A fabric air permeability performance determination method based on fabric aperture optical characteristic analysis

Abstract

The application discloses a fabric air permeability performance determination method based on fabric pore optical characteristic analysis, and particularly relates to the technical field of textile fabric performance detection, and comprises the following steps: tensionless and lossless fixing of the fabric, in-situ environment balance and optical interference elimination, and synchronous collection of in-situ parameters; multi-modal optical coaxial synchronous scanning of the same area, extraction of air permeability core features and grouping; construction of a special function based on airflow permeation physical laws, calculation of pore effective air permeability contribution, through resistance correction and form resistance correction values; nonlinear coupling output of corresponding values through adaptive weight calibration and environment correction factor calculation; and output of air permeability performance final determination values and quantitative grading through secondary fusion correction combined with physical constraint loss. The application realizes non-contact precise detection of the air permeability performance of the fabric in a native state, and overcomes the defects of the traditional method, such as destruction of pore structures, inability to analyze three-dimensional connected features and poor environmental adaptability.

Description

Technical Field

[0001] This invention relates to the field of textile fabric performance testing technology, and more specifically, to a method for determining the air permeability of fabrics based on the analysis of the optical characteristics of fabric pores. Background Technology

[0002] Breathability is an important indicator of fabric comfort, especially for sportswear, protective clothing, and high-end apparel fabrics, as their breathability directly affects the body's thermal and moisture balance. Currently, fabric breathability testing mainly uses the pressure difference method, which calculates the breathability rate by measuring the airflow rate passing vertically through a unit area of ​​the sample per unit time under specified pressure difference conditions. This method is a standardized testing method commonly used in the industry, with advantages such as simple operation and intuitive results, and is widely used in fabric production, research and development, and quality control. In addition, some studies attempt to indirectly infer the breathability of fabrics through methods such as microscopic image analysis and porosity testing, trying to establish a correlation model between structure and breathability.

[0003] However, in practical use, it still has some drawbacks. For example, the traditional differential pressure method is a destructive test. During the test, the airflow is forced through the fabric, which can easily cause irreversible deformation of the pore structure and cannot reflect the true breathability behavior of the fabric in its natural state. Static analysis methods based on microscopic images usually only focus on the two-dimensional pore morphology and ignore the continuity of pores in the thickness direction and the three-dimensional connectivity. Moreover, the test process often requires cutting the sample and applying tension to fix it, which changes the original structure of the pores and causes the analysis results to deviate from the true breathability performance. Existing technologies mostly use a single-modal data acquisition method, which fails to achieve effective fusion of multi-source features and lacks an environmental adaptive correction mechanism. The accuracy, reproducibility and generalization ability of the test results are difficult to meet the refined needs of high-end fabric research and development and production. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method for determining the air permeability of fabrics based on the optical characteristics analysis of fabric pores, and solves the problems mentioned in the background art through the following scheme.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for determining the air permeability of fabrics based on the optical characteristics analysis of fabric pores, comprising:

[0006] S1: Without changing the original pore structure of the fabric under test or causing irreversible deformation, complete the tension-free and non-destructive fixation of the fabric under test, the in-situ balance and adaptation of the test environment, the elimination of optical imaging interference, and simultaneously collect and bind the in-situ basic parameters of the test area to complete the pre-processing qualification verification and obtain the pre-processed qualified fabric under test and the corresponding in-situ basic parameters.

[0007] S2: Perform multimodal optical coaxial synchronous scanning on the same detection area, collect raw data related to pore morphology, three-dimensional connectivity, and material spectrum, extract core features related to fabric breathability, and accurately group the features according to subsequent calculation purposes to obtain a grouped breathability-related feature dataset.

[0008] S3: Based on the physical laws of airflow permeation in textile fabrics, and combined with the air permeability-related feature dataset obtained in S2, construct multiple sets of exclusive physical functions corresponding to pore air permeability contribution and air permeability resistance correction, and output multiple core feature values ​​directly related to the air permeability performance of the fabric through standardized calculation and data verification.

[0009] S4: Call the core feature value obtained in S3, combine it with the in-situ basic parameters collected in S1, complete the adaptive calibration of coupling weight and the calculation of environmental correction factor, complete the collaborative fusion of multiple core feature values ​​through adaptive nonlinear coupling algorithm, and output the coupling corresponding value directly related to the fabric air permeability after standardization.

[0010] S5: Call the coupling corresponding value obtained in S4, combine the pre-calibrated fabric breathability physical constraint parameters with the breathability-related feature dataset obtained in S2, calculate the physical constraint loss term and complete the secondary fusion correction, output the final judgment value of fabric breathability performance, and at the same time complete the quantitative grading of fabric breathability performance based on the preset grading standard.

[0011] The technical effects and advantages of this invention are as follows:

[0012] 1. This invention achieves multimodal optical detection without altering the original pore structure of the fabric or causing irreversible deformation by using tension-free in-situ clamping, pneumatic adsorption fixation, and polarized light extinction pretreatment. It solves the problems of traditional pressure difference method detection damaging the pore structure and microscopic analysis requiring cutting and fixing, which leads to structural distortion. This ensures that the test results truly reflect the breathability of the fabric in its natural state.

[0013] 2. This invention employs a multimodal coaxial synchronous scanning technology that combines transmission-reflection dual-path microscopy, laser confocal tomography, and near-infrared spectroscopy. This technology enables the simultaneous acquisition of two-dimensional morphology, three-dimensional connectivity, and material spectral characteristics of pores in the same detection area. It can accurately distinguish between through-holes, semi-open holes, and closed holes, and fully extract 12 core features. This overcomes the shortcomings of existing technologies that only focus on two-dimensional pore morphology and ignore the continuity in the thickness direction.

[0014] 3. Based on the physical laws of airflow permeation in textile fabrics, this invention constructs three sets of exclusive physical functions: effective air permeability contribution of pores, correction of pore penetration resistance, and correction of pore morphology resistance. It establishes a physical relationship between microscopic pore characteristics and macroscopic air permeability performance, and achieves multi-feature fusion through an adaptive nonlinear coupling algorithm, thus solving the problems of fixed feature fusion methods and lack of physical meaning in existing technologies.

[0015] 4. This invention introduces an environmental adaptive correction mechanism, which dynamically calculates the environmental correction factor based on the fiber moisture absorption expansion coefficient and the relative humidity of the detection environment, and adaptively calibrates the coupling weight in combination with the actual thickness of the fabric, effectively eliminating the influence of environmental temperature and humidity changes on the detection results and improving the environmental adaptability and detection accuracy of the method.

