Method for measuring arc length, measuring device and computer readable storage medium

Abstract

The application discloses a kind of electric arc length measurement method and its measuring device and computer readable storage medium.The measurement method includes the following steps: via the image of electric arc to be measured, the electric arc detection frame of electric arc to be measured is obtained;Based on the two edge straight line sections of the weld image in non-welding state, the weld edge interval actual physical length and the distance pixel conversion ratio between the interval pixel point number of two edge straight line sections;And via distance pixel conversion ratio, the actual physical length of the electric arc to be measured corresponding to the pixel point number of electric arc detection frame is obtained.The application does not need to contact welding environment, and omits the step of prior calibration, not only can be suitable for a variety of automatic welding welding environment, has good adaptability and universality, but also has strong anti-jitter and anti-interference, realizes efficient and stable electric arc length measurement, to improve the effectiveness, accuracy and reliability of measurement result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of welding automation, and in particular relates to an arc length measurement method, an arc length measurement device, and a computer readable storage medium. BACKGROUND

[0002] The arc length is a key parameter affecting the welding quality and the stability of the welding process. A suitable arc length can ensure the uniformity and consistency of the weld. An excessively long arc can cause overheating or deformation of the material, and an excessively short arc can affect the penetration and fusion of the weld. Therefore, an excessively long or short arc can cause arc fluttering or instability, thereby affecting the continuity of the welding operation and the quality of the weld.

[0003] With the development of automatic welding arc length measurement technology, there are mainly two types of contact measurement and non-contact measurement. Contact measurement can include using mechanical or electronic contact sensors to directly measure the distance between the welding torch and the workpiece to determine the arc length, but this measurement method will interfere with the welding process, is not suitable for use in high-temperature environments, and has low measurement accuracy. Non-contact measurement can include using electrical signal analysis, i.e., calculating the arc length by the arc voltage, but this measurement method is susceptible to electromagnetic interference and can also result in inaccurate measurement results.

[0004] Further, with the combination of robotic automatic welding and machine vision, by monitoring the arc length in real time, the welding parameters (such as voltage, current, or welding torch height) can be adjusted in time to ensure process consistency and reduce human error. Therefore, the non-contact measurement method based on visual measurement can be suitable for use in complex welding environments such as high temperature, strong light, and smoke, and will not interfere with the welding process.

[0005] However, the existing non-contact measurement method based on visual measurement needs to use a coplanar linear camera calibration method. After setting the calibration board and the measurement object to be strictly coplanar, the conversion relationship between the image coordinates, the camera coordinates, and the world coordinates is described by the camera intrinsic matrix and the extrinsic matrix. Once the calibration board and the measurement object are not strictly coplanar, non-coplanar errors will occur, thereby affecting the accuracy and stability of the measurement results. In addition, complex mathematical calculations are required when solving the camera parameters, which further leads to low measurement efficiency and poor stability of the measurement results.

[0006] In order to solve the above problems in the prior art, the technical field urgently needs an arc length measurement technology, which does not need to contact the welding environment and omits the prior calibration step, can be applied to various automatic welding environments, has good adaptability and universality, has strong anti-jitter and anti-interference, realizes efficient and stable arc length measurement, and thus improves the effectiveness, accuracy and reliability of the measurement results. SUMMARY

[0007] The following presents a simplified summary of one or more aspects in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects, and is intended to neither identify key or critical elements of all aspects nor delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.

[0008] In order to overcome the above-mentioned defects in the prior art, the present application provides an arc length measurement method, an arc length measurement device, and a computer readable storage medium, which do not need to contact the welding environment and omit the prior calibration step, can be applied to various automatic welding environments, have good adaptability and universality, have strong anti-jitter and anti-interference, realize efficient and stable arc length measurement, and thus improve the effectiveness, accuracy and reliability of the measurement results.

[0009] Specifically, the arc length measurement method according to the first aspect of the present application comprises the following steps: obtaining an arc detection frame of a to-be-measured arc via a to-be-measured arc image; obtaining two edge straight line segments of a weld seam based on a non-welding state weld seam image; obtaining a distance pixel conversion ratio between an actual physical length of a weld seam edge spacing and a number of interval pixels between the two edge straight line segments; and obtaining an actual physical length of the to-be-measured arc corresponding to a number of pixel points of the arc detection frame via the distance pixel conversion ratio.

