An aluminum cap sorting monitoring system and method based on visual detection

Abstract

The application relates to the technical field of aluminum cap sorting, and discloses an aluminum cap sorting and monitoring system and method based on visual detection, which comprises a collection and processing module configured to perform feature extraction on an image data set to obtain a surface image feature set of an aluminum cap to be sorted; a comparison module configured to generate a defect label of the aluminum cap to be sorted and calculate the coincidence degree of the aluminum cap to be sorted and a defect template; a detection module configured to compare the size feature set with a standard template in a standard object database and determine the quality grade of the aluminum cap to be sorted according to a comparison result; and a calibration module configured to judge whether to calibrate the quality grade of the aluminum cap to be sorted based on an analysis result, calculate an environmental influence coefficient according to running environment data, calibrate the quality grade according to the environmental influence coefficient, and obtain a final quality grade. The application can realize efficient and accurate sorting of aluminum caps and improve sorting efficiency and accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aluminum cap sorting, in particular to an aluminum cap sorting monitoring system and method based on visual detection. BACKGROUND

[0002] With the increasing strictness of the pharmaceutical, food and other industries on packaging quality, the quality of aluminum caps as an important packaging component directly affects the sealing and safety of products. The traditional aluminum cap sorting method mainly relies on manual detection, which is not only inefficient, but also easily affected by human factors, resulting in instability and unreliability of the sorting results.

[0003] Therefore, it is necessary to design an aluminum cap sorting monitoring system and method based on visual detection to solve the problems existing in the current technology. SUMMARY

[0004] In view of this, the present application proposes an aluminum cap sorting monitoring system and method based on visual detection, aiming to realize efficient and accurate sorting of aluminum caps through automation and intelligent means.

[0005] In one aspect, the present application proposes an aluminum cap sorting monitoring system based on visual detection, comprising:

[0006] The acquisition and processing module is configured to determine the aluminum cap to be sorted, acquire a plurality of image data of the aluminum cap to be sorted within a preset time length through a visual detection device, and form an image data set; perform feature extraction on the image data set to obtain a surface image feature set of the aluminum cap to be sorted; wherein the surface image feature set is composed of surface feature values, texture feature values and reflection feature values;

[0007] The comparison module is configured to compare the surface image feature set with a defect template in a standard object database, and determine whether the aluminum cap to be sorted has defects according to the comparison result; if so, generate a defect label of the aluminum cap to be sorted, and calculate the coincidence degree of the aluminum cap to be sorted and the defect template;

[0008] The detection module is configured to determine whether the aluminum cap to be sorted is qualified according to the coincidence degree; if so, obtain a size feature set of the aluminum cap to be sorted according to the image data set, compare the size feature set with a standard template in the standard object database, and determine the quality grade of the aluminum cap to be sorted according to the comparison result;

[0009] The calibration module is configured to collect running environment data of the visual inspection device, analyze the running environment data, judge whether to calibrate the quality grade of the aluminum cap to be sorted based on the analysis result, calculate an environmental influence coefficient based on the running environment data if yes, calibrate the quality grade based on the environmental influence coefficient, and obtain a final quality grade.

[0010] Further, the collection processing module performs feature extraction on the image data set to obtain a surface image feature set of the aluminum cap to be sorted, and the feature extraction includes:

[0011] The image data set is preprocessed, and the preprocessing includes denoising and contrast enhancement.

[0012] An image recognition algorithm is used to extract feature points from the preprocessed image data set to obtain contour features and color features of the surface of the aluminum cap to be sorted.

[0013] Based on the contour features and color features, an image processing technique is used to calculate surface feature values, texture feature values, and reflection feature values of the aluminum cap to be sorted.

[0014] Further, the comparison module compares the surface image feature set with a defect template in a standard object database, and judges whether the aluminum cap to be sorted has defects based on a comparison result, and the judgment includes:

[0015] The difference between each feature value in the surface image feature set and the corresponding feature value in the defect template is calculated one by one and recorded as a feature value difference;

[0016] When there is a feature value difference greater than or equal to a corresponding feature value difference threshold, it is determined that the aluminum cap to be sorted has defects.

[0017] Otherwise, it is determined that the aluminum cap to be sorted has no defects.

[0018] Further, the comparison module generates a defect label of the aluminum cap to be sorted and calculates the degree of coincidence between the aluminum cap to be sorted and the defect template, and the generation and calculation include:

[0019] The defect label includes a defect type and a defect position.

[0020] The degree of coincidence is obtained by the following formula:

[0021]

[0022] Wherein, R represents the coincidence degree; n represents the total number of features; ωi represents the weight of the ith feature; F1 i represents the value of the aluminum cover image to be sorted on the ith feature; F2 i represents the value of the defect template on the ith feature; Similarity(F1 i, F2 i) represents the similarity measure of the ith feature.

