Pipe material feeding direction identification method, pipe material processing method and laser processing machine tool

Abstract

The present application relates to a pipe feeding direction recognition method, a pipe processing method and a laser processing machine tool. The pipe feeding direction recognition method comprises: a) collecting a profile point cloud of a pipe after feeding by a visual sensing device; b) performing registration based on the profile point cloud and a drawing point cloud to obtain a deviation angle of the profile point cloud relative to the drawing point cloud; c) determining whether the deviation angle meets a threshold deviation requirement; d) if the threshold deviation requirement is met, determining the deviation angle as an actual deviation angle of the pipe after feeding relative to the drawing in a direction; and e) if the threshold deviation requirement is not met, performing mirror processing on the profile point cloud or the drawing point cloud, and again performing steps b) and c), and determining the deviation angle as an actual deviation angle of the pipe after reverse installation relative to the drawing in a direction when the deviation angle meets the threshold deviation requirement.

Description

Technical Field

[0001] This invention relates to the field of pipe processing, and more specifically to a method for identifying the feeding direction of pipes, a pipe processing method, and a laser processing machine tool. Background Technology

[0002] In the field of pipe processing, such as laser cutting and laser welding, pipes are typically fed into the chuck of a machine tool by an automatic feeding mechanism for clamping and fixing. If the orientation of the clamped pipe after feeding is inconsistent with its corresponding drawing, that is, if the orientation of the pipe after feeding deviates from the orientation on its theoretical drawing, it will directly affect the processing accuracy. In fact, errors in the position and orientation of the holes, grooves, or feature structures may ultimately lead to the scrapping or rework of the pipe.

[0003] Therefore, after the pipe is loaded, it is necessary to determine the deviation between the actual position of the pipe and the theoretical position on the drawing, and to perform position correction accordingly. This is also known in the industry as "centering". However, for non-standard pipes with unconventional cross-sections, due to the uncertainty of their cross-sections, it is very troublesome and time-consuming to manually judge whether the direction of the pipe clamping matches the drawing and whether there is a situation where the pipe is loaded backwards after loading.

[0004] Existing technologies sometimes employ monocular vision-based solutions to replace manual judgment. This approach uses a 2D industrial camera to capture cross-sectional images of the pipe at its front end, extracting cross-sectional contour features and matching them with blueprints. However, this solution has a limited applicability; while maintaining resolution, the field of view is limited, making it unable to identify large pipes, thus struggling to balance field of view and accuracy. Furthermore, shooting in front of the pipe poses risks to the camera; the camera installation must avoid the pipe, and sparks during cutting can damage the camera, requiring structural protection or contamination avoidance, increasing equipment complexity and maintenance costs. In addition, monocular vision-based solutions are also affected by the pipe's thickness and other interference within the field of view during image capture.

[0005] Existing technologies also employ another capacitance-based approach, where the profile of the pipe cross-section is detected by capacitive tracking and then matched with a drawing. This approach is not very accurate, especially when detecting small pipes, where the capacitance is insufficient to identify features on the small pipe.

[0006] Therefore, there is an urgent need for a new technology that can safely, conveniently and accurately identify the orientation of various pipes after loading. Summary of the Invention

[0007] The present invention aims to overcome the above-mentioned and / or other problems in the prior art. Through the present invention, the deviation angle between the feeding direction and the drawing direction of various feeding pipes—especially special-shaped pipes—can be easily and accurately identified, and adjustments can be made according to the identification results to match the direction of the feeding pipe with the drawing direction, thereby significantly improving the processing accuracy and efficiency of the pipes.

[0008] According to a first aspect of the present invention, a method for identifying the feeding direction of a pipe is provided, which may include the following steps: a) acquiring a contour point cloud of the pipe after feeding using a visual sensing device; b) performing registration based on the contour point cloud and the drawing point cloud corresponding to the pipe to obtain a point cloud deviation angle of the contour point cloud relative to the drawing point cloud; c) determining whether the point cloud deviation angle meets a threshold deviation requirement; d) if the point cloud deviation angle meets the threshold deviation requirement, determining the point cloud deviation angle as the actual deviation angle of the pipe after feeding relative to its corresponding drawing in the direction; and e) if the point cloud deviation angle does not meet the threshold deviation requirement, performing mirroring processing on the contour point cloud or the drawing point cloud, and performing steps b) and c) again, and when the point cloud deviation angle meets the threshold deviation requirement, determining the point cloud deviation angle as the actual deviation angle of the pipe after feeding relative to its corresponding drawing in the direction after reverse installation.