[0010] In addition, the arc length measurement device according to the second aspect of the present application comprises a memory and a processor. The processor is connected to the memory and is configured to implement the arc length measurement method according to the first aspect of the present application.

[0011] In addition, the third aspect of the present application further provides a computer readable storage medium having computer instructions stored thereon. When the computer instructions are executed by a processor, the arc length measurement method according to the first aspect of the present application is implemented. BRIEF DESCRIPTION OF DRAWINGS

[0012] The above features and advantages of the present application will be better understood by reading the following detailed description of the embodiments of the application in conjunction with the drawings, in which:

[0013] Figure 1 A flow chart of a method for measuring arc length is shown according to some embodiments of the present application;

[0014] Figure 2 A flow chart of a method for measuring arc length is shown according to some embodiments of the present application;

[0015] Figure 3 A schematic diagram of an image to be measured is shown according to some embodiments of the present application;

[0016] Figure 4 A flow chart of a straight line fitting of a weld edge in a non-welding state is shown according to some embodiments of the present application;

[0017] Figure 5 A schematic diagram of a line support domain in a straight line fitting process is shown according to some embodiments of the present application;

[0018] Figure 6 A block diagram of a measuring device for arc length is shown according to some embodiments of the present application.

[0019] Reference signs:

[0020] S110-S140 steps;

[0021] 300 image sequence;

[0022] 310 arc detection block;

[0023] 410 original weld image;

[0024] 420 gradient image;

[0025] 430 straight line detection image;

[0026] 440 weld edge fitting image;

[0027] 510 image one;

[0028] 520 image two;

[0029] 530 image three;

[0030] 531 region one;

[0031] 532 region two;

[0032] 533 Region three;

[0033] 600 Measurement device of arc length;

[0034] 610 Memory;

[0035] 620 Processor. DETAILED DESCRIPTION

[0036] The advantages and benefits of the present application will become apparent upon reading the following description in conjunction with the appended claims and drawings, of which:

[0037] In the description of the present application, it should be noted that the terms "mounted", "connected" and "linked" should be construed broadly in accordance with their usage in the present application unless otherwise explicitly defined and limited, for example, they can be fixed connection, detachable connection, or integral connection; can be mechanical connection, electrical connection; can be direct connection, or indirect connection through an intermediate medium; can be internal connection of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0038] In addition, "up", "down", "left", "right", "top", "bottom", "horizontal", "vertical" used in the following description should be understood as the orientation shown in the paragraph and the related drawings. The relative terms are only used for the convenience of description, and do not mean that the device described thereby should be manufactured or operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0039] It can be understood that although the terms "first", "second", "third" and the like can be used herein to describe various components, regions, layers and / or parts, these components, regions, layers and / or parts should not be limited by these terms, and these terms are only used to distinguish different components, regions, layers and / or parts. Therefore, the first component, region, layer and / or part discussed below can be referred to as the second component, region, layer and / or part without departing from some embodiments of the present application.

[0040] As described above, the existing non-contact measurement method based on visual measurement needs to use a coplanar linear camera calibration method, and after setting the calibration board and the measurement object to be strictly coplanar, the conversion relationship between the image coordinates, the camera coordinates and the world coordinates is described by the camera intrinsic matrix and the extrinsic matrix. Once the calibration board and the measurement object are not strictly coplanar, non-coplanar error will be generated, thereby affecting the accuracy and stability of the measurement result. In addition, complex mathematical calculations are involved in solving the camera parameters, which will further lead to low measurement efficiency, poor stability of the measurement result and other problems.

[0041] In order to solve the above problems existing in the prior art, the present application provides a measurement method of arc length, a measurement device of arc length and a computer readable storage medium, which do not need to contact the welding environment and omit the prior calibration step, can be applied to various automatic welding environments, have good adaptability and universality, have strong anti-jitter and anti-interference performance, realize efficient and stable arc length measurement, and thus improve the effectiveness, accuracy and reliability of the measurement result.

[0042] In some non-limiting embodiments, the above-mentioned measurement method of arc length provided by the first aspect of the present application can be implemented via the above-mentioned measurement device of arc length provided by the second aspect of the present application.

