Method and system for evaluating quality of sanding machine woven fabric combined with machine vision

Abstract

The application provides a method and system for evaluating the quality of brushed woven fabric combined with machine vision, and relates to the technical field of brushed woven fabric. The method comprises the following steps: introducing a multi-dimensional model for evaluating the quality of fabric, combining the fabric production demand and the fabric production image to perform multi-dimensional evaluation of the quality of fabric, and constructing a fabric quality evaluation atlas; self-adaptively adjusting the fabric production control scheme; performing optimization of the quality of fabric according to the multi-dimensional model for evaluating the quality of fabric; guiding the first fabric production adjustment domain to perform multiple reproduction optimizations according to the first adjustment guide population; and optimizing the control of the brushing machine according to the production control optimization result. The application solves the technical problem that the quality evaluation accuracy is not enough due to the lack of comprehensive evaluation of multiple dimensions in the prior art, which affects the stability of the quality of fabric. The self-adaptive adjustment of the production control scheme of the brushing machine is realized through the multi-dimensional model for evaluating the quality of fabric, and the stability of the quality of fabric is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of brushed woven fabric, and particularly relates to a brushed woven fabric quality evaluation method and system combining machine vision. BACKGROUND

[0002] In the traditional fabric production process, especially in the operation of the brushing machine, the fabric quality evaluation usually relies on a single-dimensional detection method, only focusing on the appearance quality, mechanical properties or comfort of the fabric, and lacks comprehensive evaluation of multiple performance dimensions of the fabric, which cannot fully reflect the performance of the fabric in actual use, and cannot achieve precise quality control when facing different production demands, thereby affecting the consistency and stability of the final product. Since the single-dimensional quality evaluation cannot identify the specific dimension of the problem, it cannot be targeted for optimization when adjusting the production parameters, resulting in fluctuations in product quality between different production batches, or even between the same batch, and lack of consistency and stability.

[0003] In summary, the prior art has the technical problem that the fabric quality evaluation relies on single-dimensional detection, lacks comprehensive evaluation of multiple dimensions, resulting in insufficient quality evaluation accuracy, and thereby affecting the stability of the fabric quality. SUMMARY

[0004] The purpose of the present application is to provide a brushed woven fabric quality evaluation method and system combining machine vision, to solve the technical problem in the prior art that the fabric quality evaluation relies on single-dimensional detection, lacks comprehensive evaluation of multiple dimensions, resulting in insufficient quality evaluation accuracy, and thereby affecting the stability of the fabric quality.

[0005] In view of the above problems, the present application provides a brushed woven fabric quality evaluation method and system combining machine vision.

[0006] In a first aspect, the application provides a method for evaluating the quality of a sanding woven fabric by combining machine vision, which is implemented by a sanding woven fabric quality evaluation system combining machine vision. The method includes the following steps: when a sanding machine executes a fabric production instruction, a fabric production image is synchronously acquired, the fabric production instruction including a fabric production control scheme corresponding to a fabric production requirement; a fabric quality evaluation multi-dimensional model is introduced, and a fabric quality multi-dimensional evaluation is performed in combination with the fabric production requirement and the fabric production image to construct a fabric quality evaluation atlas; the fabric production control scheme is adaptively adjusted according to the fabric quality evaluation atlas to obtain a first fabric production adjustment domain; the fabric quality evaluation multi-dimensional model is used to perform fabric quality evaluation optimization on the first fabric production adjustment domain to obtain a first adjustment guide population; based on the fabric quality evaluation multi-dimensional model, the first adjustment guide population is used to guide the first fabric production adjustment domain to perform multiple reproduction optimizations to obtain a production control optimization result; and the sanding machine is controlled based on the production control optimization result.

[0007] Optionally, when the sanding machine executes the fabric production instruction, a fabric production monitoring image is synchronously acquired according to a machine vision device; and the fabric production monitoring image is subjected to denoising processing to obtain the fabric production image.

[0008] Optionally, a multi-dimensional defect is captured from the fabric production image according to the fabric production requirement to obtain a fabric surface defect grid, a fabric texture defect grid and a fabric color defect grid; the fabric quality evaluation multi-dimensional model is activated, the fabric quality evaluation multi-dimensional model including a surface defect evaluation model, a texture defect evaluation model and a color defect evaluation model; the fabric surface defect grid is input into the surface defect evaluation model to obtain a surface defect evaluation coefficient; the fabric texture defect grid is input into the texture defect evaluation model to obtain a texture defect evaluation coefficient; and the fabric color defect grid is input into the color defect evaluation model to obtain a color defect evaluation coefficient; and the fabric surface defect grid, the fabric texture defect grid, the fabric color defect grid, the surface defect evaluation coefficient, the texture defect evaluation coefficient and the color defect evaluation coefficient are collated to obtain the fabric quality evaluation atlas.

[0009] Optionally, surface feature convolution is performed according to the fabric production image to obtain a fabric surface feature grid; qualified fabric image sample retrieval is performed according to the fabric production requirement to obtain a qualified fabric image set; surface feature convolution is performed according to the qualified fabric image set to obtain a plurality of qualified surface feature grids; feature aggregation is performed on the plurality of qualified surface feature grids to obtain a surface feature reference grid; and surface defect detection is performed on the fabric surface feature grid according to the surface feature reference grid to obtain the fabric surface defect grid.

[0010] Optionally, surface defect path tracing is performed on the fabric quality evaluation graph according to the fabric brushing machine monitoring data to obtain a fabric surface defect path; texture defect path tracing is performed on the fabric quality evaluation graph according to the fabric brushing machine monitoring data to obtain a fabric texture defect path; color defect path tracing is performed on the fabric quality evaluation graph according to the fabric brushing machine monitoring data to obtain a fabric color defect path; and correlation feature analysis is performed on the fabric production control scheme according to the fabric surface defect path, the fabric texture defect path, and the fabric color defect path, respectively, to construct a fabric defect control correlation graph; and multi-parameter adjustment is performed on the fabric production control scheme according to the fabric defect control correlation graph to obtain the first fabric production adjustment domain.

[0011] Optionally, a fabric production adjustment first scheme is extracted according to the first fabric production adjustment domain; fabric feature prediction is performed based on the fabric production adjustment first scheme to obtain a first fabric feature prediction result; a first fabric quality evaluation sequence is obtained according to the fabric quality evaluation multi-dimensional model based on the first fabric feature prediction result; and if the first fabric quality evaluation sequence satisfies a fabric quality evaluation constraint, the fabric production adjustment first scheme is added to the first adjustment guide population.

[0012] Optionally, the fabric quality evaluation constraint includes a surface defect evaluation constraint, a texture defect evaluation constraint, and a color defect evaluation constraint.

[0013] Optionally, based on the fabric quality evaluation multi-dimensional model, the first fabric production adjustment domain is bred and optimized according to the first adjustment guide population, to obtain a second adjustment guide population; based on the fabric quality evaluation multi-dimensional model, the first fabric production adjustment domain is continuously bred and optimized according to the second adjustment guide population, until a plurality of adjustment guide populations satisfying a predetermined number of breeding and optimization are obtained; the plurality of adjustment guide populations are calculated for fabric comprehensive quality according to a fabric quality evaluation multi-dimensional weight, to obtain a fabric comprehensive quality distribution; based on the fabric comprehensive quality distribution, the plurality of adjustment guide populations are optimized for comprehensive quality according to a fabric comprehensive quality threshold, to obtain a second fabric production adjustment domain; the production control optimization result is obtained by performing energy consumption minimization optimization according to the second fabric production adjustment domain.

[0014] Optionally, the first fabric production adjustment domain is identified for difference according to the first adjustment guide population, to obtain a first adjustment difference vector set; the first adjustment difference vector set is randomly mutated, to obtain a first adjustment mutation vector set; the first fabric production adjustment domain is mutated and bred according to the first adjustment mutation vector set, to obtain a first adjustment mutation breeding space; the first adjustment mutation breeding space is evaluated for fabric quality according to the fabric quality evaluation multi-dimensional model, to generate the second adjustment guide population.

