A Visual Analysis Method for Defects in Spunlace Nonwoven Fabric

Abstract

This invention relates to a visual analysis method for defects in spunlace nonwoven fabrics, belonging to the field of defect inspection in spunlace nonwoven fabrics. The method includes acquiring a top-view infrared thermal image in response to a fabric arrival signal, whereby the fabric arrival signal is the signal indicating that the spunlace nonwoven fabric to be inspected has been placed on a preset defect detection device; analyzing the temperature distribution and average temperature of the spunlace nonwoven fabric using the top-view infrared thermal image; comparing the temperature distribution with the average temperature to obtain abnormal temperature zones and corresponding abnormal temperatures; analyzing the defects and their corresponding defect categories based on the abnormal temperatures and the average temperature; and outputting a defect signal formed by the abnormal temperature zones and defect categories. This application has the advantages of efficiently determining whether defects exist in nonwoven fabrics and accurately analyzing the defect types.

Description

Technical Field

[0001] This invention relates to the field of defect inspection in spunlace nonwoven fabrics, and in particular to a visual analysis method for defects in spunlace nonwoven fabrics. Background Technology

[0002] Currently, spunlace nonwoven fabrics, with their softness, breathability, strong moisture absorption, and environmentally friendly biodegradability, are widely used in medical protection, hygiene care, industrial filtration, and household cleaning, becoming an indispensable basic material in modern life. However, in actual use, some products have been found to have problems such as pore blockage leading to decreased water permeability and residual water marks on the surface affecting aesthetics. Therefore, it is necessary to conduct quality inspections on finished products to reduce the defect rate.

[0003] Currently, defect detection for spunlace nonwoven fabrics mainly relies on manual inspection. Operators visually scan the fabric surface under natural light or standard light sources to identify obvious defects such as holes, dirt, and fiber agglomeration.

[0004] Regarding the aforementioned technologies, relying on workers to visually inspect surface defects is inefficient, susceptible to subjective factors, and has a high rate of missed detections, especially making it difficult to detect hidden defects such as internal pore blockage. Moreover, the entire process depends on individual sensory acuity and experience, and is easily affected by fatigue, changes in lighting, and differences in subjective judgment, and it is difficult to quantify the micro-pore structure or deep fiber distribution. Summary of the Invention

[0005] To more scientifically and accurately identify defects in spunlace nonwoven fabrics, this invention provides a visual analysis method for defects in spunlace nonwoven fabrics.

[0006] This invention provides a visual analysis method for defects in spunlace nonwoven fabrics, employing the following technical solution:

[0007] A visual analysis method for defects in spunlace nonwoven fabrics, characterized by comprising:

[0008] Step 1: In response to the fabric arrival signal, acquire a top-view infrared thermal image. The fabric arrival signal is the signal that the spunlace nonwoven fabric to be inspected is placed on a preset defect detection device. The defect detection device includes a conveyor wheel for placing the spunlace nonwoven fabric and a nozzle for spraying hot air. There is a space between the upper and lower conveyor wheels for the spunlace nonwoven fabric to pass through. The nozzle is located on the lower side of the spunlace nonwoven fabric and faces the spunlace nonwoven fabric, and sprays hot air evenly onto the spunlace nonwoven fabric. The defect detection device also includes an infrared detector located above the spunlace nonwoven fabric.

[0009] Step 2: Analyze the temperature distribution and average temperature of the spunlace nonwoven fabric using the top-view infrared thermal imaging image;

[0010] Step 3: Compare the temperature distribution with the average temperature to obtain the abnormal temperature zone and the corresponding abnormal temperature;

[0011] Step 4: Analyze the defects and their corresponding defect categories based on the abnormal temperature and the average temperature;

[0012] Step 5: Output the defect signal by forming a defect signal from the abnormal temperature zone and defect category.

[0013] By employing the aforementioned technical solution, efficient detection of defects in spunlace nonwoven fabrics was successfully achieved using dynamic hot gas penetration and infrared thermal imaging technology. The core of this method lies in uniformly spraying hot gas into the spunlace nonwoven fabric, combined with an infrared detector to capture the material's temperature distribution in real time. Through comparison of coarse and fine grid segments, multi-level temperature range determination, and analysis of the thermodynamic characteristics of defect areas, the method accurately identifies defect types such as pore blockage, holes, and watermarks. This method uses scientifically quantified temperature differences as the criterion, replacing the subjectivity and inefficiency of traditional manual visual inspection. It quickly locates abnormal temperature areas and classifies defects through thermal imaging, not only improving detection efficiency and coverage but also penetrating the surface to identify internal structural defects. This effectively solves the problem of traditional methods struggling to balance efficiency, accuracy, and comprehensiveness, providing objective and reliable technical support for nonwoven fabric quality control.

[0014] Optionally, the specific method for obtaining the abnormal temperature zone by comparing the temperature distribution with the average temperature includes:

[0015] Step 30: Divide the top-view infrared thermal image into multiple intervals according to the preset coarse grid segmentation method, and define the divided areas as coarse intervals;

[0016] Step 31: Divide the temperature distribution according to the coarse interval to obtain the average temperature of the coarse interval;

[0017] Step 32: Calculate the coarse-division interval temperature range based on the average temperature and the preset coarse-division fluctuation parameters;

[0018] Step 33: Define the coarse interval whose average temperature does not fall within the coarse interval temperature range as the coarse interval abnormal temperature range;

[0019] Step 34: Divide the coarse-divided abnormal temperature range into multiple ranges according to the preset fine-grid segmentation method, and define the divided ranges as subdivided ranges;

[0020] Step 35: Divide the temperature distribution according to the subdivided intervals to obtain the average temperature of the subdivided intervals;

[0021] Step 36: Obtain the subdivided temperature range based on the average temperature and the preset subdivided fluctuation parameters;

[0022] Step 37: Define the subdivided interval whose average temperature does not fall within the subdivided interval temperature range as the abnormal temperature zone and output it.

[0023] By adopting the above technical solution and utilizing a coarse-to-fine grid hierarchical segmentation strategy, rapid screening and precise location of abnormal temperature zones are achieved. First, a coarse grid segmentation method is used to determine the approximate location of defects. Then, a fine grid segmentation method is used to further subdivide the coarsely segmented abnormal zone, precisely locating the defect area caused by minute temperature differences. This solves the problem that a single segmentation method cannot accurately obtain the defect location and is prone to resulting in segments that are too large or too small. This method of first coarse segmentation and then fine segmentation significantly improves the reliability of detection and its practicality in engineering.

[0024] Optionally, the specific method for analyzing the defect and its corresponding defect category based on the abnormal temperature and the average temperature includes:

[0025] Step 40: Traverse all the abnormal temperature zones to form defect interval groups. Within each defect interval group, there is at least one path connecting any two defect intervals through adjacent defect intervals.

[0026] Step 41: The defect area group is modified using the top-view infrared thermal image to obtain the complete defect area;

[0027] Step 42: Divide the temperature distribution according to the intact area of ​​the defect to obtain the average temperature of the defect area;

[0028] Step 43: Expand the defective area outward by a predetermined heat diffusion width to obtain a surrounding defect area;

[0029] Step 44: Divide the surrounding defect area according to the temperature distribution to obtain the average temperature of the surrounding area;

[0030] Step 440: When the average temperature of the surrounding area exceeds the temperature range of the subdivision interval and the average temperature of the defect area is lower than the temperature range of the subdivision interval, the defect is defined as a blockage defect and a blockage defect signal is output.

[0031] Step 441: When the average temperature of the surrounding area falls within the temperature range of the subdivision interval and the average temperature of the defective area is lower than the temperature range of the subdivision interval, the defect is defined as a hole defect and a hole defect signal is output.

[0032] Step 442: When the average temperature of the surrounding area is lower than the temperature range of the subdivision interval and the average temperature of the defect area is lower than the temperature range of the subdivision interval, the defect is defined as a watermark defect and a watermark defect signal is output.