[0009] The pipe feeding direction identification method of the present invention acquires the contour point cloud of the pipe using a visual sensing device and registers it with the corresponding drawing point cloud. This automatically obtains the actual deviation angle of the pipe feeding direction relative to the theoretical drawing direction without manual intervention, significantly improving identification efficiency and accuracy. Furthermore, if the initial registration result does not meet the threshold deviation requirement, the method performs mirroring and re-registering, effectively distinguishing between forward and reverse installation states of the pipe. This is particularly suitable for irregularly shaped pipes with asymmetrical features, allowing for timely detection of reversed pipe installation and subsequent correction.

[0010] Optionally, step b) above may include: b1) setting N initial rotation angles, where N is an integer ≥ 4; b2) rotating the acquired contour point cloud or the drawing point cloud by each of the N initial rotation angles, and performing registration based on the rotated contour point cloud and the drawing point cloud after each rotation to obtain the rotated point cloud deviation angle of the rotated contour point cloud relative to the drawing point cloud, or performing registration based on the rotated drawing point cloud and the acquired contour point cloud after each rotation to obtain the rotated point cloud deviation angle of the acquired contour point cloud relative to the rotated drawing point cloud; and b3) selecting the one with the smallest error from all the obtained N rotated point cloud deviation angles as the optimal rotated point cloud deviation angle, and superimposing the corresponding initial rotation angle on the optimal rotated point cloud deviation angle as the point cloud deviation angle output.

[0011] After the pipe is loaded, there may be a situation where the initial angle deviation from the drawing is too large. This can cause the centering and registration algorithm to have difficulty converging or produce incorrect results. The sub-steps b1) to b3) above cleverly rotate the collected pipe contour point cloud or drawing point cloud by multiple initial angles before registration. Then, the one with the smallest error is selected from these registration results and its corresponding initial rotation angle is superimposed as the final deviation angle. This can avoid the situation where the initial deviation angle between the loaded pipe (especially irregular pipes) and the drawing is too large, causing the registration result to only meet the local optimum rather than the global optimum, and further improve the accuracy of recognition.

[0012] In addition, step a) above can be acquired using either single-frame acquisition or scanning acquisition, which can be flexibly adapted to different scenarios.

[0013] When single-frame acquisition is used, step a) may include: a11) rotating the pipe after loading M times, each time rotating 360 / M degrees, and pausing after each rotation and acquiring a point cloud of the pipe once through the visual sensing device, where M is an integer ≥3; and a12) stitching together the point cloud frames acquired after the M rotations to obtain the outline point cloud of the pipe.

[0014] When scanning acquisition is used, step a) may include: a21) rotating the loaded pipe continuously for one revolution, and continuously acquiring point clouds of the pipe through the vision sensing device during the rotation; and a22) stitching the continuously acquired point clouds together to form the outline point cloud of the pipe.

[0015] In addition, step e) above may also include: when the point cloud deviation angle does not meet the threshold deviation requirement, indicating that the pipe material does not match its corresponding drawing.

[0016] According to a second aspect of the present invention, a pipe processing method is provided, which may include: obtaining an actual deviation angle of the pipe after loading relative to its corresponding drawing by means of the method of the present invention for identifying the loading direction of the pipe; correcting the direction of the pipe according to the obtained actual deviation angle, reloading the pipe in the reverse direction, or mirroring the drawing corresponding to the pipe; and processing the pipe with the corrected direction or the pipe with the corrected direction after reloading in the reverse direction based on the drawing, or processing the pipe with the corrected direction based on the mirrored drawing.

[0017] If the feeding pipe is installed backwards, then the actual deviation angle obtained above can be understood as the point cloud deviation angle obtained in step e plus 180 degrees.

