Welding seam tracking method, system, medium and equipment for carbon steel sheet
By using auxiliary lighting devices and light optimization technology during the welding process of carbon steel thin plates, the problem of blurred weld seam images was solved, clear imaging of the weld seam area and high-precision fitting of the welding trajectory line were achieved, thus improving the weld seam tracking effect.
Patent Information
- Application Number
- CN202511132575.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-13
- Publication Date
- 2025-11-21
AI Technical Summary
In the existing technology, during the welding process of carbon steel thin plates, the welding arc light, spatter and ambient light interference result in low contrast and blurred weld seam images captured by industrial cameras. As a result, deep learning models have difficulty accurately identifying the region of interest in the weld seam, which affects the accuracy of weld seam tracking.
By adding auxiliary lighting devices, the illumination angle and light intensity are automatically adjusted according to the diameter of the weldment and the welding arc length. Combined with the ambient light compensation factor and dynamic weight coefficient, the illumination calculation is optimized to achieve clear imaging of the weld area. The recognition accuracy of the welding trajectory line is improved through mean filtering.
This improved the accuracy of deep learning models in identifying weld areas, enhanced the fitting precision of welding trajectory lines, and ensured the stability and accuracy of weld tracking.
Smart Images
Figure CN120997281A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application belongs to the technical field of image processing, and particularly relates to a carbon steel sheet welding seam tracking method, system, medium and equipment. BACKGROUND
[0002] The statements in this section merely provide background information related to the present application and do not necessarily constitute prior art.
[0003] The carbon steel sheet seamless butt welding technology based on deep learning technology usually acquires the welding seam image of the welding piece through an industrial camera for deep learning, and uses the obtained deep learning model to visually recognize the welding fixed point and the welding seam, so as to quickly locate the welding seam area to extract the welding trajectory line and improve the welding seam tracking effect.
[0004] However, the carbon steel sheet itself is relatively thin, and is more sensitive to the interference of the welding environment. Due to the welding arc light, spatter and environmental light interference, the contrast of the image collected by the industrial camera is low, and the welding seam area is blurred. The deep learning model originally suitable for conventional plates cannot accurately recognize the welding seam region of interest when facing such images, thereby affecting the fitting precision of the welding trajectory line and reducing the precision of the welding seam tracking. SUMMARY
[0005] In order to solve the technical problems existing in the background art, the present application provides a carbon steel sheet welding seam tracking method, system, medium and equipment, which adds an auxiliary lighting device, automatically adjusts the irradiation angle and illumination intensity of the auxiliary lighting device according to different welding piece diameters and welding arc lengths, so that the industrial camera can acquire clear welding seam images of the welding seam area, improves the accuracy of the deep learning model in recognizing the welding seam area, and further improves the precision of fitting the welding trajectory line, so that the welding seam tracking can have a good effect.
[0006] In order to achieve the above purpose, the present application adopts the following technical solutions: The first aspect of the present application provides a carbon steel sheet welding seam tracking method, which comprises: acquiring the welding piece diameter and the welding arc length; adjusting the irradiation angle and the illumination intensity of the lighting device based on the welding piece diameter and the welding arc length; the illumination intensity is proportional to the welding arc length; the irradiation angle is the tangent value mapping of the ratio of the welding piece diameter to the distance from the lighting device to the welding piece, so that the irradiation angle increases with the increase of the welding piece diameter and / or the decrease of the distance from the lighting device to the welding piece; acquiring the to-be-recognized image, extracting the welding trajectory line, calculating the deviation value of the horizontal coordinate of the position of the welding gun and the horizontal coordinate of the welding trajectory line, and driving the welding gun to move towards the welding trajectory line at a certain speed.
[0007] Further, the lighting device is fixed with the camera that collects the image to be identified.
[0008] Further, the light intensity , wherein k is a proportional coefficient, h is the welding arc length, is an ambient light compensation factor.
[0009] Further, the step of extracting the welding trajectory includes: For the image to be identified, the welding seam area and the welding fixed point area are identified by the visual recognition model; After the mean filtering processing of the welding seam area, the welding trajectory in the welding seam area is extracted.
[0010] Further, the window size used in the mean filtering is: n = round[n0 + k x (D / D0-1)]; wherein n0 is the reference window size, the dynamic weight coefficient k = 1 + 0.3 x max(0, |θ-θ0| / θ a -1), θ0 is the standard incident angle, θ a is the angle deviation threshold, θ is the irradiation angle, D0 is the reference welding piece diameter, and D is the welding piece diameter.
