A smart clothing modification design method, system, device and medium

Abstract

The present application relates to an intelligent clothing modification design method, system, device and medium, belonging to the field of clothing design technology, and the method includes: receiving clothing style images and modification requirement texts uploaded by a user terminal; identifying and obtaining various component areas of the clothing style image; semantically parsing the modification requirement text to obtain the component area to be modified, modification requirement parameters and modification style labels; marking the modification of the component area to be modified; determining the corresponding modification design elements based on the design element database; optimizing the modification design elements to obtain optimized design elements; replacing the original design elements of the component area to be modified with optimized design elements; performing visual optimization processing based on a boundary repair algorithm to obtain a clothing style image after partial modification; performing global style optimization based on the modification style label to obtain a clothing style image after modification. The present application can meet the user's needs for high-quality and highly personalized clothing modification design.

Description

Technical Field

[0001] The present application relates to the technical field of clothing design, and in particular to an intelligent clothing modification design method, system, device and medium. Background Art

[0002] In recent years, with the rapid growth of consumer demand for personalized clothing design, the efficiency and flexibility of traditional clothing design methods have been unable to meet market demand. Traditional design processes usually require designers to manually draw styles, adjust parts, match colors, and modify patterns, which is not only time-consuming and labor-intensive, but also requires a high level of experience from designers, limiting clothing design innovation and product development efficiency. In addition, for scenarios with mass and customized needs, such as clothing style updates on e-commerce platforms and collaborative design by designer teams, traditional methods lack intelligent and efficient tool support and are difficult to respond quickly to market changes.

[0003] In order to improve design efficiency, some design tools have been able to perform a certain degree of auxiliary processing on clothing style drawings, such as providing basic color replacement or pattern addition functions. However, such tools are usually limited to global color adjustment or pattern coverage, and lack the ability to intelligently understand and analyze user needs. Users' design needs often require complex manual operations to be realized in style drawings, which may lead to inconsistency between the local design and the overall clothing style after the modification, thus affecting the final design effect and making it difficult to meet users' needs for high-quality and highly personalized clothing modification designs. Summary of the invention

[0004] In order to meet users' demand for high-quality and highly personalized clothing modification design, the present application provides an intelligent clothing modification design method, system, device and medium.

[0005] In the first aspect, the present application provides a smart clothing modification design method, which adopts the following technical solutions:

[0006] A smart clothing modification design method, the method comprising:

[0007] Receive clothing style images and modification requirement text uploaded by user terminals;

[0008] Extracting features from the clothing style image to identify various component areas in the clothing style image;

[0009] Performing semantic analysis on the modification requirement text to obtain the component area to be modified, modification requirement parameters and modification style labels;

[0010] Segmenting the clothing style image according to each of the component regions, and marking the component regions to be modified based on the modification requirement parameters, to obtain the clothing style image after modification marking;

[0011] Based on the design element database, determining corresponding facelift design elements according to the facelift requirement parameters and the facelift style labels;

[0012] According to the original design elements of the component area to be modified, the modified design elements are optimized to obtain optimized design elements;

[0013] Based on the clothing style image after the modification mark, the original design elements of the component area to be modified are replaced with the optimized design elements to obtain the modified component area;

[0014] Performing visual optimization processing on the modified component area based on a boundary repair algorithm to obtain a clothing style image after partial modification;

[0015] Based on the modified style label, a global style optimization is performed on the clothing style image after the partial modification to obtain the clothing style image after the modification.

[0016] By adopting the above technical solutions, accurate identification of each component of clothing and accurate mapping of modification requirements are achieved, ensuring a high degree of automation in the design process. By converting the user's modification requirements into efficient, coordinated and personalized clothing design results, and using the matching, optimization and boundary repair algorithms of design elements, the visual coordination and natural integration of local design and overall style are ensured. Finally, global style optimization further ensures that the overall design meets the target style expected by the user, meets the user's needs for efficient, high-quality and highly personalized clothing modification design, and improves the design's completion and market adaptability.

[0017] Optionally, based on the design element database, the step of determining corresponding facelift design elements according to the facelift requirement parameters and the facelift style labels includes:

[0018] Loading a design element database; the design element database includes colors, patterns and component styles;

[0019] According to the model modification requirement parameters, matching corresponding design elements in the design element database to obtain design element matching results and corresponding similarity scores;

[0020] Sorting the design element matching results according to the similarity scores, and filtering based on the facelift style labels to obtain a sorted and filtered design element list;

[0021] Sending the sorted and screened design element list to the user terminal for selection;

[0022] Receive and respond to the design element selection instruction sent by the user terminal, and determine the corresponding modified design element.

[0023] By adopting the above technical solution, accurate matching and determination of modification requirements to design elements are achieved. Through multi-dimensional feature matching and style screening, the fit between design materials and modification requirements is improved. Through user terminal interaction, the personalization of design results and user satisfaction are enhanced. At the same time, this technical solution has significant advantages in efficiency, flexibility and accuracy, and is suitable for application scenarios of personalized clothing design and intelligent modification.

