Intelligent detection method for sanitary napkin production defects based on multi-modal image analysis

Abstract

The present application belongs to the field of visual detection, and proposes a sanitary napkin production defect intelligent detection method based on multi-modal image analysis, which comprises the following steps: laying the sanitary napkin on a transparent plane and irradiating the sanitary napkin with light to obtain a light transmission image of the irradiated sanitary napkin; carrying out gray scale processing on the light transmission image of the sanitary napkin to obtain a sanitary napkin gray scale image; obtaining a light transmission display grid according to the sanitary napkin gray scale image, and obtaining a real light transmission area through the light transmission display grid; dividing the sanitary napkin gray scale image into two parts to obtain a left gray scale image and a right gray scale image, comparing the real light transmission areas of the left gray scale image and the right gray scale image, and obtaining a side leakage area; dividing the gray scale image into two parts, comparing the left gray scale image and the right gray scale image, and obtaining a thin defect area. According to the intelligent detection method of the present application, the presence of side leakage defects in the structure of the sanitary napkin can be detected through visual detection of the light transmission information of the sanitary napkin.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of visual inspection, and particularly relates to a sanitary napkin production defect intelligent detection method based on multi-modal image analysis. BACKGROUND

[0002] At present, in the production and manufacturing process of sanitary napkins, in order to guarantee product quality, the industry usually relies on image recognition technology, artificial visual inspection, or traditional machine vision algorithms based on color, shape contour and other features to perform defect detection and quality control; these methods mainly focus on surface appearance defects of the product, such as pattern printing deviation, indentation asymmetry, edge cutting error, surface structure damage and other explicit defects; these defects are easily identified by camera shooting due to their strong visual distinguishability, and are suitable for online detection systems on high-speed assembly lines.

[0003] However, the structure of sanitary napkins is complex, usually composed of multiple layers of materials such as surface non-woven fabric, absorbent layer, flow guide layer and bottom film, and the functionality is mainly reflected in the compactness of the internal structure and the uniformity of the material distribution. When the absorbent layer has phenomena such as material sparseness, insufficient compression, and loose fiber arrangement in some local areas, even if there is no obvious difference in appearance, these thin areas may become the starting point for liquid to break through the absorption barrier and leak out of the edge during actual use, causing side leakage problems.

[0004] These implicit defects cannot be effectively detected by conventional image vision algorithms because they do not directly manifest as color abnormalities or edge shape changes. Especially in a high-speed production environment, thin products with perfect appearance are easily misjudged as qualified products by traditional systems. SUMMARY

[0005] The present application aims to at least partially solve one of the technical problems in the related art. To this end, the purpose of the present application is to propose a sanitary napkin production defect intelligent detection method based on multi-modal image analysis, which can detect whether the structure of the sanitary napkin has a side leakage defect through the light transmission information of the sanitary napkin through visual inspection, especially those hidden problems that are difficult to find through appearance detection.

[0006] To achieve the above purpose, the present application proposes a sanitary napkin production defect intelligent detection method based on multi-modal image analysis, which comprises the following steps:

[0007] S100, laying the sanitary napkin on a transparent plane and illuminating the sanitary napkin with light to obtain a light transmission image of the illuminated sanitary napkin;

[0008] S200, performing gray-scale processing on the light transmission image of the sanitary napkin to obtain a sanitary napkin gray-scale image;

[0009] S300, obtaining a light transmission display grid according to the sanitary napkin gray scale image, and obtaining a real light transmission area through the light transmission display grid;

[0010] S400, obtaining a left gray scale image and a right gray scale image by dividing the sanitary napkin gray scale image into two parts, comparing the real light transmission areas of the left gray scale image and the right gray scale image, and obtaining a side leakage area. Figure One

[0011] According to the intelligent detection method, whether the sanitary napkin structure has a side leakage defect can be detected through the light transmission information of the sanitary napkin through visual detection.

[0012] Further, in step S100, the sanitary napkin is laid flat on a transparent plane and irradiated with light to obtain a light transmission image of the irradiated sanitary napkin, including:

[0013] The sanitary napkin product to be detected is laid flat on a transparent optical work platform, the platform is made of high-transmittance glass or acrylic plate, which ensures that the bottom light can uniformly penetrate, and the sanitary napkin is ensured to be wrinkle-free and non-overlapping during the laying process, and the long axis direction is parallel to the platform coordinate axis, so that subsequent image partitioning and symmetric processing are facilitated; then the top lighting and bottom transmission light source system is started, a visible light LED light source with a wavelength range of 400-700 nm is used, and the optical response of the sanitary napkin structure is excited by vertical irradiation; wherein the bottom light source penetrates the sanitary napkin and forms a light transmission image on the top imaging device, reflecting the light density difference of the internal absorbing material, fiber distribution and pressing structure of the product; wherein the top imaging device selects a high-resolution industrial camera or a line array camera, collects complete light transmission image data, and saves it as an original image file for subsequent image processing.

[0014] Further, in step S200, the light transmission image of the sanitary napkin is grayed to obtain a sanitary napkin gray scale image, including:

[0015] The obtained light transmission image is input into the image preprocessing software OpenCV for gray scale operation; the original color image is converted into a single-channel gray scale image by using an image processing algorithm, and the brightness component of each pixel is extracted as its gray scale value; wherein the gray scale processing is realized by using a weighted average method, a maximum value method or an OpenCV standard gray scale conversion function, so that the value of each pixel point in the image is in the interval of 0-255, the larger the value, the stronger the light transmission, and the smaller the value, the more dense or opaque the structure; the finally obtained sanitary napkin gray scale image can clearly reflect the structure light transmission characteristics of the product under light excitation, and serve as an input basis for subsequent light transmission grid analysis and defect recognition.

