Digital image technology-based color fastness rating method for single-color sample

Abstract

The invention provides a digital image technology-based color fastness rating method for a single-color sample. The method comprises the following steps of: placing an original sample and a variable / stained sample on a white card which is visually uniform and consistent in parallel; judging whether the color of the variable / stained sample is uniform or not by using an automatic judgment method; if so, the measurement area is selected at the center position of the sample; if the color difference is not uniform, non-uniform parts are processed through digital images to judge whether the color difference is larger than a preset threshold value, and if the color difference is larger than the preset threshold value, the area with the most remarkable color change is adopted as the measurement area; if the color difference is smaller than or equal to the preset threshold value, the measurement area selects all test areas of the sample; and setting the test area of the original sample to be the same as the test area of the variable / stained sample, and evaluating the determined test area based on a digital image technology. According to the method, the non-uniform phenomenon occurring in the actual variable / stained sample is fully considered, the automatic selection of the test area is realized, and the rating accuracy is improved.

Description

Technical Field

[0001] The invention relates to the technical field of textile color fastness rating, in particular to a color fastness rating method for monochrome samples based on digital image technology. Background Art

[0002] The color fastness of textiles, that is, the persistence of their color when subjected to various external factors, is an important performance indicator for measuring textile quality. There are generally two core indicators for evaluating color fastness: one is the degree of discoloration of the fabric itself, and the other is the color staining on undyed adjacent fabric. Subjective evaluation of color fastness includes visual inspection, in which a trained evaluator compares colors using a standard color chart to determine the degree of color change. Objective evaluation methods include color measurement with a spectrophotometer and grading based on digital image recognition and processing. Subjective evaluation methods are simple and direct, but are easily influenced by subjective factors such as the evaluator's psychology. Objective evaluation methods are highly objective, repeatable, and efficient.

[0003] Currently, third-party testing agencies still rely on visual evaluation by experienced, visually calibrated professionals for colorfastness testing. This is primarily due to the fact that objective evaluation methods based on spectrophotometers (GB / T 32598-2016 and GB / T 32616-2016) are limited by the instrument's own colorimetric aperture (which determines the measurement area). This prevents the entire measurement area from being selected for measurement on samples exhibiting non-uniform color change or staining. While colorfastness rating methods based on digital imaging technology allow for flexible determination of the measurement area, GB / T 39648-2020, "Textiles - Color Fastness Tests - Digital Image Technology Rating," only stipulates "manual selection of the measurement area" for single-color specimens and does not provide detailed instructions for selecting the measurement area. Currently, colorfastness instruments based on digital imaging technology also require the tester to select the test area themselves. Practice has shown that arbitrarily determining the location and size of the test area significantly affects rating accuracy.

[0004] Patent application number 202310145478.2 discloses a method for automatically rating the color fastness of textiles based on spectral reconstruction. The method comprises the following steps: preparing textile samples; conducting color fastness tests according to different color fastness test requirements; visually rating the color fastness and obtaining the visual color fastness rating results of the textile samples; under specified conditions, using a digital camera to capture digital images of the textile samples and spectral modeling samples, and extracting digital response values ​​of the photographed objects; using the spectral modeling samples to perform spectral modeling on the camera to obtain a spectral reconstruction model; using the spectral reconstruction model to reconstruct the spectrum of the textile samples; using the spectrum of the textile samples to calculate the corresponding color data; using the color data of the textile samples to calculate the color difference and color attribute difference corresponding to the color fastness test; constructing a color fastness prediction model for the textile samples; and using the color fastness prediction model to predict the color fastness grade of the target textile samples. The invention eliminates subjective differences and overcomes the problem of inaccurate ratings caused by color cast in existing rating methods. However, regarding the selection of the measurement area, the invention only mentions that the shooting area of ​​the digital camera is the designated area, but does not specify the method for selecting the designated area. Its accuracy is based on the condition that the textile samples are uniform, and does not take into account the complex situation of uneven textile samples that may occur during the actual color fastness test. Summary of the Invention

[0005] To address the technical issue that existing colorfastness evaluation methods lack detailed information on how to select the measurement area, this paper proposes a colorfastness rating method for monochrome specimens based on digital imaging technology. By implementing standardized test area selection and automatically selecting the test area and area, the proposed rating method can accurately rate monochrome specimens. This method fully accounts for the unevenness that can occur in actual color shifting / staining samples, enabling automated selection of the test area. This method effectively improves the consistency between instrumental and visual ratings, enhancing rating accuracy.