[0024] Optionally, the step of matching corresponding design elements in the design element database according to the model modification requirement parameters to obtain design element matching results and corresponding similarity scores includes:

[0025] Classifying the model change demand parameters and generating corresponding matching rules for each demand parameter category;

[0026] Retrieving corresponding candidate design elements in the design element database according to the model modification requirement parameters to obtain a list of candidate design elements;

[0027] Based on the matching rule, calculating a similarity score between each candidate design element in the candidate design element list and the model modification requirement parameter;

[0028] The candidate design elements whose similarity scores are greater than a preset threshold are screened to obtain design element matching results and corresponding similarity scores.

[0029] By adopting the above technical solutions, the modification requirements are accurately matched with the materials in the design element database, the requirements are clarified through classification, the similarity is accurately calculated based on multiple algorithms, and the design elements with the highest degree of fit are screened out, thereby ensuring that the design results not only meet the user's demand parameters, but also have high precision and reliability, thereby significantly improving the efficiency and accuracy of clothing modification design.

[0030] Optionally, the matching rules include color matching rules, pattern matching rules and component style matching rules, and the similarity scores include color similarity scores, pattern similarity scores and component style similarity scores.

[0031] Optionally, the step of optimizing the modified design elements according to the original design elements of the component area to be modified to obtain the optimized design elements includes:

[0032] Extracting characteristics according to the original design elements of the component area to be modified to obtain original design characteristics;

[0033] Performing characteristic analysis on the facelift design elements to obtain facelift design characteristics;

[0034] Calculating the characteristic difference between the original design characteristic and the modified design characteristic to obtain a characteristic difference analysis result;

[0035] Based on the characteristic difference analysis result, the facelift design characteristics of the facelift design element are dynamically adjusted to obtain an optimized design element.

[0036] By adopting the above technical solutions, the precise optimization and natural integration of the facelift design elements and the original area in terms of color, texture and geometric shape are achieved, which improves the visual coordination, geometric adaptability and overall aesthetic of the design elements and the original area. At the same time, the efficiency and applicability of the facelift design are significantly improved through the optimization process, meeting the needs of personalized and high-precision design.

[0037] Optionally, the step of performing visual optimization processing on the modified component region based on a boundary repair algorithm to obtain a clothing style image after partial modification includes:

[0038] Based on an edge detection algorithm, a boundary curve of the modified component region is extracted to determine a corresponding boundary region;

[0039] Determining the original component area and the modified component area according to the areas on both sides of the boundary area;

[0040] extracting differences in color and texture characteristics between the original component region and the modified component region;

[0041] Performing repair processing on the boundary area based on the color and texture characteristic differences to obtain an optimized boundary area;

[0042] The optimized boundary area is replaced into the clothing style image to obtain a clothing style image after partial modification.

[0043] By adopting the above technical solutions, high-quality optimization of images of partially modified clothing styles is achieved. By enhancing the continuity and smoothness of the boundaries, the natural transition between the new and old areas is achieved by using color gradient and texture fusion technology. At the same time, global adjustments are made to ensure the visual consistency between the partially modified areas and the overall design, thereby significantly improving the overall beauty and design completion of the clothing styles.

[0044] Optionally, the step of performing global style optimization on the clothing style image after partial modification based on the modified style label to obtain the clothing style image after the modification includes:

[0045] Parsing the facelift style label and mapping it to obtain target style characteristics;

[0046] Extracting global features based on the clothing style image after partial modification to obtain current global features;

[0047] Calculating a style characteristic difference between the target style characteristic and the current global characteristic;

[0048] The current global characteristics are optimized and adjusted based on the style characteristic differences to obtain a clothing style image after the modification.

[0049] By adopting the above technical solutions, the comprehensive unification of the image of the clothing style after partial modification and the target style characteristics is accurately achieved. Through the optimization of global color, pattern and texture, the local and overall designs are visually highly coordinated, which significantly enhances the overall beauty and design completion of the clothing style, while improving the intelligence and efficiency of the modification design.

[0050] In the second aspect, the present application provides an intelligent clothing modification design system, which adopts the following technical solutions:

[0051] An intelligent clothing modification design system, the system comprising:

[0052] A receiving module, used to receive clothing style images and modification requirement texts uploaded by a user terminal;

[0053] A component region extraction module is used to extract features from the clothing style image and identify each component region in the clothing style image;

[0054] A semantic analysis module, used to perform semantic analysis on the modification requirement text to obtain the component area to be modified, modification requirement parameters and modification style labels;

[0055] A style change marking module, used for performing region segmentation on the clothing style image according to each of the component regions, and performing style change marking on the component regions to be changed based on the style change requirement parameters, to obtain a clothing style image after the style change marking;

[0056] A facelift design element determination module, configured to determine corresponding facelift design elements based on a design element database and according to the facelift requirement parameters and facelift style labels;

[0057] A design element optimization module, used to optimize the modified design elements according to the original design elements of the component area to be modified, so as to obtain optimized design elements;

[0058] A design element replacement module, for replacing the original design elements of the component region to be modified with the optimized design elements based on the modified clothing style image, so as to obtain the modified component region;

[0059] A boundary repair module, used for performing visual optimization processing on the modified component area based on a boundary repair algorithm to obtain a clothing style image after partial modification;

[0060] The global optimization module performs global style optimization on the clothing style image after partial modification based on the modification style label to obtain the clothing style image after the modification.