[0011] The second aspect of the present application provides a welding seam tracking system for carbon steel sheet, which includes: A data acquisition module configured to acquire the welding piece diameter and the welding arc length; A lighting device adjustment module configured to adjust the irradiation angle and the light intensity of the lighting device based on the welding piece diameter and the welding arc length; the light intensity is proportional to the welding arc length; the irradiation angle is the tangent value mapping of the ratio of the welding piece diameter to the distance between the lighting device and the welding piece, so that the irradiation angle increases with the increase of the welding piece diameter and / or the decrease of the distance between the lighting device and the welding piece; A welding seam tracking module configured to acquire the image to be identified, extract the welding trajectory, calculate the deviation value of the horizontal coordinate of the position of the welding gun and the horizontal coordinate of the welding trajectory, and drive the welding gun to move towards the welding trajectory at a certain speed.
[0012] Further, the lighting device is fixed with the camera that collects the image to be identified.
[0013] Further, the light intensity , wherein k is a proportional coefficient, h is the welding arc length, is an ambient light compensation factor.
[0014] The third aspect of the present application provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the steps of the weld tracking method of the carbon steel sheet as described above.
[0015] The fourth aspect of the present application provides a computer device, which comprises a computer readable storage medium, a processor and a computer program stored on the computer readable storage medium and executable on the processor, and the processor executes the program to implement the steps of the weld tracking method of the carbon steel sheet as described above.
[0016] Compared with the prior art, the present application has the following beneficial effects: The present application adds an auxiliary lighting device, automatically adjusts the irradiation angle and illumination intensity of the auxiliary lighting device according to different welding part diameters and welding arc lengths, makes the industrial camera obtain a clear weld image of the weld area, improves the accuracy of the deep learning model in identifying the weld area, and further improves the precision of the fitted welding trajectory line, so that the weld tracking can have a good effect.
[0017] The present application utilizes the dynamic correlation between the welding arc length and the arc light intensity change in the welding process: when the arc length increases (such as the welding gap becomes wider), the illumination calculation value is increased to compensate for the insufficient brightness of the weld area caused by the dispersion of the arc light; when the arc length decreases, the calculated illumination is reduced to avoid overexposure.
[0018] The optimal incidence angle of the light source of the present application is the dynamic adaptation result of the distance from the light source to the surface of the welding part and the diameter of the welding part: the focusing angle (small angle) when the distance is far, the diffusion angle (large angle) when the distance is short; the widening angle (large angle) when the diameter is large, and the tightening angle (small angle) when the diameter is small, finally realizing uniform illumination of the weld area.
[0019] In the calculation of the illumination intensity, the present application introduces an ambient light compensation factor, which can offset the fluctuation of the workshop ambient light (such as strong light during the day and weak light at night), so that the calculated illumination intensity always adapts to the imaging needs of the weld, providing a stable and accurate illumination basis for subsequent welding trajectory line identification and extraction.
[0020] The present application realizes the precise adjustment of the filter window size by dynamically correlating the incidence angle with the weight coefficient and adapting the welding part diameter with a nonlinear correction term: for large-diameter welding parts, the filter effect can be enhanced by increasing the window size to avoid noise interference such as welding spatter; for small-diameter welding parts, the window size can be reduced to effectively retain the fine weld features and prevent feature loss; when the incidence angle deviates from the standard value, the dynamic adjustment of the weight coefficient can specifically strengthen the adaptation ability of the window to uneven illumination (shadows or overexposure), reducing the impact of extreme illumination on image quality. BRIEF DESCRIPTION OF DRAWINGS
[0021] The accompanying drawings, which form a part of this specification, are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification. The embodiments of the application, and their
[0022] Figure 1 is a flow chart of a welding seam tracking method of a carbon steel sheet plate according to an embodiment of the application; Figure 2 is a schematic diagram of a welding seam image acquired by a camera without an illuminating device according to an embodiment of the application; Figure 3 is a schematic diagram of a welding seam image acquired by a camera with an illuminating device according to an embodiment of the application; Figure 4 is a schematic diagram of a welding seam tracking method of a carbon steel sheet plate according to an embodiment of the application; Figure 5 is a schematic diagram of a computer device according to an embodiment of the application. DETAILED DESCRIPTION
[0023] In order to make the objects, technical solutions and advantages of the embodiments of the application clearer, the technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application.
