A large model-based intelligent design method for graphic publicity materials

Abstract

The application discloses a kind of based on big model's graphic propaganda material intelligent design method, it is related to graphic-text relationship modeling technical field, including the following steps: S1, constructs graphic propaganda material design input description set;S2, using design input description set carries out graphic-text expression relationship analysis;S3, using graphic-text expression relationship structure carries out graphic-text layout structure deduction;S4, using graphic-text layout structure description carries out design constraint uniformity;S5, using graphic propaganda material design constraint set carries out design consistency verification.The application is analyzed by setting based on design input description set and carried out graphic-text expression relationship, to graphic-text expression relationship structure, structured determination mechanism of corresponding relationship, attachment relationship and emphasis relationship is introduced between text expression unit and image expression unit, so that the semantic association between graphic and text is classified under information hierarchy constraint and output as stable graphic-text expression relationship structure.

Description

Technical Field

[0001] This invention relates to the field of graphic-text relationship modeling technology, specifically to an intelligent design method for graphic-text promotional materials based on a large model. Background Technology

[0002] With the widespread application of large-scale modeling technology in content generation, the design of graphic promotional materials is gradually shifting from manual layout to intelligent generation. Current technologies typically involve semantically encoding text content and image materials, then using similarity calculations or generative models to output a combined text and image result for rapid promotional material generation. However, graphic promotional materials not only require semantic relevance between text and images, but also consistency in information hierarchy, expressive emphasis, and visual presentation to ensure the accurate communication of promotional objectives.

[0003] In existing technologies, the construction of image-text relationships is mostly limited to the ranking of semantic similarity between text and images or the end-to-end generation of results. It does not fully combine the existing information hierarchy structure and design object boundary constraints in the design input, resulting in a lack of clear expression type distinction in image-text relationships. The generated results are difficult to support subsequent layout structure derivation and consistency verification. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an intelligent design method for graphic promotional materials based on large models, thereby solving the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, embodiments of the present invention provide an intelligent design method for graphic promotional materials based on a large model, comprising the following steps: S1. Construct a set of input descriptions for graphic and text promotional material design; S2. Use the design input description set to parse the graphic and textual expression relationships to obtain the graphic and textual expression relationship structure; S3. Use the text-image relationship structure to derive the text-image layout structure and obtain the text-image layout structure description. S4. Use graphic and text layout structure descriptions to unify design constraints and obtain a set of graphic and text promotional material design constraints. S5. Use the design constraint set of graphic promotional materials to verify the design consistency and obtain a description of the design results of the graphic promotional materials.

[0006] To further optimize this technical solution, step S1 transforms the original, natural language design requirements into a unified, explicit, and parsable set of design input descriptions; Step S1, during the construction of the input description set: Design target semantic parsing; Information content type identification and hierarchical classification; Analysis of publicity carrier formats and display constraints; The design aims to unify the expression of input information.

[0007] To further optimize this technical solution, the graphic promotional material design input description set obtained in step S1 is represented as follows: ; in: This is a subset describing the boundaries between the target audience and the information attributes. A subset describing the information content type and hierarchical structure; This is a subset describing the forms and display constraints of promotional media.

[0008] To further optimize this technical solution, step S2 involves the design input description set already generated in step S1. Based on this, the semantic relationships between text and image representation units are structurally analyzed to form a text-image representation relationship structure. ; Step S2, in parsing the relationship between the text and graphics, includes the following steps: Analysis of text and image representation units; Text-image semantic matching calculation; Determining the relationship between text and image.

[0009] To further optimize this technical solution, in step S2, when parsing text and image representation units, the information in the design input description set is decomposed into the smallest parsable unit; from The text information hierarchy is analyzed to form a set of text expression units. ; from and The range of images that can be represented is determined, forming a set of image representation units. ; pass Limit the semantic relevance of text and image units, and exclude units that are irrelevant to the target audience; Derivation of textual expression units With image expression unit .