[0061] In a third aspect, the present application provides a computer device, which adopts the following technical solution:

[0062] A computer device comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method as described in the first aspect.

[0063] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution:

[0064] A computer-readable storage medium stores a computer program that can be loaded by a processor and execute any one of the methods in the first aspect.

[0065] To summarize, the present application includes at least one of the following beneficial technical effects: by converting the user's modification needs into efficient, coordinated and personalized clothing design results, and utilizing the matching, optimization and boundary repair algorithms of design elements to ensure the visual coordination and natural integration of local design and overall style, it meets the user's needs for efficient, high-quality and highly personalized clothing modification design. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a first flow chart of the intelligent clothing modification design method according to one of the embodiments of the present application.

[0067] Figure 2 This is a second flow chart of the intelligent clothing modification design method according to one of the embodiments of the present application.

[0068] Figure 3 This is a third flow chart of the smart clothing modification design method according to one of the embodiments of the present application.

[0069] Figure 4 This is a fourth process diagram of the intelligent clothing modification design method according to one of the embodiments of the present application.

[0070] Figure 5 This is a fifth flow chart of the intelligent clothing modification design method according to one of the embodiments of the present application.

[0071] Figure 6 This is the sixth flow chart of the intelligent clothing modification design method according to one of the embodiments of the present application. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figure 1-6 It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0073] The embodiment of the present application discloses a smart clothing modification design method.

[0074] Reference Figure 1 , a smart clothing modification design method, the method comprising:

[0075] Step S101, receiving clothing style images and style change requirement text uploaded by a user terminal;

[0076] Among them, users upload clothing style images and modification requirement texts through the interactive interface. Clothing style images are the core basic input data, which are used for subsequent feature extraction and modification operations; modification requirement texts are input through natural language, which is an expression of user design intentions and needs.

[0077] For example, the image upload module can receive user-uploaded clothing style pictures and perform format verification to ensure that the image file is complete and clear and can be processed. For example, if a user uploads an image of a dress and inputs the text "change the neckline to a round neckline and the overall style to a retro style", the system can extract the three key requirements of "neckline", "round neckline" and "retro style".

[0078] Step S102, extracting features from the clothing style image to identify various component regions in the clothing style image;

[0079] Among them, image feature extraction is to separate the main parts of the clothing (such as neckline, cuffs, body) by analyzing the geometric characteristics of the clothing image (such as contour, edge, texture, etc.), which serves as the basis for realizing clothing modifications. By accurately identifying each part, it is convenient for subsequent local modifications.

[0080] In an embodiment of the present application, an edge detection algorithm (such as Canny edge detection) can be used to extract the contour information of the image, and a region growing algorithm can be used to segment different component areas of the clothing. For example, the body of the clothing can be separated from the cuffs and collar according to the symmetry or texture distribution of the clothing. After the components are identified, the region annotations are generated through vectorization to record the boundaries, area, color and other characteristics of each region.

[0081] For example, in the uploaded dress image, the system will identify the "neckline area", "cuff area" and "hem area", and generate a labeling file to record the area information of each part.

[0082] Step S103, semantically analyzing the modification requirement text to obtain the component area to be modified, modification requirement parameters and modification style labels;

[0083] The core of semantic analysis is to understand the user's intention of changing the model and convert it into machine-operable instructions. This includes identifying the specific parts that need to be changed, specific change parameters (such as color, pattern) and style orientation (such as retro, modern).

[0084] Exemplarily, the user's text description is matched with a predefined demand parameter library. For example, if the user inputs "change the cuffs to lace style", the system will identify "cuffs" as the modification target and "lace style" as the design parameter; if the user inputs "change the color to pink and add floral pattern", the system will parse "color = pink" and "pattern = floral pattern".

[0085] Step S104, segmenting the clothing style image according to each component region, and marking the component region to be modified based on the modification requirement parameter, to obtain the clothing style image after modification marking;

[0086] Among them, on the basis of image segmentation, the area of ​​the component to be modified is further highlighted according to the needs. The modification marking is to clearly map the user's modification requirement parameters to the target area, so as to facilitate the precise execution of the subsequent modification design.

[0087] Specifically, combined with the segmentation results, corresponding to the names of the modified parts extracted, a label map of the target area is generated, which records information such as the area boundary, original color, texture, pattern, etc., and adds the user's modification requirement parameters. For example, the "cuff area" is labeled, and its boundaries and modification parameters (such as "the pattern is lace pattern") are recorded.

[0088] Step S105, determining corresponding facelift design elements based on the design element database according to the facelift requirement parameters and the facelift style labels;

[0089] Among them, the design element database contains a large number of materials such as colors, patterns, component styles, etc. By matching design elements with modification requirement parameters and modification style labels, the system can quickly find the candidate materials that are closest to the requirements.

[0090] Specifically, elements in the design element database are called according to the facelift design parameters, and material matching is achieved through feature similarity calculation. For example, the color space (such as the Lab color model) is used to calculate the matching degree between the facelift target color and the database material color, and the style label is used to filter the results. Only materials that meet the target style are retained for users to choose, and the final facelift design elements are determined.

