A Method for Real-time Measuring the Dry Extension of Welding Wire by a Molten Pool Monitoring Camera

Abstract

The present invention discloses a method for real-time measurement of the dry length of a welding wire by a molten pool monitoring camera, comprising the following steps: the molten pool monitoring camera takes real-time photos to collect molten pool images; the images are read and preprocessed to maximize the difference in image gray values between the welding wire and the background; the straight line at the bottom of the nozzle and the characteristic points at the bottom end of the welding wire are analyzed and determined; the distance from the characteristic points at the bottom end of the welding wire to the straight line at the bottom of the nozzle is calculated; according to the pre-stored calibration value, this distance is calibrated to obtain the actual dry length of the welding wire. The method of the present invention can automatically identify the position information of the welding wire and the nozzle through machine vision, calculate the result of the dry length, and has high measurement accuracy and small error; by collecting images with the molten pool camera, the photographing speed is fast, and with preprocessing and calculation, the dry length of the welding wire can be quickly calculated, ensuring that the current photographed image can be processed to obtain the result before the molten pool camera takes the next image.

Description

Technical Field

[0001] The present invention relates to a method for real-time measuring the dry extension length of a welding wire by a molten pool monitoring camera, and particularly to automatically identifying the welding wire through machine vision and calculating the result of the dry extension length based on the nozzle position information, belonging to the field of welding technology. Background Art

[0002] In production welding, to ensure the production quality, the quality of the welding process requires the operator to adjust the welding parameters by observing the welding area, the melting position of the welding wire, and the molten pool state, so as to ensure the consistency of the weld formation. As is well known, the welding environment is extremely harsh, accompanied by smoke, strong arc light radiation, and spatter, etc. The operator needs to wear strict protective measures. In the past, operators generally relied on welding helmets and protective screens to avoid direct contact with the harmful radiation generated by the welding arc. With the progress and development of technology, nowadays, when monitoring the welding process, it is already possible to rely on the welding molten pool monitoring camera to solve the problem.

[0003] The molten pool camera only solves the problem that there is no need for the human eye to wear strict protective measures to observe the melting state of the welding wire. However, in actual production, to ensure good weld formation, different diameters of welding wires have the optimal wire extension length (dry extension length). The dry extension length of the welding wire refers to the distance L from the end of the welding wire 1 to the end of the conductive nozzle 2, as Figure 1 shown. This section of the welding wire will generate resistance heat during surfacing. The melting speed of the welding wire is jointly determined by the arc 3 and the resistance heat. The melting speed of the welding wire is proportional to the dry extension length of the welding wire, that is, the longer the dry extension length, the faster the melting speed of the welding wire. How can the size of the dry extension length be visually and real-time identified through the video monitored by the molten pool monitoring camera, so as to provide reference information for the producer to adjust the equipment parameters? Currently, even when using the molten pool monitoring camera to monitor the molten pool state during the welding process, the size information of the dry extension length still depends on manual judgment with the naked eye, and the accuracy is affected by manual experience and the working state of the personnel.

[0004] The difficulties in measuring the dry extension length of the molten pool welding wire are as follows:

[0005] 1. The human eye observation is only an approximate distance, and accurate size information cannot be obtained.

[0006] 2. Affected by the welding arc interference, the identified dry extension length of the welding wire is unstable.

[0007] 3. The molten pool camera is a high-speed camera with a very fast shooting speed. Therefore, high-speed algorithms are required for visual recognition and processing. Before the molten pool camera captures the next image, it is required that the visual recognition and processing algorithm must be able to process the current captured image to obtain the result, and the requirements for the speed of the visual recognition and processing algorithm and the response of the hardware system of the processing algorithm are very high. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide a solution for real-time measurement of the dry length of the welding wire in cooperation with a molten pool camera. The position information of the welding wire and the nozzle can be automatically recognized by machine vision, and the result of the dry length can be calculated. Specific dimension information is provided to the producer to facilitate the adjustment of equipment parameters and improve the production quality of products.

