Clothing pattern denim art style migration method

Abstract

The invention discloses a clothing pattern jean art style migration method. The method comprises the following steps: (1) calculating the widths of warp yarns and weft yarns according to a pattern image resolution and an actual printing size; (2) carrying out denim warp yarn texture stylization on the pattern image; (3) performing denim weft yarn texture stylization on the pattern image; and (4) according to the position of each pixel, carrying out denim warp and weft color stylization. According to the method, any pattern can be converted into the jean style so as to assist a designer in solving the problem in the aspect of pattern generation, and the method is applied to clothing pattern printing and dyeing of various non-jean fabrics and has good convenience and flexibility.

Description

Technical Field

[0001] This invention belongs to the field of digital image processing technology, specifically relating to a method for transferring the artistic style of denim clothing patterns. Background Technology

[0002] Currently, clothing pattern design still heavily relies on designers' experience and handcrafted creation. Designers must complete complex processes such as pattern conception, color matching, and layout planning within a limited time, often requiring several weeks to refine a single design. This traditional model exposes a sharp contradiction between creative efficiency and market demand when faced with the weekly new product launches required by the fast fashion industry. With the advancement of computer technology, pattern style transfer technology, with its unique advantages, provides strong support for the sustainable development of the textile industry. This technology can effectively reuse existing design resources, reduce over-reliance on new materials, and thus lessen the environmental burden of production activities, aligning with the concept of a circular economy. In actual clothing pattern selection, some users like denim-style patterns but dislike denim fabric, or wish to have denim patterns printed on T-shirts, cotton-padded jackets, and other garments.

[0003] Existing research on art style transfer has achieved certain results and formed some algorithm models. In terms of theoretical research, the literature [Jiang Wenqin, Yan Jingwen, Luo Qianqian. Research on color transfer of sweater pattern based on butterfly color style [J]. Knitting Industry, 2024(5)] carried out research on color transfer of sweater pattern based on butterfly color style. Three different color transfer algorithms were selected to automatically transfer the color style of the butterfly to the target pattern. The output pattern was combined with the knitted sweater design and different butterfly color styles were presented in the form of virtual design. The literature [Zhang Yuyu, Wan Pengbo, Zhu Xiaofeng. Application of style transfer algorithm in personalized design of shoe and boot pattern [J]. Leather Science and Engineering, 2025, 35(05): 83-90] used style transfer to realize personalized design of shoe and boot pattern. In order to solve the problem that the existing style transfer model is not effective in shoe and boot products, a personalized style transfer method for shoe and boot area control was proposed. References [Tao Ye, Wang Shana. Research on Service Design of Xianju Lantern Handicraft Based on Style Transfer Algorithm [J]. Packaging Engineering, 2025, 46(16): 243-252] analyze image datasets related to Xianju lantern handicraft, deeply explore traditional handicraft image data, and propose a testing method for Xianju lantern style transfer algorithm. References [He Yanfen, Hou Jue, Qiu Xuelin. Construction of Yunjin Image Generation System Based on Style Transfer [J]. Wool Textile Technology, 2025, 53(08): 45-53] propose a local style transfer method based on color optimization to improve the clarity and color restoration effect of traditional style transfer model in Yunjin image generation, and construct a Yunjin style image generation system. References [Dong Zhengyu. Research on Unsupervised Style Transfer of Chinese Traditional Painting Based on Generative Adversarial Network [D]. Yantai University, 2025] focus on Chinese traditional painting. It conducts optimization research on network models to address the problems of existing deep learning models in different artistic expressions of Chinese painting.