[0033] By employing the aforementioned technical solution, the temperature difference between abnormal temperature zones and the surrounding environment is systematically analyzed, enabling accurate classification of defects in spunlace nonwoven fabrics. First, discrete abnormal temperature zones are integrated into connected, complete defect areas, and their average temperature is calculated. Then, a predetermined width is expanded outward to obtain the surrounding area, and its corresponding temperature is calculated. Finally, by comparing the temperatures of the two regions and combining this with subdivided temperature range rules, the defects are categorized as blockages, holes, or watermarks. This approach solves the problem of pore blockage defects, which cannot be efficiently identified by methods such as light analysis and machine learning, due to temperature differences. This method improves the scientific rigor and objectivity of defect classification, provides data support for targeted process optimization, and significantly reduces the risk of processing errors caused by confusion in defect types.

[0034] Optionally, the method further includes a method for not outputting a blockage defect signal when the average temperature of the surrounding area exceeds the subdivision temperature range and the average temperature of the defect area is lower than the subdivision temperature range, the method comprising:

[0035] Step 4400: Calculate the temperature difference of the blocked area by using the average temperature of the defective area and the average temperature of the surrounding area;

[0036] Step 4401: Obtain the blockage degree value from the preset temperature blockage degree table based on the temperature difference of the blockage area;

[0037] Step 4402: Obtain the defect area based on the complete defect area;

[0038] Step 4403: Obtain the degree of local blockage based on the defect area and the blockage degree value;

[0039] Step 44030: When the local blockage degree is higher than the preset blockage threshold, the defective complete area corresponding to the blockage degree value is defined as the completely blocked area, and a blockage defect signal is output.

[0040] Step 44031: When the local blockage level is lower than the blockage threshold, the defective complete area corresponding to the blockage level value is defined as a partially blocked area, and no blockage defect signal is output.

[0041] By employing the aforementioned technical solution, the degree of blockage is quantified based on the temperature difference between the defective area and its surroundings. Simultaneously, the degree of local blockage is comprehensively assessed by combining the defect area with the assessment. Finally, a threshold is used to automatically distinguish between completely and partially blocked areas. Its core solution addresses the problem of traditional manual inspection's difficulty in accurately quantifying the severity of blockage. By using data from both temperature difference and area, it achieves graded determination of blockage defects. This method eliminates subjective judgment bias with a standardized temperature-blockage comparison table and balances the relationship between the degree of local blockage and the scope of influence through a local blockage degree model.

[0042] Optional, also includes:

[0043] Step 4404: Obtain the total number of all the partially blocked areas, defined as the number of partially blocked areas;

[0044] Step 4405: Obtain the blockage degree value corresponding to all the partially blocked areas, and define it as the partially blocked degree value;

[0045] Step 4406: Calculate the air permeability value of the spunlace nonwoven fabric based on the partial blockage degree value and the number of partial blockage areas;

[0046] Step 4407: If the air permeability value of the spunlace nonwoven fabric is lower than the preset air permeability threshold, an overall air permeability defect signal is generated and output.

[0047] By adopting the above technical solution, a quantitative index is established and the overall air permeability value of the spunlace nonwoven fabric is calculated by statistically analyzing the number of partially blocked areas and their corresponding degree of blockage, ultimately determining whether it is lower than a preset air permeability threshold. This approach solves the problem of air permeability assessment bias caused by neglecting the cumulative effect of partial blockage in traditional testing. Through a mathematical model, discrete partial blockage defects are transformed into systematic air permeability performance indicators, achieving accurate identification of material defects.

[0048] Optionally, when the average temperature of the surrounding area falls within the subdivided temperature range and the average temperature of the defective area is lower than the subdivided temperature range, a method may be used to not output a hole defect signal. This method includes:

[0049] Step 4410: Obtain the perimeter of the complete area of ​​the defect based on the top-view infrared thermal image, and define it as the defect perimeter;

[0050] Step 4411: Obtain the porosity of the intact area of ​​the defect based on the defect area and the defect perimeter;

[0051] Step 4412: When the roundness of the pore is greater than the preset roundness threshold, the corresponding intact area of ​​the defect is defined as a normal pore area, and the hole defect signal may not be output.

[0052] By employing the above technical solution, the roundness of pores in defective areas is calculated to filter out abnormally shaped areas with regular shapes. When the roundness exceeds a threshold, it is judged as a normal pore rather than a hole defect. This solves the problem of misjudging regular pores as holes due to abnormal temperatures. Through geometric feature analysis, it effectively distinguishes between inherent process pores and actual damage defects, significantly reducing the false alarm rate while ensuring the hole detection rate and improving the reliability of the quality inspection system's results.

[0053] Optional, also includes:

[0054] Step 4413: Obtain the pore core coordinates of all the normal pore regions, where the pore core coordinates refer to the coordinates of the geometric center of the normal pore region;

[0055] Step 4414: Arrange all the pore core coordinates to form a pore core coordinate group;

[0056] Step 4415: Calculate the lateral distance and longitudinal distance between any two adjacent pore core coordinates in the pore core coordinate group, and define them as the pore lateral coordinate spacing and the pore longitudinal coordinate spacing.

[0057] Step 4416: Obtain the standard pore transverse spacing based on the statistical analysis of all the pore transverse coordinate spacings;

[0058] Step 4417: Obtain the standard pore longitudinal spacing based on the statistical analysis of all the pore longitudinal coordinate spacings;

[0059] Step 4418: Construct a pore arrangement grid using the coordinates of any of the pore cores, the standard pore lateral spacing, and the standard pore longitudinal spacing;

[0060] Step 4419: Obtain the alignment matching degree based on the coordinates of the pore core and the pore arrangement grid;

[0061] Step 4420: When the alignment matching degree is lower than the preset matching degree threshold, the corresponding normal pore area is defined as an abnormal pore area, the defect is defined as an abnormal pore defect, and a preset abnormal pore defect signal is output.

[0062] By employing the above technical solution, the core coordinates of normal pores are extracted and their arrangement regularity is analyzed to construct a standard pore mesh model. This quantifies the matching degree between the actual pore distribution and the ideal arrangement, thereby identifying structural anomalies caused by pore disorder. This solution solves the problem of identifying abnormal circular holes as standard circular holes. By statistically establishing a reference mesh through horizontal and vertical standard spacing, and combining it with a matching degree threshold to determine disordered areas, it effectively captures hidden structural defects caused by process fluctuations.

[0063] Optionally, a method is included that does not output the abnormal temperature zone, the method comprising:

[0064] Step 370: Obtain edge correction temperature parameters based on the average temperature of the subdivided intervals;

[0065] Step 371: Obtain the edge subdivision region temperature range based on the edge correction temperature parameters and the subdivision interval temperature range;

[0066] Step 372: When the defect interval coordinates fall into the preset edge coordinate set, the average temperature of the corresponding subdivided area is defined as the edge average temperature;

[0067] Step 3720: If the average edge temperature falls within the temperature range of the edge subdivision region, the defect signal may not be output.

[0068] Step 3721: If the average edge temperature does not fall within the temperature range of the edge subdivision region, then continue to output the defect signal.

[0069] By adopting the above technical solution and introducing edge correction temperature parameters and temperature ranges for edge subdivision regions, the defect judgment logic for edge areas has been specifically optimized. This solves the problem of false alarms caused by uneven heat dissipation and environmental interference in traditional detection. While maintaining strict detection standards for core areas, a more lenient judgment threshold is applied to edge areas. This avoids invalid defect signals caused by natural temperature gradients at the edges, while ensuring that real defects can still be effectively identified, significantly improving the anti-interference capability and reliability of the quality inspection system.

[0070] Optionally, the method for outputting the defect signal further includes:

[0071] Step 50: Obtain the total number of all the coarse-division abnormal temperature intervals, defined as the number of coarse-division abnormal intervals;

[0072] Step 51: When the number of coarse-divided abnormal intervals exceeds the preset abnormal interval threshold, the defect is defined as an overall quality defect, and a preset overall quality defect signal is output.