[0016] 5. This invention calculates the physical constraint loss term and performs secondary fusion correction, linearly regressing the coupled corresponding value, the intensity of the near-infrared spectral characteristic absorption peak, the warp / weft pore orientation difference, and the fabric type label, and introduces physical constraint weights for correction, ensuring that the final judgment value conforms to the physical laws of the breathability of textile fabrics. At the same time, a monthly incremental update mechanism is established to continuously optimize the regression coefficient and physical constraint weights, thereby improving the method's generalization ability to new fabrics and new processes.

[0017] 6. The present invention outputs a fabric breathability performance judgment value that is consistent with the national standard breathability unit, and establishes a 5-level quantitative grading standard. At the same time, it performs strict data reliability verification to ensure that the test results are accurate, reproducible and traceable, providing a scientific basis for fabric research and development, production quality control and functional evaluation. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0019] Figure 2 This is a schematic diagram of the preprocessing and in-situ parameter acquisition structure of the present invention.

[0020] Figure 3 This is a schematic diagram of the multimodal scanning and feature grouping structure of the present invention.

[0021] Figure 4 This is a schematic diagram of the core computation, coupling, secondary fusion, and hierarchical structure of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] refer to Figures 1-4 The method for determining the air permeability of fabrics based on the optical characteristics analysis of fabric pores, as shown, includes:

[0024] S1: Without changing the original pore structure of the fabric under test or causing irreversible deformation, complete the tension-free and non-destructive fixation of the fabric under test, the in-situ balance and adaptation of the test environment, the elimination of optical imaging interference, and simultaneously collect and bind the in-situ basic parameters of the test area to complete the pre-processing qualification verification and obtain the pre-processed qualified fabric under test and the corresponding in-situ basic parameters.

[0025] S2: Perform multimodal optical coaxial synchronous scanning on the same detection area, collect raw data related to pore morphology, three-dimensional connectivity, and material spectrum, extract core features related to fabric breathability, and accurately group the features according to subsequent calculation purposes to obtain a grouped breathability-related feature dataset.

[0026] S3: Based on the physical laws of airflow permeation in textile fabrics, and combined with the air permeability-related feature dataset obtained in S2, construct multiple sets of exclusive physical functions corresponding to pore air permeability contribution and air permeability resistance correction, and output multiple core feature values ​​directly related to the air permeability performance of the fabric through standardized calculation and data verification.

[0027] S4: Call the core feature value obtained in S3, combine it with the in-situ basic parameters collected in S1, complete the adaptive calibration of coupling weight and the calculation of environmental correction factor, complete the collaborative fusion of multiple core feature values ​​through adaptive nonlinear coupling algorithm, and output the coupling corresponding value directly related to the fabric air permeability after standardization.

[0028] S5: Call the coupling corresponding value obtained in S4, combine the pre-calibrated fabric breathability physical constraint parameters with the breathability-related feature dataset obtained in S2, calculate the physical constraint loss term and complete the secondary fusion correction, output the final judgment value of fabric breathability performance, and at the same time complete the quantitative grading of fabric breathability performance based on the preset grading standard.

[0029] S1: In-situ non-destructive pretreatment of fabric samples and synchronous environmental acquisition: Without changing the original pore structure of the fabric or causing irreversible deformation, the fabric to be tested is fixed non-destructively, adapted to the detection environment in situ, and the source of imaging interference is eliminated. The in-situ basic parameters required for subsequent calculations are acquired simultaneously. The specific operation process and parameter control are as follows:

[0030] S101. Tension-free in-situ clamping: Achieves non-destructive fixing of fabrics without stretching, wrinkles, or deformation of the pore structure, eliminating irreversible changes in pore characteristics caused by clamping stress. Specific implementation details:

[0031] Sample marking and equipment calibration: The finished / work-in-process fabric to be tested is not cut. After confirming that the fabric is undamaged and has no large-area defects, select ≥3 parallel detection areas on its flat area. The size of a single detection area is not less than 10mm×10mm. Avoid non-homogeneous areas such as fabric seams, printing, coating and sealing. Mark the detection areas to ensure that all subsequent parameter acquisition and optical scanning are completed in the same marked area.

[0032] Fabric naturally relaxed and laid flat: The fabric to be tested is laid naturally and flat on the optical-grade pneumatic adsorption stage of the testing chamber. The flatness of the stage is ≤0.002mm / m, and the surface is uniformly covered with micron-level adsorption holes. The fabric is kept in a natural relaxed state in both the warp and weft directions, without artificial stretching or forced flattening, which completely simulates the natural state of the fabric in actual use and avoids artificially introducing internal stress.

[0033] Precise alignment of the detection area: The parallel detection area marked on the fabric is precisely aligned with the effective area of ​​the adsorption through-hole of the stage and the center of the field of view of the subsequent optical system to ensure that the detection area completely covers the effective adsorption range, with no edge suspension and no local areas exceeding the adsorption area, thus avoiding local warping and wrinkles of the fabric during the subsequent adsorption process.

[0034] Gradient-pressure pneumatic adsorption fixation: The pneumatic adsorption device (including a pressure sensor with a measurement accuracy of ±1Pa and a laser thickness gauge with a measurement accuracy of ±0.001mm) is activated, and a three-stage gradient pressure mode is used for fixation to avoid fabric stretching and deformation caused by instantaneous high pressure.

[0035] The first stage involves increasing the pressure to 10Pa and maintaining it for 30 seconds to allow the fabric to initially adhere to the stage.

[0036] The second stage increases the pressure to 30Pa and holds it for 30s to eliminate slight local warping of the fabric and achieve uniform surface bonding.

[0037] The third stage ultimately stabilizes at 50 Pa (maximum adsorption pressure not exceeding 50 Pa), and is maintained for 60 seconds to complete the tension-free fixation of the fabric.

[0038] Non-deformation compliance verification: After fixation, a machine vision system is used to perform a global scan of the detection area to visually confirm that the fabric is free from stretching, wrinkles, and warping; at the same time, a laser thickness gauge is used to collect the thickness of 5 feature points (center + four corners) in the detection area and compare it with the initial thickness of the fabric in its natural relaxed state. If the thickness deviation is ≤0.5%, the clamping is considered valid; if the thickness deviation exceeds the range, the adsorption pressure is immediately released and the clamping operation is repeated until the verification is valid.

