Methods, systems, computers, and storage media for detecting burrs on the edges of metal sheets

Abstract

This invention discloses a method, system, computer, and storage medium for detecting burrs on the edges of metal plates. The method includes: acquiring a contour image containing at least two adjacent processed edges of the metal plate to be tested; binarizing the contour image to obtain a contour binary image, and sharpening the contour binary image to obtain a contour comparison image to highlight the processed edges; determining whether burrs exist; if not, extending the sampling period according to a predetermined periodic change strategy; if yes, performing continuous sampling detection and calculating the number of burrs; if the number of burrs exceeds a first predetermined threshold, determining tool wear and controlling the production line to stop; and if the number of consecutive burr-free determinations reaches a second predetermined threshold, re-performing periodic sampling detection. This invention achieves intelligent burr detection, effectively improving detection efficiency and enabling dynamic adjustment of the detection period to meet the needs of large-scale processing detection.

Description

Technical Field

[0001] This invention relates to the field of metal plate inspection technology, and in particular to a method, system, computer, and storage medium for detecting burrs on the edge of metal plates. Background Technology

[0002] Steel plates and aluminum plates are commonly used metal sheets in industries such as automobiles and home appliances. The current metal sheet processing process is as follows: after the coil is placed on the production line, it is unrolled and gradually unwound to form a metal strip. During the unrolling process, the metal strip is leveled by pressure rollers. Then, one or both sides of the metal strip are cut to a fixed width. Next, the cutter periodically presses down to punch and cut, thus forming a metal sheet of the predetermined size. The metal sheets are then transferred out sequentially on the conveyor belt for subsequent stacking, packaging and other work.

[0003] Because the production line operates continuously, the cutting tools may wear down after prolonged operation, resulting in burrs at the edges of the finished metal sheets. This can easily cause defects in products requiring high machining precision. Currently, burrs are typically detected manually on the conveyor belt by visual inspection, touch, or the use of tools. This method is inefficient and relies heavily on manual experience and skills, making it difficult to meet the demands of long-term, mass production. Summary of the Invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of this invention is to provide a method, system, computer and storage medium for detecting burrs on the edge of metal plates, which has the advantage of being able to quickly detect burrs on the edge of metal plates.

[0005] The objective of this invention is achieved through the following technical solution:

[0006] According to a first aspect of the present disclosure, a method for detecting burrs on the edge of a metal plate is provided, comprising:

[0007] Obtain a contour image containing at least two adjacent processed edges of the metal plate under test;

[0008] The contour image is binarized to obtain a contour binary image, and the contour binary image is sharpened to obtain a contour comparison image to highlight the processed edge;

[0009] The contour comparison image is compared with a standard image to determine whether there are burrs by the smoothness of the processed edge;

[0010] If not, the sampling period will be extended according to the predetermined periodic change strategy for periodic sampling and detection;

[0011] If so, continuous sampling detection is performed, and the cusp feature data of the contour comparison map is obtained to calculate the number of burrs;

[0012] If the number of burrs exceeds a first predetermined threshold, the tool is determined to be worn and the production line is stopped.

[0013] If the number of consecutive times without burrs reaches the second predetermined threshold, then periodic sampling and detection will be performed again.

[0014] In some exemplary embodiments, obtaining a contour image that includes at least two adjacent processed edges of the metal plate under test specifically includes:

[0015] A high-speed camera acquires a first set of images at a predetermined angle, which includes at least two adjacent sides of the metal plate under test.

[0016] Extract edge features from the first image set and perform correction processing on the edge features to form at least two correction planar images corresponding to adjacent two sides of the metal plate to be tested, respectively.

[0017] At least two of the aforementioned correction planar images are synthesized to obtain a contour image containing two adjacent processed edges of the metal plate under test.