[0009] To solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0010] A method for real-time measurement of the dry length of a welding wire by a molten pool monitoring camera, comprising the following steps:

[0011] Step 1: The molten pool monitoring camera takes real-time photos to collect molten pool images;

[0012] Step 2: Read the image and perform preprocessing to maximize the difference in image gray values between the welding wire and the background;

[0013] Step 3: Analyze and determine the straight line at the bottom of the nozzle and the characteristic points at the bottom end of the welding wire;

[0014] Step 4: Calculate the distance |QE| from the characteristic points at the bottom end of the welding wire to the straight line at the bottom of the nozzle;

[0015] Step 5: Calibrate the distance |QE| according to the pre-stored calibration value to obtain the actual dry length of the welding wire T = |QE| * Scale * cosθ;

[0016] Wherein, Scale is the calibration value of the actual length in millimeters corresponding to each pixel of the molten pool monitoring camera, and θ is the calibration value of the installation angle of the molten pool monitoring camera.

[0017] Further, the preprocessing step of Step 2 is: in the read image, draw a region of interest ROI that includes a part of the front end of the welding wire in the front end region of the welding wire. The parameters of the region of interest ROI are: X ROI , Y ROI , W, H; wherein, X ROI is the starting point X-direction coordinate in the image coordinate system; Y ROI is the starting point Y-direction coordinate in the image coordinate system; W is the width value of the rectangular area; H is the height value of the rectangular area; the units of each parameter are all pixels;

[0018] According to the different materials of the welding wire, one or more algorithm modules are used to preprocess the image in the region of interest to maximize the difference in gray values between the welding wire image and the background image.

[0019] Further, in Step 2, the read image is preprocessed using an image equalization algorithm module.

[0020] Further, when the material of the welding wire is carbon steel, the Gaussian filtering module, dilation morphology module and image equalization algorithm module are used for preprocessing;

[0021] When the material of the welding wire is stainless steel, the Soble algorithm module and image equalization algorithm module are used for preprocessing;

[0022] When the material of the welding wire is copper, the mean filtering module, Soble algorithm module and image equalization algorithm module are used for preprocessing.

[0023] Further, in step 3, the steps for determining the straight line at the bottom of the nozzle are specifically as follows:

[0024] Find two points P1(X1, Y1) and P2(X2, Y2) at the bottom of the nozzle, and fit them into the straight line equation of the bottom of the nozzle Ax + By + C = 0, where A = Y2 - Y1, B = X1 - X2, C = X2 * Y1 - X1 * Y2, A, B, and C are the parameters of the fitted straight line equation, and A and B are not all zero, and x and y are the horizontal and vertical coordinates of any point on the straight line.

[0025] Further, in step 3, the steps for determining the characteristic point at the bottom end of the welding wire are specifically as follows:

[0026] Step 31, reduce the ROI image area: For the region of interest ROI in step 2, calculate the parameters of the reduced image area: X ROI -M, Y ROI -N, W - P, H - Q; where M and N are the offsets of the starting point coordinates in the X and Y directions in the image coordinate system respectively, and P and Q are the reduction values of the width and height of the rectangular area, and the units are all pixels;

[0027] Step 32, image gray value: Calculate the average gray value of the reduced image area, and compare it with the set first threshold. If it is greater than the set first threshold, when proceeding to the next step of scanning for points, execute according to the process with bright arc light; if it is less than or equal to the set first threshold, when proceeding to the next step of scanning for points, execute according to the process without bright arc light;

[0028] Step 33, scan for intersection points: Scan and search for all intersection points of the welding wire with bright arc light and without bright arc light in the reduced image area;

[0029] Compare the Y values of the vertical coordinates of all intersection points, and select the point with the largest Y value of the vertical coordinate as the characteristic point at the bottom end of the welding wire, and its coordinates are (Xq, Yq), where Yq is the point with the largest vertical coordinate value, and Xq is the corresponding horizontal coordinate value of the point with the largest vertical coordinate value.