[0023] Further, when the detection module determines whether the aluminum cover to be sorted is qualified according to the coincidence degree, comprising:

[0024] Set the coincidence degree threshold, compare the coincidence degree with the coincidence degree threshold, and determine whether the aluminum cover to be sorted is qualified according to the comparison result;

[0025] When the coincidence degree is greater than or equal to the coincidence degree threshold, it is determined that the aluminum cover to be sorted is unqualified;

[0026] When the coincidence degree is less than the coincidence degree threshold, it is determined that the aluminum cover to be sorted is qualified.

[0027] Further, when the detection module compares the size feature set with the standard template in the standard object database to determine the quality grade of the aluminum cover to be sorted according to the comparison result, comprising:

[0028] The size feature set consists of thickness, diameter and edge flatness of the aluminum cover to be sorted;

[0029] The standard template consists of thickness standard value, diameter standard value and edge flatness standard value;

[0030] The deviation value of each feature value in the size feature set and the corresponding feature value in the standard template is calculated one by one and recorded as the size deviation value;

[0031] According to the size deviation value, the comprehensive deviation value of the aluminum cover to be sorted is calculated by using the weighted average method;

[0032] Compare the comprehensive deviation value with the first comprehensive deviation value and the second comprehensive deviation value, and determine the quality grade of the aluminum cover to be sorted according to the comparison result; wherein the first comprehensive deviation value is less than the second comprehensive deviation value;

[0033] When the comprehensive deviation value is less than or equal to the first comprehensive deviation value, it is determined that the quality grade of the aluminum cover to be sorted is first class;

[0034] When the comprehensive deviation value is greater than the first comprehensive deviation value and less than or equal to the second comprehensive deviation value, it is determined that the quality grade of the aluminum cover to be sorted is second class;

[0035] When the comprehensive deviation value is greater than the second comprehensive deviation value, it is determined that the quality level of the aluminum cap to be sorted is level three.

[0036] Further, when the calibration module determines whether to calibrate the quality level of the aluminum cap to be sorted based on the analysis result, the method comprises:

[0037] determining an environment characteristic value corresponding to the running environment data, and determining a corresponding environment characteristic threshold value;

[0038] If all the environment characteristic values are within the corresponding environment characteristic threshold value, it is determined that the quality level of the aluminum cap to be sorted does not need to be calibrated;

[0039] If one or more environment characteristic values are outside the corresponding environment characteristic threshold value, it is determined that the quality level of the aluminum cap to be sorted needs to be calibrated.

[0040] Further, when the calibration module calculates the environment influence coefficient according to the running environment data, the method comprises:

[0041] determining an environment standard value corresponding to each environment characteristic value;

[0042] extracting all environment characteristic values greater than the environment standard value, and calculating a deviation value of each environment characteristic value and the corresponding environment standard value;

[0043] determining all the deviation values, and generating a first deviation value set according to all the deviation values less than or equal to a preset deviation value;

[0044] generating a second deviation value set according to all the deviation values greater than the preset deviation value;

[0045] calculating a first average deviation value of the first deviation value set;

[0046] calculating a second average deviation value of the second deviation value set;

[0047] calculating the environment influence coefficient based on the first average deviation value and the second average deviation value.

[0048] Further, when the calibration module calibrates the quality level according to the environment influence coefficient to obtain a final quality level, the method comprises:

[0049] comparing the environment influence coefficient with a first environment influence coefficient and a second environment influence coefficient, and calibrating the quality level according to the comparison result; wherein the first environment influence coefficient is less than the second environment influence coefficient;

[0050] When the environment influence coefficient is less than or equal to the first environment influence coefficient, it is determined that the quality level does not need to be adjusted, and is the final quality level;

[0051] when the environmental influence coefficient is greater than the first environmental influence coefficient and less than or equal to the second environmental influence coefficient, the quality grade is downgraded by one level as a final quality grade;

[0052] when the environmental influence coefficient is greater than the second environmental influence coefficient, the quality grade is downgraded by two levels as a final quality grade.

[0053] Compared with the prior art, the beneficial effects of the present application are that the aluminum cap sorting monitoring system based on visual detection provided by the present application can realize efficient and accurate sorting of aluminum caps, and improve sorting efficiency and accuracy. Through feature extraction of image data of the aluminum caps to be sorted by the acquisition and processing module, a surface image feature set of the aluminum caps can be obtained, and then the comparison module is compared with the defect template in the standard object database to determine whether the aluminum cap has defects; whether the aluminum cap is qualified is determined according to the coincidence degree, and the detection module is further used to compare the size features of the qualified aluminum cap to determine its quality grade; the calibration module can also calibrate the quality grade of the aluminum cap according to the running environment data of the visual detection equipment, so as to ensure the accuracy and reliability of the sorting result.