[0004] In terms of patent application, Chinese patent application CN120765451A proposes a bidirectional fusion image style transfer method and system based on content and style decoupling. This method uses a cross-attention mechanism to bidirectionally fuse style features and content features to generate fused bidirectional features. Based on the acquired text description, text features are extracted and concatenated with the bidirectional fusion features to generate a conditional control vector as guiding information for image generation. The acquired target image is then subjected to a denoising inverse diffusion process guided by the conditional control vector, and finally, the stylized image is decoded. Based on the bidirectional feature fusion of style and content and the text-guided diffusion generation mechanism, the method achieves an image style transfer effect that is refined in style expression, clear in semantic structure, and highly controllable. Chinese patent application CN120635247A proposes a method and system for image generation and style transfer based on a large model. This method acquires user-uploaded content images, style images, and slider parameters, and captures the slider trajectory to generate a timestamped intensity sequence, which is then dynamically sampled and compressed into a frequency domain feature vector. A large model is used to parse the semantic contours of the content image and the texture regions of the style image in parallel, establishing semantic associations and matching target texture features. The target features are progressively weighted and fused according to the style intensity parameters. Finally, the frequency domain vector and the fused features are injected into the model's latent space as control signals. Simultaneously, the iteration depth of the latent space is dynamically adjusted based on device performance, and the output frame rate is kept stable by constraining the noise attenuation path, generating a continuous stylized preview stream.

[0005] Most of the theoretical research and patented technologies mentioned above have not been combined with practical applications in the textile industry, and there are no style transfer methods specifically for denim fabrics. Summary of the Invention

[0006] In response to the practical application needs faced in the above-mentioned clothing pattern design, this invention provides a method for transferring clothing pattern denim art style, which can convert any pattern into a denim style to help designers solve problems in pattern generation, and can be applied to the printing and dyeing of clothing patterns on various non-denim fabrics.

[0007] A method for transferring denim art style from clothing patterns includes the following steps: (1) Calculate the warp and weft widths based on the pattern image resolution and the actual printing size; (2) The denim warp texture of the floral pattern image is stylized; (3) The denim weft texture of the floral pattern image is stylized; (4) According to the pixel position, the denim warp and weft yarn color stylization processing is performed on the floral image.

[0008] Further, in step (1), the warp width p is calculated using the following expression. v and weft yarn width p h: , Where: d v and d h These represent the number of warp and weft yarns per inch (set by the user), y v and y h These are the actual height and width of the floral pattern image during printing (in inches, set by the user, generally consistent with the aspect ratio of the image resolution), while w and h are the width and height of the floral pattern image resolution (this parameter is determined by the floral pattern image).

[0009] Furthermore, the specific implementation of step (2) is as follows: S21: Divide the flower pattern image into multiple p v ×6p v A rectangular block of a certain size; S22: Starting from the second column of rectangular blocks in the flower pattern image, each column is shifted down 2 pixels from the left column; S23: For any rectangular block, divide the pixels in the block into 3 categories, and let the outermost pixel be P. w The outermost pixel is P z The remaining pixels are P n ; S24: Stylize the brightness of each pixel in the floral image based on the pixel's position and category to simulate a cowboy style.

[0010] Furthermore, step S24 performs stylistic correction on the brightness of each pixel in the flower pattern image using the following relationship: Where: Y(P) w ) and Y'(P w ) represent the pixel P before and after correction. w The brightness value, Y(P) z ) and Y'(P z ) represent the pixel P before and after correction. z The brightness value, Y(P) n ) and Y'(P n ) represent the pixel P before and after correction. n The brightness value, t1 is the preset value, L is the distance of the current pixel in the vertical direction from the center point of the rectangle, and Y is the value of the brightness value. bw Y bz Y bn To simulate the preset values ​​of warp yarn texture in denim fabric, P fP represents the pixels of the flower pattern. b R represents the background pixels. bw R bz R bn To simulate random numbers with a sense of unevenness.

[0011] Furthermore, the specific implementation of step (3) is as follows: S31: Divide the flower pattern image into multiple p h ×6p v A rectangular block of a certain size; S32: Starting from the second column of rectangular blocks in the flower pattern image, each column is shifted down p based on the left column. h +1 pixel; S33: For any rectangular block, divide the pixels in the block into 3 categories: For the topmost pixel in the rectangular block... h ×p h A square of size P, with the outermost pixels excluding the four vertices set to P. s The internal pixels are set to P. x Apart from these two types of pixels, the remaining pixels in the rectangular block are P. y ; S34: Analyze pixel P in the flower pattern image according to its position and category. s and P x The brightness was stylized to simulate a cowboy style.