[0043] The steps of the above-mentioned measurement method of arc length will be described below in combination with some embodiments of the measurement device of arc length. Those skilled in the art can understand that these embodiments of the measurement device of arc length are only some non-limiting embodiments provided by the present application, which are intended to clearly demonstrate the main concept of the present application and provide some specific schemes for facilitating the public to implement, and do not limit the implementation subject of the steps of the measurement method of arc length. Similarly, the measurement method of arc length is only a non-limiting embodiment provided by the present application, and is not used to limit all working modes or all functions of the measurement device of arc length.

[0044] Please refer to Figure 1 , Figure 1 A flowchart of a measurement method of arc length according to some embodiments of the present application is shown.

[0045] As Figure 1 shown, in some embodiments of the present application, the measurement method of arc length can mainly include steps S110-S140. First, step S110 can be performed: acquiring an arc detection frame of the to-be-measured arc via a to-be-measured arc image.

[0046] Specifically, it can be understood Figure 2 together that Figure 2A flowchart of a method for measuring arc length is shown. Step S110 can be embodied as steps S111-S116.

[0047] As shown in Figure 2 some embodiments, step S111 can be performed first. Specifically, the molten pool images of the entire welding process, including the arc striking stage, the welding stage, and the arc extinguishing stage, can be collected in an offline state first, and the images can be adjusted to a uniform pixel size to adapt the training model. For example, the pixel size can be unified to 256x256.

[0048] Optionally, the quality of the collected molten pool images can be preliminarily screened to remove irrelevant and low-quality images to facilitate subsequent computer recognition. This preliminary screening process can be performed manually or automatically controlled by a computer.

[0049] After the preliminary screening is completed, the arc region can be labeled in the molten pool image by a bounding box as sample data. Specifically, in some optional embodiments, the LabelImg software (LabelImg is a visual image labeling tool) can be used to label the arc region in the molten pool image, and a bounding box can be created to label and display the arc region as sample data.

[0050] Further, optionally, since the Darknet network can be used to perform computer vision tasks, especially for target detection with high detection speed and efficiency, the Darknet network can be selected as a training tool and technical carrier to train an optimal arc detection model. Specifically, the above-mentioned labeling information can be converted into a label format suitable for Darknet network training to establish an automated arc detection database. In this embodiment, the automated arc detection database can be used as a training set to train the arc detection model.

[0051] In some preferred embodiments, as shown in Figure 2 to improve the generalization ability of the model, after step S111 is completed, step S112 can be performed to expand the number of training samples based on the collected sample data using an image data augmentation algorithm. Optionally, the image data augmentation algorithm can use the Mosaic data (splicing) augmentation algorithm. Then, step S113 can be performed to train the arc detection model based on the Darknet network using the training set after the training sample expansion, so that the optimal arc detection model based on the Darknet network can be obtained (step S114).

[0052] Continuing as Figure 2As shown, step S115 can be executed next, which allows for the acquisition of the original images of the real-time video stream or real-time image sequence of the arc under test obtained during the welding process in an online state. Specifically, as shown... Figure 3 As shown, Figure 3 An image sequence 300 of the arc under test acquired during welding is shown, according to some embodiments of the present invention. When the original images in the acquired image sequence 300 have inconsistent image sizes, their pixel dimensions can be normalized to [a specific size]. For example, some original images with a size of 1920×1080 pixels can be normalized to a size of 256×256 pixels.

[0053] Continue as Figure 2 As shown, step S116 can be executed, whereby the image of the electric arc to be tested is input into the trained optimal electric arc detection model. Electric arc detection is performed based on the Darknet network, utilizing the optimal electric arc detection model to process real-time image sequences, thereby detecting the arc region of the electric arc to be tested. Further, as... Figure 3 As shown, the optimal arc detection model can obtain the bounding box data of the arc to be tested, i.e., the arc detection box 310, from the image of the arc to be tested. For arc detection, usually only a rectangular box representing the arc region can be obtained, but the edge that perfectly fits the shape of the arc cannot be obtained. Therefore, the bounding box here represents the detected arc region.

[0054] Back Figure 1 As shown, the arc length detection method provided by the present invention may further include step S120: obtaining two straight line segments of the weld edge based on the weld image in a non-welding state.

[0055] Specifically, such as Figure 2 As shown, step S120 may further include steps S121 to S124.