[0039] S102. In-situ Environmental Balancing and Multi-Parameter Synchronous Acquisition: Achieve rapid balancing between the fabric and the testing environment, and synchronously acquire in-situ basic data required for subsequent pore feature correction. This solves the problems of asynchronous acquisition and inaccurate correction of environmental parameters and pore features in existing technologies. Specific implementation details:

[0040] Pre-balancing of the testing chamber and adaptation to the fabric environment: The constant temperature and humidity testing chamber is started 30 minutes in advance and preset and stabilized to the standard testing environment parameters: temperature 20℃±2℃, relative humidity 65%±2%, so that the internal environment of the chamber reaches a stable state in advance.

[0041] After the clamping is effective, immediately seal the constant temperature and humidity testing chamber and keep the above environmental parameters constant. Allow the fabric to be tested to undergo environmental equilibration for 15 minutes to ensure that the temperature, humidity, and fiber moisture absorption / release state of the fabric are completely consistent with the testing environment, thus avoiding changes in the pore structure caused by fiber moisture absorption expansion and contraction.

[0042] Simultaneous acquisition of multiple parameters in the same area: After the environmental equilibration time, without opening the cavity, changing the fabric clamping state, or moving the fabric position, the acquisition of three types of core parameters is completed simultaneously in the same detection area:

[0043] Environmental parameter acquisition: The temperature and humidity of the detection area are collected in real time through the built-in temperature and humidity sensors (temperature measurement accuracy ±0.2℃, relative humidity measurement accuracy ±1%RH). relative humidity The sampling frequency is 1Hz, 10 sets of data are collected continuously, the arithmetic mean is taken as the final environmental parameter, and recorded and bound to the corresponding detection area.

[0044] Fabric thickness measurement: Using a laser thickness gauge, the thickness of five marked feature points within the detection area is measured. Each feature point is measured five times consecutively, and the average value is taken as the thickness value of that feature point. The arithmetic mean of the five feature points is then taken as the true thickness of the fabric in the detection area. Record and bind to storage;

[0045] Fiber intrinsic parameter matching: Based on the type of fabric to be tested, the corresponding pre-stored fiber moisture absorption and expansion coefficient is retrieved. relative humidity collected simultaneously Fabric type label Binding storage is used for subsequent calculation of environmental correction factors.

[0046] S103. Surface Impurity and Optical Imaging Interference Elimination: Remove surface impurities and fiber optical reflections to ensure clear and distinguishable aperture boundaries in subsequent optical imaging. Simultaneously, complete pre-processing closed-loop verification and process flow. Specific implementation details:

[0047] Ion wind non-contact dust removal: Keep the cavity sealed, the fabric clamped, and the cavity environmental parameters completely unchanged. Start the ion wind dust removal device built into the cavity, using low-pressure gentle parallel ion wind with the wind speed controlled at ≤0.5m / s. Scan the detection area back and forth along the parallel direction of the fabric surface to remove dust. Scan 3 times, with each scan lasting 10 seconds, to thoroughly remove floating dust and loose lint from the fabric surface. At the same time, the ion wind eliminates static electricity on the fabric surface to prevent secondary dust adsorption. After dust removal, let the cavity stand for 30 seconds to allow the internal environment to return to stability.

[0048] Orthogonal extinction preprocessing: Keeping the detection area position unchanged, the polarization module of the multi-mode optical system is activated, and the polarization angles of the polarizer and analyzer are adjusted to the orthogonal extinction state to eliminate the specular reflection gloss on the fiber surface. The pore imaging effect is observed in real time through microscopic imaging, and the polarization angle is finely adjusted until the grayscale contrast between the pore and the fiber matrix is ​​≥30dB, the pore boundary is clear without halo or reflective obstruction, and the optical interference is eliminated, providing a high-quality imaging substrate for subsequent optical feature acquisition.

[0049] Pretreatment verification and process flow: After dust removal and matting are completed, a pretreatment qualification verification is performed simultaneously, which must meet the following four requirements at the same time:

[0050] The fabric is free from stretching, wrinkles, and warping, and the thickness after clamping deviates from the initial thickness in its natural state by ≤0.5%.

[0051] The ambient temperature was kept stable at 20℃±2℃, and the relative humidity was kept stable at 65%±2%.

[0052] Microscopic imaging observation and detection area, free from dust and free lint obstructing the pore area;

[0053] The grayscale contrast between the pores and the fiber matrix is ​​≥30dB, the pore boundaries are clearly distinguishable, and there is no reflective halo.

[0054] After the verification is passed, keep the fabric clamping state, cavity environment parameters, and polarization light module parameters completely unchanged, and immediately proceed to the multimodal optical feature acquisition process; if the verification fails, repeat the corresponding preprocessing process for the unqualified items until the verification is passed; samples that still fail to meet the qualified standard after 3 repeated preprocessing are marked as abnormal samples.

[0055] The S2 multimodal optical feature acquisition and precise grouping: Based on a multimodal optical system, coaxial synchronous scanning is performed on the same marked detection area to accurately distinguish between three types of pores: closed pores, semi-open pores, and through pores. Twelve core original features are fully acquired, and precise grouping is performed strictly according to their specific applications. The specific operation process and parameter control are as follows:

[0056] S201. Simultaneous Multimodal Scanning and Raw Data Acquisition in the Same Area: Maintain the constant temperature and humidity of the detection chamber, ensure the fabric's pneumatic adsorption pressure remains constant, and prevent the detection area from shifting. Activate all modules of the multimodal optical system and complete preheating. Based on the acquired actual fabric thickness... Fabric type label Pre-set the acquisition parameters for each module, pre-load the pore type determination algorithm and feature extraction algorithm, and complete the necessary calibration for coaxial common field of view and synchronous triggering of the system; use the marked detection area as the only acquisition area, and define 3 parallel detection fields of view within this area, with the size of each field of view being 5mm×5mm. There is no overlap or blind spot between the fields of view. The cavity environment and fabric state remain unchanged throughout the process. Perform synchronous scanning of the three modes on the locked detection fields of view and acquire the corresponding raw data. Specific implementation content:

[0057] S2011. Simultaneous Acquisition of Transmission-Reflection Dual-Path Microscopic Imaging

[0058] An infinity-corrected optical system is used, with objectives of 10× / 20× magnification (numerical aperture NA ≥ 0.3), and an imaging resolution ≤ 1. The pixel matrix is ​​set to 2048×2048 to ensure sharp capture of pixels as small as 1. Pore ​​morphology;

[0059] The transmission and reflection optical paths are triggered synchronously to perform synchronous imaging of the same field of view. Three sets of transmission images and three sets of reflection images are continuously acquired for each field of view. Noise reduction is performed through a multi-frame averaging algorithm, and finally the transmission morphology grayscale map and reflection morphology grayscale map of the field of view are output, which completely preserves the original information such as the boundary, shape, orientation and distribution of the aperture.

