A method and system for detecting air permeability and color fastness of a knitted garment fabric

Abstract

The application discloses a kind of knitted garment air permeability and color fastness detection method and system, applied to textile detection technical field, by obtaining the multi-band spectral reflectance data and three-dimensional space geometry data of the surface of material to be measured, according to spectral data extraction optical characteristic parameter, according to geometry data extraction geometric characteristic parameter, these parameters are input into performance evaluation model, and the air permeability prediction value and color fastness performance prediction grade are calculated, so as to avoid the partial non-uniform damage caused by preposition air permeability test in sequential test to sample, with the accuracy and repeatability of subsequent color fastness test result, guarantee the beneficial effects of product marketing certification and consumer experience.

Description

Technical Field

[0001] This invention relates to the field of textile testing technology, and in particular to a method and system for testing the air permeability and color fastness of knitted garment fabrics. Background Technology

[0002] In the quality control of high-end sports or outdoor knitted fabrics, breathability and colorfastness are key indicators. Traditional methods require separate testing, which is time-consuming and wasteful of samples. To improve efficiency, existing technologies typically employ sequential testing, that is, first testing breathability, and then using the same sample to test colorfastness. This method works well on regular pure cotton or polyester knitted fabrics.

[0003] However, this testing method is prone to problems when applied to novel functional knitted fabrics with microporous surface finishing and dyeing with environmentally friendly plant dyes. The color fastness to rubbing measured after sequential testing is worse than the results obtained by testing directly with brand-new samples, and the fluctuations are extremely large, resulting in a loss of repeatability.

[0004] The study found that the problem stemmed from the air permeability test itself: the ring clamp of the tester applied excessive pressure to prevent air leakage, which compacted and deformed the knitted loop structure in the clamped area, damaging the microporous finishing layer. Simultaneously, the high-speed airflow continuously blowing through the fibers in the test center may strip away some loosely bonded plant dye particles. This resulted in an uneven surface on the same sample, with reduced dye adhesion in the ring area pressed by the clamp, while the loose dye in the central area was prematurely removed. In subsequent rubbing fastness tests, if the sample was taken from the compressed ring area, severe staining resulted in a failing grade; conversely, if taken from the central area, the results were better than the original sample. Due to the random sampling location, the final data fluctuated wildly and was unreliable.

[0005] Therefore, the urgent technical problem to be solved is: how to establish a method that can accurately assess both breathability and color fastness for this type of special fabric, so as to avoid localized non-uniform damage to the sample caused by the pre-test breathability test in the sequential test, thereby ensuring the accuracy and repeatability of the subsequent color fastness test results, and guaranteeing product certification and consumer experience. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, the present invention aims to provide a method for testing the air permeability and color fastness of knitted garment fabrics, which can simultaneously and accurately evaluate both air permeability and color fastness. This solves the problem in the existing sequential testing process where the air permeability test, due to the pressure applied by the annular clamp and the impact of the high-speed airflow in the center of the test area, causes localized and non-uniform changes in the microstructure and surface state of the fabric sample, thus seriously interfering with the accuracy and repeatability of the subsequent color fastness test results.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect, a method for testing the air permeability and color fastness of knitted garment fabrics, the method comprising the following steps: S1: Acquire multi-band spectral reflectance data of the surface of the material to be tested, and acquire the three-dimensional spatial geometric data of the surface of the material to be tested; S2: Based on the multi-band spectral reflectance data, extract optical characteristic parameters that characterize the dye distribution state and binding strength; S3: Based on the three-dimensional spatial geometric data, extract geometric feature parameters that characterize the micro-pore structure and surface morphology of the material surface; S4: Input the optical characteristic parameters and the geometric characteristic parameters into a preset performance evaluation model, and calculate the correlation between the optical characteristic parameters, the geometric characteristic parameters and the physical performance indicators through the performance evaluation model to obtain the predicted value of the air permeability performance and the predicted level of the color fastness performance of the material to be tested.

[0008] Preferably, step S1 includes: S11: Before the material to be tested enters the fabric rolling process, the material to be tested is scanned online in full width using a non-contact optical scanning module; S12: Within the wavelength range of 400 nm to 1000 nm, the reflectance spectral data of the surface of the material under test is collected at preset wavelength intervals and used as the multi-band spectral reflectance data. S13: Project an coded structured light pattern onto the surface of the material under test, and capture the deformation image generated by the coded structured light pattern on the surface of the material under test to reconstruct the three-dimensional geometric shape of the surface of the material under test, wherein the three-dimensional geometric shape is the three-dimensional spatial geometric data.

[0009] Preferably, step S2 includes: S21: Extract the color uniformity index and dye spectral characteristic value of the surface of the material to be tested from the reflection spectrum data, and use them as optical characteristic parameters to characterize the dye distribution state and binding strength, respectively.

[0010] Preferably, step S3 includes: S31: Calculate the micropore structure parameters and surface roughness of the material under test based on the three-dimensional geometric morphology, and use them as geometric feature parameters characterizing the micropore structure and surface morphology of the material surface, respectively.

