Paper-cut pattern automatic generation system based on artificial intelligence

Abstract

The invention relates to the technical field of pattern generation, in particular to an artificial intelligence-based paper-cut pattern automatic generation system, which comprises a geometric feature analysis module, a texture dynamic adjustment module, a paper-cut style analysis module, a user interaction adjustment module and a paper-cut pattern optimization module, the geometric feature analysis module collects paper-cut pattern boundary curve data and positions a paper-cut area by using internal partition coordinates. According to the method, by automatically collecting and analyzing the boundary curve data of the pattern and recognizing the local contour gradient, the basic structure of the pattern is precisely processed, the design precision is improved, automatic adjustment of the texture is automatically optimized according to geometric features, harmony and consistency of the texture and the pattern are ensured, the attractiveness of the pattern is improved, and the design efficiency is improved. User interaction data is intelligently analyzed to allow the system to perform real-time adjustment according to user feedback, so that the design process is more flexible, the response speed is higher, and the efficiency and adaptability of paper-cut pattern design are improved through automatic and intelligent processing on the whole.

Description

Technical Field

[0001] The present invention relates to the technical field of pattern generation, and in particular to an automatic paper-cutting pattern generation system based on artificial intelligence. Background Art

[0002] Pattern generation technology mainly involves the use of software tools and algorithms to automatically or semi-automatically create visual patterns. This technology is widely used in graphic design, animation, game development, cultural and artistic creation and other fields. With the help of artificial intelligence, pattern generation technology has achieved a leap from simple geometric graphics to complex artistic style imitation. Through AI algorithms, it can learn specific artistic styles and create novel pattern designs, greatly expanding the boundaries of creative expression.

[0003] Among them, the artificial intelligence paper-cutting pattern automatic generation system refers to a system that uses artificial intelligence technology to automatically create paper-cutting art patterns. Paper-cutting is a traditional art form that creates decorative or symbolic patterns by cutting paper. The system uses AI technology to automatically analyze and learn the styles, patterns and construction methods of traditional paper-cutting, and then generate new paper-cutting designs. It can be used as an auxiliary tool for artistic creation, saving time and energy. The application of the system can also be extended to commercial fields such as advertising, product packaging and fashion design.

[0004] Existing technologies have limited capabilities in automatically identifying complex textures and fine patterns, resulting in automatically generated paper-cut patterns lacking in artistry and precision. This limitation is mainly due to the inability of the algorithm to fully reproduce the craftsmen's ability to handle details. Existing technologies are inflexible in processing user feedback, making it difficult to adjust patterns in real time according to the user's specific needs, limiting the potential for personalized design. Existing systems use relatively fixed algorithms for texture adaptability, which means that when faced with designs with complex geometric features, the system's adaptability is insufficient, resulting in the inability of automatically generated paper-cut patterns to achieve professional levels of sophistication and beauty. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an automatic paper-cutting pattern generation system based on artificial intelligence.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: an automatic generation system of paper-cutting patterns based on artificial intelligence, the system comprising: The geometric feature analysis module collects the boundary curve data of the paper-cut pattern, locates the paper-cut area using the internal partition coordinates, identifies the local contour gradient by analyzing the pattern data, compares the curve with the gradient data, analyzes the gradient direction and value distribution, screens the curvature mutation area, and obtains the local curvature characteristic index; The texture dynamic adjustment module determines the texture density change trend based on the local curvature characteristic index and the basic paper-cut texture, analyzes the corresponding relationship between the texture parameters and the curvature data, and adjusts the texture density and direction along the local curvature gradient distribution to obtain the texture adaptability parameter; The paper-cutting style analysis module calls the texture adaptability parameter, analyzes the fluctuation trend of the paper-cutting style in different time periods, selects key style feature values, evaluates their adaptability in the current trend, and compares the offset direction of the style data to obtain the style trend analysis result; Based on the style trend analysis results, the user interaction adjustment module counts the range of changes in user operation parameters, evaluates the fluctuation range of user zoom and rotation parameters, analyzes the style deviation degree of color selection, screens operation parameters that match the current trend, and obtains user preference adaptation values.

[0007] The improvements of the present invention are that the local curvature characteristic indicators include curvature extremes, gradient peaks, and regional distribution; the texture adaptability parameters include texture intensity index and directional consistency index; the style trend analysis results include trend adaptation scores and style evolution vectors; and the user preference adaptation values ​​include user operation sensitivity and style consistency ratings.