[0073] Step 52: Perform standard deviation analysis on the temperature of the top-view infrared thermal image between the coarse divisions, and calculate the corresponding standard deviation of the coarse division regions;

[0074] Step 53: When the standard deviation of the coarse segmentation region exceeds the preset abnormal standard deviation threshold, the defect is also defined as the overall quality defect, and the overall quality defect signal is output.

[0075] By employing the aforementioned technical solution, the number of coarsely segmented abnormal temperature ranges is statistically analyzed, and their standard deviations are examined. Defects are assessed from two dimensions: abnormal distribution density and dispersion. When the number of abnormal ranges exceeds a threshold or the standard deviation exceeds the abnormal range, it indicates that the fabric has multiple small-area defects or significant thickness unevenness, thus triggering an overall quality defect signal. This approach solves the problem of misjudgment caused by a single indicator in traditional testing. For example, scattered small-area anomalies may accumulate to form a systemic risk, or excessively high temperature dispersion may reflect material uniformity defects.

[0076] Optionally, the method for outputting the overall quality defect signal further includes:

[0077] Step 54: Obtain the number of standard abnormal regions and the standard deviation threshold of standard samples based on the preset standard top-down infrared spectrum;

[0078] Step 55: If the number of coarse-divided abnormal regions is greater than the number of standard abnormal regions or the standard deviation of the coarse-divided regions is greater than the standard deviation threshold of the standard sample, the defect is defined as the overall quality defect, and the overall quality defect signal is output.

[0079] By adopting the above technical solution, standard top-view infrared spectra are introduced to obtain the number of standard abnormal regions and the standard deviation threshold of standard samples as benchmarks. The coarse-grained abnormal data from actual detection are then scientifically compared with the standard values. This approach solves the problem of threshold setting relying on subjective experience in traditional methods. By comparing quantitative standards with real-time data, overall quality defects are objectively determined, significantly improving the accuracy and consistency of detection and reducing the risk of misjudgment or missed detection due to threshold deviations.

[0080] In summary, this application includes at least one of the following beneficial technical effects:

[0081] 1. This method uses infrared thermal imaging technology combined with temperature distribution analysis to efficiently determine whether there are defects in non-woven fabrics. Its scientific basis lies in utilizing the temperature field difference generated when hot gas penetrates the material. By comparing average temperatures, dividing the grid into coarse and fine sections, and determining multi-level temperature ranges, it can accurately locate abnormal temperature areas, avoiding the subjectivity and inefficiency of traditional manual inspection, and realizing objective quantitative analysis based on differences in physical properties.

[0082] 2. This method can effectively identify pore blockage defects. By comparing the temperature difference and blockage degree between the defective area and the surrounding area, combined with the air permeability calculation, it can distinguish between complete blockage and partial blockage, and quantitatively assess the impact of blockage on overall air permeability. This analysis method based on thermal conductivity and temperature gradient overcomes the technical limitations of traditional visual inspection or permeability testing in accurately judging the internal pore state.

[0083] 3. This method can detect both local defects and assess overall quality. By statistically analyzing the number of coarsely divided abnormal intervals, performing standard deviation analysis, and comparing with preset standard maps, it can determine whether there are systematic defects in the material. At the same time, by adjusting the temperature distribution of the subdivided mesh and calculating the mesh matching degree of the pore arrangement, it can penetrate surface features to capture internal fiber structure anomalies, achieving multi-level defect identification from micropores to macrotexture. Attached Figure Description

[0084] Figure 1 This is a flowchart of a visual analysis method for defects in spunlace nonwoven fabric according to an embodiment of this application.

[0085] Figure 2 This is a schematic diagram of the device for obtaining a top-view infrared thermal image in an embodiment of this application.

[0086] Figure 3 This is a schematic diagram of the defect interval group, the defect complete area, and the area surrounding the defect in the embodiments of this application. Detailed Implementation

[0087] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0088] This application discloses a method for visual analysis of defects in spunlace nonwoven fabrics. (Refer to...) Figure 1 A visual analysis method for defects in spunlace nonwoven fabrics includes:

[0089] A visual analysis method for defects in spunlace nonwoven fabrics, characterized by comprising:

[0090] Step 1: Acquire a top-view infrared thermal image in response to the fabric arrival signal. The fabric arrival signal is the signal indicating that the spunlace nonwoven fabric to be inspected is placed on a preset defect detection device.

[0091] Reference Figure 2 The defect detection equipment includes a conveyor wheel for placing the spunlace nonwoven fabric and a nozzle for spraying hot air. There is a space between the upper and lower conveyor wheels for the spunlace nonwoven fabric to pass through. The nozzle is located on the lower side of the spunlace nonwoven fabric and faces the spunlace nonwoven fabric, so as to spray hot air evenly onto the spunlace nonwoven fabric. The defect detection equipment also includes an infrared detector located above the spunlace nonwoven fabric for detecting the temperature distribution on the surface of the fabric.

[0092] The fabric arrival signal is the initial signal indicating that the spunlace nonwoven fabric to be inspected has been placed on the defect inspection equipment and is ready to trigger the subsequent inspection process.

[0093] A top-down infrared thermal image is a thermal image captured vertically downwards by an infrared detector positioned above the spunlace nonwoven fabric, including the surface temperature distribution of the spunlace nonwoven fabric after it has been heated.

[0094] The complete process for obtaining a top-down infrared thermal image is as follows: When the spunlace nonwoven fabric is conveyed to the predetermined position and triggers the fabric arrival signal, a high-speed hot gas stream of preset temperature is sprayed into the hot gas station through a nozzle, and a thermal image of the spunlace nonwoven fabric is acquired after a preset duration. The fabric arrival signal can be emitted by an infrared detector or by devices such as a light sensor; this is only for illustrative purposes.

[0095] Step 2: Analyze the temperature distribution and average temperature of the spunlace nonwoven fabric using top-down infrared thermal imaging.

[0096] Temperature distribution refers to the uneven heating phenomenon on the surface of spunlace nonwoven fabric caused by defects. Temperature distribution can be intuitively presented as the spatial distribution characteristics of different color temperature regions through top-view infrared thermal imaging. The analysis process is the basic method of thermal imaging and temperature conversion, which is common knowledge in the field and will not be elaborated here.

[0097] The average temperature is the overall mean temperature of all pixels in the thermal image, used to characterize the overall thermal state of the fabric after it has been heated. It is calculated by averaging all the temperature data.

[0098] Step 3: Compare the temperature distribution with the average temperature to obtain the abnormal temperature zones and their corresponding abnormal temperatures.

[0099] An abnormal temperature zone refers to a localized area on the surface of a spunlace nonwoven fabric where the temperature deviates significantly from the overall average temperature. This deviation is usually caused by defects in the spunlace nonwoven fabric, such as blockages, holes, and watermarks, and thus appears as a "high temperature zone" or "low temperature zone" in an infrared thermal imaging image. The temperature distribution can be compared with the average temperature simply by numerical comparison.

[0100] Step 4: Analyze the defects and their corresponding defect categories based on abnormal and average temperatures.

[0101] Defects refer to localized structural abnormalities caused by process anomalies or material defects during the production of spunlace nonwoven fabrics. Defect categories are different types of defects, classified according to their corresponding temperature distribution patterns, geometric morphological characteristics, and physical causes.

[0102] The system first merges adjacent abnormal temperature zones into a connected whole, eliminating isolated areas to ensure defect integrity. Then, it corrects the boundaries through temperature gradient fitting, forming a continuous region that matches the actual shape, and calculates the average temperature of this region. Simultaneously, based on preset thermal diffusion parameters, it extends outwards into a transition zone to obtain the average temperature of the surrounding area. Finally, the system determines the defect type by comparing the temperature relationship between the two: if the surrounding area experiences high-temperature accumulation due to impaired heat conduction while the defect area remains low, it is classified as a blockage defect; if the surrounding temperature is normal but the defect area is low due to structural defects, it is classified as a hole defect; if the surrounding area is low due to moisture absorption and the defect area shows an abnormally high specific heat capacity, it is identified as a watermark defect.

