Fabric high-fidelity color-changing soft proofing method and system based on image segmentation

Abstract

The invention relates to a fabric high-fidelity color-changing soft proofing method and system based on image segmentation. The method comprises the following steps: shooting a digital image of a fabric sample; performing color correction on the fabric sample image; decomposing the fabric image into a cartoon part and a texture part; carrying out clustering segmentation on the cartoon part to obtain a fabric image color separation graph; selecting a part to be subjected to color changing from the color separation image, and respectively converting the image to be subjected to color changing and the target color image into a YIQ space and a CIELab space to respectively obtain a color changing image P1 and a color changing image P2; the P1 and the P2 are converted to a Lab space, and L components of the P1 and the P2 are reconstructed to obtain a new L component; combining the new L component with the ab component of the P1, and converting into an RGB space to generate a color changing result of the color separation image; and combining all the color separation images to obtain a final fabric color changing result. According to the method, the influence of fabric textures on segmentation can be effectively reduced, and during color changing, the texture strength can be consistent with that of a real fabric in visual perception while the color is kept close to the real fabric.

Description

Technical Field

[0001] The invention belongs to the technical field of computer digital image processing, and in particular relates to a fabric high-fidelity color-changing soft proofing method and system based on image segmentation. Background Art

[0002] Textile fabrics need to go through a complex process from design to proofing, and designers will adjust patterns and colors according to the proofing effect. It often takes multiple proofing to achieve satisfactory results. This traditional manual proofing method has a long proofing cycle, a large workload, and is prone to waste of resources. Therefore, researchers have explored a variety of digital online color change methods, which are mainly divided into two categories: traditional methods and deep learning methods. The traditional method uses color mapping technology to replace the fabric color with the target color after segmenting the different color regions of the textile fabric. The deep learning method can achieve more refined expression and natural and realistic image color change effects by training a large number of samples. However, this method is highly dependent on samples and requires a large number of sample support. Therefore, the traditional method is more suitable for the field of textile fabric color change.

[0003] However, high-fidelity color change of textile fabrics faces many challenges. First, textile fabrics have complex texture structures, which will affect the segmentation accuracy. Second, some existing color change methods do not fully consider the impact of color on texture perception, resulting in the color-changed image still retaining the characteristics of the original fabric in terms of texture strength, but not reflecting the texture difference that the target color fabric should have. Therefore, although the color of the image after color change has changed, the texture effect is not consistent with the real target color fabric. Summary of the invention

[0004] The purpose of the present invention is to solve the problem described in the background technology and to propose a fabric high-fidelity color-changing soft proofing method based on image segmentation.

[0005] In view of the problems existing in the above-mentioned existing research, the present invention proposes a method for solving the problem. First, a digital camera is used to take a digital image of a fabric sample in an experimental light box, and color correction is performed on it to ensure the accuracy of the color. Then, the fabric image is decomposed into a cartoon part and a texture part using image decomposition technology, and then the cartoon part is clustered and segmented by a mean-shift clustering algorithm to obtain a color separation diagram of the fabric. After selecting the part to be changed in color from the color separation diagram, the image to be changed in color and the target color image are respectively converted to the YIQ space and the CIELab space, and the color change diagrams P1 and P2 are obtained by using the TB method and the Reinhard method respectively. Then, P1 and P2 are converted to the CIELab space, and the L components of P1 and P2 are reconstructed to obtain a new L component. The new L component is combined with the ab component of P1 and converted to the RGB space to generate the color change result of the color separation diagram. Finally, all the color separation diagrams are merged to obtain the final fabric color change result.

[0006] The technical solution of the present invention is a fabric high-fidelity color-changing soft proofing method based on image segmentation, which specifically comprises the following steps:

[0007] Step 1, taking a digital image of a fabric sample;

[0008] Step 2, color correction is performed on the digital image of the fabric sample;

[0009] Step 3, decomposing the color-corrected fabric image into a cartoon part and a texture part;

[0010] Step 4, clustering and segmenting the cartoon part to obtain a color separation map of the fabric image;

[0011] Step 5, select the part to be changed in color from the color separation image, convert the image to be changed in color and the target color image into the YIQ space, and obtain the color change image P1;

[0012] Step 6, convert the image to be color-changed and the target color image into CIELab space, and use the Reinhard method to obtain the color-changed image P2;

[0013] Step 7, convert P1 and P2 to CIELab space, separate the L component into low-frequency and high-frequency components using Gaussian filtering, and reconstruct the low-frequency component of P1 and the high-frequency component of P2 to obtain a new L component;

[0014] Step 8, merge the new L component with the a and b components of P1 and convert them into RGB space to obtain the color separation image color change result;

[0015] Step 9, merging the color-changing result of the color separation image in step 8 with other color separation images into a complete color-changing result image.