[0091] Step S106, optimizing the modified design elements according to the original design elements of the component area to be modified to obtain optimized design elements;

[0092] Among them, by analyzing the original design characteristics of the target area (such as color and texture), the matched design elements are adjusted to make them visually consistent with the original image.

[0093] Specifically, the edge features can be used to adjust the size, direction and position of the design elements, and perform color and texture blending algorithms, such as generating a seamless effect with the original background based on color gradients. For example, the optimized lace pattern size is adjusted to fit the cuff boundary to ensure the natural integration of the design material and the original image, avoiding abrupt or conflicting visual effects.

[0094] Step S107, based on the clothing style image after the modification mark, the original design elements of the component area to be modified are replaced with the optimized design elements to obtain the modified component area;

[0095] Among them, the optimized design elements are actually applied to the target area to complete the partial facelift design. Specifically, the image replacement technology can be used to superimpose the optimized design elements to the target area, repair the boundaries and adjust the transparency so that the new elements are completely integrated with the original design. For example, the lace pattern is superimposed on the cuff area to replace the original design, realizing the specific implementation of the facelift requirements at the image level.

[0096] Step S108, performing visual optimization processing on the modified component area based on a boundary repair algorithm to obtain a clothing style image after partial modification;

[0097] Among them, the boundaries and textures of the newly designed area are smoothly transitioned to ensure the naturalness and coordination of the local modification. Specifically, the boundary repair algorithm (such as Poisson fusion) can be used to optimize the boundaries of the local area, and the texture smoothing algorithm can be executed to ensure the texture continuity between the new and old areas. For example, by optimizing the boundary between the lace pattern and the main area of ​​the cuff, the transition is natural, the visual quality of the modification effect is improved, and hard edges or color conflicts are avoided.

[0098] Step S109, performing global style optimization on the clothing style image after partial modification based on the modified style label to obtain the clothing style image after the modification.

[0099] Among them, after partial modification, the global design characteristics need to be adjusted according to the style label to make the overall style conform to the target style. Specifically, by performing optimization algorithms (such as style mapping) on ​​the global color distribution, pattern repetition rate, etc., it is ensured that the modified local area is consistent with the overall design in terms of vision and style.

[0100] For example, after modifying the partial cuffs to lace, the global color tone is adjusted to warm colors in a retro style to achieve the coordination and unity of the local and overall styles.

[0101] Through the above implementation, accurate identification of each component of clothing and accurate mapping of modification requirements are achieved, ensuring a high degree of automation in the design process. By converting the modification requirements of users into efficient, coordinated and personalized clothing design results, and using the matching, optimization and boundary repair algorithms of design elements, the visual coordination and natural integration of local design and overall style are ensured. Finally, global style optimization further ensures that the overall design meets the target style expected by the user, meets the user's needs for efficient, high-quality and highly personalized clothing modification design, and improves the design completion and market adaptability.

[0102] Reference Figure 2 As an implementation of step S105, based on the design element database, the step of determining the corresponding facelift design element according to the facelift requirement parameter and the facelift style label includes:

[0103] Step S201, loading the design element database;

[0104] The design element database is the core resource of the modification process, which contains basic design element information such as color, pattern and component style. The database loading process is designed to ensure that the modification requirements can be accessed and matched with high-quality design materials in real time. The database needs to support multi-dimensional query capabilities, including color coding, texture features, shape descriptions and style tags.

[0105] For example, the design element database can be loaded from a storage medium (such as a cloud or a local server) and structured into three main categories: color, pattern, and component style. Each category contains multiple fields, such as a color table containing the RGB value of the color, Lab model data, name, etc., a pattern table including the texture features of the vector pattern, classification labels (such as geometry, plants, abstract) and style descriptions, and a component style table including the geometric outline of the component, functional description (such as cuff type: round sleeve, trumpet sleeve) and style label.

[0106] It is understandable that in order to improve the efficiency of subsequent queries, field indexes are established, such as indexes based on color similarity and directional histograms of texture features, to support fast retrieval. In addition, style tags are important limiting conditions for modification requirements. Each database element is associated with a style tag, providing direct style screening capabilities.

[0107] Step S202, matching corresponding design elements in a design element database according to the model modification requirement parameters, and obtaining design element matching results and corresponding similarity scores;

[0108] Among them, the matching of design elements is achieved by analyzing the parameters of the facelift requirements, and searching for qualified materials in the database one by one. Each requirement parameter (such as color, pattern or component style) corresponds to a different matching logic, and the fit between the material and the requirements is judged based on the similarity score.

[0109] Exemplarily, the modification requirement parameters are classified, for example, "color = pink" is mapped to the color table, and "cuff = lace pattern" is mapped to the pattern table; and a matching rule is generated for each requirement parameter. For example, color matching uses the color difference calculation of the Lab color space, pattern matching uses the texture feature histogram similarity, and component style matching calculates the style contour similarity based on the geometric shape descriptor (such as the shape context algorithm).

[0110] It can be understood that each matched design element generates a corresponding similarity score and records its corresponding database ID and feature information.

[0111] Step S203, sorting the design element matching results according to the similarity scores, and filtering based on the facelift style tags to obtain a sorted and filtered design element list;

[0112] Among them, sorting and filtering the matching results are key steps to improve design efficiency and accuracy. The similarity score is used to measure the degree of fit between the design elements and the modification requirements, while the style label filtering ensures that the results meet the user's overall design requirements.