[0060] After acquisition, the image quality is verified in real time. The grayscale contrast between the pores and the fiber matrix must be ≥30dB, and there must be no reflective obstruction, haloing, or blurring at the pore boundaries. If the image quality does not meet the requirements, the polarization parameters are readjusted and the image is acquired again until it meets the requirements.

[0061] S2012. Laser confocal tomography three-dimensional scanning acquisition

[0062] Synchronously triggered with dual-optical-path imaging, a laser confocal optical path coaxial with the dual-optical-path module is used; continuous tomographic scanning is performed along the fabric thickness direction (Z-axis), with the upper surface of the fabric as the 0 point, and the actual fabric thickness acquired. The scanning limit is set to cover the entire thickness direction of the fabric, with no blind spots.

[0063] The scanning step size is adaptively adjusted according to the fabric thickness: thin fabrics ( <0.2mm) Step size 1 Thick fabric ( ≥0.2mm) Step size 2 After scanning, output a 3D pore tomography dataset for that field of view, including the position, area, and shape of each pore in the X / Y plane, as well as its start position, end position, and continuity status in the Z-axis direction; verify the integrity of the scan in real time, requiring the total length of the tomographic scan along the Z-axis to be equal to the actual thickness of the fabric. The deviation should be ≤1%, with no tomography or frame loss. If the requirements are not met, the tomography scan should be repeated until the verification is successful.

[0064] S2013. Near-infrared spectroscopy combined with synchronous acquisition

[0065] A near-infrared diffuse reflection optical path coaxial with the system's principal optical axis is adopted, with the spectral acquisition range set to 1000nm-2500nm and the spectral resolution ≤ The system performs 32 scans, using multi-scan averaging and noise reduction to ensure a signal-to-noise ratio of ≥1000:1 for the spectral data. The sampling spot completely covers the current detection field of view without offset or exceeding the range, acquiring the raw near-infrared diffuse reflectance spectrum data for that field of view. This data includes complete information such as the characteristic absorption peaks of the fiber material and the scattering absorption peaks corresponding to the pore structure, while also enabling qualitative identification of the fabric material. The spectral quality is verified in real time, requiring no baseline drift, no sharp noise interference, and clearly distinguishable characteristic absorption peaks. If the requirements are not met, the spectral acquisition is repeated until the verification is successful.

[0066] S202. Core Feature Extraction and Precise Grouping Based on Design Logic

[0067] The collected raw data underwent standardized preprocessing without altering its physical nature: adaptive threshold binarization was applied to the microscopic images to separate the pore region from the fiber matrix region; connected component analysis was performed on the 3D tomography dataset to complete the 3D modeling of the pores; baseline correction and normalization were performed on the near-infrared spectral data to provide a standardized data foundation for feature extraction; strictly following the design logic of the original scheme, 12 core features were extracted, and the specific purpose, extraction method, and physical meaning of each feature group were clearly defined, as follows:

[0068] S2021. Feature Extraction of Function Construction Group: Statistically analyze the relevant features of through-holes that can form airflow channels, completely eliminating interference from invalid closed or semi-open holes. The extraction method for each feature is as follows:

[0069] Through porosity ( Based on a laser confocal 3D tomography dataset, through connected component analysis, only the total area of ​​through holes that completely penetrate the upper and lower surfaces of the fabric is counted, and the proportion of this area to the total area of ​​the detection field of view is calculated. The formula is as follows: The area statistics of closed-cell and semi-open-cell holes were completely excluded.

[0070] Equivalent hydraulic aperture Based on the matching results of dual-path microscopic images and three-dimensional tomographic datasets, only the diameters of the identified through-holes were statistically analyzed. The equivalent hydraulic diameter of each through-hole was calculated using the equivalent circle diameter method, and the arithmetic mean of the equivalent diameters of all through-holes was taken as the final value. Values ​​that exclude closed-hole and semi-open-hole aperture data;

[0071] Pore ​​size variation coefficient ( Based on the equivalent hydraulic aperture dataset of the above through-holes, the ratio of the standard deviation to the mean of the aperture is calculated using the following formula: This characterizes the uniformity of the diameter of the through holes, and is only statistically analyzed for through holes;

[0072] Pore ​​penetration ( Based on a 3D tomography dataset, the average extension length of all through holes in the Z-axis direction is statistically analyzed, and the ratio of this length to the actual fabric thickness h is calculated using the following formula: The value range is 0≤ ≤1, =1 represents the thickness of the fabric with pores completely connected.

[0073] Closed-pore ratio ( Based on a 3D tomography dataset, the total area of ​​closed pores completely enclosed by fibers and without any connectivity is statistically analyzed. The proportion of this area to the total area of ​​all pores (through pores + semi-open pores + closed pores) within the detection field of view is calculated using the following formula: Quantify the proportion of ineffective pores;

[0074] Fractal dimension of pores ( Based on the planar morphology of the through-hole in dual-path microscopic images, the box-dimensional method is used for fractal calculation to obtain the complexity of the through-hole channel. The range of values ​​is 2≤ ≤3, The closer the value is to 2, the straighter the pore channels and the lower the air permeability resistance. The closer a value is to 3, the more tortuous the pore channels and the greater the air permeability resistance.

[0075] S2022. Coupling Correction Group Feature Extraction:

[0076] Directly call the actual fabric thickness collected by S1 Fiber moisture absorption and expansion coefficient ; Detecting ambient relative humidity .

[0077] S2023. Feature extraction of secondary fusion groups:

[0078] Near-infrared spectral pore characteristic absorption peak intensity ( Based on the preprocessed near-infrared spectral data, characteristic absorption peaks corresponding to the microstructure of the pores are extracted, and the normalized peak area is taken as the intensity of the characteristic absorption peak. It captures deep structural information of nanoscale pores that cannot be identified by tomographic imaging;

[0079] Warp / Weft Pore Orientation Difference ( Based on dual-path microscopic images, the principal orientation angles of pores in the warp and weft directions of the fabric are extracted using a two-dimensional Fourier transform method. The absolute value of the difference between the principal orientation angles in the two directions is calculated as the warp / weft pore orientation difference. Characterizes the anisotropic properties of fabric pores;

[0080] Fabric type label Based on the material qualitative results from near-infrared spectroscopy, combined with the basic information of the fabric to be tested, the label standardization and assignment were completed: cotton fabric =1, synthetic fiber fabric =2, blended fabrics =3, Special Functional Fabrics =4, providing a basis for subsequent adaptive weight adjustment and physical constraint term calibration.