[0016] ​At present, in the production and manufacturing process of sanitary napkins, product quality control mainly relies on image recognition, manual detection or machine vision algorithm based on color and contour. However, these methods can only be used for surface appearance, pattern integrity or edge misplacement of explicit defects, and it is difficult to effectively identify implicit defects caused by uneven distribution of absorbent material or insufficient structure compaction, especially the local thin area located at the edge or inside of the sanitary napkin. These areas are prone to become the starting point of liquid leakage during product use, thereby causing side leakage problem. In order to solve the above problems, the step S300 is provided.

[0017] Further, in step S300, the light transmission revealing grid is obtained according to the gray scale image of the sanitary napkin, which comprises:

[0018] The gray scale image is divided into K grids, wherein K=1800, and the gray scale value of the i-th grid of the gray scale image is represented by sut(i), i is [1, K], and K is the number of grids after the gray scale image is divided. The median of the gray scale values of each grid in sut(i) is obtained, the average of the gray scale values of each grid in sut(i) is obtained, and the maximum of the gray scale values of each grid in sut(i) is obtained and recorded as ura; wherein the gray scale value of the grid is the average of the gray scale values of all pixels in the grid; the balanced light transmission gray scale urp is calculated, and the calculation method of the balanced light transmission gray scale is:

[0019]

[0020] Further, the grid with a gray scale value greater than urp is recorded as a light transmission revealing grid, and the grid with a gray scale value less than urp is recorded as a light transmission grid;

[0021] Further, in actual production, the local thin area of the absorbent layer does not necessarily exhibit color or edge abnormalities, but due to insufficient material density, light is more easily transmitted, showing a higher gray scale value. By using urp, the grid in the gray scale image exceeding urp is determined as a "light transmission revealing grid", that is, an area where the material is sparse. Although these areas may not have obvious defect characteristics in appearance, they are the potential starting point of side leakage during use of the sanitary napkin.

[0022] The beneficial effect of this step is that: since the traditional method is usually based on fixed threshold for image segmentation, it is easy to be disturbed by environmental light change, lighting angle and other factors and fail, while the urp defined in the present application is self-adaptively adjusted according to the actual distribution characteristics of each image, and can maintain consistency in discrimination when the overall brightness of the image changes, and since the value integrates the median, mean and extreme value three statistical characteristics, it not only ensures stable judgment of the overall image, but also can guide sensitive response to local high gray scale (thin) area.

[0023] Further, in step S300, obtaining the real light transmission area through the light transmission grid includes the following steps:

[0024] S301, let zj(j) represent the gray value of the jth light transmission grid, and j is in the range of [1, H], where H is the number of light transmission grids; zg(j) represents the gray value of the light blocking grid closest to the center point of the jth light transmission grid; for each zg(j), obtain the median and maximum value, and the average of the median and maximum value is denoted as zjxp;

[0025] S302, define an integer variable k, and set the initial value to 1, and create two variables sutPTRa and sutPTRb with initial values of zero, which are used for subsequent calculation and comparison;

[0026] S303, create a blank sequence TYR1, denoted as the real light transmission grid sequence, and create a blank sequence TYR2, denoted as the fuzzy light transmission grid sequence;

[0027] S304, calculate the values of sutPTRa and sutPTRb, where the calculation formula of sutPTRa is zg(k)-zj(k), and the calculation formula of sutPTRb is [zg(k)+zjxp] / 2-urp; compare the values of sutPTRa and sutPTRb: if sutPTRa is greater than sutPTRb, then add the grid sequence number where zj(k) is located to the real light transmission grid sequence TYR1; if sutPTRa is less than or equal to sutPTRb, then add the grid sequence number where zg(k) is located to the fuzzy light transmission grid sequence TYR2;

[0028] S305, if the current variable k is less than H, then increase k by 1 and return to step S304; if the current variable k is not less than H, then jump to S306;

[0029] S306, obtain the median puy in the fuzzy light transmission grid sequence TYR2, and add the gray values greater than puy in the fuzzy light transmission grid sequence TYR2 to the real light transmission grid sequence TYR1; the area composed of the grids corresponding to the gray values in the real light transmission grid sequence TYR1 is denoted as the real light transmission area.

[0030] The beneficial effects of this step are as follows: By analyzing grayscale differences, comparing neighboring blocking grids, and using a median compensation mechanism, regions with true physical thinness characteristics are accurately selected from the transparent grids. Specifically, this step establishes a local contrast relationship between transparent and blocking grids, avoiding the bias of simply using the average value of the entire image for judgment. The difference sutPTRa = zg(j) - zj(j) is defined as the degree of light transmission prominence of the transparent grid. Simultaneously, a reference value sutPTRb = ([zg(j) + zjxp] / 2 - urp) is calculated, combining the characteristic point value zjxp of the blocking grayscale and the balanced transparent grayscale urp to reflect the boundary stability of the grayscale distribution of the structure. When sutPTRa > sutPTRb, it indicates a significant difference in grayscale between the transparent grid and its neighboring blocking grids, and it can be determined as a true transparent grid. Its grid number is then added to the TYR1 sequence. Conversely, areas with poor light transmission are marked as fuzzy transparent grids and added to the TYR2 sequence. However, considering that some edge areas or fiber gradient areas may have gray-level perturbations, it is necessary to further filter out fuzzy grids with gray-level values ​​significantly higher than the median gray-level value puy in TYR2 and add them to TYR1 to reduce the risk of missed detection and enhance the model's discrimination tolerance. This real transparent area can be used to construct representative thin transparent candidate areas, providing accurate regional input for subsequent left-right image comparison and side leakage risk assessment. Compared with existing technologies, this method has the following significant advantages: by using a comparison model of "light transmission and adjacent blocking", it avoids excessive false detection and missed detection caused by simple threshold judgment; by introducing dynamic gray-level reference values ​​and compensation strategies, the recognition algorithm can adapt to different batches, lighting changes and material differences, and the identified areas reflect anomalies at the physical structure level, such as insufficient compaction of the absorption layer and thin material thickness, which are key sources of side leakage.