[0018] In some exemplary embodiments, the binarization of the contour image to obtain a contour binary map specifically includes:

[0019] The obtained contour image is converted to grayscale to obtain a grayscale image;

[0020] Calculate the mean grayscale values ​​of the foreground and background in the grayscale image to dynamically adjust the foreground and background segmentation thresholds;

[0021] The grayscale image is binarized based on the division threshold to obtain a contour binary image.

[0022] In some exemplary embodiments, the step of extending the sampling period according to a predetermined periodic variation strategy for periodic sampling detection specifically includes:

[0023] When sampling and testing is performed during the initial sampling period, it is determined whether the number of consecutively qualified metal plates reaches the preset qualified threshold.

[0024] If not, then maintain the initial sampling period and perform periodic sampling and detection;

[0025] If so, the corresponding periodic variation strategy is obtained according to the material of the metal plate to be tested to extend the sampling period. The periodic variation strategy is as follows: the sampling period is gradually increased from the initial sampling period to the maximum sampling period and then gradually decreased to the initial sampling period. After the initial sampling period is reached again, the sampling period is gradually increased again.

[0026] In some exemplary embodiments, the periodic variation strategy is calculated and executed according to the following formula:

[0027] T1 = kx + T0, x = 0, 1, 2, ..., n;

[0028] T2=-kx+T0+2kn, x= n,n+1,n+2,…,m;

[0029] Where T0 is the initial sampling period, T1 is the number of periods in the increasing segment, and T2 is the number of periods in the decreasing segment, with the number of periods represented by the number of interval metal plates; k is the sampling coefficient, x is the number of interval metal plates, and x+1 is added after each sampling; n and m are both positive integers.

[0030] In some exemplary embodiments, obtaining the cusp feature data of the contour comparison map to calculate the number of burrs specifically includes:

[0031] The contour comparison map is processed using an image feature processing model to obtain cusp feature data, which includes cusp peak data.

[0032] The peak data of the cusp is compared with a preset reference parameter to mark data that exceeds the preset reference parameter as burrs;

[0033] The total number of burrs is calculated to obtain the number of burrs.

[0034] In some exemplary embodiments, the cusp feature data further includes cusp projection data, and the method further includes:

[0035] The burr type is determined based on the apex projection data, and a tool wear warning level is generated based on the burr type.

[0036] According to a second aspect of the present disclosure, a burr detection system for the edge processing of metal plates is provided, comprising:

[0037] The image acquisition unit is used to acquire a contour image containing at least two adjacent processed edges of the metal plate under test;

[0038] An image processing unit is used to perform binarization processing on the contour image to obtain a contour binary image, and to perform sharpening processing on the contour binary image to obtain a contour comparison image to highlight the processed edge;

[0039] An image analysis unit is used to compare the contour comparison image with a standard image to determine whether there are burrs by the smoothness of the processed edge;

[0040] The first judgment unit is used to extend the sampling period according to a predetermined periodic change strategy to perform periodic sampling detection when it is determined that there are no burrs.

[0041] The second judgment unit is used to perform continuous sampling detection when it is determined that burrs exist, and to obtain the cusp feature data of the contour comparison map to calculate the number of burrs; wherein, if the number of burrs exceeds a first predetermined threshold, the tool wear is judged and the production line is stopped; and if the number of consecutive burr-free judgments reaches a second predetermined threshold, the periodic sampling detection is repeated.

[0042] According to a third aspect of the present disclosure, a computer is provided, including a memory and a processor, wherein the memory stores computer-readable instructions that, when executed by the processor, cause the processor to perform the steps of the detection method as described in the first aspect.

[0043] According to a fourth aspect of the present disclosure, a storage medium storing computer-readable instructions is provided, which, when executed by one or more processors, cause the one or more processors to perform the steps of the detection method as described in the first aspect.

