Quality control method and device for knitted fabric production process

Abstract

The invention provides a knitted fabric production process quality control method and device, and belongs to the technical field of textile, and the method comprises the steps: determining the structure type of a knitted fabric based on a first detection result of a first visual detection unit; comparing the structure type of the knitted fabric with a preset knitted fabric structure type to determine a second detection mode of the knitted fabric; determining a quality parameter of the knitted fabric based on a second detection mode, and performing quality detection on the knitted fabric based on the second detection mode to obtain a second detection result; and updating control parameters in the knitted fabric production process based on the second detection result and the quality parameters of the knitted fabric. According to the strip-shaped mosquito net cloth quality detection device, the strip-shaped mosquito net cloth quality detection accuracy and the product production efficiency can be improved.

Description

Technical Field

[0001] The present disclosure belongs to the technical field of textiles, and more particularly, relates to a method and device for quality control in the production process of knitted fabrics. Background Art

[0002] Due to its unique softness, breathability, and good wearing comfort, knitted fabrics have been widely used in many fields such as clothing, household items, and industrial textiles. With the continuous improvement of consumers' requirements for product quality and the increasingly fierce market competition, the quality of knitted fabrics has become a key factor for enterprises to gain a foothold in the market. Although there are currently various quality detection methods for knitted fabrics, these methods still have certain limitations, with low accuracy in quality detection and difficulty in reflecting the product quality status. Summary of the Invention

[0003] The purpose of the present disclosure is to provide a method and device for quality control in the production process of knitted fabrics to improve the accuracy of quality detection of knitted fabrics.

[0004] In the first aspect of the embodiments of the present disclosure, a method for quality control in the production process of knitted fabrics is provided, including: Determining the structural type of the knitted fabric based on the first detection result of the first vision detection unit; Comparing the structural type of the knitted fabric with a preset knitted fabric structural type to determine the second detection method for the knitted fabric; Determining the quality parameters of the knitted fabric based on the second detection method, and performing quality detection on the knitted fabric based on the second detection method to obtain a second detection result; Updating the control parameters in the production process of the knitted fabric based on the second detection result and the quality parameters of the knitted fabric.

[0005] In the second aspect of the embodiments of the present disclosure, a device for quality control in the production process of knitted fabrics is provided, including: A first detection module for determining the structural type of the knitted fabric based on the first detection result of the first vision detection unit; An analysis module for comparing the structural type of the knitted fabric with a preset knitted fabric structural type to determine the second detection method for the knitted fabric; A second detection module for determining the quality parameters of the knitted fabric based on the second detection method, and performing quality detection on the knitted fabric based on the second detection method to obtain a second detection result; A parameter update module for updating the control parameters in the production process of the knitted fabric based on the second detection result and the quality parameters of the knitted fabric.

[0006] In a third aspect of the embodiments of the present disclosure, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the above-mentioned quality control method for the knitted fabric production process are implemented.

[0007] In a fourth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned quality control method for the knitted fabric production process are implemented.

[0008] The beneficial effects of the quality control method and device for the knitted fabric production process provided by the embodiments of the present disclosure are as follows: On the one hand, by introducing the first visual detection unit, the present disclosure can quickly and accurately determine the structure type of the knitted fabric. At the same time, by automatically selecting the corresponding second detection method according to the structure type of the knitted fabric, the quality detection process becomes more targeted and efficient. By precisely matching the structure type of the knitted fabric with the preset detection method, the quality parameters of the knitted fabric can be evaluated more accurately, thereby realizing the fine control of the production process. This can not only improve production efficiency but also significantly improve the quality control accuracy.

[0009] On the other hand, based on the second detection result and the quality parameters of the knitted fabric, the present disclosure updates the control parameters in the knitted fabric production process in real time. This dynamic adjustment mechanism can ensure that the production process always remains in the best state, effectively avoiding quality problems caused by improper parameters. By continuously optimizing the production parameters, the control method provided by the present disclosure can significantly reduce the defective rate and improve the product qualification rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0011] Figure 1 It is a schematic flowchart of the quality control method for the knitted fabric production process provided by an embodiment of the present disclosure; Figure 2 It is a structural block diagram of the quality control device for the knitted fabric production process provided by an embodiment of the present disclosure; Figure 3 It is a schematic block diagram of the electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0012] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present disclosure. However, those skilled in the art should clearly understand that the present disclosure can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present disclosure.

[0013] To make the objectives, technical solutions, and advantages of the present disclosure clearer, the following will be described through specific embodiments in conjunction with the accompanying drawings.

[0014] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of a quality control method for the production process of knitted fabrics provided by an embodiment of the present disclosure. The method includes: S101: Determine the structure type of the knitted fabric based on the first detection result of the first vision detection unit.

[0015] In this embodiment, the first vision detection unit is a device or system component specifically used for visually detecting knitted fabrics. The first vision detection unit may include a camera, a light source device, etc. The camera can acquire image information of the knitted fabric, and the light source device is used to provide appropriate lighting conditions for image acquisition to ensure clear and accurate capture of the characteristics of the knitted fabric.