[0008] The present invention is improved in that the geometric feature analysis module comprises: The boundary curve acquisition submodule acquires the boundary curve data of the paper-cut pattern, analyzes the paper-cut area, obtains the partition coordinates of each area, establishes a coordinate index, and locates the paper-cut area according to the partition coordinates to generate the boundary data of the paper-cut area; The local contour gradient calculation submodule is based on the paper-cutting area boundary data, analyzes the pattern data, collects the gradient information of the boundary point set, and adopts the formula: ; Calculate the local contour gradient of each point to obtain the local gradient distribution characteristics, where: Representative The local contour gradient of the point, and Respectively represent The adjacent point Pointed and The coordinate difference of the direction, For the The sum of the points in the point's neighborhood; The curvature mutation screening submodule analyzes the gradient direction and value distribution according to the local gradient distribution characteristics, compares the curve with the gradient data, screens the curvature mutation area, and obtains the local curvature characteristic index.

[0009] The present invention is improved in that the texture dynamic adjustment module comprises: The texture density calculation submodule is based on the local curvature characteristic index and the basic texture of paper cutting to determine the texture density change trend, using the formula: ; Calculate the texture density change rate of each area and establish the texture density change parameters ,in, Representative The curvature value of the region, Representative The curvature value of the region, Representative The curvature gradient of a region, represents the total number of regions; The texture direction adjustment submodule calls the texture density variation parameter, analyzes the corresponding relationship between the texture parameter and the curvature data, and adjusts the texture density and texture direction along the local curvature gradient distribution to obtain the texture adaptability parameter.

[0010] The present invention is improved in that the paper-cutting style analysis module comprises: The style data collection submodule calls the texture adaptability parameter, collects paper-cut design style data, collects paper-cut style information in different time periods, and classifies the data according to the time interval, counts the variation range of the paper-cut style parameters, and obtains paper-cut style time series data; The key style screening submodule calls the paper-cut style time series data, compares the style parameter changes in different time periods, and uses the formula: ; Calculating Style Volatility , and filter the key style feature values, where Representative The change range of style parameters in each time period, Representative The duration of a time period, represents the total number of time periods; The style trend evaluation submodule calculates the adaptability of the key style feature value in the current style trend according to the key style feature value, compares the offset direction of the style data, analyzes the evolution trend over time, and obtains the style trend analysis result.

[0011] The present invention is improved in that the user interaction adjustment module comprises: The operation parameter statistics submodule analyzes the user interaction data based on the style trend analysis result, counts the user's zoom, rotation and color selection operations, and identifies the variation range of the operation parameters to obtain the user operation variation range; The interaction parameter analysis submodule analyzes the fluctuation range of the scaling and rotation parameters based on the user operation variation range, evaluates the deviation degree of the color selection, and obtains the user interaction fluctuation range; The adaptation value calculation submodule analyzes the volatility of parameters and the deviation trend of color selection according to the user interaction fluctuation interval, and adopts the formula: ; Get user preference adaptation value ,in, Representative Secondary scaling parameter value, represents the mean value of the scaling parameter, represents the standard deviation of the scaling parameter, Representative Secondary rotation parameter value, represents the mean value of the rotation parameter, represents the standard deviation of the rotation parameter, Representative Secondary color selection value, Represents the color reference value under the current style trend, represents the weight coefficient of color selection deviation, Represents the total number of operations.

[0012] The present invention is improved in that the system further comprises: The paper-cut pattern optimization module analyzes the pattern morphology data according to the user preference adaptation value, evaluates the edge curvature change of the pattern contour data, determines the symmetry offset of the morphology data, screens the morphology with an offset exceeding the normal range, and then intelligently adjusts and corrects the contour with abnormal curvature change to generate a paper-cut design file; The paper-cut design file includes design accuracy, detail completeness, and visual effect scores.