[0113] Specifically, the generated matching results are sorted in descending order by similarity score, with the material with the highest similarity being given priority. If multiple materials have the same similarity score, they are further sorted according to priority rules (such as color takes precedence over pattern, and pattern takes precedence over component style). In addition, during the style tag screening process, if a style tag (such as "retro") is specified in the modification requirement, only the materials associated with the style tag are retained. It should be noted that for requirements without specified style tags, all sorting results can be retained.

[0114] Step S204, sending the sorted and filtered design element list to the user terminal for selection;

[0115] Among them, the sorted and screened design materials will be displayed in a user-visual form to ensure the user's active participation and satisfaction with the design change. Users can directly select the design elements that best meet their expectations from the candidate materials.

[0116] Step S205, receiving and responding to the design element selection instruction sent by the user terminal, and determining the corresponding facelift design element.

[0117] Among them, the user's selection result is the key input for determining the final facelift design elements. The system updates the final design plan based on the user's selection and uses it for subsequent facelift modules.

[0118] Specifically, the complete material data is extracted from the filter list according to the design element ID selected by the user, and the final design element file is generated, including all necessary parameters (such as color coding, pattern file path, style outline, etc.).

[0119] In the above implementation, accurate matching and determination of modification requirements to design elements are achieved, and the fit between design materials and modification requirements is improved through multi-dimensional feature matching and style screening; through user terminal interaction, the personalization of design results and user satisfaction are enhanced; at the same time, this technical solution has significant advantages in efficiency, flexibility and accuracy, and is suitable for application scenarios of personalized clothing design and intelligent modification.

[0120] Reference Figure 3 As an implementation of step S202, according to the model modification requirement parameters, the corresponding design elements are matched in the design element database to obtain the design element matching results and the corresponding similarity scores, including:

[0121] Step S301, classifying the model change demand parameters and generating corresponding matching rules for each demand parameter category;

[0122] Among them, the modification requirements usually include multiple design elements (such as color, pattern, and component style). Through classification, the nature of each requirement parameter can be clarified, and matching rules can be generated for each category. The matching rules include color matching rules, pattern matching rules, and component style matching rules to ensure the pertinence and accuracy of subsequent database retrieval.

[0123] In some embodiments, the color matching rule is: using the Lab color space to calculate the color difference, based on the Lab value of the target color, and comparing it with the Lab value of each color in the database; the pattern matching rule is: extracting the texture feature histogram of the target pattern, and calculating the similarity of the pattern texture histogram during matching; the component style matching rule is: comparing the target component with the style outline in the database based on a geometric shape descriptor (such as a shape context).

[0124] Step S302, searching for corresponding candidate design elements in a design element database according to the model change requirement parameters to obtain a list of candidate design elements;

[0125] Among them, the modification requirement parameters and matching rules provide the conditions for retrieving design elements. By searching the qualified materials in the database one by one, a list of candidate design elements can be generated, providing a basis for subsequent similarity calculations.

[0126] Step S303, calculating the similarity score between each candidate design element in the candidate design element list and the model modification requirement parameter based on the matching rule;

[0127] Specifically, the degree of fit between each material in the candidate design element list and the model modification requirement parameters is quantitatively expressed by a similarity score. The requirement parameters of different categories use corresponding matching rules to calculate the similarity, generate a similarity score for each candidate material, and record the material ID and feature information. The similarity score includes color similarity score, pattern similarity score and component style similarity score.

[0128] Step S304: screening candidate design elements with similarity scores greater than a preset threshold to obtain design element matching results and corresponding similarity scores.

[0129] Among them, the similarity score is the core indicator to measure the degree of fit between candidate design elements and modification requirements. By screening candidate design elements with scores greater than the threshold, materials that do not meet the requirements can be excluded and the accuracy of the design results can be improved.

[0130] Exemplarily, different categories of similarity score thresholds can be defined, for example, the color similarity threshold is set to 0.85, the pattern similarity threshold is set to 0.8, and the component style similarity threshold is set to 0.75; the similarity scores of the candidate design elements are determined one by one to see whether they are greater than the thresholds, and the candidate design elements that do not meet the conditions are eliminated.

[0131] It is understandable that through the screening process, the final design element matching results are ensured to be highly consistent with the needs, which significantly improves the accuracy of the facelift design and user satisfaction.

[0132] In the above implementation, the modification requirements are accurately matched with the materials in the design element database, the requirements are clarified through classification, the similarity is accurately calculated based on multiple algorithms, and the design elements with the highest degree of fit are screened out, thereby ensuring that the design results not only meet the user's demand parameters, but also have high precision and reliability, thereby significantly improving the efficiency and accuracy of clothing modification design.

[0133] Reference Figure 4 As an implementation method of step S106, the step of optimizing the modified design elements according to the original design elements of the component area to be modified to obtain the optimized design elements includes:

[0134] Step S401, extracting characteristics based on the original design elements of the component area to be modified to obtain original design characteristics;

[0135] Among them, the original design characteristics of the component area to be modified are the basis for optimizing the modified design elements. By extracting characteristics such as color, texture, and geometric boundaries, the system can accurately describe the visual attributes and structural information of the original area, providing data basis for subsequent optimization.