[0015] In a second aspect, the application further provides a fabric quality evaluation system of a sanding machine combined with machine vision, for executing the fabric quality evaluation method of the sanding machine combined with machine vision as described in the first aspect, wherein the fabric quality evaluation system of the sanding machine combined with machine vision comprises: a fabric production data acquisition module, for synchronously acquiring a fabric production image when a sanding machine executes a fabric production instruction, the fabric production instruction comprising a fabric production control scheme corresponding to a fabric production requirement; a multi-dimensional evaluation module, for introducing a fabric quality evaluation multi-dimensional model, performing multi-dimensional evaluation of fabric quality in combination with the fabric production requirement and the fabric production image, and constructing a fabric quality evaluation atlas; an adaptive adjustment module, for performing adaptive adjustment of the fabric production control scheme according to the fabric quality evaluation atlas, to obtain a first fabric production adjustment domain; a quality evaluation optimization module, for performing evaluation of fabric quality of the first fabric production adjustment domain according to the fabric quality evaluation multi-dimensional model, to obtain a first adjustment guide population; a multi-time breeding optimization module, for guiding the first fabric production adjustment domain to perform multi-time breeding optimization according to the first adjustment guide population based on the fabric quality evaluation multi-dimensional model, to obtain a production control optimization result; and an optimization control module, for performing optimization control of the sanding machine according to the production control optimization result.

[0016] The one or more technical solutions provided in the application have at least the following beneficial effects: by synchronously acquiring a fabric production image when a sanding machine executes a fabric production instruction, the fabric production instruction includes a fabric production control scheme corresponding to a fabric production requirement; a fabric quality evaluation multidimensional model is introduced, and fabric quality multidimensional evaluation is performed in combination with the fabric production requirement and the fabric production image to construct a fabric quality evaluation atlas; the fabric production control scheme is adaptively adjusted according to the fabric quality evaluation atlas to obtain a first fabric production adjustment domain; the fabric quality evaluation multidimensional model is used to perform fabric quality evaluation optimization on the first fabric production adjustment domain to obtain a first adjustment guide population; based on the fabric quality evaluation multidimensional model, the first adjustment guide population is used to guide the first fabric production adjustment domain to perform multiple reproduction optimizations to obtain a production control optimization result; and the sanding machine is controlled according to the production control optimization result. That is, by using the fabric quality evaluation multidimensional model, multiple dimensions of the fabric are comprehensively considered, the fabric quality evaluation atlas is constructed, adaptive adjustment of the sanding machine production control scheme is realized, the production process can be dynamically adjusted, instability in the production process is avoided, the stability of the fabric quality is improved, and the production efficiency is improved.

[0017] The above description is only a summary of the technical solutions of the application. In order to more clearly understand the technical means of the application, the application can be implemented according to the content of the specification, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the application, nor is it intended to limit the scope of the application. Other features of the application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the application or the prior art, the following will briefly introduce the drawings needed in the embodiments or the prior art description. Obviously, the drawings in the following description are only exemplary, and those skilled in the art can obtain other drawings without creative labor on the basis of the provided drawings.

[0019] Figure 1 The flowchart of the fabric quality evaluation method of the sanding machine combined with machine vision of the application.

[0020] Figure 2 The structural schematic diagram of the fabric quality evaluation system of the sanding machine combined with machine vision of the application.

[0021] Explanation of reference signs: fabric production data acquisition module 11, multi-dimensional evaluation module 12, adaptive adjustment module 13, quality evaluation optimization module 14, multiple propagation optimization module 15, and optimization control module 16. DETAILED DESCRIPTION

[0022] The present application provides a fabric quality evaluation method and system combined with machine vision, which solves the technical problem in the prior art that the quality evaluation of the fabric depends on single-dimensional detection and lacks comprehensive evaluation of multiple dimensions, resulting in insufficient accuracy of the quality evaluation and affecting the stability of the fabric quality. Through the multi-dimensional model of the fabric quality evaluation, multiple dimensions of the fabric are comprehensively considered, a fabric quality evaluation atlas is constructed, adaptive adjustment of the production control scheme of the fabric is realized, the production process can be dynamically adjusted, the instability in the production process is avoided, the stability of the fabric quality is improved, and the production efficiency is improved.

[0023] The technical solutions in the present application will be described clearly and completely below with reference to the drawings. Obviously, the described embodiments are only some of the embodiments of the present application, not all embodiments of the present application. It should be understood that the present application is not limited by the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application. In addition, it should be noted that, for convenience of description, only parts related to the present application are shown in the drawings, not all.

[0024] Embodiment one, please refer to the attached Figure 1 The present application provides a fabric quality evaluation method combined with machine vision, wherein the fabric quality evaluation method combined with machine vision is executed by a fabric quality evaluation system combined with machine vision, and the fabric quality evaluation method combined with machine vision specifically includes the following steps:

[0025] When the fabric production instruction is executed by the fabric raising machine, a fabric production image is synchronously acquired, and the fabric production instruction includes a fabric production control scheme corresponding to a fabric production requirement.

[0026] Further, the present application further includes the following steps: when the fabric production instruction is executed by the fabric raising machine, a fabric production monitoring image is synchronously acquired according to a machine vision device; and the fabric production monitoring image is subjected to denoising processing to obtain the fabric production image.

[0027] Specifically, when the production process starts, the sanding machine executes corresponding operations according to preset fabric production instructions. The fabric production instructions include a production control scheme formulated according to fabric production requirements. That is, when processing the fabric, the sanding machine adjusts parameters such as speed, pressure, and grinding head tension according to the fabric production instructions to achieve the target quality requirements. While the sanding machine executes the production instructions, the machine vision device installed on the sanding machine synchronously collects production monitoring images of the fabric. The machine vision device usually uses a high-resolution industrial camera to record the real-time state of the fabric in the production process, including information such as fabric surface defects, sanding effects, and pile distribution, by capturing image data of the fabric surface in real time. The machine vision device is a device based on camera and image processing technology, usually used to automatically acquire and analyze image data in the production process, capable of image capture, processing, and analysis, helping to achieve automatic detection, quality assessment, and other functions. The fabric production monitoring image refers to the image obtained by the machine vision device, reflecting the actual state of the fabric at each moment in the production process, including surface quality, structure, defects, and other information.

[0028] Since there may be noise in the image acquisition process, such as uneven lighting, lens flaws, or camera noise, it is necessary to perform denoising processing on the collected fabric production monitoring images to remove unnecessary interference and ensure that useful information in the fabric production monitoring images, such as fabric surface defects and pile distribution, is clearly visible. For example, Gaussian filtering or median filtering algorithms are used to denoise the image, removing 5% of noise points and improving the quality of the fabric production monitoring image. For example, the machine vision device collects fabric production images, each image is 1024*1024 pixels in size, the image resolution is 0.1 millimeters per pixel, an image is collected every 30 seconds, and Gaussian filtering is used for denoising. After denoising, the noise of the image is reduced to 3%, ensuring that the edges of the image are clear and facilitating subsequent processing.

[0029] The fabric production instructions include a fabric production control scheme corresponding to the fabric production requirements. The production control scheme sets corresponding machine parameters according to the quality requirements of the target fabric, ensuring that the fabric always maintains within the predetermined quality standard range during the production process.

[0030] A multi-dimensional fabric quality evaluation model is introduced, and a multi-dimensional fabric quality evaluation is performed in combination with the fabric production requirements and the fabric production images to construct a fabric quality evaluation atlas.

[0031] Further, the application further comprises the following steps: performing multi-dimensional defect capture on the fabric production image according to the fabric production requirements to obtain a fabric surface defect grid, a fabric texture defect grid, and a fabric color defect grid; activating the fabric quality evaluation multi-dimensional model, wherein the fabric quality evaluation multi-dimensional model comprises a surface defect evaluation model, a texture defect evaluation model, and a color defect evaluation model; inputting the fabric surface defect grid into the surface defect evaluation model to obtain a surface defect evaluation coefficient; inputting the fabric texture defect grid into the texture defect evaluation model to obtain a texture defect evaluation coefficient; inputting the fabric color defect grid into the color defect evaluation model to obtain a color defect evaluation coefficient; and collating the fabric surface defect grid, the fabric texture defect grid, the fabric color defect grid, the surface defect evaluation coefficient, the texture defect evaluation coefficient, and the color defect evaluation coefficient to obtain the fabric quality evaluation atlas.

