Cloth color difference comparison method and system based on color matching lamp box

Abstract

The invention discloses a cloth color difference comparison method and system based on a color matching lamp box, and relates to the technical field of image recognition, and the method comprises the following steps: S1, shooting a sample image of a sample under a preset color matching condition; s2, analyzing the sample image, and judging a chromatic aberration comparison value of the sample through a gray value to obtain the chromatic aberration comparison value of the sample; s3, when the color matching condition is changed, the sample image is shot again and analyzed, and the color difference change ratio between the samples before and after the preset color matching condition is changed is obtained; and S4, executing the steps S1-S3 on a plurality of samples to obtain a plurality of color difference change ratios of different samples under different color matching conditions. According to the invention, based on image identification, comprehensive quantitative evaluation of the cloth color difference is realized, the problem that the color difference relation is difficult to quantify when the color matching environment of the cloth is changeable is effectively solved, and the accuracy and reliability of cloth color difference evaluation are greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and more specifically, to a method and system for comparing color difference of fabrics based on a color matching light box. Background Art

[0002] A color matching light box is an artificial lighting device, which is used to place standard samples for testing. Specifically, it is used to display standard samples in the fields of textile production, clothing design, printing and publishing, textile print testing, textile, clothing display and textile advertising, sample products, display and testing of finished products, display of textile color styles and textile design communication to ensure that the color provided as a sample can reproduce its original color at any time, and can reproduce its original color during the design, production, inspection, management or display of products such as textiles, clothing, printed materials, packaging materials, indoor goods and industrial products.

[0003] Color difference testing refers to the process of comparing the color of the product to be tested with its standard color sample using a colorimeter or visual inspection to assess the degree of color difference. In the textile industry, the usual way to quantify color is to compare the textile with a small sample with a color code and mark the color difference level.

[0004] However, in specific applications, the color matching environment of fabrics is changeable (such as different light source color temperatures), so it is necessary to quantify the relationship between fabric color difference changes under different light source color temperatures and its color difference relationship with standard fabrics. Therefore, a method for quantifying fabric color difference changes is needed to analyze the color difference relationship between samples when the color matching conditions change under color matching conditions. Provide reliable quantitative data for subsequent process settings and process control. However, there is currently no reliable technology that can meet this requirement. In view of this, we propose a fabric color difference comparison method and system based on a color matching light box. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for comparing fabric color difference based on a color matching light box. The technical problem to be solved is to quantify the relationship between fabric color difference changes and its color difference relationship with standard fabric under different light source color temperature conditions, analyze the color difference relationship between samples when the color matching conditions change, and provide reliable quantitative data for subsequent process settings and process control.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for comparing color difference of fabrics based on a color matching light box, comprising the following steps: S1: taking a sample image of the sample under the preset color matching conditions; S2: Analyze the sample image, determine the color difference comparison value of the sample by the gray value, and obtain the color difference comparison value of the sample; S3: When the color matching conditions are changed, re-shoot the sample images and analyze them to obtain the color difference change ratio between the samples before and after the preset color matching conditions are changed; S4: Execute steps S1-S3 with multiple samples to obtain multiple color difference change ratios between different samples under different color matching conditions; The present invention realizes a comprehensive quantitative evaluation of fabric color difference by shooting sample images under preset color matching conditions and analyzing color difference comparison values, and reshooting images to analyze color difference change ratios when the color matching conditions change. In actual operation, multiple groups of sample images can be shot under preset color matching conditions according to specified steps, and then processed by an analysis module to accurately obtain color difference data under different conditions, thereby effectively solving the problem that it is difficult to quantify the color difference relationship under different light source color temperature conditions and provide data for process control when the fabric color matching environment is changeable, greatly improving the accuracy and reliability of fabric color difference evaluation, providing reliable quantitative data for subsequent process settings and process control, and filling the gap in the prior art in this regard.

[0007] Wherein, the step S2 includes the following sub-steps: S201: photographing the boundary positions of different color areas in the sample image to obtain a color block segmentation map; S202: Convert the color blocks in the color block segmentation image into RGB values; S203: Analyze the color difference comparison value of the sample image, including the color difference comparison value of the color block and the overall color difference comparison value of the sample image.

