Aluminum profile extrusion forming monitoring method and system based on machine vision

Abstract

The invention relates to the field of image processing, in particular to an aluminum profile extrusion forming monitoring method and system based on machine vision. The method comprises the following steps: obtaining finished images at a plurality of angles; taking the finished product image at any angle as a reference image, and obtaining a distribution model of each pixel in the reference image; obtaining a plurality of category combinations according to the distribution model, and obtaining an adjusted reference image based on the category combinations; calculating a signal description capability; taking the adjusted reference image with the maximum signal description capability as a first corrected image; obtaining a plurality of sub-category combinations, and obtaining an adjusted reference image based on the sub-category combinations; calculating the signal description capability, and taking the adjusted reference image with the maximum signal description capability as a second corrected image; and so on, until the signal description capability converges, obtaining a final corrected image. And aluminum profile extrusion forming monitoring is achieved based on the final corrected image. And the accuracy of aluminum profile extrusion forming monitoring is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing, and in particular to a monitoring method and system for aluminum profile extrusion molding based on machine vision. Background Art

[0002] During the extrusion molding process of aluminum profiles, various quality problems are likely to occur if the control parameters are set improperly. For example, improper setting of temperature parameters may cause pitting or pitted surfaces on the surface of aluminum profiles, and these pitting or pitted surfaces will affect the appearance and use effect of aluminum profiles. Therefore, it is necessary to monitor the quality of the aluminum profile extrusion molding process, so that when quality problems occur, the control parameters can be adjusted in time to prevent the continuous occurrence of quality problems and cause more losses.

[0003] The finished products obtained by aluminum profile extrusion molding generally have a relatively high surface smoothness and strong metallic luster. Therefore, reflective areas are likely to appear on the surface of the finished products obtained by aluminum profile extrusion molding. Such reflective areas will affect the accuracy of quality monitoring. For example, the brightness at the pitting areas is relatively large and has certain texture information, and this information has similar characteristics to the reflective areas. Therefore, when extracting defects through color and texture, the reflective areas are easily misjudged as defect areas. In order to improve the accuracy of quality monitoring, it is necessary to remove the reflective areas on the surface of the aluminum profile extrusion molding before quality monitoring. Therefore, how to accurately remove the reflective areas on the surface of aluminum profiles has become the research focus of the present invention. The patent document with the publication number CN118122804B discloses a remote monitoring and intelligent diagnosis aluminum profile extrusion device and control method. The method in this patent document mainly shows the process of fault diagnosis of the aluminum profile extrusion device and does not involve the content of how to remove the reflective areas. Therefore, the method in this patent document cannot solve the technical problems in the present invention. Summary of the Invention

[0004] In order to solve the problem of how to accurately remove the reflective areas on the surface of aluminum profiles, the present invention provides a monitoring method and system for aluminum profile extrusion molding based on machine vision.

[0005] In a first aspect, the present invention provides a monitoring method for aluminum profile extrusion molding based on machine vision, adopting the following technical solution: The monitoring method for aluminum profile extrusion molding based on machine vision includes the steps of: Collecting finished product images of aluminum profile extrusion molding at several angles; Taking the finished product image at any angle as the reference image, matching the finished product images at other angles with the reference image, obtaining the set of pixel values of each pixel in the reference image according to the pixel matching relationship, and fitting a distribution model to the set of pixel values; performing clustering processing on the distribution model to obtain several categories, combining the categories of all pixels in the reference image to obtain several category combinations, and taking the median of each category in the category combination as the pixel value of the corresponding pixel in the reference image to obtain the adjusted reference image based on this category combination; calculating the signal description ability , comparing the edge pixels in the adjusted reference image with the edge pixels in the structural design image to set the difference flag values of each edge pixel in the structural design drawing, and classifying the pixels in the adjusted reference image into structural pixels and non-structural pixels according to the difference flag values. S represents the number of edge pixels with a difference flag value of 0 in the structural design drawing, represents the gradient value of the i-th non-structural pixel in the adjusted reference image, N represents the number of non-structural pixels in the adjusted reference image, M represents the signal description ability of the adjusted reference image, and tanh() represents the hyperbolic tangent function; taking the adjusted reference image with the maximum signal description ability as the first corrected image; Recording the categories in the category combination corresponding to the first corrected image as the reference categories; performing clustering processing on the data in the reference categories to obtain several sub-categories, combining the sub-categories of all pixels in the reference image to obtain several sub-category combinations, and taking the median of each sub-category in the sub-category combination as the pixel value of the corresponding pixel in the reference image to obtain the adjusted reference image based on this sub-category combination; calculating the signal description ability, and taking the adjusted reference image with the maximum signal description ability as the second corrected image; and so on until the signal description ability converges to obtain the final corrected image; Realizing the monitoring of aluminum profile extrusion forming based on the final corrected image.