[0016] The structure types of knitted fabrics include single-sided fabric, double-sided fabric, jacquard fabric, etc. Different fabrics have different knitting rules and appearance characteristics. From the perspective of appearance morphology, the structure types can be classified according to the texture characteristics, density, coil shape, etc. of the knitted fabric. For example, there are fine plain structures, patterned structures with a concave-convex feeling, and loose mesh structures.

[0017] The first detection result is the output information obtained after the first vision detection unit detects the knitted fabric. This output information is presented in various forms, such as the image data of the knitted fabric (including visual information such as the appearance and texture of the knitted fabric). The output information can also be data obtained after preliminary analysis of the image data of the knitted fabric, such as certain characteristic parameters (needle gauge, coil density, texture direction, printing characteristics, etc.).

[0018] In this embodiment, determining the structure type of the knitted fabric based on the first detection result of the first vision detection unit includes: Extract features from multiple images collected by the first vision detection unit to obtain the overall features and local features of the knitted fabric; Input the overall features and local features of the knitted fabric into a classification and recognition model to obtain the structure type of the knitted fabric. The classification and recognition model is a trained neural network model.

[0019] The overall characteristics of the knitted fabric include texture complexity, color distribution, and density. The overall characteristics of the knitted fabric are obtained by feature extraction from multiple images collected by the first visual detection unit, including: texture complexity, color distribution, and density are obtained by feature extraction from multiple images.

[0020] Among them, the calculation method of texture complexity is as follows: determine the first distance and the first angle, and count the frequency of the gray-scale combinations of pixel pairs with a distance of the first distance and an angle of the first angle in multiple images to form a gray-level co-occurrence matrix; calculate feature statistics based on the gray-level co-occurrence matrix, and there are multiple feature statistics; weight the multiple feature statistics to obtain the texture complexity. The multiple feature statistics are at least two of contrast, correlation, energy, and homogeneity.

[0021] The local characteristics of the knitted fabric include loop morphology and pattern features. The local characteristics of the knitted fabric obtained by feature extraction from multiple images collected by the first visual detection unit include: the loop morphology and pattern features of the knitted fabric are obtained by feature extraction from multiple images collected by the first visual detection unit. The loop morphology includes the size, shape, and uniformity of the loops, etc. Different knitted fabric structure types have different loop morphology characteristics. For example, the loops of single-sided fabric have obvious front and back sides. The front loops present loop columns, which are relatively smooth, neat, and have a strong sense of line; the back loops are loop arcs, with a relatively rough appearance and certain undulations. Both sides of double-sided fabric are composed of loops, without obvious front and back differences, and the appearances of both sides are relatively similar. The pattern complexity is relatively low, generally mainly simple geometric patterns, stripes, grids, etc. The pattern complexity of double-sided fabric is moderate, and some patterns richer than those of single-sided fabric can be formed, such as combinations of stripes with different widths, splicing of simple geometric patterns, etc. The pattern complexity of jacquard fabric is high, and very complex and delicate pattern effects can be achieved. Various details can be accurately reproduced, the lines of the patterns are smooth, and the color transitions are natural, achieving an effect like a painting. It can not only have complex contours and shapes, but also incorporate multiple colors and hierarchical changes in one pattern.

[0022] Input the overall features and local features of the knitted fabric into the classification and recognition model to obtain the structural type of the knitted fabric. It can be understood that: First, normalize the extracted overall features and local features, and map the feature values to a fixed range, such as the interval [0, 1]. Commonly used normalization methods include min-max normalization and Z-score normalization. Normalization can eliminate the dimensional differences between different features and improve the training effect and stability of the model. Second, feature selection and dimensionality reduction processing are required. Methods such as correlation analysis and principal component analysis can be used to select the features that contribute most to the discrimination of the structural type, reduce the feature dimension, and retain important information at the same time. Finally, input the preprocessed overall features and local features into the trained classification and recognition model to obtain the structural type of the knitted fabric.

[0023] To ensure the reliability of the classification results, the confidence of the classification results can be evaluated. For example, judge the confidence of the model in the classification results through the probability values of the output layer of the classification and recognition model. If the probability value is close to 1, it means that the model has a high confidence in the classification result; if the probability value is low, further inspection is required.

[0024] S102: Compare the structural type of the knitted fabric with the preset knitted fabric structural types to determine the second detection method of the knitted fabric.

[0025] In this embodiment, the preset knitted fabric structural types are a series of standard structural types and their corresponding quality characteristics and detection requirements set according to production requirements, product standards or experience summaries before the start of the production process. These preset knitted fabric structural types contain detailed descriptions of knitted fabrics of different structural types, such as coil density, pattern regularity, elasticity requirements, etc.