[0008] Preferably, the sub-step S201 includes the following sub-steps: S201a, identifying different color areas of the sample image; Among them, the color clustering algorithm is used to identify different color areas of the sample image and minimize the objective function: ; In the formula, is the number of clusters, It's a pixel. It is A set of clusters, It is The center of the cluster; S201b, performing image segmentation on the sample image by using a grayscale segmentation method; S201c, using a boundary recognition algorithm to capture the boundary positions of different color blocks after image segmentation, and obtain a color block segmentation map.

[0009] Preferably, the step S201b includes the following sub-steps: A1. Grayscale histogram of the sample image , where represents the gray level, Indicates grayscale The corresponding number of pixels; A2. According to the preset segmentation threshold , segment the sample image; A3, by calculating the intra-class variance and inter-class variance of the image after segmentation at each segmentation threshold, the optimal segmentation threshold is obtained; A4. Segment the sample image with the optimal segmentation threshold to obtain a color block segmentation map of the color image.

[0010] Preferably, the step A3 includes the following sub-steps: A301, calculate the segmentation interval set: Each segmentation threshold , the corresponding between-class variance is ; Among them, the between-class variance is expressed by the formula Calculate, where , is the ratio of the two types of pixels, , is the average gray value of the two types of pixels, is the average gray value of the entire image; A302. Calculate each segmentation threshold The corresponding intra-class variance of the image; Among them, the intra-class variance is expressed by the formula Calculate, where , are the variances within the two types of pixels; A303, taking the maximum value among the intra-class variances corresponding to each segmentation threshold, and calculating the corresponding inter-class variance; A304. Take the maximum value of the inter-class variance as the optimal segmentation threshold.

[0011] Preferably, the step S203 includes the following sub-steps: S203a: Acquire a standard image of the sample; S203b: Analyze the color difference between the sample image and the standard image to obtain a color difference ratio of a single color block; Among them, the color difference ratio of a single color block is calculated by the color difference formula Calculate, where , , They are the differences of brightness, red-green axis, and yellow-blue axis in CIELab color space. The color difference between the sample image and the standard image is analyzed to obtain the color difference ratio of a single color block. S203c: Sum the color difference ratios of the individual color blocks to obtain the overall color difference ratio of the sample image.

[0012] Preferably, in step S3, the position of the sample and the incident angle of the light are the same before and after the color matching conditions are changed.

[0013] Preferably, in step S3, steps S1-S2 are repeatedly performed to obtain the color difference comparison values ​​of the samples under different color matching conditions. , where , They are the color difference comparison values ​​before and after the change respectively.

[0014] A fabric color difference comparison system based on a color matching light box comprises an image acquisition module and an analysis module, wherein the image acquisition module is used to capture a sample image of a sample under a preset color matching condition, and the analysis module is used to analyze the sample image to obtain a color difference comparison value of the sample; Preferably, the image acquisition module comprises a colorimeter and a light source, the colorimeter is used to capture the sample image, and the light source is used to provide preset color matching conditions.

[0015] Preferably, the analysis module includes a segmentation submodule, a conversion submodule and an analysis submodule; The segmentation submodule is used to capture the boundary positions of different color areas in the sample image to obtain a color block segmentation map; The conversion submodule is used to convert the color blocks in the color block segmentation image into RGB values; The analysis submodule is used to analyze the color difference comparison value of the sample image, including the color difference comparison value of the color block and the overall color difference comparison value of the sample image.

[0016] Preferably, the segmentation submodule includes a region recognition submodule, an image segmentation submodule and a boundary recognition submodule; The region recognition submodule is used to recognize different color regions of the sample image; The image segmentation submodule is used to segment the sample image by grayscale segmentation method; The boundary recognition submodule is used to capture the boundary positions of different color blocks after image segmentation to obtain a color block segmentation map.

[0017] Preferably, the image segmentation submodule includes a test image acquisition submodule, a calculation submodule and an optimal segmentation threshold submodule; The test image acquisition submodule is used to capture the grayscale histogram of the sample image; The calculation submodule is used to segment the sample image according to a preset segmentation threshold; The optimal segmentation threshold submodule is used to obtain the optimal segmentation threshold by calculating the intra-class variance and inter-class variance of the image after segmentation by each segmentation threshold; The optimal segmentation threshold submodule is also used to segment the sample image with the optimal segmentation threshold to obtain a color block segmentation map of the color image.

