoid unreasonable or unnatural picture effects. Through the analysis of behavioral performance and scene prediction, possible situations can be predicted more accurately, so that more possibilities can be taken into account during the imaging process, improving the adaptability and flexibility of imaging.

[0009] Optionally, the picture elements include object type and object form; and the determining the scene corresponding to the imaging keyword in combination with the picture elements and the possible scenes includes: Determining the scene atmosphere according to the object type and the behavior performance; Determining the degree of scene rendering according to the object form; Determining the scene tone according to the scene atmosphere and the scene rendering degree; Based on the scene keynote, the possible scenes are screened to obtain the scenes corresponding to the imaging keywords.

[0010] Through this solution, by analyzing the object type and behavior performance, imaging keywords can be extracted more accurately, thereby improving the accuracy of imaging; it can more accurately identify the object type and form in the picture, providing a reliable data basis for picture design and scene construction; through the analysis of picture elements, it can more reasonably construct the picture logic to ensure that the relationship between elements in the picture conforms to the logic of the real world; through the analysis of behavior performance and scene prediction, it can more accurately predict possible situations, thereby considering more possibilities in the imaging process; it can generate customized scenes according to different imaging keywords and picture elements to meet specific needs; it can handle a variety of scenes and needs, whether it is static images or dynamic videos, and can provide flexible imaging solutions; the automated imaging process reduces the time of manual design and improves imaging efficiency when processing a large number of images; through precise imaging keyword analysis and reasonable picture logic construction, the generated imaging effect is more natural and realistic, and the overall quality is improved.

[0011] Optionally, according to the screen logic, several screen modules are designed, including: Determine the picture level according to the picture logic; Based on the picture hierarchy, the scene and the picture elements are separated into layers to obtain an overall picture module; Determine an occlusion situation based on the behavior performance; Determining an element picture module of the picture element according to the occlusion condition; According to the picture logic, predict the movement trajectory of the picture elements; Determining a special effect module according to the motion trajectory; The overall picture module, the element picture module, and the special effect module are determined as a plurality of picture modules.

[0012] Through this solution, the details of the scene can be captured and reproduced more accurately through the analysis of the picture logic, thereby improving the accuracy of imaging; through layered and modular design, it is ensured that the elements in the picture are arranged in a logical order, enhancing the overall sense and coherence of the picture; through sophisticated layer design and special effects application, the visual quality of the picture is improved, making it more vivid and fascinating; modular design allows the picture to be quickly adjusted and customized according to different needs, improving the flexibility of imaging technology; through reasonable picture hierarchy and module allocation, the use of computing resources is optimized and imaging efficiency is improved.

[0013] Optionally, determining the occlusion situation according to the behavior performance includes: Determining the relationship between the elements of the picture according to the behavior performance; Based on the preset element model, according to the scene, the basic form of the screen elements is screened and determined; The occlusion situation is determined according to the basic form and the relationship.

[0014] Through this scheme, by analyzing the behavior performance, the occlusion situation can be predicted and determined more accurately, thereby improving the accuracy of imaging; considering the occlusion situation during the imaging process can ensure that the elements in the picture are arranged in a logical order, enhancing the overall sense and coherence of the picture; through fine occlusion processing, the visual quality of the picture is improved, making it more vivid and fascinating; determining the occlusion situation based on the behavior performance enables the imaging system to flexibly respond to different scenes and needs; through reasonable occlusion processing, the use of computing resources is optimized and the imaging efficiency is improved; by accurately predicting and determining the occlusion situation, the scene can be reproduced more realistically and the realism of imaging can be improved; in dynamic scenes, determining the occlusion situation based on the behavior performance can make the motion effect more natural and smooth; by analyzing the behavior performance and the occlusion situation, the visual hierarchy of the picture elements can be arranged more reasonably, and the readability and attractiveness of the picture can be improved; by determining the occlusion situation based on the behavior performance, customized design can be carried out according to different scenes and needs to meet diverse imaging tasks; through reasonable occlusion processing, the use of computing resources can be optimized and the imaging efficiency can be improved.

[0015] Optionally, combining and / or adjusting the plurality of picture modules according to the scenario to generate a picture includes: According to the described scene, determine the basic position of each screen element; Obtaining element features of each screen element; Determine an occlusion area according to the occlusion situation and the element characteristics; Determining a shielding ratio according to the basic station position and the shielding area; According to the occlusion ratio, the overall picture module and the element picture module are combined and / or adjusted to obtain a picture outline; According to the special effect module, the picture outline is rendered to generate a picture.

[0016] Through this solution, through precise occlusion processing and station layout, the generated picture can be made closer to the real world and the authenticity of the visual effect can be improved. By extracting and combining detail features, the objects in the picture can be made more three-dimensional and textured. Ensure that the elements of the picture are arranged in a logical order to avoid visual errors and confusion. Through the application of special effects modules, the artistic expression of the picture can be enhanced to make it more attractive. According to the specific needs of users, personalized imaging services are provided to meet diverse needs. Automated processing procedures can improve imaging efficiency and reduce waiting time. It can adapt to different scenes and environments and improve its stability and reliability.

[0017] Optionally, determining the picture hierarchy according to the picture logic includes: Obtain and analyze imaging requirements, and determine imaging size based on the results of the requirements analysis; Determining the screen layout of the screen elements according to the imaging size; Determine the number of picture layers according to the picture layout and the picture elements; The picture level is determined according to the number of picture layers.

[0018] Through this scheme, the audience's attention can be attracted by placing the most important visual elements at the most prominent level of the picture; a reasonable hierarchical structure helps to effectively convey information and make the content of the picture clearer and easier to understand; hierarchical design can enhance the artistic effect of the picture and make it more expressive and creative; in scene reproduction, hierarchical design helps to enhance the realism of the scene; according to different imaging technologies and platforms, hierarchical design can better adapt to technical requirements.