[0026] Knitted fabrics of different structural types may have different quality defects and problems. Therefore, different detection methods need to be adopted to ensure the accuracy and effectiveness of the detection. In this embodiment, by comparing the actually detected structural type of the knitted fabric with the preset knitted fabric structural types, the second detection method of this knitted fabric can be selected according to the matching situation and differences between the two, so as to more accurately detect the quality problems existing in the knitted fabric.

[0027] S103: Determine the quality parameters of the knitted fabric based on the second detection method, and perform quality detection on the knitted fabric based on the second detection method to obtain the second detection result.

[0028] In this embodiment, the quality parameters of the knitted fabric include physical performance parameters, appearance quality parameters and internal quality parameters. The physical performance parameters include coil density, yarn linear density, fabric thickness. The appearance quality parameters include pattern quality defects, color uniformity, etc. The internal quality parameters include internal broken yarn parameters, hole and gap parameters, etc.

[0029] After the second detection method is determined, the quality parameters of the knitted fabric can be determined based on the second detection method. For example, if the second detection method is only visual inspection, the quality parameters of the knitted fabric can be coil density, color uniformity, etc. If the second detection method includes both visual inspection and ultrasonic inspection, the quality parameters of the knitted fabric can be coil density, yarn linear density, fabric thickness, pattern quality defects, color uniformity, internal broken yarn parameters, hole and gap parameters, etc.

[0030] The quality inspection of the knitted fabric is carried out based on the second detection method to obtain the second detection result, including: If the second detection method is visual inspection, the detection equipment can be a high-resolution industrial camera, a suitable light source, and image processing software. The industrial camera can collect images of the knitted fabric, the light source ensures clear images and uniform illumination, and the image processing software analyzes and processes the images. The image processing software can use specific algorithms to identify coils and count the number of coils per unit length or area to determine the coil density; it can also evaluate the color uniformity by calculating statistical quantities such as the average value and standard deviation of colors through the color analysis function, so as to obtain the corresponding quality inspection result.

[0031] If the second detection method is a combination of visual inspection and ultrasonic inspection, in addition to the equipment required for visual inspection, the ultrasonic inspection equipment needs to be started. First, use visual inspection to obtain the appearance quality parameters of the knitted fabric, such as coil density, pattern quality defects, color uniformity, etc. Then, use the ultrasonic inspection equipment to emit ultrasonic signals to the knitted fabric, receive the reflected waves and analyze their characteristics to detect internal quality parameters such as yarn linear density, fabric thickness, internal broken yarn parameters, hole and gap parameters, etc. Finally, integrate the results obtained by the two detection methods to form a comprehensive second detection result, providing a basis for the quality evaluation of the knitted fabric.

[0032] S104: Update the control parameters in the production process of the knitted fabric based on the second detection result and the quality parameters of the knitted fabric.

[0033] In this embodiment, each specific index in the second detection result, such as the number of defects, density deviation, color difference, etc., is compared one by one with the pre-set quality parameter standard of the knitted fabric. Determine the quality parameters that meet the standard and the quality parameters that do not meet the standard. Calculate the deviation value of the quality parameters that do not meet the standard, and update the control parameters in the production process of the knitted fabric based on the deviation value.

[0034] Different deviation values of quality parameters often correspond to different control parameters. For example, if the color and luster uniformity does not meet the standard, the control parameters in the dyeing process can be adjusted, such as dyeing temperature, dyeing time, dye concentration, etc.; if problems such as internal broken yarn are found, the control parameters such as the yarn feeding tension and the state of knitting needles of the knitting machine can be checked and adjusted. At the same time, the mutual influence between various control parameters should be comprehensively considered to avoid adjusting one parameter and causing other quality problems.

[0035] After determining the control parameters, update the control parameters in the production process of the knitted fabric according to the determined adjustment direction and amplitude. During the update process, the quality of the produced knitted fabric is monitored in real time. For example, a small-scale test method can be adopted to observe the effect after adjustment, and the control parameters can be further optimized according to the feedback results. Continuously cycle this process until the quality of the knitted fabric meets the preset quality parameter standard, realizing the stability of the production process and the improvement of product quality.

[0036] As can be seen from the above, on the one hand, in the embodiment of the present disclosure, by introducing the first visual detection unit, the structural type of the knitted fabric can be quickly and accurately determined. At the same time, the corresponding second detection method is automatically selected according to the structural type of the knitted fabric, making the quality detection process more targeted and efficient. By accurately matching the structural type of the knitted fabric with the preset detection method, the quality parameters of the knitted fabric can be more accurately evaluated, thereby realizing the fine control of the production process. This can not only improve production efficiency but also significantly improve the quality control accuracy.

[0037] On the other hand, based on the second detection result and the quality parameters of the knitted fabric, the control parameters in the production process of the knitted fabric are updated in real time. This dynamic adjustment mechanism can ensure that the production process always remains in the best state, effectively avoiding quality problems caused by improper parameters. By continuously optimizing the production parameters, the control method provided by the present disclosure can significantly reduce the defective rate and improve the product qualification rate.