[0011] Preferably, the procedure before step S4 includes: S041: Standardize the optical feature parameters and the geometric feature parameters to scale the feature values ​​of different dimensions to a uniform preset value range. S042: The optical feature parameters and the geometric feature parameters are smoothed using a noise reduction algorithm to remove random noise interference during the data acquisition process.

[0012] Preferably, step S4 includes: S41: Input the microporous structure parameters and the surface roughness as geometric feature parameters, and the color uniformity index and the dye spectral feature values ​​as optical feature parameters into the performance evaluation model; S42: In the performance evaluation model, the microporous structure parameters and the surface roughness are used to characterize the air flow resistance, calculate the correlation between them and the air permeability, and output the predicted value of the air permeability. S43: In the performance evaluation model, the color uniformity index and the dye spectral characteristic value are used to characterize the dye adhesion state, calculate the correlation between the dye and the color fastness performance, and output the color fastness performance prediction level.

[0013] Preferably, the method further includes: S5: Compare the color uniformity index with a preset uniformity threshold. When the color uniformity index is lower than the preset uniformity threshold, determine that the material under test has a risk of uneven dye distribution and mark the corresponding risk area. S6: Compare the micropore structure parameters with the preset structure range. When the micropore structure parameters deviate from the preset structure range, determine that the material under test has a risk of micropore abnormality and mark the corresponding risk area.

[0014] Preferably, after step S4, the method further includes: S7: The predicted values ​​of air permeability and color fastness performance are automatically compared with preset quality standards to determine whether the current batch of the material to be tested meets the quality requirements. S8: When the predicted value of air permeability or the predicted level of color fastness is within a preset critical range, a warning signal is issued and the corresponding potential risk area is highlighted on the control interface.

[0015] Preferably, the method further includes: S9: When the predicted air permeability value is lower than the preset quality standard, an adjustment command for the temperature, pressure or chemical additive concentration in the micropore formation process is output according to the deviation of the micropore structure parameters. S10: When the predicted color fastness performance level is lower than the preset quality standard, an adjustment instruction for the amount of fixing agent or the fixing time in the dyeing and fixing process is output based on the evaluation result of the dye adhesion state.

[0016] Secondly, a system for testing the air permeability and color fastness of knitted garment fabrics, characterized in that the system is used to implement the steps of any of the above methods, and the system includes: The data acquisition module is used to acquire multi-band spectral reflectance data of the surface of the material under test, and to acquire three-dimensional spatial geometric data of the surface of the material under test. The optical feature extraction module is used to extract optical feature parameters characterizing the distribution state and binding strength of the dye based on the multi-band spectral reflectance data. The geometric feature extraction module is used to extract geometric feature parameters that characterize the micropore structure and surface morphology of the material surface based on the three-dimensional spatial geometric data. The performance evaluation module is used to input the optical characteristic parameters and the geometric characteristic parameters into a preset performance evaluation model, and to calculate the correlation between the optical characteristic parameters, the geometric characteristic parameters and the physical performance indicators through the performance evaluation model, so as to obtain the predicted value of the air permeability performance and the predicted level of the color fastness performance of the material under test.

[0017] Compared with existing technologies, the beneficial effects of this invention are as follows: by acquiring multi-band spectral reflectance data and three-dimensional spatial geometric data of the surface of the material to be tested, optical feature parameters are extracted based on the spectral data, and geometric feature parameters are extracted based on the geometric data. These parameters are then input into a performance evaluation model to calculate the predicted values ​​of air permeability and color fastness performance. This avoids the local non-uniform damage to the sample caused by the pre-test air permeability test in the sequential test, and ensures the accuracy and repeatability of the subsequent color fastness test results, thus guaranteeing the product market certification and consumer experience. Attached Figure Description

[0018] Figure 1 The flowchart illustrates a method for testing the air permeability and color fastness of knitted garment fabrics provided by this invention.

[0019] Figure 2 The present invention provides a structural diagram of a system for testing the air permeability and color fastness of knitted garment fabrics.

[0020] Figure 3 This is a schematic diagram of a system for testing the air permeability and color fastness of knitted garment fabrics provided by the present invention.

[0021] The reference numerals are as follows: 201, Data Acquisition Module; 202, Optical Feature Extraction Module; 203, Geometric Feature Extraction Module; 204, Performance Evaluation Module. Detailed Implementation

[0022] The present invention will now be further described with reference to the accompanying drawings and specific embodiments: Please refer to Figure 1 A method for testing the air permeability and color fastness of knitted garment fabrics, the method comprising the following steps: S1: Acquire multi-band spectral reflectance data of the surface of the material under test, and acquire three-dimensional spatial geometric data of the surface of the material under test; S2: Based on multi-band spectral reflectance data, extract optical characteristic parameters that characterize the dye distribution state and binding strength; S3: Extract geometric feature parameters that characterize the micropore structure and surface morphology of the material surface based on three-dimensional spatial geometric data; S4: Input the optical and geometric characteristic parameters into the preset performance evaluation model. Calculate the correlation between the optical and geometric characteristic parameters and the physical performance indicators through the performance evaluation model to obtain the predicted values ​​of the air permeability and color fastness performance of the material under test.

