Clothing moisture drying rate testing method, system and device

Abstract

The invention provides a clothing material moisture drying rate testing method, system and device, and relates to the technical field of clothing material moisture drying rate testing.The method mainly comprises the steps that a first testing scene is built; after test water with a preset volume is uniformly injected into the dress material sample through the pipette, the dress material sample is heated through the heating plate; collecting an image of the clothing material sample through a CMOS camera, and synchronously recording the weight of the clothing material sample through a weighing sensor; extracting a region-of-interest image in the clothing sample image through an image segmentation model; extracting humidity features in the image of the region of interest through a CNN (Convolutional Neural Network); constructing a regression model based on the humidity characteristics and the clothing weight data; constructing a physical constraint equation; and constructing and training an LSTM neural network. According to the scheme, a test scene conforming to a real wearing scene of a human body can be built through a simple test tool; the scheme belongs to non-contact measurement, and the testing efficiency and effect of the moisture drying rate of the clothing can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of clothing moisture drying rate testing technology, and in particular to a clothing moisture drying rate testing method, system and device. Background Art

[0002] Traditional drying rate testing relies primarily on gravimetric methods (such as those described in the national standard GB / T 21655.1-2008). This involves dropping a specified amount of water onto a fabric sample, hanging it until dry, and then weighing it to calculate the drying rate. This method operates at a standard ambient temperature of approximately 23°C, which doesn't reflect the actual wearing experience of a normal person. Furthermore, this method suffers from the following drawbacks: It lacks the ability to obtain real-time spatial distribution information on drying rate and other data; contact-based measurements interfere with the natural drying process and are cumbersome and error-prone; and single-point data struggles to reflect complex mass transfer mechanisms and overall drying progress, making it unsuitable for increasingly complex testing scenarios involving spliced and composite fabrics. Summary of the Invention

[0003] The object of the present invention is to provide a method, system and device for testing the drying rate of moisture in clothing materials, so as to solve at least one of the above-mentioned technical problems existing in the prior art.

[0004] In a first aspect, to solve the above technical problems, the present invention provides a method for testing the drying rate of clothing moisture, comprising the following steps: Step 1: Set up the first test scene, specifically including setting up an LED cold light source, a heating plate, a pipette, a weighing sensor, a CMOS camera, an infrared thermal imager, etc. in a constant temperature and humidity environment; The heating plate is used to heat the clothing sample at a preset temperature (e.g., 37°C); The pipette is used to inject a preset volume (e.g., 0.1 mL / cm²) of test water (e.g., deionized water) into the clothing sample; The weighing sensor is used to weigh the clothing sample; The CMOS camera is used to photograph the area where the clothing sample is located; The infrared thermal imager is used to collect the actual temperature of the clothing sample; This makes it easy and effective to build a test scene that simulates a human body wearing clothes.

[0005] In a feasible embodiment, the first test scene also includes a fan and a wind speed sensor, which are used to blow air to the clothing sample at a preset wind speed (for example, 1.5 m / s) to simulate the wind speed in a real wearing environment and accelerate the moisture evaporation process.

[0006] Step 2: After evenly injecting a preset volume of test water into the fabric sample using a pipette, heat the fabric sample using a heating plate. Use a CMOS camera to capture images of the fabric sample in a time series at a preset shooting frequency, and simultaneously record the fabric sample weight using a weighing sensor until the weighing sensor value stops changing (i.e., the fabric is completely dry). This allows for simultaneous collection of fabric image data and fabric weight data during the drying process, providing data support for subsequent processing.

[0007] Step 3: Extract a region of interest (ROI) from the fabric sample image using an image segmentation model (e.g., a threshold mask method or a U-Net image segmentation model). The ROI image only includes the fabric sample. This allows for capturing and sampling fabric samples of different sizes without adjusting the CMOS camera's viewfinder, while automatically avoiding background noise interference in the fabric sample image. Then, extract the humidity features from the ROI image using a CNN (convolutional neural network) network.