[0081] The S3 exclusive physical function construction and core feature value calculation: Based on the physical nature of airflow permeability in textile fabrics, three sets of exclusive physical functions are constructed, and three core feature values ​​directly related to air permeability are output through standardized calculations. , , The specific operation process and parameter control are as follows:

[0082] S301. Construction of Dedicated Physics Functions, Specific Implementation Details:

[0083] The first set of functions is used to calculate the effective air permeability contribution value of the pores. Based on penetrating porosity Equivalent hydraulic aperture Combining Poiseuille's law of airflow infiltration, a physical function quantifying the contribution of effective pore permeability is constructed, and the formula is: Among them, the aperture normalization coefficient =100 The value range is 0 < ≤10; the higher the proportion of through-pores and the larger the pore size, the stronger the contribution to air permeability.

[0084] The second set of functions is used to calculate the correction value for pore penetration resistance. Based on porosity closed-pore ratio The physical function that quantifies the effect of pore connectivity on air permeability resistance is constructed, and the formula is as follows: The value range is 0 < ≤1; the higher the porosity and the lower the proportion of closed pores, the smaller the air permeability resistance and the larger the correction coefficient.

[0085] The third set of functions: used to calculate the correction value for morphological resistance in the pore thickness direction. Based on pore size variation coefficient pore fractal dimension The physical function that modifies the air permeability resistance by quantifying pore morphology (uniformity, tortuosity) is defined by the following formula: The value range is 0 < ≤1; the more uniform the pore size and the straighter the pore channels ( The closer to 2), the smaller the air resistance and the larger the correction coefficient.

[0086] 302. Core Eigenvalues , , Standardized calculation, specific implementation details:

[0087] Single-field feature value calculation: For each parallel detection field of view, the six functions corresponding to that field of view are called to construct a set of feature data, which are then substituted into three sets of dedicated physical functions to complete the calculation step by step. , , The calculation, specific steps:

[0088] calculate : The penetrating porosity of this field of view Equivalent hydraulic aperture Substituting the first set of functions, the single field of view is calculated precisely. The value is 1, retaining 4 significant digits to ensure calculation accuracy;

[0089] calculate 2: Aperture penetration of the field of view closed-pore ratio Substituting into the second set of functions, we can calculate the single field of view. The value is 2, and 4 significant figures are retained to ensure the accuracy of the correction factor;

[0090] calculate 3: The aperture variation coefficient of this field of view pore fractal dimension Substituting into the third set of functions, the single field of view is calculated. The value is set to 2, retaining 4 significant digits to avoid the accumulation of calculation errors.

[0091] Multi-field data aggregation and outlier removal: completing single-field analysis of 3 parallel detection fields. , , After calculation, all data are summarized, and outliers are removed using the Grubbs criterion (single-field data with a relative deviation exceeding 2 standard deviations) to ensure data reliability. If there are fewer than 2 sets of valid data after removal, feature extraction in step S2 and the calculation in this step are repeated until at least 2 sets of valid data are obtained.

[0092] Final core feature value determination: For the valid single-view data after removing outliers, calculate respectively , , The average value is used as the final core feature value of the detection area. , , Retain 4 significant digits.

[0093] S303. Full-Dimensional Compliance Verification of Core Feature Values

[0094] For the final determination , , Perform comprehensive verification of physical boundaries, logical consistency, and parallelism to ensure that feature values ​​are true, reliable, and fully meet the requirements of subsequent coupled calculations. Simultaneously, complete data storage and process flow. Specific implementation details include:

[0095] Physical boundary verification: Verify one by one , , The range of values ​​must conform to the physical meaning and the preset interval: >0 (no ineffective breathability contribution), 0< 2≤1 (resistance correction factor within a reasonable range), 0< 3≤1 (morphological correction coefficient is within a reasonable range), no negative values, no outliers; if outliers exist, recheck the function calculation process and the feature data of S2, and re-extract and calculate features if necessary.

[0096] The S4 eigenvalue adaptive nonlinear coupling and coupling correspondence value calculation: The modified adaptive nonlinear coupling algorithm completes eigenvalue collaborative fusion and environmental correction, outputting a coupling correspondence value directly related to the fabric's breathability. To address the shortcomings of existing technologies, such as fixed feature fusion methods, lack of environmental adaptive correction, and weak correlation between coupling results and air permeability, the specific operation process and parameter control are as follows:

[0097] S401. Adaptive Calibration of Coupling Weights

[0098] The final core feature value output by the call Combining the extracted coupling correction group of 3 features (true fabric thickness) Fiber moisture absorption expansion coefficient Detecting the relative humidity of the environment Based on the physical laws of fabric breathability, the coupling weight adaptive calibration is completed to ensure that the weights are accurately matched with fabric characteristics and environmental conditions. Specific implementation details are as follows:

[0099] Input data locking: Strictly call output. 1 (Effective air permeability contribution value of pores) 2 (Correction value for pore penetration resistance) 3 (pore morphology resistance correction value), and extracted coupled correction group features ( ).

[0100] Weighting criteria: Based on the core logic of "effective breathability as the primary factor, resistance correction as a secondary factor, and environmental condition compensation as the secondary factor," and combined with the mechanism by which the breathability performance of textile fabrics is affected, the weighting rules are determined as follows: 1 is the core indicator with the highest weighting. 2. 3. Weighting based on fabric thickness Adaptive adjustment; environmental correction item weights based on and Determined collaboratively.

[0101] Core coupling weight ( Calibration:

[0102] Basic weight settings: Preset 1. Basic weight , 2. Basic Weights , 3 Basic Weights ,satisfy ;

[0103] Thickness adaptive correction: based on the extracted actual fabric thickness. ,right Adaptive correction is performed, and the correction formula is:

[0104]

[0105]

[0106]

[0107] Correction logic: The thicker the fabric ( The larger the pore connectivity resistance and shape resistance, the less marginal the effect of their influence on breathability, and the corresponding weight should be appropriately reduced to conform to the physical laws of breathability resistance of thick fabrics.

[0108] Weight boundary constraints: Ensure after calibration There are no negative weights or out-of-bounds weights. If a weight exceeds the range, the adjustment coefficient will be applied. Adjusted to Find suitable values ​​within the range until the requirements are met.