[0016] Furthermore, in step 1, a digital camera is used to capture a digital image of the fabric sample in the experimental light box. The specific method is as follows: the digital camera is installed on the top of the experimental light box to form an overall system, the digital camera is connected to a computer, and then the fabric sample is placed in a drawer of the experimental light box, the drawer is closed, the camera shooting parameters are adjusted, and a shooting button is pressed to obtain a digital image of the sample.

[0017] Furthermore, the specific implementation method of color correction of the digital image of the fabric sample in step 2 is as follows:

[0018] Take a digital image of a standard color card in a closed light box, build a color correction model, use a spectrophotometer to get the standard XYZ tristimulus color data C of the color card, and get the standard RGB data S of the color card based on the XYZ to RGB color space conversion matrix M. The formula is as follows:

[0019]

[0020] The standard RGB data S is obtained by matrix multiplication:

[0021] S=MC (2)

[0022] Using the captured color card digital image, extract the RGB data D of each color block, and use the polynomial model to expand the obtained RGB data value matrix. Taking the second-order polynomial model as an example, its expansion form is as follows:

[0023] V=(1,R,G,B,R 2 ,G 2 ,B 2 ,RG,RB,GB) T (3)

[0024] Among them, R, G and B correspond to the values ​​of the three color channel components of the color block respectively, and V is the color value vector after polynomial expansion; according to the expansion matrix V obtained by the color card standard RGB color data S and the RGB data D obtained by the color card shooting, the following mapping relationship is established:

[0025] S=V×Q T (4)

[0026] Among them, Q is the 3×K dimensional color correction coefficient matrix to be determined, K is the number of terms in the polynomial expansion, and the matrix is ​​obtained by the least squares method, that is:

[0027]

[0028] Where J is the number of color blocks in the color card, ∥∥ F is the F norm; the color correction model Q obtained through the above process, the final correction method is as follows:

[0029] I c =IQ T (6)

[0030] Among them, I is the color data of the digital image of the textile fabric, Q is the color correction model, and I c is the corrected digital image color data.

[0031] Furthermore, in step 3, the fabric image is decomposed into a cartoon part and a texture part using image decomposition technology. The specific implementation method is as follows:

[0032] The goal of image decomposition is to transform the image f∈R n Decomposed into corresponding texture part and cartoon part, that is,

[0033] f=u+v (7)

[0034] Among them, f is the original image, u,v∈Rn , u is the cartoon part of the image, which contains the main structure and general outline of the image, and v is the texture part, which contains the detail information and noise in the image; the CLRP model is used to decompose the image, and the specific formula is as follows:

[0035]

[0036] Among them, u is the cartoon part, v is the texture part, b0 is the input image, is the absolute value of the gradient of the cartoon part u, τ and μ are two regularization parameters used to balance the cartoon and texture parts. ∥∥1 is the l1 norm, ∥∥ * is the nuclear norm. ∥∥2 is the L2 norm, which is used to measure the difference between two images.

[0037] Furthermore, in step 4, the mean-shift clustering algorithm is used to cluster and segment the cartoon part. The specific implementation method is as follows:

[0038] Convert the cartoon part from RGB color space to CIELab color space, apply mean-shift algorithm in CIELab color space, cluster the pixels of the cartoon part, and get the color segmentation result. i , calculate its mean shift vector:

[0039]

[0040] Among them, x i represents the position of the i-th data point, S = {x j ∈X|||x i -x j ∥≤h} represents a neighborhood set, including all distances x i Points that do not exceed the bandwidth parameter h are used to calculate the mean shift vector, X is the set of all pixel points in the image, K h (‖x i -x j ‖) is a Gaussian kernel function, which is used to calculate the weights between data points. The formula is as follows:

[0041]

[0042] Where h is the bandwidth parameter; the algorithm iteratively updates the position of each data point until the moving distance of all data points is less than the set threshold:

[0043] x i t+1 =x i t +m(x i ) (11)

[0044] Where t represents the number of iterations, and the updated position is the new position of the data point after the offset. By adjusting the bandwidth parameter h, the fineness of the segmentation can be controlled to adapt to different image characteristics. Finally, similar cluster merging and morphological operations are used to obtain the final fabric image color separation map.