[0010] To further optimize this technical solution, in step S2, when performing text-image semantic matching calculation, in order to quantify the potential semantic association between text and image, cross-modal semantic embedding technology is used to embed text units... With image unit Mapping to a unified semantic vector space: Text vector: ; Image vector: ; And calculate the text-image correlation: ; in Calculate vector similarity; This is a set of semantic matching degrees between text and images.

[0011] To further optimize this technical solution, in step S2, when determining the relationship between the text and image expressions, the semantic matching result is transformed into a clear type of text and image expression relationship. Combining hierarchical information of text units Matching degree Perform rule-based judgment: Core information corresponds to highly correlated images → Correspondence relationship ; Supplementary information is attached to the image → Dependency relationship ; Emphasizing key information visually → Highlighting relationships ; This leads to the set of relationships between text and graphics:

[0012] .

[0013] To further optimize this technical solution, step S3 takes the graphic and textual expression relationship structure output by step S2 and transforms the clearly defined graphic and textual expression relationship into a structural constraint description of the relative organization of graphics and text in the layout of promotional materials; Step S3: Final output of graphic layout structure description The output fully depicts the relative arrangement, adjacency, and visual hierarchy of text and images in promotional materials.

[0014] To further optimize this technical solution, step S4 describes the graphic and text layout structure output in step S3. The order of arrangement in Adjacency description Description of the relationship with visual hierarchy They are uniformly organized into a set of design constraints that can be directly invoked in subsequent design generation steps. .

[0015] To further optimize this technical solution, in step S5, after the graphic layout structure is derived and the design constraints are unified, the consistency of the graphic promotional material design results is verified, and the verification results are transformed into a structured graphic promotional material design result description. Step S5, when performing design conformity verification, includes the following steps: The design results are presented in a state analysis; Constraint-level consistency determination; Overall consistency results convergence; Design result description generated.

[0016] Secondly, embodiments of the present invention provide a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, they implement the steps of an intelligent design method for graphic promotional materials based on a large model as described in the first aspect of the present invention.

[0017] Thirdly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, they implement the steps of an intelligent design method for graphic promotional materials based on a large model as described in the first aspect of the present invention.

[0018] Compared with existing technologies, this invention provides an intelligent design method for graphic promotional materials based on large models, which has the following beneficial effects: This intelligent design method for graphic promotional materials based on a large model analyzes the graphic-textual relationship based on the design input description set, and introduces a structured judgment mechanism for correspondence, dependency, and emphasis relationships between textual and image expression units. This allows the semantic association between text and images to be clearly classified under information hierarchy constraints and output as a stable graphic-textual relationship structure. This provides a clear, interpretable, and reproducible input foundation for subsequent graphic-text layout structure derivation, effectively solving the problems of ambiguous graphic-textual relationship expression and difficulty in supporting structured design in existing technologies. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating an intelligent design method for graphic promotional materials based on a large model, as proposed in this invention. Figure 2This is a schematic diagram illustrating the graphic expression relationship analysis process of an intelligent design method for graphic promotional materials based on a large model proposed in this invention. Figure 3 This is a schematic diagram illustrating the derivation process of the graphic layout structure of an intelligent design method for graphic promotional materials based on a large model, as proposed in this invention. Figure 4 This is a schematic diagram illustrating the design consistency verification process of an intelligent design method for graphic promotional materials based on a large model, as proposed in this invention. Detailed Implementation

[0021] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0022] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0023] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0024] Example 1: Reference Figures 1-4 This is the first embodiment of the present invention, which provides an intelligent design method for graphic promotional materials based on a large model, including the following steps: S1. Construct a set of input descriptions for graphic and text promotional material design; Step S1 is used to structurally define the design problem of graphic promotional materials in the entire intelligent design method of graphic promotional materials based on large models; and to transform the original, natural language design requirements into a unified, clear and parsable set of design input descriptions.