[0056] like Figure 2 As shown, in some embodiments, step S121 can be performed to obtain the original image of the weld in a non-welding state, and to perform image denoising preprocessing on the original weld image. Figure 4 As shown, a raw image 410 of the weld in a non-welding state can be acquired using a camera. Furthermore, the pixel size of the raw weld image 410 can be adjusted to fit the optimal arc detection model. For example, if the size of the acquired raw weld image 410 is 1920... 1080 pixels, we can first normalize its pixel size to For example, normalized to 256 256 pixel size image. Then, the image can be converted into a grayscale image, and Gaussian filtering can be performed on the grayscale image to remove noise. Those skilled in the art can understand that in some other embodiments, other filtering and noise removal methods can also be selected.

[0057] Then, gradient processing can be performed on the preprocessed image to obtain a plurality of weld fitting candidate line segments.

[0058] Specifically, as shown in Figure 2 , in some embodiments, step S122 can be performed to calculate the gradient of the preprocessed image to obtain the gradient angle of each pixel point. Preferably, a hybrid difference algorithm combining forward difference and central difference can be selected to calculate the gradient of each pixel point i in the x direction and the gradient in the y direction. The calculation formulas of the gradient and the gradient may be as follows:

[0059] In the process of gradient calculation of the pixel point, the single forward difference algorithm is sensitive to noise, and it actually reflects the gradient in the interval [x, x+1]. It only depends on the difference value of two adjacent pixels. If one of the pixels is disturbed by noise (such as random brightness jump), it will directly lead to gradient estimation error. The single central difference algorithm is sensitive to high-frequency noise. Since it estimates the gradient only through the difference value of the function values of the two symmetric points, the noise difference of the two symmetric points will be significantly increased due to the sharp fluctuation of the high-frequency noise in a short distance, and the difference value is directly included in the gradient calculation, thereby masking the true gradient.

[0060] To this end, in the preferred embodiment described above, a hybrid difference algorithm combining forward difference and central difference is adopted, and the difference value of two adjacent pixel pairs is averaged, for example combining and , which essentially performs "small-range smoothing" on the local gradient, thereby offsetting part of the random fluctuations caused by noise (the positive / negative deviation of the noise is weakened after averaging), and making the gradient estimation more stable, that is, the hybrid difference noise suppression capability is stronger.

[0061] Then, according to the gradient in the x direction and the gradient in the y direction, the gradient amplitude and the gradient angle can be obtained. The gradient amplitude and the gradient angle The formulas are as follows: , .

[0062] Due to gradient angle The accuracy depends on and The stability of the gradient obtained using the hybrid difference algorithm mentioned above is related to the following: and gradient The grayscale change trends in the local x and y directions can be combined separately, such as... By averaging the differences in the x-direction between two adjacent rows (y and y+1), the grayscale changes in both the horizontal and vertical directions can be more comprehensively reflected. Compared to traditional operators that rely on only a single pixel pair (such as forward difference and center difference algorithms), this "multi-pixel pair synthesis" method enables... and ratio It is more stable, with smaller gradient angle estimation errors, more stable edge positioning, and more reliable direction judgment, especially for weak edges or noisy areas.

[0063] Gradient processing (calculating image gradients) is the core operation for extracting edge information. For example... Figure 4 As shown, in the original weld image 410, the weld edge exhibits abrupt changes in grayscale values, but it is easily obscured by complex background textures and smooth transition areas. However, the gradient processing described above, by calculating the rate of change of pixel grayscale, can directly "peel off" and highlight the weld edge from the original weld image 410, thereby obtaining a gradient image 420 that makes the weld edge the visual focus of the image.

[0064] Furthermore, returning to Figure 2 As shown, step S123 can be performed to establish a line support domain based on gradient image 420 to obtain multiple candidate line segments for weld edge fitting.

[0065] Specifically, in some optional embodiments, a threshold for the gradient angle difference can be preset. For example, a gradient angle difference threshold can be defined. For arcsine(0.5). The threshold is determined by the gradient angle difference. Adjacent elements with a gradient angle difference less than a threshold are selected. The pixels are aggregated and considered to belong to the same line support domain. The line support domain is established, thereby obtaining the initial result set L of multiple candidate line segments for fitting weld edges.

[0066] For example, it can be combined Figure 5 Common understanding Figure 5 A schematic diagram illustrating the calculation of the line support domain during a line fitting process provided according to some embodiments of the present invention is shown.

[0067] like Figure 5 As shown, Image 1 (510) illustrates local details of the grayscale image. By calculating the gradient of each pixel, a series of line segments in different directions can be obtained. Image 2 (520) shows local details of the pixel gradient. Image 3 (530) shows local details of the line support domain, where regions 1 (531), 2 (532), and 3 (533) represent three different line support domains, corresponding to three different line segments. Various line segments may exist throughout the entire image.