[0031] Because traditional technologies have not yet used light transmittance as a core criterion for judging the structural stability or defects of sanitary napkins, this physical characteristic lacks a modeling and quantification mechanism in existing quality inspection systems. Those skilled in the art mostly use grayscale and texture to identify visible surface defects, rather than establishing a standardized method for judging the structural integrity of products based on light transmittance distribution and symmetry. Therefore, existing testing systems suffer from low accuracy, vague judgment criteria, and difficulty in identifying asymmetric defects in complex, uneven structures. To address these issues, this invention proposes step S400.

[0032] Furthermore, in step S400, the grayscale of the sanitary napkin is... Figure One The process involves dividing the image into two parts, obtaining a left grayscale image and a right grayscale image. The actual light-transmitting areas of the left and right grayscale images are then compared to determine the side leakage area. This process includes the following steps:

[0033] S401, the acquired sanitary towel gray scale is vertically divided with the horizontal width midpoint as the symmetry axis to obtain a left gray scale and a right gray scale, which represent the left and right symmetrical areas of the product structure respectively;

[0034] Based on the symmetry of the sanitary towel, the symmetry axis obtained by vertically dividing the sanitary towel image according to the horizontal width midpoint is used to divide the gray scale into a left gray scale and a right gray scale, which are used to compare the real light transmission grid distribution in the respective areas to identify potential structural asymmetry and side leakage risk. By comparing the real light transmission areas in the left and right gray scales, a side transmission area is obtained: each grid in the left gray scale is checked to determine whether it belongs to a real light transmission area.

[0035] S402, the real light transmission area is matched and compared to obtain a side transmission area;

[0036] Specifically, for the real light transmission area in the left gray scale, it is checked whether the corresponding position in the right gray scale is a real light transmission area; if not, the left side of the area is marked as a side transmission area.

[0037] Further, for the real light transmission area in the right gray scale, it is checked whether the corresponding position in the left gray scale is a real light transmission area; if not, the right side of the area is marked as a side transmission area.

[0038] S403, the areas in the left gray scale and the right gray scale that are at the same position and are both real light transmission grids are extracted and marked as the same transmission area, and the gray scale value sequences of the area in the left gray scale and the right gray scale are extracted to form a left same transmission sequence and a right same transmission sequence;

[0039] Specifically, the grids at the same position in the left gray scale and the right gray scale that belong to the real light transmission area are marked as the same transmission area, the grids of the same transmission area in the left gray scale are marked as the left same transmission grid, and the grids of the same transmission area in the right gray scale are marked as the right same transmission grid;

[0040] zyp(f) represents the gray scale value of the fth left same transmission grid in the left gray scale; zhg(f) represents the gray scale value grid at the same position as the fth left same transmission grid in the right gray scale, f is the serial number, f takes values [1, T], where T is the number of left same transmission grids in the left gray scale; an empty sequence is created, zyp(f) is sequentially imported into the sequence and marked as the left same transmission sequence, an empty sequence is created, zhg(f) is sequentially imported into the sequence and marked as the right same transmission sequence; the average value of the left same transmission sequence is obtained and marked as zym.

[0041] S404, the light transmission shock interval is obtained by the left same transmission sequence, and the light transmission difference limit is calculated;

[0042] Specifically, the upper quartile ZTJ and the lower quartile ZTK of the left homochromatic sequence are obtained; when an element is less than or equal to ZTK, the element is recorded as a low-order light transmission shock range value; when an element is greater than ZTK and less than ZTJ, the element is recorded as a middle-order light transmission shock range value; when an element is greater than or equal to ZTJ, the element is recorded as a high-order light transmission shock range value; the minimum value of the left homochromatic sequence is recorded as SOU, the maximum value of the left homochromatic sequence is recorded as MAX, and the average value of the left homochromatic sequence is recorded as zym; the value range of all low-order light transmission shock ranges, middle-order light transmission shock ranges and high-order light transmission shock ranges is recorded as [SOU, ZTK], (ZTK, ZTJ) and [ZTJ, MAX] respectively; the difference between the average value zym of the left homochromatic sequence and the minimum shock value SOU is recorded as a first shock difference fluctuation Fos1; the difference between the maximum shock point MAX and the average value zym of the left homochromatic sequence is recorded as a second shock difference fluctuation Fos2; the difference between SOU and ZTK is recorded as a lower boundary extension Fxer; the difference between MAX and ZTJ is recorded as an upper boundary extension Rxer;

[0043] The smaller value of Fos1 and Fos2 is used as the numerator, the larger value of Fos1 and Fos2 is used as the denominator, and the ratio of the numerator and the denominator is used as the light transmission difference limit coefficient YTER; when Fxer is less than or equal to Rxer, the light transmission difference limit UYATR is Fxer x YTER; when Fxer is greater than Rxer, the light transmission difference limit UYATR is Rxer x YTER; when the number of light transmission area grids is insufficient, the difference between the average light transmission gray scale urm and urp is used as the light transmission difference limit UYATR by default.