[0109] Environmental Correction Factor ( Calibration:

[0110] The original distributed weight design is abandoned, and a single environmental correction factor is integrated, with the following formula:

[0111]

[0112] Of which, 65% For standard humidity testing, The extracted fiber hygroscopic expansion coefficient ( ), To detect the relative humidity of the environment;

[0113] Post-calibration verification: ensure There are no superphysical boundary values, if If the absolute value is greater than 0.2, it is truncated by ±0.2 to avoid excessive distortion due to environmental correction.

[0114] S402. Adaptive Nonlinear Coupling Calculation: Based on the calibrated coupling weights and environmental correction factors, the core feature value is fused using a modified nonlinear coupling algorithm to calculate the initial coupling value. Specific implementation details:

[0115] Coupling algorithm determination: An adaptive nonlinear coupling formula with clear physical meaning is adopted, combined with the weight calibration results. The formula is as follows:

[0116] in, Strengthening the core role of effective breathability To mitigate the impact of extreme values ​​in morphological resistance, An environmental correction factor is used to ensure that the coupling results conform to the actual detection environment and are free from distortion caused by environmental interference.

[0117] Step-by-step coupled computation:

[0118] Step 1: Calculate the core feature value coupling term and substitute it into the calibrated value. With the corresponding The formula is: The calculation result should be rounded to 4 significant figures.

[0119] Step 2: Calculate the initial coupling value , core coupling items With environmental correction factors Multiplication, the formula is: Retain 4 significant figures and record the calculation process and intermediate results;

[0120] Single-view and multi-view coupled calculation:

[0121] Single field-of-view calculation: For the three parallel detection fields defined by S2, call the corresponding fields of view respectively. The data is used to calculate the initial coupling value for each field of view as described above. ;

[0122] Multi-field summary: Summarizes 3 fields of view The values ​​are then processed using the Grubbs criterion to remove outliers (single-field data with a relative deviation exceeding 2 standard deviations). If there are fewer than 2 sets of valid data after removal, the eigenvalues ​​and coupling calculations are repeated until at least 2 sets of valid data are obtained.

[0123] S403. Coupling Corresponding Value Standardization and Correction: Initial Coupling Values Perform dimensional standardization and anomaly correction, and output the final coupled corresponding value. Specific implementation details:

[0124] Dimensional standardization: Initial coupling values ​​of dimensionless quantities Converted to dimensions consistent with national standard air permeability, the standardized formula is: in Dimension conversion factor ( To ensure that it is calibrated according to 100+ categories of standard fabrics The value range matches the breathability of conventional fabrics: ).

[0125] Outlier correction:

[0126] like The problem was determined to be a coupling calculation error. The weight calibration, the substitution process of the coupling formula, and the preceding feature data were rechecked, and the calculation was recalculated.

[0127] like The air permeability exceeds the physical limit of conventional textile fabrics and is corrected to 1000 mm / s. It is also marked as an extremely high air permeability sample for physical constraint correction.

[0128] like No need to modify, just keep it.

[0129] Parallelism Correction and Final Value determined:

[0130] Effective field of view after outlier removal Calculate the relative standard deviation (RSD) and ensure that the RSD is ≤3% to guarantee data repeatability. If the RSD exceeds the range, recalculate the eigenvalues, weight calibration, and coupling calculation in this step until the requirements are met.

[0131] Take effective field of view The arithmetic mean of the values ​​is used as the final coupling correspondence value for the detection region. .

[0132] Coupling corresponding value compliance verification: final verification The physical boundary of the value ( Parallelism (RSD≤3%) is checked to ensure no anomalies; after passing the verification, the corresponding values ​​will be finally coupled. Together with the preceding feature data, weight calibration results, coupling calculation process, and standardized records, they are bound to a unique sample number and encrypted storage in the detection area; keeping the fabric clamping state and cavity environment parameters unchanged, the process immediately enters the physical constraint secondary fusion and final judgment process of air permeability; if it still fails after 3 repeated calculations and verifications, it is marked as an abnormal sample and the detection process is terminated.

[0133] The S5 physical constraint secondary fusion and final air permeability performance determination: Combining pre-calibrated parameters and the physical constraint mechanism, the secondary fusion calculation and final air permeability performance determination are completed, and the final air permeability performance value of the fabric is output. The quantitative grading results are used, and a dynamic incremental correction mechanism for the secondary fusion function is established to ensure long-term stability of detection accuracy. The specific operation process and parameter control are as follows:

[0134] S501. Calculation of Physical Constraint Loss Term: Call the output coupling correspondence value Extracted fabric type labels Based on the breathability of the fabric, the physical constraint loss term is calculated. This is used to correct the secondary fusion results and ensure that the final judgment value conforms to the physical laws of breathability of textile fabrics. Specific implementation details are as follows:

[0135] Input data matching with parameters: the final coupled value of the output call. Extracted fabric type labels Based on fabric type label Automatically matches pre-calibrated parameters, including regression coefficients. (Based on standard fabric calibration across all categories, the goodness of fit R² ≥ 0.995), and physical constraint weights. (cotton fabric) =0.8, chemical fiber fabrics =0.6, blended fabrics =0.7, Special functional fabrics =1.0), ensuring that the parameters are accurately matched with the fabric type, with no mismatches or omissions.

[0136] Dedicated physics function call: Employing a breathable physics mechanism, a dedicated physics function is called, based on the extracted secondary fusion group features and their corresponding coupling values. Calculate the theoretical value of fabric breathability.

[0137] Calculation of physical constraint loss term: Calculate the theoretical value of fabric air permeability and the corresponding coupling value. The absolute deviation is used as the physical constraint loss term. The calculation formula is: The theoretical air permeability value is in mm / s to ensure that the physical meaning of the loss term is clear.

[0138] Through physical constraint weights right We perform weighting to obtain the weighted constraint terms. This is used for subsequent secondary fusion correction, constraining the final judgment value to conform to the physical nature of the breathability of textile fabrics, and avoiding distortion caused by pure data fitting.

[0139] Compliance verification of loss items: After calculation, verification is performed. The rationality, to ensure ,like If the value exceeds this range, it indicates a significant deviation between the theoretical value and the corresponding coupling value. The parameters of the specific physical function should be rechecked. The value calculation process and the matching status of the fabric type label M were recalculated. .

[0140] S502. Single field of view Value calculation: based on regression coefficients, secondary fusion group features, and coupled corresponding values. Including physical constraint loss terms, the final judgment value for a single detection field of view is completed. Calculations are performed to ensure both accuracy and physical plausibility. Specific implementation details include:

[0141] Linear regression basic term calculation: matching regression coefficients , and the output Extracted Substituting into the linear regression formula, we can calculate the basic terms of linear regression. The formula is as follows: .