[0030] Further, in step 33, the steps of scanning and searching for all intersection points between the welding wire and the areas with bright arc light in the reduced image area are as follows: Set the second threshold value of the gray value change amount, calculate the change amount between the gray value at any position with bright arc light and the gray value of the welding wire, compare the change amount of the two with the set second threshold value. If the change amount is greater than the set second threshold value, the point at this position is the intersection point between the welding wire and the area with bright arc light; repeat the above process to traverse and scan all the intersection points between the welding wire and the areas with bright arc light in the reduced image area.

[0031] Further, in step 33, the steps of scanning and searching for all intersection points between the welding wire and the areas without bright arc light in the reduced image area are as follows: Set the third threshold value of the gray value change amount, calculate the change amount between the gray value at any position without bright arc light and the gray value of the welding wire, compare the change amount of the two with the set third threshold value. If the change amount is greater than the set third threshold value, the point at this position is the intersection point between the welding wire and the area without bright arc light; repeat the above process to traverse and scan all the intersection points between the welding wire and the areas without bright arc light in the reduced image area.

[0032] Further, in step 4, calculate the distance from the characteristic point at the bottom end of the welding wire to the straight line at the bottom of the nozzle

[0033]

[0034] Further, in step 5, the calibration value Scale of the actual length in millimeters corresponding to each pixel of the molten pool monitoring camera is obtained through the following calibration steps:

[0035] According to the actually measured diameter R of the welding wire and the width value K visually measured by the molten pool monitoring camera, calculate the calibration value Scale of the actual length in millimeters corresponding to each pixel of the molten pool monitoring camera as Scale = R / K.

[0036] Further, in step 5, the calibration value θ of the installation angle of the molten pool monitoring camera is obtained through the following calibration steps:

[0037] According to the actually measured length L of the welding wire extending from the nozzle and the length L1 of the welding wire visually measured by the molten pool monitoring camera, use the formula cosθ = L1 / L to calculate the calibration value θ of the installation angle of the molten pool monitoring camera.

[0038] The beneficial effects achieved by the present invention:

[0039] The method for the molten pool monitoring camera of the present invention to measure the wire dry elongation in real time can automatically identify the wire and nozzle position information through machine vision, calculate the result of the dry elongation, with high measurement accuracy and small error. This method collects images through the molten pool camera, and the photographing speed is very fast. Combined with preprocessing and calculation, the dry elongation of the wire can be quickly calculated, ensuring that the result of the current photographed image can be processed before the molten pool camera captures the next image. It can provide producers with the dry elongation size information of the wire with high precision, facilitating the adjustment of equipment parameters and improving the production quality of products. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a schematic diagram of the existing wire dry elongation and nozzle position;

[0041] Figure 2 is the calibration flowchart of the method of this embodiment;

[0042] Figure 3 is the schematic diagram of the selected point position at the bottom of the nozzle in this embodiment;

[0043] Figure 4 is the flowchart of the method steps of this embodiment;

[0044] Figure 5 is the image of the front end area of the aluminum alloy wire read by this method;

[0045] Figure 6 is to adopt this method for Figure 5 the effect diagram after image preprocessing. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] The present invention will be further described below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and cannot be used to limit the protection scope of the present invention.

[0047] Embodiment 1

[0048] Combined with Figure 4 , the method for the molten pool monitoring camera of this embodiment to measure the wire dry elongation in real time is described.

[0049] The method for the molten pool monitoring camera of this embodiment to measure the wire dry elongation in real time includes the following steps:

[0050] Step 1. During the normal operation of the molten pool, the molten pool monitoring camera takes pictures in real time to collect molten pool images and transmits and saves the images to the specified folder frame by frame in real time.

[0051] Step 2. Read the images from the specified folder and preprocess the images.

