Stainless steel chain product quality inspection method and system

Abstract

The application relates to the technical field of metal product inspection, and provides a stainless steel chain product quality inspection method and system.The method comprises the following steps: image acquisition is performed on a plurality of stainless steel chain products to be inspected; appearance defect detection is performed on the plurality of stainless steel chain products to be inspected based on a plurality of image acquisition results; screening is performed on the plurality of stainless steel chain products to be inspected based on a plurality of appearance defect detection results; production parameter acquisition is performed on the plurality of screened stainless steel chain products; comparison is performed on a plurality of product production parameter sets based on a standard production parameter information database; quality inspection sample setting is performed based on a plurality of quality inspection indexes and a plurality of product production parameter comparison results; and quality inspection is performed based on the plurality of quality inspection indexes and a quality inspection sample data set.The method can solve the technical problem that the quality inspection accuracy is low due to the single quality inspection method for the stainless steel chain products in the prior art.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of metal product quality inspection, in particular to a stainless steel chain product quality inspection method and system. BACKGROUND

[0002] In recent years, with the continuous development of China's economy, the demand for metal products has also continued to grow, driving the market size of the metal product industry to increase continuously. The metal product industry includes structural metal product manufacturing, metal tool manufacturing, container and metal packaging container manufacturing, stainless steel and similar daily metal product manufacturing, etc.

[0003] In order to improve the production and manufacturing quality of metal structural parts, strengthen the supervision and inspection and quality control of the related links of the metal product production process, and ensure the complete implementation of the design requirements of the structural parts, it is particularly important to have a complete quality inspection method.

[0004] In real life, the quality inspection of stainless steel chain products is mainly carried out by quality inspection indicators, which has low accuracy and often cannot meet the quality inspection requirements of users.

[0005] In summary, the existing technology has the technical problem of low quality inspection accuracy caused by the single quality inspection method of stainless steel chain products. SUMMARY

[0006] Therefore, it is necessary to provide a method and system capable of improving the quality inspection accuracy of stainless steel chain products in order to solve the above technical problems.

[0007] A stainless steel chain product quality inspection method, which is applied to a stainless steel chain product quality inspection system, the system is in communication connection with an image acquisition device, and the method comprises the following steps: based on the image acquisition device, image acquisition is performed on a plurality of to-be-inspected stainless steel chain products to obtain a plurality of image acquisition results; based on the plurality of image acquisition results, appearance defect detection is performed on the plurality of to-be-inspected stainless steel chain products to obtain a plurality of appearance defect detection results; based on the plurality of appearance defect detection results, screening is performed on the plurality of to-be-inspected stainless steel chain products to obtain a plurality of screened stainless steel chain products; based on the plurality of screened stainless steel chain products, production parameter acquisition is performed to obtain a plurality of product production parameter sets; based on a standard production parameter information database, the plurality of product production parameter sets are compared respectively to obtain a plurality of product production parameter comparison results; a plurality of quality inspection indicators are obtained, quality inspection sample setting is performed based on the plurality of quality inspection indicators and the plurality of product production parameter comparison results to obtain a quality inspection sample dataset; and quality inspection is performed on the plurality of screened stainless steel chain products based on the plurality of quality inspection indicators and the quality inspection sample dataset.

[0008] In an embodiment, based on the plurality of image acquisition results, appearance defect detection is performed on the plurality of stainless steel chain products to be inspected to obtain a plurality of appearance defect detection results, and the method further comprises: obtaining a plurality of preset appearance defect detection indicators; based on the plurality of preset appearance defect detection indicators, an appearance defect detection image knowledge base is constructed; based on the plurality of preset appearance defect detection indicators, a multi-level appearance defect comparison is performed on the appearance defect detection image knowledge base and the plurality of image acquisition results to obtain the plurality of appearance defect detection results, wherein the plurality of appearance defect detection results have a corresponding relationship with the plurality of image acquisition results.

[0009] In an embodiment, based on the plurality of appearance defect detection results, screening is performed on the plurality of stainless steel chain products to be inspected to obtain a plurality of screened stainless steel chain products, and the method further comprises: based on the plurality of appearance defect detection results, appearance quality evaluation is performed on the plurality of stainless steel chain products to be inspected to obtain a plurality of appearance quality evaluation coefficients; a preset appearance quality evaluation coefficient is obtained; based on the preset appearance quality evaluation coefficient, the plurality of appearance quality evaluation coefficients are screened to obtain a plurality of preferred appearance quality evaluation coefficients that meet the preset appearance quality evaluation coefficient; based on the plurality of preferred appearance quality evaluation coefficients, the plurality of stainless steel chain products to be inspected are matched to obtain the plurality of screened stainless steel chain products.