[0025] Step S1, during the construction of the input description set: Design goal semantic analysis: Semantic analysis is performed on the original design requirements of graphic promotional materials to clarify the target audience and information boundaries of the design objectives; The parsing process uses mature natural language semantic parsing technology to parse the design requirement text at the sentence level, identify the object references, functional descriptions and limiting conditions, and perform semantic unification processing on ambiguous expressions according to general disambiguation rules. This analysis process yields a clear scope of the target audience and the boundaries of the information it can involve, which is used to define the semantic scope of subsequent content analysis.

[0026] Information content type identification and hierarchical classification; After completing semantic parsing, the information content contained in the design requirements is classified and hierarchically divided. Using mature text topic classification and semantic dependency analysis techniques, the parsed content is identified according to its information function, divided into core information, explanatory information, and supplementary information, and the hierarchical order of the above information is determined according to semantic dependency relationships. When the same content simultaneously satisfies multiple information characteristics, it is uniquely classified according to its main function in the overall expression to ensure a clear and stable information hierarchy structure.

[0027] Analysis of publicity carrier formats and display constraints; After clarifying the content structure, the display carrier format of the graphic promotional materials is analyzed to extract the constraints on the graphic design. By employing mature media attribute analysis and layout specification parsing technologies, we can identify the characteristics of the display area corresponding to the carrier, the limitations of the graphic and text presentation methods, and the insurmountable display conditions. The resulting display constraints are expressed in the form of descriptive constraints, and distinguish between immutable constraints and arrangement restrictions that must be followed, which are used to constrain the feasible range of text-image relationships and layout structure in subsequent steps.

[0028] Design a unified representation of input information; After completing the above analysis, the results are organized and presented in a standardized format.

[0029] By using mature structured description modeling methods, semantic boundaries, information levels, and display constraints are integrated according to preset fields to form input description content with clear structure and explicit field meanings.

[0030] The input description set for graphic and text promotional material design is represented as follows: ; in: This is a subset describing the boundaries between the target audience and information attributes of promotional materials. It is used to characterize the design object of graphic and textual promotional materials and the scope of information that can be involved, including: The target audience for the publicity campaign was clearly defined. The range of information attributes that the target audience is allowed to express; Information not included in the scope of publicity and expression; This is a subset describing the information content types and hierarchical structure, used to describe the internal structural relationships of the promotional content, including: The set of identified information content types; The hierarchical order between different information content types; A description of the hierarchical relationship between information content in the overall expression; This is a subset describing the forms and constraints of promotional media, used to describe the display media of graphic and textual promotional materials and the constraints on their formation, including: Description of the formal characteristics of the publicity medium; Display restrictions that must be followed when presenting text and images; Description of unbreakable boundary conditions.

[0031] S2. Use the design input description set to parse the graphic and textual expression relationships to obtain the graphic and textual expression relationship structure; The design input description set generated in step S1 Based on this, the semantic relationships between text and image representation units are structurally analyzed to form a text-image representation relationship structure. This provides a referenceable and stable input for the subsequent derivation of the graphic and text layout structure; Step S2, in parsing the relationship between the text and graphics, includes the following steps: Analysis of text and image representation units: Decompose the information in the design input description set into the smallest resolvable unit; from The text information hierarchy is analyzed to form a set of text expression units. ; from and The range of images that can be represented is determined, forming a set of image representation units. ; pass Limit the semantic relevance of text and image units, and exclude units that are irrelevant to the target audience; Derivation of textual expression units With image expression unit .

[0032] Text-image semantic matching calculation: To quantify the potential semantic relationships between text and images, based on mature cross-modal semantic embedding technology, text units are... With image unit Mapping to a unified semantic vector space: Text vector: ; Image vector: ; Calculate text-image correlation: ; in Calculate vector similarity (e.g., cosine similarity); This is a set of semantic matching degrees between text and images.