[0008] In a feasible implementation, the method for extracting humidity features specifically includes: Step 31: Convert the image format of the region of interest (e.g., RGB) to LAB format using a color space conversion method, separating the L (luminance) channel, the A (green-red) channel, and the B (blue-yellow) channel. This allows the L channel to be highly correlated with the lightness perceived by the human eye, allowing for simple capture of details such as the darkening of the fabric caused by moisture. Specifically, when the fabric is wet, the increased moisture reduces the surface diffuse reflectance, resulting in a significant decrease in the L value. Step 32: Calculate the mean brightness and dynamic contrast based on the L channel: The brightness mean is used to directly reflect the overall brightness change. That is, when it is wet, the brightness mean decreases, and when it is dry, the brightness mean recovers, thereby directly reflecting the reflectivity change of the clothing surface. The specific calculation formula includes: ; in, represents the mean brightness; Represents the total number of pixels in a single frame of clothing sample image; Indicates the The brightness value of each pixel; The dynamic contrast is used to quantify local brightness fluctuations by calculating the ratio of the brightness standard deviation to the brightness mean, thereby reflecting the contrast reduction caused by the homogenization of the water film in the wet area; Step 33: Convert the image of the region of interest into a grayscale image and calculate the information entropy (Shannong entropy) to quickly and easily reflect the complexity of the texture. That is, when the fabric is wet, water stains will cause the texture of the fabric fiber structure to be smoothed, thereby reducing the information entropy value. When the fabric is dry, the fabric fiber structure is restored, thereby increasing the information entropy value. The specific formula includes: ; in, represents information entropy; Indicates the The probability distribution of the gray value of each pixel.

[0009] In a feasible implementation manner, after calculating the brightness mean and the dynamic contrast, step 32 further includes: Step 321: Calculate chromaticity based on the A channel and the B channel. The specific formula includes: ; in, Indicates chromaticity; Indicates the component value of channel A (before channel offset, the value range is -128 to 127, with positive values tending to be red and negative values tending to be green; after channel offset, the value range is 0 to 255); Represents the component value of the B channel (before channel offset, the value range is -128 to 127, with positive values appearing yellowish and negative values appearing bluish; after channel offset, the value range is 0 to 255). This allows for both the darker color (red) and higher chroma (color) of dark fabrics to be more evenly distributed and less chroma (color) when wet, thereby improving the applicability of the humidity feature to dark and multicolored fabrics. Step 322: Calculate comprehensive features. The specific formula includes: ; in, All represent corresponding weights; This will help significantly expand the applicability of humidity characteristics to clothing of different colors and improve generalization.

[0010] In a feasible implementation manner, step 33 may also be replaced by: Step 34: Convert the region of interest image into a grayscale image and calculate the gray level co-occurrence matrix (GLCM) contrast , the specific formula is: ; in, Indicates the Grayscale value and The joint probability of a pixel pair composed of grayscale values in a preset direction and a preset spacing; Represents the total number of grayscale values; compared with information entropy, this can more effectively quantify the dynamic changes of clothing surface texture during the wetting-drying process. Combined with multi-directional and multi-spacing feature extraction and subsequent physical constraints, it can significantly improve the accuracy and robustness of drying rate prediction, but the computational complexity will increase.

[0011] Step 4: Build a regression model based on humidity characteristics and clothing weight data , which is used to establish the relationship between humidity characteristics and moisture content of clothing. The specific expressions include: ; in, Indicates the moisture content of the fabric; Represents humidity characteristics (depending on the application scenario, it can include one or more combined features such as the aforementioned brightness mean, dynamic contrast, information entropy, chromaticity, comprehensive features, gray-level co-occurrence matrix contrast, etc.).

[0012] Step 5: Construct the physical constraint equations, including the mixed drying equation. The specific formula is: ; in, Indicates the equilibrium moisture content (i.e. the moisture content of the clothing sample when it is in equilibrium with the ambient humidity); Indicates time; Indicates the surface evaporation coefficient (unit: ), Used to describe the surface evaporation process, which conforms to the linear mass transfer law; represents the diffusion correlation coefficient (unit is ), Used to describe the diffusion process inside porous media, which conforms to the nonlinear mass transfer law; as long as these two coefficients are predicted, the drying rate of the cloth can be calculated. .