[0038] In an embodiment of the present disclosure, the second detection result obtained by performing quality detection on the knitted fabric based on the second detection method includes: If the second detection method is visual detection, the second detection result is obtained by performing quality detection on the knitted fabric based on the second visual detection unit, and the detection accuracy of the second visual detection unit is greater than that of the first visual detection unit; If the second detection method is visual detection and ultrasonic detection, the second detection result is obtained by performing quality detection on the knitted fabric based on the second visual detection unit and the ultrasonic detection unit.

[0039] In this embodiment, the first visual detection unit is mainly used to judge the structural type of the knitted fabric, and its detection accuracy is relatively low. The first visual detection unit can be an ordinary industrial camera equipped with a low-resolution lens. There are two ordinary industrial cameras and two low-resolution lenses, which are respectively arranged on the front and back sides of the knitted fabric. The ordinary industrial camera can obtain the overall images of the front and back sides of the knitted fabric. Cooperating with the low-resolution lens, it can quickly obtain the general structural information of the knitted fabric, such as the basic arrangement mode of the coils, the texture direction of the fabric, etc., which is sufficient to make a preliminary judgment on the structural type of the knitted fabric. For example, the coil arrangements on the front and back sides of a single-sided knitted fabric are different. By analyzing the images of the front and back sides of the knitted fabric, it can be determined whether the knitted fabric is of single-sided organization. Although the first visual detection unit can obtain the structural type of the knitted fabric, it is difficult to clearly see the fine details of the knitted fabric. Therefore, a second visual detection unit is needed to detect the quality of the knitted fabric.

[0040] The second visual detection unit can be a high-resolution industrial CCD camera, a machine vision system, and a 3D visual detection device. The CCD camera has the characteristics of high sensitivity and low noise, and can capture the fine features on the surface of the knitted fabric. With a high-resolution lens and a suitable light source, it can clearly capture details such as the coil morphology and yarn texture of the knitted fabric. The machine vision system consists of a camera, a lens, a light source, an image acquisition card, and image processing software, etc., and can perform high-precision processing and analysis on the knitted fabric image. Through advanced image processing algorithms, the machine vision system can accurately identify various details of the knitted fabric, such as tiny defects and color differences, etc., to achieve precise detection of the quality of the knitted fabric. The 3D visual detection device can obtain the three-dimensional image information of the knitted fabric, and can not only detect the details on the surface of the knitted fabric, but also measure and analyze its three-dimensional features such as thickness and concavity and convexity, with relatively high detection accuracy.

[0041] The ultrasonic detection unit includes devices such as an ultrasonic transmitter, a transducer, an ultrasonic receiver, a signal processing circuit, and a detection host. Its working principle is as follows: First, the ultrasonic transmitter generates a high-frequency electrical signal, which is transmitted to the transducer. The transducer uses the piezoelectric effect to convert the electrical signal into ultrasonic waves and emits them. The ultrasonic waves propagate in the object to be detected such as the knitted fabric. When encountering the interfaces of objects with different densities, such as the yarn intersections in the knitted fabric, internal holes or gaps, and broken yarn positions, etc., phenomena such as reflection, refraction, and scattering will occur. The reflected ultrasonic waves are received by the transducer and are converted into electrical signals again through the piezoelectric effect. The electrical signals are processed by circuits such as multi-stage amplification, filtering, and demodulation in the receiver, and then transmitted to the signal processing circuit for further analysis. Finally, the detection host judges the quality parameters of the knitted fabric according to the processed signals. For example, by analyzing the signal intensity, propagation time, frequency change, etc., parameters such as yarn linear density, fabric thickness, internal broken yarn parameters, hole and gap parameters, etc. are obtained.

[0042] As can be seen from the above, in this embodiment, by introducing a high-precision second vision detection unit, the accuracy and reliability of the quality inspection of knitted fabrics are improved; at the same time, by combining the two methods of vision detection and ultrasonic detection, the quality of knitted fabrics can be more comprehensively detected, effectively avoiding the problem of inaccurate detection by a single detection method and improving the overall quality inspection efficiency of knitted fabrics.

[0043] In an embodiment of the present disclosure, based on the second vision detection unit to perform quality inspection on the knitted fabric to obtain a second detection result, including: Preprocess the knitted fabric image collected by the second vision detection unit, and the preprocessing includes at least one of denoising, grayscale conversion, and contrast enhancement operations; Extract the features in the preprocessed image, compare the extracted features with the preset quality standard features, and generate a second detection result according to the comparison result. The second detection result includes whether there are defects on the surface of the knitted fabric, the position of the defects, and the type of defects.