[0006] In order to achieve the above object, the technical solution of the present invention is implemented as follows: a color fastness rating method for monochrome samples based on digital image technology, the steps of which are as follows:

[0007] Step S1, placing the original sample and the color change / staining sample after the color fastness test side by side on a white opaque paper or white card that is visually uniform and does not contain fluorescence;

[0008] Step S2, using an automatic determination method to determine whether the color of the discolored / stained sample is uniform;

[0009] Step S3: If it is determined to be uniform, the measurement area is selected at the center of the sample; if it is determined to be non-uniform, proceed to step S4;

[0010] Step S4: Determine whether the color difference of the uneven area is greater than a preset threshold by digital image processing. If the color difference is greater than the preset threshold, the measurement area of ​​the discolored / stained sample is the area with the most significant color change. If the color difference is less than or equal to the preset threshold, the measurement area of ​​the discolored / stained sample is the entire test area of ​​the sample.

[0011] Step S5: The test area of ​​the original sample is set to the same position and test area as the test area of ​​the discoloration / staining sample, and the automatically determined test area is transmitted to an instrument based on digital imaging technology for evaluation and then a color fastness grade is given.

[0012] Preferably, the automatic determination method comprises the following steps:

[0013] 1). Using a digital imaging system instrument to collect a surface image I of the discolored / stained sample, and preprocessing the surface image I;

[0014] 2) Segment the preprocessed surface image I into equally spaced regions, calculate the color difference between any two sub-regions, and obtain multiple color difference pairs;

[0015] 3) Extract the maximum value of the color difference pair as the maximum color difference value;

[0016] 4) Compare the maximum color difference value with the preset threshold value to determine whether the color of the discolored / stained sample is uniform.

[0017] Preferably, in step 4), the maximum color fastness grade is calculated according to the maximum color difference value, and the maximum color fastness grade is compared with a preset threshold value to determine whether the color of the discoloration / staining sample is uniform.

[0018] Preferably, the method of preprocessing the surface image I includes denoising and white balance correction;

[0019] The preprocessed surface image I is divided into n sub-regions using a clustering algorithm. The center point of each sub-region is used as a sampling point, its chromaticity parameters are read, and the CIEDE2000 color difference ΔE between each of the n sub-regions is calculated. 00 , a total of m color difference pairs ΔE are obtained 00_m ;

[0020] CIEDE2000 color difference ΔE 00 DE2000 brightness coefficient k in the calculation L =1.0, DE2000 saturation coefficient k C =0.5.

[0021] Preferably, the method for calculating the maximum color fastness grade according to the maximum color difference value is: the maximum color fastness grade of the discoloration sample is: GRC max=0.88+3.89e -0.2ΔEGRC ;

[0022] In the formula, the intermediate variable ΔL 00 is the brightness difference of CIEDE2000; ΔE max is the maximum color difference value;

[0023] The maximum color fastness grade of the stained specimen is:

[0024] In the formula, the intermediate variable ΔL 00 CIEDE2000 brightness difference.

[0025] Preferably, if the maximum color difference value ΔE max If the maximum color difference value ΔE is greater than the threshold value I, it can be judged as uneven; max If it is less than or equal to the threshold I, it can be judged as uniform;

[0026] If the maximum color fastness grade GRC max If the color fastness is greater than the threshold II, it can be judged as uneven; if the maximum color fastness grade GRC max If it is less than or equal to the threshold II, it can be judged as uniform.

[0027] Preferably, if it is judged to be uniform, for discolored samples, the measuring area of ​​the measuring area is set to a rectangle of 3.2 cm long and 1.6 cm wide; for stained samples, the measuring area of ​​the measuring area is set to a circle with a diameter of 1 cm; if the sample area is smaller than the measuring area of ​​the measuring area, the measuring area of ​​the measuring area is automatically reduced until the test area is filled with the sample color.