[0045] Furthermore, the specific implementation method in step 5 is as follows:

[0046] First, convert the image to the YIQ color space and extract the Y component Y of the image to be changed. t (x,y) and the Y component Y of the target color image c (x, y), calculate the Y component deviation of the image to be changed, as shown in formula (12):

[0047]

[0048] Among them, the YIQ color space represents a color image by decomposing the color information into brightness Y and chrominance IQ components. is the average value of the Y component of the image to be changed; then, the deviation is superimposed on the Y component of the target color image to generate a new Y component:

[0049] Y′(x,y)=ΔY t (x,y)+Y c (x,y) (13)

[0050] Combined with the IQ components of the target color image, the output image P1 is obtained by converting it into the RGB color space after fusion.

[0051] Furthermore, the specific implementation method of obtaining the color-changing graph P2 using the Reinhard method in step 6 is as follows:

[0052] First, the image to be changed and the target color image are converted to the CIELab color space, denoted as L t , a t , b t and L c , a c , b c , respectively calculate the mean and standard deviation of the Lab channel of the image to be color-changed and the target color image. Then, adjust the Lab channel of the image to be color-changed according to the formula so that its mean and standard deviation match the target color image:

[0053]

[0054] in, are the mean and standard deviation of the L, a, and b channels of the image to be changed, respectively. is the mean and standard deviation of the target color image L, a, b channels; Lr ,a r ,b r To adjust the image, convert it to RGB space to get the output image P2.

[0055] Furthermore, the specific implementation method in step 7 is as follows:

[0056] Convert images P1 and P2 to CIELab color space, obtain their brightness components L1 and L2 respectively, decompose the brightness components into low-frequency and high-frequency components through Gaussian filter, use the low-frequency component of P1 and the high-frequency component of P2 for brightness fusion, and obtain new brightness components:

[0057]

[0058] Among them, L(x,y) is the newly reconstructed brightness component, is the low-frequency component of the L component of P1, is the high frequency component of the L component of P2, and x and y are the horizontal and vertical coordinates respectively.

[0059] Furthermore, the specific implementation method of fusing the new L component in step 8 with the a and b components of P1 is as follows:

[0060] P(x,y)=Lab(L(x,y),a1(x,y),b1(x,y)) (16)

[0061] Among them, L(x,y) is the reconstructed L component, a1(x,y) and b1(x,y) are the a and b components of P1 respectively, P(x,y) is the fusion result, and x and y are the horizontal and vertical coordinates respectively.

[0062] The present invention also provides a fabric high-fidelity color-changing soft proofing system based on image segmentation, comprising:

[0063] A processor and a memory, the memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute a high-fidelity color-changing soft proofing method for fabrics based on image segmentation as described in the above technical solution.

[0064] The present invention aims at the problem that the existing color-changing methods for textile fabrics do not fully consider the interaction between color and texture, resulting in that although the color of the image after color change has changed, the texture strength is inconsistent with the real target color fabric. After color correction of the fabric image, the image is divided into a cartoon part and a texture part using image decomposition technology, and the cartoon part is segmented using a mean-shift clustering algorithm. The TB and Reinhard methods are applied to the image to be color-changed to generate a color-changing image, and then the L component is separated and reconstructed, and the new L component and the ab component are reconstructed to obtain the color-changing result. Finally, the color separation images are merged to obtain the final fabric color-changing result. The present invention can effectively reduce the influence of fabric texture on segmentation, and when changing colors, while keeping the color close to the real fabric, the texture strength can also achieve visual perception consistency with the real fabric. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 is a flow chart of an embodiment of the present invention.

[0066] Figure 2 This is the experimental light box used in the embodiments of the present invention.

[0067] Figure 3 This is the color change result of the textile sample according to the embodiment of the present invention. DETAILED DESCRIPTION

[0068] When the technical solution of the present invention is specifically implemented, it can be operated by those skilled in the art using computer software technology.