[0023] The working principle of this method is based on the close correspondence between the macroscopic physical properties of a fabric and its microscopic optical and geometric characteristics. Breathability is essentially determined by the ease with which air passes through the fabric's microscopic pore structure. Therefore, the size, density, and uniformity of the micropores on the fabric surface, as well as the undulating shape of the yarn structure, collectively determine its breathability. Colorfastness directly depends on the adhesion state of dye molecules or particles on the fiber, including the uniformity of dye distribution and the bonding strength between the dye and the fiber. While this microscopic information is difficult to discern with the naked eye, it can be precisely captured using advanced optical sensing technology.

[0024] This method cleverly avoids the uncontrollable damage caused by mechanical clamping and airflow impact in traditional physical testing methods by introducing non-contact optical scanning technology. Acquiring multi-band spectral reflectance data is equivalent to performing a detailed chemical composition analysis on the fabric, revealing the microscopic distribution of dye particles on the fibers and identifying areas of localized sparseness or aggregation—areas often weak points in colorfastness. Acquiring three-dimensional spatial geometric data is like creating a high-precision microscopic topographic map of the fabric surface, enabling accurate measurement of key factors determining breathability, such as the shape, size, depth, and distribution density of micropores, as well as the density of the knitted loop structure.

[0025] Once these crucial microscopic feature data are collected, the subsequent processing is handled by a pre-defined performance evaluation model. In one specific implementation, the performance evaluation model is pre-constructed using a supervised learning algorithm. The construction process includes: collecting multiple batches of knitted garment fabric samples, extracting optical feature parameters (color uniformity index, dye spectral feature values) and geometric feature parameters (microporous structure parameters, surface roughness) through non-contact optical scanning, and determining the actual air permeability (unit: mm / s) and color fastness grade (1-5) of each sample according to standard physical testing methods (such as ASTM D737, AATCC 8). Using the above feature parameters as input and the measured performance index as output, an ensemble model of random forest regression and gradient boosting machine (XGBoost) is used for training. Hyperparameters are optimized through five-fold cross-validation, ultimately obtaining a prediction model with stable mapping relationships.

[0026] The performance evaluation model takes a standardized and denoised feature vector as input, including: equivalent micropore diameter, micropore distribution density, root mean square value of surface roughness, coefficient of variation of color uniformity, intensity and half-width at half-maximum of dye characteristic absorption peaks. The model contains two parallel processing branches: the first branch, based on geometric feature parameters, simulates the flow resistance of air through micropores according to fluid dynamics principles, outputting a continuous prediction of air permeability; the second branch, based on optical feature parameters, analyzes the morphological changes of dye absorption peaks to assess the dye-fiber binding strength, and, combined with the color uniformity index, outputs a discrete prediction of color fastness performance (mapped to levels 1-5 using the Softmax function). The model supports online incremental learning; when new fabrics or processes are introduced, the model can be fine-tuned using new samples to maintain prediction accuracy.

[0027] By mapping and predicting microscopic features to macroscopic performance, this method shifts the quality control process from time-consuming and labor-intensive offline laboratory testing to real-time online evaluation at the end of the production line. This shift not only fundamentally solves the unique problem of non-uniform sample damage caused by pre-testing for air permeability, which interferes with color fastness test results, but also enables rapid response and closed-loop control of the production process, greatly improving production efficiency and product quality stability.

[0028] Preferably, step S1 includes: S11: Before the material to be tested enters the fabric rolling process, the material to be tested is scanned online in full width using a non-contact optical scanning module; S12: Collect the reflection spectrum data of the surface of the material under test in the wavelength range of 400 nm to 1000 nm at preset wavelength intervals, as multi-band spectral reflection data; S13: Project an coded structured light pattern onto the surface of the material to be tested, and capture the deformation image generated by the coded structured light pattern on the surface of the material to be tested, so as to reconstruct the three-dimensional geometric shape of the surface of the material to be tested, which is three-dimensional spatial geometric data.

[0029] Specifically, the steps for acquiring multi-band spectral reflectance data and three-dimensional spatial geometric data of the material under test involve performing an online full-width scan of the material before it enters the fabric rolling process using a non-contact optical scanning module. In a real production line environment, this non-contact optical scanning module is typically integrated into a gantry-type frame structure spanning above the fabric conveyor belt. As the fabric, having undergone dyeing and finishing processes, passes beneath this frame at a stable speed, the scanning module continuously and comprehensively acquires data. This online full-width scanning method ensures that every area of ​​the entire fabric can be inspected, avoiding omissions that may occur with traditional sampling inspections and achieving 100% quality control.

[0030] During the scanning process, reflectance spectral data of the material surface is first collected within the wavelength range of 400nm to 1000nm at preset wavelength intervals, serving as multi-band spectral reflectance data. The selection of this wavelength range has clear physical significance. The visible light band of 400nm to 700nm is primarily used to obtain color information of the fabric, forming the basis for evaluating the uniformity of dye distribution. The near-infrared band of 700nm to 1000nm reflects the chemical composition of the material, particularly the binding state between dye molecules and fiber macromolecules. By analyzing the spectral absorption peaks or reflectance valleys in specific bands, the binding strength of the dye can be indirectly assessed. A preset wavelength interval, for example, set to 5nm, means that more than 100 data points at different wavelengths are collected across the entire spectrum, thus forming a fine spectral curve and providing rich information for subsequent feature extraction.