[0044] The principle of the light transmittance difference limit UYATR is as follows: in the present application, the calculation principle of the light transmittance difference limit UYATR is based on statistical modeling of the gray scale distribution of the real light transmittance area in the left gray scale image. The gray scale values in the same light transmittance area on the left side are formed into a gray scale sequence, and the minimum value SOU, the maximum value MAX, the average value zym, the lower quartile ZTK and the upper quartile ZTJ are calculated to construct a three-section distribution model of the gray scale fluctuation: [SOU, ZTK] is defined as the low-order light transmittance oscillation interval, (ZTK, ZTJ) is the middle-order interval, and [ZTJ, MAX] is the high-order interval. The oscillation gap between the average value and the boundary is calculated to obtain the first oscillation fluctuation Fos1=zym-SOU and the second oscillation fluctuation Fos2=MAX-zym. The smaller value and the larger value are used to construct a distribution balance coefficient. At the same time, the lower boundary extension value Fxer=ZTK-SOU and the upper boundary extension value Rxer=MAX-ZTJ are calculated to measure the low and high tolerance capacity of the gray scale distribution. Finally, the light transmittance difference limit UYATR is determined according to the following rules: when Fxer is less than or equal to Rxer, UYATR=Fxer*YTER; otherwise, UYATR=Rxer*YTER. The limit value is used to determine whether the gray scale difference at the same position in the left and right gray scale images exceeds the reasonable structural deviation range. Once the gray scale difference value is greater than UYATR, it is determined that the area has serious light transmittance asymmetry, which may constitute a structural thin defect or a potential side leakage risk. Through the calculation model, the judgment accuracy can be improved by dynamically adapting to different image gray scale distribution characteristics, and the risk of false judgment caused by fixed threshold setting can be avoided.

[0045] S405, obtaining a side leakage grid through the light transmittance difference limit;

[0046] The difference between zyp(f) and zhg(f) is calculated in sequence, and the absolute value of the difference between zyp(f) and zhg(f) is denoted as chaz(f). chaz(f) is compared with the light transmittance difference limit UYATR. If chaz(f) is greater than the light transmittance difference limit UYATR, zyp(f) and zhg(f) are compared in size. The grid corresponding to the gray scale value greater between zyp(f) and zhg(f) is marked as a side leakage grid.

[0047] S406, obtaining a side leakage area through the side leakage grid;

[0048] The number of side leakage grids of the left gray image and the right gray image is obtained, and the number of side leakage grids of the left gray image is recorded as TUEW, and the number of side leakage grids of the right gray image is recorded as YUEW. The area composed of the side leakage grids of the left gray image is recorded as the left side leakage area, and the area composed of the side leakage grids of the right gray image is recorded as the right side leakage area. The number of side leakage grids of the left gray image and the right gray image is compared. If the number of side leakage grids of the right gray image YUEW is greater than 2*TUEW, it indicates that there is a defect on the right side of the sanitary napkin, and the right side leakage area will leak during the use of the sanitary napkin. The right side leakage area is recorded as the side leakage area. If TUEW is greater than 2*YUEW, it indicates that the sanitary napkin has a defect that will leak, and it indicates that there is a defect on the left side of the sanitary napkin. The left side leakage area will leak during the use of the sanitary napkin. The left side leakage area is recorded as the side leakage area.

[0049] The principle of this step is to calculate the difference between the true light transmission grid gray values of the same position in the left gray image and the right gray image in sequence to construct a difference sequence chaz(f), and compare it with the light transmission difference limit UYATR calculated in the foregoing. When the gray difference chaz(f) of a certain position is greater than UYATR, it indicates that the left and right structures at this point are significantly asymmetric in light transmission performance, which is a potential thin area. At this time, it is further judged which gray value is larger in zyp(f) and zhg(f), and the grid corresponding to the larger gray value is marked as a side leakage grid, because the side area structure is weaker and has stronger light transmission, and is more likely to leak liquid. Then the number of side leakage grids in the left and right gray images is counted respectively, and recorded as TUEW and YUEW. According to the comparison result, the structural integrity is judged. If the number of side leakage grids on one side is more than twice that on the other side, it indicates that there is a significant structural imbalance on this side, which is easy to leak during use. The core basis of this determination logic is the natural symmetry of the sanitary napkin structure design. Any significant asymmetric light transmission abnormality usually reflects functional defects such as uneven distribution of absorbent material, poor edge sealing or insufficient compaction.

[0050] The beneficial effect of this step is that by establishing a dynamic recognition mechanism based on the combination of gray difference and limit deviation, this step can realize high sensitivity and high reliability of side leakage defect positioning and judgment, and overcome the technical bottlenecks of unclear recognition of local structural hidden defects and weak detection ability of asymmetric thin defects in existing detection systems.

[0051] The present application is not obvious for the following reasons: the present application introduces the physical quantity of "light transmittance", which is originally used for material optical performance evaluation, into sanitary napkin structure defect identification, constructs a complete set of light transmittance gray scale quantification, grid dynamic discrimination and structure symmetry comparison algorithm system, and realizes high-precision identification of functional implicit defects which are difficult to be identified by traditional visual systems. This method has not been disclosed or inspired by the prior art, which is embodied in the following aspects: the prior art mainly focuses on color, edge or texture and other surface features for defect identification, while the present application directly links the gray value with the material density based on the response characteristics of the product internal structure to the transmitted light, proposes a real light transmittance area identification method, has a unique physical modeling basis, and belongs to a substantial breakthrough in detection mechanism. Traditional structure integrity judgment mainly depends on size deviation or edge misplacement and other shape parameters, while the present application models the symmetric gray difference, dynamically generates the light transmittance difference limit UYATR, and then compares the gray distribution on both sides point by point, proposes the idea of measuring the internal structure consistency by gray symmetry, and breaks through the limitation that the invisible is not measurable in traditional detection. The present application adopts multi-level statistical quantity fusion judgment, constructs a multi-order oscillation judgment mechanism different from the simple threshold method, which considers local abnormal fluctuations and has an automatic compensation mechanism, significantly improves the fault tolerance and detection robustness, and is difficult for conventional engineering personnel to naturally think of. And the industry generally believes that the light transmittance of sanitary napkin has no direct functional index significance and is not used as a core parameter for defect judgment, so the present application promotes it to an important identification dimension of structural defects, breaks through the limitation of traditional quality inspection that only visible defects can be judged, and is a reverse inspiration and breakthrough to the existing technology.