[0052] The specific preprocessing process is as follows:

[0053] In the read image, a region of interest ROI that includes a part of the front end of the welding wire is drawn in the front end region of the welding wire. In this embodiment, the region of interest ROI is a rectangular region.

[0054] The parameters of the rectangular region are: X ROI , Y ROI , W, H. Among them, X ROI is the starting point X-direction coordinate in the image coordinate system; Y ROI is the starting point Y-direction coordinate in the image coordinate system; W is the width value of the rectangular region; H is the height value of the rectangular region; the unit of each parameter is pixel.

[0055] Use the image equalization algorithm module to preprocess the image in the region of interest, so that the gray value difference between the welding wire image and the background image is the largest and there is an obvious difference.

[0056] The gray value of the image has a standard definition in vision. The brightest value is 255, and the darkest value is 0. The image is composed of pixel points, and each pixel point has a value within 0-255. For example, if the gray value of the welding wire is 80, then the gray value of the area in the region of interest that is not the welding wire is preferably greater than 130.

[0057] When the normal molten pool camera takes pictures, it will filter out the arc light. There is no bright arc light in the image, and the image equalization algorithm module plays an important equalization effect. Occasionally, when there is a very bright arc light that cannot be filtered out, at this time, the arc light gray value in the bright arc light region is near 255. The image equalization algorithm module equalizes the gray value not near 255 in the bright arc light region to near 255, so that the gray value difference between the background image and the welding wire image is the largest and the distinction effect is the most obvious.

[0058] Taking the aluminum alloy welding wire as an example, Figure 5 , Figure 6 The effect diagrams before and after processing are respectively shown in the small squares of .

[0059] According to the different materials of the welding wire, such as carbon steel, stainless steel, copper and other materials, algorithm modules such as Soble algorithm module, opening / closing operation module, dilation / erosion morphology module, Gaussian filter module, median / mean filter module, etc. can be selected to perform single or combined preprocessing on the region of interest image.

[0060] For example, when the material of the welding wire is carbon steel, the Gaussian filter module + dilation morphology module + image equalization algorithm module can be adopted;

[0061] When the material of the welding wire is stainless steel, the Soble algorithm module + image equalization algorithm module can be adopted;

[0062] When the material of the welding wire is copper, the average filtering module + Soble algorithm module + image equalization algorithm module can be adopted.

[0063] The purpose of performing single or combined preprocessing on the image of the region of interest using different algorithm modules is to maximize the difference in gray values between the welding wire, the bright arc light, and the background image, making them significantly distinguishable. By using different preprocessing algorithm modules for different welding wire materials, the measurement of the dry elongation of welding wires with different materials can be compatible, making the compatibility and practicality of this method stronger.

[0064] Step 3: Analyze and determine the straight line at the bottom of the nozzle and the characteristic points at the bottom end of the welding wire.

[0065] In this embodiment, the steps for determining the straight line at the bottom of the nozzle are specifically as follows:

[0066] Select two points P1(X1, Y1) and P1(X2, Y2) at the bottom of the nozzle. In this embodiment, points P1 and P2 are respectively selected near about 1 / 3 and 2 / 3 of the bottom of the nozzle 2, as Figure 3 shown. Since the camera moves synchronously with the welding torch and the position of the camera relative to the welding torch is fixed and unchanged, it is not necessary to reselect the points unless the nozzle is replaced. Fit the straight line at the bottom of the nozzle into Ax + By + C = 0, where A = Y2 - Y1, B = X1 - X2, C = X2 * Y1 - X1 * Y2, A, B, and C are the parameters of the fitted straight line equation, and A and B are not all zero, and x and y are the horizontal and vertical coordinates of any point on the straight line. This step is not affected by aspects such as the material, size, and shape of the nozzle. When the nozzle is replaced, only the coordinates of points P1 and P2 need to be updated, and there is no need to adjust the algorithm, which is convenient for personnel operation.