[0069] like Figure 1 As shown, the embodiment of the present invention proposes a fabric high-fidelity color change soft proofing method based on image segmentation, which specifically includes the following steps:

[0070] Step 1, using a digital camera to take a digital image of a fabric sample in an experimental light box;

[0071] Step 2, color correction of the fabric sample image;

[0072] Step 3, decomposing the fabric image into a cartoon part and a texture part using image decomposition technology;

[0073] Step 4, using mean-shift clustering algorithm to cluster and segment the cartoon part to obtain a color separation map of the fabric image;

[0074] Step 5, select the part to be changed in color from the color separation image, convert the image to be changed in color and the target color image into the YIQ space, and obtain the color change image P1;

[0075] Step 6, convert the image to be color-changed and the target color image into CIELab space, and use the Reinhard method to obtain the color-changed image P2;

[0076] Step 7, convert P1 and P2 to CIELab space, separate the L component into low-frequency and high-frequency components using Gaussian filtering, and reconstruct the low-frequency component of P1 and the high-frequency component of P2 to obtain a new L component;

[0077] Step 8, merge the new L component with the a and b components of P1 and convert them into RGB space to obtain the color separation image color change result;

[0078] Step 9, merging the color-changing result of the color separation image in step 8 with other color separation images into a complete color-changing result image.

[0079] The following is a specific example to illustrate the embodiment of the present invention.

[0080] Embodiments The method of the present invention was tested based on a self-developed enclosed daylight lighting box, a Nikon D7200 digital camera, a ColorChecker SG140 color card, a ColorChecker 24 color card and textile fabrics.

[0081] In step 1, the specific process of using a digital camera to capture a digital image of a fabric sample in an experimental light box in the embodiment is as follows: install the digital camera on the top of the experimental light box to form an integrated system, connect the digital camera to the computer, and then place the fabric sample in the effective photographing area of ​​the experimental light box, close the drawer, adjust the camera shooting parameters, and press the shooting button to obtain the sample digital image. The experimental light box is illuminated by a full-spectrum daylight light source, which can meet the requirements of photographic color measurement for full-spectrum visible light lighting conditions; at the same time, the light box is a closed light box, which is not affected by the external lighting environment, and can ensure the uniformity of lighting in the imaging area at the bottom of the light box; and the light box adopts completely diffuse reflection lighting, which can effectively avoid the shortcomings of direct lighting of existing light boxes. Then, the effective area of ​​the fabric in the captured image is cropped. The experimental light box used in the present invention is as shown in the attached Figure 2 shown.

[0082] In step 2, a digital image of a standard color card is taken in a closed light box to construct a color correction model. The standard XYZ tristimulus value color data C of the color card is obtained using a spectrophotometer, and the standard RGB data S of the color card is obtained based on the XYZ to RGB color space conversion matrix M, as shown in the following formula:

[0083]

[0084] The standard RGB data S is obtained by matrix multiplication:

[0085] S=MC (2)

[0086] Using the captured color card digital image, extract the RGB data D of each color block, and use the polynomial model to expand the obtained RGB data value matrix. Taking the second-order polynomial model as an example, its expansion form is as follows:

[0087] V=(1,R,G,B,R 2 ,G 2 ,B 2 ,RG,RB,GB) T (3)

[0088] Among them, R, G and B correspond to the values ​​of the three color channel components of the color block respectively, and V is the color value vector after polynomial expansion. According to the expansion matrix V obtained by the color card standard RGB color data S and the RGB data D obtained by color card shooting, the following mapping relationship is established:

[0089] S=V×Q T (4)

[0090] Among them, Q is the 3×K dimensional color correction coefficient matrix to be determined, K is the number of terms in the polynomial expansion, and the matrix is ​​obtained by the least squares method, that is:

[0091]

[0092] Where J is the number of color blocks in the color card, ∥∥ F is the F norm; the color correction model Q obtained through the above process can be used to perform color correction on digital images of textile fabrics obtained under the same shooting conditions, further improving the color accuracy. The final correction method is as follows:

[0093] I c =IQ T (6)

[0094] Among them, I is the digital image color data of the textile fabric, Q is the color correction model, and I c is the corrected digital image color data.

[0095] In the embodiment, digital images of the ColorChecker SG140 color card and the ColorChecker 24 color card are captured in a closed light box, the ColorChecker SG140 color card is used as a training sample, and the ColorChecker 24 color card is used as a test sample, the RGB data of all pixels in the m×m pixel area at the center of each color block are extracted, and the RGB data of the m×m pixels are averaged, and a third-order polynomial expansion is used to obtain a color vector V.