[0142] Physical constraint correction and single field of view Value determination: Combine the above linear regression basic terms with the physical constraint weighted correction term ( The values ​​are superimposed to obtain the final decision value for a single field of view. The superposition formula is: (The correction direction conforms to the laws of physics, making) (The value approaches the theoretical air permeability value).

[0143] Single field of view Value verification: After calculation, verify the single field of view immediately. The physical rationality of the value ensures (No negative values ​​for breathability), if The calculation was determined to be abnormal, and the regression coefficients were rechecked for matching. The calculation process for the basic terms of linear regression, and the recalculation of the single field of view. value.

[0144] S503. Final Value determination: Single field of view for 3 parallel detection fields of view The values ​​are used to remove outliers and calculate the average value to determine the final judgment value for the fabric's breathability performance. To ensure data repeatability and reliability, the specific implementation details are as follows:

[0145] Multi-view Value summary: Summarize the single field of view corresponding to the three defined parallel detection fields of view. Values ​​to ensure the accuracy of each field of view. All values ​​have been verified by a single field of view, with no outliers or missing values, and correspond one-to-one with the fields of view in the previous steps.

[0146] Outlier Removal: Outlier removal rules (Grubbs' criterion) are used to remove outliers in the three parallel fields of view whose relative deviation exceeds twice the standard deviation. Value, ensuring the remaining The reliability of the value; if it is valid after removal. If there are fewer than two sets of values, recalculate the coupling, physical constraint loss term, and single-viewpoint calculations. Value calculation continues until at least two valid sets are obtained. value.

[0147] final Value calculation and verification: Retain valid values ​​after removing outliers. The arithmetic mean of the values ​​is used as the final judgment value for the air permeability of the fabric in that test area. Simultaneously verify the final The coupling between value and output corresponds to the value The deviation should be within ±5%. If the deviation exceeds this range, the calculation of the physical constraint loss term and the single field of view should be rechecked. The value calculation process continues until the requirements are met; if the deviation requirements are still not met after three repeated calculations, the sample is marked as abnormal and the testing process is terminated.

[0148] S504. Quantitative Grading of Breathability and Reliability Verification of Results: Based on the final The breathability performance was quantitatively graded, and reliability verification was conducted to ensure the accuracy and reproducibility of the test results. Specific implementation details are as follows:

[0149] Quantitative grading: based on the final judgment value The air permeability performance should be quantitatively graded according to the following standards:

[0150] Level 1 (non-breathable): ;

[0151] Level 2 (Low breathability): ;

[0152] Level 3 (Medium breathability): ;

[0153] Level 4 (High Breathability): ;

[0154] Level 5 (Extremely High Breathability): After grading is completed, the grading results are recorded and linked to the sample number and testing area to ensure traceability of the grading.

[0155] Reliability verification:

[0156] Standard deviation and expanded uncertainty calculation: based on 3 effective parallel fields of view Value, calculate the experimental standard deviation, and take the coverage factor. According to the formula Calculate the expanded uncertainty U;

[0157] Re-inspection rule execution: when expanded uncertainty At the same time, while keeping the fabric clamping state and the constant temperature and humidity chamber environmental parameters unchanged, re-execute the full process calculation of S2-5 (from multimodal optical feature acquisition to final...). (Value determined); after the re-inspection is completed, the re-inspection value is calculated. Value and initial inspection If the relative deviation of the values ​​is ≤3%, the initial inspection and re-inspection shall be considered. The arithmetic mean of the values ​​is used as the final Values ​​and grading criteria; if the relative deviation is greater than 3%, change the testing area and re-execute the full process testing;

[0158] Verification pass / fail criteria: When the expanded uncertainty If the relative deviation is ≤3% after retesting, the reliability verification is deemed qualified; if the verification requirements still cannot be met after changing the testing area, the sample is marked as abnormal and the testing process is terminated.

[0159] S505. Dynamic Incremental Correction Mechanism for Secondary Fusion Function: Establish a monthly incremental update mode to dynamically correct relevant parameters of secondary fusion without changing the core structure and physical constraint logic of the function, thereby improving the model's generalization ability for new fabrics and fabrics using new processes. Specific implementation details are as follows:

[0160] Correction mode and scope: A monthly incremental update mode is adopted, and only the regression coefficients are updated. and physical constraint weights The modifications are made without altering the core structure, physical constraint logic, and calculation process of the quadratic fusion function, ensuring the stability and continuity of the detection method.

[0161] Valid sample collection and screening: Relevant data on newly added fabrics are collected monthly, including collected multimodal optical feature data and data calculated by this method. The data includes the measured breathability values ​​according to national standards; valid samples are selected based on the following criteria: ≥50 valid samples of newly added fabrics in a single product category, with complete data and verified and reliable measured breathability values ​​according to national standards; samples with missing data or abnormal measured values ​​are removed.

[0162] Incremental parameter correction and validation: Based on the selected valid samples, the regression coefficients are corrected using a generalized regression incremental learning algorithm. Physical constraint weights Correction; after correction, the model fit for all categories is verified to be R²≥0.995, and the extreme samples ( or If the prediction deviation is ≤3%, the original parameters will be retained and the correction will not be performed. The corrected parameters will be updated synchronously to the pre-calibration parameter library for parameter matching in subsequent tests.

[0163] Full-sample cross-validation and optimization: After every 6 monthly incremental updates, full-sample cross-validation is performed, integrating all historical valid samples (including newly added samples and existing standard samples), and optimizing the physical constraint weights. The adaptive rules further enhance the model's generalization ability to new fabrics and new process fabrics (such as new blended fabrics and special functional fabrics), ensuring long-term stable detection accuracy. After cross-validation, if the model fit or extreme sample prediction deviation does not meet the standard, the incremental learning algorithm is re-optimized, and the parameters are corrected and validated again.

[0164] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0165] In conclusion, 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 within the protection scope of the present invention.