[0054] In another aspect, the present application also provides an aluminum cap sorting monitoring method based on visual detection, comprising the following steps:

[0055] S100: determining aluminum caps to be sorted, acquiring a plurality of image data of the aluminum caps to be sorted within a preset time length by a visual detection device, and forming an image data set; performing feature extraction on the image data set to obtain a surface image feature set of the aluminum caps to be sorted; wherein the surface image feature set is composed of surface feature values, texture feature values and reflection feature values;

[0056] S200: comparing the surface image feature set with a defect template in a standard object database, and determining whether the aluminum caps to be sorted have defects according to the comparison result; if yes, generating a defect label of the aluminum caps to be sorted, and calculating the coincidence degree of the aluminum caps to be sorted and the defect template;

[0057] S300: determining whether the aluminum caps to be sorted are qualified according to the coincidence degree; if yes, obtaining a size feature set of the aluminum caps to be sorted according to the image data set, comparing the size feature set with a standard template in the standard object database, and determining the quality grade of the aluminum caps to be sorted according to the comparison result;

[0058] S400: Collect the running environment data of the visual detection device, analyze the running environment data, and determine whether to calibrate the quality level of the aluminum cap to be sorted based on the analysis result; if yes, calculate the environmental influence coefficient according to the running environment data, calibrate the quality level according to the environmental influence coefficient, and obtain the final quality level.

[0059] It can be understood that the above-mentioned aluminum cap sorting monitoring system and method based on visual detection have the same beneficial effects, which will not be repeated here. BRIEF DESCRIPTION OF DRAWINGS

[0060] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of preferred embodiments and are not meant to limit the present application. Furthermore, the same reference numerals in different drawings identify the same components. In the drawings:

[0061] Figure 1 A structural block diagram of the aluminum cap sorting monitoring system based on visual detection provided by the embodiment of the present application is shown in the figure;

[0062] Figure 2 A flowchart of the aluminum cap sorting monitoring method based on visual detection provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0063] Exemplary embodiments of the present disclosure will be described in greater detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0064] Referring to Figure 1 As shown in the figure, in some embodiments of the present application, the present embodiment provides an aluminum cap sorting monitoring system based on visual detection, which comprises:

[0065] The acquisition processing module is configured to determine the aluminum cap to be sorted, acquire a plurality of image data of the aluminum cap to be sorted within a preset time length by a visual detection device, and form an image data set; perform feature extraction on the image data set to obtain a surface image feature set of the aluminum cap to be sorted; wherein the surface image feature set is composed of surface feature values, texture feature values and reflection feature values;

[0066] The comparison module is configured to compare the surface image feature set with a defect template in a standard object database, judge whether the aluminum cover to be sorted has defects according to a comparison result, if yes, generate a defect label of the aluminum cover to be sorted, and calculate a coincidence degree of the aluminum cover to be sorted and the defect template;

[0067] The detection module is configured to judge whether the aluminum cover to be sorted is qualified according to the coincidence degree, if yes, obtain a size feature set of the aluminum cover to be sorted according to the image data set, compare the size feature set with a standard template in the standard object database, and determine a quality grade of the aluminum cover to be sorted according to a comparison result.

[0068] The calibration module is configured to collect running environment data of the visual detection equipment, analyze the running environment data, judge whether to calibrate the quality grade of the aluminum cover to be sorted based on an analysis result, if yes, calculate an environmental influence coefficient according to the running environment data, calibrate the quality grade according to the environmental influence coefficient, and obtain a final quality grade.

[0069] It can be understood that the aluminum cover sorting monitoring system based on visual detection provided by the embodiment can realize efficient and accurate sorting of aluminum covers, and improve sorting efficiency and accuracy. By feature extraction of image data of the aluminum cover to be sorted by the acquisition and processing module, a surface image feature set of the aluminum cover can be obtained, and then the comparison module is compared with a defect template in a standard object database to judge whether the aluminum cover has defects. According to the coincidence degree, it is judged whether the aluminum cover is qualified, and the detection module further compares the size features of the qualified aluminum cover to determine its quality grade. The calibration module can also calibrate the quality grade of the aluminum cover according to the running environment data of the visual detection equipment to ensure the accuracy and reliability of the sorting result.