[0031] To achieve this goal, a hyperspectral imaging unit is integrated within the non-contact optical scanning module. This unit can employ a pushbroom hyperspectral camera, which works by imaging a line along the width of the conveyor belt at every instant the fabric moves. Simultaneously, a built-in grating or prism disperses the light signal from each spatial point along this line into a complete spectrum. As the fabric continues to move, these line scan data are continuously acquired and stitched together, ultimately forming a three-dimensional data cube, where two dimensions are spatial coordinates and the third dimension is the spectral wavelength.

[0032] Simultaneously, to acquire three-dimensional spatial geometric data, the scanning module projects an coded structured light pattern onto the surface of the material under test and captures the deformed images produced by the coded structured light pattern on the surface of the material under test to reconstruct the three-dimensional geometric topography of the surface of the material under test. This three-dimensional geometric topography is the three-dimensional spatial geometric data. This process is completed by the three-dimensional structured light scanning unit inside the module. This unit typically contains a high-frequency pattern projector and one or more high-speed industrial cameras. The projector projects a series of specially coded grating patterns, such as sinusoidal stripes or Gray code patterns, onto the fast-moving fabric surface. Due to the microscopic undulations on the fabric surface, these projected patterns will be distorted and deformed accordingly on the fabric. The high-speed industrial camera synchronously captures these deformed pattern images from a different angle than the projector. By analyzing the degree of deformation of the pattern in the image and using the principle of triangulation, the three-dimensional spatial coordinates of every point on the fabric surface can be accurately calculated, thereby reconstructing the microscopic geometric topography of the entire fabric surface with sub-millimeter accuracy.

[0033] As a specific implementation method, in the detection scenario of ultra-high elastic knitted fabrics, due to the ease with which the fabric can develop edge curling and local surface undulations during transport, an adaptive frequency Gray code composite sinusoidal fringe projection scheme is introduced during the reconstruction process. The projection device dynamically adjusts the period width of the projected stripes based on feedback from the fabric's movement speed. When the industrial camera captures images of severe deformation caused by fabric wrinkles, the algorithm processes the phase information using unwrapping technology, filtering out macroscopic fabric surface jitter interference and retaining only microscopic data on knitted loops and micropore depth. This method of maintaining high-precision geometric reconstruction in a dynamically changing environment demonstrates that the 3D acquisition scheme can adapt to fabrics with different mechanical strengths and elastic properties, effectively solving the measurement errors caused by fabric deformation during dynamic continuous production and ensuring the data foundation for subsequent air permeability prediction.

[0034] After acquiring the raw multi-band spectral reflectance data, the next step is to extract optical characteristic parameters that characterize the dye distribution and binding strength. Specifically, this step includes: S21: Extracting the color uniformity index of the surface of the material under test and the dye spectral characteristic values ​​from the reflectance spectral data, which are used as optical characteristic parameters characterizing the dye distribution and binding strength, respectively.

[0035] The logic for extracting binding strength is based on the inverse application of the Lambert-Beer law. The reflectance of the test material at the main absorption peak wavelength of a specific plant dye is analyzed and normalized to the reflectance of a non-absorption reference band. Since the strength of the dye-fiber bond affects the vibrational energy level of the chromophore, thus altering the full width at half maximum (FWHM) and peak intensity of the absorption spectrum, the strength of the dye bond between fiber molecules can be quantitatively assessed by extracting the slope of the first derivative and the integral area of ​​the absorption peak. The distribution pattern is determined by calculating the spatial correlation function of pixels in the multi-band image; a short correlation length and large variance indicate uneven dye distribution.

[0036] The purpose of extracting color uniformity indicators is to quantify the consistency of dye distribution on the fabric at both the macro and micro scales. A specific calculation method involves first dividing the full-width scan image of the fabric into several small computational units, for example, one unit per square centimeter. Within each unit, a characteristic wavelength most sensitive to the color of the dye used is selected, typically the wavelength corresponding to the main absorption peak of the dye in the visible light range. Then, the reflectance values ​​of all pixels within the unit at this characteristic wavelength are calculated, and further statistical parameters of these values, such as the standard deviation or coefficient of variation, are calculated. The coefficient of variation, the ratio of the standard deviation to the mean, is a dimensionless index of relative dispersion. A larger coefficient of variation for a region indicates greater color depth variation within that region, meaning a more uneven dye distribution, which may indicate that the region is more prone to color fading during subsequent friction or washing.

[0037] The core of extracting the spectral characteristic values ​​of dyes lies in assessing the strength of the bond between the dye and the fiber. This is typically achieved by analyzing the specific morphology of the spectral curve. For example, a specific plant dye may have a characteristic absorption peak in a certain near-infrared band, while the fiber substrate itself also has a relatively stable reflection characteristic at another different wavelength. By calculating the intensity ratio of these two characteristic peaks, a parameter indirectly reflecting the amount of dye adhered can be obtained. Under ideal fixation processes, the dye and fiber are fully bonded, and this ratio will be at a stable and high level. If this ratio is low, it may mean that the dye fixation is insufficient, and a large amount of dye is merely physically attached to the fiber surface, i.e., there is a lot of floating dye. Such fabrics will inevitably have poor colorfastness.