[0103] Step 5: Output the defect signal by forming a defect signal from the abnormal temperature zone and defect category.

[0104] A defect signal refers to information containing the coordinates of a defect and its corresponding defect category. The system automatically collects the coordinate range of the abnormal temperature zone, labels it with the corresponding defect category, and integrates it into an information signal output containing defect location and type. The purpose of this integrated output is to achieve efficient problem localization and provide accurate defect identification basis for subsequent quality control by combining the coordinate location of the abnormal area with the defect type label.

[0105] Specific methods for comparing temperature distribution with average temperature to identify anomalous temperature zones include:

[0106] Step 30: Divide the top-view infrared thermal image into multiple intervals according to the preset coarse grid segmentation method, and define the divided areas as coarse intervals.

[0107] The coarse grid segmentation method refers to a segmentation method that divides a top-view infrared thermal image into multiple rectangular regions of equal size or proportion based on a uniform grid generation rule with fixed rows and columns.

[0108] A coarse-divided interval is a rectangular region divided by a coarse-grid partitioning method. By using coarse-grid partitioning, the global temperature distribution is discretized into a finite number of intervals, reducing the complexity of data processing.

[0109] Step 31: Divide the temperature distribution into coarse intervals to obtain the average temperature of the coarse intervals.

[0110] The coarse-grid average temperature refers to the arithmetic mean of the temperature data of all pixels within each coarse-grid segment after dividing the top-view infrared thermal image into multiple coarse-grid segments of equal size or proportion using a preset coarse-grid segmentation method. This value reflects the overall heating state of the spunlace nonwoven fabric surface within the corresponding coarse-grid segment and is a fundamental indicator for subsequent comparison with the global average temperature to screen for potential abnormal temperature areas.

[0111] Step 32: Calculate the temperature range of the coarse division interval based on the average temperature and the preset coarse division fluctuation parameters.

[0112] The coarse-grained fluctuation parameter is a parameter set based on the material properties, production process, and historical test data of spunlace nonwoven fabric to obtain the allowable temperature fluctuation range. After conducting numerous tests on spunlace nonwoven fabric, the staff set the upper and lower fluctuation parameters corresponding to the maximum acceptable temperature range as the coarse-grained fluctuation parameter.

[0113] The coarse-division temperature range is a temperature range calculated based on the global average temperature and coarse-division fluctuation parameters. It is used to determine whether the average temperature of the coarse-division range falls within the normal fluctuation range. The coarse-division temperature range includes an upper and a lower limit. The upper limit of the coarse-division temperature range is equal to the average temperature of the coarse-division range plus the average temperature of the coarse-division range multiplied by the coarse-division fluctuation parameter. The lower limit of the coarse-division temperature range is equal to the average temperature of the coarse-division range minus the average temperature of the coarse-division range multiplied by the coarse-division fluctuation parameter.

[0114] Step 33: Define the coarse interval whose average temperature does not fall into the coarse interval temperature range as the coarse interval abnormal temperature range.

[0115] The coarse-divided abnormal temperature range refers to the coarse-divided range after being divided by the coarse grid segmentation method, in which the average temperature value exceeds the preset coarse-divided temperature range.

[0116] When the average temperature of the coarse mesh interval does not fall within the temperature range of the coarse mesh interval, it indicates that the temperature of the coarse mesh interval is significantly abnormal and is initially judged to be possibly due to material defects, such as holes or blockages. Further fine mesh segmentation is required for precise positioning.

[0117] Step 34: Divide the coarse-divided abnormal temperature range into multiple ranges according to the preset fine-grid segmentation method, and define the divided ranges as subdivided ranges.

[0118] The fine mesh segmentation method is a high-density mesh generation rule that further divides the coarse-divided anomaly temperature range. By increasing the number of rows and columns, the coarse-divided anomaly region is divided into smaller rectangular regions of equal size.

[0119] Subdivision intervals are rectangular areas divided by a fine grid segmentation method. By using smaller interval sizes, the specific boundaries and coordinates of defects can be located.

[0120] Step 35: Divide the temperature distribution into subdivided intervals to obtain the average temperature of each subdivided interval.

[0121] The average temperature of a subdivision interval refers to the value obtained by arithmetically averaging the temperature data of all pixels within a single subdivision interval.

[0122] Step 36: Obtain the subdivided temperature range based on the average temperature and the preset subdivided fluctuation parameters.

[0123] The subdivision fluctuation parameter is a parameter set based on the material properties, production process, and historical test data of the spunlace nonwoven fabric to obtain the allowable temperature fluctuation range under subdivision conditions. Its acquisition method is the same as step 32 and will not be repeated here.

[0124] The subdivided temperature range is a temperature range calculated based on the global average temperature and subdivided fluctuation parameters. It includes the upper and lower limits of the subdivided temperature range. The specific calculation method is the same as in step 32, and will not be repeated here.

[0125] Step 37: Define the subdivision interval whose average temperature does not fall within the subdivision interval temperature range as the abnormal temperature zone and output it.

[0126] If the average temperature of a subdivided interval does not fall within the temperature range, it indicates that the temperature in that area deviates significantly from the normal range, suggesting possible defects such as pores or abnormal fiber structure. These defects are marked by the algorithm and output for subsequent analysis.

[0127] Reference Figure 3 Specific methods for analyzing defects and their corresponding defect categories based on abnormal and average temperatures include:

[0128] Step 40: Traverse all abnormal temperature zones to form defect interval groups. Within any two defect intervals in a defect interval group, there is at least one path that connects them through adjacent defect intervals.

[0129] A defect interval group is a connected set consisting of all anomalous temperature regions, satisfying the condition that any two anomalous temperature regions within the group are connected through adjacent anomalous temperature regions. This ensures that all anomalous temperature regions within the group form a single connected region through adjacency relationships, such as vertical, horizontal, or diagonal adjacency, without isolated partitions. Simultaneously, the defect interval group includes all local anomalous temperature regions belonging to the same macroscopic defect, avoiding defect fragmentation caused by mesh segmentation.

[0130] Step 41: Use the top-view infrared thermal image to refine the defect area group to obtain the complete defect area.

[0131] The complete defect region is a continuous closed region obtained after morphological modification of the defect interval group. Its boundary is determined by the temperature gradient or curve fitting algorithm in the thermal image, rather than relying solely on the original mesh division. By identifying the temperature abrupt boundary between the defect area and the normal area, points in adjacent meshes where the temperature difference exceeds a preset threshold are connected to form continuous isotherms. Finally, the isotherms are closed to form a complete region that matches the actual defect shape, and its boundary is composed of critical points with significant internal and external temperature differences.

[0132] Step 42: Divide the temperature distribution according to the intact area of ​​the defect to obtain the average temperature of the defect area.

[0133] The average temperature of the defective area refers to the arithmetic mean of the temperature values ​​of all pixels within the defective area after dividing the temperature distribution data in the infrared thermal image into complete defective areas.

[0134] Step 43: Expand the defective area outwards by a preset heat diffusion width to surround the defective area.

[0135] The thermal diffusion width is a preset fixed distance parameter used to expand outward from the boundary of the intact defect area, forming a region surrounding the defect. Its value is set by the staff based on the material's thermal conductivity characteristics, heating process parameters, and the accuracy of the testing equipment, and is used to infer the shape and location of the region surrounding the defect based on the intact defect area.

[0136] The area surrounding the defect is a ring-shaped region formed by the uniform outward expansion of the intact defect area, extending across its thermal diffusion width. It includes the transition zone around the defect affected by heat conduction. The area surrounding the defect does not include the intact defect area.

[0137] Step 44: Divide the temperature distribution according to the area surrounding the defect to obtain the average temperature of the surrounding area.

[0138] The average temperature of the surrounding area is the result of calculating the arithmetic mean of the temperature values ​​of all pixels within the defect area. This temperature value is used to help determine the type of defect, such as blockages, holes, or watermarks.