[0013] The present invention is improved in that the paper-cut pattern optimization module comprises: The pattern morphology analysis submodule analyzes the morphology data of the paper-cut pattern according to the user preference adaptation value, extracts the edge curvature of the contour, calculates the curvature variation interval, and determines the reference value of the curvature variation to obtain the pattern contour curvature data; The symmetry offset detection submodule evaluates the symmetry of the pattern according to the pattern contour curvature data, identifies the offset area exceeding the normal range by measuring the morphological offset on both sides of the pattern symmetry axis, and generates the symmetry offset; The contour curvature correction submodule analyzes the contour area with abnormal curvature change according to the symmetry offset, adjusts and corrects the curvature of the area, and adopts the formula: ; Get the corrected contour curvature , optimize the pattern outline and generate paper-cut design files, where Representative The original curvature of the contour points, represents the mean value of the contour curvature, represents the standard deviation of the curvature, Representative The offset of the points, represents the standard deviation of the offset, Represents the total number of contour points.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by automatically collecting and analyzing the boundary curve data of the pattern and identifying the local contour gradient, the basic structure of the pattern is accurately processed, thereby improving the design accuracy. The automatic adjustment of the texture is automatically optimized according to the geometric features to ensure the harmony between the texture and the graphics, thereby improving the aesthetics of the pattern. Intelligent analysis of user interaction data allows the system to make real-time adjustments based on user feedback, making the design process more flexible and responsive. On the whole, the efficiency and adaptability of paper-cut pattern design are improved through automated and intelligent processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a system flow chart of the present invention; Figure 2 It is a flow chart of the geometric feature analysis module in the present invention; Figure 3 This is a flow chart of the texture dynamic adjustment module in the present invention; Figure 4 This is a flow chart of the paper-cutting style analysis module in the present invention; Figure 5 A flowchart of a user interaction adjustment module in the present invention; Figure 6 It is a flow chart of the paper-cut pattern optimization module in the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0018] Example: See also Figure 1 The present invention provides a technical solution: an automatic generation system of paper-cutting patterns based on artificial intelligence comprises: The geometric feature analysis module collects the boundary curve data of the paper-cut pattern, locates the paper-cut area using the internal partition coordinates, identifies the local contour gradient by analyzing the pattern data, compares the curve with the gradient data, analyzes the gradient direction and value distribution, screens the curvature mutation area, and obtains the local curvature characteristic index; The texture dynamic adjustment module is based on the local curvature characteristic index and the basic texture of paper cutting to determine the trend of texture density change, analyze the correspondence between texture parameters and curvature data, and adjust the texture density and direction along the local curvature gradient distribution to obtain texture adaptability parameters; The paper-cut style analysis module calls texture adaptability parameters, collects paper-cut design style data, analyzes the fluctuation trend of paper-cut styles in different time periods, selects key style feature values, evaluates their adaptability in the current trend, and compares the offset direction of style data to obtain style trend analysis results; The user interaction adjustment module analyzes user interaction data based on the style trend analysis results, counts the range of changes in user operation parameters, evaluates the fluctuation range of user zoom and rotation parameters, and analyzes the style deviation of color selection, screens the operation parameters that match the current trend, and obtains the user preference adaptation value; The paper-cut pattern optimization module analyzes the pattern morphological data according to the user's preferred adaptation value, evaluates the edge curvature changes of the pattern contour data, and determines the symmetry offset of the morphological data. It screens out the shapes whose offsets exceed the normal range, and then intelligently adjusts and corrects the contours with abnormal curvature changes to generate paper-cut design files.

[0019] Local curvature characteristic indicators include curvature extreme value, gradient peak, and regional distribution. Texture adaptability parameters include texture intensity index and directional consistency index. Style trend analysis results include trend adaptation score and style evolution vector. User preference adaptation values ​​include user operation sensitivity and style consistency rating. Paper-cutting design files include design accuracy, detail completeness, and visual effect score.

[0020] See also Figure 2 , the geometric feature analysis module includes: The boundary curve acquisition submodule acquires the boundary curve data of the paper-cut pattern, analyzes the paper-cut area, obtains the partition coordinates of each area, establishes a coordinate index, and locates the paper-cut area according to the partition coordinates to generate the boundary data of the paper-cut area; The paper-cut pattern is scanned or photographed with high precision to obtain the original image data containing complete boundary information. Subsequently, the image data is binarized to distinguish the paper-cut area from the background area. The edge detection algorithm, such as Canny edge detection, is used to extract the outline of the paper-cut pattern to obtain all boundary point sets. The partition coordinates of each area are determined by dividing the extracted boundary point set into regions. The division method is usually based on the topological structure or spatial distribution characteristics, and the adjacent and connected boundary points are classified into the same region. In each partition, the geometric center of the boundary points is further calculated, and the outermost envelope of the partition is determined to establish a coordinate index. The process of establishing the coordinate index involves constructing a spatial data structure, such as an R-tree index. The role of the index is to accelerate the positioning of the paper-cut area so that the subsequent boundary feature analysis can be carried out efficiently. The method of locating the paper-cut area based on the partition coordinates is to match the overall boundary data of the paper-cut pattern with the partition coordinate index to determine the specific location of each partition. For example, if the paper-cut pattern is a floral pattern, the boundary curve of each petal can be used as a partition, and the position of each petal in the overall pattern is determined by the coordinate index to generate the paper-cut area boundary data.