[0032] Further, the application further comprises the following steps: performing surface feature convolution on the fabric production image to obtain a fabric surface feature grid; performing qualified fabric image sample retrieval according to the fabric production requirements to obtain a qualified fabric image set; performing surface feature convolution on the qualified fabric image set to obtain a plurality of qualified surface feature grids; performing feature aggregation on the plurality of qualified surface feature grids to obtain a surface feature reference grid; and performing surface defect detection on the fabric surface feature grid according to the surface feature reference grid to obtain the fabric surface defect grid.

[0033] Specifically, surface feature convolution is performed on the fabric production image, details are extracted from each part of the image through a sliding window method, information such as texture, nap distribution, and surface unevenness of the fabric is captured, a fabric surface feature grid is formed, and feature information of a specific area of the fabric surface is reflected. The convolution operation decomposes the image into a plurality of small areas, extracts local features such as texture, nap distribution, and surface flatness through a sliding convolution kernel, and integrates the features into a surface feature grid of the fabric through a convolution network, and each grid cell represents a part of the surface features of the fabric.

[0034] According to the fabric production requirements, qualified fabric image samples that meet the standards of the fabric production requirements are retrieved from an existing image library to obtain a qualified fabric image set, and fabric features that meet the quality standards are displayed. The qualified fabric image set is a collection of images that meet the production requirements filtered from the existing fabric image library, and displays the fabric surface that meets the quality standards. The same surface feature convolution is performed on the retrieved qualified fabric image set, the surface features of the qualified fabric image set are extracted, and a plurality of qualified surface feature grids are generated, and the feature standards of ideal fabrics are displayed.

[0035] The feature aggregation is performed on the plurality of qualified surface feature grids to combine them into a surface feature reference grid, which integrates all the standard fabric features in the qualified images as a standard reference for subsequent fabric quality detection.

[0036] According to the generated surface feature reference grid, the fabric surface feature grid in actual production is compared to detect whether there is a defect on the fabric surface. Through difference analysis, the part of the fabric surface that does not meet the standard, such as uneven pile, scratch, color difference, etc., is identified, and a fabric surface defect grid is generated to record the defect type and location of the fabric surface. For example, if the pile density of a part of the fabric is 6 pile / cm2, which is lower than the standard 8 pile / cm2, the corresponding defect grid is generated to indicate the uneven pile problem of this part.

[0037] In addition to surface defects, texture defects and color defects are also captured according to the fabric production requirements to obtain a fabric texture defect grid and a fabric color defect grid, respectively. The texture defect grid captures the unevenness of the fabric surface texture, and the color defect grid detects the inconsistency of the fabric color. For example, if a part of the fabric has color difference, such as color deviation exceeding the standard range, the color defect grid will record this information and show the color difference position and size; similarly, if some parts of the fabric have uneven texture, the texture defect grid will also record the corresponding information.

[0038] The fabric quality evaluation multidimensional model is a machine learning model that comprehensively considers various aspects of fabric quality, including a surface defect evaluation model, a texture defect evaluation model, and a color defect evaluation model. The surface defect evaluation model detects whether there is a defect on the fabric surface, such as scratch, uneven pile, stain, etc., by analyzing the surface features in the fabric image; the texture defect evaluation model is used to identify whether the fabric texture is uniform and regular; and the color defect evaluation model detects whether there is a color-related quality problem in the fabric through the color information of the image, including color difference, uneven dyeing, fading, etc.

[0039] To build a surface defect evaluation model, a large number of fabric surface image sample data are collected, and the defect types in each image, such as scratches, color differences, uneven fuzz, etc., as well as the location and area of the defects, are labeled. The image data is standardized to ensure uniform size, and the image data is normalized so that the pixel value range is between 0 and 1. For example, assume there are 2000 fabric surface images, of which 500 contain scratch defects, 500 contain uneven fuzz defects, and the remaining 1000 are defect-free images. Each image is 256x256 pixels in size. All images are labeled, defining different types of defect categories and assigning corresponding labels to each image. For example, scratch defect images are labeled as 1, defect-free images are labeled as 0, and uneven fuzz images are labeled as 2. The processed fabric surface image sample data is divided into training set and validation set, usually 80% training set and 20% validation set. Select an appropriate convolutional neural network architecture, such as a CNN model, including input layer, convolutional layer, fully connected layer, and output layer. The input layer receives preprocessed image data, i.e., the pixel values of each image. Multiple convolutional layers are used to extract local features of the image, with a convolution kernel size of 3x3 or 5x5 and a stride of 1. Each convolutional layer is followed by a pooling layer, such as a max pooling layer, to reduce the size of the feature map and reduce the amount of computation. The flattened feature map is fed into the fully connected layer to abstract high-level features, and the final output is the classification information of the defect. After flattening, pass through 2 fully connected layers, each with 256 neurons. The output layer classifies the surface defects according to their types. The softmax activation function is used to output the probability values of each defect category.

[0040] The labeled training set is input into the CNN model, and the backpropagation algorithm is used for training. During the training process, the model learns the characteristics of different defects and adjusts the network parameters. For example, through 10 epochs of training, the model can predict the type of defect through the input image. After training, the model is evaluated using the validation set, and the accuracy, recall rate, F1-score, and other indicators are calculated. According to the evaluation results, the model parameters are adjusted or different network architectures are tried to improve the model performance. For example, the model accuracy is 85%, the recall rate is 80%, and the F1-score is 82%. If the results do not meet the requirements, more training data or deeper network structures need to be used for training. Through further optimization of the model, such as adjusting the number of convolution layers, adding data augmentation, etc., the test set accuracy is improved to 90%. The trained surface defect evaluation model is deployed in the actual production environment, and real-time images collected by the machine vision device during the fabric production process are received, and the surface defects in the images are automatically detected. The input of the surface defect evaluation model is the real-time collected image during the fabric production process, i.e. the fabric surface defect grid; according to the data of the grid, the type and severity of the defect are evaluated, and the surface defect evaluation coefficient is output. The surface defect evaluation coefficient is used to quantify the severity of the fabric surface defect, which is usually a value between 0 and 1, and the smaller the value, the more serious the defect and the worse the quality. A coefficient higher than 0.8 may indicate a relatively minor defect, while a coefficient lower than 0.2 may indicate a more serious defect.

[0041] The training process of the texture defect evaluation model and the color defect evaluation model is the same. The input of the texture defect evaluation model is the fabric texture defect grid of the fabric, which focuses on the uniformity and regularity of the texture. The image resolution is usually 1024x1024, and the image may contain uneven texture distribution, texture distortion, etc. Similar to the surface defect model, the image needs to be denoised, standardized before input, and may be subjected to texture feature extraction to enhance the model's perception of texture details. A numerical value is output to represent the severity of the fabric texture defect, usually between 0 and 1, and the smaller the value, the more serious the texture defect. For example, 0.7 indicates that the texture defect is relatively small, and there may be some uneven texture areas. The training process uses a regression loss function, such as mean square error.

[0042] The input of the color defect evaluation model is a fabric color defect grid, focusing on the color uniformity and color difference of the fabric. The image resolution is usually 1024x1024, and there may be color difference, uneven dyeing, etc. in the image. Color information needs to be converted in color space, such as from RGB to HSV or Lab, so that the model can better understand and evaluate the color difference. The output color defect evaluation coefficient quantifies the severity of fabric color defects, which is usually a value between 0 and 1, and the smaller the value, the more serious the color difference problem. For example, 0.25 indicates that the fabric has obvious color difference, which will affect its appearance. If the specific location of the color defect needs to be provided, the color difference area in the fabric will be marked. The CNN model is used, and the training process uses a regression loss function or a cross-entropy loss function.