[0070] Specifically, when the acquisition and processing module extracts features from the image data set to obtain the surface image feature set of the aluminum cover to be sorted, it includes:

[0071] The image data set is preprocessed, and the preprocessing includes denoising and contrast enhancement;

[0072] Feature points are extracted from the preprocessed image data set by using an image recognition algorithm to obtain contour features and color features of the surface of the aluminum cover to be sorted;

[0073] Based on the contour features and color features, image processing technology is used to calculate surface feature values, texture feature values and reflection feature values of the aluminum cover to be sorted.

[0074] In this embodiment, the surface feature value refers to the quantified value of the physical characteristics such as the concave-convex degree, scratches, and wear of the aluminum cover surface; the texture feature value refers to the quantified value of the complexity and regularity of the pattern and texture on the aluminum cover surface; and the reflection feature value reflects the quantified value of the glossiness and reflection ability of the aluminum cover surface.

[0075] In this embodiment, first, advanced image filtering algorithms are used to remove noise in the image and enhance the contrast of the image, making the surface features of the aluminum cover clearer. Then, through edge detection and contour extraction technology, the contour features of the aluminum cover surface are accurately identified, which helps subsequent feature value calculation. At the same time, color space conversion and color analysis techniques are used to extract the color features of the aluminum cover surface, which is of great significance for judging whether the aluminum cover has color abnormalities or contamination. Based on the contour features and color features, feature matching and image analysis algorithms are used to calculate the surface feature value, texture feature value, and reflection feature value of the aluminum cover. These feature values not only reflect the physical and optical properties of the aluminum cover surface, but also provide key basis for subsequent defect detection and quality grade determination.

[0076] Specifically, when the comparison module compares the surface image feature set with the defect templates in the standard object database to determine whether the aluminum cover to be sorted has defects, the method comprises the following steps:

[0077] The difference between each feature value in the surface image feature set and the corresponding feature value in the defect template is calculated one by one and recorded as a feature value difference;

[0078] When there is a feature value difference greater than or equal to the corresponding feature value difference threshold, it is determined that the aluminum cover to be sorted has defects;

[0079] Otherwise, it is determined that the aluminum cover to be sorted has no defects.

[0080] It can be understood that the calculation of the feature value difference is based on the similarity or difference between the aluminum cover to be sorted and the defect template in each feature dimension. By setting a reasonable feature value difference threshold, sensitive and accurate detection of aluminum cover defects can be achieved. When a certain feature value difference exceeds the threshold, it means that the aluminum cover to be sorted has significant differences with the defect template in that feature dimension, and thus it is determined that the aluminum cover has defects. This defect detection method based on feature value comparison not only improves the accuracy of detection, but also greatly shortens the detection time, achieving rapid identification of aluminum cover defects. At the same time, the system provided in this embodiment can also generate corresponding defect labels according to the specific type and degree of the defect, providing convenience for subsequent processing and classification.

[0081] Specifically, when the comparison module generates the defect label of the aluminum cover to be sorted and calculates the coincidence degree of the aluminum cover to be sorted with the defect template, it includes:

[0082] The defect label includes a defect type and a defect position.

[0083] The coincidence degree is obtained by the following formula:

[0084]

[0085] Wherein, R represents the coincidence degree; n represents the total number of features; ωi represents the weight of the ith feature; F1i represents the value of the aluminum cover image on the ith feature; F2i represents the value of the defect template on the ith feature; Similarity(F1i, F2i) represents the similarity measure of the ith feature.

[0086] In this embodiment, the similarity measure adopts the cosine similarity calculation method, which can measure the directional consistency of the aluminum cover image to be sorted and the defect template in each feature dimension, thereby accurately reflecting the similarity between them.

[0087] It can be understood that the generation of the defect label is crucial for the identification and classification of aluminum cover defects. By recording the type and position of the defect in detail, not only can valuable information be provided for subsequent quality control and process improvement, but also it helps to realize the accurate traceability of the aluminum cover. At the same time, the calculation of the coincidence degree is an important indicator to measure the similarity between the aluminum cover to be sorted and the defect template. By considering the weight and similarity measure of each feature, an objective and accurate coincidence degree value can be obtained, thereby providing strong support for the qualification of the aluminum cover.

[0088] Specifically, when the detection module determines whether the aluminum cover to be sorted is qualified according to the coincidence degree, it includes:

[0089] Set a coincidence degree threshold, compare the coincidence degree with the coincidence degree threshold, and determine whether the aluminum cover to be sorted is qualified according to the comparison result;

[0090] When the coincidence degree is greater than or equal to the coincidence degree threshold, it is determined that the aluminum cover to be sorted is unqualified;

[0091] When the coincidence degree is less than the coincidence degree threshold, it is determined that the aluminum cover to be sorted is qualified.