[0021] The local contour gradient calculation submodule is based on the paper-cut area boundary data, analyzes the pattern data, collects the gradient information of the boundary point set, and uses the formula: ; Calculate the local contour gradient of each point to obtain the local gradient distribution characteristics, where: Representative The local contour gradient of the point, and Respectively represent The adjacent point Pointed and The coordinate difference of the direction, For the The sum of the points in the point's neighborhood; A certain point The coordinate in the pattern is (100, 200), and there are 5 adjacent points in its neighborhood, namely (102, 201), (98, 199), (101, 202), (99, 198) and (100, 203). Calculate the relative coordinate change: , ; Calculate the gradient modulus: ; After the gradient value is calculated, the gradient direction is determined by calculating the relative angle. For example, if the gradient vector of a point is (2, 1), the gradient direction is calculated as: ; It shows that the local contour gradient value of the boundary point reflects the smoothness of the boundary curve of the paper-cut pattern and the intensity of local changes. When the gradient value is small, it means that the boundary change of the area where the point is located is relatively gentle and the edge transition is relatively natural. When the gradient value is large, it means that the point is in an area with drastic boundary changes, corresponding to sharp corners, small depressions or complex texture areas in the paper-cut pattern. The gradient value directly affects the accuracy of subsequent curvature calculation, because areas with large gradient changes often correspond to areas of sudden curvature changes. By counting and classifying local gradient values, we can further screen out areas with the most drastic boundary changes and use them as the focus of subsequent curvature calculations.

[0022] The curvature mutation screening submodule analyzes the gradient direction and value distribution according to the local gradient distribution characteristics, compares the curve with the gradient data, screens the curvature mutation area, and obtains the local curvature characteristic index; To classify local gradients, gradient direction data is usually organized by angle interval division. For example, the gradient direction can be classified into different intervals such as 0°–30°, 30°–60°, 60°–90°, and the gradient value distribution density in each interval is counted. In the statistical process, it is necessary to calculate the proportion of points with different gradient directions in each area, as well as the mean and standard deviation of the gradient values ​​in each direction, so as to obtain the trend characteristics of gradient changes. Suppose a paper-cut pattern contains multiple different boundary curve areas, and there are boundary points with different gradient directions in each area. By counting the gradient direction distribution of the boundary points, it can be determined which areas have a larger gradient change and which areas have a smaller gradient change. When comparing the curve with the gradient data, it is necessary to calculate the local curvature of the boundary curve. The calculation of the local curvature depends on the coordinate change of the boundary points. For each boundary point, it is first necessary to calculate the coordinate offset of its front and rear adjacent points, and then calculate the curvature of the curve based on the offset. Areas with large curvature changes usually correspond to complex contours in paper-cut patterns, such as tortuous edges, sharp corners, or fine hollow parts. When screening areas with curvature mutations, it is necessary to set a certain curvature change threshold. The setting of this threshold is usually based on the overall complexity of the paper-cut pattern. For example, in a floral pattern, the edge of the leaf has a large curvature change, while the edge of the petal is relatively smooth. Therefore, a curvature change threshold can be set, such as 0.1. When the curvature change of a boundary point exceeds the threshold, it can be judged that the point is in the curvature mutation area. After all boundary points are calculated, all boundary point sets that meet the curvature mutation conditions are stored and organized into a curvature mutation distribution data table to obtain local curvature characteristic indicators.

[0023] See also Figure 3 , the texture dynamic adjustment module includes: The texture density calculation submodule is based on the local curvature characteristic index and the basic texture of paper cutting to determine the trend of texture density change. The formula is: ; Calculate the texture density change rate of each area and establish the texture density change parameters ,in, Representative The curvature value of the region, Representative The curvature value of the region, Representative The curvature gradient of a region, represents the total number of regions; The variation trend of texture density can be determined by the basic texture data of paper-cutting. The grid data model of paper-cutting texture can be established by high-resolution scanning. The variation rate of texture density in each area can be calculated by taking the grid unit as the basic unit. The variation rate can be calculated by the sampling data of the edge and internal texture of the paper-cutting. For example, the curvature data of the edge area of ​​a paper-cutting work are , and The corresponding curvature gradients are , and , substitute into the formula and get: ; Calculated Represents the local texture density change parameter of the paper-cut work. In the set paper-cut work density change threshold, if Exceeding the set If the threshold is smaller than 0.05, it indicates that the area has a large texture density change and needs to be paid special attention in subsequent adjustments. The calculation result can quantify the local density change characteristics of the paper-cut works and establish the texture density change parameters.