[0096] In step 3, the specific implementation method of decomposing the fabric image into a cartoon part and a texture part using image decomposition technology is as follows:

[0097] The goal of image decomposition is to transform the image f∈R n Decomposed into corresponding texture part and cartoon part, that is,

[0098] f=u+v (7)

[0099] Among them, f is the original image, u,v∈R n , u is the cartoon part of the image, which contains the main structure and general outline of the image, and v is the texture part, which contains the details and noise in the image. The CLRP model (customized low-rank prior model) is used to decompose the fabric image into the cartoon part and the texture part. The specific formula is as follows:

[0100]

[0101] Among them, u is the cartoon part, v is the texture part, b0 is the input image, is the absolute value of the gradient of the cartoon part u, τ and μ are two regularization parameters used to balance the cartoon and texture parts. ∥∥1 is the l1 norm, ∥∥ * is the nuclear norm. ∥∥2 is the L2 norm, which is used to measure the difference between two images. is the data fidelity term, ensuring that the decomposed image is as close to the original image as possible. From formula (8), we can see that this is an optimization problem, and the goal is to minimize this objective function and solve the variables u and v.

[0102] In step 4, the cartoon part of the fabric image is converted from the RGB color space to the CIELab color space, the mean-shift algorithm is applied in the CIELab color space to cluster the cartoon part pixels, and then similar cluster merging and morphological operations are used to obtain the final fabric image color separation map. The specific implementation method is as follows:

[0103] Convert the cartoon part from RGB color space to CIELab color space, apply mean-shift algorithm in CIELab color space, cluster the pixels of the cartoon part, and get the color segmentation result. i , calculate its mean shift vector:

[0104]

[0105] Among them, x i represents the position of the i-th data point, S = {x j ∈X||x i -x j ∥≤h} represents a neighborhood set, including all distances x iPoints that do not exceed the bandwidth parameter h are used to calculate the mean shift vector, X is the set of all pixel points in the image, K h (||x i -x j ||) is a Gaussian kernel function, which is used to calculate the weights between data points. The formula is as follows:

[0106]

[0107] Where h is the bandwidth parameter. The algorithm iteratively updates the position of each data point until the moving distance of all data points is less than the set threshold:

[0108] x i t+1 =x i t +m(x i ) (11)

[0109] Where t represents the number of iterations, and the updated position is the new position of the data point after the offset. By adjusting the bandwidth parameter h, the fineness of the segmentation can be controlled to adapt to different image characteristics. Finally, similar cluster merging and morphological operations are used to obtain the final fabric image color separation map.

[0110] In step 5, the image is converted to the YIQ color space, the Y component Yt(x, y) of the image to be changed and the Y component Yc(x, y) of the target color image are extracted, and the Y component deviation of the image to be changed is calculated:

[0111]

[0112] Among them, the YIQ color space represents color images by decomposing color information into brightness (Y) and chrominance (IQ) components. Y represents brightness, I represents In-phase, and the colors range from orange to cyan, and Q represents Quadrature-phase, and the colors range from purple to yellow-green. is the average value of the Y component of the image to be changed. Then, the deviation is added to the Y component of the target color image to generate a new Y component:

[0113] Y′(x,y)=ΔY t (x,y)+Y c (x,y) (13)

[0114] Combined with the IQ components of the target color image, the output image P1 is obtained by converting it into the RGB color space after fusion.

[0115] In step 6, the image to be changed and the target color image are converted to the CIELab color space, denoted as L t , a t , bt and L c , a c , b c , respectively calculate the mean and standard deviation of the image to be color-changed and the target color image in the Lab channel, and then adjust the Lab channel of the image to be color-changed according to the formula:

[0116]

[0117]

[0118] in, are the mean and standard deviation of the L, a, and b channels of the image to be changed, respectively. is the mean and standard deviation of the target color image L, a, b channels. r ,a r ,b r To adjust the image, convert it to RGB space to get the output image P2.

[0119] In step 7, images P1 and P2 are converted to CIELab color space to obtain their brightness components L1 and L2 respectively. The brightness component is decomposed into low-frequency and high-frequency components by Gaussian filter, and the low-frequency component of P1 and the high-frequency component of P2 are used for brightness fusion to obtain a new brightness component:

[0120]

[0121] Among them, L(x,y) is the reconstructed new brightness component, is the low-frequency component of the L component of P1, is the high frequency component of the L component of P2, and x and y are the horizontal and vertical coordinates respectively.