[0018] Preferably, the analysis submodule includes a standard image acquisition submodule, a single color block comparison submodule and a sample image comparison submodule; The standard image acquisition submodule is used to acquire a standard image of the sample; The single color block comparison submodule is used to analyze the color difference between the sample image and the standard image to obtain the color difference ratio of the single color block; The sample image comparison submodule is used to sum the color difference ratios of individual color blocks to obtain the overall color difference ratio of the sample image.

[0019] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention realizes a comprehensive quantitative evaluation of fabric color difference by shooting sample images under preset color matching conditions and analyzing color difference comparison values, and reshooting images to analyze color difference change ratios when the color matching conditions change. In actual operation, multiple groups of sample images can be shot under preset color matching conditions according to specified steps, and then processed by the analysis module to accurately obtain color difference data under different conditions, thereby effectively solving the problem that it is difficult to quantify the color difference relationship under different light source color temperature conditions and provide data for process control when the fabric color matching environment is changeable, greatly improving the accuracy and reliability of fabric color difference evaluation, providing reliable quantitative data for subsequent process settings and process control, and filling the gap in the prior art in this regard.

[0020] 2. The coordinated work of the image acquisition module and the analysis module in the system of the present invention realizes a highly automated and intelligent operation process. The colorimeter captures the sample image under specific light source conditions, and the analysis module processes it according to the image features. For different types of images, color analysis or texture recognition can be used respectively, which not only improves the efficiency of color difference comparison, but also further enhances the accuracy of evaluation, better adapts to the complex and diverse fabric detection needs, and ensures that accurate color difference comparison results can be stably obtained in various practical scenarios.

[0021] 3. The system of the present invention is closely combined with actual application needs through strict sample selection standards and flexible color matching condition settings. In actual operation, suitable samples can be selected according to different fabric characteristics to simulate various real color matching environments, ensuring that the comparison results are highly consistent with the actual production situation, so that the system can not only accurately evaluate the color difference, but also flexibly adjust the detection strategy according to the changes in different processes and color matching conditions, providing more targeted and reliable color difference control solutions for fabrics in multiple links such as design, production, and inspection, effectively improving the quality and efficiency of the entire fabric production and testing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic block diagram of the system of the present invention. DETAILED DESCRIPTION

[0023] Embodiment 1: The present invention relates to a method for comparing color difference of fabrics based on a color matching light box, comprising the following steps: S1: taking a sample image of the sample under the preset color matching conditions; S2: Analyze the sample image to obtain the color difference comparison value of the sample; S3: When the color matching conditions are changed, re-shoot the sample images and analyze them to obtain the color difference change ratio between the samples before and after the preset color matching conditions are changed; S4: Execute steps S1-S3 with multiple samples to obtain multiple color difference change ratios between different samples under different color matching conditions.

[0024] Specifically, by photographing the sample under preset color matching conditions, a sample image of the sample is obtained, and the sample image is analyzed to obtain the color difference comparison value of the sample. When the comparison between the samples is completed, the color matching conditions are changed to obtain the color difference change ratio between the samples before and after changing the preset color matching conditions. The above process is repeated to finally obtain multiple color difference change ratios between different samples under different color matching conditions.

[0025] Among them, multiple samples are included, and multiple color difference change ratios are obtained by comparing the color differences of multiple samples, which is more scientific.

[0026] In an embodiment of the present invention, step S2 includes the following sub-steps: S201: photographing the boundary positions of different color areas in the sample image to obtain a color block segmentation map; In an embodiment of the present invention, the sub-step S201 includes the following sub-steps: S201a, identifying different color areas of the sample image; Among them, the color clustering algorithm is used to identify different color areas of the sample image and minimize the objective function: ; In the formula, is the number of clusters, It's a pixel. It is A set of clusters, It is The center of the cluster; S201b, performing image segmentation on the sample image by using a grayscale segmentation method; In an embodiment of the present invention, step S201b includes the following sub-steps: A1. Grayscale histogram of the sample image , where represents the gray level, Indicates grayscale The corresponding number of pixels; A2. According to the preset segmentation threshold , segment the sample image; A3, by calculating the intra-class variance and inter-class variance of the image after segmentation at each segmentation threshold, the optimal segmentation threshold is obtained; In an embodiment of the present invention, step A3 includes the following sub-steps: A301, calculate the segmentation interval set: Each segmentation threshold , the corresponding between-class variance is ; Among them, the between-class variance is expressed by the formula Calculate, where , is the ratio of the two types of pixels, , is the average gray value of the two types of pixels, is the average gray value of the entire image; A302. Calculate each segmentation threshold The corresponding intra-class variance of the image; Among them, the intra-class variance is expressed by the formula Calculate, where , are the variances within the two types of pixels; A303, taking the maximum value among the intra-class variances corresponding to each segmentation threshold, and calculating the corresponding inter-class variance; A304. Take the maximum value of the inter-class variance as the optimal segmentation threshold.