[0012] Furthermore, step S34 uses the following relationship to determine the pixel P in the flower pattern image. s and P x Stylize the brightness: Where: Y(P) s ) and Y'(P s ) represent the pixel P before and after correction. s The brightness value, Y(P) x ) and Y'(P x ) represent the pixel P before and after correction. x The brightness value, t2 is the preset value, L is the distance of the current pixel in the vertical direction from the center point of the rectangle, and Y is the value of the brightness value. bs and Y bx To simulate the preset values ​​of the weft yarn texture of denim fabric, P f P represents the pixels of the flower pattern. b R represents the background pixels. bs and R bx To simulate random numbers with a sense of unevenness.

[0013] Furthermore, the specific implementation of step (4) is as follows: First, for the pixel P of the flower shape portion in the flower shape image f Its chromaticity is stylized using the following relationship: Then, for the pixels in the background portion of the flower-shaped image, their chroma is further stylized according to the pixel category using the following formula: Where: Cb(P) f ) and Cb'(P f ) represent the pixel P before and after correction. f The blueness component value, Cr(P) f ) and Cr'(P f ) represent the pixel P before and after correction. f The redness component value, Cb'(P s ), Cb'(P x ), Cb'(P w ), Cb'(P z ), Cb'(P n ) represent the pixels P in the corrected background area. s P x P w P z P n The blueness component value, Cr'(P) s ), Cr'(P x ), Cr'(P w ), Cr'(P z ), Cr'(P n ) represent the pixels P in the corrected background area. s P x P w P z P nThe redness component values, t3 and t4 are preset values, R fcb R fcr R scb R scr R xcb R xcr R wcb R wcr R zcb R zcr R ncb and R ncr All are random numbers simulating a sense of unevenness. Cb1 and Cr1 are the chromaticity component values ​​of the warp yarn reference color, and Cb2 and Cr2 are the chromaticity component values ​​of the weft yarn reference color.

[0014] A computer device includes a memory and a processor, wherein the memory stores a computer program and the processor executes the computer program to implement the above-described method for transferring denim art style in clothing patterns.

[0015] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for transferring the denim art style of clothing patterns.

[0016] Based on the above technical solution, the present invention has the following beneficial technical effects: 1. This invention can convert any floral image into a denim-style floral pattern, offering excellent flexibility and adaptability.

[0017] 2. The denim floral pattern generated using this invention can be directly used for pattern printing on clothing such as T-shirts and sweatshirts, which can greatly improve design and production efficiency.

[0018] 3. The method of the present invention has good effect on various fabrics and has good versatility. Attached Figure Description

[0019] Figure 1 This is an illustration of denim-style fabric.

[0020] Figure 2 This is a schematic diagram of the flower pattern to which this invention applies.

[0021] Figure 3 This is a schematic diagram of rectangular blocks used to stylize the warp yarn texture.

[0022] Figure 4 This is a schematic diagram illustrating the pixel classification within a rectangular block stylized as a warp texture.

[0023] Figure 5 This is a schematic diagram illustrating the division of rectangular blocks into stylized weft yarn textures.

[0024] Figure 6A schematic diagram illustrating pixel classification within a rectangular block stylized for weft yarn texture.

[0025] Figure 7 This is a schematic diagram of the actual printing effect after style transfer in an embodiment of the present invention. Detailed Implementation

[0026] To describe the present invention in more detail, the technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0027] This embodiment provides a method for transferring the denim art style of clothing patterns, including the following steps: (1) Calculate the width of the warp and weft yarns based on the resolution of the pattern image and the actual printing size.