[0010] In an embodiment, the plurality of appearance quality evaluation coefficients are obtained, and the method further comprises: an appearance quality evaluation model is constructed; the plurality of appearance defect detection results are input as input information into the appearance quality evaluation model to obtain the plurality of appearance quality evaluation coefficients.

[0011] In an embodiment, based on a standard production parameter information database, a plurality of product production parameter sets are compared respectively to obtain a plurality of product production parameter comparison results, and the method further comprises: based on the plurality of stainless steel chain products to be inspected, product type analysis is performed to obtain a plurality of product type analysis results; based on the plurality of product type analysis results, standard production parameter acquisition is performed to obtain the standard production parameter information database; a plurality of production parameter comparison dimensions are obtained, wherein the plurality of production parameter comparison dimensions include production material parameter comparison dimension, production structure parameter comparison dimension, and production performance parameter comparison dimension; based on the plurality of production parameter comparison dimensions and the standard production parameter information database, a plurality of product production parameter sets are compared to obtain the plurality of product production parameter comparison results.

[0012] In one embodiment, the quality inspection sample dataset is obtained, and the method further comprises: performing historical data query based on the plurality of quality inspection indexes, obtaining a plurality of historical quality inspection samples and a plurality of historical product production parameter comparison results, and the plurality of historical quality inspection samples and the plurality of historical product production parameter comparison results have a corresponding relationship; performing correlation evaluation on the plurality of historical product production parameter comparison results and the plurality of product production parameter comparison results, obtaining a plurality of correlation evaluation coefficients; based on a correlation evaluation coefficient threshold, screening the plurality of correlation evaluation coefficients, and adding the plurality of historical inspection samples corresponding to the plurality of correlation evaluation coefficients satisfying the correlation evaluation coefficient threshold to the quality inspection sample database; performing principal component analysis based on the quality inspection sample database to obtain the quality inspection sample dataset.

[0013] In one embodiment, the quality inspection sample dataset is obtained based on the principal component analysis of the quality inspection sample database, and the method further comprises: obtaining a first quality inspection sample feature dataset from the quality inspection sample database; performing decentralization processing on the first quality inspection sample feature dataset to obtain a second quality inspection sample feature dataset; obtaining a covariance matrix of the first quality inspection sample feature according to the second quality inspection sample feature dataset; obtaining a first eigenvalue and a first eigenvector according to the covariance matrix of the first quality inspection sample feature; and obtaining the quality inspection sample dataset according to the first eigenvalue and the first eigenvector.

[0014] A stainless steel chain product quality inspection system, the system is in communication connection with an image acquisition device, and the system comprises:

[0015] An image acquisition module, the image acquisition module is used for image acquisition of a plurality of stainless steel chain products to be inspected based on the image acquisition device, and a plurality of image acquisition results are obtained;

[0016] A defect detection module, the defect detection module is used for appearance defect detection of the plurality of stainless steel chain products to be inspected based on the plurality of image acquisition results, and a plurality of appearance defect detection results are obtained;

[0017] A product screening module, the product screening module is used for screening the plurality of stainless steel chain products to be inspected based on the plurality of appearance defect detection results, and a plurality of screened stainless steel chain products are obtained;

[0018] A parameter acquisition module, the parameter acquisition module is used for production parameter acquisition based on the plurality of screened stainless steel chain products, and a plurality of product production parameter sets are obtained;

[0019] A parameter comparison module is configured to compare a plurality of product production parameter sets based on a standard production parameter information database, and obtain a plurality of product production parameter comparison results.

[0020] A sample setting module is configured to obtain a plurality of quality inspection indexes, set quality inspection samples based on the plurality of quality inspection indexes and the plurality of product production parameter comparison results, and obtain a quality inspection sample dataset.

[0021] A quality inspection module is configured to perform quality inspection on the plurality of screened stainless steel chain products based on the plurality of quality inspection indexes and the quality inspection sample dataset.