[0068] Furthermore, the minimum bounding rectangle of the line support domain can be used as the candidate line segments for fitting the detected weld edges, thereby obtaining an initial result set L including multiple candidate line segments for fitting weld edges, wherein the initial result set... .

[0069] exist Figure 4 In the illustrated embodiment, an initial result set is obtained. It contains four candidate line segments for weld seam fitting. The line detection image 430 clearly shows the four detected candidate line segments for weld seam edge fitting.

[0070] Continue back Figure 2 As shown, step S124 can then be executed, which allows for the selection of two edge straight line segments of the actual weld from multiple weld fitting candidate line segments based on the constraint conditions.

[0071] Specifically, in some optional embodiments, since the method of calculating the weld line segment per pixel is used in step S123 above, there may be short line segments caused by noise, or other line segments unrelated to the weld. To address this, a length threshold can be used. As the first constraint, the initial result set L of the candidate line segments fitted to the weld edge is traversed, and the candidate line segments that do not meet the requirements in the initial result set L are discarded, thereby filtering out the length constraint result set. ,in, , This represents the candidate line segment fitted to the edge of the j-th weld.

[0072] For example, suppose the first constraint has a length threshold. If the length is 200 pixels, then candidate line segments for fitting weld edges with a length less than 200 pixels can be selected, forming a length-constrained result set. .

[0073] Furthermore, since there may be multiple candidate line segments for fitting weld edges in the image, but the starting position of the weld is fixed, in some preferred embodiments, fixing the starting position can further remove a large number of interfering candidate line segments. To this end, the starting position region B can be used as a second constraint condition, and the length constraint result set can be traversed. Candidate line segments are fitted based on the weld edge. starting pixel position The candidate line segments for fitting that remain within the starting position region B are selected to obtain the starting position constraint result set. ,in, It can be understood that "starting position region B" here refers to a specific pixel area on the image.

[0074] For example, suppose the starting position area Then, candidate line segments for fitting the weld edge within the pixel region where the ordinate of the starting pixel is between [0, 20] can be retained to form the starting position constraint result set. .

[0075] Furthermore, in some preferred embodiments, to filter out non-weld line segments after line detection and obtain the left and right edges of the weld, the distance between two line segments can be used as a third constraint condition. The starting position constraint result set can be traversed using the distance between two candidate line segments as the third constraint condition. From these, candidate line segments are fitted to the two weld edges with the smallest spacing. As two straight line segments at the edges of the actual weld, among which, . Represents the result set of initial position constraints Any two straight line segments in the array, This represents the candidate line segment fitted to the edge of the k-th weld. This represents the distance between any two line segments. This formula represents the distance between any two line segments belonging to the initial position constraint result set. Calculate the minimum distance between any two line segments, and retain the two line segments with the smallest distance, namely: and .

[0076] like Figure 4 As shown, in the weld edge fitting image 440, the candidate line segments 441 and 442 are the two edge line segments of the actual weld. That is, by using the three constraints of length and / or starting position and / or spacing mentioned above, other interfering line segments can be removed, and the two edge line segments of the actual weld can be selected.

[0077] Next, return to Figure 1As shown, the method for measuring the arc length provided by the present application can further include step S130 of obtaining a distance pixel conversion ratio between the actual physical length of the weld edge spacing and the spacing pixel points between the edge straight line segments.

[0078] As shown, step S130 can be implemented as steps S131-S133. Figure 2

[0079] Specifically, in some embodiments, step S131 can be performed to obtain the pixel point number N of the minimum spacing between the two edge straight line segments of the actual weld. Then, step S132 can be performed to obtain the actual physical length (i.e., the physical distance value) len of the weld edge spacing. Then, the distance pixel conversion ratio w between the image pixel number and the actual distance value can be obtained by the following formula:

[0080] Then, as shown, step S140 can be performed to obtain the actual physical length of the to-be-measured arc corresponding to the pixel point number of the arc detection frame via the distance pixel conversion ratio. Figure 1