[0006] The present invention selects the pixel value with the least light interference among the pixel values at multiple angles, thereby effectively removing the reflective area in the finished image and providing a basis for subsequent accurate defect detection. Further, when selecting the pixel value with the least light interference among the pixel values at multiple angles, the single pixel value is replaced by a category, thereby effectively reducing the number of pixel value combinations and effectively improving the analysis efficiency. Further, when selecting the pixel value with the least light interference among the pixel values at multiple angles, the situation where the adjusted reference image obtained from the selected pixel value accurately describes the finished product information is analyzed, so as to accurately evaluate the selected pixel value each time and provide a basis for accurately selecting the pixel value with the least light interference. Further, when analyzing the situation where the adjusted reference image obtained from the selected pixel value accurately describes the finished product information, the integrity of the structural information that can be extracted from the adjusted reference image is accurately evaluated by introducing the number of edge pixels with a difference flag value of 0. At the same time, the situation where the adjusted reference image is affected by light interference is accurately evaluated by introducing the gradient value of non-structural pixels, providing a basis for accurately evaluating the ability to describe the finished product information in the adjusted reference image. Further, when selecting the pixel value with the least light interference among the pixel values at multiple angles, through layer-by-layer clustering analysis, the positioning range of the pixel value with the least light interference is gradually narrowed, so as to select the pixel value with the least light interference faster and more accurately.

[0007] Preferably, obtaining the set of pixel values of each pixel in the reference image according to the matching relationship of the pixels, and fitting a distribution model to the set of pixel values includes: Denote any pixel in the reference image as the target pixel, obtain the image containing the target pixel in the finished images at all other angles as the matching image, obtain the pixel matching the target pixel in all matching images as the matching pixel, and use the set composed of the pixel values of the target pixel and all matching pixels as the set of pixel values of the target pixel; Fit a Gaussian model to all pixel values in the set of pixel values of the target pixel and denote it as the distribution model of the target pixel.

[0008] The present invention describes by analyzing the distribution characteristics of the pixel values of the corresponding pixels at multiple angles, providing a basis for accurately positioning the pixel value with the least light interference subsequently.

[0009] Preferably, combining the categories of all pixels in the reference image to obtain several category combinations, and using the median of each category in the category combination as the pixel value of the corresponding pixel in the reference image to obtain the adjusted reference image based on this category combination includes: Randomly select a category from all the categories of each pixel in the reference image, combine the categories selected by all pixels in the reference image together to obtain a category combination, and obtain all category combinations; The median of all data for each category is obtained and recorded as the reference data for each category. The reference data for each category in any category combination is used as the pixel value of the corresponding pixel to obtain an adjusted reference image based on the category combination.