[0018] If the feeding pipe is not installed backwards, it can be corrected to align with the drawing direction based on the obtained actual deviation angle. If the feeding pipe is installed backwards, it can be re-fed in reverse and corrected based on the obtained actual deviation angle; alternatively, the drawing can be mirrored, and the pipe corrected to align with the mirrored drawing direction based on the obtained actual deviation angle. The pipe processing method of this invention can process pipes whose orientation has been corrected, which significantly improves processing accuracy.

[0019] According to a third aspect of the present invention, a laser processing machine tool is provided, comprising a laser processing head, a chuck, and a control unit. A vision sensing device may be mounted on the laser processing head for acquiring a contour point cloud of a loaded tube. The chuck is used to clamp the loaded tube. The control unit is used to control the laser processing head, the vision sensing device, and the chuck, and can be configured to perform a method for identifying the loading direction of the tube or a tube processing method according to the present invention.

[0020] The aforementioned visual sensing device may be, for example, a line laser sensor, a laser rangefinder, or a 3D camera.

[0021] According to a fourth aspect of the invention, a computer-readable storage medium is provided having encoded instructions recorded thereon, which, when executed, implement the method for identifying the feeding direction of a pipe or the pipe processing method according to the invention.

[0022] According to a fifth aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed, implements the method for identifying the feeding direction of a pipe or a pipe processing method according to the present invention.

[0023] Other features and aspects of the invention will become clearer from the following detailed description taken in conjunction with the accompanying drawings. Attached Figure Description

[0024] The invention can be better understood by describing exemplary embodiments of the invention in conjunction with the accompanying drawings, in which: Figure 1 A flowchart of a method for identifying the feeding direction of pipes according to an embodiment of the present invention is shown; Figure 2 A flowchart illustrating an example of a method for identifying the feeding direction of pipes according to an embodiment of the present invention is shown; Figure 3 A flowchart illustrating another example of a method for identifying the feeding direction of pipes according to an embodiment of the present invention is shown; Figure 4 A flowchart illustrating another example of a method for identifying the feeding direction of pipes according to an embodiment of the present invention is shown; Figure 5 A flowchart of a pipe processing method according to an embodiment of the present invention is shown; and Figure 6 A schematic block diagram of a laser processing machine tool according to an embodiment of the present invention is shown. Detailed Implementation

[0025] The present invention will be further described below with reference to specific embodiments and accompanying drawings. More details are set forth in the following description in order to provide a full understanding of the present invention. However, the present invention can obviously be implemented in many other ways different from those described herein. Those skilled in the art can make similar extensions and derivations based on actual application situations without departing from the spirit of the present invention. Therefore, the scope of protection of the present invention should not be limited by the content of this specific embodiment.

[0026] Unless otherwise defined, the technical or scientific terms used in the claims and description shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in the description and claims of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. The terms "an" or "a" and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms "comprising" or "including" and similar terms mean that the element or object preceding "comprising" or "including" encompasses the element or object listed following "comprising" or "including" and its equivalents, and do not exclude other elements or objects. The terms "connected" or "linked" and similar terms are not limited to physical or mechanical connections, nor are they limited to direct or indirect connections.

[0027] According to an embodiment of the present invention, a method for identifying the feeding direction of pipes is provided.

[0028] Figure 1A flowchart of a method 100 for identifying the feeding direction of pipes according to an embodiment of the present invention is shown. Figure 1 As shown, the method 100 may include steps 110 to 150.

[0029] In step 110, the outline point cloud of the pipe after loading can be collected by a visual sensing device.

[0030] The aforementioned pipes can be of any type, including not only standard types (such as square pipes, rectangular pipes, round pipes, I-beams, channel steel, angle steel, etc.), but also special-shaped pipes. In pipe processing, after a pipe is cut, a new pipe can be moved from the preparation area to the processing area by a feeding device, and clamped by a main clamp to facilitate subsequent processing. This process is called "feeding".