[0044] In this embodiment, the knitted fabric image data collected by the second vision detection unit will be interfered by the environment. Therefore, it is necessary to preprocess the knitted fabric image data after collection. The preprocessing includes at least one of denoising, grayscale conversion, and contrast enhancement operations. Denoising: The knitted fabric image may be interfered by various noises during the collection process, such as electronic component noise, ambient light noise, etc. Denoising is to use various algorithms and technologies, such as mean filtering, median filtering, etc., to remove the noise in the image and improve the quality and clarity of the image. Grayscale conversion: Convert the color knitted fabric image into a grayscale image. When analyzing the image, only the brightness information of the image needs to be utilized. Grayscale conversion can reduce the amount of data, improve the processing speed, and at the same time can also highlight some structural features of the image, facilitating subsequent feature extraction and analysis. Grayscale conversion is generally achieved by weighted averaging the RGB three channels of the color image. Contrast enhancement: By changing the brightness distribution of the image, expanding the difference between different gray levels in the image, and making the details in the image more clearly visible. For example, the defects of some knitted fabrics may have a small gray difference from the normal part. Through contrast enhancement, these defects can be more easily identified and distinguished. Commonly used methods include histogram equalization, gamma correction, etc.

[0045] The preset quality standard features are a set of feature parameters or feature templates preset according to the quality requirements and standards of the knitted fabric. For example, for a knitted fabric without defects, its loops should have a certain shape and size, and the arrangement should be neat and uniform, which can all be used as quality standard features.

[0046] In this embodiment, after preprocessing the collected images, specific image feature extraction algorithms are used, such as texture feature extraction algorithms based on gray-level co-occurrence matrices, geometric feature extraction algorithms based on edge detection, etc., to extract various features of the knitted fabric. The extracted features of the knitted fabric are compared with the preset quality standard features, and methods such as pattern matching and similarity calculation are used to determine whether the knitted fabric meets the quality standards. If a difference is found between the extracted features and the quality standard features, it indicates that there are defects on the surface of the knitted fabric, and the device can mark the defective images or directly give an early warning to avoid major problems.

[0047] As can be seen from the above, in this embodiment, by preprocessing the knitted fabric images, the image quality is improved; by comparing the feature extraction with the preset quality standard features, the defects of the knitted fabric can be accurately identified, including the presence or absence, location, and type of defects, which improves the quality detection efficiency and accuracy and helps to timely discover and solve quality problems.

[0048] In an embodiment of the present disclosure, based on the second vision detection unit and the ultrasonic detection unit, quality detection of the knitted fabric is performed to obtain a second detection result, including: Feature extraction is performed on the knitted fabric images collected by the second vision detection unit to obtain a first feature set, and feature extraction is performed on the information of the knitted fabric collected by the ultrasonic detection unit to obtain a second feature set; The features in the first feature set and the second feature set are fused to obtain a multi-source fusion feature set, and quality detection of the knitted fabric is performed based on the multi-source fusion feature set to obtain a second detection result.

[0049] Among them, fusing the features in the first feature set and the second feature set to obtain a multi-source fusion feature set includes: Obtaining the multi-source fusion feature set according to the first formula, and the first formula is;

[0050] Among them, represents the multi-source fusion feature set, , F 1 represents the first feature set, represents the i-th feature in the first feature set, and m represents the number of features in the first feature set; , F 2 is the second feature set, represents the j-th feature in the second feature set, and n represents the number of features in the second feature set; represents a two-dimensional weight coefficient matrix, represents a function for mapping a single feature in the first feature set, represents mapping a single feature in the second feature set, denotes an activation function, denotes assigning a single feature to the first feature set the weight coefficient of the feature after mapping, denotes assigning a single feature to the second feature set the weight coefficient of the feature after mapping, denotes a feature interaction function.

[0051] In this embodiment, the first feature set is a set of features extracted from the knitted fabric image collected by the second vision detection unit. denotes the i-th feature in the first feature set, such as the texture feature of the image (such as the local binary pattern feature value), the color feature (such as the color moment feature), etc. The second feature set is a set of features extracted from the knitted fabric information collected by the ultrasonic detection unit. denotes the j-th feature in the second feature set.

[0052] denotes a function that maps a single feature in the first feature set. It can transform the original image feature into a more expressive feature space. For example, using a deep neural network as to perform a non-linear transformation on the input image feature. denotes mapping a single feature in the second feature set to map the ultrasonic feature to a suitable feature space, which can also be a deep neural network structure. and can be the same neural network structure or different neural network structures. denotes an activation function, usually a non-linear activation function such as the ReLU (Rectified Linear Unit) function whose role is to introduce non-linear factors into the fusion process, enhance the expression ability of the model, and enable the model to learn more complex feature relationships. denotes a two-dimensional weight coefficient matrix, reflecting the importance degree of the combination of the i-th feature in the first feature set and the j-th feature in the second feature set in the fusion process, , .

[0053] denotes a feature interaction function used to measure the interaction between the i-th feature in the first feature set and the j-th feature in the second feature set. It can be defined as , denotes the inner product, denotes the norm of the vector, and reflects the interaction by calculating the correlation between features.