[0010] Preferably, the comparison and analysis of the edge pixels in the adjusted reference image and the structural design image to set the difference flag value of each edge pixel in the structural design drawing includes: Perform edge detection on the adjusted reference image to obtain an edge image. The edge image and the adjusted reference image are respectively matched with the structural design drawing, and based on the matching results, the edge image and the adjusted reference image are adjusted into images with the same perspective and scale as the structural design drawing; The edge pixels in the edge image after perspective and scale adjustment are recorded as comparison pixels. Obtain the preset window of any edge pixel in the structural design drawing, and determine whether there are comparison pixels in the corresponding area of the edge image for the preset window of each edge pixel. If there are, set the difference flag value of the edge pixel in the structural design drawing to 1, otherwise set the difference flag value to 0.

[0011] The present invention uses the structural design image for comparison and accurately sets the difference flag value according to the difference between pixels, providing a basis for accurately analyzing the information description ability. Preferably, the pixels in the adjusted reference image are divided into structural pixels and non-structural pixels according to the difference flag value, including: The pixels in the corresponding area of the adjusted reference image after perspective and scale adjustment for the preset window of each edge pixel are recorded as structural pixels, and vice versa as non-structural pixels.

[0012] Preferably, the obtaining of the preset window of any edge pixel in the structural design drawing includes: In the structural design drawing, a window with a length of L is obtained in the vertical direction of the edge line where the edge pixel is located with the edge pixel as the center, and is recorded as the window of the edge pixel in the structural design drawing. L represents the preset size.

[0013] Preferably, the "until the signal description ability converges" includes: Calculate the absolute value of the difference between each signal description ability and the previous signal description ability. If the absolute value of the difference is less than the preset difference threshold, it is considered that the signal description ability converges.

[0014] The present invention judges the convergence situation through the absolute value of the difference, and this implementation method is relatively simple and has high implementation efficiency.

[0015] Preferably, the monitoring of the aluminum profile extrusion forming based on the final corrected image includes: Perform edge detection on the final corrected image. If the edge pixel in the edge image of the final corrected image is a non-structural pixel, mark the edge pixel as a defective pixel; If the number of defective pixels is greater than a preset number, an abnormal warning is issued.

[0016] In a second aspect, the present invention provides a monitoring system for the extrusion forming of aluminum profiles based on machine vision, adopting the following technical solutions: The monitoring system for the extrusion forming of aluminum profiles based on machine vision includes: a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned monitoring method for the extrusion forming of aluminum profiles based on machine vision is implemented.

[0017] By adopting the above technical solutions, the above-mentioned monitoring method for the extrusion forming of aluminum profiles based on machine vision is generated into a computer program and stored in the memory to be loaded and executed by the processor, so as to manufacture a terminal device according to the memory and the processor, which is convenient to use.

[0018] The present invention has the following technical effects: The present invention selects the pixel value with the least light interference from the pixel values at multiple angles, thereby effectively removing the reflective area in the finished product image and providing a basis for subsequent accurate defect detection; Further, when selecting the pixel value with the least light interference from the pixel values at multiple angles, the single pixel value is replaced by a category, thereby effectively reducing the number of pixel value combinations and effectively improving the analysis efficiency; Further, when selecting the pixel value with the least light interference from the pixel values at multiple angles, the situation where the adjusted reference image obtained from the selected pixel value accurately describes the finished product information is analyzed, so as to accurately evaluate the selected pixel value each time and provide a basis for accurately selecting the pixel value with the least light interference; Further, when analyzing the situation where the adjusted reference image obtained from the selected pixel value each time accurately describes the finished product information, the integrity of the structural information that can be extracted from the adjusted reference image is accurately evaluated by introducing the number of edge pixels with a difference flag value of 0, and at the same time, the situation of the adjusted reference image being affected by light interference is accurately evaluated by introducing the gradient value of non-structural pixels, providing a basis for accurately evaluating the ability of the adjusted reference image to describe the finished product information; Further, when selecting the pixel value with the least light interference from the pixel values at multiple angles, through layer-by-layer clustering analysis, the positioning range of the pixel value with the least light interference is gradually narrowed, so as to select the pixel value with the least light interference faster and more accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become readily understood. In the drawings, several embodiments of the present invention are shown in an exemplary but not restrictive manner, and the same or corresponding reference numerals denote the same or corresponding parts.