[0067] In this embodiment, the steps for determining the characteristic points at the bottom end of the welding wire are specifically as follows:

[0068] Step 31: Reduce the ROI image area: For the region of interest ROI in Step 2, calculate the parameters of the reduced image area: X ROI -M, Y ROI -N, W - P, H - Q. Where M and N are the offsets of the starting point coordinates in the X and Y directions in the image coordinate system respectively, and P and Q are the reduction values of the width and height of the rectangular area, and the units are all pixels.

[0069] Step 32: Image gray value: Calculate the average value of the gray values of the reduced image area, and compare it with the set threshold one. If it is greater than the set threshold one, when entering the next step of scanning for the intersection point, execute according to the process with bright arc light; if it is less than or equal to the set threshold one, when entering the next step of scanning for the intersection point, execute according to the process without bright arc light.

[0070] Step 33. Scan to find the intersection points: Scan and search for all intersection points between the welding wire and the areas with and without bright arc light within the reduced image area.

[0071] For the process with bright arc light, set the threshold two for the change in gray value. Calculate the change in gray value between the gray value at any position with bright arc light and the gray value of the welding wire, and compare the change between the two with the set threshold two. If the change is greater than the set threshold two, then the point at this position is the intersection point between the welding wire and the area with bright arc light; repeat the above process to traverse and scan all intersection points within the reduced image area.

[0072] In this embodiment, the minimum change range parameter of the image gray value is set to be around 150, and the threshold two for the change in gray value is set to 150.

[0073] For example, if the set threshold for the change in gray value is 150, the gray value at a certain position with bright arc light is 250, and the gray value of the welding wire is 80, the change in gray value between the two is 170, which is greater than the set threshold 150. Then the point at this position is the intersection point between the welding wire and the arc light. Traverse and scan all points within the reduced image area, and find the points where the change in gray value is greater than the set threshold 150 as the intersection points between the welding wire and the arc light.

[0074] For the process without bright arc light, set the threshold three for the change in gray value. Calculate the change in gray value between the gray value at any position without bright arc light and the gray value of the welding wire, and compare the change between the two with the set threshold three. If the change is greater than the set threshold three, then the point at this position is the intersection point between the welding wire and the area without bright arc light; repeat the above process to traverse and scan all intersection points within the reduced image area.

[0075] In this embodiment, the minimum change range parameter of the gray value is set to be around 40; the threshold three for the change in gray value is set to 40.

[0076] For example, if the set threshold for the change in gray value is 40, the gray value at a certain position without bright arc light is 130, and the gray value of the welding wire is 80, the change in gray value between the two is 50, which is greater than the set threshold 40. Then the point at this position is the intersection point between the welding wire and the area without bright arc light. Traverse and scan all points within the reduced image area, and find the points where the change in gray value is greater than the set threshold 40 as the intersection points between the welding wire and the position without bright arc light.

[0077] Scan and find all intersection points between the welding wire and the areas with and without bright arc light within the reduced image area searched, and then compare the Y - coordinate values of all intersection points. Select the intersection point with the largest Y - coordinate value as the feature point at the bottom of the welding wire, and its coordinates are (Xq, Yq), where Yq is the point with the largest Y - coordinate value, and Xq is the corresponding X - coordinate value of the point with the largest Y - coordinate value.

[0078] Step 4: Calculate the distance from the feature point at the bottom end of the welding wire to the straight line at the bottom of the nozzle

[0079] Step 5: Calibrate the distance |QE| according to the pre-stored calibration values to obtain the dry elongation T of the welding wire: T = |QE| * Scale * cosθ.

[0080] Among them, Scale is the calibration value of the actual length in millimeters corresponding to each pixel of the molten pool monitoring camera, and θ is the calibration value of the installation angle of the molten pool monitoring camera. The calibration values Scale and θ are obtained through pre-calibration steps and pre-stored for use when calculating the dry elongation T of the welding wire.