[0028] Preferably, the method for determining the color difference of the uneven parts in step S4 is:

[0029] (1) Extract the average tristimulus values ​​or chromaticity parameters of the area with obvious color difference in the uneven color change / staining sample;

[0030] (2) Calculate the color difference or color fastness grade between the area with the most significant color change and the original area using the obtained average tristimulus values ​​or colorimetric parameters. Compare the color difference or color fastness grade with the preset threshold to determine whether the color difference of the uneven area is obvious.

[0031] Preferably, the method for extracting the area with obvious color difference in the uneven sample is: using an image feature extraction algorithm to identify the area with significant color change in the pre-processed surface image I to locate the area with the largest color difference, and the implementation method is: for the pre-processed surface image I, according to the resolution information, calculate the number of pixels equivalent to 1 cm, and use this as a unit to divide the entire pre-processed surface image I into several adjacent and non-overlapping fixed-size sub-areas, each sub-area has a size of 1×1 (cm 2 ); for each sub-region, extract the color data of all its pixels; calculate the color standard deviation of the sub-region based on the color values ​​of all pixels contained in the sub-region; compare the values ​​of the standard deviation of all sub-regions, and select the region with the largest standard deviation as the image region with the most significant color change.

[0032] Preferably, the discolored sample is automatically identified to obtain the lighter area compared to the original sample, and the stained sample is automatically identified to obtain the darker area compared to the original sample; the average tristimulus value or colorimetric parameter of the area with the most significant color change is read by OCR identification;

[0033] The CIEDE2000 color difference between the area with the most significant color change and the original area is calculated using the average tristimulus value or chromaticity parameter, and recorded as color difference ΔE 00_2 If the color difference ΔE 00_2 > preset threshold I, the color difference of the sample is judged to be obvious; if the color difference ΔE 00_2 If the value is less than the preset threshold value I, the color difference of the discolored / stained sample is judged to be not obvious;

[0034] According to the color difference ΔE 00_2 The corresponding color fastness grade GRC is calculated. If the color fastness grade GRC> the preset threshold II, it is judged that the color difference of the color change / staining sample is relatively obvious; if the color fastness grade GRC≤the preset threshold II, it is judged that the color difference of the color change / staining sample is not obvious.

[0035] The beneficial effects of the present invention are as follows: the present invention fixes the original sample and the color-changing / stained sample on a visually uniform and non-fluorescent opaque white paper or white card; then uses a software program to automatically determine whether the color of the color-changing / stained sample is uniform; if it is determined to be uniform, the measurement area is selected near the center of the sample; for uneven samples, if the color difference of the uneven parts is more obvious, the sample measurement area is the area with the most severe color change; if the color difference of the uneven parts is not obvious, the sample measurement area should be the entire test area of ​​the sample; after determining the measurement area of ​​the sample, the test area of ​​the original sample is set to the same position and area as the test area of ​​the sample, and then an instrument based on digital image technology is used for evaluation to give a color fastness grade. The method of selecting the test area of ​​the present invention can effectively improve the consistency between the instrument rating results and the visual results, and improve the accuracy of the rating; the method of automatically selecting the test area can provide a basis for the automatic evaluation of color fastness. The present invention provides a method for instrumental evaluation and intelligent evaluation of color fastness, improving the accuracy of instrumental evaluation. Driven by the high-quality development of new quality productivity, it is a general trend to adopt accurate objective testing methods to replace the subjective evaluation methods of textiles that are highly subjective, poorly repeatable and inefficient.

[0036] The color fastness rating method designed by the present invention, based on digital imaging technology and designed for monochrome samples, is an objective method. Compared to visual rating methods, it is not affected by factors such as the user's psychological, physiological, and proficiency levels, and provides stable and accurate results. The present invention has a wide range of applications and is applicable to any instrument used to evaluate color fastness based on digital imaging technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 It is the overall flow chart of the present invention. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without creative work are within the scope of protection of the present invention.

[0040] like Figure 1As shown in the figure, a color fastness rating method based on digital image technology for monochrome samples is an intelligent rating method based on color fastness testing instruments of digital image recognition type. The steps are as follows:

[0041] Step S1: Place the original sample and the color change / staining sample after the color fastness test side by side on a visually uniform and non-fluorescent opaque white paper or white card to prevent the sample from being affected by the backing color.

[0042] Step S2: using an automatic determination method to determine whether the color of the discolored / stained sample is uniform.