[0019] Optionally, determining the relationship between the screen elements according to the behavior performance includes: Analyze the image elements to determine dynamic objects and static objects; Analyzing the static object and the dynamic object to determine the object properties of the static object and the behavior characteristics of the dynamic object; According to the object attributes and the behavior characteristics, the relationship between the static object and the dynamic object is determined, and the relationship between the picture elements is obtained.

[0020] Through this solution, the relationship between dynamic objects and static objects is properly handled, so that the elements in the picture can set off each other and form a harmonious whole; through behavioral performance, the depth and details of the plot can be increased to make the audience more engaged; behavioral performance can guide the audience's sight and make the key elements in the picture more prominent; the behavior of dynamic objects and the properties of static objects can jointly create a specific scene atmosphere; in interactive media, the relationship determined according to behavioral performance can enhance the interactivity between users and the picture.

[0021] Optionally, after designing a plurality of picture modules according to the picture logic and combining the plurality of picture modules according to the scene, and / or adjusting to generate a picture, the method further includes: Analyze the generated images and determine the rationality of the images based on the analysis results; Determining whether to re-combine and / or re-adjust the plurality of image modules according to the rationality of the image; If it is determined to re-combine and / or re-adjust the plurality of image modules, then an unreasonable area is determined according to the image analysis result; Analyze each unreasonable area, determine the unreasonable cause, and perform secondary combination and / or secondary adjustment on each unreasonable area according to the unreasonable cause.

[0022] Through this solution, the visual continuity between the picture elements can be ensured through reasonable combination and adjustment of picture modules; through meticulous picture analysis, visual problems can be identified and solved, thereby improving the overall quality of the picture; ensuring the rationality of the picture logic, making the picture content more in line with the laws of the real world, and enhancing the audience's sense of conviction. The rationality of the picture logic helps to tell a coherent story and enhance the audience's sense of immersion; through secondary combination and adjustment of picture modules, the audience's visual experience can be optimized to make the picture more attractive; the optimized picture can better guide the audience's line of sight and highlight the key content; customized design can adapt to different scenarios and needs, making the picture more in line with specific environments and backgrounds.

[0023] In a second aspect, the present application provides an automated imaging system based on a large model, the system comprising: An information analysis module is used to obtain and analyze imaging information and determine imaging keywords; A keyword analysis module, used to analyze the imaging keywords to determine the image elements, image logic and scenes; The screen generation module is used to design a plurality of screen modules according to the screen logic and to combine and / or adjust the plurality of screen modules according to the scenario to generate a screen.

[0024] Optionally, when the keyword analysis module analyzes the imaging keywords to determine the picture elements, picture logic and scenes, it is used to: Analyzing the imaging keywords to determine keyword semantic information; Determining screen elements according to the keyword semantic information; Filtering imaging keywords according to the keyword semantic information to obtain non-object keywords; Analyzing the non-object keywords to determine the behavior performance of the screen elements; Determine the screen logic based on the behavior performance; Analyze the behavior and predict possible scenarios; The scene corresponding to the imaging keyword is determined by combining the picture elements, the behavior performance and the possible scene.

[0025] Optionally, the picture elements include object type and object form; when the keyword analysis module determines the scene corresponding to the imaging keyword by combining the picture elements, the behavior performance and the possible scene, it is used to: Determining the scene atmosphere according to the object type and the behavior performance; Determining the degree of scene rendering according to the object form; Determining the scene tone according to the scene atmosphere and the scene rendering degree; Based on the scene keynote, the possible scenes are screened to obtain the scenes corresponding to the imaging keywords.

[0026] Optionally, when the screen generation module designs a plurality of screen modules according to the screen logic, it is used to: Determine the picture level according to the picture logic; Based on the picture hierarchy, the scene and the picture elements are separated into layers to obtain an overall picture module; Determine an occlusion situation based on the behavior performance; Determining an element picture module of the picture element according to the occlusion condition; According to the picture logic, predict the movement trajectory of the picture elements; Determining a special effect module according to the motion trajectory; The overall picture module, the element picture module, and the special effect module are determined as a plurality of picture modules.

[0027] Optionally, when the picture generation module determines the occlusion situation according to the behavior performance, it is used to: Determining the relationship between the elements of the picture according to the behavior performance; Based on the preset element model, according to the scene, the basic form of the screen elements is screened and determined; The occlusion situation is determined according to the basic form and the relationship.

[0028] Optionally, when the screen generation module combines the plurality of screen modules according to the scenario and / or adjusts them to generate a screen, it is used to: According to the described scene, determine the basic position of each screen element; Obtaining element features of each screen element; Determine an occlusion area according to the occlusion situation and the element characteristics; Determining a shielding ratio according to the basic station position and the shielding area; According to the occlusion ratio, the overall picture module and the element picture module are combined and / or adjusted to obtain a picture outline; According to the special effect module, the picture outline is rendered to generate a picture.

[0029] Optionally, when the screen generation module determines the screen hierarchy according to the screen logic, it is used to: Obtain and analyze imaging requirements, and determine imaging size based on the results of the requirements analysis; Determining the screen layout of the screen elements according to the imaging size; Determine the number of picture layers according to the picture layout and the picture elements; The picture level is determined according to the number of picture layers.

[0030] Optionally, when the screen generation module determines the relationship between screen elements according to the behavior performance, it is used to: Analyze the image elements to determine dynamic objects and static objects; Analyzing the static object and the dynamic object to determine the object properties of the static object and the behavior characteristics of the dynamic object; According to the object attributes and the behavior characteristics, the relationship between the static object and the dynamic object is determined, and the relationship between the picture elements is obtained.