[0044] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0045] This invention provides a method, system, computer, and storage medium for detecting burrs on the edges of metal plates. During metal plate processing, a high-speed camera or other visual inspection device acquires the contour image of the metal plate. Since steel plate processing typically involves cutting along both the length and width directions, the contour image must include at least two adjacent processing edges to simultaneously reflect the wear of both sets of cutting tools in the length and width directions. Binarization of the contour image, using only black and white, provides a simpler and more intuitive representation of edge burrs. Sharpening the contour image further highlights the processing edges, providing a more accurate basis for subsequent burr detection. By comparing the contour image with a standard image, the presence of burrs is first determined. If no burrs are found, the metal plate processing is considered successful and the cutting tools are unworn. Since tool wear is typically minimal over a considerable period, the sampling period can be extended to reduce computer processing load. However, if burrs are found, a certain amount is usually permissible for metal plates with lower precision requirements. Therefore, to determine the presence of burrs... The first step is to determine if the metal sheet meets the processing requirements. The second step is to count the number of burrs. This is done by analyzing the burr characteristic data and comparing it to a first predetermined threshold. Since tool wear is a gradual process, there will be periods where the number of burrs on metal sheets is less than the first predetermined threshold. These metal sheets still meet the processing requirements, or can be easily repaired to meet them. In such cases, the machine can be stopped after the coil processing is complete before tool inspection. However, if the number of burrs exceeds the first predetermined threshold, the metal sheet no longer meets the processing requirements, and further processing is unnecessary. In this case, the machine needs to be stopped for tool inspection. Errors may occur during the inspection process. If no burrs are continuously detected, it indicates that the tool is normal and has no wear. In this case, the continuous inspection process can be canceled, and periodic sampling and inspection can be restarted, effectively reducing the system's inspection pressure. This process achieves intelligent burr detection, effectively improving inspection efficiency, saving labor costs, and enabling dynamic adjustment of the inspection cycle to maintain long-term effectiveness and meet the needs of large-scale processing inspection. Attached Figure Description

[0046] Figure 1 This is a flowchart of the method for detecting burrs on the edge of a metal plate in an embodiment of the present invention.

[0047] Figure 2 This is a schematic diagram of the process for obtaining a contour image in an embodiment of the present invention.

[0048] Figure 3This is a diagram showing the spatial relative positions of the steel plate and the high-speed camera when acquiring the contour image in an embodiment of the present invention.

[0049] Figure 4 This is a schematic diagram of the process for obtaining a binary contour map in an embodiment of the present invention.

[0050] Figure 5 This is a schematic diagram of the process for calculating the number of burrs in an embodiment of the present invention.

[0051] Figure 6 This is a schematic diagram of the basic structure of the metal plate edge processing burr detection system in an embodiment of the present invention.

[0052] Figure 7 This is a basic structural block diagram of the computer in an embodiment of the present invention.

[0053] The numbers and letters in the diagram represent the names of the corresponding components:

[0054] 10. Image acquisition unit; 20. Image processing unit; 30. Image analysis unit; 40. First judgment unit; 50. Second judgment unit. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] like Figures 1 to 5 As shown, the first aspect of the present invention provides a method for detecting burrs on the edge of a metal plate, comprising:

[0057] S100: Obtain a contour image containing at least two adjacent processed edges of the metal plate to be tested.

[0058] Specifically, such as Figure 2 As shown, step S100 specifically includes:

[0059] S101. A high-speed camera acquires a first image set at a predetermined angle, containing at least two adjacent sides of the metal plate under test. In practical applications, the first image set may include one, two, or more images. The predetermined angle can be set to any angle capable of capturing the sides of the metal plate under test. Figure 3 As shown, if only one set of high-speed cameras is used, the high-speed camera can be set at one corner of the metal plate to be measured, and preferably at the same angle as the diagonal of the metal plate to be measured, such as... Figure 3As shown in Figure a, when the metal plate to be tested moves to the detection position, the high-speed camera can capture an image including both sides, which can be used as the first image set. If two sets of high-speed cameras are set, they can be positioned on one side of the length direction and one side of the width direction, respectively. Figure 3 As shown in Figure b, when the metal plate to be tested moves to the detection position, two images, each containing one of the two sides, can be obtained. These two images can then be used as the first image set. Alternatively, one set of high-speed cameras can be positioned at the corner, and another set can be positioned on one of the sides, as shown in Figure b. Figure 3 As shown in Figure c, you can obtain two images containing three sides, or you can set the two high-speed cameras at opposite corners to obtain two images containing four sides. Regardless of how the high-speed cameras are set up, it is sufficient that the first image set contains images of two adjacent sides.