[0052] The present application has the following beneficial effects: by constructing a multi-modal structure analysis model based on light transmittance gray scale image, the detection method relying on traditional surface image recognition is broken through, the structural implicit defects caused by uneven compaction of the absorbent layer, insufficient material thickness and the like can be identified, and it is especially suitable for positioning potential side leakage sources such as sanitary napkin edges or internal "thin transmission areas"; the abnormal area can be stably identified under different illumination and image noise conditions, and the light transmittance difference limit UYATR model designed based on the symmetry analysis of left and right gray scale images improves the sensitive discrimination ability for structural asymmetry defects, makes up for the low recognition rate and fuzzy standard of the existing detection system for defects such as uneven structure or imbalance of pressing, and thus realizes the overall upgrade of intelligent identification of sanitary napkin defects. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 The flowchart of a sanitary napkin production defect intelligent detection method based on multi-modal image analysis is shown. DETAILED DESCRIPTION

[0054] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0055] Figure 1 The diagram shows a flowchart of an intelligent detection method for sanitary napkin production defects based on multimodal image analysis.

[0056] Reference Figure 1 This invention proposes an intelligent detection method for sanitary napkin production defects based on multimodal image analysis. The method includes the following steps:

[0057] S100: Lay the sanitary napkin flat on a transparent surface and shine a light on the sanitary napkin to obtain a translucent image of the illuminated sanitary napkin;

[0058] S200: Grayscale image of sanitary napkin is obtained by converting the light-transmitted image of the sanitary napkin to grayscale.

[0059] S300 obtains the light-transmitting display grid based on the grayscale image of the sanitary napkin, and obtains the actual light-transmitting area through the light-transmitting display grid;

[0060] S400, the grayscale of sanitary napkins Figure One The image is divided into two parts, a left grayscale image and a right grayscale image. The actual light-transmitting areas of the left grayscale image and the right grayscale image are compared to obtain the side leakage area.

[0061] According to the intelligent detection method of the present invention, the presence of side leakage defects in the structure of a sanitary napkin can be detected by visual inspection using the light transmittance information of the sanitary napkin.

[0062] Furthermore, in step S100, laying the sanitary napkin flat on a transparent surface and illuminating it with a light to obtain a translucent image of the illuminated sanitary napkin includes:

[0063] The sanitary napkin product to be detected is laid flat on a transparent optical work platform, the platform is made of high light transmittance glass to ensure that the bottom light can uniformly penetrate, and the laying process ensures that the sanitary napkin is wrinkle-free, non-overlapping, and the long axis direction is parallel to the platform coordinate axis for subsequent image partitioning and symmetric processing; then the top lighting and bottom transmission light source system is started, a visible light LED light source with a wavelength range of 600 nm is used, and the optical response of the sanitary napkin structure is excited by vertical irradiation; wherein the bottom light source penetrates the sanitary napkin and forms a transmission image on the top imaging device, reflecting the light density difference of the internal absorbing material, fiber distribution and compression structure of the product; the top imaging device selects a high-resolution industrial camera Vieworks VC-4MC-M180 to collect complete transmission image data and save it as an original image file for subsequent image processing.

[0064] Further, in step S200, the transmission image of the sanitary napkin is grayed to obtain a sanitary napkin grayscale image, including:

[0065] The obtained transmission image is input into the image preprocessing software OpenCV for gray processing; the original color image is converted into a single-channel grayscale image by using an image processing algorithm, and the brightness component of each pixel is extracted as its grayscale value; wherein the gray processing is realized by using a weighted average method, a maximum value method or an OpenCV standard grayscale conversion function, so that the value of each pixel point in the image is in the interval of 0-255, and the larger the value is, the stronger the transmission is, and the smaller the value is, the more dense or opaque the structure is; finally, the obtained sanitary napkin grayscale image can clearly reflect the structural transmission characteristics of the product under light excitation, serving as the input basis for subsequent transmission grid analysis and defect identification.

[0066] Further, in step S300, the transmission grid is obtained according to the sanitary napkin grayscale image, including:

[0067] The grayscale image is divided into K grids, where K=1800, and sut(i) represents the grayscale value of the i-th grid of the grayscale image, i is in the range of [1, K], and K is the number of grids after the grayscale image is divided. The median ur of the grayscale values of each grid in sut(i) is obtained, the average urm of the grayscale values of each grid in sut(i) is obtained, and the maximum value ura of the grayscale values of each grid in sut(i) is obtained and recorded; wherein the grayscale value of the grid is the average value of the grayscale values of all pixels in the grid; the balanced transmission grayscale urp is calculated, and the calculation method of the balanced transmission grayscale is:

[0068]

[0069] Further, the grid with a grayscale value greater than urp is recorded as a transmission grid, and the grid with a grayscale value less than urp is recorded as a block grid.