[0022] The above-described stainless steel chain product quality inspection method and system can construct an appearance defect detection knowledge base according to historical appearance defect image data, compare the appearance defect detection knowledge base with the plurality of image collection results in multiple levels, and find out appearance defect products that are difficult to be detected by naked eyes, thereby improving the accuracy of screening the plurality of stainless steel chain products to be inspected based on the plurality of appearance defect detection results. The quality inspection sample dataset is obtained by setting quality inspection samples based on the plurality of quality inspection indexes and the plurality of product production parameter comparison results, and the plurality of screened stainless steel chain products are inspected based on the plurality of quality inspection indexes and the quality inspection sample dataset, thereby solving the technical problem of low accuracy of the existing stainless steel chain product quality inspection method, and improving the accuracy of stainless steel chain product quality inspection.

[0023] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the specific embodiments of the present application can be implemented according to the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 A flowchart of a stainless steel chain product quality inspection method is provided for the present application.

[0025] Figure 2 A structural diagram of a stainless steel chain product quality inspection system is provided for the present application.

[0026] Explanation of reference signs: image collection module 1, defect detection module 2, product screening module 3, parameter collection module 4, parameter comparison module 5, sample setting module 6, quality inspection module 7. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0028] As shown in Figure 1 The present application provides a stainless steel chain product quality inspection method, which is applied to a stainless steel chain product quality inspection system, the system is in communication connection with an image acquisition device, and the method comprises the following steps:

[0029] Step S100: image acquisition of a plurality of stainless steel chain products to be inspected is performed based on the image acquisition device, and a plurality of image acquisition results are obtained;

[0030] Specifically, the image acquisition device refers to a device for real-time image acquisition of stainless steel chain products, which comprises a lens, a frame memory, an information transmission module and the like. The real-time images collected are transmitted to the quality inspection system through the information transmission module. The image acquisition device is used to perform multi-angle and omnidirectional image acquisition of a plurality of stainless steel chain products that have not been subjected to quality inspection. A plurality of image acquisition results are obtained. The plurality of image acquisition results include image acquisition results of the overall surface, parts and positions of the stainless steel chain products, such as chain plates, chain pins, shaft sleeves and other components. Image acquisition of the products provides data support for the next step of appearance detection.

[0031] Step S200: appearance defect detection of the plurality of stainless steel chain products to be inspected is performed based on the plurality of image acquisition results, and a plurality of appearance defect detection results are obtained;

[0032] In one embodiment, the step S200 of the present application further comprises:

[0033] Step S210: a plurality of preset appearance defect detection indexes are obtained;

[0034] Step S220: an appearance defect detection image knowledge base is constructed based on the plurality of preset appearance defect detection indexes;

[0035] Step S230: multi-level appearance defect comparison is performed on the appearance defect detection image knowledge base and the plurality of image acquisition results based on the plurality of preset appearance defect detection indexes, and the plurality of appearance defect detection results are obtained, wherein the plurality of appearance defect detection results have a corresponding relationship with the plurality of image acquisition results.

[0036] Specifically, the appearance defect detection indicators include chain plate cracks, chain plate connection fractures, chain pin missing, chain pin too large, shaft sleeve cracks and other factors that can be found by appearance and affect the normal use of the product. According to the plurality of preset appearance defect detection indicators, the stainless steel chain product appearance defect image data conforming to the indicators is obtained in the historical image data, and the appearance defect detection image knowledge base includes a plurality of stainless steel product historical appearance defect image data. Based on the plurality of preset appearance defect detection indicators, by comparing the plurality of stainless steel chain product image acquisition results with the historical appearance defect image data in the appearance defect detection image knowledge base, an comparison result is obtained, and the plurality of appearance defect detection results are obtained by judging whether the stainless steel chain product has an appearance defect based on the comparison result, wherein the plurality of appearance defect detection results correspond one-to-one with the plurality of image acquisition results. The first round of quality detection of the stainless steel chain product is performed by obtaining the appearance defect detection result, thereby improving the efficiency of the overall quality inspection.

[0037] Step S300: screening the plurality of stainless steel chain products to be inspected based on the plurality of appearance defect detection results, and obtaining a plurality of screened stainless steel chain products;

[0038] In one embodiment, the step S300 of the present application further comprises:

[0039] Step S310: based on the plurality of appearance defect detection results, performing appearance quality evaluation on the plurality of stainless steel chain products to be inspected, and obtaining a plurality of appearance quality evaluation coefficients;

[0040] In one embodiment, the step S310 of the present application further comprises:

[0041] Step S311: constructing an appearance quality evaluation model;

[0042] Step S312: inputting the plurality of appearance defect detection results as input information into the appearance quality evaluation model to obtain the plurality of appearance quality evaluation coefficients.