[0043] The fabric surface defect grid is input into the surface defect evaluation model, which records the location and type of surface defect area, including defect information of each area on the fabric surface, such as scratches, stains, or uneven pile. The surface defect evaluation model learns the characteristics of the fabric surface defects and evaluates the severity of the defects according to the characteristics of the defect area. Through the processing of the surface defect evaluation model, a surface defect evaluation coefficient is generated, which is a value between 0 and 1, reflecting the severity of the fabric surface defects. A coefficient higher than 0.8 indicates that the defect is relatively minor, while a coefficient lower than 0.2 indicates that the defect is relatively serious.

[0044] The fabric texture defect grid is input into the texture defect evaluation model, which records the area of uneven or distorted texture, including the distribution of fabric surface texture, texture uniformity, and whether there are irregular textures such as texture distortion and fracture. The texture defect evaluation model analyzes the input grid and identifies abnormal or defective areas of the fabric texture. By comparing the differences between the actual texture and the ideal texture, the model evaluates the severity of the texture defect. The texture defect evaluation model outputs a texture defect evaluation coefficient, which quantifies the severity of the fabric texture defect, and the smaller the value, the more serious the texture defect.

[0045] The fabric color defect grid is input into the color defect evaluation model, which records the color difference or color unevenness information of different areas on the fabric surface, checks for color difference, color unevenness, or color variation. The color defect evaluation model analyzes the input color grid and evaluates whether there are color difference, color unevenness, etc. on the fabric surface, and judges the severity of the color defect by learning the color pattern of the fabric. The color defect evaluation model outputs a color defect evaluation coefficient, which represents the degree of fabric color defect. The smaller the value, the more serious the color difference problem.

[0046] The fabric surface defect grid, fabric texture defect grid, and fabric color defect grid of the fabric are integrated with the corresponding evaluation coefficients, i.e., surface defect evaluation coefficient, texture defect evaluation coefficient, and color defect evaluation coefficient, to form a comprehensive quality evaluation model. Each dimension of defect information is recorded in detail to obtain a fabric quality evaluation atlas that contains defect information of each dimension of the fabric surface, texture, and color, including the type, location, and evaluation coefficient of each defect. For example, a certain area shows that the surface defect scratch is relatively light, with an evaluation coefficient of 0.9, the texture defect is uneven, with an evaluation coefficient of 0.4, and the color defect is color difference, with an evaluation coefficient of 0.2.

[0047] Through the multi-dimensional defect capture and quality evaluation model, the quality of the fabric is comprehensively evaluated from multiple dimensions, ensuring that any quality problem can be discovered in a timely manner. By generating the fabric quality evaluation atlas, the quality of the fabric can be fed back in real time, and the production process can be adjusted in a timely manner to avoid the generation of unqualified products and improve the quality stability of the fabric.

[0048] According to the fabric quality evaluation atlas, the fabric production control scheme is adaptively adjusted to obtain a first fabric production adjustment domain.

[0049] Further, the application further includes the following steps: according to the fabric quality evaluation atlas, the surface defect path of the fabric is traced according to the fabric quality evaluation atlas, the texture defect path of the fabric is traced according to the fabric quality evaluation atlas, and the color defect path of the fabric is traced according to the fabric quality evaluation atlas; the fabric defect control correlation atlas is constructed by correlating the fabric production control scheme according to the fabric surface defect path, the fabric texture defect path, and the fabric color defect path; and the fabric production control scheme is adjusted according to the fabric defect control correlation atlas to obtain the first fabric production adjustment domain.

[0050] Specifically, the sanding machine monitoring data, i.e. the data collected in real time by the sanding machine equipment, including the equipment operating parameters and the processing state of the fabric, such as the speed of the sanding machine, the pressure, the degree of sanding wear, the tension of the fabric, the production speed and other parameters, reflect the processing state of the fabric in the production process, and further affect the quality of the fabric. Through the sanding machine monitoring data, the surface defect path of the fabric quality evaluation map is traced back, i.e. the formation process and possible causes of the surface defects of the fabric are traced back. The purpose of defect path tracing is to find the specific causes of the defects and to provide a basis for subsequent production adjustment. For example, if there are obvious defects on the surface of the fabric in a certain area, such as scratches or stains, the correlation between the production data and the quality evaluation results of that area is analyzed. Assuming that the surface defect evaluation coefficient of a certain area is 0.4, indicating that there are obvious surface scratches in that area, by comparing the sanding machine data of that area, it is found that the sanding machine speed of that area is too high or the pressure is too large, thus inferring the cause of the defect and finding the path leading to the surface defect. Through path tracing, the generated fabric surface defect path will include the specific factors and path leading to the defect. For example, assuming that the sanding machine has a speed setting of 1500 rpm and a pressure of 120 N in a certain production process, and the fabric quality evaluation map indicates that the surface defect evaluation coefficient of a certain area is 0.4, indicating that there are scratch problems in that area; the sanding machine speed of that area is 1500 rpm, which is consistent with the known relationship between high speed and surface defects, so it is considered that high speed is the main cause of surface scratches, and it is recommended to reduce the speed to 1200 rpm to reduce the occurrence of similar defects.

[0051] For texture defect path tracing and color defect path tracing, the fabric quality evaluation map is traced back according to the sanding machine monitoring data, the specific causes of the defects are determined, and the fabric texture defect path and the fabric color defect path are obtained. That is, the texture defect path tracing analyzes the formation reasons of uneven fabric texture or other texture defects, and can trace the source of the texture defects. The formation process of uneven fabric color, color difference and other problems is analyzed, and by combining the sanding machine monitoring data, the source of the color defect and related production factors such as dyeing process, fabric tension, etc. are traced back.

[0052] The fabric surface defect path, the fabric texture defect path, and the fabric color defect path are associated with the fabric production control scheme to construct a fabric defect control correlation map that shows the relationship between different defects and production parameters. The fabric defect control correlation map associates the defect paths of the fabric with the production control scheme for analyzing and optimizing the production process. By analyzing the fabric defect control correlation map, it is determined which production parameters need to be adjusted to reduce a specific type of defect. For example, if the fabric defect control correlation map shows that high rotation speed leads to surface scratches, the rotation speed of the sanding machine is reduced, or the fabric surface quality is optimized by adjusting the pressure of the sanding machine. For texture defects and color defects, the fabric tension, production speed, or sanding wear state is adjusted to improve the fabric quality. By adjusting multiple parameters of the fabric production control scheme, a first fabric production adjustment domain is obtained, which includes multiple fabric production adjustment schemes, and the fabric quality is optimized by adjusting the production parameters. For example, the abrasive type, speed setting, and temperature control of the sanding machine are adjusted to improve the fabric quality. By tracing the path of the fabric defect and constructing the fabric defect control correlation map, the problems in the production process are accurately identified, and precise parameter adjustment is performed.

[0053] According to the fabric quality evaluation multi-dimensional model, the first fabric production adjustment domain is subjected to fabric quality evaluation optimization to obtain a first adjustment guide population.

[0054] Further, the application further includes the following steps: extracting a fabric production adjustment first scheme according to the first fabric production adjustment domain; performing fabric feature prediction based on the fabric production adjustment first scheme to obtain a first fabric feature prediction result; obtaining a first fabric quality evaluation sequence according to the fabric quality evaluation multi-dimensional model based on the first fabric feature prediction result; and if the first fabric quality evaluation sequence satisfies a fabric quality evaluation constraint, adding the fabric production adjustment first scheme to the first adjustment guide population.

[0055] Further, the application further includes the following steps: the fabric quality evaluation constraint includes a surface defect evaluation constraint, a texture defect evaluation constraint, and a color defect evaluation constraint.