[0092] It can be understood that the setting of the coincidence threshold is based on the comprehensive consideration of the defect tolerance of the aluminum cover. Through a large number of experiments and data analysis, a reasonable coincidence threshold is determined to ensure the accuracy and reliability of the sorting result. When the coincidence of the aluminum cover to be sorted exceeds the threshold, it means that it is highly similar to the defect template and may have obvious defects, so it is judged as unqualified. On the contrary, when the coincidence is lower than the threshold, it is considered that the aluminum cover to be sorted is significantly different from the defect template, and it can be judged as qualified and continue to perform subsequent quality level judgment. This qualified judgment method based on coincidence not only improves the accuracy of sorting, but also enhances the adaptability and flexibility of the system, which can cope with multiple same types and degrees of aluminum cover defects.

[0093] Specifically, when the detection module compares the size feature set with the standard template in the standard object database and determines the quality level of the aluminum cover to be sorted according to the comparison result, it includes:

[0094] The size feature set consists of the thickness, diameter and edge flatness of the aluminum cover to be sorted;

[0095] The standard template consists of a thickness standard value, a diameter standard value and an edge flatness standard value;

[0096] The deviation value of each feature value in the size feature set from the corresponding feature value in the standard template is calculated one by one and recorded as a size deviation value;

[0097] According to the size deviation value, a weighted average method is used to calculate the comprehensive deviation value of the aluminum cover to be sorted;

[0098] The comprehensive deviation value is compared with the first comprehensive deviation value and the second comprehensive deviation value, and the quality level of the aluminum cover to be sorted is determined according to the comparison result; wherein the first comprehensive deviation value is less than the second comprehensive deviation value;

[0099] When the comprehensive deviation value is less than or equal to the first comprehensive deviation value, the quality level of the aluminum cover to be sorted is determined as level one;

[0100] When the comprehensive deviation value is greater than the first comprehensive deviation value and less than or equal to the second comprehensive deviation value, the quality level of the aluminum cover to be sorted is determined as level two;

[0101] When the comprehensive deviation value is greater than the second comprehensive deviation value, the quality level of the aluminum cover to be sorted is determined as level three.

[0102] In the embodiment, the levels from low to high are three, two and one, wherein three represents that the size deviation of the aluminum cover is relatively large and the quality is relatively low; two represents that the size deviation of the aluminum cover is moderate and the quality is moderate; one represents that the size deviation of the aluminum cover is the smallest and the quality is the highest.

[0103] It can be understood that the quality level determination method based on the comprehensive deviation value can objectively and accurately reflect the actual quality condition of the aluminum cover, and provides an important basis for subsequent quality control and use. At the same time, by setting different comprehensive deviation value thresholds, flexible sorting of aluminum covers with different quality requirements can be realized, and the applicability and practicability of the system are improved.

[0104] Specifically, when the calibration module determines whether to calibrate the quality level of the aluminum cover to be sorted based on the analysis result, the method comprises:

[0105] determining the environment characteristic value corresponding to the running environment data, and determining the corresponding environment characteristic threshold value;

[0106] If all the environment characteristic values are within the corresponding environment characteristic threshold value, it is determined that the quality level of the aluminum cover to be sorted is not calibrated;

[0107] If one or more environment characteristic values are outside the corresponding environment characteristic threshold value, it is determined that the quality level of the aluminum cover to be sorted is calibrated.

[0108] In the embodiment, the running environment data includes temperature, humidity and light intensity.

[0109] It can be understood that the running environment data of the visual detection device has an important influence on the determination of the quality level of the aluminum cover. Changes in environmental factors such as temperature, humidity and light intensity may cause errors in the image acquisition and processing process, thereby affecting the final sorting result. Therefore, by using the calibration module to monitor and analyze the running environment data in real time, these errors can be found and corrected in time, and the accuracy and reliability of the sorting result can be ensured. When all the environment characteristic values are within the corresponding environment characteristic threshold value, it is considered that the current environment has little influence on the quality level of the aluminum cover, and the quality level can not be calibrated. However, when one or more environment characteristic values exceed the corresponding environment characteristic threshold value, it means that the current environment may have a significant influence on the determination of the quality level of the aluminum cover, and the quality level needs to be calibrated.

[0110] Specifically, when the calibration module calculates the environment influence coefficient according to the running environment data, the method comprises:

[0111] determining the environment standard value corresponding to each environment characteristic value;

[0112] extracting all environmental characteristic values greater than the environmental standard values, and calculating a deviation value of each environmental characteristic value from a corresponding environmental standard value;

[0113] determining all deviation values, and generating a first deviation value set according to all deviation values less than or equal to a preset deviation value;

[0114] generating a second deviation value set according to all deviation values greater than the preset deviation value;

[0115] calculating a first average deviation value of the first deviation value set;

[0116] calculating a second average deviation value of the second deviation value set;

[0117] calculating an environmental influence coefficient based on the first average deviation value and the second average deviation value.