[0033] Determining the relationship between text and image; Transform semantic matching results into explicit graphic and textual relationship types.

[0034] Combining hierarchical information of text units Matching degree Perform rule-based judgment: Core information corresponds to highly correlated images → Correspondence relationship ; Supplementary information is attached to the image → Dependency relationship ; Emphasizing key information visually → Highlighting relationships ; This is only used to exclude image units that are not allowed to appear, and does not participate in spatial arrangement; This leads to the set of relationships between text and graphics:

[0035] .

[0036] Unlike existing technologies that mainly rely on text and image similarity ranking or end-to-end generation of text-image correspondences, this step introduces the design object boundary and information hierarchy structure output from step S1 to explicitly constrain the semantic matching process and refine the text-image relationship results into distinguishable relationship types such as correspondence, dependence, and emphasis. Ultimately, a structured text-image expression relationship structure that can be directly referenced by subsequent steps is formed, thereby achieving higher interpretability and reproducibility in the overall method.

[0037] S3. Use the text-image relationship structure to derive the text-image layout structure and obtain the text-image layout structure description. Step S3 follows the graphic-text expression relationship structure output from step S2, further transforming the already defined graphic-text expression relationship into a structural constraint description of the relative organization of graphics and text in the layout of promotional materials.

[0038] Step S3, in deriving the graphic layout structure, includes the following steps: The relationship between text and graphics and the deterministic mapping of layout constraints: For the types of expressions of relation that have been clearly defined in step S2 Establish fixed mapping rules, convert them into layout constraints, and form a set of layout constraints: ; in: when When representing correspondences, relative position constraints are generated. This is used to define the parallel or aligned relationship between text units and image units on the page; when When representing dependency relationships, adjacency constraints are generated. This is used to define the proximity organization method between graphic and text units; when When expressing an emphasis relationship, hierarchical constraints are generated. This is used to define the visual priority of text and image units on the page layout; This constraint only applies to the set of relations. The graph and text units that have relationships are bound together, but no constraints are imposed on unrelated units.

[0039] Construction of the text and image layout relationship matrix: In the set of layout constraints Based on this, construct a matrix of text and image layout relationships: ; Among them, matrix elements From the correspondence triple ( , , The constraints mapped to form a combination used to describe text units. With image unit The relative positional relationships, adjacency relationships, and hierarchical relationships in the layout structure.

[0040] Image and text layout structure description generation: Based on the image and text layout relationship matrix The text and image organization relationships in the overall layout are abstracted and summarized to form a text and image layout structure description, denoted as: ; in: This describes the overall arrangement order of text and image units on the page, derived from the layout constraint set. The relative position constraint subset in and layout relationship matrix The corresponding positional relationship information; through the matrix elements By organizing and comparing the overall positional constraint information, the arrangement order of text and image units in the page layout, such as sequential, parallel, or aligned, is abstracted. The adjacency descriptions between text and image units are derived from the layout constraint set. Adjacency constraint subset in and layout relationship matrix The matrix elements that represent adjacency relationships; through the matrix All elements containing adjacency constraints are extracted and summarized to form a structural description of whether text and graphic units need to be organized into adjacent units; The description of the visual hierarchy between text and image units is derived from the layout constraint set. Hierarchical constraint subsets in and layout relationship matrix The hierarchical priority information recorded in the matrix; through the matrix By comparing and summarizing the relevant hierarchical information, the relative visual priority relationships between graphic and textual units are obtained.

[0041] Step S3: Final output of graphic layout structure description The output fully depicts the relative arrangement, adjacency, and visual hierarchy of text and images in promotional materials.

[0042] Unlike existing technologies that commonly generate layouts directly based on preset templates or rule bases, or implicitly output complete layout results through end-to-end models, step S3 does not directly generate specific layout instances. Instead, it forms an interpretable intermediate layout structure description based on the parsed graphic-text relationship structure through a clear relationship mapping and constraint derivation process. This approach breaks down the layout design process into traceable structure derivation steps, thereby achieving a clear connection between graphic-text relationship parsing and layout organization in the overall method.