[0028] In ordinary fabrics, warp yarns are the yarns arranged along the length (longitudinal direction) of the fabric, while weft yarns are the yarns arranged along the width (transverse direction) of the fabric, interweaving with the warp yarns to form the fabric structure. However, in denim fabric weaving, the warp and weft yarns interweave perpendicularly, but after leaving the machine, the weft and warp yarns on the fabric surface exist in a parabolic shape at a certain angle. Therefore, the weft yarn changes after leaving the machine; the angle between the weft and warp yarns changes from 90 degrees to less than 90 degrees, which is commonly referred to as weft skew. Figure 1 The left image shows ordinary denim fabric, while the right image is an enlarged version, showing the warp yarns as vertically spaced, discontinuous dark blue rectangular blocks, and the weft yarns as diagonally discontinuous white dotted blocks. The purpose of this embodiment is to transfer and introduce similar [fabric] patterns into ordinary floral images. Figure 1 The base pattern is in the style of denim.

[0029] Warp density is the number of warp yarns per unit length along the vertical direction of the selvage, and weft density is the number of weft yarns per inch along the horizontal direction of the selvage. Denim fabric is generally measured in imperial units, with units of warp yarns per inch × weft yarns per inch. Common units include 48×30, 72×44, and 106×58. In this embodiment, this parameter is user-set (not exceeding 106×58), and is set to d. v ×d h d v and d h These are the number of warp yarns and weft yarns per inch, respectively. Set the actual printing and dyeing size y of the pattern image. v and y h , respectively, represent the actual height and width during printing, in inches. This dimension is user-defined and generally maintains the aspect ratio of the image resolution. The pattern image resolution is w and h, representing the image width and height, respectively. This parameter is determined by the pattern image.

[0030] warp width p v and weft yarn width p h The calculation method is as follows: , Where: p v and p h The unit is pixels, and the result is rounded down. To achieve better results, the resolution of the floral image and the aspect ratio of the printed pattern should be as high as possible, and the w / y ratio should be... h Greater than 600, h / y v Greater than 300.

[0031] (2) Stylize the denim warp texture of the floral pattern image.

[0032] The flower pattern image processed in this embodiment is as follows: Figure 2 The flower pattern and background shown are clearly distinguishable, and the background color is a solid color. For each pixel P in the flower pattern image, the pixels belonging to the flower pattern are set as P0. f The pixels belonging to the background area are set to P. b .

[0033] Starting from the top left of the flower image, divide all pixels into p v ×6p v To simulate the staggered texture of warp threads in denim fabric, starting from the second column of rectangles from the left, each column is shifted down 2 pixels from the column on the left, as shown below. Figure 3 As shown. Within each rectangular block, pixels are divided into three categories based on their position, with the outermost red pixels designated as P. w The outermost green pixel is set to P. z The remaining pixels are set to P. n ,like Figure 4 As shown.

[0034] When stylizing the denim warp texture of a floral image, the denim style is simulated by changing the brightness value Y of each pixel; the stylization calculation formula is different for each type of pixel in the rectangular block, and the calculation method is as follows: Where: t1 is a preset value, L is the pixel distance of this pixel from the center point of this rectangle in the vertical direction; Y bw Y bz Y bn These are preset values ​​that simulate the warp texture of denim fabric; R bw R bz R bn These are random numbers, simulating a sense of unevenness; the final brightness value Y' is rounded to the nearest integer.

[0035] (3) Stylize the denim weft texture of the floral pattern image.

[0036] Starting from the top left of the flower image, divide all pixels into p h ×6p v The rectangular blocks, designed to simulate the slanted texture of the weft yarns in denim fabric, are arranged starting from the second column from the left, with each column shifted downwards by p units from the column on the left. h +1 pixel, such as Figure 5 As shown. Within each rectangular block, they are divided into three categories based on pixel position, with the topmost p... h ×p h In the block, the outermost red pixels excluding the four vertices are set to P. s The internal green pixels are set to P. x The remaining white pixels are not included in the calculation, such as Figure 6 As shown.

[0037] When stylizing the denim weft texture of a floral image, the denim style is simulated by changing the brightness value Y of each pixel. The stylization calculation formula differs for each type of pixel within the rectangular block; the calculation method is as follows: Where: t2 is a preset value, L is the pixel distance of this pixel from the center point of this rectangle in the vertical direction; Y bs Y bx These are preset values ​​that simulate the weft texture of denim fabric; R bs R bx These are random numbers, simulating a sense of unevenness; the final brightness value Y' is rounded to the nearest integer.