[0081] Specifically, in this embodiment, combined with Figure 2 , the calibration values Scale and θ can be obtained through the following calibration steps:

[0082] Step 51: According to the actually measured diameter R of the welding wire and the width value K visually measured by the molten pool monitoring camera, calculate the calibration value Scale of the actual length in millimeters corresponding to each pixel of the molten pool monitoring camera: Scale = R / K, unit: millimeter per pixel (mm / pixel).

[0083] Step 52: According to the actually measured length L of the welding wire extending from the nozzle and the length L1 of the welding wire visually measured by the molten pool monitoring camera, use the formula cosθ = L1 / L to calculate the calibration value θ of the installation angle of the molten pool monitoring camera.

[0084] Step 53: Calculate the calibration values Scale and θ and pre-store these calibration values.

[0085] After the molten pool monitoring camera is installed and fixed at a position, it is calibrated once. If the subsequent molten pool monitoring camera does not change the measurement angle position, there is no need to calibrate again.

[0086] Embodiment 2

[0087] Suppose the acquisition of the molten pool monitoring camera is 60fps, and the interval time between two frames is 1000 / 60 = 16.67ms. The visual high-speed algorithm can process the result within 16.67ms to ensure that the result of real-time measurement is consistent with the real-time state of the molten pool monitored by the monitoring camera.

[0088] To verify the data accuracy of this method, data at ten positions are extracted for comparison during the working process. The data is shown in the following table.

[0089]

[0090] The visual value is the wire dry extension T calculated by this method, and the theoretical value is the actual dry extension measured manually. It can be seen that the precision error between the wire dry extension calculated by this method and the actual dry extension measured manually is 0.25 mm.

[0091] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and deformations can be made, and these improvements and deformations should also be regarded as the protection scope of the present invention.

Claims

1. A method for real-time measurement of the dry extension of a welding wire by a molten pool monitoring camera, characterized in that, It includes the following steps: Step 1: The molten pool monitoring camera takes real-time photos to collect molten pool images; Step 2: Read the images for preprocessing. According to the different materials of the welding wire, one or more algorithm modules are used to preprocess the images in the region of interest to maximize the difference in image gray values between the welding wire and the background; When the material of the welding wire is carbon steel, the Gaussian filtering module, dilation morphology module, and image equalization algorithm module are used for preprocessing; When the material of the welding wire is stainless steel, the Soble algorithm module and image equalization algorithm module are used for preprocessing; When the material of the welding wire is copper, the mean filtering module, Soble algorithm module, and image equalization algorithm module are used for preprocessing; Step 3: Analyze and determine the straight line at the bottom of the nozzle and the characteristic points at the bottom end of the welding wire; Step 4: Calculate the distance |QE| from the characteristic points at the bottom end of the welding wire to the straight line at the bottom of the nozzle; Step 5: According to the pre-stored calibration value, calibrate the distance |QE| to obtain the actual dry elongation T of the welding wire = |QE| * Scale * cosθ; Among them, Scale is the calibration value of the actual length in millimeters corresponding to each pixel of the molten pool monitoring camera, and θ is the calibration value of the installation angle of the molten pool monitoring camera; In Step 3, the specific steps for determining the characteristic points at the bottom end of the welding wire are as follows: Step 31. Narrow the ROI image area: For the region of interest ROI in Step 2, calculate the parameters of the narrowed image area: , , W P, H Q; where M and N are the offsets of the starting point coordinates in the X and Y directions in the image coordinate system respectively, and P and Q are the reduction values of the width and height of the rectangular area respectively, and the unit of all is pixel. Step 32: Image gray value: Calculate the average gray value of the reduced image area and compare it with the set first threshold. If it is greater than the set first threshold, when entering the next step of scanning for points, execute according to the process with bright arc light; if it is less than or equal to the set first threshold, when entering the next step of scanning for points, execute according to the process without bright arc light; Step 33: Scan and find the intersection points: Scan and search for all intersection points between the welding wire and the regions with and without bright arc light in the reduced image area; Compare the Y values of all intersection points, and select the point with the largest Y value as the characteristic point at the bottom end of the welding wire, and its coordinates are (Xq, Yq), where Yq is the point with the largest Y value, and Xq is the abscissa value corresponding to the point with the largest Y value.