[0031] Optionally, the automated imaging system further includes an adjustment module, configured to: Analyze the generated images and determine the rationality of the images based on the analysis results; Determining whether to re-combine and / or re-adjust the plurality of image modules according to the rationality of the image; If it is determined to re-combine and / or re-adjust the plurality of image modules, then an unreasonable area is determined according to the image analysis result; Analyze each unreasonable area, determine the unreasonable cause, and perform secondary combination and / or secondary adjustment on each unreasonable area according to the unreasonable cause. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0033] Figure 1 A schematic diagram of an application scenario provided for an embodiment of the present application; Figure 2 A flowchart of an automated imaging method based on a large model provided in one embodiment of the present application; Figure 3 A schematic diagram of the structure of an automated imaging system based on a large model is provided for one embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0035] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0036] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0037] Due to the limitations of the design, traditional automated imaging methods often fail to achieve ideal imaging effects. Imaging effects are affected by factors such as the accuracy of imaging keywords and the rationality of the image logic, resulting in unsatisfactory image effects. Existing technologies are difficult to cope with imaging tasks in different scenarios and with different needs. In practical applications, customized designs are often required based on specific scenarios and needs, but existing technologies are difficult to meet this flexibility requirement.

[0038] Based on this, the present application provides an automated imaging method and system based on a large model, which acquires and analyzes imaging information, determines imaging keywords; analyzes imaging keywords, determines picture elements, picture logic and scenes; designs several picture modules according to the picture logic and combines several picture modules according to the scene, and / or adjusts to generate a picture. By using advanced image recognition and natural language processing technology, it is possible to automatically extract and analyze imaging keywords from imaging information, reducing recognition errors. This helps to ensure the accuracy of imaging keywords, thereby providing a more reliable basis for subsequent picture design. By using a large model to conduct in-depth analysis of imaging keywords, picture elements (such as object type, form, etc.), picture logic (such as the relationship between objects, actions, etc.) and scenes (such as background, atmosphere, etc.) can be more accurately determined. This analysis is not only based on the keywords themselves, but also takes into account contextual information, thereby improving the rationality of picture design. Automatically designing picture modules and combining and adjusting them according to scene requirements can ensure that the generated pictures are more visually coordinated and natural. The system can automatically handle complex situations such as occlusion, motion trends, special effects, etc. according to pre-set rules and algorithms to generate high-quality imaging pictures.

[0039] Figure 1A schematic diagram of an application scenario provided by the present application, when using a large model for automated imaging, the method provided by the present application is applied, specifically, the method provided by the present application is applied to any server, the server interacts with the interactive terminal, and by using advanced image recognition and natural language processing technology, it can automatically extract and analyze imaging keywords from the imaging information input by the interactive terminal, reducing the recognition error. This helps to ensure the accuracy of imaging keywords, thereby providing a more reliable basis for subsequent screen design. Using a large model to conduct in-depth analysis of imaging keywords, it is possible to more accurately determine screen elements (such as object type, form, etc.), screen logic (such as the relationship between objects, actions, etc.) and scenes (such as background, atmosphere, etc.). This analysis is not only based on the keywords themselves, but also takes into account contextual information, thereby improving the rationality of screen design. Automatically designing screen modules and combining and adjusting them according to scene requirements can ensure that the generated screen is more visually coordinated and natural. The system can automatically handle complex situations such as occlusion, motion trends, special effects, etc. according to pre-set rules and algorithms to generate high-quality imaging screens.

[0040] For specific implementation methods, please refer to the following embodiments.

[0041] Figure 2 This is a flowchart of a large model-based automated imaging method and system provided in one embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes: S201, acquiring and analyzing imaging information, and determining imaging keywords; Imaging information can be visual data contained in an image or video, as well as contextual information related to the visual data.

[0042] Imaging keywords may be words or phrases extracted from imaging information that can represent image content or user intent.

[0043] Specifically, traditional methods usually rely on manual recognition and classification of image content, and then extract keywords; this process may result in insufficient keyword accuracy due to complex image content or recognition errors. For example, if a picture contains multiple objects, some important objects may be missed or objects may be incorrectly recognized, thus affecting subsequent imaging effects. By introducing machine learning algorithms to assist in the recognition and classification of image content, the accuracy of keyword extraction can be improved.

[0044] S202, analyzing imaging keywords to determine picture elements, picture logic and scenes; Picture elements can be the basic elements that make up the picture in an image or video, including objects, colors, shapes, textures, etc.

[0045] Picture logic can be the relationship and interaction between the various elements in the picture and between the elements and the whole.

[0046] A scene can be a specific environment or situation depicted by an image or video.

[0047] Specifically, due to design limitations, the picture logic may be unreasonable or the scene selection may be inappropriate, which in turn affects the imaging effect. For example, if the keyword analysis is inaccurate, the relationship between objects may be misunderstood, resulting in a picture logic that does not meet expectations. For the extracted keywords, the semantic information of the keywords is analyzed through natural language processing technology to more accurately determine the picture elements and behavior logic to determine the elements in the picture (such as object type, form, etc.), picture logic (such as the relationship between objects, actions, etc.), and scenes (such as background, atmosphere, etc.).

[0048] S203 . Designing a plurality of picture modules according to picture logic and combining and / or adjusting the plurality of picture modules according to scenes to generate a picture.

[0049] The picture module can be a set of reusable visual elements designed to generate the final picture during the imaging process.

[0050] Specifically, the generated images may not be able to cope well with different scenarios and needs. For example, for dynamic scenes, it may be difficult to accurately predict the behavior and movement trends of objects, resulting in the lack of physical rationality of the generated images; for scenes that require a high degree of customization, it may not be able to adapt quickly, resulting in imaging effects that do not meet the specific needs of users. According to the previously determined picture logic and scenes, different picture modules are manually designed, and these modules are combined to generate the final imaging picture. Computer graphics technology is used to automatically design picture modules, and the combination and layout of modules are dynamically adjusted according to scene requirements. For example, an algorithm can be used to automatically determine the movement trend of an object, and adjust the rendering effect of the module according to the scene atmosphere.