[0038] Similarly, after acquiring the three-dimensional geometric topography data, it is necessary to extract geometric feature parameters that characterize the microscopic pore structure and surface morphology of the material surface. Specifically, step S3 includes: S31: Calculate the micropore structure parameters and surface roughness of the material under test based on the three-dimensional geometric morphology, and use them as geometric characteristic parameters to characterize the micropore structure and surface morphology of the material surface.

[0039] By meshing the 3D point cloud data, regions with negative height values ​​in the local coordinate system are identified as micropores. The boundary perimeter and internal pixel area of ​​the connected domains are calculated to obtain the equivalent circle diameter. Gaussian curvature is used to analyze surface undulations, and the root mean square height is used as a quantitative indicator of surface roughness, thus transforming the abstract morphology into a feature vector that can participate in numerical calculations. For connectivity evaluation, a Delaunay triangulation is constructed at the center points of the micropores to calculate the topological distance between adjacent pores and the throat diameter, serving as the basis for calculating air penetration resistance.

[0040] The calculation of micropore structure parameters is based on reconstructed 3D point cloud data or depth images. Image processing algorithms, such as setting a height threshold, are applied to segment the recessed areas on the fabric surface, and these connected recessed areas are identified as micropores. Once the micropores are identified, the geometric parameters of each micropore can be quantified, including calculating its equivalent diameter, average depth, area, and shape regularity. Based on this, the number of micropores per unit area, i.e., the micropore density, can be statistically analyzed, and the connectivity between micropores can be assessed, such as calculating the average distance between adjacent micropores. These parameters collectively constitute a comprehensive description of the material's microporous structure and are the direct physical basis for predicting air permeability.

[0041] Surface roughness calculation is also based on three-dimensional geometric topography data. A commonly used quantitative indicator is the root mean square (RMS) value of the height of all points on the fabric surface within a specific region, relative to the average height. The larger this RMS value, the more pronounced the surface undulations, i.e., the rougher the surface. Surface roughness not only affects the resistance of airflow across the fabric surface, thus indirectly affecting breathability, but it can also affect colorfastness to rubbing, as a rougher surface may exert a greater mechanical stripping effect on the dye during rubbing.

[0042] Preferably, the procedure before step S4 includes: S041: Standardize the optical and geometric feature parameters to scale the feature values ​​of different dimensions to a uniform preset range. S042: Use a noise reduction algorithm to smooth the optical and geometric feature parameters in order to remove random noise interference during the data acquisition process.

[0043] Before inputting the extracted optical and geometric feature parameters into the performance evaluation model, data preprocessing is typically required to improve the model's computational stability and prediction accuracy. This preprocessing includes standardizing the optical and geometric feature parameters to scale the feature values ​​of different dimensions to a uniform, predefined numerical range. Standardization is necessary because the extracted feature parameters often have different physical units and numerical ranges. For example, the unit for micropore diameter might be micrometers, with values ​​ranging from tens to hundreds; while the coefficient of variation for color uniformity is a dimensionless decimal, typically between 0 and 0.1. Without standardization, features with larger numerical ranges would dominate the model's computation, while features with smaller ranges might be ignored. Standardization, such as through min-max scaling that linearly maps all feature values ​​to the zero-to-one range, eliminates the influence of these dimensional and scale differences, ensuring that all features are treated fairly in the model.

[0044] Furthermore, the preprocessing process includes smoothing optical and geometric feature parameters using noise reduction algorithms to remove random noise interference during data acquisition. An adaptive median filtering algorithm is employed for spatial domain processing of the feature matrix. For each pixel's feature value, the median is found within a preset sliding window, and anomalous impulse noise is replaced. This suppresses isolated outliers caused by scanning environment flicker or machine vibration while preserving abrupt changes in fabric edges and the true microporous structure. For periodic stripe noise in the geometric data, a fast Fourier transform is used to convert the data to frequency band space. A notch filter is used to remove mechanical vibration interference at specific frequencies, and then an inverse transform is used to restore the true surface morphology parameters. During high-speed online scanning, due to the sensor's own electronic noise, minor fluctuations in ambient light, or slight fabric vibrations, the acquired raw data inevitably contains some random noise points. This noise interferes with accurate feature extraction. Therefore, after feature extraction, smoothing filtering algorithms, such as median filtering or Gaussian filtering, can be applied to smooth the spatial distribution of feature parameters. For example, median filters can effectively remove isolated impulse noise while preserving edge information well, making the processed feature data more reflective of the true state of the fabric.

[0045] After preprocessing, these clean and scale-uniform feature parameters can be used as input to the performance evaluation model for final performance prediction. Specifically, step S4 includes: S41: Input the microporous structure parameters and surface roughness as geometric feature parameters, and the color uniformity index and dye spectral feature values ​​as optical feature parameters into the performance evaluation model; S42: In the performance evaluation model, the air flow resistance is characterized by microporous structure parameters and surface roughness, the correlation between them and air permeability is calculated, and the predicted value of air permeability is output. S43: In the performance evaluation model, the color uniformity index and dye spectral characteristic values ​​are used to characterize the dye adhesion state, calculate the correlation between it and the color fastness performance, and output the color fastness performance prediction level.