[0139] Step 440: When the average temperature of the surrounding area exceeds the temperature range of the subdivision interval and the average temperature of the defect area is lower than the temperature range of the subdivision interval, the defect is defined as a blocked defect and a blocked defect signal is output.

[0140] Blockage defects refer to the thermal conduction obstruction caused by fiber agglomeration, impurity accumulation, or excessively high local density in spunlace nonwoven fabrics. A blockage defect signal is an alarm message output by the system after detecting a blockage defect. It includes a defect type identifier, location information, and associated data, such as the average temperature of the defect area, the average temperature of the surrounding area, and a comparison of temperature ranges within the subdivided areas.

[0141] When the average temperature of the surrounding area exceeds the temperature range of the subdivision zone while the average temperature of the defective area is lower than the temperature range of the subdivision zone, it indicates that the surrounding area is experiencing heat accumulation, while the defective area is in a state of impeded heat conduction. The surrounding area is high because the blockage area obstructs the normal heat conduction path, forcing heat to accumulate in a roundabout manner in the surrounding area. The defective area, on the other hand, has low heat conduction efficiency due to dense fibers or the presence of foreign objects, so its surface temperature rises slowly when heated.

[0142] Step 441: When the average temperature of the surrounding area falls within the subdivision interval temperature range and the average temperature of the defect area is lower than the subdivision interval temperature range, define the defect as a hole defect and output a hole defect signal.

[0143] A hole defect refers to a penetrating defect in spunlace nonwoven fabric caused by fiber breakage, structural defects, or excessively low density in certain areas. A hole defect signal is an alarm message output by the system after detecting a hole defect, including defect type identification, location information, and associated data.

[0144] When the average temperature of the surrounding area falls within the subdivided temperature range while the average temperature of the defective area is lower than the subdivided temperature range, it indicates that the material has localized abnormal heat capacity due to fiber breakage or voids. The system defines this as a hole defect. The high temperature in the defective area is because the hole area lacks structure, preventing heat retention and causing the surface temperature to not rise. Meanwhile, the area surrounding the hole has an intact structure, and the heat conduction path is not significantly disturbed, hence the normal temperature in the surrounding area.

[0145] Step 442: When the average temperature of the surrounding area is lower than the temperature range of the subdivision interval and the average temperature of the defect area is lower than the temperature range of the subdivision interval, the defect is defined as a watermark defect and a watermark defect signal is output.

[0146] Watermark defects refer to localized abnormalities in the thermal properties of spunlace nonwoven fabrics caused by residual moisture, such as in humid environments, incomplete drying after cleaning, or water penetration during the production process. A watermark defect signal is an alarm message output by the system after detecting a watermark defect, including defect type identification, location information, and associated data.

[0147] When the average temperature of the surrounding area is detected to be lower than the temperature range of the subdivided interval, and the average temperature of the defect area is also lower than the temperature range of the subdivided interval, it indicates that the material has localized changes in thermal properties due to residual moisture, which the system defines as a watermark defect. The watermark area, due to its high water content, has a higher specific heat capacity than anhydrous materials, resulting in a lower temperature rise when absorbing the same amount of heat. Meanwhile, the surrounding area, because its heat is carried away by the area surrounding the defect, has a temperature lower than the temperature range of the subdivided interval.

[0148] It also includes a method for not outputting a blockage defect signal when the average temperature of the surrounding area exceeds the temperature range of the subdivision interval and the average temperature of the defect area is lower than the temperature range of the subdivision interval, the method comprising:

[0149] Step 4400: Calculate the temperature difference of the blocked area by using the average temperature of the defective area and the average temperature of the surrounding area.

[0150] The temperature difference in the blocked area is the difference between the average temperature of the surrounding area and the average temperature of the defective area. It can be obtained by subtracting the average temperature of the defective area from the average temperature of the surrounding area. The temperature difference in the blocked area quantifies the localized heat accumulation effect caused by the blockage defect; a larger difference indicates a more severe obstruction of heat conduction due to the blockage.

[0151] Step 4401: Obtain the blockage degree value from the preset temperature blockage degree table based on the temperature difference of the blockage area.

[0152] The temperature clogging level table is a pre-defined mapping table used to convert the temperature difference of clogging areas into discrete clogging level values. The table was obtained through extensive experimentation by staff. By analyzing the temperature differences of clogging areas in a large number of spunlace fabrics with varying degrees of clogging, different clogging levels were correlated with the temperature differences of clogging areas, and the correspondence was ultimately stored in the table to obtain the temperature clogging level table.

[0153] The blockage severity value is a discrete numerical value obtained from the temperature blockage severity table, representing the severity level of the defect corresponding to the temperature difference in the current blockage area. The blockage severity value can be directly obtained from the temperature blockage severity table based on the temperature difference in the blockage area.

[0154] Step 4402: Obtain the defect area based on the complete defect area.

[0155] The defect area refers to the area of ​​the complete defect region obtained using infrared thermal imaging, and the unit is square millimeter or square centimeter.

[0156] Step 4403: Obtain the degree of local blockage based on the defect area and the degree of blockage value.

[0157] The degree of local blockage is a comprehensive indicator used to quantify the impact of defects on the material's functionality. Its value is calculated by weighting the defect area and the degree of blockage using a preset formula.

[0158] ;

[0159] Where S 瑕疵 Area of ​​defect; V 堵塞 : Clogging degree value; K: Correction coefficient. The correction coefficient is determined by researchers through experiments combining different defect areas and clogging degrees, simulating defect distribution in actual production, testing changes in key material properties such as permeability and strength, and analyzing the synergistic effect of both on functionality. After numerous experiments, researchers obtain the optimal parameter that correlates defect area, clogging degree, and local clogging degree, and use this parameter as the correction coefficient.

[0160] Step 44030: When the local blockage level is higher than the preset blockage threshold, the defective complete area corresponding to the blockage level value is defined as the completely blocked area, and a blockage defect signal is output.

[0161] A completely blocked area refers to a region where the blockage has significantly affected the material's functions, such as breathability and strength.

[0162] The clogging threshold is a value set by staff to measure the degree of clogging. When the degree of clogging in a local area exceeds the preset clogging threshold, it indicates that the combined impact of the defect, its area and the degree of clogging have exceeded the allowable range of the process, which may lead to the functional failure of the material. Therefore, a clogging signal is required.

[0163] Step 44031: When the local blockage level is lower than the blockage threshold, the defective complete area corresponding to the blockage level value is defined as a partial blockage area, and no blockage defect signal is output.

[0164] Partial blockage refers to areas where blockage is within an acceptable range. When the degree of local blockage is below a preset blockage threshold, it indicates that the defect is within a controllable range, has a minimal impact on material properties, and can be allowed to continue production or be reserved for subsequent sampling inspection.

[0165] Also includes:

[0166] Step 4404: Obtain the total number of all partial congestion areas, defined as the number of partial congestion areas.

[0167] The number of partially blocked areas refers to the total number of defective areas that are identified as "partially blocked areas" during the inspection process.

[0168] Step 4405: Obtain the congestion level value corresponding to all partial congestion areas, and define it as the partial congestion level value.

[0169] The partial blockage degree value is the set of blockage degree values ​​corresponding to all partial blockage areas. The blockage degree value of each partial blockage area is obtained from the temperature blockage degree table and is used to indicate the obstruction of heat conduction in that area due to blockage. The specific operation has been described in step 4401 and will not be repeated here.

[0170] Step 4406: Calculate the air permeability value of the spunlace nonwoven fabric based on the partial blockage degree value and the number of partial blockage areas.

[0171] The air permeability value of spunlace nonwoven fabric is a comprehensive index calculated by a preset algorithm, combining the number of partially blocked areas and the degree of partial blockage. It is used to quantify the overall air permeability performance of the material. Specifically, the air permeability value of spunlace nonwoven fabric is obtained by subtracting the total value of all localized blockages on the spunlace nonwoven fabric from the baseline air permeability value. The baseline air permeability value is set by the staff.

[0172] Step 4407: If the air permeability value of the spunlace nonwoven fabric is lower than the preset air permeability threshold, an overall air permeability defect signal is generated and output.