[0013] In a feasible implementation manner, the physical constraint equation also includes a thermodynamic-mass transfer coupling equation, and the specific formula is: ; in, Indicates the actual temperature of the clothing sample (Unit: Kelvin, obtained by converting data collected by infrared thermal imager) dependent evaporation coefficient; represents the latent heat of vaporization of water; Indicates the thickness of the fabric; The spatial second-order derivative of moisture content is used to describe the diffusion behavior of moisture in the fabric (if the value is greater than 0, moisture will diffuse from the area to the surrounding area; if the value is less than 0, the surrounding moisture will converge to the area); The temperature-dependent diffusion coefficient can be calculated using the Arrhenius equation as follows: ; in, represents the pre-factor related to the pore structure of the fabric; represents the diffusion activation energy, which is used to reflect the energy barrier that needs to be overcome for diffusion; represents the ideal gas constant; In this way, the temperature variable can be added to the physical constraint equation, which can take into account the impact of temperature on the moisture drying rate, making the physical constraint more in line with the actual application scenario when the human body wears clothes.

[0014] Step 6: Construct and train an LSTM neural network (Long Short-Term Memory Network) to predict the coefficients in the physical constraint equation based on the humidity characteristics based on the time series, so that the physical constraint equation can accurately calculate the drying rate of the clothing moisture. The physical constraint loss function of the LSTM neural network specifically includes: ; in, Represents the value of the physical constraint loss function; All represent the corresponding first weight coefficients; Indicates the predicted water content; Indicates the actual water content.

[0015] In a feasible implementation, the physical constraint loss function of the LSTM neural network specifically further includes: ; in, Represents the second weight coefficient (which can be set based on experience and prediction results. When , the LSTM neural network is driven by pure data fitting; when , the LSTM neural network is driven by pure physical constraints); In this way, the temperature variable can be added to the physical constraint loss function, thereby taking into account the effect of temperature on the moisture drying rate.

[0016] In a feasible embodiment, the application method of the clothing moisture drying rate test method includes: A second test scenario is set up, specifically including setting up an LED cold light source, a heating plate, a pipette and a CMOS camera in a constant temperature and humidity environment; in the second test scenario, a preset volume of test water is evenly injected into the tested fabric through the pipette, the tested fabric is heated through the heating plate, and the tested fabric image is captured in a time series according to a preset shooting frequency through the CMOS camera, and the image is input into the image segmentation model to extract the image of the region of interest; then, the humidity features in the image of the region of interest are extracted through the CNN network; then, the coefficients of the physical constraint equation are predicted through the LSTM neural network, and the moisture drying rate of the fabric is calculated through the physical constraint equation; in this way, through a simple test scenario and a trained model, the moisture drying rate of various complex fabrics can be non-contact tested, which greatly improves the test efficiency and effect.

[0017] In a second aspect, based on the same inventive concept, the present application also provides a clothing moisture drying rate testing system using the above method, comprising a data acquisition module, a data processing module, and a result generation module; The data acquisition module is used to acquire images of the tested clothing; The data processing module includes an image segmentation unit, a CNN network unit and an LSTM neural network unit; The image segmentation unit is used to extract an image of a region of interest from the image of the measured clothing; The CNN network unit is used to extract humidity features from the image of the region of interest; The LSTM neural network unit predicts the moisture drying rate of the tested clothing based on the humidity characteristics; The result generation module is used to send the prediction results externally.

[0018] On the third aspect, based on the same inventive concept, the present application also provides a clothing moisture drying rate testing device, including a processor, a memory and a bus, wherein the memory stores instructions and data that can be read by the processor, and the processor is used to call the instructions and data in the memory to execute the clothing moisture drying rate testing method as described above, and the bus connects the functional components for transmitting information.

[0019] In a feasible embodiment, the device further comprises a constant temperature and humidity chamber; in the constant temperature and humidity chamber, an LED cold light source, a heating plate, a pipette gun and a CMOS camera are provided; The heating plate is horizontally arranged in the center of the inner cavity of the constant temperature and humidity chamber; the CMOS camera is arranged above the heating plate, with the lens of the CMOS camera facing the upper surface of the heating plate; the LED cold light sources are evenly distributed around the CMOS camera; the pipette is retractably arranged at one end of the inner cavity of the constant temperature and humidity chamber so as to flexibly inject test water into various positions on the tested clothing without blocking the CMOS camera from shooting.

[0020] In a feasible embodiment, a weighing sensor is further provided under the heating plate for weighing the heating plate and its load.