[0056] Specifically, any fabric production adjustment first scheme is extracted from the first fabric production adjustment domain, which contains specific production parameter settings, such as the speed of the sanding machine, the pressure, the temperature, etc. Fabric feature prediction is performed according to the fabric production adjustment first scheme, and a first fabric feature prediction result is obtained, including surface defect feature prediction, texture defect feature prediction, and color defect feature prediction. The first fabric feature prediction result is evaluated by the fabric quality evaluation multidimensional model to obtain a first fabric quality evaluation sequence. That is, the surface defect feature prediction is input into the surface defect evaluation model to obtain a surface defect evaluation coefficient, the texture defect feature prediction is input into the surface defect evaluation model to obtain a texture defect evaluation coefficient, and the color defect feature prediction is input into the surface defect evaluation model to obtain a color defect evaluation coefficient. The surface defect evaluation coefficient, the texture defect evaluation coefficient, and the color defect evaluation coefficient together constitute the first fabric quality evaluation sequence. For example, the surface defect evaluation coefficient is 0.9, the texture defect evaluation coefficient is 0.6, and the color defect evaluation coefficient is 0.8.

[0057] The fabric quality evaluation constraint is a set of limiting conditions for the fabric quality evaluation result, including surface defect evaluation constraints, texture defect evaluation constraints, and color defect evaluation constraints, which are used to ensure that the produced fabric meets the quality standards. The surface defect evaluation constraint specifies the maximum tolerance of fabric surface defects, the texture defect evaluation constraint specifies the maximum tolerance of fabric texture defects, and the color defect evaluation constraint specifies the maximum tolerance of fabric color defects. For example, the fabric quality evaluation constraint includes that the surface defect evaluation coefficient needs to be greater than 0.7, the texture defect evaluation coefficient needs to be greater than 0.5, and the color defect evaluation coefficient needs to be greater than 0.6.

[0058] According to the fabric quality evaluation constraint, it is checked whether the first fabric quality evaluation sequence meets the fabric quality evaluation constraint. If it meets, the fabric production adjustment first scheme corresponding to the first fabric quality evaluation sequence is added to the first adjustment guide population. For other fabric production adjustment schemes in the first fabric production adjustment domain, the foregoing steps are also performed, and finally all fabric production adjustment schemes that meet the fabric quality evaluation constraint are added to the first adjustment guide population.

[0059] Exemplarily, it is assumed that the first fabric production adjustment domain includes three adjustment schemes: adjustment scheme 1 with a rotation speed of 1200 rpm, a pressure of 100 N, and a fabric tension of 12 N; adjustment scheme 2 with a rotation speed of 1400 rpm, a pressure of 120 N, and a fabric tension of 14 N; and adjustment scheme 3 with a rotation speed of 1600 rpm, a pressure of 130 N, and a fabric tension of 15 N. The fabric feature prediction for adjustment scheme 1 is that the surface has almost no defects; the texture prediction for adjustment scheme 1 is that the texture uniformity is moderate and there is a certain unevenness; and the color prediction for adjustment scheme 1 is that the color consistency is good but there is a slight color difference. The first fabric feature prediction result includes surface defect feature prediction, texture defect feature prediction, and color defect feature prediction. The first fabric feature prediction result is input into the fabric quality evaluation multidimensional model to calculate the evaluation coefficients of the surface, texture, and color defects, respectively, to obtain a surface defect evaluation coefficient of 0.9, a texture defect evaluation coefficient of 0.6, and a color defect evaluation coefficient of 0.8. The fabric quality evaluation constraints are that the surface defect evaluation coefficient is greater than 0.7, the texture defect evaluation coefficient is greater than 0.5, and the color defect evaluation coefficient is greater than 0.6. The first fabric quality evaluation sequence is compared with the quality evaluation constraints: the surface defect evaluation coefficient 0.9 is greater than 0.7, which meets the constraint; the texture defect evaluation coefficient 0.6 is greater than 0.5, which meets the constraint; and the color defect evaluation coefficient 0.8 is greater than 0.6, which meets the constraint. Since the first fabric quality evaluation sequence meets all the quality evaluation constraints, this scheme meets the quality requirements. The first fabric production adjustment scheme (rotation speed of 1200 rpm, pressure of 100 N, and fabric tension of 12 N) is added to the first adjustment guide population. The foregoing steps are repeated to obtain a surface defect evaluation coefficient of 0.8, a texture defect evaluation coefficient of 0.7, and a color defect evaluation coefficient of 0.9 for adjustment scheme 2; and a surface defect evaluation coefficient of 0.6, a texture defect evaluation coefficient of 0.4, and a color defect evaluation coefficient of 0.7 for adjustment scheme 3. The quality evaluation sequence of adjustment scheme 2 meets the fabric quality evaluation constraints, and adjustment scheme 2 is added to the first adjustment guide population. Since the quality evaluation sequence of adjustment scheme 3 does not meet the surface defect evaluation constraint and the texture defect evaluation constraint, adjustment scheme 3 is not added to the first adjustment guide population.

[0060] Through the adjustment scheme prediction based on the quality evaluation, the quality characteristics of each fabric can be accurately controlled to ensure that each link in the production process meets the best quality standards. It is ensured that each adjustment scheme meets the quality evaluation constraints, maximally reduces fabric defects, and improves the stability and consistency of fabric quality.

[0061] Based on the fabric quality evaluation multidimensional model, the first fabric production adjustment domain is guided to perform multiple reproduction optimizations according to the first adjustment guide population to obtain a production control optimization result.

[0062] Further, the application further comprises the following steps: based on the fabric quality evaluation multi-dimensional model, breeding optimization is performed on the first fabric production adjustment domain according to the first adjustment guide population to obtain a second adjustment guide population; based on the fabric quality evaluation multi-dimensional model, breeding optimization is continuously performed on the first fabric production adjustment domain according to the second adjustment guide population until a plurality of adjustment guide populations satisfying a predetermined breeding optimization number of times are obtained; fabric comprehensive quality calculation is performed on the plurality of adjustment guide populations according to a fabric quality evaluation multi-dimensional weight to obtain a fabric comprehensive quality distribution; based on the fabric comprehensive quality distribution, comprehensive quality optimization is performed on the plurality of adjustment guide populations according to a fabric comprehensive quality threshold to obtain a second fabric production adjustment domain; energy consumption minimization optimization is performed according to the second fabric production adjustment domain to obtain the production control optimization result.

[0063] Further, the application further comprises the following steps: difference identification is performed on the first fabric production adjustment domain according to the first adjustment guide population to obtain a first adjustment difference vector set; random variation is performed according to the first adjustment difference vector set to obtain a first adjustment variation vector set; variation breeding is performed on the first fabric production adjustment domain according to the first adjustment variation vector set to obtain a first adjustment variation breeding space; fabric quality evaluation optimization is performed on the first adjustment variation breeding space according to the fabric quality evaluation multi-dimensional model to generate the second adjustment guide population.

[0064] Specifically, the first adjustment guide population contains a plurality of adjustment schemes conforming to fabric quality evaluation constraints. The first fabric production adjustment domain includes a plurality of adjustment schemes obtained after multi-parameter adjustment of fabric production control schemes. By comparing each adjustment scheme in the first adjustment guide population with the adjustment schemes in the first fabric production adjustment domain, adjustment schemes with significant differences can be identified, which can be differences in parameters or differences in quality evaluation results. The first adjustment difference vector set is a vector set obtained through difference identification, each vector representing the difference between an adjustment scheme and other schemes, and being able to represent information of adjustment scheme diversity.

[0065] According to the first adjustment difference vector set, random variation is performed to change the parameters of part of the adjustment schemes, such as randomly adjusting the values of production parameters such as rotating speed, pressure, fabric tension, etc., to generate a new adjustment vector, and a first adjustment variation vector set is generated. Random variation generates a new scheme by randomly adjusting certain parameters of the adjustment scheme, which is used to explore new production adjustment schemes to improve fabric quality. According to the first adjustment variation vector set, the first fabric production adjustment domain is varied and propagated to combine and expand the adjustment schemes in the first fabric production adjustment domain to generate a plurality of new schemes. By combining different variation vectors in the first adjustment variation vector set, a new adjustment scheme is formed. For example, the parameters of adjustment scheme A and adjustment scheme B can be combined through the propagation operation to generate a new adjustment scheme C, such as the rotating speed of adjustment scheme A being 1500 rpm, the pressure being 100 N, and the tension being 12 N, the rotating speed of adjustment scheme B being 1300 rpm, the pressure being 110 N, and the tension being 14 N, and after variation and propagation, the rotating speed of the new adjustment scheme C is 1400 rpm, the pressure is 105 N, and the tension is 13 N. The first adjustment variation propagation space contains all possible adjustment schemes, representing new production parameter combinations generated from the variation and propagation process. The schemes in the first adjustment variation propagation space are the schemes obtained after variation and propagation of the schemes in the first adjustment variation vector set in a better direction.