[0118] In this embodiment, the environmental influence coefficient is obtained by weighted average calculation of the first average deviation value and the second average deviation value.

[0119] It can be understood that the calculation of the environmental influence coefficient fully considers the influence degree of different environmental characteristic values on the aluminum cover quality grade determination. By respectively calculating the first average deviation value and the second average deviation value, the fluctuation of environmental factors can be more carefully measured. The first average deviation value reflects the average contribution of those characteristic values with small deviation from the environmental standard value to the overall environmental influence, while the second average deviation value reflects the significant influence of characteristic values with large deviation on the environment. Through weighted average calculation, the environmental influence coefficient can comprehensively reflect these factors, thereby providing a scientific basis for quality grade calibration. When the environmental influence coefficient is large, it means that the current environment has a more significant influence on the aluminum cover quality grade, and a larger calibration adjustment is needed; when the environmental influence coefficient is small, it means that the environment has a smaller influence on the quality grade, and the calibration amplitude can also be correspondingly reduced. This calibration method based on the environmental influence coefficient not only improves the accuracy of calibration, but also enhances the adaptability and robustness of the system, and can meet the aluminum cover sorting needs in different environments.

[0120] Specifically, when the calibration module calibrates the quality grade according to the environmental influence coefficient to obtain a final quality grade, the method comprises:

[0121] comparing the environmental influence coefficient with a first environmental influence coefficient and a second environmental influence coefficient, and calibrating the quality grade according to the comparison result; wherein the first environmental influence coefficient is smaller than the second environmental influence coefficient;

[0122] when the environmental influence coefficient is smaller than or equal to the first environmental influence coefficient, it is determined that the quality grade does not need to be adjusted, and the quality grade is the final quality grade;

[0123] When the environmental influence coefficient is greater than the first environmental influence coefficient and less than or equal to the second environmental influence coefficient, the quality grade is downgraded by one level as the final quality grade;

[0124] When the environmental influence coefficient is greater than the second environmental influence coefficient, the quality grade is downgraded by two levels as the final quality grade.

[0125] It can be understood that when the quality grade is three, it does not need to be downgraded, and the lowest level of the final quality grade is three. When the environmental influence coefficient is at a low level, i.e., less than or equal to the first environmental influence coefficient, it means that the current environment has little influence on the quality grade of the aluminum cover, so there is no need to adjust the quality grade, and the current quality grade is directly taken as the final quality grade. This approach not only ensures the accuracy of the sorting result, but also avoids unnecessary calibration operations, improving work efficiency. However, when the environmental influence coefficient exceeds the first environmental influence coefficient but is less than or equal to the second environmental influence coefficient, it means that the influence of the current environment on the quality grade of the aluminum cover begins to appear, but has not yet reached a serious degree of influence. At this time, in order to ensure the reliability of the sorting result, the system will downgrade the quality grade by one level as the final quality grade. This approach takes into account the influence of environmental factors and ensures the reasonableness of the sorting result. Finally, when the environmental influence coefficient is greater than the second environmental influence coefficient, it means that the influence of the current environment on the quality grade of the aluminum cover is very significant and has seriously interfered with the accuracy of the sorting result. In this case, the system will downgrade the quality grade by two levels as the final quality grade. This large adjustment aims to ensure that a relatively accurate sorting result can be obtained even in a harsh environment, thereby improving the robustness and practicality of the system.

[0126] Referring to Figure 2 In some embodiments of the present application, the present embodiment provides a visual detection-based aluminum cover sorting monitoring method, comprising the following steps:

[0127] S100: determining a to-be-sorted aluminum cover, acquiring a plurality of image data of the to-be-sorted aluminum cover within a preset time length by a visual detection device, and constructing an image data set; performing feature extraction on the image data set to obtain a surface image feature set of the to-be-sorted aluminum cover; wherein the surface image feature set is composed of a surface feature value, a texture feature value, and a reflection feature value;

[0128] S200: comparing the surface image feature set with a defect template in a standard object database, judging whether the to-be-sorted aluminum cover has defects according to the comparison result; if yes, generating a defect label of the to-be-sorted aluminum cover, and calculating the coincidence degree of the to-be-sorted aluminum cover and the defect template;

[0129] S300: judging whether the aluminum cap to be sorted is qualified according to the coincidence degree; if yes, obtaining a size feature set of the aluminum cap to be sorted according to the image data set, comparing the size feature set with a standard template in the standard object database, and determining a quality grade of the aluminum cap to be sorted according to a comparison result;

[0130] S400: collecting running environment data of the visual detection equipment, analyzing the running environment data, judging whether to calibrate the quality grade of the aluminum cap to be sorted based on an analysis result; if yes, calculating an environmental influence coefficient according to the running environment data, calibrating the quality grade according to the environmental influence coefficient, and obtaining a final quality grade.