[0021] In a feasible implementation manner, an infrared thermal imager is further provided in the constant temperature and humidity chamber, and a lens of the infrared thermal imager faces the upper surface of the heating plate.

[0022] In a feasible implementation manner, a fan and a wind speed sensor are further provided in the constant temperature and humidity chamber, the air outlet of the fan faces the upper surface of the heating plate, and the wind speed sensor is provided at the air outlet.

[0023] By adopting the above technical solution, the present invention has the following beneficial effects: The present invention provides a method, system, and device for testing the moisture drying rate of clothing. Using simple testing tools, a test scenario that aligns with a real-life human wearing scenario can be constructed. By taking and processing an overall photograph of the tested clothing, spatial distribution information of data such as the drying rate can be obtained, reflecting complex mass transfer mechanisms and overall drying progress. The method is suitable for testing complex fabrics such as spliced and composite fabrics. This solution employs non-contact measurement, is easy to operate, and can effectively improve test efficiency and effectiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 A flow chart of a method for testing the drying rate of moisture in clothing provided by an embodiment of the present invention; Figure 2 A flow chart of the humidity feature extraction method provided in an embodiment of the present invention; Figure 3 A diagram of a clothing moisture drying rate testing system provided by an embodiment of the present invention; Figure 4A three-dimensional structural diagram of a device for testing the drying rate of moisture in clothing provided by an embodiment of the present invention; Figure 5 for Figure 4 Another angle view of Reference numerals: 1- Constant temperature and humidity chamber; 2- LED cold light source; 3- Heating plate; 4- Pipette; 5- CMOS camera. DETAILED DESCRIPTION

[0026] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0027] In the description of the present invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0028] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0029] The present invention will be further explained below with reference to specific embodiments.

[0030] It should also be noted that the following specific embodiments or specific implementations are a series of optimized settings listed in the present invention to further explain the specific content of the invention, and these settings can be combined or used in association with each other.

[0031] Example 1: like Figure 1 As shown, this embodiment provides a method for testing the drying rate of moisture in clothing, comprising the following steps: Step 1: Set up the first test scene, specifically including setting up an LED cold light source, a heating plate, a pipette, a weighing sensor, a CMOS camera, an infrared thermal imager, etc. in a constant temperature and humidity environment (e.g., 28±0.5°C and 50%±2% humidity); The heating plate is used to heat the clothing sample at a preset temperature (e.g., 37°C); The pipette is used to inject a preset volume (e.g., 0.1 mL / cm²) of test water (e.g., deionized water or simulated sweat) into the clothing sample; The weighing sensor is used to weigh the clothing sample; The CMOS camera is used to photograph the area where the clothing sample is located; The infrared thermal imager is used to collect the actual temperature of the clothing sample; This makes it easy and effective to build a test scene that simulates a human body wearing clothes.

[0032] Furthermore, the first test scenario also includes a fan and a wind speed sensor, which are used to blow air to the clothing sample at a preset wind speed (for example, 1.5 m / s) to simulate the wind speed in a real wearing environment and accelerate the water evaporation process.

[0033] Step 2: After evenly injecting a preset volume of test water (e.g., 2 mL or the saturated water absorption capacity of the fabric sample) into different types of fabric samples using a pipette, heat the fabric samples using a heating plate. Use a CMOS camera to capture fabric sample images in a time series at a preset shooting frequency (e.g., 2 Hz in the early stages of drying and 0.5 Hz in the later stages), and simultaneously record the fabric sample weight using a weighing sensor until the weighing sensor value stops changing (i.e., the fabric is completely dry). This allows for the simultaneous collection of fabric image data and fabric weight data during the drying process, providing data support for subsequent processing.

[0034] Step 3: Using an image segmentation model (e.g., a conventional threshold masking method or a U-Net image segmentation model), extract a region of interest (ROI) image from the clothing sample image. The ROI image only includes the clothing sample. This allows for capturing and sampling clothing samples of different sizes without adjusting the CMOS camera's viewfinder, while automatically avoiding background noise interference in the clothing sample image. Then, using a conventional CNN network (three-layer structure), extract the humidity features from the ROI image.