[0136] Specifically, the color characteristics include primary hue, secondary hue and color distribution ratio, the texture characteristics include texture direction and texture type, and the geometric boundary characteristics include size and shape characteristics.

[0137] Step S402, analyzing the characteristics of the facelift design elements to obtain facelift design characteristics;

[0138] Among them, the characteristics of the modified design elements describe the properties of the material to be applied. Through the analysis of color, texture and geometric characteristics, the difference between the material and the original area can be quantified to provide input for optimization.

[0139] Step S403, calculating the characteristic difference between the original design characteristic and the modified design characteristic to obtain a characteristic difference analysis result;

[0140] Among them, the characteristic difference quantifies the fit between the original area and the design elements in terms of color, texture and geometric characteristics, and the difference results provide optimization direction for subsequent dynamic adjustments.

[0141] Specifically, color difference calculation can calculate the color difference between the original main color and the modified main color through the Lab color space, texture difference calculation can calculate the similarity between the original texture and the modified texture through the texture histogram cross entropy, and geometric difference calculation can calculate the size ratio difference between the original area and the modified design elements.

[0142] Step S404 , dynamically adjusting the facelift design characteristics of the facelift design elements based on the characteristic difference analysis results to obtain optimized design elements.

[0143] The design elements of the renovation were optimized in terms of color, texture and geometry to blend seamlessly with the original area both visually and structurally.

[0144] In the above implementation, precise optimization and natural integration of the facelift design elements and the original area in terms of color, texture and geometric shape are achieved, which improves the visual coordination, geometric adaptability and overall aesthetics of the design elements and the original area. At the same time, the efficiency and applicability of the facelift design are significantly improved through the optimization process, meeting the needs of personalized and high-precision design.

[0145] Reference Figure 5 As an implementation method of step S108, the step of performing visual optimization processing on the modified component area based on the boundary repair algorithm to obtain the clothing style image after partial modification includes:

[0146] Step S501, based on the edge detection algorithm, extract the boundary curve of the modified component area and determine the corresponding boundary area;

[0147] Among them, the boundary curve is the dividing line between the new design area and the original design area, and its accurate extraction is the basis for subsequent color and texture optimization.

[0148] Specifically, the Canny edge detection algorithm can be used to identify edge pixels in the design image and generate a binary edge map. According to the pixel position of the edge map, closed or continuous boundary curve data is extracted. With the boundary curve as the center, a boundary area (such as an area with a width of 10 pixels) is generated for subsequent feature analysis and repair.

[0149] Step S502, determining the original component area and the modified component area according to the areas on both sides of the boundary area;

[0150] Among them, the areas on both sides of the boundary correspond to the original design elements and the new design elements respectively. Accurately distinguishing the areas on both sides can provide a boundary reference for color and texture feature extraction.

[0151] Step S503, extracting the differences in color and texture characteristics between the original component region and the modified component region;

[0152] Among them, the differences in color and texture characteristics between the original area and the modified area are the core basis for optimization. By quantitatively analyzing the color distribution and texture characteristics of the two areas, a clear repair target can be provided.

[0153] Step S504, performing repair processing on the boundary area based on the difference in color and texture characteristics to obtain an optimized boundary area;

[0154] Among them, the boundary area is repaired by color gradient and texture fusion technology to make the transition between the new and old areas natural. The boundary repair algorithm (such as Poisson fusion) can be used to achieve high-quality connection in color and texture.

[0155] In some embodiments, the specific repair process includes: generating a gradient from the main color of the original area to the main color of the modified area in the boundary area, using a texture weighted blending algorithm to generate a texture transition, and repairing the boundary area based on a Poisson fusion algorithm to ensure pixel continuity and smoothness of the transition boundary.

[0156] Step S505, replacing the optimized boundary area into the clothing style image to obtain the clothing style image after partial modification.

[0157] Among them, the optimized boundary area is seamlessly integrated into the complete image to generate a visually natural and high-quality clothing style image; specifically, the pixel overlay method can be used to replace the optimized boundary area with the corresponding position of the original image to ensure that the boundary area is aligned with the new and old areas without offset or breakage.

[0158] In the above implementation, high-quality optimization of the image of the partially modified clothing style is achieved. By enhancing the continuity and smoothness of the boundaries, the natural transition between the new and old areas is achieved by using color gradient and texture fusion technology. At the same time, global adjustments are made to ensure the visual consistency between the partially modified area and the overall design, thereby significantly improving the overall beauty and design completion of the clothing style.

[0159] Reference Figure 6 As an implementation method of step S109, the step of performing global style optimization on the clothing style image after partial modification based on the modified style label to obtain the clothing style image after the modification includes:

[0160] Step S601, parsing the facelift style label and mapping it to obtain the target style characteristics;

[0161] Among them, the style label of the modification (such as "retro", "simple", "modern") is usually an abstract description of the overall design goal of the clothing. By parsing the style label, it is mapped into quantifiable target style characteristics, such as color scheme, texture characteristics and pattern style, thus providing a clear guidance direction for subsequent optimization.