[0038] (4) Stylize the denim warp and weft yarn colors according to the position of each pixel.

[0039] To simulate the style of denim fabric, the colors of the pattern image need to be adjusted: First, set the chromaticity component values ​​Cb1 and Cr1 for the warp reference color and Cb2 and Cr2 for the weft reference color; then calculate the chromaticity component values ​​of all pixels in the following order.

[0040] Pixel P belonging to the flower shape f The method for calculating colorimetry is as follows: Where: t3 and t4 are preset values, R fcb and R fcr It is a random number to simulate a sense of variation; finally, the chromaticity values ​​Cb' and Cr' are rounded to the nearest integer.

[0041] Of the remaining pixels, pixel P belongs to the background area. s and P x The method for calculating colorimetry is as follows: Where: R scb R scr R xcb R xcr It is a random number to simulate a sense of variation; finally, the chromaticity values ​​Cb' and Cr' are rounded to the nearest integer.

[0042] Of the remaining pixels, pixel P belongs to the background area. w P z and P n The method for calculating colorimetry is as follows: Where: R wcb R wcr R zcb R zcr R ncb and R ncr It is a random number to simulate a sense of variation; finally, the chromaticity values ​​Cb' and Cr' are rounded to the nearest integer.

[0043] The parameter setting value in this embodiment is: d v =72, d H =44, y v =3, y h =3, w=2560, h=1440, T1=1.21, T2=1.14, T3=0.87, T4=0.83, Y bw =44, Y bz =46, Y bn =48, R bw R is a random number within the range [0, 1]. bz R is a random number within the range [0.4, 1.7]. bn Y is a random number within the range [0.5, 2]. bs =210, Y bx =217, Rbs R is a random number within the range [0.3, 1]. bx Given random numbers within the range [0.4, 1.2], Cb1=142, Cr1=117, Cb2=203, Cr2=197, R fcb R is a random number within the range [0.7, 1.8]. fcr R is a random number within the range [0.7, 2]. scb R is a random number within the range [0.3, 1.3]. scr R is a random number within the range [0.3, 1.2]. xcb R is a random number within the range [0.4, 1.2]. xcr R is a random number within the range [0.3, 1.2]. wcb R is a random number within the range [1, 2]. wcr R is a random number within the range [1, 2]. zcb R is a random number within the range [1, 2, 2]. zcr Let R be a random number within the range [1.1, 2.3]. ncb R is a random number within the range [1.1, 2.2]. ncr A random number within the range [1.2, 2.4].

[0044] The effect of printing denim-style patterns on cotton fabric in this embodiment is as follows: Figure 7 As shown in the figure, the denim texture is generated by printing, and its visual effect is very similar to that of denim fabric generated by textiles. It can be seen that the present invention can effectively transform the pattern into a denim style and flexibly apply it to actual fabric printing.

[0045] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. Those skilled in the art can readily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A method of migrating a costume motif cowboy art style, characterized by, The method comprises the following steps: (1) calculating the warp and weft width according to the pattern image resolution and the actual printing size; (2) performing a denim warp texture stylization processing on the pattern image; (3) performing a denim weft texture stylization processing on the pattern image; (4) performing a denim warp and weft color stylization processing on the pattern image according to the pixel position.

2. The costume motif cowboy art style transfer method of claim 1, wherein, The warp width p in the step (1) is calculated by the following expression v and the weft width p h : ， wherein: d v and d h are the number of warp and weft threads per inch, respectively, y v and y h are the height and width of the pattern image when actually printed, and w and h are the width and height of the pattern image resolution.

3. The costume motif cowboy art style transfer method of claim 2, wherein, The specific implementation of the step (2) is as follows: S21: dividing the pattern image into a plurality of p v ×6p v rectangular blocks of size S22: starting from the second column of the rectangular blocks of the pattern image, each column is moved downward by 2 pixels on the basis of the left column; S23: For any rectangular block, divide the pixels in the block into 3 classes, let the outermost pixels be P w , the next outermost pixels be P z , and the rest of the pixels be P n ; S24: performing a stylization correction on the brightness of each pixel in the pattern image according to the position and category of the pixel to simulate the denim style.