[0043] Specifically, the appearance quality evaluation model is used to evaluate the plurality of appearance defect detection results. A plurality of appearance defect weights are preset, the plurality of appearance defect weights correspond one-to-one to the plurality of appearance defect indicators, the various appearance defects in the plurality of appearance defect detection results are multiplied by the corresponding weights to obtain products, and then the products are added to obtain a plurality of summation results. The plurality of appearance quality evaluation coefficients are the plurality of summation results. By obtaining the plurality of appearance quality evaluation coefficients, data support is provided for the next step of screening the plurality of stainless steel chain products to be inspected.

[0044] Step S320: obtaining a preset appearance quality evaluation coefficient threshold;

[0045] Step S330: screening the plurality of appearance quality evaluation coefficients based on the preset appearance quality evaluation coefficient threshold to obtain a plurality of preferred appearance quality evaluation coefficients satisfying the preset appearance quality evaluation coefficient threshold;

[0046] Step S340: matching the plurality of to-be-inspected stainless steel chain products based on the plurality of preferred appearance quality evaluation coefficients to obtain the plurality of screened stainless steel chain products.

[0047] Specifically, a plurality of historical appearance images with appearance quality meeting the standard are evaluated by the appearance quality evaluation model to obtain a plurality of historical appearance image evaluation coefficients, the preset appearance quality evaluation coefficient threshold refers to a threshold meeting the plurality of historical appearance image evaluation coefficients, the plurality of appearance quality evaluation coefficients are screened by the preset appearance quality evaluation coefficient threshold, the plurality of preferred appearance quality evaluation coefficients refer to evaluation coefficients satisfying the preset appearance quality evaluation coefficient threshold, and the plurality of to-be-inspected stainless steel chain products are one-to-one matched based on the plurality of preferred appearance quality evaluation coefficients to obtain the plurality of screened stainless steel chain products. By obtaining the plurality of screened stainless steel chain products, raw data is provided for the next step of production parameter collection.

[0048] Step S400: collecting production parameters based on the plurality of screened stainless steel chain products to obtain a plurality of product production parameter sets;

[0049] Specifically, a plurality of product production parameter sets are obtained by collecting parameters of the plurality of screened stainless steel chain products, and the product production parameters include parameters such as raw materials, specifications, sizes, and weights of the stainless steel chain products. By obtaining the plurality of product production parameter sets, data support is provided for the next step of comparison with standard production parameters.

[0050] Step S500: comparing the plurality of product production parameter sets based on a standard production parameter information database to obtain a plurality of product production parameter comparison results;

[0051] In one embodiment, the step S500 of the present application further includes:

[0052] Step S510: performing product type analysis based on the plurality of to-be-inspected stainless steel chain products to obtain a plurality of product type analysis results;

[0053] Step S520: collecting standard production parameters based on the plurality of product type analysis results to obtain the standard production parameter information database;

[0054] In step S530, a multi-level production parameter comparison dimension is obtained, wherein the multi-level production parameter comparison dimension includes a production material parameter comparison dimension, a production structure parameter comparison dimension, and a production performance parameter comparison dimension.

[0055] In step S540, based on the multi-level production parameter comparison dimension and the standard production parameter information database, a plurality of product production parameter sets are compared to obtain a plurality of product production parameter comparison results.

[0056] Specifically, the plurality of stainless steel chain products to be inspected are classified according to different product types, for example: according to the product type, the stainless steel chain products can be divided into toothed chain, plate chain, hollow pin shaft chain, double pitch transmission chain, double pitch conveying chain, etc. to obtain a plurality of product classification results. The standard production parameters of the plurality of product classification results are obtained. The standard production parameter information database includes the standard production parameters of the plurality of product classification results. A multi-level production parameter comparison dimension is set, and the multi-level production parameter includes a production material parameter, a production structure parameter, and a production performance parameter. The plurality of product production parameter sets and the standard production parameter information database are compared through the three dimensions of production material parameter, production structure parameter, and production performance parameter to obtain comparison results. The plurality of product production parameter comparison results include production material parameter comparison results, production structure parameter comparison results, and production performance parameter comparison results. Through obtaining the plurality of product production parameter comparison results, data support is provided for the next step of setting quality inspection samples.