[0043] S4. Use graphic and text layout structure descriptions to unify design constraints and obtain a set of graphic and text promotional material design constraints. Step S4 describes the graphic and text layout structure output in step S3. The order of arrangement in Adjacency description Description of the relationship with visual hierarchy These are uniformly organized into a set of design constraints that can be directly invoked in subsequent design generation steps.

[0044] Step S4, when unifying design constraints, includes the following steps: Transformation of structural relationships into design constraint items: Specifically for , , The structural relationships contained therein are transformed into design constraint entries in a unified format using mature rule mapping and template expression methods: Will The arrangement order relationship is transformed into a sequence class design constraint, which is used to limit the organizational order of graphic and text units in the overall page layout; Will The adjacency relationships in the diagram are transformed into adjacency class design constraints, which are used to limit the proximity organization requirements between graphic and text units; Will The visual hierarchy in the image is transformed into hierarchical class design constraints, which are used to limit the relative visual priority between graphic and text units; Integration and consistency of design constraints: After generating various design constraint items, all constraints are integrated and processed in a unified manner. This process uses a mature constraint merging and consistency sorting method to merge duplicate items that come from the same graphic unit pair and belong to the same constraint type, while maintaining the parallel existence relationship between different types of constraints. The final expression for the set of design constraints for graphic promotional materials obtained in step S4 is: ; in, Arrangement order constraint set: ; Representing graphic and text units With text and image units The required order of arrangement; Adjacency constraint set: ; Representing graphic and text units With text and image units Adjacency requirements Visual hierarchy constraint set Elements in each constraint set , , All can be directly traced back to the output of step S3. , , Correspondence.

[0045] S5. Use the design constraint set of graphic promotional materials to verify the design consistency and obtain a description of the design results of the graphic promotional materials. Step S5 involves deriving the graphic and text layout structure and unifying the design constraints. Then, the consistency of the graphic and text promotional material design results is verified, and the verification results are transformed into a structured description of the graphic and text promotional material design results.

[0046] Step S5, when performing design conformity verification, includes the following steps: Design Result Presentation Status Analysis: For the generated graphic and text promotional material design, the graphic and text units are analyzed, including graphic and text unit boundary recognition, graphic and text unit spatial relationship analysis, and graphic and text unit visual hierarchy presentation state analysis. The actual presentation state of each graphic and text unit in the design result is extracted to form a set of design result presentation states:

[0047] in: Representing graphic and text units and The actual arrangement order in the design results; Representing graphic and text units and The actual spatial adjacency status in the design results; Representing graphic and text units and The actual visual hierarchy presented in the design results; This process is achieved through mature layout structure analysis and spatial relationship recognition technologies.

[0048] Constraint-level consistency determination: For each constraint element in the graphic and text promotional material design constraint set output in step S4, perform a consistency judgment between it and the corresponding presentation state to form a single constraint consistency result. The consistency determination is based on existing mature rule matching and logic verification technologies. The consistency determination uses the design constraints defined in step S4 as the determination benchmark. The actual presentation state obtained from the parsing of the design results is compared with the corresponding constraints item by item. Through condition satisfaction judgment, relationship consistency check and other methods, it is determined whether the design results meet the constraint requirements. For any constraint element The consistency determination result can be expressed as:

[0049] in: Indicates specific arrangement order constraints Adjacency constraints or visual hierarchy constraints ; This indicates the actual presentation state corresponding to the constraint; This represents the rule matching and logical verification process based on the constraint definition itself. This indicates the determination result of whether the constraint is satisfied by the design result; Set of permutation order constraints ,judge Does it meet the requirements? The defined order relationship; For the set of adjacency constraints ,judge Does it meet the requirements? The defined adjacency requirements; visual hierarchy constraint set ,judge Does it meet the requirements? The defined hierarchical priority relationship.