[0043] The automatic judgment method specifically includes the following steps:

[0044] 1. Use a digital imaging system compliant with GB / T 39648-2020 to capture surface image I of the discolored / stained specimen. To avoid interference from ambient light, illuminate the sample with a uniform light source. Perform preprocessing on surface image I, including denoising and white balance correction, to reduce random noise in the image, improve analysis accuracy, eliminate light source color cast, and ensure faithful color reproduction.

[0045] 2. Perform equal-interval region segmentation on the preprocessed surface image I. Specifically, a clustering algorithm (such as K-means or DBSCAN) is used to divide the preprocessed surface image I into n sub-regions, for example, n circular regions with a diameter of 1 cm. The center point of each sub-region is used as a sampling point, and its color parameters are read. The CIEDE2000 color difference ΔE between any two sub-regions can be calculated. 00 , a total of m color difference pairs ΔE are obtained 00_m .

[0046] 3. Extract color difference ΔE 00_m The maximum value ΔE in max If ΔE max If ΔE is greater than the threshold, it can be judged as uneven; max If the color difference is less than or equal to the threshold, it can be judged as uniform. The size of this threshold can be determined by mutual agreement. When judging whether it is uniform or not, you can set a color difference threshold or a color fastness level threshold. Using formula (1) and formula (2), you can calculate the color difference ΔE by 00 Calculate the corresponding color fastness grade GRC, which can be obtained by the maximum value ΔE max Calculate the maximum GRC max Similarly, if the color fastness grade is GRC max If the color fastness grade is greater than the threshold value, it can be judged as uneven; if the color fastness grade is the maximum value GRC maxIf the color is less than or equal to the threshold, it can be judged as uniform. The threshold value can also be determined by mutual agreement. Based on experience, the color fastness grade GRC = 4-5 can be selected, indicating small to medium differences.

[0047] The color fastness grade of the discoloration sample is calculated by formula (1):

[0048]

[0049] Where, ΔL 00 CIEDE2000 brightness difference; CIEDE2000 color difference ΔE 00 DE2000 brightness coefficient k in the calculation L =1.0, DE2000 saturation coefficient k C =0.5.

[0050] The color fastness grade of the staining sample is calculated by formula (2):

[0051]

[0052] Where, ΔL 00 CIEDE2000 brightness difference; CIEDE2000 color difference ΔE 00 DE2000 brightness coefficient k in the calculation L =1.0, DE2000 saturation coefficient k C =0.5.

[0053] Step S3: If the surface is judged to be uniform in step S2, then step S4 is executed; if the surface is judged to be non-uniform, then step S5 is executed.

[0054] In step S4, if it is determined to be uniform, the measurement area is selected at the center of the sample, and the center position is automatically located according to the geometric center of the image to ensure representativeness and avoid the influence of edge anomalies.

[0055] For discoloration samples, the measurement area is set as a 3.2cm x 1.6cm rectangle, the same size as the color change chart used for visual color fastness evaluation. For staining samples, the measurement area is set as a 1cm diameter circle, the same size as the color change / staining chart used for visual color fastness evaluation. The consistent size of the measurement area and the color change / staining chart ensures a consistent reference for automated evaluation and visual evaluation, reducing evaluation bias caused by size differences. If the sample area is smaller than this rectangle / circle, the test area is automatically reduced until the sample color fills the test area.

[0056] In step S5, if the color difference is determined to be uneven, digital image processing is used to further determine whether the color difference of the uneven area is obvious, that is, whether it is greater than a preset threshold. If the color difference is greater than the preset threshold, the sample measurement area is the area with the most significant color change, which is recorded as Area I. If the color difference is not obvious, that is, less than or equal to the preset threshold, the sample measurement area is the entire test area of ​​the sample, which is recorded as Area II. Specifically:

[0057] 1. Extract the average tristimulus values ​​or colorimetric parameters of areas with significant color differences in the non-uniform sample. Specifically, use an image feature extraction algorithm (analysis of variance) to identify areas of significant color variation in the preprocessed surface image I obtained in step S1. This identifies areas of maximum color difference and improves the sensitivity and accuracy of the assessment. First, for the preprocessed surface image I, the number of pixels equivalent to 1 cm is calculated based on the resolution information of the system or display device. The physical size of the image is mapped to a pixel scale to ensure that the size of the segmented regions meets actual measurement requirements and provide a consistent measurement standard for different operators. Based on this unit, the entire surface image I is divided into several adjacent, non-overlapping subregions of a fixed size, each approximately 1×1 cm in size. For each subregion in surface image I, color data for all pixels is extracted. Each pixel contains values ​​for the three color channels: red (R), green (G), and blue (B). For each subregion, the color standard deviation is calculated based on the color values ​​of all pixels within it. The color standard deviation refers to the degree of dispersion of the pixel color distribution within that region. The larger the standard deviation, the more complex or dramatic the color change in that area; the smaller the standard deviation, the flatter the area and the more evenly distributed the color. The standard deviation values ​​of all sub-areas are compared, and the area with the largest standard deviation is selected as the image area with the most significant color change. This image area is considered to be the most visually complex or significant part of the image and is recorded as Area I. The discolored sample is automatically identified as a lighter area compared to the original sample, and the stained sample is automatically identified as a darker area compared to the original sample. The average tristimulus values ​​or colorimetric parameters (such as L*, a*, and b* values) of Area I are read through OCR recognition.

[0058] 2. Determine whether the color difference is obvious. Specifically, calculate the CIEDE2000 color difference between area I and the original area using the obtained average tristimulus values ​​or chromaticity parameters, and record it as color difference ΔE 00_2 Compare it with the preset threshold, if the color difference ΔE 00_2 >Preset threshold value, the sample color difference is judged to be more obvious; if the color difference ΔE 00_2 ≤ threshold, the color difference of the sample is judged to be not obvious. The size of this color difference threshold can be determined by mutual agreement. It can also be determined based on the color difference ΔE 00_2The corresponding color fastness grade (GRC) is calculated. If GRC is greater than the threshold, the color difference is considered significant; if GRC is less than the threshold, the color difference is considered less significant. The color fastness grade threshold can be determined by mutual agreement. For example, based on experience, a threshold of GRC = 4 can be selected.

[0059] 3. If the difference is obvious, use area I as the test area of ​​the specimen.

[0060] 4. If the difference is not obvious, use Area II as the test area of ​​the specimen.

[0061] Step S6: After determining the measurement area of ​​the sample, the original test area is set to the same position and area as the test area of ​​the sample, and then the automatically determined test area is transmitted to an instrument based on digital imaging technology (such as a FastnessGrade color fastness tester) for evaluation and then a color fastness grade is given.

[0062] In order to verify the accuracy of the method of the present invention, samples submitted for inspection by a national textile testing agency were selected, and a total of 11 samples with uneven color change / staining were selected. Three rating experts from the national textile testing agency who had undergone visual calibration and passed the assessment rated the samples with uneven color change / staining to obtain visual grades. The instrument rating was performed using the FastnessGrade color fastness tester based on digital imaging technology. When the instrument rating used a fixed measurement area of ​​a 20mm diameter circle, its color fastness grade and visual grade are shown in Table 1. The results show that the difference between the instrument rating and the visual rating is 1. After using the method of the present invention, the grades obtained by re-measurement are shown in Table 2. The difference with the visual rating grade is within 0.5 grades, which is in line with the tolerance of the color fastness grade difference for manual visual evaluation.

[0063] Table 1 The difference between the instrument measurement level and the visual measurement level when the fixed area is 20 mm

[0064]

[0065] Table 2 Results of the difference between instrument measurement level and visual measurement level when adjusting the area

[0066]

[0067] Example 1

[0068] like Figure 1 As shown, a color fastness rating method based on digital image technology for monochrome samples includes the following steps:

[0069] Step 1: Fix the original sample and the discolored sample on a white paper or white card that is visually uniform and does not contain fluorescence;

[0070] Step 2: Use a color fastness meter based on digital image technology that complies with the GB / T 39648-2020 standard to collect the surface image I of the discolored sample. De-noise the image by Gaussian filtering and divide it into 25 circular areas with a diameter of 1 cm. Use optical character recognition (OCR) to automatically obtain the tristimulus values ​​X, Y, and Z of the center point of each sub-area measured by the color fastness meter based on digital image technology.

[0071] Step 3: Calculate the CIEDE2000 color difference between each of the 25 regions to obtain a total of 300 color difference pairs. The maximum value of the statistical color difference pairs is 1.2CIEDE2000, and the color change sample is judged to be uniform.