[0046] In one specific implementation, the performance evaluation model can be a multiple linear regression model. For predicting air permeability, the model can be expressed as a linear equation; for example, the predicted air permeability value equals a constant term, plus the square of the average micropore diameter multiplied by coefficient A, plus the micropore distribution density multiplied by coefficient B, and minus the surface roughness multiplied by coefficient C. These coefficients A, B, and C are obtained through regression analysis on the training data. The physical meaning of this model is clear: air permeability is positively correlated with the sum of the squares of the pore size and its density, and negatively correlated with surface roughness.

[0047] In addition to outputting the final performance prediction, this method also includes real-time identification and labeling of potential risks. Specifically, the method also includes: S5: Compare the color uniformity index with the preset uniformity threshold. When the color uniformity index is lower than the preset uniformity threshold, it is determined that the material under test has a risk of uneven dye distribution, and the corresponding risk area is marked. S6: Compare the micropore structure parameters with the preset structure range. When the micropore structure parameters deviate from the preset structure range, it is determined that the material under test has a risk of micropore abnormality, and the corresponding risk area is marked.

[0048] For example, if the color variation coefficient of a certain area exceeds a preset 5%, even if its predicted overall color fastness level is still within an acceptable range, that area will be marked as a high-risk area. This provides quality control personnel with more refined diagnostic information, enabling them to focus on potential local defects.

[0049] Similarly, this method also includes comparing the micropore structure parameters with a preset structural range. When the micropore structure parameters deviate from the preset range, it is determined that the material under test has a risk of micropore abnormalities, and the corresponding risk area is marked. For example, according to process requirements, the average diameter of the micropores in the fabric should be controlled between 0.08 mm and 0.12 mm. If the average diameter of the micropores in a certain area is detected to be less than 0.08 mm, there may be a risk of micropore blockage, which will lead to a decrease in air permeability; if it is greater than 0.12 mm, it may mean that the finishing process is out of control, which may affect the strength and appearance of the fabric. Through this comparison, process fluctuations in the production process can be detected in a timely manner.

[0050] After obtaining the performance prediction results and risk markers, automated quality assessment and early warning will be performed. Specifically, after step S4, the following steps will also be taken: S7: Automatically compare the predicted values ​​of air permeability and color fastness with the preset quality standards to determine whether the current batch of the material to be tested meets the quality requirements. S8: When the predicted value of air permeability or the predicted level of color fastness is within the preset critical range, a warning signal is issued and the corresponding potential risk area is highlighted on the control interface.

[0051] The quality standards can be preset and stored according to different product orders or customer requirements. The system will compare the real-time predicted values ​​with the standard values ​​and directly display the pass or fail judgment results on the control interface.

[0052] Furthermore, as a specific implementation, in full-width fabric inspection, the highlighting function is achieved by overlaying a rendered heat map onto a virtual fabric map. When the color uniformity index at a certain location is detected to be below the safety threshold, a red semi-transparent spot will appear in real time at the corresponding coordinates on the control interface. The intensity of the spot represents the severity of the risk; that is, the lower the predicted colorfastness grade, the brighter the color patch. Quality inspectors can click on the highlighted area to view the abnormal fluctuations in the spectral reflectance curve at that location. This visualization method supports the precise location of minute dye aggregation points in rolls of fabric thousands of meters long, providing guidance for subsequent cutting and grading even if the relevant area has not yet constituted a macroscopically obvious defect. This solution demonstrates the spatial positioning capability of the early warning logic in complex production environments, providing an intuitive decision-making basis for refined quality management on the production floor. This critical range, for example, is a range of 5% above or below the pass line. When the predicted result falls into this range, it means that the product quality is on the edge of pass and fail, posing a significant risk. At this point, an audible and visual alarm will be triggered, and these critical areas will be highlighted in a striking color, such as yellow, on the interface displaying the full-width quality map of the fabric. This will prompt operators and quality inspectors to pay close attention or conduct small-batch sampling verification, thereby achieving proactive management of quality risks.

[0053] The ultimate goal of this method is not merely detection and alarm, but to provide data-driven decision support for optimizing production processes, forming a complete quality control closed loop. To this end, the method also includes: S9: When the predicted air permeability value is lower than the preset quality standard, an adjustment command for the temperature, pressure or chemical additive concentration in the micropore formation process is output according to the deviation of the micropore structure parameters. S10: When the predicted color fastness performance level is lower than the preset quality standard, an adjustment instruction for the amount of fixing agent or the fixing time in the dyeing and fixing process is output based on the evaluation results of the dye adhesion status.

[0054] For example, if the reason for the failure to meet the air permeability standard is that the micropore diameter is generally too small, a suggested instruction may be issued, prompting the operator to appropriately increase the heat setting temperature during the finishing process or reduce the pressure of the rollers to facilitate the full formation of micropores.