[0066] According to the fabric quality evaluation multi-dimensional model, the first adjustment variation propagation space is subjected to fabric quality evaluation optimization, i.e., fabric feature prediction is performed on each scheme in the first adjustment variation propagation space to obtain corresponding surface defect feature prediction, texture defect feature prediction, and color defect feature prediction. The fabric quality evaluation multi-dimensional model is used to evaluate the surface defect feature prediction, the texture defect feature prediction, and the color defect feature prediction to obtain the corresponding surface defect evaluation coefficient, the texture defect evaluation coefficient, and the color defect evaluation coefficient of each scheme. It is determined whether the surface defect evaluation constraint, the texture defect evaluation constraint, and the color defect evaluation constraint in the fabric quality evaluation constraint are satisfied. If so, it is added to the second adjustment guide population. The second adjustment guide population is a new population generated after quality evaluation optimization, which contains better adjustment schemes. Through difference identification, random variation, and propagation operation, diversified adjustment schemes are generated. The new schemes generated are subjected to quality evaluation, and adjustment schemes that satisfy the fabric quality evaluation constraint are selected to ensure that each adjustment scheme meets the quality requirements, thereby improving the consistency and quality stability of the final product.

[0067] According to the second adjustment guide population, the breeding optimization is continued on the first fabric production adjustment domain, and the adjusted scheme after breeding is evaluated in quality by the fabric quality evaluation multi-dimensional model, and the foregoing process is repeatedly until a predetermined breeding optimization number is met. The predetermined breeding optimization number is a preset breeding number in the optimization process, and is usually to ensure that enough adjusted schemes are generated, and through multiple breeding optimizations, the most suitable adjusted scheme is finally found. Multiple breeding and optimization are performed according to the predetermined breeding optimization number. After each breeding, a new adjusted scheme is generated and evaluated in quality until a preset breeding number, such as 10 times, is reached. Each breeding generates a new adjusted scheme, which is screened and optimized by the fabric quality evaluation multi-dimensional model, and a better adjusted scheme is gradually screened out, and finally multiple adjusted guide populations are formed. The multiple adjusted guide populations are a set of optimal adjusted schemes based on quality evaluation in the production adjustment process.

[0068] In the fabric quality evaluation, each dimension has a different weight on the final quality, and a fabric quality evaluation multi-dimensional weight is obtained. The higher the weight value, the greater the contribution of the dimension to the overall quality of the fabric. Specifically, the fabric quality evaluation multi-dimensional weight includes a surface defect evaluation weight, a texture defect evaluation weight and a color defect evaluation weight. The fabric overall quality of the multiple adjusted guide populations is calculated to obtain multiple fabric overall quality coefficients, which constitute a fabric overall quality distribution. The fabric overall quality coefficient of each adjusted scheme is calculated using the following formula: fabric overall quality coefficient = surface defect evaluation coefficient x surface defect evaluation weight + texture defect evaluation coefficient x texture defect evaluation weight + color defect evaluation coefficient x color defect evaluation weight. The fabric overall quality coefficient of each adjusted scheme reflects its overall performance in all quality dimensions.

[0069] The fabric comprehensive quality threshold is a preset quality standard value, representing the quality requirements of the fabric in various dimensions. Only the adjustment scheme that exceeds the fabric comprehensive quality threshold is considered to meet the quality requirements. The multiple fabric comprehensive quality coefficients corresponding to the multiple adjustment guide populations in the fabric comprehensive quality distribution are compared with the fabric comprehensive quality threshold. If it is greater than or equal to the fabric comprehensive quality threshold, the scheme meets the quality requirements, and it is added to the second fabric production adjustment domain. The second fabric production adjustment domain is optimized for minimum energy consumption, the energy consumption of each scheme is evaluated, and the adjustment scheme with the minimum energy consumption is found, obtaining the production control optimization result. For example, in a certain example, the preset surface defect evaluation weight, texture defect evaluation weight and color defect evaluation weight are 0.4, 0.3 and 0.3 respectively; the surface defect evaluation coefficient of the adjustment guide population 1 is 0.9, the texture defect evaluation coefficient is 0.6, and the color defect evaluation coefficient is 0.7, and the comprehensive quality coefficient is 0.75; the rotational speed of the adjustment guide population 2 is 1500 rpm, the pressure is 120 N, the fabric tension is 14 N, the surface defect evaluation coefficient is 0.7, the texture defect evaluation coefficient is 0.6, and the color defect evaluation coefficient is 0.5, and the comprehensive quality coefficient is 0.61; the rotational speed of the adjustment guide population 3 is 1700 rpm, the pressure is 140 N, the fabric tension is 15 N, the surface defect evaluation coefficient is 0.5, the texture defect evaluation coefficient is 0.7, and the color defect evaluation coefficient is 0.4, and the comprehensive quality coefficient is 0.53. Assuming that the fabric comprehensive quality threshold is 0.6, the adjustment guide population 1 and the adjustment guide population 2 both meet the quality requirements and enter the second fabric production adjustment domain. Assuming that the energy consumption of the adjustment guide population 1 is 80 kWh and the energy consumption of the adjustment guide population 2 is 90 kWh, the adjustment guide population 1 is selected as the production control optimization result. Through comprehensive quality evaluation and energy optimization, it is ensured that the finally selected adjustment scheme not only meets the quality requirements, but also maximally reduces the energy consumption and improves the production efficiency.

[0070] According to the production control optimization result, the fuzzing machine is optimized and controlled.

[0071] Specifically, according to the production control optimization result, the parameters of the sanding machine are adjusted, including the speed, pressure, sanding degree, fabric tension, production speed, etc. The sanding machine produces fabric according to the modified operating parameters, while monitoring the fabric production process in real time, and adjusting the parameters according to real-time feedback data, such as fabric surface quality, texture uniformity, etc., to ensure that the quality of the fabric in each production cycle always meets the requirements. If real-time feedback indicates that production parameters need to be adjusted, such as due to equipment wear or environmental changes, dynamically adjust the operating parameters of the sanding machine to ensure continuous optimization of quality and energy efficiency. Based on production monitoring data, adaptive adjustments are made according to the actual operating state of the sanding machine, such as adjusting the speed to adapt to different characteristics of the fabric or production conditions. Through the optimized control of the sanding machine, the surface quality, texture and color of the fabric are ensured to meet the predetermined standards. Instead of relying on human experience during production, dynamic adjustments are made based on data, thereby improving the consistency and stability of fabric quality. Under the premise of ensuring quality, energy consumption is minimized through optimization and control, reducing energy consumption and production costs during production.

[0072] In summary, the sanding machine fabric quality evaluation method combined with machine vision provided by the present application has the following beneficial effects: by synchronously acquiring fabric production images when the sanding machine executes fabric production instructions, the fabric production instructions include a fabric production control scheme corresponding to fabric production requirements; a fabric quality evaluation multi-dimensional model is introduced to conduct multi-dimensional fabric quality evaluation in combination with the fabric production requirements and the fabric production images, and a fabric quality evaluation atlas is constructed; the fabric production control scheme is adaptively adjusted according to the fabric quality evaluation atlas to obtain a first fabric production adjustment domain; the first fabric production adjustment domain is evaluated for fabric quality optimization according to the fabric quality evaluation multi-dimensional model to obtain a first adjustment guide population; based on the fabric quality evaluation multi-dimensional model, the first fabric production adjustment domain is guided to reproduce and optimize multiple times according to the first adjustment guide population to obtain a production control optimization result; and the sanding machine is controlled according to the production control optimization result. That is, through the fabric quality evaluation multi-dimensional model, multiple dimensions of the fabric are considered comprehensively, a fabric quality evaluation atlas is constructed, adaptive adjustment of the sanding machine production control scheme is realized, the production process can be dynamically adjusted, instability in the production process is avoided, the stability of fabric quality is improved, and production efficiency is thus improved.