[0024] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the application. Unless otherwise indicated, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the application pertains.
[0025] Embodiment I The embodiment provides a welding seam tracking method of a carbon steel sheet plate.
[0026] As introduced in the background, the current seamless butt welding of carbon steel sheet plates based on deep learning has the problem that the welding seam region of the welding seam image acquired by the industrial camera is not clear enough, such as Figure 2 as shown, which affects the accuracy of visual recognition of the welding seam region of interest, reduces the fitting accuracy of the welding trajectory line, and further affects the accuracy of welding seam tracking.
[0027] The welding seam tracking method of the carbon steel sheet plate provided in the embodiment adds an auxiliary illuminating device, fixes the illuminating device and the camera together, adjusts the irradiation angle and illumination intensity of the auxiliary illuminating device according to different welding piece diameters and welding arc lengths, so that the industrial camera acquires a clear welding seam image of the welding seam region, improves the accuracy of welding seam region of interest recognition, further improves the accuracy of fitting the welding trajectory line, and enables the welding seam tracking to have a good effect.
[0028] The welding seam tracking method of the carbon steel sheet plate provided in the embodiment, as Figure 1As shown, comprising the following steps: Step 1, making data set, training visual recognition model.
[0029] Step 101, labeling the weld image, respectively labeling the weld solid point and the weld with a rectangle, obtaining the labeled data, packing and arranging the labeled picture and the labeled data into a data set, and using a format conversion tool to convert the data set into coco (Common Objects in Context) format.
[0030] Step 102, training the arranged data set, wherein 80% of the data set is used to train the convolutional neural network, and 20% of the data set is used to verify the convolutional neural network, to obtain the visual recognition model.
[0031] Step 2, obtaining the diameter data of the welding part to determine the welding process.
[0032] The welding part is fixed on the welding equipment using a clamp, the diameter of the welding part is measured using a measuring tool, the diameter data of the welding part is input into the welding controller, the welding controller calls the pre-set welding process parameters, adjusts the position of the welding gun, and makes the tungsten needle of the welding gun centered with the carbon steel sheet weld, and prepares for welding.
[0033] Step 3, obtaining the welding arc length information.
[0034] Start welding, start welding, and obtain arc length information through an arc sensor, and feed back to the welding controller in real time.
[0035] Step 4, the welding controller calculates the illumination intensity and the illumination angle, and adjusts the position and angle of the illumination.
[0036] During the welding process, the illumination angle and the illumination intensity of the lighting device are automatically adjusted according to the obtained arc length information and the diameter information of the welding part.
[0037] Step 401, calculating the illumination intensity , , wherein k is a proportional coefficient, h is the real-time welding arc length, is an environmental light compensation factor, which is adjusted according to the intensity of the workshop environmental light.
[0038] That is, the arc length dynamically correlates the change of the arc light intensity in the welding process - when the arc length increases (such as the welding gap becomes wider), the illumination calculation value is increased to compensate for the insufficient brightness of the weld area caused by the dispersion of the arc light; when the arc length decreases, the calculated illumination is reduced to avoid overexposure.
[0039] , wherein the environmental light compensation factor can offset the fluctuation of the workshop environmental light (such as strong light during the day and weak light at night), so that the calculated illumination intensity always adapts to the imaging needs of the weld, and provides a stable and accurate illumination basis for subsequent welding trajectory line recognition and extraction.
[0040] Step 402, calculating the optimal incident angle of the light source , wherein L is the distance from the light source to the surface of the welding piece, and D is the diameter of the welding piece.
[0041] That is, the optimal incident angle of the light source is the dynamic adaptation result of L and D: a focusing angle (small angle) when the distance is far, a diffusion angle (large angle) when the distance is close; a widening angle (large angle) when the diameter is large, and a tightening angle (small angle) when the diameter is small, so as to finally realize uniform illumination of the weld area.
[0042] Step 403, the welding controller automatically adjusts the angle and illumination intensity of the lighting device according to the calculated illumination intensity and optimal incident angle, so that a bright band region can be formed on the weld image, as shown in Figure 3 .
[0043] Step 5, obtaining a to-be-recognized image and extracting a welding trajectory line.
[0044] Step 501, the visual recognition model obtained by step 1 can obtain the weld area and the welding fixed point area of the to-be-recognized image, and can quickly distinguish the welding fixed point and the weld, so as to position the weld, as shown in Figure 4 .