[0057] In step S600, a plurality of quality inspection indexes are obtained, and based on the plurality of quality inspection indexes and the plurality of product production parameter comparison results, quality inspection sample setting is performed to obtain a quality inspection sample dataset.

[0058] In one embodiment, step S600 of the present application further includes:

[0059] In step S610, based on the plurality of quality inspection indexes, historical data is queried to obtain a plurality of historical quality inspection samples and a plurality of historical product production parameter comparison results, and the plurality of historical quality inspection samples and the plurality of historical product production parameter comparison results have a corresponding relationship.

[0060] In step S620, the plurality of historical product production parameter comparison results and the plurality of product production parameter comparison results are associated to obtain a plurality of association evaluation coefficients.

[0061] In step S630, based on an association evaluation coefficient threshold, the plurality of association evaluation coefficients are screened, and the plurality of historical inspection samples corresponding to the plurality of association evaluation coefficients satisfying the association evaluation coefficient threshold are added to the quality inspection sample database.

[0062] Step S640: Principal component analysis is performed based on the quality inspection sample database to obtain the quality inspection sample dataset.

[0063] Specifically, the plurality of quality inspection indexes refer to type information of quality tests performed on different types of stainless steel chain products, such as parameters of quality tests such as tension test, pressure test, and load capacity test. The stainless steel chain products that have completed production and meet quality inspection standards are obtained according to the plurality of quality inspection indexes. The plurality of historical quality inspection samples refer to parameters of quality tests performed on products that meet quality requirements. The production parameters and standard parameters are queried to obtain a plurality of historical product production parameter comparison results, wherein the plurality of historical quality inspection samples and the plurality of historical product production parameter comparison results correspond one-to-one. The correlation evaluation refers to mining the relationship between the plurality of historical product production parameter comparison results and the plurality of product production parameter comparison results to obtain a plurality of correlation evaluation coefficients. The correlation evaluation coefficients are used to measure the coefficient index of the relationship between the two. A correlation evaluation coefficient threshold is set, the plurality of correlation evaluation coefficients are screened according to the correlation evaluation coefficient threshold, a quality inspection sample database is constructed, and the plurality of historical inspection samples corresponding to the plurality of correlation evaluation coefficients within the threshold range are added to the quality inspection sample database. Then, the feature data in the quality inspection sample database is processed by dimension reduction to obtain the quality inspection sample dataset obtained by dimension reduction. By obtaining the quality inspection sample dataset, a reference standard is provided for the next step of quality inspection of the plurality of screened stainless steel chain products.

[0064] In one embodiment, the step S640 of the present application further comprises:

[0065] Step S641: According to the quality inspection sample database, a first quality inspection sample feature dataset is obtained.

[0066] Step S642: The first quality inspection sample feature dataset is processed by decentralization to obtain a second quality inspection sample feature dataset.

[0067] Step S643: According to the second quality inspection sample feature dataset, a covariance matrix of the first quality inspection sample feature is obtained.

[0068] Step S644: According to the covariance matrix of the first quality inspection sample feature, a first eigenvalue and a first eigenvector are obtained.

[0069] Step S645: According to the first eigenvalue and the first eigenvector, the quality inspection sample dataset is obtained.

[0070] Specifically, first, the feature data in the quality inspection sample database is subjected to data standardization processing, and a feature data set matrix is constructed to obtain the first quality inspection sample feature data set. Then, each feature data in the first quality inspection sample feature data set is subjected to decentralization processing, which refers to solving the average value of each feature in the first quality inspection sample feature data set, and then subtracting the average value of each feature from each feature of all samples to obtain new feature values. The second quality inspection sample feature data set is a data matrix composed of the new feature values. The second quality inspection sample feature data set is operated through a covariance formula to obtain the covariance matrix of the first quality inspection sample feature of the second quality inspection sample feature data set, and then the eigenvalues and eigenvectors of the covariance matrix of the first quality inspection sample feature are solved through matrix operation, and each eigenvalue corresponds to an eigenvector. In the solved first eigenvector, the first K largest eigenvalues and the corresponding eigenvectors are selected, and the original features in the first quality inspection sample feature data set are projected onto the selected eigenvectors to obtain the first quality inspection sample feature data set after dimension reduction. Through principal component analysis of the quality inspection sample database, the feature data in the quality inspection sample database can be reduced in dimension, redundant data can be removed under the premise of ensuring the information amount, so that the sample amount of the feature data in the quality inspection sample database is reduced, thereby accelerating the operation speed of the training model for data.