[0020] Figure 1 is a flowchart of the method in the method for monitoring the extrusion forming of aluminum profiles based on machine vision according to an embodiment of the present invention. Specific embodiments

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0022] It should be understood that when the claims, specifications, and drawings of the present invention use terms such as "first" and "second", they are only used to distinguish different objects and not to describe a specific order. The terms "comprising" and "including" used in the specifications and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0023] An embodiment of the present invention discloses a method for monitoring the extrusion forming of aluminum profiles based on machine vision. Referring to Figure 1 , it includes steps S1 - S4: S1: Collect finished product images of the extrusion forming of aluminum profiles at several angles.

[0024] Specifically, collecting the images of the finished products of the extrusion forming of aluminum profiles at several angles is denoted as the finished product images of each angle.

[0025] S2: Take the finished product image at any one angle as the reference image, perform matching processing on the finished product images at other angles and the reference image, obtain the set of pixel values of each pixel in the reference image according to the pixel matching relationship, and fit a distribution model to the set of pixel values; perform clustering processing on the distribution model to obtain several categories, combine the categories of all pixels in the reference image to obtain several category combinations, take the median of each category in the category combination as the pixel value of the corresponding pixel in the reference image to obtain the adjusted reference image based on the category combination; calculate the signal description ability; take the adjusted reference image with the maximum signal description ability as the first corrected image.

[0026] It should be noted that, due to the different relative angles between the camera and the finished product, the light reflected from the finished product into the camera is different. Therefore, the positions of the reflective areas in the finished product images at different angles are different. In other words, a position on the finished product appears as a reflective phenomenon at some angles and does not appear as a reflective phenomenon at other angles. Non-reflective pixels can be extracted from the finished product images at some of the multiple angles. Therefore, the reflective areas can be removed based on this.

[0027] It should be further noted that it is still impossible to determine which angle's pixel information should be selected among the multiple angles so that it can effectively remove the reflective information and at the same time better preserve the useful information of the finished product. Therefore, it is necessary to analyze the descriptive ability of the pixel information selected for each angle, so as to select better pixel information.

[0028] S20: Take the finished product image at any one angle as the reference image, perform matching processing on the finished product images at other angles and the reference image, obtain the set of pixel values of each pixel in the reference image according to the pixel matching relationship, and fit a distribution model to the set of pixel values.

[0029] It should be noted that since one position corresponds to the pixel information at multiple angles, if all the pixel information at all angles for each position is traversed and analyzed, a large number of combination methods will appear, which will greatly increase the amount of calculation. For example, if one position corresponds to the pixel values at 36 angles and there are 10,000 positions in total, then there will be kinds of combination methods. To reduce the amount of calculation, the pixel information can be roughly classified first according to the distribution characteristics of the pixel information at all angles for one position, and the classification results are combined, which will greatly reduce the number of combinations and thus reduce the amount of calculation. For example, if the pixel values at 36 angles for one position are clustered into 3 categories, then there will only be kinds of combination methods, which will greatly reduce the amount of calculation. First, obtain the distribution model of the pixel values at all angles for one position.

[0030] Preferably, as an example, take the finished product image at any one angle as the reference image, perform matching processing on the finished product images at other angles and the reference image, obtain the set of pixel values of each pixel in the reference image according to the pixel matching relationship, and fit a distribution model to the set of pixel values, including: Take the finished product image at any one angle as the reference image, perform matching processing on the finished product images at other angles and the reference image. Denote any one pixel in the reference image as the target pixel, obtain the image containing the target pixel in all the finished product images at other angles as the matching image, obtain the pixel matching the target pixel in all the matching images as the matching pixel, and take the set composed of the pixel value of the target pixel and all the matching pixels as the set of pixel values of the target pixel.