[0024] The texture direction adjustment submodule calls the texture density variation parameter, analyzes the correspondence between the texture parameter and the curvature data, and adjusts the texture density and texture direction along the local curvature gradient distribution to obtain the texture adaptability parameter; During the texture direction adjustment process, the distribution of the curvature gradient is relied upon to decide how to adjust the texture direction. Areas with larger curvature gradients usually need to adjust the texture direction to maintain a visual sense of naturalness and balance. When adjusting the texture density and texture direction, it is necessary to calculate the local texture adaptability parameters of each area to ensure that the texture direction and density match the surrounding local curvature changes. For example, in areas with large changes in local curvature gradients, by adjusting the texture direction, the texture can appear to conform to changes in the terrain. This adjustment process makes the overall texture display more natural and meets the performance requirements of the actual terrain, and the texture adaptability parameters are obtained.

[0025] See also Figure 4 ,The paper-cutting style analysis module includes: The style data collection submodule calls the texture adaptability parameters, collects paper-cut design style data, collects paper-cut style information in different time periods, and classifies the data according to the time interval, counts the variation range of paper-cut style parameters, and obtains paper-cut style time series data; The paper-cut works are converted into digital images, and key feature parameters in the images are extracted, such as color proportion, edge complexity and symmetry. For paper-cut works in different time periods, a fixed time interval can be set, such as 5 or 10 years, in order to classify the data and analyze its changing trends. In this process, it is necessary to count the variation range of paper-cut style parameters in each time period. For example, in a certain time period, the edge complexity of the paper-cut works increased by 10% and the symmetry decreased by 5%. By calculating the fluctuations of the characteristic parameters in each time period, the evolution direction of the paper-cut style can be reflected. After data collection, the data is classified according to the time interval to form time series data, and the variation range of paper-cut style parameters in each time period is counted to obtain paper-cut style time series data.

[0026] The key style screening submodule calls the paper-cut style time series data, compares the changes in style parameters in different time periods, and uses the formula: ; Calculating Style Volatility , and filter the key style feature values, where Representative The change range of style parameters in each time period, Representative The duration of a time period, represents the total number of time periods; Assume that the time periods are 1950-1970, 1970-1990, and 1990-2010, and count the changes in the style characteristics of paper-cut works in each time period. The characteristic parameters Represents the style complexity change and duration In years: 1950-1970: Stylistic Complexity , duration Year; 1970-1990: Stylistic complexity , duration Year; 1990-2010: Stylistic complexity , duration Year; Substitute the data into the formula to calculate: ; Calculated Style Volatility , if the value exceeds the set threshold, e.g. , it indicates that the paper-cutting style has a large range of changes in this period, and it is necessary to further screen the key style feature values. For example, the style changes from 1950 to 1970 are mainly reflected in the increase of texture details, while from 1990 to 2010, the pattern edges are more complicated, and the style adjustment occurs during 1970-1990. The calculated volatility can be used to quantify the evolution trend of the style, screen out the key style feature values, and establish the key style feature values.

[0027] The style trend assessment submodule calculates the adaptability of key style feature values ​​in the current style trend, compares the offset direction of style data, analyzes the evolution trend over time, and obtains the style trend analysis results; If a key style feature value shows an upward trend in multiple time periods, such as symmetry gradually increasing from 80% to 90%, it can be determined that the style feature is more popular in modern paper-cutting design. On the other hand, if a certain feature value shows a significant decrease, such as the complexity of traditional patterns decreasing from 70% to 50%, it indicates that the style is gradually being replaced by a simplified style. To further analyze the style trend, it is also necessary to compare the offset direction of the style data to determine whether the style change is cyclical. For example, some traditional styles decline in a certain period of time, but become popular again in subsequent time periods. By calculating the evolution trend of each key style feature value over time, a time series prediction model for style changes is established to obtain the style trend analysis results.