[0035] Furthermore, if Figure 2 As shown, the method for extracting humidity features specifically includes: Step 31. Convert the image format of the region of interest (e.g., RGB) to LAB format using a color space conversion method (e.g., OpenCV's cvtColor function), separating the L (luminance) channel, the A (green-red) channel, and the B (blue-yellow) channel. This allows us to exploit the high correlation between the L channel and the human eye's perceived brightness to simply capture the detail of the darkening caused by wetness. Specifically, when the fabric is wet, the increased moisture reduces the surface diffuse reflectance, resulting in a significant decrease in the L value. Step 32: Calculate the mean brightness and dynamic contrast based on the L channel: The brightness mean is used to directly reflect the overall brightness change. That is, when it is wet, the brightness mean decreases, and when it is dry, the brightness mean recovers, thereby directly reflecting the reflectivity change of the clothing surface. The specific calculation formula includes: ; in, represents the mean brightness; Represents the total number of pixels in a single frame of clothing sample image; Indicates the The brightness value of each pixel (the range before quantization is 0~100, and the range after quantization is 0~255); The dynamic contrast is used to quantify local brightness fluctuations by calculating the ratio of the brightness standard deviation to the brightness mean, thereby reflecting the contrast reduction caused by the homogenization of the water film in the wet area; For example, under different drying conditions, the brightness mean, brightness standard deviation and dynamic contrast values are shown in Table 1;

[0036] Step 33: Convert the image of the region of interest into a grayscale image and calculate the information entropy to quickly and easily reflect the texture complexity. That is, when the clothing is wet, water stains will cause the texture of the clothing fiber structure to be smoothed, thereby causing the information entropy value to decrease. When the clothing is dry, the clothing fiber structure is restored, resulting in an increase in the information entropy value. The specific formula includes: ; in, represents information entropy; Indicates the The probability distribution of the gray value of each pixel.

[0037] Furthermore, after calculating the brightness mean and the dynamic contrast, step 32 further includes: Step 321: Calculate chromaticity based on the A channel and the B channel. The specific formula includes: ; in, Indicates chromaticity; Indicates the component value of channel A (the value range before quantization is -128~127, with positive values tending to be red and negative values tending to be green; the value range after quantization is 0~255); Represents the component value of the B channel (the value range before quantization is -128 to 127, with positive values tending to be yellowish and negative values tending to be bluish; the value range after quantization is 0 to 255). This can take into account the characteristics of dark fabrics such as red that deepen and increase in chroma when wet, and the characteristics of multicolored fabrics that become more uniform and have lower chroma when wet, thereby improving the applicability of the humidity feature to dark and multicolored fabrics. For example, the chromaticity and its normalized value under different drying conditions are shown in Table 2;

[0038] Step 322: Calculate comprehensive features. The specific formula includes: ; in, All represent corresponding weights; This helps to significantly increase the applicability of the humidity characteristics to clothing of different colors.

[0039] Furthermore, in other embodiments, step 33 may also be replaced by: Step 34: Convert the region of interest image into a grayscale image and calculate the gray level co-occurrence matrix (GLCM) contrast , the specific formula is: ; in, Indicates the Grayscale value and The joint probability of a pixel pair composed of grayscale values in a preset direction and a preset spacing; Represents the total number of grayscale values. Compared with information entropy, this method can more effectively quantify the dynamic changes of clothing surface texture during the wetting-drying process. Combining multi-directional and multi-spacing feature extraction with subsequent physical constraints can significantly improve the accuracy and robustness of drying rate prediction, but the computational complexity will increase. For example, under different drying conditions, when the preset distance is 1 gray level and the preset angle is 0°, the contrast The values are shown in Table 3;

[0040] Step 4: Build a regression model based on humidity characteristics and clothing weight data (A general mathematical model that quantitatively describes the statistical relationship of a sample) is used to establish the relationship between humidity characteristics and moisture content of clothing. The specific expressions include: ; in, Indicates the moisture content of the fabric; Indicates humidity characteristics.

[0041] Step 5: Construct the physical constraint equations, including the mixed drying equation. The specific formula is: ; in, Indicates the equilibrium moisture content (i.e. the moisture content of the clothing sample when it is in equilibrium with the ambient humidity); Indicates time; Indicates the surface evaporation coefficient (unit: ), Used to describe the surface evaporation process, which conforms to the linear mass transfer law; represents the diffusion correlation coefficient (unit is ), Used to describe the diffusion process inside porous media, which conforms to the nonlinear mass transfer law; as long as these two coefficients are predicted, the drying rate of the cloth can be calculated. .