[0031] The aforementioned visual sensing device can be mounted on the processing head or at a location other than the processing head. This visual sensing device can be a line laser sensor, a laser ranging device (e.g., a point laser ranging device), or a 3D camera (e.g., a structured light camera).

[0032] Next, in step 120, registration can be performed based on the contour point cloud and the drawing point cloud corresponding to the pipe to obtain the point cloud deviation angle of the contour point cloud relative to the drawing point cloud.

[0033] The drawing corresponding to the pipe can be imported into, for example, CNC software, and the point cloud of the drawing can be parsed. By registering the contour point cloud and the point cloud of the drawing, the translation matrix and rotation matrix between the two point clouds can be obtained, that is, the translation amount and deviation angle (i.e., the point cloud deviation angle) between the two point clouds can be obtained.

[0034] Next, in step 130, it can be further determined whether the point cloud deviation angle meets the threshold deviation requirement. This threshold deviation requirement refers to an angle deviation threshold preset according to the specific processing scenario.

[0035] If the point cloud deviation angle is determined to meet the threshold deviation requirement, then method 100 proceeds to step 140: the point cloud deviation angle is determined as the actual deviation angle of the pipe after loading relative to its corresponding drawing in the direction.

[0036] If the point cloud deviation angle does not meet the threshold deviation requirement, then mirroring is performed on the contour point cloud or the drawing point cloud, and steps 120 and 130 are executed again. Then, method 100 proceeds to step 150: when the point cloud deviation angle meets the threshold deviation requirement, the point cloud deviation angle is determined as the actual deviation angle in the direction of the pipe after it has been installed in reverse relative to its corresponding drawing. That is, the point cloud deviation angle is the deviation angle in the direction of the drawing after the pipe has been rotated 180 degrees.

[0037] In pipe processing, the orientation of the pipe after loading may not necessarily match the orientation in the corresponding theoretical drawing, which can significantly affect the accuracy of subsequent processing. Especially for non-standard, irregularly shaped pipes, the uncertainty of the pipe's cross-section can lead to the scrapping of the entire pipe due to the deviation between the actual orientation after loading and the orientation in the theoretical drawing. The pipe loading orientation identification method of this invention, by registering the point cloud of the loaded pipe with the point cloud of the drawing, can automatically and efficiently identify the deviation angle between the actual orientation of the pipe after loading and the orientation in the theoretical drawing without manual intervention, thus providing a favorable guarantee for subsequent precise processing. Moreover, the pipe loading orientation identification method of this invention is applicable to various processing scenarios, including large and small pipes, is not affected by the thickness of the pipe itself, and eliminates the risk of equipment damage from collisions with the pipe or sparks during cutting. It does not require consideration of avoiding the pipe or additional protective devices. Compared with existing identification methods, it greatly improves the breadth and safety of application scenarios and saves costs.

[0038] Furthermore, reversed pipe loading frequently occurs on the processing site, meaning the pipes are installed backwards (front to back) during the loading process. This is especially problematic for irregularly shaped pipes, whose front and back shapes are often asymmetrical; if installed backwards, it will directly cause problems in subsequent processing. The pipe loading direction identification method of this invention fully considers the possibility of reversed pipe loading. If the initial point cloud deviation angle does not meet the threshold deviation requirement, it does not blindly prompt errors such as discrepancies between the actual pipe size and the drawing size. Instead, it automatically mirrors the pipe point cloud or the drawing point cloud after loading and then re-registers them, verifying whether the deviation angle obtained from the registration meets the threshold deviation requirement. This significantly improves the ability to handle anomalies in pipe processing, avoiding processing scrap due to reversed pipe loading. Furthermore, in the case of timely detection of reversed pipe loading, it can identify the deviation angle between the pipe and the drawing direction after correcting the pipe's orientation, improving efficiency and providing further assurance for accurate subsequent processing.

[0039] Optionally, step 150 may further include: when the point cloud deviation angle does not meet the threshold deviation requirement, indicating that the pipe material does not match its corresponding drawing.