[0027] Specifically, through the above method, the image is segmented to obtain the optimal segmentation threshold and obtain a color block segmentation map.

[0028] A4. Segment the sample image using the optimal segmentation threshold to obtain a color block segmentation map of the color image.

[0029] Specifically, by segmenting the sample image, the sample image is converted into a color block segmentation map, which is more convenient for analysis.

[0030] S201c, using a boundary recognition algorithm to capture the boundary positions of different color blocks after image segmentation, and obtain a color block segmentation map.

[0031] Specifically, different color regions may have different color differences, so the sample image is first converted into a color block segmentation map, and then the color block segmentation map is processed and analyzed.

[0032] S202: Convert the color blocks in the color block segmentation image into RGB values; S203: Analyze the color difference comparison value of the sample image, including the color difference comparison value of the color block and the overall color difference comparison value of the sample image.

[0033] Specifically, the sample image may be a monochrome image or an image composed of different color blocks. The sample image is converted into RGB values ​​to facilitate analysis of the color block color difference comparison value and the overall color difference comparison value of the sample image.

[0034] In an embodiment of the present invention, step S203 includes the following sub-steps: S203a: Acquire a standard image of the sample; S203b: Analyze the color difference between the sample image and the standard image to obtain a color difference ratio of a single color block; Among them, the color difference ratio of a single color block is calculated by the color difference formula Calculate, where , , They are the differences of brightness, red-green axis, and yellow-blue axis in CIELab color space. The color difference between the sample image and the standard image is analyzed to obtain the color difference ratio of a single color block. S203c: Sum the color difference ratios of the individual color blocks to obtain the overall color difference ratio of the sample image.

[0035] Specifically, the above method facilitates the analysis of the color difference comparison value of the color block and the overall color difference comparison value of the sample image, thereby avoiding misjudgment caused by naked eye judgment.

[0036] In an embodiment of the present invention, in step S3, the position of the sample and the incident light angle are the same before and after the color matching conditions are changed. By changing the color matching conditions and the position of the sample and the incident light angle, multiple color difference comparisons are made more scientific.

[0037] In an embodiment of the present invention, when the color matching conditions change, the color matching conditions of the image acquisition module change, and the color difference change ratio between the samples before and after the preset color matching conditions are changed is obtained. The new sample image is obtained after the color matching conditions are changed by analyzing the image acquisition module, and the new sample image is analyzed to obtain the color difference change ratio between the samples before and after the color matching conditions are changed.

[0038] In the embodiment of the present invention, in the step S2, the color difference comparison value of the sample is determined by the gray value, and the color difference comparison value of the sample is determined by the gray value, so as to avoid human judgment and improve the accuracy of judgment.

[0039] In the embodiment of the present invention, in step S3, steps S1-S2 are repeatedly performed to obtain the color difference comparison value of the sample under different color matching conditions. The color difference comparison value , where , The color difference comparison values ​​before and after the change are obtained respectively, and the color difference comparison values ​​of the samples under different color matching conditions are obtained as a reference for the final color difference change ratio to improve the accuracy of the judgment. The color difference change ratio obtained by analysis is made into a color difference change ratio table for easy subsequent search.

[0040] The present invention realizes a comprehensive quantitative evaluation of fabric color difference by shooting sample images under preset color matching conditions and analyzing color difference comparison values, and reshooting images to analyze color difference change ratios when the color matching conditions change. In actual operation, multiple groups of sample images can be shot under preset color matching conditions according to specified steps, and then processed by an analysis module to accurately obtain color difference data under different conditions, thereby effectively solving the problem that it is difficult to quantify the color difference relationship under different light source color temperature conditions and provide data for process control when the fabric color matching environment is changeable, greatly improving the accuracy and reliability of fabric color difference evaluation, providing reliable quantitative data for subsequent process settings and process control, and filling the gap in the prior art in this regard.