[0173] The air permeability threshold is a preset critical value used to determine whether the air permeability of spunlace nonwoven fabric is up to standard. The air permeability threshold is obtained by staff through extensive testing. By simulating different degrees of clogging defects, the lowest value that can represent the air permeability value is finally obtained and used as the air permeability threshold.

[0174] When the air permeability value of spunlace nonwoven fabric is lower than the preset air permeability threshold, it indicates that the cumulative effect of some blockage areas is significant. Unlike traditional local defects, an air permeability value higher than the threshold indicates that the defect has developed from an isolated problem into a systemic risk affecting the overall performance, requiring a comprehensive investigation of the production batch or adjustment of process parameters.

[0175] The overall air permeability defect signal refers to the overall defect alarm signal output by the system when the air permeability value of the spunlace nonwoven fabric is lower than the air permeability threshold.

[0176] A method for not outputting a hole defect signal when the average temperature of the surrounding area falls within the subdivision temperature range and the average temperature of the defect area is lower than the subdivision temperature range, the method comprising:

[0177] Step 4410: Obtain the perimeter of the complete defect area based on the top-view infrared thermal image and define it as the defect perimeter.

[0178] The perimeter of a defect refers to the length of the closed curve of the boundary of the intact defect region. The perimeter of a defect can be obtained by using a computer contour extraction algorithm to obtain the contour size of the intact defect region, and thus the perimeter of the intact defect region.

[0179] Step 4411: Obtain the pore roundness of the complete area of ​​the defect based on the defect area and defect perimeter.

[0180] Pore ​​roundness is a dimensionless parameter used to quantify how closely the shape of a defective area approximates an ideal circle. Its calculation formula is:

[0181] ;

[0182] The value of pore roundness ranges from 0 to 1. The larger the pore roundness value, the closer the shape is to a circle.

[0183] Step 4412: When the roundness of the pore is greater than the preset roundness threshold, the corresponding defective complete area is defined as a normal pore area, and the hole defect signal may not be output.

[0184] The normal porosity region refers to an area where there are normal pores rather than holes or defects. The roundness threshold is set by the operator and represents the value at which the intact area of ​​the defect can be considered a uniformly formed hole in the process.

[0185] When the roundness of the pore is higher than the preset threshold, it means that the defective area is close to a circle, which is consistent with the basic characteristics of the pore formed by the process. Therefore, the system judges it as a normal pore and does not output a hole defect signal to avoid false alarms.

[0186] Also includes:

[0187] Step 4413: Obtain the pore core coordinates of all normal pore regions. The pore core coordinates refer to the coordinates of the geometric center of the normal pore region.

[0188] The pore core coordinates are the geometric center coordinates of a normal pore region, usually represented by two-dimensional planar coordinates (x, y). For each normal pore region, its smallest bounding rectangle is extracted using image processing algorithms, and the coordinates of the rectangle's center are calculated as the pore core coordinates. Alternatively, the core coordinates can be obtained through centroid calculation methods from binary images.

[0189] Step 4414: Arrange all the pore core coordinates to form a pore core coordinate group.

[0190] A pore core coordinate set is a collection of coordinates of all normal pore cores, typically stored as a list or array. All pore core coordinates are arranged column-majorly, from left to right by x-coordinate, and within the same column, sorted from top to bottom by y-coordinate.

[0191] Step 4415: Calculate the lateral distance and longitudinal distance between any two adjacent pore core coordinates in the pore core coordinate group, and define them as the pore lateral coordinate spacing and the pore longitudinal coordinate spacing.

[0192] The pore horizontal coordinate spacing refers to the difference in the x-axis coordinates of two adjacent pore core coordinates in the coordinate system. The pore vertical coordinate spacing refers to the difference in the y-axis coordinates of two adjacent pore core coordinates in the coordinate system.

[0193] Step 4416: Obtain the standard pore transverse spacing based on the statistical analysis of all pore transverse coordinate spacings.

[0194] The standard lateral pore spacing is the mode of the longitudinal distances between the coordinates of all adjacent pore cores, used to describe the ideal uniform distribution of pores in the vertical direction. The standard lateral pore spacing is determined by analyzing the lateral distances between the coordinates of all adjacent pore cores, and using the values ​​of points with a clear central tendency in the statistical distribution as the standard lateral pore spacing.

[0195] Step 4417: Obtain the standard longitudinal spacing of pores based on the statistical analysis of all pore longitudinal coordinate spacings.

[0196] The standard longitudinal spacing of pores is the mode of the longitudinal distances between the coordinates of all adjacent pore cores, used to describe the ideal uniform distribution spacing of pores in the vertical direction. The method for obtaining it is the same as that for obtaining the standard transverse spacing of pores in step 4416, and will not be repeated here.

[0197] Step 4418: Construct a pore arrangement grid using the coordinates of any pore core, the standard pore lateral spacing, and the standard pore longitudinal spacing.

[0198] The pore arrangement grid is a regular two-dimensional grid constructed based on standard transverse spacing, standard longitudinal spacing, and coordinates of any pore core, used to simulate the ideally uniformly distributed pore locations.

[0199] First, select any pore core coordinate as the grid origin, usually the first or center pore. Expand the grid horizontally and vertically according to the standard pore horizontal and vertical spacing to generate regular grid points. During the expansion process, the expansion range is controlled according to the size of the thermal image to ensure that the grid covers all pore core coordinates.

[0200] Step 4419: Obtain the alignment matching degree based on the pore core coordinates and the pore arrangement grid.

[0201] Alignment matching degree is a quantitative indicator of the deviation between the actual pore core coordinates and the ideal grid points, used to assess the regularity of pore distribution. It is calculated by matching each actual pore core coordinate with the nearest grid point and then calculating the offset, for example, by calculating the Euclidean distance. All the obtained Euclidean distances are then substituted into the root mean square error to calculate the numerical value, which is the alignment matching degree.

[0202] Step 4420: When the alignment matching degree is lower than the preset matching degree threshold, the corresponding normal pore area is defined as an abnormal pore area, the defect is defined as an abnormal pore defect, and the preset abnormal pore defect signal is output.

[0203] Abnormal porosity defects refer to the phenomenon where the regularity of the distribution of normal pore regions in a material is disrupted. When there is a significant deviation between the actual arrangement of pore core coordinates and a regular grid constructed based on standard spacing, it indicates that the pore distribution in that region does not conform to the expected uniformity or regularity, and is thus defined as an abnormal porosity defect.

[0204] An abnormal hole defect signal is a warning or identification information output by the system after detecting an abnormal hole defect, including defect type identification, location information and associated data.

[0205] If the alignment matching degree is lower than the preset threshold, it indicates that the actual distribution of pore core coordinates deviates from the ideal regular grid beyond an acceptable range, reflecting the disorder or non-uniformity of the pore arrangement. This suggests that the regular pores were not intentionally designed, and therefore an abnormal pore defect signal needs to be issued.

[0206] This includes a method that does not output abnormal temperature regions, the method comprising:

[0207] Step 370: Obtain edge correction temperature parameters based on the average temperature of the subdivided intervals.

[0208] Edge correction temperature parameters are temperature correction parameters calculated based on the differences in thermal conductivity characteristics of fabric edge areas. They are used to correct temperature measurement deviations caused by edge heat dissipation effects. These edge correction temperature parameters are obtained from a pre-set correction temperature parameter table. This table was developed by staff through extensive experimentation. Staff correlated the average temperature of the fabric with the edge correction temperature parameters based on the thermal conductivity characteristics of the fabric edge areas and saved the corresponding relationship in the correction temperature parameter table.

[0209] Step 371: Obtain the temperature range of the edge subdivision region based on the edge correction temperature parameters and the subdivision temperature range.

[0210] The edge subdivision temperature range refers to a reasonable temperature range after overlaying edge correction on top of the subdivision temperature range. The edge subdivision temperature range includes the upper and lower limits of the edge subdivision temperature range. The specific calculation method is the same as in step 32, and will not be repeated here.