[0071] Step S700: based on the plurality of quality inspection indexes and the quality inspection sample data set, performing quality inspection on the plurality of screened stainless steel chain products.

[0072] Specifically, according to the plurality of quality inspection indexes and the obtained quality inspection sample data set, quality testing and inspection are performed on the plurality of screened stainless steel chain products.

[0073] In one embodiment, as Figure 2 shown, a stainless steel chain product quality inspection system is provided, which is in communication connection with an image acquisition device, and the system comprises an image acquisition module 1, a defect detection module 2, a product screening module 3, a parameter acquisition module 4, a parameter comparison module 5, a sample setting module 6, and a quality inspection module 7, wherein:

[0074] The image acquisition module 1 is used for acquiring images of a plurality of stainless steel chain products to be inspected based on the image acquisition device to obtain a plurality of image acquisition results.

[0075] a defect detection module 2 configured to perform appearance defect detection on the plurality of stainless steel chain products to be inspected based on the plurality of image acquisition results, and obtain a plurality of appearance defect detection results;

[0076] a product screening module 3 configured to screen the plurality of stainless steel chain products based on the plurality of appearance defect detection results, and obtain a plurality of screened stainless steel chain products;

[0077] a parameter acquisition module 4 configured to acquire production parameters based on the plurality of screened stainless steel chain products, and obtain a plurality of product production parameter sets;

[0078] a parameter comparison module 5 configured to compare the plurality of product production parameter sets based on a standard production parameter information database, and obtain a plurality of product production parameter comparison results;

[0079] a sample setting module 6 configured to obtain a plurality of quality inspection indexes, set quality inspection samples based on the plurality of quality inspection indexes and the plurality of product production parameter comparison results, and obtain a quality inspection sample data set;

[0080] a quality inspection module 7 configured to perform quality inspection on the plurality of screened stainless steel chain products based on the plurality of quality inspection indexes and the quality inspection sample data set.

[0081] In one embodiment, the system further comprises:

[0082] an index obtaining module configured to obtain a plurality of preset appearance defect detection indexes;

[0083] a knowledge base construction module configured to construct an appearance defect detection image knowledge base based on the plurality of preset appearance defect detection indexes;

[0084] a result obtaining module configured to perform multi-level appearance defect comparison on the appearance defect detection image knowledge base and the plurality of image acquisition results based on the plurality of preset appearance defect detection indexes, and obtain the plurality of appearance defect detection results, wherein the plurality of appearance defect detection results have a corresponding relationship with the plurality of image acquisition results.

[0085] In one embodiment, the system further comprises:

[0086] a quality evaluation module configured to perform appearance quality evaluation on the plurality of stainless steel chain products to be inspected based on the plurality of appearance defect detection results, and obtain a plurality of appearance quality evaluation coefficients;

[0087] a coefficient obtaining module, configured to obtain preset appearance quality evaluation coefficients;

[0088] a coefficient screening module, configured to screen the plurality of appearance quality evaluation coefficients based on the preset appearance quality evaluation coefficients, to obtain a plurality of preferred appearance quality evaluation coefficients meeting preset appearance quality evaluation coefficients;

[0089] a coefficient matching module, configured to match the plurality of stainless steel chain products to be inspected based on the plurality of preferred appearance quality evaluation coefficients, to obtain the plurality of screened stainless steel chain products.

[0090] In an embodiment, the system further comprises:

[0091] a model constructing module, configured to construct an appearance quality evaluation model;

[0092] a coefficient obtaining module, configured to input the plurality of appearance defect detection results as input information into the appearance quality evaluation model, to obtain the plurality of appearance quality evaluation coefficients.

[0093] In an embodiment, the system further comprises:

[0094] a type analysis module, configured to perform product type analysis based on the plurality of stainless steel chain products to be inspected, to obtain a plurality of product type analysis results;

[0095] a parameter collecting module, configured to collect standard production parameters based on the plurality of product type analysis results, to obtain the standard production parameter information database;

[0096] a dimension obtaining module, configured to obtain a plurality of production parameter comparison dimensions, wherein the plurality of production parameter comparison dimensions comprise a production material parameter comparison dimension, a production structure parameter comparison dimension, and a production performance parameter comparison dimension;

[0097] a result obtaining module, configured to compare a plurality of product production parameter sets based on the plurality of production parameter comparison dimensions and the standard production parameter information database, to obtain a plurality of product production parameter comparison results.