[0041] Embodiment 2: Figure 1 As shown, a fabric color difference comparison system based on a color matching light box includes an image acquisition module and an analysis module, wherein the image acquisition module is used to capture a sample image of a sample under a preset color matching condition, and the analysis module is used to analyze the sample image to obtain a color difference comparison value of the sample; Among them, when the color matching conditions change, the image acquisition module is used to re-shoot the sample image, and the analysis module is used to analyze and obtain the color difference change ratio between the samples before and after the preset color matching conditions are changed. The above steps are performed with multiple samples, and the analysis module obtains multiple color difference change ratios between different samples under different color matching conditions.

[0042] Specifically, the sample is photographed under preset color matching conditions to obtain a sample image of the sample, and the sample image can be analyzed to obtain the color difference comparison value of the sample. After the comparison between the samples is completed, the color matching conditions are changed and the sample is re-photographed and analyzed to obtain the color difference change ratio between the samples before and after changing the preset color matching conditions. The above process is repeated to finally obtain multiple color difference change ratios between different samples under different color matching conditions.

[0043] As another embodiment of the present invention, the image acquisition module includes a colorimeter and a light source, the colorimeter is used to capture the sample image, and the light source is used to provide a preset color matching condition.

[0044] Specifically, after the sample image is photographed by the colorimeter, the analysis module may analyze the sample image.

[0045] As another embodiment of the present invention, the analysis module includes a segmentation submodule, a conversion submodule and an analysis submodule; The segmentation submodule is used to capture the boundary positions of different color areas in the sample image to obtain a color block segmentation map; The conversion submodule is used to convert the color blocks in the color block segmentation image into RGB values; The analysis submodule is used to analyze the color difference comparison value of the sample image, including the color difference comparison value of the color block and the overall color difference comparison value of the sample image.

[0046] Specifically, the sample image may be a monochrome image or an image composed of different color block images. The analysis module 2 first converts the sample image into a color block segmentation map, and then processes and analyzes the color block segmentation map.

[0047] The coordinated work of the image acquisition module and the analysis module in the system of the present invention realizes a highly automated and intelligent operation process. The colorimeter captures the sample image under specific light source conditions, and the analysis module processes the image according to the image features. For different types of images, color analysis or texture recognition can be used respectively, which not only improves the efficiency of color difference comparison, but also further enhances the accuracy of evaluation, better adapts to complex and diverse fabric detection needs, and ensures that accurate color difference comparison results can be stably obtained in various practical scenarios.

[0048] As another embodiment of the present invention, the segmentation submodule includes a region recognition submodule, an image segmentation submodule and a boundary recognition submodule; The region recognition submodule is used to recognize different color regions of the sample image; The image segmentation submodule is used to segment the sample image by grayscale segmentation method; The boundary recognition submodule is used to capture the boundary positions of different color blocks after image segmentation to obtain a color block segmentation map.

[0049] As another embodiment of the present invention, the image segmentation submodule includes a test image acquisition submodule, a calculation submodule and an optimal segmentation threshold submodule; The test image acquisition submodule is used to capture the grayscale histogram of the sample image; The calculation submodule is used to segment the sample image according to a preset segmentation threshold; The optimal segmentation threshold submodule is used to obtain the optimal segmentation threshold by calculating the intra-class variance and inter-class variance of the image after segmentation by each segmentation threshold; The optimal segmentation threshold submodule is also used to segment the sample image with the optimal segmentation threshold to obtain a color block segmentation map of the color image.

[0050] As another embodiment of the present invention, the analysis submodule includes a standard image acquisition submodule, a single color block comparison submodule and a sample image comparison submodule; The standard image acquisition submodule is used to acquire a standard image of the sample; The single color block comparison submodule is used to analyze the color difference between the sample image and the standard image to obtain the color difference ratio of the single color block; The sample image comparison submodule is used to sum the color difference ratios of individual color blocks to obtain the overall color difference ratio of the sample image.

[0051] The system of the present invention is closely combined with actual application needs through strict sample selection standards and flexible color matching condition settings. In actual operation, suitable samples can be selected according to different fabric characteristics to simulate various real color matching environments, ensuring that the comparison results are highly consistent with the actual production situation, so that the system can not only accurately evaluate the color difference, but also flexibly adjust the detection strategy according to the changes in different processes and color matching conditions, providing more targeted and reliable color difference control solutions for fabrics in multiple links such as design, production, and inspection, effectively improving the quality and efficiency of the entire fabric production and testing process.