[0081] As shown, step S140 can be implemented as steps S141-S142. Figure 2

[0082] Specifically, in some embodiments, step S141 can be performed first to obtain the pixel point numbers of the arc region corresponding to the to-be-measured arc in the horizontal direction and the vertical direction of the image, i.e., x and y, according to the arc detection frame data of the to-be-measured arc obtained in step S110. Then, step S142 can be performed to convert the pixel length in the horizontal direction and the vertical direction into the actual physical length corresponding thereto respectively via the distance pixel conversion ratio w, to obtain the actual physical length of the to-be-measured arc in the horizontal direction and the actual physical length of the to-be-measured arc in the vertical direction , wherein the calculation formulas of the actual physical length of the to-be-measured arc in the horizontal direction and the actual physical length of the to-be-measured arc in the vertical direction are as follows:

[0083] For example, it is assumed that the pixel point number N of the minimum spacing between the two weld edge straight line segments is 95. The actual physical length of the weld edge spacing is measured as len = 7 mm, and the calculation method of the distance pixel conversion ratio w is . Further, according to the arc detection frame data, the pixel point numbers of the arc region corresponding to the to-be-measured arc in the x and y directions of the image are obtained as 46,​​​​​​​​ is 81. The actual physical length can be converted from the pixel length by the calculation formula and the actual physical length in the vertical direction The real-time physical length and width of the arc to be measured are obtained by converting the pixel length into the actual physical length by the calculation formula, wherein the real-time physical length is is 3mm, and the actual physical width is is 6mm.

[0084] According to the above-mentioned arc length measurement method provided by the above-mentioned embodiment, by constructing a deep learning network and an edge straight line detection algorithm, the image content is analyzed and calculated, the dynamic and effective calculation of the arc length is realized, the original calibration method can be effectively simplified, the efficient and stable measurement of the arc length is realized, and even when the distance between the camera and the target object changes slightly or the imaging process shakes during the camera imaging process, the measurement result will not be affected, thereby ensuring the effectiveness and reliability of the measurement result and improving the accuracy of the measurement result. And the present application can be adapted to various automatic welding environments, such as gas shielded welding, argon arc welding and other arc welding environments, thereby improving the safety and stability of the system.

[0085] Although the above-mentioned method is illustrated and described as a series of actions in order to simplify the explanation, it should be understood and appreciated that the method is not limited by the order of the actions, because according to one or more embodiments, some actions can occur in a different order and / or concurrently with other actions illustrated and described herein or not illustrated and described herein but can be understood by those skilled in the art.

[0086] So far, the steps of the above-mentioned arc length measurement method provided by the first aspect of the present application have been basically introduced. Next, please refer to Figure 6 , Figure 6 a structural block diagram of an arc length measurement device according to some embodiments of the present application is shown.

[0087] As shown in Figure 6 , in some embodiments, the arc length measurement device 600 can include a memory 610; and a processor 620, which can be connected to the memory 610 and configured to implement the arc length measurement method including the above-mentioned steps S110~S140. The steps corresponding to each computer program instruction in the memory 610 have been described above, and will not be repeated here.

[0088] Further, in some non-limiting embodiments, the third aspect of the present application provides the above computer readable storage medium having stored thereon computer instructions. When the computer instructions are executed by the processor 620, the computer instructions can be used to implement the above method for measuring the length of the electric arc provided by the second aspect of the present application. Alternatively, the user can also acquire the computer readable storage medium by purchase or other means to acquire the computer program instructions of each step of the above method for measuring the length of the electric arc.

[0089] Those skilled in the art will further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans can implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

[0090] The various illustrative logical blocks, modules, and circuits described in connection with the embodiments disclosed herein can be implemented or performed with a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor can be a microprocessor, but in the alternative, the processor can be any conventional processor, controller, microcontroller, or state machine. A processor can also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0091] The steps of a method or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium can be integral to the processor. The processor and the storage medium can reside in an ASIC. The ASIC can reside in a user terminal. In the alternative, the processor and the storage medium can reside as discrete components in a user terminal.

[0092] In one or more exemplary embodiments, the functions described can be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functions can be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media can be any available media that can be accessed by a computer. By way of example, and not limitation, such computer-readable media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk and disc, as used herein, includes compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and blu-ray disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

[0093] In summary, the present application provides a method for measuring arc length, a device for measuring arc length, and a computer readable storage medium, without contacting the welding environment and omitting the step of prior calibration, not only can be applied to various automatic welding environments, with good adaptability and versatility, but also has strong anti-jitter and anti-interference, realizing efficient and stable arc length measurement, thereby improving the effectiveness, accuracy and reliability of the measurement results.