[0040] If, after mirroring the point cloud of the pipe or the point cloud of the drawing after loading, considering the possibility of the pipe being installed backwards, the point cloud deviation angle that meets the threshold deviation requirement cannot be obtained through the above registration operation, it indicates that the pipe may not match its corresponding drawing. In this case, the operator can be prompted that the pipe does not match the drawing by displaying a warning message on the screen, issuing an audible and visual alarm signal, or sending an abnormal notification to the remote management system. Manual inspection and handling are required.

[0041] It is important to note that after the pipe is fed, the deviation angle from its drawing orientation can sometimes be too large, especially when using automatic feeding. This often makes it impossible to guarantee the initial angle at which the pipe is clamped by the chuck, which can severely affect the centering of irregularly shaped pipes. Because of the irregular cross-section of the irregularly shaped pipe, the algorithm used for point cloud registration, as described in step 120 above, will target local features. If the deviation angle between the fed pipe and its drawing orientation is too large, the registration result will only satisfy local optima rather than global optima, ultimately making it difficult for the registration algorithm to converge or producing erroneous results.

[0042] Therefore, in order to further improve the accuracy of the pipe feeding direction identification of the present invention, the above step 120 may optionally include sub-steps 1220 to 1260, such as... Figure 2 As shown in the image.

[0043] In substep 1220, N initial rotation angles can be set, where N is an integer ≥ 4.

[0044] The N initial rotation angles can be uniformly distributed within the range of 0° to 360°. For example, when N=4, the initial rotation angles can be set to 0°, 90°, 180°, and 270°; when N=8, the initial rotation angles can be set to 0°, 45°, 90°, 135°, 180°, 225°, 270°, and 315°; when N=12, the initial rotation angles can be set to 0°, 30°, 60°, 90°, and so on. It can be understood that the N initial rotation angles can also be non-uniformly distributed within the range of 0° to 360°, but a uniform distribution makes it easier to obtain high-precision point cloud deviation angles. It can also be understood that the larger the value of N, the denser the angle coverage and the higher the robustness of registration, but the computational load will also increase accordingly. Therefore, in practical applications, an appropriate value of N can be selected based on the symmetry characteristics of the pipe cross-section and the availability of computational resources. For example, for pipes with relatively symmetrical cross-sections, N can be a smaller value (such as 4); while for pipes with strong asymmetry in cross-sections, N can be a larger value (such as 8).

[0045] Next, in sub-step 1240, the acquired contour point cloud or the drawing point cloud can be rotated by each of the N initial rotation angles, and after each rotation, registration is performed based on the rotated contour point cloud and the drawing point cloud to obtain the rotational point cloud deviation angle of the rotated contour point cloud relative to the drawing point cloud, or after each rotation, registration is performed based on the rotated drawing point cloud and the acquired contour point cloud to obtain the rotational point cloud deviation angle of the acquired contour point cloud relative to the rotated drawing point cloud.

[0046] That is, the rotation angles set in sub-step 1220 can be applied to the acquired contour point cloud or drawing point cloud respectively. For example, taking the application of each rotation angle to the contour point cloud as an example, and N is 8, these 8 initial rotation angles are 0°, 45°, 90°, 135°, 180°, 225°, 270° and 315° respectively. The acquired contour point cloud can be rotated by the 8 angles respectively, and after each rotation, registration is performed on the contour point cloud 1, contour point cloud 2, contour point cloud 3, contour point cloud 4, contour point cloud 5, contour point cloud 6, contour point cloud 7 and contour point cloud 8 (corresponding to the contour point clouds after rotation of 0°, 45°, 90°, 135°, 180°, 225°, 270° and 315° respectively) and the drawing point cloud to obtain each contour point cloud. 1~8 The rotational deviation angles are 1, 2, 3, 4, 5, 6, 7, and 8 relative to the point cloud in the drawing. It's understandable that if each rotation angle were applied separately to the point cloud in the drawing, the situation would be similar, and will not be elaborated further here.

[0047] Next, in sub-step 1260, the one with the smallest error among all N obtained rotation point cloud deviation angles can be selected as the optimal rotation point cloud deviation angle, and the corresponding initial rotation angle is superimposed on the optimal rotation point cloud deviation angle to output the point cloud deviation angle.