[0051] Through this solution, by using advanced image recognition and natural language processing technology, imaging keywords can be automatically extracted and analyzed from imaging information, reducing recognition errors. This helps to ensure the accuracy of imaging keywords, thereby providing a more reliable basis for subsequent screen design. Using large models to conduct in-depth analysis of imaging keywords can more accurately determine screen elements (such as object type, form, etc.), screen logic (such as the relationship between objects, actions, etc.), and scenes (such as background, atmosphere, etc.). This analysis is not only based on the keywords themselves, but also takes into account contextual information, thereby improving the rationality of screen design. Automated design of screen modules and combining and adjusting them according to scene requirements can ensure that the generated screen is more visually coordinated and natural. The system can automatically handle complex situations such as occlusion, motion trends, special effects, etc. according to pre-set rules and algorithms to generate high-quality imaging screens.

[0052] In some embodiments, imaging keywords are analyzed to determine keyword semantic information; based on the keyword semantic information, picture elements are determined; based on the keyword semantic information, imaging keywords are screened to obtain non-object keywords; non-object keywords are analyzed to determine behavioral performance of picture elements; based on the behavioral performance, picture logic is determined; behavioral performance is analyzed to predict possible scenarios; and the scene corresponding to the imaging keyword is determined by combining the picture elements, behavioral performance and possible scenarios.

[0053] Keyword semantic information can be the meaning and contextual association contained in the keyword, including the basic definition of the keyword, synonyms, antonyms, related words, and their usage and emotional color in a specific context.

[0054] Non-object keywords can be words that do not directly describe specific objects or scenes in an image, but rather describe the overall atmosphere, emotion, action, or other non-visual entities of the image.

[0055] Behavior can be the action, state or dynamic characteristics of objects in the image, such as how objects move, interact or change, and how these behaviors affect the overall effect of the picture, including the collection of time, place, events or situations.

[0056] Possible scenes may be scenes that may be depicted by an image, inferred based on imaging keywords and behavioral performance.

[0057] Specifically, keywords are automatically extracted from the imaging information, which may involve natural language processing (NLP) technology, such as named entity recognition (NER) or keyword extraction algorithm; the extracted keywords are semantically analyzed using word sense disambiguation (WSD) technology and semantic networks to determine their meanings and contextual relationships. Based on the semantic information of keywords, the key elements in the image are identified through image recognition technology, such as object detection and image segmentation; the identified elements are classified, such as people, objects, environment, etc.; the keywords that are irrelevant or unimportant to the picture are screened out from the extracted keywords, and the keywords include background noise or irrelevant information; the screened non-object keywords are analyzed to determine their impact on the behavioral performance of the picture elements; how the non-object keywords affect the behavior of the picture elements, such as speed, direction, action, etc.; based on the identified behavioral patterns, a model is established to describe the behavior of the picture elements; based on the behavioral performance of the picture elements, the logical relationship between the elements in the picture is inferred; the behavioral performance is mapped to the picture logic, such as cause and effect, sequence relationship, etc.; based on the behavioral performance, possible scenes or scenarios are predicted; based on the predicted scenes, corresponding scene descriptions or scene graphs are generated; the imaging keywords are matched with the predicted scenes to determine the most appropriate scene; based on the semantic information of the imaging keywords, the details of the scene are refined, such as time, place, atmosphere, etc.

[0058] Through this solution, through deep learning models and natural language processing technology, the extraction of imaging keywords is more accurate, errors are reduced, and the accuracy of keywords is improved. Using image recognition technology, it is possible to more accurately identify elements in the picture, such as people, objects, and environments, thereby providing a reliable data foundation for subsequent picture design and scene construction. By analyzing keyword semantic information, the system can more reasonably construct the picture logic, ensure that the relationship between elements in the picture conforms to the logic of the real world, and avoid unreasonable or unnatural picture effects. Through the analysis of behavioral performance and scene prediction, possible situations can be predicted more accurately, so that more possibilities can be taken into account during the imaging process, improving the adaptability and flexibility of imaging.

[0059] In some embodiments, the scene atmosphere is determined based on the object type and behavior performance; the scene rendering degree is determined based on the object form; the scene tone is determined based on the scene atmosphere and the scene rendering degree; based on the scene tone, possible scenes are screened to obtain scenes corresponding to imaging keywords.

[0060] The mood of a scene can be a feeling or emotional state created in an image, conveyed through visuals such as color, light, shadow, movement, etc.

[0061] The degree of scene rendering can be the detail and complexity of the objects in an image or video; a high degree of rendering means that the image has more details and complex textures, while a low degree of rendering means that the image is simpler and has less details.

[0062] The tone of a scene can be the overall style and feel of an image, which is the result of the atmosphere of the scene, the degree of rendering, and other visual elements.

[0063] Specifically, convolutional neural network (CNN) object type recognition is used to identify the type of object in the picture, and the behavior of the object is analyzed through a behavior recognition algorithm. According to the object type and behavior performance, the atmosphere of the scene is inferred, and the morphological characteristics of the object, such as shape, size, and color, are analyzed. The morphological characteristics of the object are analyzed using a feature extraction method, and the complexity and degree of detail of the rendering are determined according to the morphological characteristics of the object and the complexity of the features. The scene atmosphere and the degree of rendering are combined, and the sentiment analysis technology is used to determine the overall tone of the scene, and a set of visual and emotional features are defined for the scene tone. According to the imaging keywords and the scene tone, a scene database or a generative model is used to generate and screen out potential scenes that meet the requirements, and a matching algorithm is used to match the imaging keywords with the screened scenes to determine the most suitable scene.

[0064] Through this solution, by analyzing the object type and behavior performance, imaging keywords can be extracted more accurately, thereby improving the accuracy of imaging; it can more accurately identify the object type and form in the picture, providing a reliable data basis for picture design and scene construction; through the analysis of picture elements, it can more reasonably construct the picture logic to ensure that the relationship between elements in the picture conforms to the logic of the real world; through the analysis of behavior performance and scene prediction, it can more accurately predict possible situations, thereby considering more possibilities in the imaging process; it can generate customized scenes according to different imaging keywords and picture elements to meet specific needs; it can handle a variety of scenes and needs, whether it is static images or dynamic videos, and can provide flexible imaging solutions; the automated imaging process reduces the time of manual design and improves imaging efficiency when processing a large number of images; through precise imaging keyword analysis and reasonable picture logic construction, the generated imaging effect is more natural and realistic, and the overall quality is improved.