[0073] Embodiment Two, based on the same inventive concept as the sanding machine fabric quality evaluation method combined with machine vision in the aforementioned Embodiment One, the present application also provides a sanding machine fabric quality evaluation system combined with machine vision, please refer to the accompanying drawings Figure 2 The sanding machine fabric quality evaluation system combined with machine vision includes:

[0074] The fabric production data acquisition module 11 is configured to synchronously acquire a fabric production image when the sanding machine executes a fabric production instruction, the fabric production instruction including a fabric production control scheme corresponding to a fabric production requirement; the multi-dimensional evaluation module 12 is configured to introduce a fabric quality evaluation multi-dimensional model, perform fabric quality multi-dimensional evaluation in combination with the fabric production requirement and the fabric production image, and construct a fabric quality evaluation atlas; the adaptive adjustment module 13 is configured to perform adaptive adjustment on the fabric production control scheme according to the fabric quality evaluation atlas, and obtain a first fabric production adjustment domain; the quality evaluation optimization module 14 is configured to perform fabric quality evaluation optimization on the first fabric production adjustment domain according to the fabric quality evaluation multi-dimensional model, and obtain a first adjustment guide population; the multiple reproduction optimization module 15 is configured to guide the first fabric production adjustment domain to perform multiple reproduction optimization based on the fabric quality evaluation multi-dimensional model and according to the first adjustment guide population, and obtain a production control optimization result; and the optimization control module 16 is configured to perform optimization control on the sanding machine according to the production control optimization result.

[0075] Further, the fabric production data acquisition module 11 in the sanding machine fabric quality evaluation system combined with machine vision is further configured to: when the sanding machine executes the fabric production instruction, synchronously collect a fabric production monitoring image according to a machine vision device; and perform denoising processing on the fabric production monitoring image to obtain the fabric production image.

[0076] Further, the multi-dimensional evaluation module 12 in the sanding machine fabric quality evaluation system combined with machine vision is further configured to: perform multi-dimensional defect capture on the fabric production image according to the fabric production requirement, to obtain a fabric surface defect grid, a fabric texture defect grid, and a fabric color defect grid; activate the fabric quality evaluation multi-dimensional model, the fabric quality evaluation multi-dimensional model including a surface defect evaluation model, a texture defect evaluation model, and a color defect evaluation model; input the fabric surface defect grid into the surface defect evaluation model to obtain a surface defect evaluation coefficient; input the fabric texture defect grid into the texture defect evaluation model to obtain a texture defect evaluation coefficient; input the fabric color defect grid into the color defect evaluation model to obtain a color defect evaluation coefficient; and collate the fabric surface defect grid, the fabric texture defect grid, the fabric color defect grid, the surface defect evaluation coefficient, the texture defect evaluation coefficient, and the color defect evaluation coefficient to obtain the fabric quality evaluation atlas.

[0077] Further, the multi-dimensional evaluation module 12 in the fabric quality evaluation system combined with machine vision is further used for: performing surface feature convolution according to the fabric production image to obtain a fabric surface feature grid; performing qualified fabric image sample retrieval according to the fabric production requirement to obtain a qualified fabric image set; performing surface feature convolution according to the qualified fabric image set to obtain a plurality of qualified surface feature grids; performing feature aggregation on the plurality of qualified surface feature grids to obtain a surface feature reference grid; and performing surface defect detection on the fabric surface feature grid according to the surface feature reference grid to obtain a fabric surface defect grid.

[0078] Further, the self-adaptive adjustment module 13 in the fabric quality evaluation system combined with machine vision is further used for: performing surface defect path tracing on the fabric quality evaluation graph according to the fabric brushing machine monitoring data to obtain a fabric surface defect path; performing texture defect path tracing on the fabric quality evaluation graph according to the fabric brushing machine monitoring data to obtain a fabric texture defect path; performing color defect path tracing on the fabric quality evaluation graph according to the fabric brushing machine monitoring data to obtain a fabric color defect path; and performing associated feature analysis on the fabric production control scheme according to the fabric surface defect path, the fabric texture defect path and the fabric color defect path respectively to construct a fabric defect control associated graph; and performing multi-parameter adjustment on the fabric production control scheme according to the fabric defect control associated graph to obtain the first fabric production adjustment domain.

[0079] Further, the quality evaluation optimization module 14 in the fabric quality evaluation system combined with machine vision is further used for: extracting a fabric production adjustment first scheme according to the first fabric production adjustment domain; performing fabric feature prediction based on the fabric production adjustment first scheme to obtain a first fabric feature prediction result; obtaining a first fabric quality evaluation sequence according to the fabric quality evaluation multi-dimensional model based on the first fabric feature prediction result; and adding the fabric production adjustment first scheme to the first adjustment guide population if the first fabric quality evaluation sequence satisfies a fabric quality evaluation constraint.

[0080] Further, the quality evaluation optimization module 14 in the fabric quality evaluation system combined with machine vision is further used for: the fabric quality evaluation constraint includes a surface defect evaluation constraint, a texture defect evaluation constraint and a color defect evaluation constraint.

[0081] Further, the multiple reproduction optimization module 15 in the fabric quality evaluation system combined with machine vision is also used for: based on the fabric quality evaluation multi-dimensional model, reproduction optimization is performed on the first fabric production adjustment domain according to the first adjustment guide population, to obtain a second adjustment guide population; based on the fabric quality evaluation multi-dimensional model, reproduction optimization is continuously performed on the first fabric production adjustment domain according to the second adjustment guide population, until a plurality of adjustment guide populations satisfying a predetermined reproduction optimization number are obtained; fabric comprehensive quality calculation is performed on the plurality of adjustment guide populations according to a fabric quality evaluation multi-dimensional weight, to obtain a fabric comprehensive quality distribution; based on the fabric comprehensive quality distribution, comprehensive quality optimization is performed on the plurality of adjustment guide populations according to a fabric comprehensive quality threshold, to obtain a second fabric production adjustment domain; and energy consumption minimization optimization is performed according to the second fabric production adjustment domain, to obtain the production control optimization result.

[0082] Further, the multiple reproduction optimization module 15 in the fabric quality evaluation system combined with machine vision is also used for: difference identification is performed on the first fabric production adjustment domain according to the first adjustment guide population, to obtain a first adjustment difference vector set; random variation is performed according to the first adjustment difference vector set, to obtain a first adjustment variation vector set; variation reproduction is performed on the first fabric production adjustment domain according to the first adjustment variation vector set, to obtain a first adjustment variation reproduction space; and fabric quality evaluation optimization is performed on the first adjustment variation reproduction space according to the fabric quality evaluation multi-dimensional model, to generate the second adjustment guide population.

[0083] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The foregoing Figure 1 The fabric quality evaluation method combined with machine vision in the first embodiment and the specific examples are also applicable to the fabric quality evaluation system combined with machine vision in the present embodiment. Through the foregoing detailed description of the fabric quality evaluation method combined with machine vision, those skilled in the art can clearly know the fabric quality evaluation system combined with machine vision in the present embodiment. Therefore, for the sake of brevity of the specification, the fabric quality evaluation system combined with machine vision in the present embodiment will not be described in detail.

[0084] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0085] Obviously, for those skilled in the art, some improvements and modifications can be made to the present application without departing from the principles of the present application, and these improvements and modifications also fall within the protection scope of the present application.