[0070] Further, in step S300, obtaining the real light transmission area through the light transmission grid includes the following steps:

[0071] S301, let zj(j) represent the gray value of the jth light transmission grid, j is in the range of [1, H], where H is the number of light transmission grids; zg(j) represents the gray value of the light blocking grid closest to the center point of the jth light transmission grid; for each zg(j), obtain the median and maximum value, and the average of the median and maximum value is recorded as zjxp;

[0072] S302, define an integer variable k, set the initial value to 1, and create two variables sutPTRa and sutPTRb initialized to zero for subsequent calculation and comparison;

[0073] S303, create a blank sequence TYR1, recorded as the real light transmission grid sequence, and create a blank sequence TYR2, recorded as the fuzzy light transmission grid sequence;

[0074] S304, calculate the values of sutPTRa and sutPTRb, where: the calculation formula of sutPTRa is zg(k)-zj(k), and the calculation formula of sutPTRb is [zg(k)+zjxp] / 2-urp; compare the values of sutPTRa and sutPTRb: if sutPTRa is greater than sutPTRb, add the grid sequence number where zj(k) is located to the real light transmission grid sequence TYR1; if sutPTRa is less than or equal to sutPTRb, add the grid sequence number where zg(k) is located to the fuzzy light transmission grid sequence TYR2;

[0075] S305, if the current variable k is less than H, increase k by 1 and return to step S304; if the current variable k is not less than H, jump to S306;

[0076] S306, obtain the median puy in the fuzzy light transmission grid sequence TYR2, add the gray values greater than puy in the fuzzy light transmission grid sequence TYR2 to the real light transmission grid sequence TYR1; the area composed of the grids corresponding to the gray values in the real light transmission grid sequence TYR1 is recorded as the real light transmission area.

[0077] Further, in step S400, the sanitary napkin gray value Figure One is divided into two, obtaining the left gray value and the right gray value, comparing the real light transmission areas of the left gray value and the right gray value, and obtaining the side leakage area includes the following steps:

[0078] S401, the acquired sanitary towel gray scale is vertically divided with the horizontal width midpoint as the symmetry axis to obtain a left gray scale and a right gray scale, which represent the left and right symmetrical areas of the product structure respectively;

[0079] Based on the symmetry of the sanitary towel, the symmetry axis obtained by vertically dividing the sanitary towel image according to the horizontal width midpoint is used to divide the gray scale into a left gray scale and a right gray scale, which are used to compare the real light transmission grid distribution in the respective areas to identify potential structural asymmetry and side leakage risk. By comparing the real light transmission areas in the left and right gray scales, a side transmission area is obtained: each grid in the left gray scale is checked to determine whether it belongs to a real light transmission area.

[0080] S402, the real light transmission area is matched and compared to obtain a side transmission area;

[0081] Specifically, for the real light transmission area in the left gray scale, it is checked whether the corresponding position in the right gray scale is a real light transmission area; if not, the left side of the area is marked as a side transmission area.

[0082] Further, for the real light transmission area in the right gray scale, it is checked whether the corresponding position in the left gray scale is a real light transmission area; if not, the right side of the area is marked as a side transmission area.

[0083] S403, the areas in the left gray scale and the right gray scale that are at the same position and are both real light transmission grids are extracted and marked as the same transmission area, and the gray scale value sequences of the same transmission area in the left gray scale and the right gray scale are extracted to form a left same transmission sequence and a right same transmission sequence;

[0084] Specifically, the grids at the same position in the left gray scale and the right gray scale that belong to the real light transmission area are marked as the same transmission area, the grids of the same transmission area in the left gray scale are marked as the left same transmission grid, and the grids of the same transmission area in the right gray scale are marked as the right same transmission grid.

[0085] zyp(f) represents the gray scale value of the fth left same transmission grid in the left gray scale; zhg(f) represents the gray scale value grid at the same position as the fth left same transmission grid in the right gray scale, f is the serial number, f takes the value [1, T], where T is the number of left same transmission grids in the left gray scale; an empty sequence is created, zyp(f) is sequentially imported into the sequence and marked as the left same transmission sequence, an empty sequence is created, zhg(f) is sequentially imported into the sequence and marked as the right same transmission sequence; the average value of the left same transmission sequence is obtained and marked as zym.

[0086] S404, the light transmission shock interval is obtained by the left same transmission sequence, and the light transmission difference limit is calculated;

[0087] Specifically, the upper quartile ZTJ and the lower quartile ZTK of the left homochromatic sequence are obtained; when an element is less than or equal to ZTK, the element is recorded as a low-order light transmission shock range value; when an element is greater than ZTK and less than ZTJ, the element is recorded as a middle-order light transmission shock range value; when an element is greater than or equal to ZTJ, the element is recorded as a high-order light transmission shock range value; the minimum value of the left homochromatic sequence is recorded as SOU, the maximum value of the left homochromatic sequence is recorded as MAX, and the average value of the left homochromatic sequence is recorded as zym, the value range of all low-order light transmission shock ranges, middle-order light transmission shock ranges and high-order light transmission shock ranges is recorded as [SOU, ZTK], (ZTK, ZTJ) and [ZTJ, MAX] respectively; the difference between the average value zym of the left homochromatic sequence and the minimum shock value SOU is recorded as a first shock difference fluctuation Fos1; the difference between the maximum shock point MAX and the average value zym of the left homochromatic sequence is recorded as a second shock difference fluctuation Fos2; the difference between SOU and ZTK is recorded as a lower boundary extension Fxer; the difference between MAX and ZTJ is recorded as an upper boundary extension Rxer;

[0088] The smaller value of Fos1 and Fos2 is used as the numerator, the larger value of Fos1 and Fos2 is used as the denominator, and the ratio of the numerator and the denominator is used as the light transmission difference limit coefficient YTER; when Fxer is less than or equal to Rxer, the light transmission difference limit UYATR is Fxer x YTER; when Fxer is greater than Rxer, the light transmission difference limit UYATR is Rxer x YTER.