[0060] S102. Extract edge features from the first image set and perform correction processing on the edge features to form at least two correction planar images corresponding to adjacent sides of the metal plate to be tested.

[0061] The step of extracting edge features from the first image set refers to extracting images of the side regions of two adjacent metal plates. The purpose of this process is to remove the main body parts that are irrelevant to the edge features. Since burr detection usually only focuses on the features of the metal plate edges, the main body parts will increase the amount of data for image analysis and interfere with the detection process. Therefore, edge features are extracted to simplify the detection of complex images and improve detection accuracy.

[0062] Edge feature correction mainly involves correcting the tilt of the edges to a vertical or horizontal state to facilitate subsequent image comparison. Since there may be slight differences in the relative positions of different metal plates with the high-speed camera during detection, correction can also eliminate these differences and further improve the accuracy of detection.

[0063] S103. At least two of the correction planar images are synthesized to obtain a contour image containing two adjacent processed edges of the metal plate to be tested.

[0064] The purpose of the synthesis process is to stitch the corrected planar images of the two processed edges into a single contour image. This way, only one comparison is needed during subsequent detection, which further simplifies the detection process.

[0065] S200. The contour image is binarized to obtain a contour binary image, and the contour binary image is sharpened to obtain a contour comparison image to highlight the processed edge.

[0066] Among them, such as Figure 4 As shown, the binarization process of the contour image to obtain a contour binary map specifically includes:

[0067] S201. The obtained contour image is converted to grayscale to obtain a grayscale image. Grayscale conversion can usually be achieved by weighted averaging and merging RGB channels.

[0068] S202. Calculate the mean grayscale values ​​of the foreground / background in the grayscale image to dynamically adjust the foreground and background segmentation threshold. This segmentation threshold can be dynamically set according to the actual situation of different images. For example, the segmentation threshold can be determined by the P-parameter method. During binarization, pixels higher than the segmentation threshold are set to 255, and the corresponding image part is white. Pixels lower than the segmentation threshold are set to 0, and the corresponding image part is black.

[0069] S203. Based on the division threshold, the grayscale image is binarized to obtain a contour binary image, which is black and white. Therefore, the smoothness of the processing edge can be more clearly distinguished, and thus it can be determined whether there are burrs.

[0070] The purpose of sharpening the binary contour image is to further locate the contour of the binary contour image. After obtaining the contour comparison image, the processing edge can be highlighted. The sharpening process can be implemented by existing algorithms such as the Laplacian sharpening method or the gradient operator sharpening method.

[0071] S300. The contour comparison image is compared with a standard image to determine whether there are burrs by the smoothness of the processed edge.

[0072] Specifically, by analyzing the concavity and convexity changes of the processed edge in the contour comparison diagram, the smoothness of the processed edge is obtained. When a burr appears, a concavity and convexity change will be formed on the processed edge, and the corresponding smoothness will be obtained. By setting a smoothness threshold, if the smoothness exceeds the threshold, it indicates that a burr exists; otherwise, there is no burr.

[0073] S400. If not, then extend the sampling period according to the predetermined periodic change strategy to perform periodic sampling and detection.

[0074] Specifically, the S400 includes:

[0075] S401. When sampling and testing under the initial sampling period, determine whether the number of consecutively qualified metal plates has reached the preset qualified threshold; wherein, the initial sampling period is the preset interval sampling period in the initial stage of processing. Since the initial stage requires a certain amount of time for processing and debugging, the initial sampling period needs to be maintained for sampling and testing in the initial stage.

[0076] S402. If not, maintain the initial sampling period and perform periodic sampling detection.