[0131] Those skilled in the art will understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. In addition, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0132] The present application is described with reference to flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device implemented in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks. Figure 1 The functions specified in one or more flows and / or blocks.

[0133] These computer program instructions can also be stored in a computer-readable memory capable of guiding a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including instruction devices, which implement the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks. Figure 1 The functions specified in one or more flows and / or blocks.

[0134] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 Figure 1

[0135] Finally, it should be noted that the above examples are merely intended to describe the technical solutions of the present application, rather than limiting the same. Although the present application has been described in detail with reference to the above examples, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and any modification or equivalent replacement without departing from the spirit and scope of the present application should be covered within the protection scope of the claims of the present application.​​

Claims

1. A visual inspection-based aluminum cap sorting and monitoring system, characterized in that, include: The acquisition and processing module is configured to determine the aluminum cap to be sorted, acquire several image data of the aluminum cap to be sorted within a preset time period through a visual inspection device, and form an image dataset; perform feature extraction on the image dataset to obtain a surface image feature set of the aluminum cap to be sorted; wherein, the surface image feature set consists of surface feature values, texture feature values ​​and reflectivity feature values; The comparison module is configured to compare the surface image feature set with the defect template in the standard object database, and determine whether the aluminum cap to be sorted has a defect based on the comparison result; if so, a defect label is generated for the aluminum cap to be sorted, and the overlap between the aluminum cap to be sorted and the defect template is calculated. The detection module is configured to determine whether the aluminum cap to be sorted is qualified based on the overlap degree; if so, it obtains the size feature set of the aluminum cap to be sorted based on the image dataset, compares the size feature set with the standard template in the standard object database, and determines the quality level of the aluminum cap to be sorted based on the comparison result. The calibration module is configured to collect the operating environment data of the visual inspection device, analyze the operating environment data, and determine whether to calibrate the quality grade of the aluminum cover to be sorted based on the analysis result; if so, it calculates the environmental impact coefficient based on the operating environment data, calibrates the quality grade based on the environmental impact coefficient, and obtains the final quality grade. The comparison module compares the surface image feature set with the defect templates in the standard object database. When determining whether the aluminum cap to be sorted has defects based on the comparison results, the module includes: Calculate the difference between each feature value in the surface image feature set and the corresponding feature value in the defect template, and record it as the feature value difference; When there is a difference in the characteristic value that is greater than or equal to the corresponding characteristic value difference threshold, it is determined that the aluminum cap to be sorted has a defect; Otherwise, the aluminum cap to be sorted is determined to be free of defects; When the comparison module generates defect labels for the aluminum caps to be sorted and calculates the overlap between the aluminum caps to be sorted and the defect template, it includes: The defect label includes the defect type and defect location; The degree of overlap is obtained by the following formula: Where R represents the overlap; n represents the total number of features; ωi represents the weight of the i-th feature; F1i represents the value of the aluminum cap image to be sorted on the i-th feature; F2i represents the value of the defect template on the i-th feature; Similarity(F1i,F2i) represents the similarity measure of the i-th feature. When the detection module determines whether the aluminum cap to be sorted is qualified based on the overlap, it includes: Set an overlap threshold, compare the overlap with the overlap threshold, and determine whether the aluminum cap to be sorted is qualified based on the comparison result. When the overlap is greater than or equal to the overlap threshold, the aluminum cap to be sorted is determined to be unqualified. When the overlap is less than the overlap threshold, the aluminum cap to be sorted is determined to be qualified. The detection module compares the set of size features with standard templates in the standard object database, and determines the quality grade of the aluminum cap to be sorted based on the comparison results, including: The set of dimensional features consists of the thickness, diameter, and edge flatness of the aluminum cap to be sorted; The standard template consists of standard values ​​for thickness, diameter, and edge flatness. Calculate the deviation value between each feature value in the size feature set and the corresponding feature value in the standard template, and record it as the size deviation value; Based on the dimensional deviation value, the comprehensive deviation value of the aluminum cap to be sorted is calculated using the weighted average method. The overall deviation value is compared with the first overall deviation value and the second overall deviation value, and the quality grade of the aluminum cap to be sorted is determined according to the comparison result; wherein, the first overall deviation value is less than the second overall deviation value; When the comprehensive deviation value is less than or equal to the first comprehensive deviation value, the quality grade of the aluminum cap to be sorted is determined to be Grade 1; When the overall deviation value is greater than the first overall deviation value and less than or equal to the second overall deviation value, the quality grade of the aluminum cap to be sorted is determined to be Grade II; When the overall deviation value is greater than the second overall deviation value, the quality grade of the aluminum cap to be sorted is determined to be level three.