[0050] Overall consistency results convergence: All single-constraint consistency determination results are aggregated to form the overall design consistency result: ; in, This represents a mature consistency assessment method that summarizes the results of multiple constraint levels, used to reflect the degree to which the design results of graphic promotional materials satisfy the set of design constraints at the overall level.

[0051] Design result description generation: Based on the overall design consistency results Description of the generated graphic promotional material design results: ; The design results description expresses the design results in a structured form how well they meet the constraints of arrangement order, adjacency, and visual hierarchy, transforming the design results of graphic promotional materials from a simple visual presentation into a technical output with clear structural semantics.

[0052] Existing mature technologies typically rely on overall visual effect evaluation or human experience to confirm the rationality of graphic promotional material design results, lacking clear constraints and item-by-item verification paths.

[0053] The difference in the analytical logic of step S5 is as follows: Using the unified set of design constraints from step S4 as the sole verification benchmark, the design results are parsed into actual presentation states that correspond one-to-one with the constraints. A structured design result description is formed through constraint-level consistency judgment, thereby enabling the design result verification process to have clear judgment criteria and a reproducible technical path.

[0054] Example 2: This embodiment provides an application scenario of an intelligent design method for graphic promotional materials based on a large model in practical use: In a new product launch scenario for a large enterprise, a set of graphic and textual promotional materials for online promotion and offline display needed to be designed within a short period of time. This included a main visual poster, product feature introduction pages, and event guidance pages. The enterprise deployed an intelligent graphic and textual promotional material design method based on a large model, as described in this invention, into its promotional material design system to assist in the overall design of the promotional materials.

[0055] First, the system constructs a design input description set for graphic promotional materials based on the publicity objectives, distribution channels, and key product information. This set provides a unified description of the hierarchical relationship of the text content, key information, and the categories and uses of available image materials, thus providing clear design boundaries for subsequent processing.

[0056] Subsequently, based on the design input description set, the system analyzes the relationship between text expression units and image expression units, distinguishing which images are used to correspond to the explanatory text content, which images are used to attach and display specific details, and which images are used to enhance the visual expression of key information, thereby forming a clear image-text expression relationship structure.

[0057] Based on this, the system derives the overall layout structure of the graphic promotional materials according to the analyzed graphic and textual expression relationship structure, and clarifies the relative position, arrangement order and visual hierarchy of each text and image on the page, so that the layout structure can accurately reflect the importance of the information and the key points of expression.

[0058] Next, the system performs unified constraint processing on the derived graphic and text layout structure, integrating design requirements such as arrangement order, spatial adjacency, and visual hierarchy into a set of graphic and text promotional material design constraints to standardize the presentation of the final design results.

[0059] Finally, the system performs consistency verification on the generated graphic and text promotional material design results, determining whether the actual design results meet the set of design constraints, and outputs a structured description of the graphic and text promotional material design results. In this way, enterprises can quickly obtain graphic and text promotional materials that meet their promotional goals, have clear text-image relationships, and stable layout structures for promotion and distribution across different channels.

[0060] Example 3: This embodiment also provides a computer device applicable to an intelligent design method for graphic promotional materials based on a large model, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the intelligent design method for graphic promotional materials based on a large model as proposed in the above embodiment.

[0061] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements an intelligent design method for graphic promotional materials based on a large model, as proposed in the above embodiment.

[0062] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0063] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0064] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0065] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0066] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0067] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An intelligent design method for graphic promotional materials based on a large model, characterized in that, Includes the following steps: S1. Construct a set of input descriptions for graphic and text promotional material design; S2. Use the design input description set to parse the graphic and textual expression relationships to obtain the graphic and textual expression relationship structure; S3. Use the text-image relationship structure to derive the text-image layout structure and obtain the text-image layout structure description. S4. Use graphic and text layout structure descriptions to unify design constraints and obtain a set of graphic and text promotional material design constraints. S5. Use the design constraint set of graphic promotional materials to verify the design consistency and obtain a description of the design results of the graphic promotional materials.