Claims

1. A method for assessing the quality of a brushed woven fabric in combination with machine vision, characterized in that, The method comprises the following steps: When the sanding machine executes the fabric production instruction, a fabric production image is synchronously acquired, and the fabric production instruction comprises a fabric production control scheme corresponding to a fabric production requirement; A fabric quality evaluation multidimensional model is introduced, and fabric quality multidimensional evaluation is performed in combination with the fabric production requirement and the fabric production image to construct a fabric quality evaluation atlas; The fabric production control scheme is adaptively adjusted according to the fabric quality evaluation atlas to obtain a first fabric production adjustment domain; The fabric quality evaluation multidimensional model is used to perform fabric quality evaluation optimization on the first fabric production adjustment domain to obtain a first adjustment guide population; Based on the fabric quality evaluation multidimensional model, the first adjustment guide population is used to guide the first fabric production adjustment domain to perform multiple reproduction optimizations to obtain a production control optimization result; The sanding machine is controlled according to the production control optimization result; Based on the fabric quality evaluation multidimensional model, the first adjustment guide population is used to guide the first fabric production adjustment domain to perform multiple reproduction optimizations to obtain a production control optimization result, which comprises: The first adjustment guide population is used to perform reproduction optimization on the first fabric production adjustment domain based on the fabric quality evaluation multidimensional model to obtain a second adjustment guide population; Based on the fabric quality evaluation multidimensional model, the second adjustment guide population is used to continue to perform reproduction optimization on the first fabric production adjustment domain until multiple adjustment guide populations that meet a predetermined reproduction optimization number are obtained; Fabric comprehensive quality calculation is performed on the multiple adjustment guide populations based on a fabric quality evaluation multidimensional weight to obtain a fabric comprehensive quality distribution; Based on the fabric comprehensive quality distribution, comprehensive quality optimization is performed on the multiple adjustment guide populations based on a fabric comprehensive quality threshold to obtain a second fabric production adjustment domain; Energy consumption minimization optimization is performed based on the second fabric production adjustment domain to obtain the production control optimization result; Based on the fabric quality evaluation multidimensional model, the first adjustment guide population is used to perform reproduction optimization on the first fabric production adjustment domain to obtain a second adjustment guide population, which comprises: Differences are identified in the first adjustment guide population for the first fabric production adjustment domain to obtain a first adjustment difference vector set; Random variation is performed based on the first adjustment difference vector set to obtain a first adjustment variation vector set; Variation reproduction is performed on the first fabric production adjustment domain based on the first adjustment variation vector set to obtain a first adjustment variation reproduction space; Fabric quality evaluation optimization is performed on the first adjustment variation reproduction space based on the fabric quality evaluation multidimensional model to generate the second adjustment guide population.

2. The method for assessing the quality of a raised fabric in conjunction with machine vision according to claim 1, wherein, A fabric quality evaluation multidimensional model is introduced, and fabric quality multidimensional evaluation is performed in combination with the fabric production requirement and the fabric production image to construct a fabric quality evaluation atlas, which comprises: Multidimensional defect capturing is performed on the fabric production image based on the fabric production requirement to obtain a fabric surface defect grid, a fabric texture defect grid and a fabric color defect grid; activating the fabric quality evaluation multi-dimensional model, the fabric quality evaluation multi-dimensional model comprising a surface defect evaluation model, a texture defect evaluation model and a color defect evaluation model; inputting the fabric surface defect grid into the surface defect evaluation model to obtain a surface defect evaluation coefficient; inputting the fabric texture defect grid into the texture defect evaluation model to obtain a texture defect evaluation coefficient; inputting the fabric color defect grid into the color defect evaluation model to obtain a color defect evaluation coefficient; sorting the fabric surface defect grid, the fabric texture defect grid, the fabric color defect grid, the surface defect evaluation coefficient, the texture defect evaluation coefficient and the color defect evaluation coefficient to obtain the fabric quality evaluation atlas.

3. The method for assessing the quality of a raised fabric in conjunction with machine vision according to claim 2, wherein, According to the fabric production requirements, the fabric production image is subjected to multi-dimensional defect capture, comprising: According to the fabric production image, surface feature convolution is performed to obtain a fabric surface feature grid; According to the fabric production requirements, qualified fabric image sample retrieval is performed to obtain a qualified fabric image set; According to the qualified fabric image set, surface feature convolution is performed to obtain a plurality of qualified surface feature grids; The plurality of qualified surface feature grids are subjected to feature aggregation to obtain a surface feature reference grid; According to the surface feature reference grid, surface defect detection is performed on the fabric surface feature grid to obtain the fabric surface defect grid.

4. The method for assessing the quality of a raised fabric in conjunction with machine vision according to claim 1, wherein, According to the fabric quality evaluation atlas, the fabric production control scheme is subjected to self-adaptive adjustment to obtain a first fabric production adjustment domain, comprising: According to the fabric quality evaluation atlas, surface defect path tracing is performed according to the fabric quality evaluation atlas based on the sanding machine monitoring data to obtain a fabric surface defect path; According to the fabric quality evaluation atlas, texture defect path tracing is performed according to the fabric quality evaluation atlas based on the sanding machine monitoring data to obtain a fabric texture defect path; According to the fabric quality evaluation atlas, color defect path tracing is performed according to the fabric quality evaluation atlas based on the sanding machine monitoring data to obtain a fabric color defect path; According to the fabric surface defect path, the fabric texture defect path and the fabric color defect path, respectively, the fabric production control scheme is subjected to associated feature analysis to construct a fabric defect control association atlas; According to the fabric defect control association atlas, multi-parameter adjustment is performed on the fabric production control scheme to obtain the first fabric production adjustment domain.

5. The method for assessing the quality of a raised fabric in conjunction with machine vision according to claim 1, wherein, According to the fabric quality evaluation multi-dimensional model, the first fabric production adjustment domain is subjected to fabric quality evaluation optimization to obtain a first adjustment guide population, comprising: According to the first fabric production adjustment domain, a fabric production adjustment first scheme is extracted; Based on the fabric production adjustment first scheme, fabric feature prediction is performed to obtain a first fabric feature prediction result; Based on the first fabric feature prediction result, a first fabric quality evaluation sequence is obtained according to the fabric quality evaluation multi-dimensional model; If the first fabric quality evaluation sequence satisfies the fabric quality evaluation constraint, the fabric production adjustment first scheme is added to the first adjustment guide population.

6. The method for assessing the quality of a raised fabric in combination with machine vision according to claim 5, characterized in that, The fabric quality evaluation constraint comprises a surface defect evaluation constraint, a texture defect evaluation constraint and a color defect evaluation constraint.

7. The method for assessing the quality of a raised fabric in conjunction with machine vision according to claim 1, wherein, When the sanding machine executes the fabric production instruction, a fabric production image is synchronously acquired, including: When the sanding machine executes the fabric production instruction, a fabric production monitoring image is synchronously acquired according to a machine vision device; The fabric production monitoring image is denoised to obtain the fabric production image.

8. A system for assessing the quality of a raised fabric in combination with machine vision, characterized by, Steps for implementing the method for evaluating fabric quality of a sanding machine combined with machine vision according to any one of claims 1 to 7, the system for evaluating fabric quality of a sanding machine combined with machine vision comprising: A fabric production data acquisition module for synchronously acquiring a fabric production image when a sanding machine executes a fabric production instruction, the fabric production instruction including a fabric production control scheme corresponding to fabric production requirements; A multi-dimensional evaluation module for introducing a multi-dimensional fabric quality evaluation model, combining the fabric production requirements and the fabric production image to perform multi-dimensional fabric quality evaluation, and constructing a fabric quality evaluation atlas; An adaptive adjustment module for adaptively adjusting the fabric production control scheme according to the fabric quality evaluation atlas to obtain a first fabric production adjustment domain; A quality evaluation optimization module for performing fabric quality evaluation optimization on the first fabric production adjustment domain according to the multi-dimensional fabric quality evaluation model to obtain a first adjustment guide population; A multiple reproduction optimization module for guiding the first fabric production adjustment domain to perform multiple reproduction optimization according to the first adjustment guide population based on the multi-dimensional fabric quality evaluation model to obtain a production control optimization result; An optimization control module for performing optimization control on the sanding machine according to the production control optimization result.

Images