[0065] In some embodiments, the picture hierarchy is determined according to the picture logic; based on the picture hierarchy, the scene and picture elements are separated into layers to obtain an overall picture module; the occlusion situation is determined according to the behavioral performance; according to the occlusion situation, the element picture module of the picture element is determined; according to the picture logic, the movement trajectory of the picture element is predicted; according to the movement trajectory, the special effect module is determined; the overall picture module, the element picture module, and the special effect module are determined as several picture modules.

[0066] A hierarchy is a sequence of elements in an image or video that are arranged in order of importance and visual impact.

[0067] Layer separation allows you to distribute different elements of an image or video onto different layers so that you can edit and adjust each layer individually without affecting the other layers.

[0068] The overall picture module can be the sum of all the layers and elements that make up the entire image or video, which represents the final visual output.

[0069] Occlusion is a situation where one element in an image blocks another element, which affects the visibility and visual priority of the elements.

[0070] An element picture module may refer to specific, independent visual elements in an image or video, such as characters, objects, backgrounds, etc., which can be edited and adjusted individually.

[0071] A motion trajectory can refer to the path along which an element in an image moves, which describes the dynamic changes of the element from its starting position to its ending position.

[0072] The special effect module can be a module for producing various visual special effects in order to enhance the visual effect of an image, such as halo, blur, shadow, particle effect, light effect, flame, etc.

[0073] Specifically, according to the picture logic, determine the hierarchical relationship of each element in the picture; based on the picture hierarchy, separate the scenes and picture elements into layers and create a layer structure; analyze the occlusion relationship of objects in the picture to determine which objects should be in the front layer and which in the back layer; according to the occlusion situation, determine the element picture modules of the picture elements; according to the behavioral performance, predict the movement trajectory of the picture elements; according to the movement trajectory, determine the special effects modules that need to be added; consider the scenes, layers, occlusion situations and other factors comprehensively to determine the overall picture module; integrate the layers and special effects of each picture element to form an element picture module; determine and integrate all necessary special effects modules.

[0074] Through this solution, the details of the scene can be captured and reproduced more accurately through the analysis of the picture logic, thereby improving the accuracy of imaging; through layered and modular design, it is ensured that the elements in the picture are arranged in a logical order, enhancing the overall sense and coherence of the picture; through sophisticated layer design and special effects application, the visual quality of the picture is improved, making it more vivid and fascinating; modular design allows the picture to be quickly adjusted and customized according to different needs, improving the flexibility of imaging technology; through reasonable picture hierarchy and module allocation, the use of computing resources is optimized and imaging efficiency is improved.

[0075] In some embodiments, the relationship between the screen elements is determined based on the behavioral performance; based on the preset element model, the basic form of the screen elements is screened and determined according to the scene; and the occlusion situation is determined based on the basic form and the relationship.

[0076] The preset element model can be a set of predefined models for describing and analyzing basic elements in a scene, which are stored in a preset database.

[0077] The basic form can be the basic shape or structure that constitutes the preset feature model, such as the simpler and more basic components of an object or scene.

[0078] Specifically, use behavior recognition technology, such as computer vision algorithms, to analyze the behavior patterns of objects in the picture. Combine time series analysis to capture the actions of the object, including movement, rotation, scaling, etc. Analyze the dynamic characteristics of the object's actions. Use scene classification algorithms to determine the scene type, analyze environmental factors in the scene, and use preset element models to screen possible screen elements based on the scene type. Match the objects in the actual captured image or video with the preset model to determine the basic form. Analyze the relationship between different objects in the picture, and infer the spatial relationship between objects based on the object's behavior and scene environment. Based on the object's behavior and dynamic analysis, predict the possible occlusion of the object. Use image processing algorithms to detect occluded areas in the image. Based on the results of occlusion detection, determine which screen elements are occluded and the degree of occlusion.

[0079] Through this scheme, by analyzing the behavior performance, the occlusion situation can be predicted and determined more accurately, thereby improving the accuracy of imaging; considering the occlusion situation during the imaging process can ensure that the elements in the picture are arranged in a logical order, enhancing the overall sense and coherence of the picture; through fine occlusion processing, the visual quality of the picture is improved, making it more vivid and fascinating; determining the occlusion situation based on the behavior performance enables the imaging system to flexibly respond to different scenes and needs; through reasonable occlusion processing, the use of computing resources is optimized and the imaging efficiency is improved; by accurately predicting and determining the occlusion situation, the scene can be reproduced more realistically and the realism of imaging can be improved; in dynamic scenes, determining the occlusion situation based on the behavior performance can make the motion effect more natural and smooth; by analyzing the behavior performance and the occlusion situation, the visual hierarchy of the picture elements can be arranged more reasonably, and the readability and attractiveness of the picture can be improved; by determining the occlusion situation based on the behavior performance, customized design can be carried out according to different scenes and needs to meet diverse imaging tasks; through reasonable occlusion processing, the use of computing resources can be optimized and the imaging efficiency can be improved.

[0080] In some embodiments, the basic position of each screen element is determined according to the scene; the element features of each screen element are obtained; the occlusion area is determined according to the occlusion situation and the element features; the occlusion ratio is determined according to the basic position and the occlusion area; according to the occlusion ratio, the overall screen module and the element screen module are combined and / or adjusted to obtain the screen outline; according to the special effects module, the screen outline is rendered to generate the screen.

[0081] The basic position can be the initial position or reference position set for each screen element (such as a person, object, building, etc.) in the scene.

[0082] In vision and image processing, an occlusion area refers to an area where certain visual elements cannot be directly seen due to obstruction by other objects or scene elements.

[0083] The occlusion ratio refers to the area ratio of the occluded area in the overall picture, usually expressed as a percentage. It is a quantitative indicator to measure the degree to which an object is occluded by another object.