[0031] Fit a Gaussian model to all pixel values in the set of pixel values of the target pixel, which is denoted as the distribution model of the target pixel; in the same way, obtain the distribution models of all pixels in the reference image.

[0032] S21: Perform clustering on the distribution model to obtain several categories, combine the categories of all pixels in the reference image to obtain several category combinations, and use the median of each category in the category combination as the pixel value of the corresponding pixel in the reference image to obtain the adjusted reference image based on this category combination.

[0033] Preferably, as an example, performing clustering on the distribution model to obtain several categories, combining the categories of all pixels in the reference image to obtain several category combinations, and using the median of each category in the category combination as the pixel value of the corresponding pixel in the reference image to obtain the adjusted reference image based on this category combination, includes: Use the DBSCAN clustering algorithm to perform clustering on the distribution model to obtain several categories; Randomly select one category from all the categories of each pixel in the reference image, combine the categories selected by all pixels in the reference image to obtain a category combination, and obtain all category combinations in the same way; Obtain the median of all data of each category, which is denoted as the reference data of each category, and use the reference data of each category in any category combination as the pixel value of the corresponding pixel to obtain the adjusted reference image based on this category combination.

[0034] It can be understood that through clustering, the number of category combinations can be effectively reduced, and at the same time, the number of adjusted reference images is also reduced, which will effectively improve the calculation efficiency.

[0035] S22: Calculate the signal description ability.

[0036] It should be noted that since the defect area is generally small, when there is no light interference, except for the mechanical information of the finished product, the pixel value differences at other positions are small; at the same time, when there is no light interference, the structural information can be extracted more completely. Therefore, it can be judged based on this that the extracted pixel information is not affected by light interference and the description ability of the finished product information.

[0037] Preferably, as an example, calculating the signal description ability includes: Obtain the structural design drawing of the finished product of aluminum profile extrusion molding; Perform edge detection on the adjusted reference image to obtain an edge image, perform matching processing on the edge image and the adjusted reference image with the structural design drawing respectively, and adjust the edge image and the adjusted reference image to the same perspective and scale as the structural design drawing based on the matching results.

[0038] The edge pixels in the edge image after perspective scale adjustment are denoted as comparison pixels. Obtain the preset window of any edge pixel in the structural design drawing, and determine whether there are comparison pixels in the corresponding area of the edge image for the preset window of each edge pixel. If there are, set the difference flag value of the edge pixel in the structural design drawing to 1; otherwise, set the difference flag value to 0. Denote the pixels in the corresponding area of the adjusted reference image after perspective scale adjustment for the preset window of each edge pixel as structural pixels, and otherwise as non-structural pixels; The signal description ability satisfies the relation:

[0039] where S represents the number of edge pixels with a difference flag value of 0 in the structural design drawing, represents the gradient value of the i-th non-structural pixel in the adjusted reference image after perspective scale adjustment, N represents the number of non-structural pixels in the adjusted reference image after perspective scale adjustment, M represents the signal description ability of the adjusted reference image, and tanh() represents the hyperbolic tangent function.

[0040] It can be understood that the number of edge pixels with a difference flag value of 0 in the structural design drawing reflects the amount of missing structural information in the adjusted reference image. The pixel information extracted this time cannot effectively exclude light interference and has a poor ability to describe the finished product information; reflects the gradient information situation at non-structural pixels in the edge image. The larger this value, the more non-structural pixels with relatively large gradients exist in the edge image. This relatively large gradient value is very likely caused by light interference. Therefore, the pixel information extracted this time cannot effectively exclude light interference and has a poor ability to describe the finished product information.

[0041] It should be added that the method for obtaining the preset window includes: In the structural design drawing, take the length L in the vertical direction of the edge line where each edge pixel is located as the window of each edge pixel in the structural design drawing, with L representing the preset size. In this embodiment, the preset size is taken as 5 for description, and other values can be taken in other embodiments, which are not specifically limited in this embodiment.