[0045] Step 502, performing mean filtering processing on the ROI area (weld area) of the to-be-recognized image, and the filtering window is n x n.
[0046] In this embodiment, the size of the filtering window is adjusted according to the diameter D of the welding piece and the calculated incident angle θ: n = round[n0+k×(D / D0-1)]; wherein when the length n of the calculated filtering window is less than 3, n = 3; when the length n of the calculated filtering window is greater than 7, n = 7; n0 is the reference window size (n0 = 3); the dynamic weight coefficient k = 1 + 0.3 x max(0, |θ-θ0| / θ a -1), θ0 is the standard incident angle (45°), and θ a is the angle deviation threshold (15°), k increases with the increase of the deviation, and the adjustment of the window size is strengthened, at this time, the weld image is easy to produce shadow or overexposure due to uneven illumination, and the noise needs to be suppressed by increasing the window; D0 is the reference welding piece diameter; for the welding piece diameter term (D / D0-1), when D > D0 (large diameter), the term is positive, and the window can be appropriately increased (the weld feature is obvious, and stronger filtering is allowed), when D < D0 (small diameter), the term is negative, and the window size is reduced (to avoid that the subtle features are filtered); round() represents taking the integer of the result, so as to ensure that the window size is a positive integer.
[0047] The embodiment realizes precise adjustment of the filter window size by dynamically associating the incident angle with the weight coefficient, and adapting the welding piece diameter with a nonlinear correction term: for large-diameter welding pieces, the filter effect can be enhanced by increasing the window size to avoid noise interference such as welding spatter; for small-diameter welding pieces, the fine weld seam features can be effectively retained by reducing the window size to prevent feature loss; when the incident angle deviates from the standard value, the dynamic adjustment of the weight coefficient can specifically strengthen the adaptation capability of the window to uneven illumination (shadows or overexposure), and reduce the influence of extreme illumination on image quality.
[0048] Step 503: For each row of the region of interest (ROI) after filtering, search for the minimum wave valley as the weld seam feature point and store it in the weld seam feature point array.
[0049] Step 503: Perform linear Hough transformation on the weld seam feature point array to obtain the welding trajectory line.
[0050] Step 6: Weld seam tracking.
[0051] Calculate the deviation i of the horizontal coordinate of the current welding gun position and the horizontal coordinate of the welding trajectory line, when the right deviation is the left deviation, and a deviation threshold is set, if the absolute value is less than the deviation threshold, drive the welding gun to move at a certain speed towards the welding trajectory line to realize weld seam tracking.
[0052] The application adds an auxiliary lighting device, automatically adjusts the irradiation angle and illumination intensity of the auxiliary lighting device according to different welding piece diameters and welding arc lengths, so that the industrial camera obtains a clear weld seam image of the weld seam area, improves the accuracy of the deep learning model in identifying the weld seam area, and further improves the precision of fitting the welding trajectory line, so that the weld seam tracking can have a good effect.
[0053] Embodiment two The embodiment provides a weld seam tracking system for carbon steel sheet, which comprises: a data acquisition module configured to acquire the welding piece diameter and the welding arc length; an illumination device adjustment module configured to adjust the irradiation angle and the illumination intensity of the illumination device based on the welding piece diameter and the welding arc length; the illumination intensity is proportional to the welding arc length; the irradiation angle is the tangent value mapping of the ratio of the welding piece diameter to the distance from the illumination device to the welding piece, so that the irradiation angle increases with the increase of the welding piece diameter and / or the decrease of the distance from the illumination device to the welding piece; a weld seam tracking module configured to acquire an image to be identified, extract a welding trajectory line, calculate the deviation value of the horizontal coordinate of the position of the welding gun and the horizontal coordinate of the welding trajectory line, and drive the welding gun to move at a certain speed towards the welding trajectory line.
[0054] Furthermore, the lighting device is fixed together with the camera that acquires the image to be identified.
[0055] Furthermore, the light intensity , Where k is the proportionality coefficient and h is the welding arc length. It is the ambient light compensation factor.
[0056] Furthermore, the step of extracting the contact trajectory includes: For the image to be identified, the weld seam area and weld solidification point area are identified through a visual recognition model; After applying mean filtering to the weld area, the welding trajectory lines within the weld area are extracted.
[0057] Furthermore, the window size used in the mean filtering is: n = round[n0 + k × (D / D0 - 1)]; Where n0 is the baseline window size, and the dynamic weighting coefficient k = 1 + 0.3 × max(0, |θ - θ0| / θ) a -1), θ0 is the standard incident angle, θ a θ is the irradiation angle, D0 is the reference weldment diameter, and D is the weldment diameter.