[0211] Step 372: When the defect interval coordinates fall into the preset edge coordinate set, the average temperature of the corresponding subdivided area is defined as the edge average temperature.

[0212] Edge average temperature refers to the average temperature value of the defect area recalculated when the defect coordinates fall into the edge area. The edge coordinate set is a specific area located at the physical edge of the fabric, preset by the staff according to the actual size of the fabric. Its width is 5% of the total width of the fabric, and it is used to identify abnormal temperature areas that may be affected by the edge thermal conductivity characteristics.

[0213] When the coordinates of the defect area fall into the set of edge coordinates, it indicates that the detected abnormal temperature area is located at the physical edge of the fabric. Therefore, in order to prevent misjudgment caused by the difference in thermal conductivity of the fabric edge area, it is necessary to analyze the temperature of the corresponding area and then make a further judgment.

[0214] Step 3720: If the average edge temperature falls within the temperature range of the edge subdivision region, then no defect signal needs to be output.

[0215] If the average edge temperature falls within the temperature range of the edge subdivision area, it indicates that although the abnormal temperature area is located at the physical edge of the fabric, its temperature deviation is within the normal fluctuation range caused by the edge's thermal conduction characteristics, such as the edge heat dissipation effect. In this case, it is determined that there is no substantial defect in this area, and there is no need to output a defect signal.

[0216] Step 3721: If the average edge temperature does not fall within the temperature range of the edge subdivision region, continue to output the defect signal.

[0217] If the average temperature at the edge does not fall within the temperature range of the subdivided edge region, it indicates that the deviation of the abnormal temperature zone exceeds the normal range allowed by the edge's thermal conductivity, suggesting the presence of a substantial defect, such as a hole, blockage, or contaminant. In this case, a defect signal should be output to indicate that the area is a genuine defect.

[0218] Also includes:

[0219] Step 50: Obtain the total number of all coarse-division anomalous temperature ranges, which is defined as the number of coarse-division anomalous ranges.

[0220] The number of coarsely segmented anomalous regions refers to the total number of anomalous temperature regions obtained through coarse segmentation in an infrared thermal image. Specifically, it refers to the number of independent regions where the temperature distribution exceeds the normal range, such as being higher or lower than the subdivided temperature range.

[0221] Step 51: When the number of coarsely divided abnormal intervals exceeds the preset abnormal interval threshold, the defect is defined as an overall quality defect, and the preset overall quality defect signal is output.

[0222] Overall quality defects refer to comprehensive quality problems in materials or products caused by widespread defects, rather than localized, single defects. An overall quality defect signal is an alarm message output by the system after detecting an overall quality defect, including defect type identification, location information, and associated data. Workers collect temperature distribution data of spunlace nonwoven fabric under normal production conditions to obtain the minimum threshold that indicates the absence of overall defects in the spunlace nonwoven fabric, which is then used as the abnormal interval threshold.

[0223] The abnormal interval threshold is a critical value used to determine whether the number of coarsely divided abnormal intervals exceeds the normal fluctuation range. Its function is to distinguish between accidental local defects and systemic overall quality defects.

[0224] When the number of coarse-grained abnormal zones exceeds a preset threshold, it indicates that the distribution density of abnormal temperature zones in the material or product has exceeded the acceptable range, suggesting that unstable production processes or equipment parameter deviations have led to widespread defects. Therefore, an overall quality defect signal is required.

[0225] Step 52: Perform standard deviation analysis on the temperature of the top-view infrared thermal image between the coarse divisions to calculate the corresponding standard deviation of the coarse division regions.

[0226] Standard deviation analysis is a statistical method used to measure the dispersion of a dataset. In infrared thermal imaging, the uniformity of the temperature field is assessed by calculating the standard deviation of the temperature distribution over a coarsely divided region. A larger standard deviation indicates a more uneven temperature distribution, which may indicate abnormal heat conduction or structural defects.

[0227] The coarse segmentation standard deviation refers to the standard deviation calculated after statistically analyzing the temperature of the infrared thermal image corresponding to the coarse segmentation anomaly interval. The coarse segmentation standard deviation is obtained by extracting the temperature values ​​of all pixels within the coarse segmentation region, calculating the average of these temperature values, and then calculating the standard deviation based on the difference between each temperature value and the average value.

[0228] Step 53: When the standard deviation of the coarse segmentation area exceeds the preset abnormal standard deviation threshold, the defect is also defined as an overall quality defect, and an overall quality defect signal is output.

[0229] When the standard deviation of the coarse division area exceeds the preset abnormal standard deviation threshold, it indicates that the temperature distribution on the spunlace nonwoven fabric is extremely uneven.

[0230] Methods for outputting overall quality defect signals also include:

[0231] Step 54: Obtain the number of standard abnormal regions and the standard deviation threshold of standard samples based on the preset standard top-down infrared spectrum.

[0232] A standard top-view infrared image refers to an infrared image of a flawless product used as a benchmark, containing a normal temperature distribution.

[0233] The standard number of abnormal areas refers to the maximum allowed number of normal abnormal areas in a preset standard top-view infrared spectrum. Staff analyze historical infrared spectra of a large number of flawless products to count the number of abnormal areas caused by permissible deviations, and then set this as the standard number of abnormal areas.

[0234] The standard deviation threshold for standard samples refers to the critical value of the standard deviation of temperature in the coarse separation abnormal area; exceeding this value is considered an overall quality abnormality. Workers calculate the standard deviation of characteristic values ​​in the normal area of ​​the standard infrared spectrum and, combined with the allowable fluctuation range of the process, set an acceptable minimum threshold, which is then set as the standard sample standard deviation threshold.

[0235] Step 55: If the number of coarse-divided abnormal regions is greater than the number of standard abnormal regions or the standard deviation of the coarse-divided regions is greater than the standard deviation threshold of the standard sample, the defect is defined as an overall quality defect, and an overall quality defect signal is output.

[0236] When the number of coarsely identified abnormal regions is greater than the number of standard abnormal regions, it indicates that the number of actually detected abnormal regions exceeds the normal fluctuation range, suggesting that the defects may be widespread rather than isolated points, and therefore an overall quality defect signal needs to be issued.

[0237] When the standard deviation of the coarse-divided region is greater than the standard deviation threshold of the standard sample, it indicates that the temperature distribution dispersion of the abnormal region is too high, exceeding the normal random noise range, indicating that there may be large-area defects in the whole, so it is necessary to issue an overall quality defect signal.