[0089] S405, obtaining a side leakage grid through the light transmission difference limit;

[0090] The difference between zyp(f) and zhg(f) is calculated in sequence, and the absolute value of the difference between zyp(f) and zhg(f) is recorded as chaz(f); chaz(f) is compared with the light transmission difference limit UYATR, if chaz(f) is greater than the light transmission difference limit UYATR, zyp(f) and zhg(f) are compared, and the corresponding grid of the larger gray value between zyp(f) and zhg(f) is marked as a side leakage grid;

[0091] S406, obtaining a side leakage area through the side leakage grid;

[0092] The number of side leakage grids of the left and right gray scale images is obtained, and the number of side leakage grids of the left gray scale image is recorded as TUEW, and the number of side leakage grids of the right gray scale image is recorded as YUEW. The area composed of the side leakage grids of the left gray scale image is recorded as the left side leakage area, and the area composed of the side leakage grids of the right gray scale image is recorded as the right side leakage area. The number of side leakage grids of the left and right gray scale images is compared. If the number of side leakage grids of the right gray scale image YUEW is greater than 2xTUEW, it indicates that there is a defect on the right side of the sanitary napkin, and the right side leakage area will leak during the use of the sanitary napkin. The right side leakage area is recorded as the leakage area. If TUEW is greater than 2xYUEW, it indicates that the sanitary napkin has a defect that will leak, which indicates that there is a defect on the left side of the sanitary napkin, and the left side leakage area will leak during the use of the sanitary napkin. The left side leakage area is recorded as the leakage area.

[0093] The embodiment process is as follows:

[0094] Implementation environment

[0095] Samples: 100 pieces of market conventional sanitary napkins (model A), of which 20 pieces are artificially manufactured thin defect samples, and the positions are random; 80 pieces are normal controls.

[0096] Equipment: High light transmittance acrylic platform, up and down adjustable visible light LED light source (600 nm waveband), industrial camera Vieworks VC-4MC-M180.

[0097] Software: OpenCV4.5+Python3.8 realizes gray scale, grid analysis, symmetry analysis, difference judgment and other steps.

[0098] Artificial defect sample implementation step process:

[0099] S100, light transmission image acquisition: uniform environmental light intensity 400 lux, image acquisition and saving.

[0100] S200, gray scale: using OpenCV4.5 to convert to gray scale, pixel value range 0-255.

[0101] S300-S306 extract real light transmission area: gray scale is divided into K=1800 grids. Statistics ur=125, urm=128, ura=200, and urp≈151 are calculated. Identify the light transmission appearance grid≈400, and further identify the real light transmission area≈250 grids.

[0102] S401-S406 symmetry and side leakage identification: after cutting, the same light transmission area sequence length T≈180 is extracted.

[0103] Statistics: SOU = 110, ZTK = 130, ZTJ = 160, MAX = 210, zym = 145. Calculate Fos1 = 35, Fos2 = 65; YTER = 0.538; Fxer = 20, Rxer = 50; UYATR = 20 * 0.538 = 10.8.

[0104] Calculate chaz(f), identify the person as a defective sample side leakage grid about 30, while the average control sample 15.

[0105] Compare the number of side leakage grids of the left and right gray scale images: TUEW = 13, YUEW = 29, YUEW > 2 * TUEW, it is determined that there is a side leakage defect on the right side.

[0106] Experimental results and comparison:

[0107]

[0108] This embodiment shows that through a series of algorithm processes based on light transmission gray scale grid analysis, symmetry difference discrimination, etc., the identification of implicit structural defects and potential side leakage risks can be realized at a higher accuracy rate than traditional visual algorithms, and the false detection and omission rates are significantly reduced, which has high industrial application value.

[0109] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logic functions, which can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute them. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by or in connection with an instruction execution system, apparatus or device, or in conjunction with these instruction execution systems, apparatus or devices. More specific examples (non-exhaustive list) of computer-readable media include the following: electrical connections having one or more wires (electronic devices), portable computer disk boxes (magnetic devices), random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memories), fiber optic devices, and portable compact disc read-only memories (CDROMs). In addition, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be electronically obtained, for example, by optical scanning of the paper or other media, followed by editing, interpretation or processing, if necessary, in other suitable ways, and then stored in a computer memory.

[0110] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the embodiments described above, various steps or methods can be implemented, for example, through software or firmware in storage media which are executed by suitable instruction executing systems. For example, if implemented in hardware, and in another embodiment, any of the following technologies, known in the art, or their combinations, can be used: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

[0111] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present description, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0112] In the description of the present application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0113] In addition, the terms "first", "second", and the like used in the embodiments of the present application are only for the purpose of description, and can not be understood as indicating or implying relative importance, or implicitly indicating the number of technical features referred to in the embodiments. Therefore, the features defined with "first", "second" and the like in the embodiments of the present application can be explicitly or implicitly indicated to include at least one of the features in the embodiments. In the description of the present application, the meaning of the word "plurality" is at least two or two or more, such as two, three, four, etc., unless otherwise specifically limited in the embodiments.

[0114] In the present application, unless otherwise explicitly specified or limited in the embodiments, the terms "mounting", "connecting", "connecting" and "fixing" and the like appearing in the embodiments should be understood broadly, for example, the connection can be fixed connection, or detachable connection, or integral, which can be understood, or mechanical connection, electrical connection, etc. Of course, it can also be directly connected, or indirectly connected through an intermediate medium, or the internal communication of two elements, or the interaction relationship of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific implementation situation.