[0054] As can be seen from the above, by integrating the features of the second vision detection unit and the ultrasonic detection unit, and using the deep neural network mapping and non-linear activation function, the present disclosure enhances the feature expression ability. At the same time, through the two-dimensional weight coefficient matrix and the feature interaction function, the importance and interaction of features are accurately measured, improving the accuracy and efficiency of knitted fabric detection.

[0055] In an embodiment of the present disclosure, the control parameters in the production process of the knitted fabric are updated based on the second detection result and the quality parameters of the knitted fabric, including: Calculating the deviation value between the second detection result and the quality parameters of the knitted fabric, where the deviation value includes the numerical deviation value of the quality index and the deviation value of the number of defects; According to the preset control strategy and the deviation value, determining the control parameters to be adjusted and the adjustment range, and adjusting the corresponding control parameters of the knitted fabric production equipment.

[0056] In this embodiment, according to the preset control strategy and the deviation value, determining the control parameters to be adjusted and the adjustment range includes: Based on the numerical deviation value of the quality index and the first control strategy in the preset control strategy with a relevance to the quality index greater than the preset relevance, determining the first control parameter; Based on the numerical deviation value of the quality index and the first control strategy in the preset control strategy with a relevance to the quality index greater than the preset relevance, determining the adjustment range of the first control parameter; Based on the deviation value of the number of defects and the second control strategy in the preset control strategy with a relevance to the defect greater than the preset relevance, determining the second control parameter; Based on the deviation value of the number of defects and the second control strategy in the preset control strategy with a relevance to the defect greater than the preset relevance, determining the adjustment range of the second control parameter.

[0057] Among them, the numerical deviation of the quality index reflects the difference between the actual quality value of the knitted fabric and the preset standard value. The preset control strategy is a set of rules preset to cope with different quality situations. By setting the relevance threshold, the control strategy with a greater impact on this quality index can be found. For example, for the quality index of the coil density of the knitted fabric, if there is a deviation between its actual value and the standard value, in the preset control strategy, control strategies such as knitting speed and yarn tension have a relatively high relevance to the coil density, exceeding the preset relevance, then the knitting speed and yarn tension can become the first control parameters to be adjusted.

[0058] After the first control parameter is determined, it is necessary to determine the adjustment range of the first control parameter. Different degrees of deviation require different degrees of adjustment of the first control parameter. For example, if the coil density deviation value is less than the first value, according to the preset first control strategy, the knitting speed is adjusted by the first step length; if the coil density deviation value is greater than or equal to the first value, the knitting speed is adjusted by the second step length; the first step length is less than the second step length. The preset control strategy includes adjustment rules for the first control parameter under different degrees of deviation, and these rules can be formulated based on experience, test data or theoretical models. Through these rules, the adjustment range of the control parameter can be matched with the numerical deviation of the quality index, so as to effectively correct the quality deviation and ensure the product quality.

[0059] The deviation value of the number of defects reflects the gap between the actual number of defects in the knitted fabric and the acceptable standard number. When there is a deviation in the number of defects, by comparing the relevance between the preset control strategy and the defects, the factor corresponding to the strategy with a higher preset relevance can be the second control parameter, such as the adjustment of the equipment cleaning cycle.

[0060] Different deviations in the number of defects require different degrees of adjustment of the control parameter. For example, if the deviation value of the number of defects is less than or equal to 2, the equipment cleaning cycle can be shortened; if the deviation value of the number of defects is greater than 2, the equipment needs to be comprehensively overhauled and the cleaning cycle is greatly shortened. The second control strategy includes adjustment rules for the second control parameter in different cases of the deviation value of the number of defects. By following these rules, the range of the second control parameter can be reasonably adjusted, so as to effectively reduce the number of defects and improve the quality of the knitted fabric.

[0061] It can be concluded from the above that in this embodiment, by calculating the deviation value and adjusting the control parameter in the production process of the knitted fabric according to the preset control strategy, the precise control of the quality of the knitted fabric is realized. This method can effectively correct the quality deviation, reduce the number of defects, ensure the product quality and improve the production efficiency of the knitted fabric.

[0062] Corresponding to the quality control method for the production process of the knitted fabric in the above embodiment, Figure 2 This is a structural block diagram of a quality control device for the production process of a knitted fabric provided by an embodiment of the present disclosure. For the sake of convenience of description, only the parts related to the embodiment of the present disclosure are shown. Refer to Figure 2 The quality control device 20 for the production process of the knitted fabric includes: a first detection module 21, an analysis module 22, a second detection module 23, and a parameter update module 24.

[0063] Among them, the first detection module 21 is used to determine the structural type of the knitted fabric based on the first detection result of the first visual detection unit; The analysis module 22 is used to compare the structural type of the knitted fabric with the preset knitted fabric structural type to determine the second detection method of the knitted fabric; The second detection module 23 is configured to determine the quality parameters of the knitted fabric based on a second detection method, and perform quality detection on the knitted fabric based on the second detection method to obtain a second detection result; The parameter update module 24 is configured to update the control parameters in the production process of the knitted fabric based on the second detection result and the quality parameters of the knitted fabric.