2. The intelligent design method for graphic promotional materials based on a large model according to claim 1, characterized in that, Step S1 transforms the original, natural language design requirements into a unified, explicit, and parsable set of design input descriptions. Step S1, during the construction of the input description set: Design target semantic parsing; Information content type identification and hierarchical classification; Analysis of publicity carrier formats and display constraints; The design aims to unify the expression of input information.

3. The intelligent design method for graphic promotional materials based on a large model according to claim 2, characterized in that, The graphic and text promotional material design input description set obtained in step S1 is represented as follows: ； in: This is a subset describing the boundaries between the target audience and the information attributes. A subset describing the information content type and hierarchical structure; This is a subset describing the forms and display constraints of promotional media.

4. The intelligent design method for graphic promotional materials based on a large model according to claim 1, characterized in that, Step S2 refers to the design input description set generated in step S1. Based on this, the semantic relationships between text and image representation units are structurally analyzed to form a text-image representation relationship structure. ; Step S2, in parsing the relationship between the text and graphics, includes the following steps: Analysis of text and image representation units; Text-image semantic matching calculation; Determining the relationship between text and image.

5. The intelligent design method for graphic promotional materials based on a large model according to claim 4, characterized in that, In step S2, when parsing text and image representation units, the information in the design input description set is decomposed into the smallest parsable unit. from The text information hierarchy is analyzed to form a set of text expression units. ; from and The range of images that can be represented is determined, forming a set of image representation units. ; pass Limit the semantic relevance of text and image units, and exclude units that are irrelevant to the target audience; Derivation of textual expression units With image expression unit .

6. The intelligent design method for graphic promotional materials based on a large model according to claim 4, characterized in that, In step S2, during the text-image semantic matching calculation, to quantify the potential semantic association between text and image, cross-modal semantic embedding technology is used to embed text units... With image unit Mapping to a unified semantic vector space: Text vector: ; Image vector: ; And calculate the text-image correlation: ； in Calculate vector similarity; This is a set of semantic matching degrees between text and images.

7. The intelligent design method for graphic promotional materials based on a large model according to claim 4, characterized in that, In step S2, when determining the relationship between the text and image, the semantic matching result is transformed into a clear type of text and image relationship. Combining hierarchical information of text units Matching degree Perform rule-based judgment: Core information corresponds to highly correlated images → Correspondence relationship ; Supplementary information is attached to the image → Dependency relationship ; Emphasizing key information visually → Highlighting relationships ; This leads to the set of relationships between text and graphics: ； 。 8. The intelligent design method for graphic promotional materials based on a large model according to claim 1, characterized in that, Step S3 follows the graphic-text expression relationship structure output in step S2, further transforming the clearly defined graphic-text expression relationship into a structural constraint description of the relative organization of graphics and text in the layout of promotional materials; Step S3: Final output of graphic layout structure description The output fully depicts the relative arrangement, adjacency, and visual hierarchy of text and images in promotional materials.

9. The intelligent design method for graphic promotional materials based on a large model according to claim 1, characterized in that, Step S4 describes the graphic and text layout structure output in step S3. The order of arrangement in the description Adjacency description Description of the relationship with visual hierarchy They are uniformly organized into a set of design constraints that can be directly invoked in subsequent design generation steps. .

10. The intelligent design method for graphic promotional materials based on a large model according to claim 1, characterized in that, After the graphic layout structure is derived and the design constraints are unified, step S5 verifies the consistency of the graphic promotional material design results and transforms the verification results into a structured description of the graphic promotional material design results. Step S5, when performing design conformity verification, includes the following steps: The design results are presented in a state analysis; Constraint-level consistency determination; Overall consistency results convergence; Design result description generated.

Images