[0042] Furthermore, the physical constraint equation also includes a thermodynamic-mass transfer coupling equation, and the specific formula is: ; in, Indicates the actual temperature of the clothing sample (Unit: Kelvin, obtained by converting data collected by infrared thermal imager) dependent evaporation coefficient; represents the latent heat of vaporization of water; Indicates the thickness of the fabric; The spatial second-order derivative of moisture content is used to describe the diffusion behavior of moisture in the fabric (if the value is greater than 0, moisture will diffuse from the area to the surrounding area; if the value is less than 0, the surrounding moisture will converge to the area); The temperature-dependent diffusion coefficient can be calculated using the Arrhenius equation as follows: ; in, represents the pre-factor related to the pore structure of the fabric; represents the diffusion activation energy, which is used to reflect the energy barrier that needs to be overcome for diffusion; represents the ideal gas constant; In this way, the temperature variable can be added to the physical constraint equation, taking into account the effect of clothing temperature on the moisture drying rate, making the physical constraint more consistent with the actual application scenario of the human body wearing clothing; For example, three kinds of clothing are dried under different conditions. 、 and The values at 300K are shown in Table 4;

[0043] Step 6: Construct and train a conventional LSTM neural network (Long Short-Term Memory Network) to input the humidity characteristics based on the time series and predict the coefficients in the physical constraint equation so that the physical constraint equation can accurately calculate the moisture drying rate of the clothing. The physical constraint loss function of the LSTM neural network specifically includes: ; in, Represents the value of the physical constraint loss function; All represent the corresponding first weight coefficients; Indicates the predicted water content; Indicates the actual water content.

[0044] Furthermore, the physical constraint loss function of the LSTM neural network specifically includes: ; in, Represents the second weight coefficient (which can be set based on experience and prediction results. When , the LSTM neural network is driven by pure data fitting; when , the LSTM neural network is driven by pure physical constraints); this takes the temperature of the fabric into account.

[0045] Furthermore, the application method of the clothing moisture drying rate test method includes: A second test scenario is set up, specifically including setting up an LED cold light source, a heating plate, a pipette and a CMOS camera in a constant temperature and humidity environment; in the second test scenario, a preset volume of test water is evenly injected into the tested fabric through the pipette, the tested fabric is heated through the heating plate, and the tested fabric image is captured in a time series according to a preset shooting frequency through the CMOS camera, and the image is input into the image segmentation model to extract the image of the region of interest; then, the humidity features in the image of the region of interest are extracted through the CNN network; then, the coefficients of the physical constraint equation are predicted through the LSTM neural network, and the moisture drying rate of the fabric is calculated through the physical constraint equation; in this way, through a simple test scenario and a trained model, the moisture drying rate of various complex fabrics can be non-contact tested, which greatly improves the test efficiency and effect.

[0046] Example 2: like Figure 3 As shown, this embodiment provides a clothing moisture drying rate testing system using the above method, including a data acquisition module, a data processing module and a result generation module; The data acquisition module is used to acquire images of the tested clothing; The data processing module includes an image segmentation unit, a CNN network unit and an LSTM neural network unit; The image segmentation unit is used to extract an image of a region of interest from the image of the measured clothing; The CNN network unit is used to extract humidity features from the image of the region of interest; The LSTM neural network unit predicts the moisture drying rate of the tested clothing based on the humidity characteristics; The result generation module is used to send the prediction results externally.

[0047] Example 3: like Figure 4-5 As shown, this embodiment provides a clothing moisture drying rate testing device, including a processor, a memory and a bus, wherein the memory stores instructions and data that can be read by the processor, and the processor is used to call the instructions and data in the memory to execute the clothing moisture drying rate testing method as described above, and the bus connects the functional components to transmit information.

[0048] Furthermore, the processor is an embedded AI processor (computing power ≥ 4TOPS).