[0072] Step 4: The measurement area is selected at the geometric center of the sample, and the measurement area is selected as a rectangle with a length of 3.2 cm and a width of 1.6 cm.

[0073] Step 5: A rectangle with a length of 3.2 cm and a width of 1.6 cm is also selected in the center of the original sample as the measurement area, and the color fastness grade of 4-5 is given by a color fastness meter based on digital imaging technology.

[0074] Example 2

[0075] like Figure 1 As shown, a color fastness rating method based on digital image technology for monochrome samples includes the following steps:

[0076] Step 1: Fix the original sample and the stained sample on a white paper or white card that is visually uniform and does not contain fluorescence;

[0077] Step 2: Use an automated method to determine the color uniformity of the stained sample. Because the stained sample is relatively small, the preprocessed surface image I is evenly divided into nine 1×1 cm square regions, forming a matrix structure with three rows and three columns. The clustering algorithm uses a fixed spatial segmentation method to divide the image region coordinates of surface image I into the following matrix:

[0078] Region division = {R1, R2, ..., R9}

[0079] Each region R i (i=1, 2, ..., 9) contains the RGB value set of the pixel point.

[0080] For each area, the X, Y, and Z values ​​displayed by the digital imaging system are read as the average tristimulus values ​​of the area.

[0081] Step 3: The maximum value of the 44 color difference pairs is 6.67CIEDE2000, indicating that the color of the discolored sample is not uniform.

[0082] Step 4: For each subregion in surface image I, extract the color data of all its pixels. Each pixel contains values ​​for the three color channels: red (R), green (G), and blue (B). For each subregion, calculate the color standard deviation of the region based on the color values ​​of all the pixels within it. Compare the standard deviation values ​​of all subregions and select the region with the largest standard deviation as the image region with the most significant color change, denoted as Region I. Extract the average tristimulus value of Region I and calculate the color difference between it and the original color. The resulting color fastness value is 3.32, which is a significant difference. Therefore, the measurement region is the region with the most severe color change.

[0083] Step 4: Set the same position and size as the staining test area in the original sample, and assign a color fastness grade of 3-4 to the staining test area using an instrument based on digital imaging technology.

[0084] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A color fastness rating method for monochrome samples based on digital image technology, characterized in that: The steps are as follows: Step S1, placing the original sample and the color change / staining sample after the color fastness test side by side on a white opaque paper or white card that is visually uniform and does not contain fluorescence; Step S2, using an automatic determination method to determine whether the color of the discolored / stained sample is uniform; Step S3: If it is determined to be uniform, the measurement area is selected at the center of the sample; if it is determined to be non-uniform, proceed to step S4; Step S4, determining whether the color difference of the uneven area is greater than a preset threshold by digital image processing. If the color difference is greater than the preset threshold, the measurement area of ​​the discolored / stained sample is the area with the most significant color change; If the color difference is less than or equal to the preset threshold, the measurement area of ​​the discolored / stained specimen selects the entire test area of ​​the sample; Step S5: The test area of ​​the original sample is set to the same position and test area as the test area of ​​the discoloration / staining sample, and the automatically determined test area is transmitted to an instrument based on digital imaging technology for evaluation and then a color fastness grade is given.

2. The color fastness rating method based on digital image technology for monochrome samples according to claim 1, characterized in that: The automatic judgment method comprises the following steps: 1). Using a digital imaging system instrument to collect a surface image I of the discolored / stained sample, and preprocessing the surface image I; 2) Segment the preprocessed surface image I into equally spaced regions, calculate the color difference between any two sub-regions, and obtain multiple color difference pairs; 3) Extract the maximum value of the color difference pair as the maximum color difference value; 4) Compare the maximum color difference value with the preset threshold value to determine whether the color of the discolored / stained sample is uniform.

3. The color fastness rating method for monochrome samples based on digital image technology according to claim 2, characterized in that: In step 4), the maximum color fastness grade is calculated based on the maximum color difference value, and the maximum color fastness grade is compared with a preset threshold value to determine whether the color of the discoloration / staining sample is uniform.