[0084] The outline of a picture can be the lines or shapes that make up the edge of the picture, and it is usually determined by the main elements and shapes in the picture.

[0085] Specifically, analyze the overall layout, environmental characteristics and themes of the scene, identify the key elements in the scene, and determine the basic position of each screen element based on deep learning and computer vision algorithms according to the scene layout and the relative positions of the elements, that is, the spatial relationship between the elements; extract visual features from each screen element, use image processing technology, such as depth map to analyze the attributes of the elements, and determine which elements are blocked by other elements according to the characteristics and positions of the elements; evaluate the degree of occlusion and determine the occlusion area; calculate the occlusion ratio, that is, the ratio of the blocked part to the area of ​​the whole element through geometric calculation or image segmentation technology; adjust the position of the element according to the occlusion ratio to optimize the visual effect; use image editing and synthesis technology to combine the overall picture module and the element picture module according to the basic position and occlusion; make necessary adjustments to the picture module according to the scene and needs, select the appropriate special effect module according to the scene and picture needs, and apply the special effect module to the picture outline to enhance the visual effect; use rendering technology to integrate the picture module and special effects to generate the final image or video; conduct quality assessment on the rendering results to ensure that the expected effect is met.

[0086] Through this solution, through precise occlusion processing and station layout, the generated picture can be made closer to the real world and the authenticity of the visual effect can be improved. By extracting and combining detail features, the objects in the picture can be made more three-dimensional and textured. Ensure that the elements of the picture are arranged in a logical order to avoid visual errors and confusion. Through the application of special effects modules, the artistic expression of the picture can be enhanced to make it more attractive. According to the specific needs of users, personalized imaging services are provided to meet diverse needs. Automated processing procedures can improve imaging efficiency and reduce waiting time. It can adapt to different scenes and environments and improve its stability and reliability.

[0087] In some embodiments, imaging requirements are acquired and analyzed, and the imaging size is determined based on the requirements analysis results; the screen layout of the screen elements is determined based on the imaging size; the number of screen layers is determined based on the screen layout and screen elements; and the screen hierarchy is determined based on the number of screen layers.

[0088] Imaging requirements can be the specific requirements and expectations of users or customers regarding imaging content, quality, and purpose. They are the specific requirements of users for the final product during the image or video production process and are the starting point for imaging design and workflow.

[0089] Image size can be the resolution of an image or video, usually measured in pixels, inches, or centimeters.

[0090] Screen layout can be how the elements in an image or video are distributed and arranged, including the placement, size, direction and relationship of the elements, as well as the composition of the entire screen.

[0091] The number of picture layers may be the number of layers constituting an image or a video.

[0092] Specifically, collect the specific imaging requirements of users or customers, analyze the imaging requirements, and determine the key imaging elements and expected effects. According to the imaging requirements and analysis results, determine the appropriate imaging size, consider the size requirements of the output format, classify the screen elements, and plan the position and size of each element in the screen according to the imaging size and element classification. According to the imaging logic and requirements, determine the number of layers required for the screen, assign specific functions to each layer, determine the hierarchy of each element in the screen according to the importance and visual weight of the elements, assign visual weights to each element to determine its display priority in the screen, analyze the occlusion relationship between elements, and determine which elements should be located before or after other elements.

[0093] Through this scheme, the audience's attention can be attracted by placing the most important visual elements at the most prominent level of the picture; a reasonable hierarchical structure helps to effectively convey information and make the content of the picture clearer and easier to understand; hierarchical design can enhance the artistic effect of the picture and make it more expressive and creative; in scene reproduction, hierarchical design helps to enhance the realism of the scene; according to different imaging technologies and platforms, hierarchical design can better adapt to technical requirements.

[0094] In some embodiments, the picture elements are analyzed to determine dynamic objects and static objects; the static objects and dynamic objects are analyzed to determine the object properties of the static objects and the behavioral characteristics of the dynamic objects; based on the object properties and behavioral characteristics, the relationship between the static objects and the dynamic objects is determined to obtain the relationship between the picture elements.

[0095] A dynamic object can be an entity that can move or exhibit dynamic behavior in the physical world or in a virtual environment.

[0096] Static objects are entities that maintain a fixed position, shape, and state for a specific period of time and do not actively undergo physical movement.

[0097] Object attributes can be various characteristics that describe an object, including physical properties, appearance characteristics, and functional characteristics.

[0098] Behavioral characteristics can be the behavior patterns, movement habits or reaction characteristics exhibited by an object or individual in a specific environment; including the frequency, method, purpose and results of the action.

[0099] Specifically, identify all elements in the picture, classify them into static objects and dynamic objects according to their properties and behaviors, determine the static characteristics of static objects, analyze the motion trajectory, speed, direction, power source, etc. of dynamic objects, determine the dynamic characteristics of dynamic objects, analyze the interaction between static objects and dynamic objects, determine the spatial position relationship between static objects and dynamic objects, analyze the temporal relationship between static objects and dynamic objects, simulate the possible behavioral performance of dynamic objects according to their behavioral characteristics, determine their reactions or states under different behavioral performances according to their properties, verify whether the relationship between static objects and dynamic objects conforms to the logic of the real world, and ensure that the relationships and dynamic performances in the scene conform to the rationality of the scene.

[0100] Through this solution, the relationship between dynamic objects and static objects is properly handled, so that the elements in the picture can set off each other and form a harmonious whole; through behavioral performance, the depth and details of the plot can be increased to make the audience more engaged; behavioral performance can guide the audience's sight and make the key elements in the picture more prominent; the behavior of dynamic objects and the properties of static objects can jointly create a specific scene atmosphere; in interactive media, the relationship determined according to behavioral performance can enhance the interactivity between users and the picture.