[0028] See also Figure 5 , the user interaction adjustment module includes: The operation parameter statistics submodule analyzes user interaction data based on the style trend analysis results, counts the user's zoom, rotation and color selection operations, and identifies the range of operation parameters to obtain the range of user operation changes; Get the zoom, rotation and color selection parameters of the user operation. For the zoom parameters, monitor the values ​​of the user's zoom operations on the interface in different time periods, calculate the amplitude change of each zoom, and store the initial and final values ​​of each zoom operation. For example, if the user zooms in from 100% to 150%, record the zoom parameter value 1.5. If it is reduced from 150% to 120%, record the zoom parameter value 0.8. After accumulating all zoom parameters, calculate the maximum and minimum values, and determine the range of zoom parameter changes. For the rotation parameters, monitor the angle changes of the user interface rotation operation. For example, if the user rotates from 0° to 30°, record the rotation angle 30°. If If the color selection parameter is changed from 30° to -15°, the rotation angle of -15° is recorded. The angle range of all rotation operations is counted to determine the range of rotation parameter changes. For color selection parameters, the changes in RGB or HSL values ​​when the user selects a color are monitored. For example, if the user selects a color (120, 200, 255) and changes it to (110, 190, 240) in subsequent operations, the color parameter change value (10, 10, 15) is recorded. All color change data is accumulated and the color selection range is calculated. Finally, the upper and lower limits of the overall change are screened by combining the scaling, rotation and color selection parameters to obtain the user operation change range.

[0029] The interaction parameter analysis submodule analyzes the fluctuation range of the scaling and rotation parameters based on the range of user operation changes, evaluates the degree of deviation of color selection, and obtains the user interaction fluctuation range; For the scaling parameters, calculate the relative change rate of each scaling. For example, a scaling operation from 100% to 150% has a change rate of 50%, and from 150% to 120% has a change rate of -20%. Calculate the change rate of all scaling operations and obtain their standard deviations. For the rotation parameters, calculate the relative angle change of each rotation. For example, if one operation rotates from 0° to 45°, the change is 45°. The next rotation is from 45° to 30°, the change is -15°. Calculate the change of all rotation operations and calculate their standard deviation. , for color selection, calculate the color shift value, such as calculating the Euclidean distance of RGB color change , accumulate the offset values ​​of all color selection changes and calculate their mean and standard deviation ,Finally, the volatility of scaling, rotation and color selection is combined to obtain the ,user interaction fluctuation interval.

[0030] The adaptation value calculation submodule analyzes the volatility of parameters and the deviation trend of color selection according to the user interaction fluctuation range, using the formula: ; Get user preference adaptation value ,in, Representative Secondary scaling parameter value, represents the mean value of the scaling parameter, represents the standard deviation of the scaling parameter, Representative Secondary rotation parameter value, represents the mean value of the rotation parameter, represents the standard deviation of the rotation parameter, Representative Secondary color selection value, Represents the color reference value under the current style trend, represents the weight coefficient of color selection deviation, Represents the total number of operations; Assume that the user's scaling parameter mean , standard deviation , the user's zoom operations are as follows (data based on 10 operations): , calculate the deviation of each scaling: ; ; Calculate the rotation parameter deviation: Mean of the rotation parameter Degree, standard deviation The user's rotation operations are as follows (data based on 10 operations): , calculate the deviation of each rotation: ; ; Calculate the color parameter deviation: Mean reference value of color parameters (such as the red channel in the RGB color value), the user's color selection data is as follows (the data is based on 10 operations): , calculate the square value of each color deviation: ; ; Calculate the final adaptation value: Color offset weight , total number of operations , substitute into the calculation: ; Calculated This value indicates that the user's interaction parameters have a high volatility, especially the deviation of color selection has a large contribution (the total deviation square value of 1168), reflecting that the user's color selection deviates far from the current style trend. In contrast, the fluctuation range of scaling and rotation parameters is smaller. The system can adjust the weight of the interaction parameters according to this value to adapt to user preferences and optimize the interface experience.

[0031] See also Figure 6 , the paper-cut pattern optimization module includes: The pattern morphology analysis submodule analyzes the morphology data of the paper-cut pattern according to the user preference adaptation value, extracts the edge curvature of the contour, calculates the curvature variation interval, and determines the reference value of the curvature variation to obtain the pattern contour curvature data; The contour line is extracted through edge detection, and the main curve features of the pattern are identified. For each contour point, its local curvature is calculated. The curvature can be calculated by the three-point method, that is, the curvature is calculated by the coordinate change of the adjacent points. The curvature calculation interval is set. For example, if the curvature is greater than 0.1 or less than -0.1, it is determined that there is a significant curvature change at the point. For the area with a relatively stable curvature change, its curvature mean is recorded, and the curvature change curve is constructed. The change interval of the entire contour curvature is calculated, the curvature distribution is analyzed, and the reference value of the curvature is determined. The reference value is selected with reference to the mean and standard deviation of the overall curvature, and the interval [ ] as the normal curvature variation range, and obtain the pattern contour curvature data.