4. The costume motif cowboy art style transfer method of claim 3, wherein, The step S24 performs a stylization correction on the brightness of each pixel in the pattern image through the following relationship: Where: Y(P) w ) and Y'(P w ) represent the pixel P before and after correction. w The brightness value, Y(P) z ) and Y'(P z ) represent the pixel P before and after correction. z The brightness value, Y(P) n ) and Y'(P n ) represent the pixel P before and after correction. n The brightness value, t1 is the preset value, L is the distance of the current pixel in the vertical direction from the center point of the rectangle, and Y is the value of the brightness value. bw Y bz Y bn To simulate the preset values ​​of warp yarn texture in denim fabric, P f P represents the pixels of the flower pattern. b R represents the background pixels. bw R bz R bn To simulate random numbers with a sense of unevenness.

5. The costume motif cowboy art style transfer method of claim 3, wherein, The specific implementation of the step (3) is as follows: S31: divide the pattern image into a plurality of p h ×6p v rectangular blocks of size S32: Starting from the second column of the rectangular blocks of the flower pattern image, each column is shifted down by p on the basis of the left column h + 1 pixel; S33: For any rectangular block, classify the pixels in the block into 3 categories: for the topmost p h x p h square of size p x p, set the pixels outside the square except the four corners to P s , set the pixels inside the square to P x ; and set the rest of the pixels in the block to P y ; S34: Perform style modification on the luminance of the pixel P in the pattern image according to the position and category of the pixel to simulate the cowboy style. s and P x 's luminance to simulate the cowboy style.

6. The costume motif cowboy art style transfer method of claim 5, wherein, The step S34 performs a stylized correction of the luminance of the pixels P in the pattern image by the following relation: s and P x . Where: Y(P) s ) and Y'(P s ) represent the pixel P before and after correction. s The brightness value, Y(P) x ) and Y'(P x ) represent the pixel P before and after correction. x The brightness value, t2 is the preset value, L is the distance of the current pixel in the vertical direction from the center point of the rectangle, and Y is the value of the brightness value. bs and Y bx To simulate the preset values ​​of the weft yarn texture of denim fabric, P f P represents the pixels of the flower pattern. b R represents the background pixels. bs and R bx To simulate random numbers with a sense of unevenness.

7. The costume motif cowboy art style transfer method of claim 5, wherein, The specific implementation of the step (4) is as follows: First, the color of the pixel P of the pattern part in the pattern image is modified by the following relation: f , and the hue is modified by the following relation: Then, for the pixels in the background part of the pattern image, further performing a stylization correction on the chroma of the pixels through the following relationship according to the pixel category: Where: Cb(P) f ) and Cb'(P f ) represent the pixel P before and after correction. f The blueness component value, Cr(P) f ) and Cr'(P f ) represent the pixel P before and after correction. f The redness component value, Cb'(P s ), Cb'(P x ), Cb'(P w ), Cb'(P z ), Cb'(P n ) represent the pixels P in the corrected background area. s P x P w P z P n The blueness component value, Cr'(P s ), Cr'(P x ), Cr'(P w ), Cr'(P z ), Cr'(P n ) represent the pixels P in the corrected background area. s P x P w P z P n The redness component values, t3 and t4 are preset values, R fcb R fcr R scb R scr R xcb R xcr R wcb R wcr R zcb R zcr R ncb and R ncr All are random numbers simulating a sense of unevenness. Cb1 and Cr1 are the chromaticity component values ​​of the warp yarn reference color, and Cb2 and Cr2 are the chromaticity component values ​​of the weft yarn reference color.

8. A computer device comprising a memory and a processor, said memory having stored therein a computer program, characterized in that: The processor is configured to execute the computer program to implement the clothing pattern denim art style migration method according to any one of claims 1-7.

9. A computer readable storage medium storing a computer program, characterized in that: The computer program is configured to be executed by the processor to implement the clothing pattern denim art style migration method according to any one of claims 1-7.

Images