[0101] In some embodiments, after designing a number of picture modules according to the picture logic and combining and / or adjusting the picture modules according to the scene to generate a picture, it also includes: analyzing the generated picture, and determining the rationality of the picture according to the picture analysis result; determining whether to re-combine and / or re-adjust the number of picture modules according to the picture rationality; if it is determined to re-combine and / or re-adjust the number of picture modules, determining unreasonable areas according to the picture analysis result; analyzing each unreasonable area, determining the unreasonable reason, and re-combining and / or re-adjusting each unreasonable area according to the unreasonable reason.

[0102] The rationality of the picture can be the consistency of the picture content in logic, vision and emotion, that is, whether the picture is real, whether it meets expectations, and whether it can effectively convey information, including the coherence of the scene, the coordination of colors, the rationality of light and shadow, etc., to ensure that the audience can accept and understand the meaning expressed by the picture.

[0103] Secondary combination can be the rearrangement or combination of different parts or elements of the picture, involving changes in the position, order or relationship of elements to enhance the appeal and expressiveness of the picture to achieve better visual or narrative effects.

[0104] Secondary adjustments can be made to certain details or the overall effect of the picture after the initial production, including color correction, light and shadow adjustment, detail enhancement, etc., to correct errors, improve quality or adapt to specific needs to enhance the professionalism and accuracy of the picture.

[0105] Unreasonable areas can be parts that exist in the picture and are inconsistent with the overall design intent or reality logic, as well as parts that appear visually abrupt, unnatural or unexpected.

[0106] The unreasonable reasons may be the basic reasons for the unreasonable phenomena in the analysis of the pictures, which are determined to be the cause of these phenomena, and may be due to factors such as design concepts, technical implementation, information processing or external factors.

[0107] Specifically, conduct a preliminary visual review of the generated image, check the overall composition, color, light and details, evaluate the logic of the image, ensure the consistency of the scene and story narration, check whether the image contains all necessary elements and information, judge whether the image is reasonable based on imaging requirements and expected effects, and compare the image with predetermined standards or references to determine whether it meets the requirements; if the image is unreasonable, identify the parts that need to be reassembled or adjusted; based on the results of problem identification, determine whether the image modules need to be reassembled or adjusted; identify unreasonable areas in the image, which may be due to problems with composition, color, lighting or other visual elements; record the location and nature of each unreasonable area in detail; analyze each unreasonable area to determine the cause of its unreasonableness; evaluate whether it is caused by technical limitations, such as rendering errors or algorithm problems; based on the analysis results, design necessary changes, including composition adjustment, color correction, lighting optimization, etc.; if necessary, reassemble the image modules to improve the overall effect; implement design changes and make secondary adjustments to the image.

[0108] Through this solution, the visual continuity between the picture elements can be ensured through reasonable combination and adjustment of picture modules; through meticulous picture analysis, visual problems can be identified and solved, thereby improving the overall quality of the picture; ensuring the rationality of the picture logic, making the picture content more in line with the laws of the real world, and enhancing the audience's sense of conviction. The rationality of the picture logic helps to tell a coherent story and enhance the audience's sense of immersion; through secondary combination and adjustment of picture modules, the audience's visual experience can be optimized to make the picture more attractive; the optimized picture can better guide the audience's line of sight and highlight the key content; customized design can adapt to different scenarios and needs, making the picture more in line with specific environments and backgrounds.

[0109] Figure 3 A schematic diagram of a structure of an automated imaging system based on a large model provided in an embodiment of the present application, such as Figure 3 As shown, the large model-based automatic imaging system 300 of this embodiment includes: an information analysis module 301 , a keyword analysis module 302 , and a picture generation module 303 .

[0110] The information analysis module 301 is used to obtain and analyze imaging information and determine imaging keywords; The keyword analysis module 302 is used to analyze the imaging keywords to determine the image elements, image logic and scenes; The screen generation module 303 is used to design a plurality of screen modules according to the screen logic and combine and / or adjust the plurality of screen modules according to the scenario to generate a screen.

[0111] Optionally, when the keyword analysis module 302 analyzes the imaging keywords to determine the picture elements, picture logic and scenes, it is used to: Analyzing the imaging keywords to determine keyword semantic information; Determining screen elements according to the keyword semantic information; Filtering imaging keywords according to the keyword semantic information to obtain non-object keywords; Analyzing the non-object keywords to determine the behavior performance of the screen elements; Determine the screen logic based on the behavior performance; Analyze the behavior and predict possible scenarios; The scene corresponding to the imaging keyword is determined by combining the picture elements, the behavior performance and the possible scene.

[0112] Optionally, the picture elements include object type and object form; when the keyword analysis module 302 determines the scene corresponding to the imaging keyword by combining the picture elements, the behavior performance and the possible scene, it is used to: Determining the scene atmosphere according to the object type and the behavior performance; Determining the degree of scene rendering according to the object form; Determining the scene tone according to the scene atmosphere and the scene rendering degree; Based on the scene keynote, the possible scenes are screened to obtain the scenes corresponding to the imaging keywords.

[0113] Optionally, when the screen generation module 303 designs a plurality of screen modules according to the screen logic, it is used to: Determine the picture level according to the picture logic; Based on the picture hierarchy, the scene and the picture elements are separated into layers to obtain an overall picture module; Determine an occlusion situation based on the behavior performance; Determining an element picture module of the picture element according to the occlusion condition; According to the picture logic, predict the movement trajectory of the picture elements; Determining a special effect module according to the motion trajectory; The overall picture module, the element picture module, and the special effect module are determined as a plurality of picture modules.

[0114] Optionally, when the image generation module 303 determines the occlusion situation according to the behavior performance, it is used to: Determining the relationship between the elements of the picture according to the behavior performance; Based on the preset element model, according to the scene, the basic form of the screen elements is screened and determined; The occlusion situation is determined according to the basic form and the relationship.

[0115] Optionally, when the screen generation module 303 combines the plurality of screen modules according to the scenario and / or adjusts them to generate a screen, it is used to: According to the described scene, determine the basic position of each screen element; Obtaining element features of each screen element; Determine an occlusion area according to the occlusion situation and the element characteristics; Determining a shielding ratio according to the basic station position and the shielding area; According to the occlusion ratio, the overall picture module and the element picture module are combined and / or adjusted to obtain a picture outline; According to the special effect module, the picture outline is rendered to generate a picture.