Claims

1. A method for determining the air permeability of fabrics based on optical characteristic analysis of fabric pores, characterized in that, include: S1: Without changing the original pore structure of the fabric under test or causing irreversible deformation, complete the tension-free and non-destructive fixation of the fabric under test, the in-situ balance and adaptation of the test environment, the elimination of optical imaging interference, and simultaneously collect and bind the in-situ basic parameters of the test area to complete the pre-processing qualification verification and obtain the pre-processed qualified fabric under test and the corresponding in-situ basic parameters. S2: Perform multimodal optical coaxial synchronous scanning on the same detection area, collect raw data related to pore morphology, three-dimensional connectivity, and material spectrum, extract core features related to fabric breathability, and accurately group the features according to subsequent calculation purposes to obtain a grouped breathability-related feature dataset. S3: Based on the physical laws of airflow permeation in textile fabrics, and combined with the air permeability-related feature dataset obtained in S2, construct multiple sets of exclusive physical functions corresponding to the contribution of pore air permeability and the correction of air permeability resistance, respectively. Through standardized calculation and data verification, output the core feature values ​​related to the air permeability performance of the fabric. S4: Call the core feature value obtained in S3, combine it with the in-situ basic parameters collected in S1, complete the adaptive calibration of coupling weight and the calculation of environmental correction factor, complete the collaborative fusion of multiple core feature values ​​through adaptive nonlinear coupling algorithm, and output the coupling corresponding value directly related to the fabric air permeability after standardization. S5: Call the coupling corresponding value obtained in S4, combine the pre-calibrated fabric breathability physical constraint parameters with the breathability-related feature dataset obtained in S2, calculate the physical constraint loss term and complete the secondary fusion correction, output the final judgment value of fabric breathability performance, and at the same time complete the quantitative grading of fabric breathability performance based on the preset grading standard.

2. The method for determining the air permeability of fabrics based on the optical characteristics analysis of fabric pores according to claim 1, characterized in that, S1 includes: S101 Tensionless In-situ Clamping: Select the test area and mark the location of the fabric to be tested. Lay the fabric naturally and loosely on the optical-grade pneumatic adsorption stage. Use a three-level gradient pressure mode to complete the pneumatic adsorption fixation. Verify that the thickness of the fabric after clamping is ≤0.5% of the initial thickness in its natural state, thus completing the non-destructive fixation. S102 In-situ Environmental Balancing and Multi-Parameter Synchronous Acquisition: The fabric and the testing environment are balanced within the constant temperature and humidity testing chamber. The testing environment parameters, actual fabric thickness, and fiber moisture absorption expansion coefficient of the same testing area are simultaneously acquired and stored as in-situ basic parameters. S103 Surface Impurities and Optical Imaging Interference Elimination: Low-pressure parallel ion wind is used to complete non-contact dust removal of the fabric, and polarized light orthogonal extinction treatment is used to eliminate the specular reflection on the fiber surface. After completion, a pre-treatment qualification verification is performed, and after passing the verification, it enters the subsequent process.

3. The method for determining the air permeability of fabrics based on optical characteristic analysis of fabric pores according to claim 1, characterized in that, The multimodal optical coaxial synchronous scanning includes: synchronously triggering transmission-reflection dual-optical-path microscopic imaging acquisition, laser confocal tomography three-dimensional scanning acquisition, and near-infrared spectroscopy acquisition for the same detection field of view, with the optical paths of the three modes being coaxial and sharing the same field of view, and being triggered synchronously.

4. The method for determining the air permeability of fabrics based on optical characteristic analysis of fabric pores according to claim 1, characterized in that, The precise grouping of features includes: The function constructs a set of features, including through porosity, equivalent hydraulic aperture, pore size variation coefficient, pore penetration, closed pore ratio, and pore fractal dimension. The coupling correction group features include the actual fabric thickness, fiber moisture absorption expansion coefficient, and relative humidity of the detection environment; Secondary fusion group characteristics include near-infrared spectral pore characteristic absorption peak intensity, warp / weft pore orientation difference, and fabric type label.

5. The method for determining the air permeability of fabrics based on the optical characteristics analysis of fabric pores according to claim 1, characterized in that, The specific physical functions include: The first set of functions is used to calculate the effective air permeability contribution value of the pores. The formula is: in, For the permeable porosity, For equivalent hydraulic aperture, Let be the aperture normalization coefficient, and =100μm; The second set of functions is used to calculate the correction value for pore penetration resistance.

2. The formula is: in, Pore ​​penetration, The percentage of closed pores; The third set of functions is used to calculate the morphological resistance correction value in the pore thickness direction. The formula is: in, Here is the pore size variation coefficient. is the fractal dimension of the pores.

6. The method for determining the air permeability of fabrics based on the optical characteristics analysis of fabric pores according to claim 1, characterized in that, The adaptive calibration includes: setting a basic weight based on the effective air permeability contribution value of the pores, adaptively adjusting the resistance correction-related weights based on the actual fabric thickness, and ensuring that the calibrated weights meet the following requirements: ,in These are the coupling weights corresponding to the effective air permeability contribution value of pores, the correction value of pore penetration resistance, and the correction value of pore morphology resistance, respectively. .

7. The method for determining the air permeability of fabrics based on optical characteristic analysis of fabric pores according to claim 1, characterized in that, The environmental correction factors include: Of which, 65% For standard humidity testing, The extracted fiber moisture absorption and expansion coefficient. To detect the relative humidity of the environment; and The value range is limited to 0.8 ≤ ≤1.

2.

8. The method for determining the air permeability of fabrics based on the optical characteristics analysis of fabric pores according to claim 1, characterized in that, The adaptive nonlinear coupling algorithm includes: in, This is the initial coupling value. For adaptively calibrated coupling weights, As the core feature value, Environmental correction factor; for initial coupling value Perform dimensional standardization to obtain the coupled corresponding values. The standardized formula is ,in The dimension conversion factor is calibrated using standard fabric and is set to 8.5 mm / s.

9. The method for determining the air permeability of fabrics based on optical characteristic analysis of fabric pores according to claim 1, characterized in that, The S5 includes: S501 calculates the absolute deviation between the theoretical value of fabric breathability and the corresponding coupled value based on the pre-calibrated regression coefficient and physical constraint weights matched with the fabric type label, and obtains the physical constraint loss term. S502 combines the coupling correspondence value, the secondary fusion group features, and the regression coefficient to calculate the basic term of linear regression. The basic term of linear regression is then corrected by the physical constraint loss term to obtain the final judgment value for a single field of view. S503 performs outlier removal and average calculation on the final judgment values ​​of a single field of view from multiple parallel testing fields to obtain the final judgment value of the fabric's breathability performance. ; S504 is based on the final judgment value. Complete the quantitative grading of the fabric's breathability performance and perform reliability verification of the results.

10. The method for determining the air permeability of fabrics based on the optical characteristics analysis of fabric pores according to claim 9, characterized in that, The quantitative grading includes: Level 1 (non-breathable): ; Level 2 (Low breathability): ; Level 3 (Medium breathability): ; Level 4 (High Breathability): ; Level 5 (Extremely High Breathability): After grading is completed, the grading results are recorded and linked to the sample number and testing area to ensure traceability of the grading.

Images