[0238] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for visual analysis of defects in spunlace nonwoven fabric, characterized in that, include: Step 1: In response to the fabric arrival signal, acquire a top-view infrared thermal image. The fabric arrival signal is the signal that the spunlace nonwoven fabric to be inspected is placed on a preset defect detection device. The defect detection device includes a conveyor wheel for placing the spunlace nonwoven fabric and a nozzle for spraying hot air. There is a space between the upper and lower conveyor wheels for the spunlace nonwoven fabric to pass through. The nozzle is located on the lower side of the spunlace nonwoven fabric and faces the spunlace nonwoven fabric, and sprays hot air evenly onto the spunlace nonwoven fabric. The defect detection device also includes an infrared detector located above the spunlace nonwoven fabric. Step 2: Analyze the temperature distribution and average temperature of the spunlace nonwoven fabric using the top-view infrared thermal imaging image; Step 3: Compare the temperature distribution with the average temperature to obtain the abnormal temperature zone and the corresponding abnormal temperature; Step 4: Analyze the defects and their corresponding defect categories based on the abnormal temperature and the average temperature; Step 5: Output the defect signal by generating a defect signal from the abnormal temperature zone and defect category; The specific method for obtaining the abnormal temperature zone by comparing the temperature distribution with the average temperature includes: Step 30: Divide the top-view infrared thermal image into multiple intervals according to the preset coarse grid segmentation method, and define the divided areas as coarse intervals; Step 31: Divide the temperature distribution according to the coarse interval to obtain the average temperature of the coarse interval; Step 32: Calculate the coarse-division interval temperature range based on the average temperature and the preset coarse-division fluctuation parameters; Step 33: Define the coarse interval whose average temperature does not fall within the coarse interval temperature range as the coarse interval abnormal temperature range; Step 34: Divide the coarse-divided abnormal temperature range into multiple ranges according to the preset fine-grid segmentation method, and define the divided ranges as subdivided ranges; Step 35: Divide the temperature distribution according to the subdivided intervals to obtain the average temperature of the subdivided intervals; Step 36: Obtain the subdivided temperature range based on the average temperature and the preset subdivided fluctuation parameters; Step 37: Define the subdivided interval whose average temperature does not fall within the subdivided interval temperature range as the abnormal temperature zone and output it; The specific method for analyzing the defect and its corresponding defect category based on the abnormal temperature and the average temperature includes: Step 40: Traverse all the abnormal temperature zones to form defect interval groups. Within each defect interval group, there is at least one path connecting any two defect intervals through adjacent defect intervals. Step 41: The defect area group is modified using the top-view infrared thermal image to obtain the complete defect area; Step 42: Divide the temperature distribution according to the intact area of ​​the defect to obtain the average temperature of the defect area; Step 43: Expand the defective area outward by a predetermined heat diffusion width to obtain a surrounding defect area; Step 44: Divide the surrounding defect area according to the temperature distribution to obtain the average temperature of the surrounding area; Step 440: When the average temperature of the surrounding area exceeds the temperature range of the subdivision interval and the average temperature of the defect area is lower than the temperature range of the subdivision interval, the defect is defined as a blockage defect and a blockage defect signal is output. Step 441: When the average temperature of the surrounding area falls within the temperature range of the subdivided interval and the average temperature of the defective area is lower than the temperature range of the subdivided interval, the defect is defined as a hole defect and a hole defect signal is output. Step 442: When the average temperature of the surrounding area is lower than the temperature range of the subdivision interval and the average temperature of the defect area is lower than the temperature range of the subdivision interval, the defect is defined as a watermark defect and a watermark defect signal is output.

2. The method for visual analysis of defects in spunlace nonwoven fabric according to claim 1, characterized in that, It also includes a method for not outputting a blockage defect signal when the average temperature of the surrounding area exceeds the subdivision interval temperature range and the average temperature of the defect area is lower than the subdivision interval temperature range, the method comprising: Step 4400: Calculate the temperature difference of the blocked area by using the average temperature of the defective area and the average temperature of the surrounding area; Step 4401: Obtain the blockage degree value from the preset temperature blockage degree table based on the temperature difference of the blockage area; Step 4402: Obtain the defect area based on the complete defect area; Step 4403: Obtain the local blockage degree based on the defect area and the blockage degree value. The local blockage degree is calculated by weighting the defect area and the blockage degree value using a preset formula. ； Where S 瑕疵 Area of ​​defect; V 堵塞 : Degree of congestion; K: Correction factor; Step 44030: When the local blockage degree is higher than the preset blockage threshold, the defective complete area corresponding to the blockage degree value is defined as the completely blocked area, and a blockage defect signal is output. Step 44031: When the local blockage level is lower than the blockage threshold, the defective complete area corresponding to the blockage level value is defined as a partially blocked area, and no blockage defect signal is output.

3. The method for visual analysis of defects in spunlace nonwoven fabric according to claim 2, characterized in that, Also includes: Step 4404: Obtain the total number of all the partially blocked areas, defined as the number of partially blocked areas; Step 4405: Obtain the blockage degree value corresponding to all the partially blocked areas, and define it as the partial blockage degree value; Step 4406: Calculate the air permeability value of the spunlace nonwoven fabric based on the partial blockage degree value and the number of partial blockage areas. The air permeability value of the spunlace nonwoven fabric is obtained by subtracting the values ​​of all local blockage degrees on the spunlace nonwoven fabric from the preset benchmark air permeability value. Step 4407: If the air permeability value of the spunlace nonwoven fabric is lower than the preset air permeability threshold, an overall air permeability defect signal is generated and output.

4. The method for visual analysis of defects in spunlace nonwoven fabric according to claim 1, characterized in that, A method for not outputting a hole defect signal when the average temperature of the surrounding area falls within the subdivided temperature range and the average temperature of the defective area is lower than the subdivided temperature range, the method comprising: Step 4410: Obtain the perimeter of the complete area of ​​the defect based on the top-view infrared thermal image, and define it as the defect perimeter; Step 4411: Obtain the porosity of the intact area of ​​the defect based on the defect area and the defect perimeter. The formula for calculating the porosity is: ； The pore roundness value is between 0 and 1; Step 4412: When the roundness of the pore is greater than the preset roundness threshold, the corresponding defective complete area is defined as a normal pore area, and the hole defect signal may not be output.

5. The method for visual analysis of defects in spunlace nonwoven fabric according to claim 4, characterized in that, Also includes: Step 4413: Obtain the pore core coordinates of all the normal pore regions, where the pore core coordinates refer to the coordinates of the geometric center of the normal pore region; Step 4414: Arrange all the pore core coordinates to form a pore core coordinate group; Step 4415: Calculate the lateral distance and longitudinal distance between any two adjacent pore core coordinates in the pore core coordinate group, and define them as the pore lateral coordinate spacing and the pore longitudinal coordinate spacing. Step 4416: Obtain the standard pore transverse spacing based on the statistical analysis of all the pore transverse coordinate spacings; Step 4417: Obtain the standard pore longitudinal spacing based on the statistical analysis of all the pore longitudinal coordinate spacings; Step 4418: Construct a pore arrangement grid using the coordinates of any of the pore cores, the standard pore lateral spacing, and the standard pore longitudinal spacing; Step 4419: Obtain the alignment matching degree based on the coordinates of the pore core and the pore arrangement grid; Step 4420: When the alignment matching degree is lower than the preset matching degree threshold, the corresponding normal pore area is defined as an abnormal pore area, the defect is defined as an abnormal pore defect, and a preset abnormal pore defect signal is output.

6. The method for visual analysis of defects in spunlace nonwoven fabric according to claim 1, characterized in that, This includes a method that does not output the abnormal temperature region, the method comprising: Step 370: Obtain edge correction temperature parameters based on the average temperature of the subdivided intervals; Step 371: Obtain the edge subdivision region temperature range based on the edge correction temperature parameters and the subdivision interval temperature range; Step 372: When the defect interval coordinates fall into the preset edge coordinate set, the average temperature of the corresponding subdivided area is defined as the edge average temperature; Step 3720: If the average edge temperature falls within the temperature range of the edge subdivision region, the defect signal may not be output. Step 3721: If the average edge temperature does not fall within the temperature range of the edge subdivision region, then continue to output the defect signal.

7. The method for visual analysis of defects in spunlace nonwoven fabric according to claim 1, characterized in that, The method for outputting the defect signal further includes: Step 50: Obtain the total number of all the coarse-division abnormal temperature intervals, defined as the number of coarse-division abnormal intervals; Step 51: When the number of coarse-divided abnormal intervals exceeds the preset abnormal interval threshold, the defect is defined as an overall quality defect, and a preset overall quality defect signal is output. Step 52: Perform standard deviation analysis on the temperature of the top-view infrared thermal image between the coarse divisions, and calculate the corresponding standard deviation of the coarse division regions; Step 53: When the standard deviation of the coarse segmentation region exceeds the preset abnormal standard deviation threshold, the defect is also defined as the overall quality defect, and the overall quality defect signal is output.

8. The method for visual analysis of defects in spunlace nonwoven fabric according to claim 7, characterized in that, The method for outputting the overall quality defect signal further includes: Step 54: Obtain the number of standard abnormal regions and the standard deviation threshold of standard samples based on the preset standard top-down infrared spectrum; Step 55: If the number of coarse-separated abnormal regions is greater than the number of standard abnormal regions or the standard deviation of the coarse-separated regions is greater than the standard deviation threshold of the standard sample, the defect is defined as the overall quality defect, and the overall quality defect signal is output.

Images