[0055] As a specific implementation method, the adjustment instruction generation process is integrated into the closed-loop control architecture of the production line. When the predicted colorfastness level is lower than the preset quality level standard, and the hyperspectral characteristics show insufficient peak intensity of dye binding on the fiber, the logic processing module automatically calculates the deviation between the current measurement value and the target value. By calling the pre-stored process sensitivity matrix, it calculates the increase in fixing agent concentration or the amount of fixing time compensation required to improve colorfastness by one level. This instruction is directly sent to the automatic dispensing unit of the dyeing equipment via the industrial bus to achieve fine-tuning of the fixing solution flow rate. During the period after the adjustment instruction is issued, the monitoring equipment continuously compares the newly acquired spectral characteristics to verify whether the process adjustment has produced the expected repair effect, thereby achieving automatic closed-loop adjustment of production quality without manual intervention, demonstrating the breadth and depth of application of this technical solution in the integrated environment of intelligent factories. For example, if the analysis of dye spectral characteristics indicates that the binding strength between the dye and the fiber is generally insufficient, it may be recommended to increase the amount of fixing agent or appropriately extend the fixing treatment time to promote more complete chemical bonding between the dye and the fiber, thereby improving colorfastness. These specific and actionable adjustment instructions transform quality control from a reactive, ex-post screening process into a proactive optimization integrated into the production process.

[0056] Please refer to Figure 2 , Figure 3 This application also provides a system for testing the air permeability and color fastness of knitted garment fabrics, characterized in that the system is used to implement the steps of any of the above methods, and the system includes: The data acquisition module 201 is used to acquire multi-band spectral reflectance data of the surface of the material to be tested, and to acquire three-dimensional spatial geometric data of the surface of the material to be tested. The optical feature extraction module 202 is used to extract optical feature parameters that characterize the distribution state and binding strength of dyes based on multi-band spectral reflectance data. The geometric feature extraction module 203 is used to extract geometric feature parameters that characterize the micro-pore structure and surface morphology of the material surface based on three-dimensional spatial geometric data. The performance evaluation module 204 is used to input optical characteristic parameters and geometric characteristic parameters into a preset performance evaluation model. The performance evaluation model is used to calculate the correlation between optical characteristic parameters, geometric characteristic parameters and physical performance indicators to obtain the predicted value of air permeability and the predicted level of color fastness of the material under test.

[0057] Specifically, the data acquisition module 201 may consist of one or more physical sensors, such as a spectral imager and a 3D structured light scanner. The spectral imager is responsible for acquiring the reflectance spectrum information of the material surface in a non-contact manner within a preset wavelength range; the 3D structured light scanner reconstructs the 3D geometry of the material surface by projecting and capturing structured light patterns. These sensors can be integrated on a unified scanning platform or operate synchronously as independent units to ensure the integrity and synchronization of data acquisition.

[0058] The optical feature extraction module 202 can be a standalone computing unit, such as an embedded processor or an industrial computer, internally running specialized spectral analysis and image processing algorithms. This module receives multi-band spectral reflectance data from the data acquisition module and processes it, for example, by calculating color uniformity indices or analyzing spectral absorption peaks at specific wavelengths, thereby quantifying the distribution of dyes on the material surface and their binding strength to fibers.

[0059] The geometric feature extraction module 203 can be integrated with the optical feature extraction module in the same computing unit, or it can exist as an independent computing unit. This module receives three-dimensional spatial geometric data from the data acquisition module and performs three-dimensional point cloud processing and surface morphology analysis on it. For example, it calculates parameters such as porosity, pore size distribution, or surface roughness of the material to accurately characterize the microstructure features of the material.

[0060] The performance evaluation module 204 can be a high-performance computing server or an industrial control system, internally deployed with a pre-trained machine learning model or deep learning model. This module receives feature parameters from the optical feature extraction module and the geometric feature extraction module, and uses these parameters to perform complex correlation calculations through the model, ultimately outputting predicted values ​​for the air permeability and color fastness performance of the material under test. This module may also include a user interface for displaying the test results and system status.

[0061] The air permeability and color fastness testing system for knitted garment fabrics proposed in this application represents a significant technological advancement compared to traditional testing methods. Traditional methods rely on contact testing, which can easily cause physical damage and dye stripping to novel functional knitted fabrics, resulting in a substantial reduction in the accuracy and repeatability of the test results. The system in this application, through integrated non-contact optical scanning technology, achieves simultaneous and non-destructive acquisition of multi-band spectral reflectance data and three-dimensional spatial geometric data of the material. These data are refined through dedicated optical feature extraction module 202 and geometric feature extraction module 203, and comprehensively predicted by performance evaluation module 204, fundamentally avoiding sample damage and ensuring the reliability of the test results. Therefore, this system provides an efficient, accurate, and non-destructive solution for the quality control of high-end sports or outdoor knitted fabrics, significantly improving the quality control level in the production process.

[0062] For those skilled in the art, various other corresponding changes and modifications can be made based on the technical solutions and concepts described above, and all such changes and modifications should fall within the protection scope of the claims of this invention.

Claims

1. A method for testing the air permeability and color fastness of knitted garment fabrics, characterized in that, The method includes the following steps: S1: Acquire multi-band spectral reflectance data of the surface of the material to be tested, and acquire the three-dimensional spatial geometric data of the surface of the material to be tested; S2: Based on the multi-band spectral reflectance data, extract optical characteristic parameters that characterize the dye distribution state and binding strength; S3: Based on the three-dimensional spatial geometric data, extract geometric feature parameters that characterize the micro-pore structure and surface morphology of the material surface; S4: Input the optical characteristic parameters and the geometric characteristic parameters into a preset performance evaluation model, and calculate the correlation between the optical characteristic parameters, the geometric characteristic parameters and the physical performance indicators through the performance evaluation model to obtain the predicted value of the air permeability performance and the predicted level of the color fastness performance of the material to be tested.