[0077] S403. If so, then obtain the corresponding periodic change strategy according to the material of the metal plate to be tested to extend the sampling period. The periodic change strategy is: periodically increase from the initial sampling period to the maximum sampling period and then gradually decrease to the initial sampling period. After the initial sampling period is reached again, gradually increase again.

[0078] Specifically, the periodic change strategy is calculated and executed according to the following formula:

[0079] T1 = kx + T0, x = 0, 1, 2, ..., n;

[0080] T2=-kx+T0+2kn, x= n,n+1,n+2,…,m;

[0081] Where T0 is the initial sampling period, T1 is the number of periods in the increasing segment, and T2 is the number of periods in the decreasing segment, with the number of periods represented by the number of interval metal plates; k is the sampling coefficient, x is the number of interval metal plates, and x+1 is added after each sampling; n and m are both positive integers.

[0082] The k value can be selected according to the different metal plate materials. For example, aluminum plates and steel plates have different hardness, and the wear of the cutting tool is also different when processing these two types of plates. Therefore, different sampling periods can be set to reduce redundant detection while meeting the batch processing inspection requirements. Usually, it is set to m=2n, and the value of n is preferably set to a positive integer between 4 and 10. Taking a specific example, we set k=1, n=5, m=10, T0=5. That is, in the initial state, sampling inspection is performed every 5 metal plates. The maximum sampling period in the incremental stage is 10. In this stage, the sampling periods are 5, 6, 7, 8, 9, and 10 in sequence. When the sampling period reaches 10, it enters the decrement stage. In this stage, the sampling periods are 10, 9, 8, 7, 6, and 5 in sequence. Then it enters the incremental stage again, and so on, until the processing is completed or a burr is detected.

[0083] By controlling the sampling period to vary periodically, the number of samples can be reduced as much as possible while ensuring a sufficient number of samples are collected. Furthermore, compared to setting a longer sampling period, it is possible to detect glitch conditions more promptly.

[0084] S500, if so, then perform continuous sampling detection and obtain the cusp feature data of the contour comparison map to calculate the number of burrs.

[0085] After calculating the number of burrs, the following steps are also performed:

[0086] S501. If the number of burrs exceeds the first predetermined threshold, it is determined that the tool is worn and the production line is stopped. Since the shutdown for maintenance has a significant impact on the production progress and the complexity of operation, setting the first predetermined threshold for comparison can reduce the occurrence of invalid shutdowns. When the number of burrs exceeds the first predetermined threshold, it is determined that the metal plate does not meet the processing requirements. This indicates that the tool is severely worn and no longer meets the requirements, so the machine needs to be stopped for maintenance. Of course, in actual application, it can also be set to stop the production line only when the number of burrs exceeds the first predetermined threshold multiple times in a row.

[0087] Among them, such as Figure 5 As shown, the step of obtaining the cusp feature data of the contour comparison map to calculate the number of burrs specifically includes:

[0088] S5031. The contour comparison map is processed using an image feature processing model to obtain cusp feature data, wherein the cusp feature data includes cusp peak data.

[0089] Among them, the image feature processing model is a data processing model that is trained to process image features. It is trained using a large amount of historical data of cusp feature data. The larger the amount of historical data, the more accurate the processing results of the model. The cusp peak data is mainly used to reflect the position of the cusp protruding edge.

[0090] S5032. The peak data of the apex is compared with the preset reference parameter to mark the data that exceeds the preset reference parameter as burr points. When the peak data of the apex exceeds the preset reference parameter, it means that the apex protrusion distance is large, and the point can be judged as a burr point. The preset reference parameter is determined based on the accuracy requirements of metal plate processing.

[0091] S5033. Calculate the total number of burrs to obtain the number of burrs. It can be understood that the total number of burrs is the sum of the number of all marked burrs.

[0092] The continuous detection process also includes the following steps:

[0093] S502. If the number of times no burrs are detected reaches the second predetermined threshold, then periodic sampling and detection will be performed again.