[0032] The symmetry offset detection submodule evaluates the symmetry of the pattern based on the pattern contour curvature data, identifies the offset area exceeding the normal range by measuring the morphological offset on both sides of the pattern symmetry axis, and generates the symmetry offset; According to the curvature data of the pattern contour, the symmetry axis of the paper-cut pattern is analyzed, and the left and right symmetrical point pairs are selected. The distance between the point pairs is calculated. For each point pair, the curvature of the points on both sides is recorded and the offset is calculated. The offset is defined as the absolute value of the curvature difference between the two points. If the offset exceeds the threshold, it is determined that the point pair has a morphological offset. The reference benchmark for the symmetry offset is set. The benchmark value selects 95% of the normal offset range, that is, the data interval remaining after removing the maximum 5% offset value. The offset area beyond this range is screened, and the total offset is calculated to generate the symmetry offset.

[0033] The contour curvature correction submodule analyzes the contour area with abnormal curvature change based on the symmetry offset, adjusts and corrects the curvature of the area, and uses the formula: ; Get the corrected contour curvature , optimize the pattern outline and generate paper-cut design files, where Representative The original curvature of the contour points, represents the average value of the profile curvature, represents the standard deviation of the curvature, Representative The offset of the points, represents the standard deviation of the offset, Represents the total number of contour points; Set the mean curvature , standard deviation of curvature , standard deviation of offset , total number of contour points . Assume that the original curvatures of the three contour points are , , , and the corresponding offsets are , , .

[0034] Compute the normalized deviation of curvature for each contour point: ; ; ; Calculate the offset influence factor for each contour point: ; ; ; Compute the modified curvature contribution for each contour point: ; ; ; Calculate the corrected curvature: ; Calculated , compared with the original mean curvature , the corrected curvature value increases, indicating that the curvature adjustment of the pattern contour is larger. The corrected curvature will be used to update the paper-cut design file to optimize the shape of the pattern contour to make it more in line with the expected paper-cut style.

[0035] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. An artificial intelligence-based paper-cutting pattern automatic generation system, characterized in that: The system comprises: The geometric feature analysis module collects the boundary curve data of the paper-cut pattern, locates the paper-cut area using the internal partition coordinates, identifies the local contour gradient by analyzing the pattern data, compares the curve with the gradient data, analyzes the gradient direction and value distribution, screens the curvature mutation area, and obtains the local curvature characteristic index; The texture dynamic adjustment module determines the texture density change trend based on the local curvature characteristic index and the basic paper-cut texture, analyzes the corresponding relationship between the texture parameters and the curvature data, and adjusts the texture density and direction along the local curvature gradient distribution to obtain the texture adaptability parameter; The paper-cutting style analysis module calls the texture adaptability parameter, analyzes the fluctuation trend of the paper-cutting style in different time periods, selects key style feature values, evaluates their adaptability in the current trend, and compares the offset direction of the style data to obtain the style trend analysis result; Based on the style trend analysis results, the user interaction adjustment module counts the range of changes in user operation parameters, evaluates the fluctuation range of user zoom and rotation parameters, analyzes the style deviation degree of color selection, screens operation parameters that match the current trend, and obtains user preference adaptation values.

2. The system for automatically generating paper-cutting patterns based on artificial intelligence according to claim 1, characterized in that: The local curvature characteristic indicators include curvature extremes, gradient peaks, and regional distribution; the texture adaptability parameters include texture intensity index and directional consistency index; the style trend analysis results include trend adaptation scores and style evolution vectors; and the user preference adaptation values ​​include user operation sensitivity and style consistency ratings.

3. The system for automatically generating paper-cutting patterns based on artificial intelligence according to claim 1, characterized in that: The geometric feature analysis module comprises: The boundary curve acquisition submodule acquires the boundary curve data of the paper-cut pattern, analyzes the paper-cut area, obtains the partition coordinates of each area, establishes a coordinate index, and locates the paper-cut area according to the partition coordinates to generate the boundary data of the paper-cut area; The local contour gradient calculation submodule is based on the paper-cutting area boundary data, analyzes the pattern data, collects the gradient information of the boundary point set, and adopts the formula: ； Calculate the local contour gradient of each point to obtain the local gradient distribution characteristics, where: Representative The local contour gradient of the point, and Respectively represent The adjacent point Pointed and The coordinate difference of the direction, For the The sum of the points in the point's neighborhood; The curvature mutation screening submodule analyzes the gradient direction and value distribution according to the local gradient distribution characteristics, compares the curve with the gradient data, screens the curvature mutation area, and obtains the local curvature characteristic index.