[0098] In an embodiment, the system further comprises:

[0099] The data query module is used to query historical data based on the multiple quality inspection indicators to obtain comparison results of multiple historical quality inspection samples and multiple historical product production parameters, and the multiple historical quality inspection samples and the multiple historical product production parameter comparison results have a corresponding relationship.

[0100] A coefficient acquisition module is used to perform a correlation assessment on the comparison results of the multiple historical product production parameters and the comparison results of the multiple product production parameters, and to obtain multiple correlation assessment coefficients:

[0101] The sample addition module is used to filter the multiple correlation evaluation coefficients based on the correlation evaluation coefficient threshold, and add the multiple historical test samples corresponding to the multiple correlation evaluation coefficients that meet the correlation evaluation coefficient threshold to the quality test sample database.

[0102] The principal component analysis module is used to perform principal component analysis based on the quality inspection sample database to obtain the quality inspection sample dataset.

[0103] In one embodiment, the system further includes:

[0104] The first feature data acquisition module is used to obtain a first quality inspection sample feature dataset based on the quality inspection sample database.

[0105] The second feature data acquisition module is used to perform decentralized processing on the first quality inspection sample feature dataset to obtain the second quality inspection sample feature dataset.

[0106] A matrix acquisition module is used to obtain the covariance matrix of the first quality inspection sample features based on the second quality inspection sample feature dataset.

[0107] An information acquisition module is used to obtain a first eigenvalue and a first eigenvector based on the covariance matrix of the features of the first quality inspection sample.

[0108] A dataset acquisition module is configured to obtain the quality inspection sample dataset based on the first feature value and the first feature vector.

[0109] In summary, this application provides a method and system for quality inspection of stainless steel chain products, which has the following technical advantages:

[0110] 1. Based on the multiple quality inspection indicators and the quality inspection sample dataset, the multiple screened stainless steel chain products are subjected to quality inspection, which solves the technical problem of low accuracy caused by the single quality inspection method for stainless steel chain products in the prior art, and improves the accuracy of quality inspection of stainless steel chain products.

[0111] 2. By comparing multi-level appearance defects with the appearance defect detection knowledge base and image acquisition results, products with appearance defects that are difficult to detect with the naked eye can be discovered, thus improving the accuracy of screening stainless steel chain products to be inspected.

[0112] 3. By performing principal component analysis on the quality inspection sample database, the dimensionality of the feature data in the quality inspection sample database can be reduced. While ensuring the amount of information, redundant data is eliminated, thereby reducing the sample size of the feature data in the quality inspection sample database and accelerating the data processing speed of the training model.

[0113] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0114] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for quality inspection of stainless steel chain products, characterized in that, The method is applied to a stainless steel chain product quality inspection system, the system being communicatively connected to an image acquisition device, and the method includes: Based on the image acquisition device, images of multiple stainless steel chain products to be inspected are acquired, and multiple image acquisition results are obtained. Based on the multiple image acquisition results, the multiple stainless steel chain products to be inspected are subjected to appearance defect detection, and multiple appearance defect detection results are obtained. Based on the results of the multiple appearance defect detections, the multiple stainless steel chain products to be inspected are screened to obtain multiple screened stainless steel chain products. Based on the collection of production parameters from the multiple screened stainless steel chain products, a set of production parameters for multiple products is obtained. Based on a standard production parameter information database, the production parameter sets of multiple products are compared to obtain the comparison results of the production parameters of multiple products. Multiple quality inspection indicators are obtained, and quality inspection samples are set based on the comparison results of the multiple quality inspection indicators and the multiple product production parameters to obtain a quality inspection sample dataset. Based on the multiple quality inspection indicators and the quality inspection sample dataset, the multiple screened stainless steel chain products are subjected to quality inspection. The process of obtaining the quality inspection sample dataset includes: Based on the aforementioned multiple quality inspection indicators, historical data is queried to obtain multiple historical quality inspection samples and multiple historical product production parameter comparison results, and the multiple historical quality inspection samples and multiple historical product production parameter comparison results have a corresponding relationship. A correlation assessment is performed on the comparison results of the production parameters of the multiple historical products and the comparison results of the production parameters of the multiple products to obtain multiple correlation assessment coefficients; Based on the correlation evaluation coefficient threshold, the multiple correlation evaluation coefficients are screened, and multiple historical test samples corresponding to multiple correlation evaluation coefficients that meet the correlation evaluation coefficient threshold are added to the quality test sample database. Principal component analysis was performed on the quality inspection sample database to obtain the quality inspection sample dataset; Specifically, principal component analysis is performed based on the quality inspection sample database to obtain the quality inspection sample dataset, including: Based on the quality inspection sample database, a first quality inspection sample feature dataset is obtained; The first quality inspection sample feature dataset is decentralized to obtain the second quality inspection sample feature dataset. Based on the second quality inspection sample feature dataset, obtain the covariance matrix of the first quality inspection sample features; Based on the covariance matrix of the features of the first quality inspection sample, obtain the first eigenvalue and the first eigenvector; The quality inspection sample dataset is obtained based on the first feature value and the first feature vector.