[0049] Furthermore, the device also includes a constant temperature and humidity chamber 1; in the constant temperature and humidity chamber 1, an LED cold light source 2 (for example, a color temperature of 5500K, an illumination of 2000±50lux), a heating plate 3, a pipette gun 4 and a CMOS camera 5 are provided; The heating plate 3 is horizontally arranged in the center of the inner cavity of the constant temperature and humidity chamber 1; the CMOS camera 5 is arranged above the heating plate 3, with the lens of the CMOS camera 5 facing the upper surface of the heating plate 3; the LED cold light sources 2 are evenly distributed around the CMOS camera 5; the pipette gun 4 (via a two-dimensional CNC guide rail) is retractably arranged at one end of the inner cavity of the constant temperature and humidity chamber 1 so as to flexibly inject test water into various locations on the tested clothing without blocking the CMOS camera 5 from shooting.

[0050] Furthermore, a weighing sensor (with an accuracy of 0.1 mg) is provided under the heating plate 3 for weighing the heating plate and its load.

[0051] Furthermore, an infrared thermal imager (spatial resolution 0.1° C., not shown in the figure) is also provided in the constant temperature and humidity chamber 1 , and the lens of the infrared thermal imager faces the upper surface of the heating plate.

[0052] Furthermore, a fan and a wind speed sensor (not shown in the figure) are also provided in the constant temperature and humidity chamber 1. The air outlet of the fan faces the upper surface of the heating plate 3, and the wind speed sensor is provided at the air outlet.

[0053] Furthermore, the heating plate 3 includes a multi-layer ceramic heating structure and a PID controller, which can accurately simulate the heat flux density of human skin (50-100W / m²).

[0054] Furthermore, the heating rate of the heating plate 3 in the initial stage is ≤2°C / s to avoid thermal shock damage to the clothes.

[0055] Furthermore, a sample clamp (not shown in the figure) is provided on the upper surface of the heating plate 3 for clamping the clothing sample.

[0056] Furthermore, the accuracy of the humidity sensor (not shown in the figure) built into the constant temperature and humidity chamber 1 is ±1% RH.

[0057] Furthermore, the CMOS camera 5 adopts a high frame rate industrial grade (eg, frame rate ≥ 120 fps, resolution ≥ 20 million pixels) to capture high-definition images at a high frequency.

[0058] Furthermore, an operation panel and a control box (not shown in the figure) are provided at the bottom of the constant temperature and humidity chamber 1; the control box is electrically connected to the operation panel, the LED cold light source 2, the heating plate 3, the pipette gun 4, the CMOS camera 5, the weighing sensor, the infrared thermal imager, the fan, the wind speed sensor and the humidity sensor; the processor and the memory are provided in the control box.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for testing the drying rate of clothing moisture, characterized in that: include: Step 1: Set up the first test scene, specifically including setting up an LED cold light source, a heating plate, a pipette, a weighing sensor, a CMOS camera, and an infrared thermal imager in a constant temperature and humidity environment; The heating plate is used to heat the clothing sample according to a preset temperature; The pipette is used to inject a preset volume of test water into the clothing sample; The weighing sensor is used to weigh the clothing sample; The CMOS camera is used to photograph the area where the clothing sample is located; The infrared thermal imager is used to collect the actual temperature of the clothing sample; Step 2: After evenly injecting a preset volume of test water into the fabric sample using a pipette, the fabric sample is heated using a heating plate. A CMOS camera is used to capture images of the fabric sample in a time series at a preset shooting frequency, and a load cell is used to simultaneously record the fabric sample weight until the load cell value stops changing. Step 3: extracting the region of interest image from the clothing sample image using an image segmentation model; The region of interest image only includes clothing samples; Then, the humidity features in the image of the region of interest are extracted through the CNN network; Step 4: Build a regression model based on humidity characteristics and clothing weight data , which is used to establish the relationship between humidity characteristics and moisture content of clothing. The specific expressions include: ; in, Indicates the moisture content of the fabric; Indicates humidity characteristics; Step 5: Construct the physical constraint equations, including the mixed drying equation. The specific formula is: ； in, Indicates equilibrium water content; Indicates time; represents the surface evaporation coefficient; represents the diffusion correlation coefficient; Indicates the drying rate of clothing moisture; Step 6: Construct and train an LSTM neural network to predict the coefficients in the physical constraint equation based on the humidity characteristics based on the time series; the physical constraint loss function of the LSTM neural network specifically includes: ； in, Represents the value of the physical constraint loss function; All represent the corresponding first weight coefficients; Indicates the predicted water content; Indicates the actual water content.