[0094] Considering the possibility of false detections or defects in certain areas of the roll material itself, if the metal plate is continuously tested without burrs, it indicates that the cutting tool is not worn. In this case, periodic sampling and testing can be repeated to reduce invalid sampling.

[0095] Furthermore, in some embodiments, the apex feature data further includes apex projection data, which mainly reflects the projected area of ​​the apex, thereby reflecting the size of the burr. The method further includes:

[0096] S600. Determine the burr type based on the cusp projection data, and generate a tool wear warning level according to the burr type. The burr type is divided according to the burr size and shape. Since different burr sizes correspond to different defect degrees and defect types, generally, the larger the burr, the higher the wear degree of the tool. Therefore, the tool wear warning level can be generated correspondingly according to the corresponding burr type, providing a reference for the subsequent tool maintenance process.

[0097] During the metal plate processing, the contour image of the metal plate to be measured is collected by a vision detection device such as a high-speed camera. Since the steel plate processing usually includes cutting actions in two directions, namely the length direction and the width direction, at least two adjacent processing edges are controlled to be included in the contour image, so as to reflect the wear conditions of the two groups of tools in the length direction and the width direction at the same time. Through the binary processing of the contour image, since the binary image has only two colors, black and white, it can more simply and intuitively reflect the burr situation of the edge contour. And by sharpening the contour binary image, the processing edge can be more prominent, providing a more accurate comparison basis for the subsequent burr judgment. By comparing the contour comparison image with the standard image, first judge whether there is a burr. If it is judged that there is no burr, it means that the metal plate processing is qualified and the tool has no wear. At this time, generally, the tool will not show wear phenomenon in a relatively long period of time. Therefore, the sampling period can be appropriately extended to reduce the processing pressure of the computer. And if it is judged that there is a burr, for the metal plate with low precision requirements, a certain burr is usually allowed. Therefore, in order to judge whether the metal plate with burr meets the processing requirements, secondly, the burr quantity is counted. The burr quantity is obtained by analyzing the cusp feature data and compared with the first predetermined threshold. Since the tool wear is a gradual process, there will be a period of time when the burr quantity of the metal plate with burr is less than the first predetermined threshold, and this part of the metal plate also meets the processing requirements or can meet the processing requirements through simple trimming. When this situation occurs, the tool can be repaired after the coil processing is completed and the machine is stopped. And once the burr quantity exceeds the first predetermined threshold, the metal plate does not meet the processing requirements, and the metal plate to be processed continuously does not meet the processing requirements, so there is no need to continue the processing. At this time, the machine needs to be stopped for tool maintenance. And there may also be a situation of misjudgment during the detection process. When it is continuously judged that there is no burr, it means that the tool is normal and there is no wear. At this time, the continuous detection process can be cancelled and the periodic sampling detection can be carried out again, effectively reducing the system detection pressure. Through the above process, intelligent burr detection is realized, effectively improving the detection efficiency, saving labor costs, and being able to dynamically adjust the detection period, maintaining the long-term continuous effectiveness of the detection process, and meeting the needs of large-scale processing detection.

[0098] Such as Figure 6As shown, a second aspect of the present invention provides a burr detection system for metal plate edge processing, comprising: an image acquisition unit 10, configured to acquire a contour image containing at least two adjacent processed edges of the metal plate to be tested; an image processing unit 20, configured to perform binarization processing on the contour image to obtain a contour binary image, and perform sharpening processing on the contour binary image to obtain a contour comparison image to highlight the processed edge; an image analysis unit 30, configured to compare the contour comparison image with a standard image to determine whether burrs exist based on the smoothness of the processed edge; a first judgment unit 40, configured to perform periodic sampling detection by extending the sampling period according to a predetermined periodic change strategy when it is determined that no burrs exist; and a second judgment unit 50, configured to perform continuous sampling detection when it is determined that burrs exist, and acquire cusp feature data of the contour comparison image to calculate the number of burrs; wherein, if the number of burrs exceeds a first predetermined threshold, tool wear is determined and the production line is stopped; and if the number of consecutive burr-free determinations reaches a second predetermined threshold, periodic sampling detection is performed again.