2. The method for testing the air permeability and color fastness of knitted garment fabrics according to claim 1, characterized in that, Step S1 includes: S11: Before the material to be tested enters the fabric rolling process, the material to be tested is scanned online in full width using a non-contact optical scanning module; S12: Within the wavelength range of 400 nm to 1000 nm, the reflectance spectral data of the surface of the material under test is collected at preset wavelength intervals and used as the multi-band spectral reflectance data. S13: Project an coded structured light pattern onto the surface of the material under test, and capture the deformation image generated by the coded structured light pattern on the surface of the material under test to reconstruct the three-dimensional geometric shape of the surface of the material under test, wherein the three-dimensional geometric shape is the three-dimensional spatial geometric data.

3. The method for testing the air permeability and color fastness of knitted garment fabrics according to claim 2, characterized in that, Step S2 includes: S21: Extract the color uniformity index and dye spectral characteristic value of the surface of the material to be tested from the reflection spectrum data, and use them as optical characteristic parameters to characterize the dye distribution state and binding strength, respectively.

4. The method for testing the air permeability and color fastness of knitted garment fabrics according to claim 3, characterized in that, Step S3 includes: S31: Calculate the micropore structure parameters and surface roughness of the material under test based on the three-dimensional geometric morphology, and use them as geometric feature parameters characterizing the micropore structure and surface morphology of the material surface, respectively.

5. The method for testing the air permeability and color fastness of knitted garment fabrics according to claim 1, characterized in that, Step S4 includes: S041: Standardize the optical feature parameters and the geometric feature parameters to scale the feature values ​​of different dimensions to a uniform preset value range. S042: The optical feature parameters and the geometric feature parameters are smoothed using a noise reduction algorithm to remove random noise interference during the data acquisition process.

6. The method for testing the air permeability and color fastness of knitted garment fabrics according to claim 4, characterized in that, Step S4 includes: S41: Input the microporous structure parameters and the surface roughness as geometric feature parameters, and the color uniformity index and the dye spectral feature values ​​as optical feature parameters into the performance evaluation model; S42: In the performance evaluation model, the microporous structure parameters and the surface roughness are used to characterize the air flow resistance, calculate the correlation between them and the air permeability, and output the predicted value of the air permeability. S43: In the performance evaluation model, the color uniformity index and the dye spectral characteristic value are used to characterize the dye adhesion state, calculate the correlation between the dye and the color fastness performance, and output the color fastness performance prediction level.

7. The method for testing the air permeability and color fastness of knitted garment fabrics according to claim 4, characterized in that, The method further includes: S5: Compare the color uniformity index with a preset uniformity threshold. When the color uniformity index is lower than the preset uniformity threshold, determine that the material under test has a risk of uneven dye distribution and mark the corresponding risk area. S6: Compare the micropore structure parameters with the preset structure range. When the micropore structure parameters deviate from the preset structure range, determine that the material under test has a risk of micropore abnormality and mark the corresponding risk area.

8. The method for testing the air permeability and color fastness of knitted garment fabrics according to claim 1, characterized in that, Step S4 is followed by: S7: The predicted values ​​of air permeability and color fastness are automatically compared with preset quality standards to determine whether the current batch of the material to be tested meets the quality requirements. S8: When the predicted value of air permeability or the predicted level of color fastness is within a preset critical range, a warning signal is issued and the corresponding potential risk area is highlighted on the control interface.

9. The method for testing the air permeability and color fastness of knitted garment fabrics according to claim 8, characterized in that, The method further includes: S9: When the predicted air permeability value is lower than the preset quality standard, an adjustment command for the temperature, pressure or chemical additive concentration in the micropore formation process is output according to the deviation of the micropore structure parameters. S10: When the predicted color fastness performance level is lower than the preset quality standard, an adjustment instruction for the amount of fixing agent or the fixing time in the dyeing and fixing process is output based on the evaluation result of the dye adhesion state.

10. A system for testing the air permeability and color fastness of knitted garment fabrics, characterized in that, The system is used to implement the steps of any one of the methods of claims 1-9 above, and the system includes: The data acquisition module is used to acquire multi-band spectral reflectance data of the surface of the material under test, and to acquire three-dimensional spatial geometric data of the surface of the material under test. The optical feature extraction module is used to extract optical feature parameters characterizing the distribution state and binding strength of the dye based on the multi-band spectral reflectance data. The geometric feature extraction module is used to extract geometric feature parameters that characterize the micropore structure and surface morphology of the material surface based on the three-dimensional spatial geometric data. The performance evaluation module is used to input the optical characteristic parameters and the geometric characteristic parameters into a preset performance evaluation model, and to calculate the correlation between the optical characteristic parameters, the geometric characteristic parameters and the physical performance indicators through the performance evaluation model, so as to obtain the predicted value of the air permeability performance and the predicted level of the color fastness performance of the material under test.

Images