[0162] For example, if the style label is retro, the corresponding design features include: warm colors (brown, orange), soft textures, and symmetrical patterns; if the style label is modern, the corresponding design features include: cool colors (gray, blue), high-contrast textures, and geometric patterns.

[0163] Step S602, extracting global characteristics based on the clothing style image after partial modification to obtain current global characteristics;

[0164] Among them, the current global characteristics are a quantitative description of the overall design features of the clothing image after partial modification, mainly including color distribution, pattern type and texture characteristics. Extracting the current global characteristics is the basis for calculating style differences and performing optimization.

[0165] Step S603, calculating the style characteristic difference between the target style characteristic and the current global characteristic;

[0166] Among them, the difference between the target style characteristics and the current global characteristics reflects the degree of deviation of the current image from the ideal style. By quantifying these differences, the optimization direction and specific adjustment goals can be clarified.

[0167] Step S604, optimizing and adjusting the current global characteristics based on the style characteristic differences to obtain a clothing style image after the modification is completed.

[0168] Exemplarily, the optimization adjustment of the current global characteristics specifically includes global color optimization, global pattern optimization and global texture optimization. Among them, global color optimization is to use a color mapping algorithm to gradually adjust the current main color to the target main color, and use a gradient algorithm to adjust the distribution ratio of the global secondary color to make the overall color closer to the target style; global pattern optimization is to replace the pattern type that does not conform to the target style, and adjust the pattern repetition rate to the target value (such as from 0.5 to 0.3); global texture optimization is to use a texture smoothing algorithm (such as Gaussian blur) to optimize the texture smoothness, and adjust the texture directionality to make it consistent with the target style.

[0169] It can be understood that through the optimization and adjustment of global colors, patterns and textures, the overall image after partial facelift can be made to conform to the target style characteristics, thereby achieving the coordination and unity of the local and overall design styles.

[0170] In the above implementation, the comprehensive unification of the clothing style image after partial modification and the target style characteristics is accurately achieved. Through the optimization of global color, pattern and texture, the local and overall designs are visually highly coordinated, which significantly enhances the overall beauty and design completion of the clothing style, while improving the intelligence and efficiency of the modification design.

[0171] The embodiment of the present application also discloses an intelligent clothing modification design system.

[0172] An intelligent clothing modification design system, the system comprising:

[0173] A receiving module, used to receive clothing style images and modification requirement texts uploaded by a user terminal;

[0174] A component region extraction module is used to extract features from clothing style images and identify various component regions in the clothing style images;

[0175] The semantic analysis module is used to perform semantic analysis on the modification requirement text to obtain the component area to be modified, modification requirement parameters and modification style labels;

[0176] A style change marking module is used to segment the clothing style image according to each component area, and mark the component area to be changed based on the style change requirement parameters to obtain the clothing style image after the style change marking;

[0177] A facelift design element determination module, used to determine corresponding facelift design elements based on a design element database, facelift requirement parameters and facelift style labels;

[0178] A design element optimization module, used to optimize the modified design elements according to the original design elements of the component area to be modified, so as to obtain the optimized design elements;

[0179] A design element replacement module is used to replace the original design elements of the component area to be modified with the optimized design elements based on the clothing style image after the modification mark, so as to obtain the modified component area;

[0180] The boundary repair module is used to perform visual optimization processing on the modified component area based on the boundary repair algorithm to obtain the clothing style image after partial modification;

[0181] The global optimization module performs global style optimization on the clothing style images after partial modification based on the modification style labels to obtain the clothing style images after the modification.

[0182] In actual application, the technical solution of this application can significantly improve the efficiency and quality of clothing modification design, reduce the dependence of traditional design on human experience, enable clothing designers to focus more on creative expression, and at the same time meet the market demand for large-scale customized design, and promote the development of the clothing industry towards intelligence and digitalization.

[0183] An intelligent clothing modification design system in an embodiment of the present application can implement any of the above-mentioned intelligent clothing modification design methods, and the specific working process of each module in an intelligent clothing modification design system can refer to the corresponding process in the above-mentioned method embodiment.

[0184] In the several embodiments provided in this application, it should be understood that the provided methods and systems can be implemented in other ways. For example, the system embodiments described above are only illustrative; for example, the division of a certain module is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed.

[0185] The embodiment of the present application also discloses a computer device.

[0186] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, an intelligent clothing modification design method as described above is implemented.

[0187] The embodiment of the present application also discloses a computer-readable storage medium.

[0188] A computer-readable storage medium stores a computer program that can be loaded by a processor and execute any one of the above-mentioned intelligent clothing modification design methods.

[0189] Among them, computer-readable storage media can be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus or device; the program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0190] It should be noted that in the above embodiments, the description of each embodiment has different emphases, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0191] The above are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any feature disclosed in this specification (including the abstract and drawings), unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example in a series of equivalent or similar features.