[0052] Example 3: Color difference evaluation under different color matching conditions; 1. Experimental purpose: To evaluate the color difference of specific fabrics under different light source color temperature conditions and provide data support for process control.

[0053] 2. Experimental equipment and materials: color matching light box (with adjustable light source color temperature function), colorimeter (used to take sample images), standard white light source (color temperature of 6500K, as one of the preset color matching conditions), standard yellow light source (color temperature of 4000K, as the changed color matching condition), test fabric samples (including a variety of colors and patterns).

[0054] 3. Experimental steps; (1) Image acquisition and analysis under preset color matching conditions; (A) Set the light source of the color matching light box to a standard white light source (6500K), place the test fabric sample at a specific position in the light box, and ensure that the light incident angle is vertical; (B) Using a colorimeter to take an image of a test fabric sample, take an image of the sample; (C) Use the color clustering algorithm to identify different color regions in the sample image and minimize the objective function: , number of clusters ; (D) Grayscale histogram of the sample image taken , according to the preset segmentation threshold , initially set to 128, segment the image, calculate the intra-class variance and inter-class variance of the image after segmentation at each segmentation threshold, and get the optimal segmentation threshold of 150; (E) Segment the sample image with the optimal segmentation threshold, and use the boundary recognition algorithm to capture the boundary positions of different color blocks after image segmentation to obtain a color block segmentation map; (F) Convert the color blocks in the color block segmentation image into RGB values ​​to obtain the standard image of the sample; (G) Analyze the color difference between the sample image and the standard image, and calculate the color difference ratio of a single color block using the color difference formula, where: , , , then the color difference ratio of the color block , and sum up the color difference ratios of individual color blocks to obtain the overall color difference ratio of the sample image as 30.

[0055] (2) Image acquisition and analysis after changing color matching conditions; (A) Switch the light source of the color matching light box to a standard yellow light source (4000K), keep the position of the test fabric sample unchanged, and keep the incident angle of the light vertical.

[0056] (B) Repeat the above image acquisition and analysis steps to obtain a color difference comparison value of 40 for the sample image under yellow light source.

[0057] (3) Calculate the color difference change ratio; The color difference comparison value under white light source is 30. The color difference comparison value under yellow light source is 40, and the calculated .

[0058] 4. Experimental results: Under the standard white light source (6500K), the overall color difference ratio of the test fabric sample image is 30; under the standard yellow light source (4000K), the overall color difference ratio becomes 40, and the color difference change ratio between the two is about 33.33%. This shows that the change in the color temperature of the light source has a significant impact on the color presentation of the test fabric. In the actual production process, this color difference change needs to be considered to ensure the consistency of product color.

[0059] Example 4: Evaluation of color difference changes of multiple samples; 1. Experimental purpose: To compare the color difference variation of different types of test fabric samples under various color matching conditions, and to provide a reference for the process control of different fabrics.

[0060] 2. Experimental equipment and materials: color matching light box (with adjustable light source color temperature), colorimeter, standard white light source (6500K), standard yellow light source (4000K), standard blue light source (5000K), test fabric sample A (solid color), test fabric sample B (multi-color pattern), test fabric sample C (dark color).

[0061] 3. Experimental steps; (A) For each test fabric sample, image acquisition and analysis were performed under a standard white light source (6500K), and the steps were performed under the preset color matching conditions in Example 1 to obtain the color difference comparison value of each sample (the overall color difference ratio of sample A under white light source was 20, sample B was 40, and sample C was 50).

[0062] (B) The light source was switched to a standard yellow light source (4000K) and a standard blue light source (5000K) in turn, the sample position and the incident angle of light were kept unchanged, and the image acquisition and analysis steps were repeated to obtain the color difference comparison values ​​of each sample under different light sources (under the yellow light source, the overall color difference ratio of sample A became 25, sample B became 50, and sample C became 60; under the blue light source, sample A became 18, sample B became 35, and sample C became 45).