[0064] In an embodiment of the present disclosure, the second detection module 23 is specifically configured to: If the second detection method is visual detection, perform quality detection on the knitted fabric based on the second visual detection unit to obtain a second detection result, and the detection accuracy of the second visual detection unit is greater than that of the first visual detection unit; If the second detection method is visual detection and ultrasonic detection, perform quality detection on the knitted fabric based on the second visual detection unit and the ultrasonic detection unit to obtain a second detection result.

[0065] In an embodiment of the present disclosure, the second detection module 23 is specifically configured to: Preprocess the knitted fabric image collected by the second visual detection unit, and the preprocessing includes at least one of denoising, grayscale conversion, and contrast enhancement operations; Extract the features in the preprocessed image, compare the extracted features with the preset quality standard features, and generate a second detection result according to the comparison result. The second detection result includes whether there are defects on the surface of the knitted fabric, the position of the defects, and the type of the defects.

[0066] In an embodiment of the present disclosure, the second detection module 23 is specifically configured to: Extract features from the knitted fabric image collected by the second visual detection unit to obtain a first feature set, and extract features from the information of the knitted fabric collected by the ultrasonic detection unit to obtain a second feature set; Fuse the features in the first feature set and the second feature set to obtain a multi-source fusion feature set, and perform quality detection on the knitted fabric based on the multi-source fusion feature set to obtain a second detection result.

[0067] In an embodiment of the present disclosure, the second detection module 23 is specifically configured to: Obtain a multi-source fusion feature set according to a first formula, and the first formula is;

[0068] Wherein, represents the multi-source fusion feature set, , F 1 represents the first feature set, represents the i-th feature in the first feature set, and m represents the number of features in the first feature set; , F2 is the second feature set, represents the j-th feature in the second feature set, and n represents the number of features in the second feature set; represents a two-dimensional weight coefficient matrix, represents a function for mapping a single feature in the first feature set, represents mapping a single feature in the second feature set, represents an activation function, represents assigning a single feature of the first feature set the weight coefficient of the feature after mapping, represents assigning a single feature of the second feature set the weight coefficient of the feature after mapping, represents a feature interaction function.

[0069] In an embodiment of the present disclosure, the first detection module 21 is specifically configured to: Extract the overall features and local features of the knitted fabric by performing feature extraction on multiple images collected by the first visual detection unit; Input the overall features and local features of the knitted fabric into the classification and recognition model to obtain the structural type of the knitted fabric, and the classification and recognition model is a trained neural network model.

[0070] In an embodiment of the present disclosure, the parameter update module 24 is specifically configured to: Calculate the deviation value between the second detection result and the knitted fabric quality parameter, and the deviation value includes the numerical deviation value of the quality index and the deviation value of the number of defects; Determine the control parameter to be adjusted and the adjustment amplitude according to the preset control strategy and the deviation value, and adjust the corresponding control parameter of the knitted fabric production equipment.

[0071] See Figure 3 , Figure 3 is a schematic block diagram of an electronic device provided by an embodiment of the present disclosure. As Figure 3 shown, the electronic device 300 in this embodiment may include: one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The above-mentioned processors 301, input devices 302, output devices 303, and memories 304 communicate with each other through the communication bus 305. The memory 304 is used to store computer programs, and the computer programs include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. Among them, the processor 301 is configured to call the program instructions to execute the functions of each module in the above device embodiments, such as Figure 2 the functions of the modules 21 to 24 shown.

[0072] It should be understood that in the embodiments of the present disclosure, the so-called processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0073] The input device 302 may include a touchpad, a fingerprint acquisition sensor (for acquiring the fingerprint information and the direction information of the fingerprint of the user), a microphone, etc., and the output device 303 may include a display (such as an LCD), a speaker, etc.

[0074] The memory 304 may include a read-only memory and a random access memory, and provide instructions and data to the processor 301. A part of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store information about the device type.

[0075] In specific implementation, the processor 301, the input device 302, and the output device 303 described in the embodiments of the present disclosure may execute the implementation manners described in the first embodiment and the second embodiment of the quality control method for the knitted fabric production process provided by the embodiments of the present disclosure, and may also execute the implementation manner of the electronic device described in the embodiments of the present disclosure, which will not be elaborated herein.

[0076] In another embodiment of the present disclosure, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and the computer program includes program instructions. When the program instructions are executed by a processor, all or part of the processes in the method of the above embodiment are implemented. It can also be completed by instructing related hardware through the computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0077] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as the hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device. Further, the computer-readable storage medium can also include both the internal storage unit and the external storage device of the electronic device. The computer-readable storage medium is used to store the computer program and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store the data that has been output or will be output.

[0078] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described according to their functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present disclosure.

[0079] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described electronic devices and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0080] In several embodiments provided by the present application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling, direct coupling, or communication connection between each other can be an indirect coupling or communication connection through some interfaces or units, or can also be in the form of electrical, mechanical, or other connections.