[0099] A third aspect of the present invention also provides a computer. See details below. Figure 7 , Figure 7 This is a basic structural block diagram of the computer in this embodiment.

[0100] like Figure 7 The diagram shows the internal structure of a computer. The computer includes a processor, non-volatile storage medium, memory, and a network interface connected via a system bus. The non-volatile storage medium stores the operating system, a database, and computer-readable instructions. The database may store control information sequences. When executed by the processor, the computer-readable instructions enable the processor to implement a method for detecting burrs on the edges of metal plates. The processor provides computational and control capabilities, supporting the operation of the entire computer. The memory stores computer-readable instructions, which, when executed by the processor, enable the processor to perform a method for detecting burrs on the edges of metal plates. The network interface is used for communication with a terminal. Those skilled in the art will understand that… Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer to which the present application is applied. A specific computer may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0101] In this embodiment, the processor is used to execute... Figure 6The image acquisition unit 10, image processing unit 20, image analysis unit 30, first judgment unit 40, and second judgment unit 50 are described in detail. The memory stores the program code and various types of data required to execute the above modules. The network interface is used for data transmission between the user terminal and the server. The server can call the server's program code and data to execute the functions of all sub-modules.

[0102] The present invention also provides a storage medium storing computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of the metal plate edge processing burr detection method of any of the above embodiments.

[0103] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0104] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations 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, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0105] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0106] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0107] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0108] This specification may be described in the general context of computer-executable instructions, such as program units, that are executed by a computer. Generally, program units include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification may also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program units may reside in local and remote computer storage media, including storage devices.

[0109] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0110] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention. These are all equivalent modifications and improvements made to the above embodiments based on the essential technology of the present invention, and all of these fall within the protection scope of the present invention.

Claims

1. A method of detecting burrs in a metal sheet edge processing, characterized by, The method comprises the following steps: acquiring a profile image containing at least two adjacent machining edges of a metal sheet to be measured; performing binaryzation on the profile image to obtain a profile binary image, and performing sharpening on the profile binary image to obtain a profile contrast image to highlight the machining edges; comparing the profile contrast image with a standard image to determine whether burrs exist by the smoothness of the machining edges; if not, extending the sampling period according to a predetermined period variation strategy to perform periodic sampling detection; if yes, performing continuous sampling detection, and acquiring sharp point feature data of the profile contrast image to calculate the number of burrs; if the number of burrs exceeds a first predetermined threshold, determining tool wear and controlling the production line to stop; if the number of times of continuously determining no burrs reaches a second predetermined threshold, re-performing periodic sampling detection; the periodic sampling detection performed according to the predetermined period variation strategy specifically comprises: when sampling detection is performed at an initial sampling period, determining whether the number of metal sheets continuously detected to be qualified reaches a preset qualified threshold; if not, maintaining the initial sampling period to perform periodic sampling detection; if yes, acquiring a corresponding period variation strategy according to the material of the metal sheet to be measured to extend the sampling period, the period variation strategy being that the sampling period is periodically increased from the initial sampling period to a maximum sampling period and then decreased to the initial sampling period, and then the sampling period is increased again when the initial sampling period is reached again; the period variation strategy is calculated and executed according to the following calculation formula: T1=kx+T0, x=0,1,2,…,n; T2=-kx+T0+2kn, x= n,n+1,n+2,…,m; wherein, T0 is the initial sampling period, T1 is the number of periods in the increasing segment, T2 is the number of periods in the decreasing segment, wherein the number of periods is represented by the number of interval metal sheets; k is a sampling coefficient, x is the number of interval metal sheets, and x+1 is completed once sampling is performed; n and m are positive integers.