2. The method for real-time measuring the dry extension of a welding wire by a molten pool monitoring camera according to claim 1, characterized in that, The preprocessing step of Step 2 is as follows: In the read image, draw a region of interest ROI that includes a part of the very front end of the welding wire. The parameters of the region of interest ROI are: , , W, H; where is the starting point X-direction coordinate in the image coordinate system; is the starting point Y-direction coordinate in the image coordinate system; W is the width value of the rectangular area; H is the height value of the rectangular area; the unit of each parameter is pixel.

3. The method for real-time measuring the dry elongation of the welding wire by the molten pool monitoring camera according to claim 2, characterized in that, In Step 2, the image equalization algorithm module is used to preprocess the read images.

4. The method for real-time measuring the dry extension of a welding wire by a molten pool monitoring camera according to claim 1, characterized in that, In Step 3, the specific steps for determining the straight line at the bottom of the nozzle are as follows: Find two points P1(X1, Y1) and P2(X2, Y2) at the bottom of the nozzle, and fit them into the straight line Ax + By + C = 0 at the bottom of the nozzle, where A = Y2 - Y1, B = X1 - X2, C = X2 * Y1 - X1 * Y2, A, B, and C are the parameters of the fitted straight line equation, and A and B are not all zero, and x and y are the abscissa and ordinate of any point on the straight line.

5. The method for real-time measurement of the wire dry elongation by the molten pool monitoring camera according to claim 1, characterized in that, In Step 33, the steps for scanning and searching for all intersection points between the welding wire and the regions with bright arc light in the reduced image area are as follows: Set the second threshold for the change in gray value, calculate the change in the gray value between any position with bright arc light and the gray value of the welding wire, and compare the change between the two with the set second threshold. If the change is greater than the set second threshold, then the point at this position is the intersection point between the welding wire and the region with bright arc light; repeat the above process to scan and find all intersection points between the welding wire and the regions with bright arc light in the reduced image area.

6. The method for real-time measuring the dry elongation of a welding wire by the molten pool monitoring camera according to claim 1, characterized in that In step 33, the steps of scanning and searching for all the intersection points between the welding wire and the area without bright arc light in the reduced image area are as follows: Set the threshold three for the change in gray value, calculate the change in gray value between the gray value at any position without bright arc light and the gray value of the welding wire, compare the change between the two with the set threshold three. If the change is greater than the set threshold three, the point at this position is the intersection point between the welding wire and the area without bright arc light; repeat the above process to traverse and scan all the intersection points between the welding wire and the area without bright arc light in the reduced image area.

7. The method for real-time measuring the dry elongation of a welding wire by a molten pool monitoring camera according to claim 1, characterized in that, In step 5, the calibration value Scale of the actual length in millimeters corresponding to each pixel of the molten pool monitoring camera is obtained through the following calibration steps: Based on the actually measured diameter R of the welding wire and the width value K visually measured by the molten pool monitoring camera, calculate the calibration value Scale of the actual length in millimeters corresponding to each pixel of the molten pool monitoring camera as Scale = R / K.

8. The method for real-time measurement of the dry elongation of the welding wire by the molten pool monitoring camera according to claim 1, characterized in that, In step 5, the calibration value θ of the installation angle of the molten pool monitoring camera is obtained through the following calibration steps: Based on the actually measured length L of the welding wire extending from the nozzle and the length L1 of the welding wire visually measured by the molten pool monitoring camera, use the formula cosθ = L1 / L to calculate the calibration value θ of the installation angle of the molten pool monitoring camera.

Images