[0116] Optionally, when the screen generation module 303 determines the screen hierarchy according to the screen logic, it is used to: Obtain and analyze imaging requirements, and determine imaging size based on the results of the requirements analysis; Determining the screen layout of the screen elements according to the imaging size; Determine the number of picture layers according to the picture layout and the picture elements; The picture level is determined according to the number of picture layers.

[0117] Optionally, when the screen generation module 303 determines the relationship between screen elements according to the behavior performance, it is used to: Analyze the image elements to determine dynamic objects and static objects; Analyzing the static object and the dynamic object to determine the object properties of the static object and the behavior characteristics of the dynamic object; According to the object attributes and the behavior characteristics, the relationship between the static object and the dynamic object is determined, and the relationship between the picture elements is obtained.

[0118] Optionally, the automated imaging system 300 further includes an adjustment module 304, which is used to: Analyze the generated images and determine the rationality of the images based on the analysis results; Determining whether to re-combine and / or re-adjust the plurality of image modules according to the rationality of the image; If it is determined to re-combine and / or re-adjust the plurality of image modules, then an unreasonable area is determined according to the image analysis result; Analyze each unreasonable area, determine the unreasonable cause, and perform secondary combination and / or secondary adjustment on each unreasonable area according to the unreasonable cause.

[0119] The system of this embodiment can be used to execute the method of any of the above embodiments. The implementation principles and technical effects are similar and will not be described in detail here.

Claims

1. An automated imaging method based on a large model, characterized in that: include: Acquire and analyze imaging information and determine imaging keywords; Analyze the imaging keywords to determine picture elements, picture logic and scenes; According to the picture logic, several picture modules are designed and combined and / or adjusted according to the scene to generate a picture.

2. The method according to claim 1, characterized in that The analyzing the imaging keywords to determine the picture elements, picture logic and scenes includes: Analyzing the imaging keywords to determine keyword semantic information; Determining screen elements according to the keyword semantic information; Filtering imaging keywords according to the keyword semantic information to obtain non-object keywords; Analyzing the non-object keywords to determine the behavior performance of the screen elements; Determine the screen logic based on the behavior performance; Analyze the behavior and predict possible scenarios; The scene corresponding to the imaging keyword is determined by combining the picture elements, the behavior performance and the possible scene.

3. The method according to claim 2, characterized in that The picture elements include object type and object form; the combining the picture elements, the behavior performance and the possible scenes to determine the scene corresponding to the imaging keyword includes: Determining the scene atmosphere according to the object type and the behavior performance; Determining the degree of scene rendering according to the object form; Determining the scene tone according to the scene atmosphere and the scene rendering degree; Based on the scene keynote, the possible scenes are screened to obtain the scenes corresponding to the imaging keywords.

4. The method according to claim 2, characterized in that: According to the screen logic, several screen modules are designed, including: Determine the picture level according to the picture logic; Based on the picture hierarchy, the scene and the picture elements are separated into layers to obtain an overall picture module; Determine an occlusion situation based on the behavior performance; Determining an element picture module of the picture element according to the occlusion condition; According to the picture logic, predict the movement trajectory of the picture elements; Determining a special effect module according to the motion trajectory; The overall picture module, the element picture module, and the special effect module are determined as a plurality of picture modules.

5. The method according to claim 4, characterized in that The determining of the occlusion situation according to the behavior performance includes: Determining the relationship between the elements of the picture according to the behavior performance; Based on the preset element model, according to the scene, the basic form of the screen elements is screened and determined; The occlusion situation is determined according to the basic form and the relationship.

6. The method according to claim 5, characterized in that The combining and / or adjusting the plurality of picture modules according to the scene to generate a picture includes: According to the described scene, determine the basic position of each screen element; Obtaining element features of each screen element; Determine an occlusion area according to the occlusion situation and the element characteristics; Determining a shielding ratio according to the basic station position and the shielding area; According to the occlusion ratio, the overall picture module and the element picture module are combined and / or adjusted to obtain a picture outline; According to the special effect module, the picture outline is rendered to generate a picture.

7. The method according to claim 4, characterized in that The step of determining the picture hierarchy according to the picture logic includes: Obtain and analyze imaging requirements, and determine imaging size based on the results of the requirements analysis; Determining the screen layout of the screen elements according to the imaging size; Determine the number of picture layers according to the picture layout and the picture elements; The picture level is determined according to the number of picture layers.

8. The method according to claim 5, characterized in that Determining the relationship between the screen elements according to the behavior performance includes: Analyze the image elements to determine dynamic objects and static objects; Analyzing the static object and the dynamic object to determine the object properties of the static object and the behavior characteristics of the dynamic object; According to the object attributes and the behavior characteristics, the relationship between the static object and the dynamic object is determined, and the relationship between the picture elements is obtained.

9. The method according to claim 5, characterized in that After designing a plurality of picture modules according to the picture logic and combining the plurality of picture modules according to the scene, and / or adjusting to generate a picture, the method further includes: Analyze the generated images and determine the rationality of the images based on the analysis results; Determining whether to re-combine and / or re-adjust the plurality of image modules according to the rationality of the image; If it is determined to re-combine and / or re-adjust the plurality of image modules, then an unreasonable area is determined according to the image analysis result; Analyze each unreasonable area, determine the unreasonable cause, and perform secondary combination and / or secondary adjustment on each unreasonable area according to the unreasonable cause.

10. An automated imaging system based on a large model, characterized in that: include: An information analysis module is used to obtain and analyze imaging information and determine imaging keywords; A keyword analysis module, used to analyze the imaging keywords to determine the image elements, image logic and scenes; The screen generation module is used to design a plurality of screen modules according to the screen logic and to combine and / or adjust the plurality of screen modules according to the scenario to generate a screen.

Images