[0122] In step 8, the specific implementation method of fusing the new L component in step 8 with the a and b components of P1 is as follows:

[0123] P(x,y)=Lab(L(x,y),a1(x,y),b1(x,y)) (16)

[0124] Among them, L(x,y) is the reconstructed L component, a1(x,y) and b1(x,y) are the a and b components of P1 respectively, P(x,y) is the fusion result, and x and y are the horizontal and vertical coordinates respectively.

[0125] On the other hand, an embodiment of the present invention provides a fabric high-fidelity color-changing soft proofing system based on image segmentation, comprising:

[0126] A processor and a memory, the memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute a high-fidelity color-changing soft proofing method for fabrics based on image segmentation as described in the above technical solution.

[0127] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A fabric high-fidelity color-changing soft proofing method based on image segmentation, characterized in that: The steps include: Step 1, taking a digital image of a fabric sample; Step 2, color correction is performed on the digital image of the fabric sample; Step 3, decomposing the color-corrected fabric image into a cartoon part and a texture part; Step 4, clustering and segmenting the cartoon part to obtain a color separation map of the fabric image; Step 5, select the part to be changed in color from the color separation image, convert the image to be changed in color and the target color image into the YIQ space, and obtain the color change image P1; Step 6, convert the image to be color-changed and the target color image into CIELab space, and use the Reinhard method to obtain the color-changed image P2; Step 7, convert P1 and P2 to CIELab space, separate the L component into low-frequency and high-frequency components using Gaussian filtering, and reconstruct the low-frequency component of P1 and the high-frequency component of P2 to obtain a new L component; Step 8, merge the new L component with the a and b components of P1 and convert them into RGB space to obtain the color change result of the color separation image; Step 9, merging the color-changing result of the color separation image in step 8 with other color separation images into a complete color-changing result image.

2. The method for high-fidelity color change soft proofing of fabric based on image segmentation as claimed in claim 1, characterized in that: In step 1, a digital camera is used to capture a digital image of the fabric sample in the experimental light box. The specific method is as follows: install the digital camera on the top of the experimental light box to form an overall system, connect the digital camera to the computer, and then place the fabric sample in the drawer of the experimental light box, close the drawer, adjust the camera shooting parameters, and press the shooting button to obtain the sample digital image.

3. The method for high-fidelity color-changing soft proofing of fabric based on image segmentation as claimed in claim 1, characterized in that: The specific implementation method of color correction of the digital image of the fabric sample in step 2 is as follows: Take a digital image of a standard color card in a closed light box, build a color correction model, use a spectrophotometer to get the standard XYZ tristimulus color data C of the color card, and get the standard RGB data S of the color card based on the XYZ to RGB color space conversion matrix M. The formula is as follows: The standard RGB data S is obtained by matrix multiplication: S=MC (2) Using the captured color card digital image, extract the RGB data D of each color block, and use the polynomial model to expand the obtained RGB data value matrix. Taking the second-order polynomial model as an example, its expansion form is as follows: V=(1,R,G,B,R 2 ,G 2 ,B 2 ,RG,RB,GB) T (3) Among them, R, G and B correspond to the values ​​of the three color channel components of the color block respectively, and V is the color value vector after polynomial expansion; according to the expansion matrix V obtained by the color card standard RGB color data S and the RGB data D obtained by the color card shooting, the following mapping relationship is established: S=V×Q T (4) Among them, Q is the 3×K dimensional color correction coefficient matrix to be determined, K is the number of terms in the polynomial expansion, and the matrix is ​​obtained by the least squares method, that is: Where J is the number of color blocks in the color card, ∥∥ F is the F norm; the color correction model Q is obtained by formula (5), and the final correction method is as follows: I c =IQ T (6) Among them, I is the digital image color data of the textile fabric, Q is the color correction model, and I c is the corrected digital image color data.

4. The method for high-fidelity color-changing soft proofing of fabric based on image segmentation as claimed in claim 1, characterized in that: In step 3, the fabric image is decomposed into a cartoon part and a texture part using image decomposition technology. The specific implementation method is as follows: The goal of image decomposition is to transform the image f∈R n Decomposed into corresponding texture part and cartoon part, that is, f=u+v (7) Among them, f is the original image, u,v∈R n , u is the cartoon part of the image, which contains the main structure and general outline of the image, and v is the texture part, which contains the detail information and noise in the image; the CLRP model is used to decompose the image, and the specific formula is as follows: Among them, τ and μ are two regularization parameters used to balance the cartoon and texture parts, b0 is the input image, is the absolute value of the gradient of the cartoon part u, ∥∥1 is the l1 norm, ∥∥ * is the nuclear norm, and ∥∥2 is the L2 norm, which is used to measure the difference between two images.