2. The method according to claim 1, characterized in that The first test scenario also includes a fan and a wind speed sensor, which are used to blow air toward the clothing sample at a preset wind speed.

3. The method according to claim 1, characterized in that The method for extracting the humidity feature specifically includes: Step 31: Convert the image format of the region of interest into LAB format by using a color space conversion method, and separate the L channel, A channel, and B channel; Step 32: Calculate the mean brightness and dynamic contrast based on the L channel: The brightness mean is used to directly reflect the overall brightness change. The specific calculation formula includes: ； in, represents the mean brightness; Represents the total number of pixels in a single frame of clothing sample image; Indicates the The brightness value of each pixel; The dynamic contrast ratio is used to quantify local brightness fluctuations by calculating the ratio of the brightness standard deviation to the brightness mean; Step 33: Convert the image of the region of interest into a grayscale image and calculate the information entropy. The specific formula includes: ； in, represents information entropy; Indicates the The probability distribution of the gray value of each pixel.

4. The method according to claim 3, characterized in that After calculating the brightness mean and the dynamic contrast, step 32 further includes: Step 321: Calculate chromaticity based on the A channel and the B channel. The specific formula includes: ； in, Indicates chromaticity; Indicates the component value of channel A; Indicates the component value of the B channel; Step 322: Calculate comprehensive features. The specific formula includes: ； in, All represent the corresponding weights.

5. The method according to claim 3, characterized in that The step 33 is replaced by: Step 34: Convert the region of interest image into a grayscale image and calculate the grayscale co-occurrence matrix contrast , the specific formula is: ； in, Indicates the Grayscale value and The joint probability of a pixel pair composed of grayscale values in a preset direction and a preset spacing; Indicates the total number of grayscale values.

6. The method according to claim 1, wherein The physical constraint equation also includes a thermodynamic-mass transfer coupling equation, and the specific formula is: ； in, Indicates the actual temperature of the clothing sample Dependent evaporation coefficient; represents the latent heat of vaporization of water; Indicates the thickness of the fabric; The spatial second-order derivative of moisture content is used to describe the diffusion behavior of moisture inside the fabric; The temperature-dependent diffusion coefficient is calculated using the Arrhenius equation. The specific expression is: ； in, represents the pre-factor related to the pore structure of the fabric; represents the diffusion activation energy, which is used to reflect the energy barrier that needs to be overcome for diffusion; represents the ideal gas constant.

7. The method according to claim 6, characterized in that The physical constraint loss function of the LSTM neural network specifically includes: ； in, Represents the second weight coefficient.

8. The method according to claim 1, characterized in that The application method of the clothing moisture drying rate test method includes: A second test scene is set up, specifically including setting up an LED cold light source, a heating plate, a pipette and a CMOS camera in a constant temperature and humidity environment; in the second test scene, a preset volume of test water is evenly injected into the tested fabric through the pipette, the tested fabric is heated by the heating plate, and the tested fabric image is captured in a time series according to a preset shooting frequency through the CMOS camera, and the image is input into the image segmentation model to extract the image of the region of interest; then, the humidity features in the image of the region of interest are extracted through the CNN network; then, the coefficients of the physical constraint equation are predicted through the LSTM neural network, and the moisture drying rate of the fabric is calculated through the physical constraint equation.

9. A system for testing the drying rate of clothing moisture using the method according to any one of claims 1 to 8, characterized in that: It includes data acquisition module, data processing module and result generation module; The data acquisition module is used to acquire images of the tested clothing; The data processing module includes an image segmentation unit, a CNN network unit and an LSTM neural network unit; The image segmentation unit is used to extract an image of a region of interest from the image of the measured clothing; The CNN network unit is used to extract humidity features from the image of the region of interest; The LSTM neural network unit predicts the moisture drying rate of the tested clothing based on the humidity characteristics; The result generation module is used to send the prediction results externally.

10. A device for testing the drying rate of clothing moisture, characterized in that: It includes a processor, a memory and a bus, the memory stores instructions and data read by the processor, the processor is used to call the instructions and data in the memory to execute the method according to any one of claims 1 to 8, and the bus connects the functional components to transmit information.

Images