4. The system for automatically generating paper-cutting patterns based on artificial intelligence according to claim 1, characterized in that: The texture dynamic adjustment module comprises: The texture density calculation submodule is based on the local curvature characteristic index and the basic texture of paper cutting to determine the texture density change trend, using the formula: ； Calculate the texture density change rate of each area and establish the texture density change parameters ,in, Representative The curvature value of the region, Representative The curvature value of the region, Representative The curvature gradient of a region, represents the total number of regions; The texture direction adjustment submodule calls the texture density variation parameter, analyzes the corresponding relationship between the texture parameter and the curvature data, and adjusts the texture density and texture direction along the local curvature gradient distribution to obtain the texture adaptability parameter.

5. The system for automatically generating paper-cutting patterns based on artificial intelligence according to claim 1, characterized in that: The paper-cutting style analysis module includes: The style data collection submodule calls the texture adaptability parameter, collects paper-cut design style data, collects paper-cut style information in different time periods, and classifies the data according to the time interval, counts the variation range of the paper-cut style parameters, and obtains paper-cut style time series data; The key style screening submodule calls the paper-cut style time series data, compares the style parameter changes in different time periods, and uses the formula: ； Calculating Style Volatility , and filter the key style feature values, where Representative The change range of style parameters in each time period, Representative The duration of a time period, represents the total number of time periods; The style trend evaluation submodule calculates the adaptability of the key style feature value in the current style trend according to the key style feature value, compares the offset direction of the style data, analyzes the evolution trend over time, and obtains the style trend analysis result.

6. The system for automatically generating paper-cutting patterns based on artificial intelligence according to claim 1, characterized in that: The user interaction adjustment module comprises: The operation parameter statistics submodule analyzes the user interaction data based on the style trend analysis result, counts the user's zoom, rotation and color selection operations, and identifies the variation range of the operation parameters to obtain the user operation variation range; The interaction parameter analysis submodule analyzes the fluctuation range of the scaling and rotation parameters based on the user operation variation range, evaluates the deviation degree of the color selection, and obtains the user interaction fluctuation range; The adaptation value calculation submodule analyzes the volatility of parameters and the deviation trend of color selection according to the user interaction fluctuation interval, and adopts the formula: ； Get user preference adaptation value ,in, Representative Secondary scaling parameter value, represents the mean value of the scaling parameter, represents the standard deviation of the scaling parameter, Representative Secondary rotation parameter value, represents the mean value of the rotation parameter, represents the standard deviation of the rotation parameter, Representative Secondary color selection value, Represents the color reference value under the current style trend, represents the weight coefficient of color selection deviation, Represents the total number of operations.

7. The system for automatically generating paper-cutting patterns based on artificial intelligence according to claim 1, characterized in that: The system further comprises: The paper-cut pattern optimization module analyzes the pattern morphology data according to the user preference adaptation value, evaluates the edge curvature change of the pattern contour data, determines the symmetry offset of the morphology data, screens the morphology with an offset exceeding the normal range, and then intelligently adjusts and corrects the contour with abnormal curvature change to generate a paper-cut design file; The paper-cut design file includes design accuracy, detail completeness, and visual effect scores.

8. The system for automatically generating paper-cutting patterns based on artificial intelligence according to claim 7, characterized in that: The paper-cut pattern optimization module comprises: The pattern morphology analysis submodule analyzes the morphology data of the paper-cut pattern according to the user preference adaptation value, extracts the edge curvature of the contour, calculates the curvature variation interval, and determines the reference value of the curvature variation to obtain the pattern contour curvature data; The symmetry offset detection submodule evaluates the symmetry of the pattern according to the pattern contour curvature data, identifies the offset area exceeding the normal range by measuring the morphological offset on both sides of the pattern symmetry axis, and generates the symmetry offset; The contour curvature correction submodule analyzes the contour area with abnormal curvature change according to the symmetry offset, adjusts and corrects the curvature of the area, and adopts the formula: ； Get the corrected contour curvature , optimize the pattern outline and generate paper-cut design files, where Representative The original curvature of the contour points, represents the average value of the profile curvature, represents the standard deviation of the curvature, Representative The offset of the points, represents the standard deviation of the offset, Represents the total number of contour points.

Images