[0094] The foregoing description of the present disclosure has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the present disclosure to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. It is intended that the scope of the present disclosure be limited not with this detailed description, but rather by the claims appended hereto.

Claims

1. A method for measuring the length of an electric arc, characterized in that, Includes the following steps: The arc detection frame of the arc to be tested is obtained from the image of the arc to be tested; Gradient processing is performed on the weld seam image in the non-welding state to obtain an initial result set L of multiple candidate line segments for weld seam edge fitting, wherein the initial result set... ; Based on the constraints, two straight line segments representing the actual weld edge are selected from the candidate line segments fitted to the multiple weld edge edges. via length threshold As the first constraint, a length constraint result set is selected from the initial result set. ,in, , This represents the candidate line segment fitted to the edge of the j-th weld. Using the starting position region B as the second constraint condition, in the length constraint result set Filter out the initial position constraint result set ,in, , , indicating the starting pixel position; and Using the spacing between two candidate line segments as a third constraint, the result set is constrained at the starting position. Candidate line segments were fitted from the two weld edges with the smallest spacing. As the two straight line segments of the actual weld seam, wherein, , This represents the candidate line segment fitted to the edge of the k-th weld. The distance-to-pixel conversion ratio is obtained by comparing the actual physical length of the weld edge spacing with the number of pixels between the two edge line segments; and The actual physical length of the arc to be tested, corresponding to the number of pixels in the arc detection frame, is obtained through the distance-to-pixel conversion ratio.

2. The measurement method as described in claim 1, characterized in that, The step of obtaining the arc detection frame of the arc to be tested from the image of the arc to be tested includes: Acquire images of the molten pool during the arc initiation, welding, and arc extinguishing stages; The arc region is marked with a bounding box in the molten pool image as sample data; and The sample data is used as a training set to train the arc detection model.

3. The measurement method as described in claim 2, characterized in that, The step of training the arc detection model using the sample data as a training set includes: Based on the aforementioned sample data, the training samples are augmented using image data augmentation algorithms; and The arc detection model is trained using the expanded training set of training samples to obtain the optimal arc detection model.

4. The measurement method as described in claim 1, characterized in that, The step of obtaining the two straight line segments of the weld edge based on the weld image in a non-welded state includes: Preprocessing is performed on the original image of the weld in its non-welding state to remove noise. Gradient processing is performed on the preprocessed image to obtain multiple candidate line segments for weld edge fitting; and Based on the constraints, two straight line segments of the actual weld edge are selected from the candidate line segments fitted to the multiple weld edge edges.

5. The measurement method as described in claim 4, characterized in that, The step of performing gradient processing on the preprocessed image to obtain multiple candidate line segments for weld edge fitting includes: Calculate the gradient of the preprocessed image to obtain the gradient angle of each pixel. ; Preset gradient angle difference threshold ;as well as Adjacent elements whose gradient angle difference is less than the threshold The pixels are aggregated and considered to belong to the same line support domain to obtain an initial result set L of multiple candidate line segments for weld edge fitting, wherein the initial result set... .

6. The measurement method as described in claim 5, characterized in that, The gradient of the preprocessed image is calculated to obtain the gradient angle of each pixel. The steps include: The gradient of the pixel in the x-direction is calculated using a hybrid difference algorithm that combines forward difference and center difference. gradient in the y direction ,in, , And i(x, y) represents the pixel value at coordinates (x, y) in the image; and Based on the gradient in the x-direction and the gradient in the y-direction Obtain the gradient angle ,in, .

7. The measurement method as described in claim 1, characterized in that, The step of obtaining the actual physical length of the arc to be measured corresponding to the number of pixels in the arc detection frame via the distance-pixel conversion ratio includes: Using the arc detection frame, the number of pixels in the horizontal and vertical directions of the arc region corresponding to the arc under test in the image are obtained respectively. and ;as well as By using the distance-to-pixel conversion ratio w, the pixel lengths in the horizontal and vertical directions are converted into their corresponding actual physical lengths, thereby obtaining the actual physical length of the arc under test in the horizontal direction. and the actual physical length in the vertical direction ,in, ； 。 8. A device for measuring the length of an electric arc, characterized in that, include: Memory; as well as A processor, connected to the memory, and configured to implement the method for measuring the arc length as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed by the processor, the method for measuring the arc length as described in any one of claims 1 to 7 is implemented.

Images