[0058] It should be noted that each module in this embodiment corresponds one-to-one with each step in Embodiment 1, and their specific implementation processes are the same, so they will not be repeated here.
[0059] Example 3 This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the weld seam tracking method for carbon steel thin plates as described in Embodiment 1 above.
[0060] Example 4 This embodiment provides a computer device, such as... Figure 5 As shown, the system includes a computer-readable storage medium 1003, a processor 1001, a communication interface 1002, and a computer program stored on the computer-readable storage medium 1003 and executable on the processor 1001. The processor 1001, communication interface 1002, and computer-readable storage medium 1003 can be connected via a bus or other means. The communication interface 1002 is used to receive and transmit data. When the processor 1001 executes the program, it implements the steps in the weld seam tracking method for carbon steel thin plates as described in Embodiment 1 above.
[0061] The above merely provides the preferred embodiments of the present application, and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the principles and technical scope of the present application shall fall into the scope of the present application.
Claims
1. A method for tracing weld seams in carbon steel thin plates, characterized in that, include: Obtain the diameter of the workpiece and the welding arc length; Based on the diameter of the workpiece and the welding arc length, the illumination angle and light intensity of the lighting device are adjusted; the light intensity is proportional to the welding arc length; the illumination angle is mapped to the tangent of the ratio of the diameter of the workpiece to the distance from the lighting device to the workpiece, so that the illumination angle increases as the diameter of the workpiece increases and / or the distance from the lighting device to the workpiece decreases. The image to be identified is acquired, the welding trajectory line is extracted, and the deviation between the horizontal coordinate of the welding torch position and the horizontal coordinate of the welding trajectory line is calculated. The welding torch is then driven to move towards the welding trajectory line at a certain speed.
2. The weld seam tracking method for carbon steel thin plates as described in claim 1, characterized in that, The lighting device is fixed together with the camera that acquires the image to be identified.
3. The weld seam tracking method for carbon steel thin plates as described in claim 1, characterized in that, The light intensity , Where k is the proportionality coefficient and h is the welding arc length. This is the ambient light compensation factor.
4. The weld seam tracking method for carbon steel thin plates as described in claim 1, characterized in that, The steps for extracting the trajectory lines include: For the image to be identified, the weld seam area and weld solidification point area are identified through a visual recognition model; After applying mean filtering to the weld area, the welding trajectory lines within the weld area are extracted.
5. The weld seam tracking method for carbon steel thin plates as described in claim 1, characterized in that, The window size used for the mean filtering is: n = round[n0 + k × (D / D0 - 1)]; Where n0 is the baseline window size, and the dynamic weighting coefficient k = 1 + 0.3 × max(0, |θ - θ0| / θ) a -1), θ0 is the standard incident angle, θ a θ is the irradiation angle, D0 is the reference weldment diameter, and D is the weldment diameter.
6. A weld seam tracking system for carbon steel thin plates, characterized in that, include: The data acquisition module is configured to acquire the diameter of the weldment and the welding arc length. The lighting device adjustment module is configured to: adjust the illumination angle and light intensity of the lighting device based on the diameter of the weldment and the welding arc length; the light intensity is proportional to the welding arc length; the illumination angle is mapped to the tangent of the ratio of the diameter of the weldment to the distance from the lighting device to the weldment, so that the illumination angle increases as the diameter of the weldment increases and / or the distance from the lighting device to the weldment decreases; The weld seam tracking module is configured to: acquire the image to be identified, extract the welding trajectory line, calculate the deviation between the horizontal coordinate of the welding torch position and the horizontal coordinate of the welding trajectory line, and drive the welding torch to move towards the welding trajectory line at a certain speed.
7. The weld seam tracking system for carbon steel thin plates as described in claim 6, characterized in that, The lighting device is fixed together with the camera that acquires the image to be identified.
8. The weld seam tracking system for carbon steel thin plates as described in claim 6, characterized in that, The light intensity , Where k is the proportionality coefficient and h is the welding arc length. This is the ambient light compensation factor.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the weld seam tracking method for carbon steel sheet as described in any one of claims 1-5.
10. A computer device comprising a computer-readable storage medium, a processor, and a computer program stored on the computer-readable storage medium and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the weld seam tracking method for carbon steel thin plates as described in any one of claims 1-5.