[0115] In the present application, unless otherwise explicitly specified and limited, the first feature is "on" or "under" the second feature. The first and second features can be in direct contact, or the first and second features can be indirectly contacted through an intermediate medium. Moreover, the first feature "above", "above" and "above" the second feature can be directly above or obliquely above the first feature, or only indicate that the horizontal height of the first feature is higher than that of the second feature. The first feature "below", "below" and "below" the second feature can be directly below or obliquely below the first feature, or only indicate that the horizontal height of the first feature is less than that of the second feature.

[0116] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. Those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A method for intelligent detection of defects in sanitary napkin production based on multimodal image analysis, characterized in that, The method includes the following steps: S100: Lay the sanitary napkin flat on a transparent surface and shine a light on the sanitary napkin to obtain a translucent image of the shone sanitary napkin; S200: Grayscale image of sanitary napkin is obtained by converting the light-transmitted image of the sanitary napkin to grayscale. S300, obtaining a light-transmitting display grid based on the grayscale image of the sanitary napkin, and obtaining the actual light-transmitting area through the light-transmitting display grid; wherein step S300 includes: S301, let zj(j) represent the gray value of the j-th light-transmitting display grid, where the value of j is in the range of [1,H], and H is the number of light-transmitting display grids; zg(j) represent the gray value of the light-blocking grid that is closest to the center point of the j-th light-transmitting display grid; for each zg(j), obtain the median and the maximum value, and record the average of the median and the maximum value as zjxp; S302, define an integer variable k, initialized to 1, and create two variables sutPTRa and sutPTRb, initialized to zero, for subsequent calculations and comparisons; S303, create a blank sequence TYR1, denoted as the real light-transmitting grid sequence, and create a blank sequence TYR2, denoted as the blurred light-transmitting grid sequence; S304, calculate the values ​​of sutPTRa and sutPTRb, where: sutPTRa is calculated as zg(k)−zj(k), and sutPTRb is calculated as [zg(k)+zjxp] / 2−urp; compare the values ​​of sutPTRa and sutPTRb: if sutPTRa is greater than sutPTRb, add the grid number of zj(k) to the real transparent grid sequence TYR1; if sutPTRa is less than or equal to sutPTRb, add the grid number of zg(k) to the blurred transparent grid sequence TYR2; S305, if the current variable k is less than H, then increment k by 1 and return to step S304; if the current variable k is not less than H, then jump to S306. S306, obtain the median puy in the fuzzy transparent grid sequence TYR2, add the gray values ​​greater than puy in the fuzzy transparent grid sequence TYR2 to the real transparent grid sequence TYR1; the area composed of the grids corresponding to the gray values ​​in the real transparent grid sequence TYR1 is recorded as the real transparent area. S400, the grayscale image of the sanitary napkin is divided into two parts to obtain a left grayscale image and a right grayscale image. The actual light-transmitting areas of the left grayscale image and the right grayscale image are compared to obtain the side leakage area; wherein step S400 includes: S401, the acquired grayscale image of the sanitary napkin is vertically divided with the midpoint of the horizontal width as the axis of symmetry to obtain a left grayscale image and a right grayscale image, which respectively represent the left and right symmetrical areas of the product structure; S402, compare and match the actual light-transmitting area to obtain the side-transmitting area; S403, extract the regions in the left and right grayscale images that are in the same position and are both real light-transmitting grids, and record them as the same-transmitting regions. Extract the grayscale value sequences of these regions in the left and right grayscale images respectively to form the left same-transmitting sequence and the right same-transmitting sequence. S404, the light transmission oscillation range is obtained through the left same light transmission sequence, and the light transmission difference limit is calculated; S405, obtaining the side leakage mesh through the limit of light transmittance difference; S406, obtain the side leakage area through the side leakage grid.

2. The intelligent detection method for sanitary napkin production defects based on multimodal image analysis according to claim 1, characterized in that, Step S100 includes: laying the sanitary napkin product to be tested flat on a transparent optical working platform, the platform being made of high-transmittance glass or acrylic sheet; then activating the top lighting and bottom transmission light source system, using a visible light LED light source with a wavelength range of 400~700nm, to excite the optical response of the sanitary napkin structure through vertical illumination; the bottom light source penetrates the sanitary napkin and forms a translucent image on the top imaging device.

3. The intelligent detection method for sanitary napkin production defects based on multimodal image analysis according to claim 1, characterized in that, Step S200 includes: inputting the translucent image into the image preprocessing software OpenCV and performing grayscale operation on it; using image processing algorithms to convert the original color image into a single-channel grayscale image and extracting the brightness component of each pixel as its grayscale value; wherein the grayscale processing is performed using the weighted average method, the maximum value method, or the OpenCV standard grayscale transformation function, so that the grayscale value of each pixel in the image is within the range of 0~255.

4. The intelligent detection method for sanitary napkin production defects based on multimodal image analysis according to claim 1, characterized in that, Step S300 includes: dividing the grayscale image into K grids, where K=1800, and sut(i) represents the grayscale value of the i-th grid in the grayscale image, where i takes the value [1, K], and K is the number of grids after the grayscale image is divided; obtaining the median ur of the grayscale values ​​of each grid in sut(i), obtaining the average urm of the grayscale values ​​of each grid in sut(i), and obtaining the maximum grayscale value of each grid in sut(i) and recording it as ura; wherein the grayscale value of the grid is the average of the grayscale values ​​of all pixels in the grid; calculating the equalized light transmission grayscale urp, wherein the calculation method for the equalized light transmission grayscale is: urp=ur×(ur+ura) / 2urm; and recording the grids with grayscale values ​​greater than urp as light-transmitting grids and the grids with grayscale values ​​less than urp as light-blocking grids.

Images