[0081] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present disclosure.

[0082] In addition, in each embodiment of the present disclosure, each functional unit can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0083] The above is only the specific implementation manner of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present disclosure can easily think of various equivalent modifications or replacements, and these modifications or replacements should be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims

1. A method for quality control of knitted fabric production process, characterized in that: include: determining a structural type of the knitted fabric based on a first detection result of the first visual detection unit; Comparing the structure type of the knitted fabric with a preset structure type of the knitted fabric to determine a second detection method of the knitted fabric; determining a quality parameter of the knitted fabric based on a second detection method, and performing a quality detection on the knitted fabric based on the second detection method to obtain a second detection result; The control parameters in the knitted fabric production process are updated based on the second detection result and the quality parameter of the knitted fabric.

2. The knitted fabric production process quality control method according to claim 1, characterized in that: The method of performing quality inspection on the knitted fabric based on the second inspection method to obtain a second inspection result includes: If the second detection method is visual detection, a second detection result is obtained by performing quality detection on the knitted fabric based on a second visual detection unit, and the detection accuracy of the second visual detection unit is greater than the detection accuracy of the first visual detection unit; If the second detection method is visual detection and ultrasonic detection, the quality of the knitted fabric is detected based on the second visual detection unit and the ultrasonic detection unit to obtain a second detection result.

3. The knitted fabric production process quality control method according to claim 2, characterized in that: The method of performing quality inspection on the knitted fabric based on the second visual inspection unit to obtain a second inspection result includes: Preprocessing the knitted fabric image collected by the second visual inspection unit, wherein the preprocessing includes at least one of denoising, graying, and contrast enhancement operations; Features in the preprocessed image are extracted, the extracted features are compared with preset quality standard features, and a second detection result is generated based on the comparison result. The second detection result includes whether there are defects on the surface of the knitted fabric, the location of the defects and the type of defects.

4. The knitted fabric production process quality control method according to claim 2, characterized in that: The method of performing quality inspection on the knitted fabric based on the second visual inspection unit and the ultrasonic inspection unit to obtain a second inspection result includes: Performing feature extraction on the knitted fabric image collected by the second visual detection unit to obtain a first feature set, and performing feature extraction on the knitted fabric information collected by the ultrasonic detection unit to obtain a second feature set; The features in the first feature set and the features in the second feature set are fused to obtain a multi-source fused feature set, and quality inspection of the knitted fabric is performed based on the multi-source fused feature set to obtain a second inspection result.

5. The knitted fabric production process quality control method according to claim 4, characterized in that: The step of fusing the features in the first feature set and the features in the second feature set to obtain a multi-source fused feature set includes: The multi-source fusion feature set is obtained according to the first formula, wherein the first formula is: in, represents the multi-source fusion feature set, , F1 represents the first feature set, represents the i-th feature in the first feature set, and m represents the number of features in the first feature set; , F2 is the second feature set, represents the jth feature in the second feature set, and n represents the number of features in the second feature set; represents a two-dimensional weight coefficient matrix, Represents a function that maps a single feature in the first feature set, Indicates mapping of a single feature in the second feature set. represents the activation function, Indicates that a single feature is assigned to the first feature set The weight coefficient of the feature after mapping, Indicates that a single feature is assigned to the second feature set The weight coefficient of the feature after mapping, represents the feature interaction function.

6. The knitted fabric production process quality control method according to claim 1, characterized in that: The determining the structural type of the knitted fabric based on the first detection result of the first visual detection unit comprises: Performing feature extraction on the multiple images collected by the first visual inspection unit to obtain overall features and local features of the knitted fabric; The overall features and local features of the knitted fabric are input into a classification and recognition model to obtain the structural type of the knitted fabric, and the classification and recognition model is a trained neural network model.

7. The knitted fabric production process quality control method according to claim 1, characterized in that: The updating of control parameters in the knitted fabric production process based on the second detection result and the quality parameter of the knitted fabric comprises: Calculating a deviation value between the second detection result and the knitted fabric quality parameter, wherein the deviation value includes a numerical deviation value of the quality index and a defect quantity deviation value; According to the preset control strategy and the deviation value, the control parameters that need to be adjusted and the adjustment range are determined, and the corresponding control parameters of the knitted fabric production equipment are adjusted.

8. A knitted fabric production process quality control device, characterized in that: include: A first detection module, used for determining the structural type of the knitted fabric based on a first detection result of the first visual detection unit; An analysis module, configured to compare the structure type of the knitted fabric with a preset structure type of the knitted fabric to determine a second detection method of the knitted fabric; A second detection module, used for determining the quality parameter of the knitted fabric based on a second detection method, and performing quality detection on the knitted fabric based on the second detection method to obtain a second detection result; A parameter updating module is used to update control parameters in a knitted fabric production process based on the second detection result and the quality parameter of the knitted fabric.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images