2. The aluminum cap sorting and monitoring system based on visual inspection according to claim 1, characterized in that, When the acquisition and processing module extracts features from the image dataset to obtain the surface image feature set of the aluminum cap to be sorted, it includes: The image dataset is preprocessed, including denoising and contrast enhancement. Image recognition algorithms are used to extract feature points from the preprocessed image dataset to obtain the contour and color features of the aluminum cap surface to be sorted. Based on the contour and color features, image processing techniques are used to calculate the surface feature values, texture feature values, and reflectivity feature values ​​of the aluminum caps to be sorted.

3. The aluminum cap sorting and monitoring system based on visual inspection according to claim 1, characterized in that, When the calibration module determines whether to calibrate the quality grade of the aluminum cap to be sorted based on the analysis results, it includes: Determine the environmental feature values ​​corresponding to the operating environment data, and determine the corresponding environmental feature thresholds; If all environmental characteristic values ​​are within the corresponding environmental characteristic threshold, it is determined that the quality grade of the aluminum cap to be sorted will not be calibrated. If one or more environmental characteristic values ​​are outside the corresponding environmental characteristic threshold, then it is determined that the quality grade of the aluminum cap to be sorted needs to be calibrated.

4. The aluminum cap sorting and monitoring system based on visual inspection according to claim 1, characterized in that, When the calibration module calculates the environmental impact coefficient based on the operating environment data, it includes: Determine the corresponding environmental standard value for each environmental characteristic value; Extract all environmental feature values ​​that are greater than the environmental standard value, and calculate the deviation value between each environmental feature value and the corresponding environmental standard value; Determine all deviation values ​​and generate a first set of deviation values ​​based on all deviation values ​​that are less than or equal to a preset deviation value; A second set of deviation values ​​is generated based on all deviation values ​​that are greater than the preset deviation value; Calculate the first average deviation value of the first set of deviation values; Calculate the second average deviation value of the second set of deviation values; The environmental impact coefficient is calculated based on the first average deviation value and the second average deviation value.

5. The visual inspection-based aluminum cap sorting and monitoring system according to claim 4, characterized in that, The calibration module calibrates the quality level according to the environmental impact coefficient to obtain the final quality level, including: The environmental impact coefficient is compared with the first environmental impact coefficient and the second environmental impact coefficient, and the quality level is calibrated based on the comparison result; wherein the first environmental impact coefficient is smaller than the second environmental impact coefficient. When the environmental impact coefficient is less than or equal to the first environmental impact coefficient, it is determined that the quality level does not need to be adjusted, which is the final quality level; When the environmental impact coefficient is greater than the first environmental impact coefficient and less than or equal to the second environmental impact coefficient, the quality level will be downgraded by one level as the final quality level. When the environmental impact coefficient is greater than the second environmental impact coefficient, the quality level will be downgraded by two levels to become the final quality level.

6. A visual inspection-based aluminum cap sorting and monitoring method, applied to the visual inspection-based aluminum cap sorting and monitoring system as described in any one of claims 1-5, characterized in that, include: The aluminum caps to be sorted are identified, and a number of image data of the aluminum caps to be sorted are collected by a visual inspection device within a preset time period, and an image dataset is formed. Feature extraction is performed on the image dataset to obtain a surface image feature set of the aluminum cap to be sorted; wherein, the surface image feature set consists of surface feature values, texture feature values, and reflectivity feature values; The surface image feature set is compared with the defect template in the standard object database. Based on the comparison result, it is determined whether the aluminum cap to be sorted has a defect. If so, a defect label is generated for the aluminum cap to be sorted, and the overlap between the aluminum cap to be sorted and the defect template is calculated. The overlap is used to determine whether the aluminum cap to be sorted is qualified; if so, the size feature set of the aluminum cap to be sorted is obtained from the image dataset, the size feature set is compared with the standard template in the standard object database, and the quality grade of the aluminum cap to be sorted is determined based on the comparison result. The operating environment data of the visual inspection device is collected and analyzed. Based on the analysis results, it is determined whether the quality grade of the aluminum cover to be sorted should be calibrated. If so, the environmental impact coefficient is calculated based on the operating environment data, the quality grade is calibrated based on the environmental impact coefficient, and the final quality grade is obtained.

Images