[0063] (C) Calculate the color difference change ratio of each sample before and after the color temperature of different light sources changes; The color difference ratio of sample A from white to yellow light source is: ; The color difference ratio of sample A from white to blue light source is: ; The color difference ratio of sample B from white to yellow light source is: ; The color difference ratio of sample B from white to blue light source is: ; The color difference ratio of sample C from white to yellow light source is: ; The color difference ratio of sample C from white to blue light source is: .

[0064] 4. Experimental results;

[0065] Different types of test fabric samples show different color difference changes under different light source color temperatures. The color difference of pure color sample A changes relatively little when the light source color temperature changes, while the color difference of multi-color pattern sample B and dark color sample C changes relatively much. This shows that in the fabric production process, the effect of light source color temperature on color needs to be considered specifically for different types of fabrics to ensure the stability and consistency of product quality.

[0066] Example 5: Verification of the effectiveness of the system in evaluating color difference of complex patterned fabrics; 1. Experimental purpose: To verify the accuracy and reliability of the color difference evaluation of complex pattern fabrics by the fabric color difference comparison system based on the color matching light box, as well as the effectiveness of the collaborative work of various modules of the system.

[0067] 2. Experimental equipment and materials: color matching light box (the color temperature of the light source can be precisely adjusted), high-precision colorimeter, standard white light source (6500K), standard green light source (5500K), complex pattern test fabric samples (including a variety of colors, textures and details).

[0068] 3. Experimental steps; (A) Under a standard white light source (6500K), a colorimeter is used to capture an image of a complex pattern test fabric sample, and the image acquisition module transmits the image to the analysis module.

[0069] (B) The segmentation submodule in the analysis module processes the image using the color clustering algorithm and the grayscale segmentation method. The region recognition submodule identifies different color regions. The image segmentation submodule obtains the optimal segmentation threshold of 130 by calculating the grayscale histogram, the preset segmentation threshold, the intra-class variance, and the inter-class variance. After the image is segmented with the optimal segmentation threshold, the boundary recognition submodule obtains the color block boundary position and obtains the color block segmentation map.

[0070] (C) The conversion submodule converts the color blocks in the color block segmentation image into RGB values. The standard image acquisition submodule in the analysis submodule obtains the standard image of the fabric sample. The single color block comparison submodule calculates the color difference ratio between the single color block and the standard image through the color difference formula. The sample image comparison submodule sums up the color difference ratios of all single color blocks to obtain an overall color difference ratio of 35 for the sample image under white light source.

[0071] (D) Switch the light source to a standard green light source (5500K), repeat the above image acquisition and analysis steps, and obtain an overall color difference ratio of 45 for the sample image under green light source.

[0072] (E) Calculate the color difference change ratio as .

[0073] 4. Experimental results; Under standard white light source (6500K), the overall color difference ratio of the complex pattern test fabric sample image is 35; under standard green light source (5500K), the overall color difference ratio becomes 45, and the color difference change ratio is about 28.57%. This shows that the fabric color difference comparison system can effectively process the image of complex pattern fabrics, accurately analyze the color difference comparison value and color difference change ratio, verify the accuracy and reliability of the system in practical applications, and the effectiveness of the collaborative work of each module, which can provide reliable data support for the production process control of complex pattern fabrics.

[0074] The embodiments of the present invention disclose preferred embodiments, but are not limited thereto. A person skilled in the art can easily understand the spirit of the present invention based on the above embodiments and make different extensions and changes. However, as long as they do not deviate from the spirit of the present invention, they are all within the protection scope of the present invention.

Claims

1. A method for comparing fabric color difference based on a color matching light box, characterized in that: The steps include: S1: taking a sample image of the sample under the preset color matching conditions; S2: Analyze the sample image, determine the color difference comparison value of the sample by the gray value, and obtain the color difference comparison value of the sample; S3: When the color matching conditions are changed, re-shoot the sample images and analyze them to obtain the color difference change ratio between the samples before and after the preset color matching conditions are changed; S4: Execute steps S1-S3 with multiple samples to obtain multiple color difference change ratios between different samples under different color matching conditions; Wherein, the step S2 includes the following sub-steps: S201: photographing the boundary positions of different color areas in the sample image to obtain a color block segmentation map; S202: Convert the color blocks in the color block segmentation image into RGB values; S203: Analyze the color difference comparison value of the sample image, including the color difference comparison value of the color block and the overall color difference comparison value of the sample image.