[0042] S23: Take the adjusted reference image with the maximum signal description ability as the first corrected image.

[0043] Preferably, as an example, taking the adjusted reference image with the maximum signal description ability as the first corrected image includes: Obtain the adjusted reference image with the maximum signal description ability among all the adjusted reference images obtained by category combinations as the first corrected image.

[0044] S3: Denote the categories in the category combination corresponding to the first corrected image as reference categories; perform clustering on the data in the reference categories to obtain several sub-categories, combine the sub-categories of all pixels in the reference image to obtain several sub-category combinations, and use the median of each sub-category in the sub-category combination as the pixel value of the corresponding pixel in the reference image to obtain an adjusted reference image based on this sub-category combination; calculate the signal description ability, and use the adjusted reference image with the maximum signal description ability as the second corrected image; and so on until the signal description ability converges to obtain the final corrected image.

[0045] It should be noted that the above process uses the median of a category to generally reflect the situation where the optimal pixel information exists in this category, so as to only determine the general position where the optimal pixel information is located. Next, it is necessary to continue performing clustering analysis on the pixel values in the category, so as to gradually narrow down the range and locate the position where the optimal pixel information is located.

[0046] S30: Denote the categories in the category combination corresponding to the first corrected image as reference categories; perform clustering on the data in the reference categories to obtain several sub-categories, combine the sub-categories of all pixels in the reference image to obtain several sub-category combinations, and use the median of each sub-category in the sub-category combination as the pixel value of the corresponding pixel in the reference image to obtain an adjusted reference image based on this sub-category combination.

[0047] Preferably, as an example, denote the categories in the category combination corresponding to the first corrected image as reference categories; perform clustering on the data in the reference categories to obtain several sub-categories, combine the sub-categories of all pixels in the reference image to obtain several sub-category combinations, and use the median of each sub-category in the sub-category combination as the pixel value of the corresponding pixel in the reference image to obtain an adjusted reference image based on this sub-category combination, including: Denote the categories in the category combination corresponding to the first corrected image as reference categories; use the DBSCAN algorithm to perform clustering on the data in the reference categories to obtain several sub-categories.

[0048] Randomly select a sub-category from all the sub-categories of each pixel in the reference image, combine the sub-categories selected by all pixels in the reference image to obtain a sub-category combination, and obtain all sub-category combinations in the same way; Obtain the median of all the data of each sub-category as the reference data of each sub-category, and use the reference data of each sub-category in any sub-category combination as the pixel value of the corresponding pixel to obtain an adjusted reference image based on this sub-category combination.

[0049] It can be understood that by further classifying the reference categories, the positioning range of the optimal pixel value can be further narrowed, and by gradually narrowing the positioning range step by step until the optimal pixel value is obtained.

[0050] S31: Calculate the signal description ability, and use the adjusted reference image with the maximum signal description ability as the second corrected image.

[0051] Calculate the signal description ability according to the method in step S22, and obtain the adjusted reference image with the maximum signal description ability among the adjusted reference images obtained by combining all sub-categories as the second corrected image.

[0052] S32: And so on until the signal description ability converges to obtain the final corrected image.

[0053] Preferably, as an example, and so on until the signal description ability converges to obtain the final corrected image, including: Calculate the absolute value of the difference between each signal description ability and the previous signal description ability. If the absolute value of the difference is less than the preset difference threshold, it is considered that the signal description ability converges. In this embodiment, the preset difference threshold is taken as 0.01 for description. Other embodiments can take other values, and this embodiment does not make specific limitations.

[0054] Use the corrected image corresponding to the convergence of the signal description ability as the final corrected image.

[0055] S4: Implement the monitoring of aluminum profile extrusion forming based on the final corrected image.