4. The color fastness rating method for monochrome samples based on digital image technology according to claim 3, characterized in that: The method for preprocessing the surface image I includes denoising and white balance correction; The preprocessed surface image I is divided into n sub-regions using a clustering algorithm. The center point of each sub-region is used as a sampling point, its chromaticity parameters are read, and the CIEDE2000 color difference ΔE between each of the n sub-regions is calculated. 00 , a total of m color difference pairs ΔE are obtained 00_m ; CIEDE2000 color difference ΔE 00 DE2000 brightness coefficient k in the calculation L =1.0, DE2000 saturation coefficient k C =0.

5.

5. The color fastness rating method for monochrome samples based on digital image technology according to any one of claims 1 to 4, characterized in that: The method for calculating the maximum color fastness grade according to the maximum color difference value is as follows: the maximum color fastness grade of the discoloration sample is: In the formula, the intermediate variable ΔL 00 is the brightness difference of CIEDE2000; ΔE max is the maximum color difference value; The maximum color fastness grade of the stained specimen is: In the formula, the intermediate variable ΔL 00 CIEDE2000 brightness difference.

6. The color fastness rating method for monochrome samples based on digital image technology according to claim 5, characterized in that: If the maximum color difference ΔE max If the maximum color difference value ΔE is greater than the threshold value I, it can be judged as uneven; max If it is less than or equal to the threshold I, it can be judged as uniform; If the maximum color fastness grade GRC max If the color fastness is greater than the threshold II, it can be judged as uneven; if the maximum color fastness grade GRC max If it is less than or equal to the threshold II, it can be judged as uniform.

7. The color fastness rating method for monochrome samples based on digital image technology according to claim 6, characterized in that: If it is judged to be uniform, for discolored samples, the measurement area of ​​the measurement area is set to a rectangle of 3.2 cm long and 1.6 cm wide; for stained samples, the measurement area of ​​the measurement area is set to a circle with a diameter of 1 cm; if the sample area is smaller than the measurement area of ​​the measurement area, the measurement area of ​​the measurement area is automatically reduced until the test area is filled with the sample color.

8. The color fastness rating method for monochrome samples based on digital image technology according to claim 5 or 6, characterized in that: The method for determining the color difference of the uneven parts in step S4 is: (1) Extract the average tristimulus values ​​or chromaticity parameters of the area with obvious color difference in the uneven color change / staining sample; (2) Calculate the color difference or color fastness grade between the area with the most significant color change and the original area using the obtained average tristimulus values ​​or colorimetric parameters. Compare the color difference or color fastness grade with the preset threshold to determine whether the color difference of the uneven area is obvious.

9. The color fastness rating method for monochrome samples based on digital image technology according to claim 8, characterized in that: The method for extracting the area with obvious color difference in the uneven sample is as follows: using an image feature extraction algorithm to identify the area with significant color change in the pre-processed surface image I to locate the area with the largest color difference, and the implementation method is as follows: for the pre-processed surface image I, based on the resolution information, the number of pixels equivalent to 1 cm is calculated, and the entire pre-processed surface image I is divided into a number of adjacent and non-overlapping fixed-size sub-areas, each of which has a size of 1×1 (cm 2 ); for each sub-region, extract the color data of all its pixels; calculate the color standard deviation of the sub-region based on the color values ​​of all pixels contained in the sub-region; compare the standard deviation values ​​of all sub-regions, and select the region with the largest standard deviation as the image region with the most significant color change.

10. The color fastness rating method for monochrome samples based on digital image technology according to claim 9, characterized in that: Automatically identify discolored samples to obtain lighter areas compared to the original sample, and automatically identify stained samples to obtain darker areas compared to the original sample. Use OCR to read the average tristimulus values ​​or colorimetric parameters of the areas with the most significant color changes. The CIEDE2000 color difference between the area with the most significant color change and the original area is calculated using the average tristimulus value or chromaticity parameter, and recorded as color difference ΔE 00_2 If the color difference ΔE 00_2 > preset threshold I, the color difference of the sample is judged to be obvious; if the color difference ΔE 00_2 If the value is less than the preset threshold value I, the color difference of the discolored / stained sample is judged to be not obvious; According to the color difference ΔE 00_2 The corresponding color fastness grade GRC is calculated. If the color fastness grade GRC> the preset threshold II, it is judged that the color difference of the color change / staining sample is relatively obvious; if the color fastness grade GRC≤the preset threshold II, it is judged that the color difference of the color change / staining sample is not obvious.

Images