Claims

1. A smart clothing design method, characterized in that: The method comprises: Receive clothing style images and modification requirement text uploaded by user terminals; Extracting features from the clothing style image to identify various component areas in the clothing style image; Performing semantic analysis on the modification requirement text to obtain the component area to be modified, modification requirement parameters and modification style labels; Segmenting the clothing style image according to each of the component regions, and marking the component regions to be modified based on the modification requirement parameters, to obtain the clothing style image after modification marking; Based on the design element database, determining corresponding facelift design elements according to the facelift requirement parameters and the facelift style labels; According to the original design elements of the component area to be modified, the modified design elements are optimized to obtain optimized design elements; Based on the clothing style image after the modification mark, the original design elements of the component area to be modified are replaced with the optimized design elements to obtain the modified component area; Performing visual optimization processing on the modified component area based on a boundary repair algorithm to obtain a clothing style image after partial modification; Based on the modified style label, a global style optimization is performed on the clothing style image after the partial modification to obtain the clothing style image after the modification.

2. The method for designing a smart clothing modification according to claim 1, characterized in that: Based on the design element database, the step of determining the corresponding facelift design element according to the facelift requirement parameter and the facelift style label includes: Loading a design element database; the design element database includes colors, patterns and component styles; According to the model modification requirement parameters, matching corresponding design elements in the design element database to obtain design element matching results and corresponding similarity scores; Sorting the design element matching results according to the similarity scores, and filtering based on the facelift style labels to obtain a sorted and filtered design element list; Sending the sorted and screened design element list to the user terminal for selection; Receive and respond to the design element selection instruction sent by the user terminal, and determine the corresponding modified design element.

3. The method for designing a smart clothing modification according to claim 2, characterized in that: According to the model modification requirement parameters, the steps of matching corresponding design elements in the design element database to obtain design element matching results and corresponding similarity scores include: Classifying the model change demand parameters and generating corresponding matching rules for each demand parameter category; Retrieving corresponding candidate design elements in the design element database according to the model modification requirement parameters to obtain a list of candidate design elements; Based on the matching rule, calculating a similarity score between each candidate design element in the candidate design element list and the model modification requirement parameter; The candidate design elements whose similarity scores are greater than a preset threshold are screened to obtain design element matching results and corresponding similarity scores.

4. The method for designing a smart clothing modification according to claim 3, characterized in that: The matching rules include color matching rules, pattern matching rules and component style matching rules, and the similarity scores include color similarity scores, pattern similarity scores and component style similarity scores.

5. A smart clothing design method according to any one of claims 1 to 4, characterized in that: According to the original design elements of the component area to be modified, the step of optimizing the modified design elements to obtain the optimized design elements comprises: Extracting characteristics according to the original design elements of the component area to be modified to obtain original design characteristics; Performing characteristic analysis on the facelift design elements to obtain facelift design characteristics; Calculating the characteristic difference between the original design characteristic and the modified design characteristic to obtain a characteristic difference analysis result; Based on the characteristic difference analysis result, the facelift design characteristics of the facelift design element are dynamically adjusted to obtain an optimized design element.

6. The method for designing a smart clothing modification according to claim 5, characterized in that: The step of performing visual optimization processing on the modified component area based on a boundary repair algorithm to obtain a clothing style image after partial modification includes: Based on an edge detection algorithm, a boundary curve of the modified component region is extracted to determine a corresponding boundary region; Determining the original component area and the modified component area according to the areas on both sides of the boundary area; extracting differences in color and texture characteristics between the original component region and the modified component region; Performing repair processing on the boundary area based on the color and texture characteristic differences to obtain an optimized boundary area; The optimized boundary area is replaced into the clothing style image to obtain a clothing style image after partial modification.

7. The method for designing a smart clothing modification according to claim 6, characterized in that: The step of performing global style optimization on the clothing style image after partial modification based on the modified style label to obtain the clothing style image after the modification includes: Parsing the facelift style label and mapping it to obtain target style characteristics; Extracting global features based on the clothing style image after partial modification to obtain current global features; Calculating a style characteristic difference between the target style characteristic and the current global characteristic; The current global characteristics are optimized and adjusted based on the style characteristic differences to obtain a clothing style image after the modification.

8. An intelligent clothing modification design system, characterized in that: The system comprises: A receiving module, used to receive clothing style images and modification requirement texts uploaded by a user terminal; A component region extraction module is used to extract features from the clothing style image and identify each component region in the clothing style image; A semantic analysis module, used to perform semantic analysis on the modification requirement text to obtain the component area to be modified, modification requirement parameters and modification style labels; A style change marking module, used for performing region segmentation on the clothing style image according to each of the component regions, and performing style change marking on the component regions to be changed based on the style change requirement parameters, to obtain a clothing style image after the style change marking; A facelift design element determination module, configured to determine corresponding facelift design elements based on a design element database and according to the facelift requirement parameters and facelift style labels; A design element optimization module, used to optimize the modified design elements according to the original design elements of the component area to be modified, so as to obtain optimized design elements; A design element replacement module, for replacing the original design elements of the component region to be modified with the optimized design elements based on the modified clothing style image, so as to obtain the modified component region; A boundary repair module, used for performing visual optimization processing on the modified component area based on a boundary repair algorithm to obtain a clothing style image after partial modification; The global optimization module performs global style optimization on the clothing style image after partial modification based on the modification style label to obtain the clothing style image after the modification.

9. A computer device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 7 when executing the program.

10. A computer-readable storage medium, characterized in that: A computer program is stored which can be loaded by a processor and execute the method according to any one of claims 1 to 7.

Images