[0056] It should be noted that if there are no defects in the finished product, it is relatively smooth at non-structural pixels. At the same time, the unevenness caused by light is also excluded through the above process. Therefore, there will be no texture at non-structural pixels in the final corrected image when there are no defects in the finished product. If there is texture at non-structural pixels in the final corrected image, it means that there are defects on the finished product. Therefore, defect detection can be carried out based on this.

[0057] Preferably, as an example, implementing the monitoring of aluminum profile extrusion forming based on the final corrected image includes: Perform edge detection on the final corrected image. If the edge pixels in the edge image of the final corrected image are non-structural pixels, mark these edge pixels as defect pixels; If the number of defect pixels is greater than the preset number, an abnormal warning is issued, so that people can adjust the control parameters in time to prevent more defective products from being produced due to improper setting of control parameters. In this embodiment, the preset number is taken as 10 for description. Other embodiments can take other values, and this embodiment does not make specific limitations.

[0058] The embodiment of the present invention also discloses a monitoring system for aluminum profile extrusion forming based on machine vision, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the monitoring method for aluminum profile extrusion forming based on the present invention is implemented.

[0059] The above system further includes other components well known to those skilled in the art, such as a communication bus and a communication interface, whose settings and functions are known in the art, and thus will not be described herein again.

[0060] In the present invention, the aforementioned memory can be any tangible medium that contains or stores a program, and the program can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory, a dynamic random access memory, a static random access memory, an enhanced dynamic random access memory, a high-bandwidth memory, a hybrid memory cube, etc., or any other medium that can be used to store the required information and can be accessed by an application program, a module, or both. Any such computer storage medium can be part of the device or accessible or connectable to the device.

[0061] Although this specification has shown and described multiple embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will think of many changes, alterations, and alternative ways without departing from the spirit and scope of the present invention. It should be understood that various alternatives to the embodiments of the present invention described herein can be employed in the practice of the present invention.

[0062] The above are all preferred embodiments of the present invention, and the protection scope of the present invention is not limited thereby. Therefore, all equivalent changes made according to the structure, shape, and principle of the present invention should be covered within the protection scope of the present invention.

Claims

1. A machine vision-based aluminum profile extrusion molding monitoring method, characterized in that: Includes steps: Capture images of finished aluminum extrusion products at several angles; A finished product image at any angle is used as a reference image, and the finished product images at other angles are matched with the reference image. A pixel value set of each pixel in the reference image is obtained according to the matching relationship of the pixels, and a distribution model is fitted to the pixel value set; a clustering process is performed on the distribution model to obtain several categories, and the categories of all pixels in the reference image are combined to obtain several category combinations, and the median of each category in the category combination is used as the pixel value of the corresponding pixel in the reference image to obtain an adjusted reference image based on the category combination; Computational signal description capabilities , compare and analyze the edge pixels in the adjusted reference image and the structural design image, set the difference mark value of each edge pixel in the structural design image, and divide the pixels in the adjusted reference image into structural pixels and non-structural pixels according to the difference mark value. S represents the number of edge pixels in the structural design image with a difference mark value of 0. represents the gradient value of the i-th non-structural pixel in the adjusted reference image, N represents the number of non-structural pixels in the adjusted reference image, M represents the signal description ability of the adjusted reference image, and tanh() represents the hyperbolic tangent function; the adjusted reference image with the largest signal description ability is used as the first corrected image; Recording the category in the category combination corresponding to the first corrected image as the reference category; Clustering the data in the benchmark category to obtain several subcategories, combining the subcategories of all pixels in the benchmark image to obtain several subcategory combinations, and using the median of each subcategory in the subcategory combination as the pixel value of the corresponding pixel in the benchmark image to obtain an adjusted benchmark image based on the subcategory combination; Calculate the signal description capability, and use the adjusted reference image with the largest signal description capability as the second corrected image; and so on, until the signal description capability converges, and obtain the final corrected image; Aluminum profile extrusion forming monitoring is achieved based on the final corrected image.