2. According to claim 1, a method for comparing color difference of fabrics based on a color matching light box is characterized in that: The sub-step S201 includes the following sub-steps: S201a, identifying different color areas of the sample image; Among them, the color clustering algorithm is used to identify different color areas of the sample image and minimize the objective function: ； In the formula, is the number of clusters, It's a pixel. It is A set of clusters, It is The center of the cluster; S201b, performing image segmentation on the sample image by using a grayscale segmentation method; S201c, using a boundary recognition algorithm to capture the boundary positions of different color blocks after image segmentation, and obtain a color block segmentation map.

3. The method for comparing color difference of fabrics based on a color matching light box according to claim 2, characterized in that: The step S201b includes the following sub-steps: A1. Grayscale histogram of the sample image , where represents the gray level, Indicates grayscale The corresponding number of pixels; A2. According to the preset segmentation threshold , segment the sample image; A3, by calculating the intra-class variance and inter-class variance of the image after segmentation at each segmentation threshold, the optimal segmentation threshold is obtained; A4. Segment the sample image using the optimal segmentation threshold to obtain a color block segmentation map of the color image.

4. The method for comparing fabric color difference based on a color matching light box according to claim 3, characterized in that: The step A3 comprises the following sub-steps: A301, calculate the segmentation interval set: Each segmentation threshold , the corresponding between-class variance is ; Among them, the between-class variance is expressed by the formula Calculate, where , is the ratio of the two types of pixels, , is the average gray value of the two types of pixels, is the average gray value of the entire image; A302. Calculate each segmentation threshold The corresponding intra-class variance of the image; Among them, the intra-class variance is expressed by the formula Calculate, where , are the variances within the two types of pixels; A303, taking the maximum value among the intra-class variances corresponding to each segmentation threshold, and calculating the corresponding inter-class variance; A304. Take the maximum value of the inter-class variance as the optimal segmentation threshold.

5. The method for comparing fabric color difference based on a color matching light box according to claim 1, characterized in that: The step S203 includes the following sub-steps: S203a: Acquire a standard image of the sample; S203b: Analyze the color difference between the sample image and the standard image to obtain a color difference ratio of a single color block; Among them, the color difference ratio of a single color block is calculated by the color difference formula Calculate, where , , They are the differences of brightness, red-green axis, and yellow-blue axis in CIELab color space. The color difference between the sample image and the standard image is analyzed to obtain the color difference ratio of a single color block. S203c: Sum the color difference ratios of the individual color blocks to obtain the overall color difference ratio of the sample image.

6. The method for comparing color difference of fabrics based on a color matching light box according to claim 1, characterized in that: In step S3, the position of the sample and the incident angle of the light are the same before and after the color matching condition is changed.

7. The method for comparing fabric color difference based on a color matching light box according to claim 1, characterized in that: In step S3, steps S1-S2 are repeatedly performed to obtain the color difference comparison values ​​of the samples under different color matching conditions. The color difference comparison values , where , They are the color difference comparison values ​​before and after the change respectively.

8. A color-matching light box-based fabric color-difference comparison system, which uses the color-matching light box-based fabric color-difference comparison method according to any one of claims 1 to 7, characterized in that: include: An image acquisition module, used for capturing a sample image of the sample under a preset color matching condition; An analysis module is used to analyze the sample image and obtain the color difference comparison value of the sample; Wherein, the image acquisition module includes a colorimeter for capturing sample images and a light source for providing preset color matching conditions; The analysis module comprises: The segmentation submodule is used to capture the boundary positions of different color areas in the sample image and obtain a color block segmentation map; The conversion submodule is used to convert the color blocks in the color block segmentation image into RGB values; The analysis submodule is used to analyze the color difference comparison value of the sample image, including the color difference comparison value of the color block and the overall color difference comparison value of the sample image.

9. The cloth color difference comparison system based on the color matching light box according to claim 8, characterized in that: The segmentation submodule comprises a region recognition submodule, an image segmentation submodule and a boundary recognition submodule, and the image segmentation submodule comprises a test image acquisition submodule, a calculation submodule and an optimal segmentation threshold submodule.

10. The cloth color difference comparison system based on the color matching light box according to claim 9, characterized in that: The analysis submodule includes a standard image acquisition submodule for acquiring a sample standard image, a single color block comparison submodule for analyzing the color difference between the sample image and the standard image, and a sample image comparison submodule for summing the color difference ratios of the single color blocks.

Images