2. The metal plate edge processing burr detection method according to claim 1, characterized by, the profile image containing at least two adjacent machining edges of the metal sheet to be measured is acquired specifically as follows: a high-speed camera collects a first image set containing at least two adjacent sides of the metal sheet to be measured at a predetermined angle; edge features in the first image set are extracted, and the edge features are corrected to form at least two corrected planar images corresponding to the two adjacent sides of the metal sheet to be measured respectively; the at least two corrected planar images are synthesized to obtain the profile image containing the two adjacent machining edges of the metal sheet to be measured.

3. The metal plate edge processing burr detection method according to claim 1 or 2, characterized by, the profile binary image is obtained by performing binaryzation on the profile image specifically as follows: performing gray scale conversion on the acquired profile image to obtain a gray scale image; calculating the foreground / background gray scale mean value in the gray scale image to dynamically adjust the division threshold of the foreground and the background; performing binaryzation on the gray scale image based on the division threshold to obtain the profile binary image.

4. The metal plate edge processing burr detection method according to claim 3, characterized by, the sharp point feature data of the profile contrast image is acquired to calculate the number of burrs specifically as follows: the profile contrast image is processed by using an image feature processing model to acquire sharp point feature data, the sharp point feature data including sharp point peak value data; comparing the peak data with preset reference parameters to mark data exceeding the preset reference parameters as burr points; counting the total amount of the burr points to obtain a burr quantity.

5. The metal plate edge processing burr detection method according to claim 4, characterized by, The peak feature data further includes peak projection data, and the method further includes: determining a burr type according to the peak projection data, and generating a tool wear warning level according to the burr type.

6. A metal sheet edge processing burr detection system characterized by comprising: comprise: an image acquisition unit configured to acquire a profile image containing at least two adjacent processing edges of a metal plate to be measured; an image processing unit configured to perform binaryzation processing on the profile image to obtain a profile binary image, and perform sharpening processing on the profile binary image to obtain a profile contrast image to highlight the processing edges; an image analysis unit configured to compare the profile contrast image with a standard image to determine whether there is a burr by the smoothness of the processing edges; a first determination unit configured to, when it is determined that there is no burr, extend a sampling period according to a predetermined period variation strategy to perform periodic sampling detection; a second determination unit configured to, when it is determined that there is a burr, perform continuous sampling detection, and acquire peak feature data of the profile contrast image to calculate a burr quantity; wherein, if the burr quantity exceeds a first predetermined threshold, it is determined that the tool is worn and the production line is stopped; and if the number of times of continuously determining no burr reaches a second predetermined threshold, periodic sampling detection is performed again; wherein, the periodic sampling detection performed according to the predetermined period variation strategy specifically comprises: when sampling detection is performed at an initial sampling period, determining whether the number of metal plates continuously detected to be qualified reaches a preset qualified threshold; if not, maintaining the initial sampling period to perform periodic sampling detection; if yes, acquiring a corresponding period variation strategy according to the material of the metal plate to be measured to extend the sampling period, the period variation strategy being that the sampling period is gradually increased from the initial sampling period to a maximum sampling period, then gradually decreased to the initial sampling period, and then gradually increased again when the initial sampling period is reached again; the period variation strategy is calculated and executed according to the following calculation formula: T1=kx+T0, x=0,1,2,…,n; T2=-kx+T0+2kn, x= n,n+1,n+2,…,m; wherein, T0 is the initial sampling period, T1 is the number of periods in the increasing segment, T2 is the number of periods in the decreasing segment, wherein the number of periods is represented by the number of interval metal plates; k is a sampling coefficient, x is the number of interval metal plates, and x+1 is completed once sampling is performed; n and m are positive integers.

7. A computer comprising a memory and a processor, wherein the memory stores computer readable instructions, and the computer readable instructions are executed by the processor to cause the processor to perform the steps of the detection method according to any one of claims 1 to 5.

8. A storage medium storing computer readable instructions, wherein the computer readable instructions are executed by one or more processors to cause the one or more processors to perform the steps of the detection method according to any one of claims 1 to 5.

Images