2. The method for monitoring aluminum profile extrusion molding based on machine vision according to claim 1, characterized in that: The step of obtaining a pixel value set of each pixel in the reference image according to the matching relationship of the pixels and fitting a distribution model to the pixel value set includes: Any pixel in the reference image is recorded as a target pixel, and images containing the target pixel are obtained from finished images at all other angles and recorded as matching images. Pixels matching the target pixel are obtained from all matching images and recorded as matching pixels. The set consisting of the pixel values ​​of the target pixel and all matching pixels is taken as the pixel value set of the target pixel. The Gaussian model fitted to all pixel values ​​in the pixel value set of the target pixel is recorded as the distribution model of the target pixel.

3. The method for monitoring aluminum profile extrusion molding based on machine vision according to claim 1, characterized in that: The step of combining the categories of all pixels in the reference image to obtain a plurality of category combinations, and using the median of each category in the category combination as the pixel value of the corresponding pixel in the reference image to obtain an adjusted reference image based on the category combination includes: Randomly select a category from all categories of each pixel in the reference image, combine the categories selected from all pixels in the reference image together to obtain a category combination, and obtain all category combinations; The median of all data of each category is obtained and recorded as the reference data of each category. The reference data of each category in any category combination is used as the pixel value of the corresponding pixel to obtain an adjusted reference image based on the category combination.

4. The method for monitoring aluminum profile extrusion molding based on machine vision according to claim 1, characterized in that: The step of comparing and analyzing the edge pixels in the adjusted reference image and the structure design image to set the difference mark value of each edge pixel in the structure design image includes: Performing edge detection on the adjusted reference image to obtain an edge image, matching the edge image and the adjusted reference image with the structural design drawing respectively, and adjusting the edge image and the adjusted reference image to images with the same viewing angle and scale as the structural design drawing based on the matching results; The edge pixels in the edge image after the viewing angle scale is adjusted are recorded as comparison pixels, and the preset window of any edge pixel in the structural design diagram is obtained. It is determined whether there is a comparison pixel in the corresponding area of ​​the edge image in the preset window of each edge pixel. If so, the difference flag value of the edge pixel in the structural design diagram is set to 1, otherwise the difference flag value is set to 0.

5. The method for monitoring aluminum profile extrusion molding based on machine vision according to claim 4, characterized in that: The step of dividing the pixels in the adjusted reference image into structural pixels and non-structural pixels according to the difference flag value comprises: The pixels in the corresponding area of ​​the preset window of each edge pixel in the adjusted reference image after the viewing angle scale is adjusted are recorded as structural pixels, otherwise they are recorded as non-structural pixels.

6. The method for monitoring aluminum profile extrusion molding based on machine vision according to claim 5, characterized in that: The step of obtaining a preset window of any edge pixel in the structure design diagram includes: In the structural design diagram, a window with a length of L obtained in a vertical direction of the edge line with the edge pixel as the center is recorded as the window of the edge pixel in the structural design diagram, and L represents a preset size.

7. The method for monitoring aluminum profile extrusion molding based on machine vision according to claim 1, characterized in that: The step of convergence of the signal description capability includes: The absolute value of the difference between the description capability of each signal and the description capability of the previous signal is calculated. If the absolute value of the difference is less than a preset difference threshold, the signal description capability is considered to have converged.

8. The method for monitoring aluminum profile extrusion molding based on machine vision according to claim 5, characterized in that: The method of realizing aluminum profile extrusion forming monitoring based on the final corrected image includes: Performing edge detection on the final corrected image, if an edge pixel in the edge image of the final corrected image is a non-structural pixel, marking the edge pixel as a defective pixel; If the number of defective pixels is greater than the preset number, an abnormal warning is issued.

9. Aluminum profile extrusion molding monitoring system based on machine vision, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the machine vision-based aluminum profile extrusion forming monitoring method according to any one of claims 1 to 8 is implemented.

Images