2. The method as described in claim 1, characterized in that, Based on the multiple image acquisition results, the method further includes: detecting appearance defects in the multiple stainless steel chain products to be inspected, obtaining multiple appearance defect detection results. Multiple preset appearance defect detection indicators are obtained; Based on the multiple preset appearance defect detection indicators, an appearance defect detection image knowledge base is constructed. Based on the multiple preset appearance defect detection indicators, the appearance defect detection image knowledge base and the multiple image acquisition results are compared at multiple levels to obtain the multiple appearance defect detection results, wherein the multiple appearance defect detection results correspond to the multiple image acquisition results.

3. The method as described in claim 1, characterized in that, The method further includes screening the multiple stainless steel chain products to be inspected based on the multiple appearance defect detection results to obtain multiple screened stainless steel chain products. Based on the multiple appearance defect detection results, the appearance quality of the multiple stainless steel chain products to be inspected is evaluated to obtain multiple appearance quality evaluation coefficients. Obtain the preset appearance quality evaluation coefficient; Based on the preset appearance quality evaluation coefficient, the plurality of appearance quality evaluation coefficients are screened to obtain a plurality of preferred appearance quality evaluation coefficients that satisfy the preset appearance quality evaluation coefficients. The plurality of stainless steel chain products to be inspected are matched based on the plurality of preferred appearance quality evaluation coefficients to obtain the plurality of screened stainless steel chain products.

4. The method as described in claim 3, characterized in that, The method for obtaining multiple appearance quality evaluation coefficients further includes: Construct an appearance quality assessment model; The multiple appearance defect detection results are used as input information and input into the appearance quality assessment model to obtain the multiple appearance quality assessment coefficients.

5. The method as described in claim 1, characterized in that, Based on a standard production parameter information database, the method compares multiple product production parameter sets to obtain comparison results for multiple product production parameters. The method further includes: Product type analysis was performed on the multiple stainless steel chain products to be inspected, and multiple product type analysis results were obtained. Based on the analysis results of the multiple product types, standard production parameters are collected to obtain the standard production parameter information database. Obtain multi-level production parameter comparison dimensions, wherein the multi-level production parameter comparison dimensions include production material parameter comparison dimensions, production structure parameter comparison dimensions, and production performance parameter comparison dimensions; Based on the multi-level production parameter comparison dimensions and the standard production parameter information database, multiple product production parameter sets are compared to obtain the comparison results of the multiple product production parameters.

6. A quality inspection system for stainless steel chain products, characterized in that, The system is used to implement the method according to any one of claims 1-5, the system is communicatively connected to an image acquisition device, and the system comprises: An image acquisition module is used to acquire images of multiple stainless steel chain products to be inspected based on the image acquisition device, and obtain multiple image acquisition results. A defect detection module is used to perform appearance defect detection on the multiple stainless steel chain products to be inspected based on the multiple image acquisition results, and obtain multiple appearance defect detection results. The product screening module is used to screen the multiple stainless steel chain products to be inspected based on the multiple appearance defect detection results, and obtain multiple screened stainless steel chain products. The parameter acquisition module is used to acquire production parameters based on the multiple screened stainless steel chain products to obtain a set of multiple product production parameters. The parameter comparison module is used to compare multiple product production parameter sets based on a standard production parameter information database to obtain comparison results of multiple product production parameters. The sample setting module is used to obtain multiple quality inspection indicators, set quality inspection samples based on the multiple quality inspection indicators and the comparison results of multiple product production parameters, and obtain a quality inspection sample dataset. A quality inspection module is used to perform quality inspection on the multiple selected stainless steel chain products based on the multiple quality inspection indicators and the quality inspection sample dataset.

Images