5. The method for high-fidelity color-changing soft proofing of fabric based on image segmentation as claimed in claim 1, characterized in that: In step 4, the mean-shift clustering algorithm is used to cluster and segment the cartoon part. The specific implementation method is as follows: Convert the cartoon part from RGB color space to CIELab color space, apply mean-shift algorithm in CIELab color space, cluster the pixels of the cartoon part, and get the color segmentation result. i , calculate its mean shift vector: Among them, x i represents the position of the i-th data point, S = {x j ∈X|||x i -x j ∥≤h} represents a neighborhood set, including all distances x i Points that do not exceed the bandwidth parameter h are used to calculate the mean shift vector, X is the set of all pixel points in the image, K h (||x i -x j ||) is a Gaussian kernel function, which is used to calculate the weights between data points. The formula is as follows: Where h is the bandwidth parameter; the algorithm iteratively updates the position of each data point until the moving distance of all data points is less than the set threshold: x i t+1 =x i t +m(x i ) (11) Where t represents the number of iterations, and the updated position is the new position of the data point after the offset. By adjusting the bandwidth parameter h, the fineness of the segmentation can be controlled to adapt to different image characteristics. Finally, similar cluster merging and morphological operations are used to obtain the final fabric image color separation map.

6. The method for high-fidelity color-changing soft proofing of fabric based on image segmentation as claimed in claim 1, characterized in that: The specific implementation method in step 5 is as follows: First, convert the image to the YIQ color space and extract the Y component Y of the image to be changed. t (x,y) and the Y component Y of the target color image c (x, y), calculate the Y component deviation of the image to be changed, as shown in formula (12): Among them, the YIQ color space represents a color image by decomposing the color information into brightness Y and chrominance IQ components. is the average value of the Y component of the image to be changed; then, the deviation is superimposed on the Y component of the target color image to generate a new Y component: Y′(x,y)=ΔY t (x,y)+Y c (x,y) (13) Combined with the IQ components of the target color image, the output image P1 is obtained by converting it into the RGB color space after fusion.

7. The method for high-fidelity color-changing soft proofing of fabric based on image segmentation as claimed in claim 1, characterized in that: The specific implementation method of obtaining the color-changing graph P2 using the Reinhard method in step 6 is as follows: First, the image to be changed and the target color image are converted to the CIELab color space, denoted as L t , a t , b t and L c , a c , b c , respectively calculate the mean and standard deviation of the Lab channel of the image to be color-changed and the target color image. Then, adjust the Lab channel of the image to be color-changed according to the formula so that its mean and standard deviation match the target color image: in, are the mean and standard deviation of the L, a, and b channels of the image to be changed, respectively. is the mean and standard deviation of the target color image L, a, b channels; L r ,a r ,b r To adjust the image, convert it to RGB space to get the output image P2.

8. The method for high-fidelity color-changing soft proofing of fabric based on image segmentation as claimed in claim 1, characterized in that: The specific implementation method in step 7 is as follows: Convert images P1 and P2 to Lab color space, obtain their brightness components L1 and L2 respectively, decompose the brightness components into low-frequency and high-frequency components through Gaussian filter, use the low-frequency component of P1 and the high-frequency component of P2 for brightness fusion, and obtain new brightness components: Among them, L(x,y) is the newly reconstructed brightness component, is the low-frequency component of the L component of P1, is the high frequency component of the L component of P2, and x and y are the horizontal and vertical coordinates respectively.

9. The method for high-fidelity color-changing soft proofing of fabric based on image segmentation as claimed in claim 1, characterized in that: The specific implementation method of fusing the new L component with the a and b components of P1 in step 8 is as follows: P(x,y)=Lab(L(x,y),a1(x,y),b1(x,y)) (16) Among them, L(x,y) is the reconstructed L component, a1(x,y) and b1(x,y) are the a and b components of P1 respectively, P(x,y) is the fusion result, and x and y are the horizontal and vertical coordinates respectively.

10. A fabric high-fidelity color-changing soft proofing system based on image segmentation, characterized in that: include: A processor and a memory, the memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute a high-fidelity color-changing soft proofing method for fabrics based on image segmentation as described in any one of claims 1 to 9.

Images