[0048] Taking the application of the aforementioned eight rotation angles to the contour point cloud as an example, by registering contour point cloud 1, contour point cloud 2, contour point cloud 3, contour point cloud 4, contour point cloud 5, contour point cloud 6, contour point cloud 7, and contour point cloud 8 with the drawing point cloud, eight sets of corresponding translation and rotation matrices can be obtained. Based on these eight sets of corresponding translation and rotation matrices, the unrotated original acquired contour point cloud can be converted into drawing point cloud, resulting in eight sets of drawing point clouds (drawing point cloud 1, drawing point cloud 2, drawing point cloud 3, drawing point cloud 4, drawing point cloud 5, drawing point cloud 6, drawing point cloud 7, and drawing point cloud 8). The errors between these eight sets of drawing point clouds and the original theoretical drawing point cloud can be calculated, and the set with the smallest error, for example, drawing point cloud 6, can be selected. Corresponding to drawing point cloud 6 is contour point cloud 6 and the rotation deviation angle 6, and contour point cloud 6 is obtained by rotating the original acquired contour point cloud by 225°. The point cloud deviation angle 6 after rotation is the optimal point cloud deviation angle after rotation. Superimposing it by 225° yields the point cloud deviation angle of the original acquired contour point cloud relative to the drawing point cloud. It is understandable that the situation would be similar if each rotation angle were applied separately to the drawing point cloud, which will not be elaborated upon here.

[0049] By employing the aforementioned multi-initial-rotation-angle registration strategy, the problem of registration algorithms getting trapped in local optima can be effectively avoided, significantly improving the accuracy and robustness of registration. Especially for irregularly shaped pipes with strong asymmetric characteristics, regardless of the deviation angle between the pipe after loading and its drawing direction (even if the deviation angle is large), the point cloud deviation angle of the contour point cloud relative to the drawing point cloud can be accurately obtained.

[0050] Optionally, step 110 above can be implemented using single-frame acquisition. In this case, step 110 may include sub-steps 1112 and 1114, such as... Figure 3 As shown in the image.

[0051] In sub-step 1112, the pipe after loading can be rotated M times, each rotation being 360 / M degrees, and after each rotation, it can pause and collect a point cloud of the pipe once through the visual sensing device, where M is an integer ≥3.

[0052] For example, when M=3, the pipe rotates 120° each time, pausing at 0°, 120°, and 240° to collect point clouds; when M=4, the pipe rotates 90° each time, pausing at 0°, 90°, 180°, and 270° to collect point clouds. M can be set to a larger integer depending on the specific situation.

[0053] In sub-step 1114, the point cloud frames acquired after the M rotations can be stitched together to form the outline point cloud of the pipe.

[0054] Taking M=4 as an example, the pipe is rotated 90° four times. After each rotation, the pipe is paused and the visual sensing device collects a point cloud image of the pipe. After the four rotations, the pipe has completed one full rotation. At this point, the four collected point cloud images are stitched together to obtain the complete outline point cloud of the pipe.

[0055] Optionally, step 110 above can also be implemented by scanning and acquisition. In this case, step 110 may include sub-steps 1122 and 1124, such as... Figure 4 As shown in the image.

[0056] In sub-step 1122, the pipe after loading can be rotated continuously for one revolution, and point cloud data of the pipe can be continuously collected by the vision sensing device during the rotation.

[0057] Continuous rotation means that there is no pause in the middle; that is, the pipe does not stop during the process of rotating once.

[0058] In sub-step 1124, the continuously collected point clouds can be stitched together to form the outline point cloud of the pipe.

[0059] Taking a line laser sensor (which can be mounted on a processing head, for example) as the visual sensing device, the line laser sensor continuously performs line laser scanning on the pipe's contour during one continuous rotation of the pipe. It should be noted that while the line laser sensor can actually obtain all point cloud data of a single line on the pipe's contour at a given moment, only the data of a single point on that line at a given moment needs to be collected here. Therefore, the line laser sensor is essentially performing a point laser function. As the pipe continues to rotate, the line laser sensor continuously collects data from each corresponding point at every moment. Finally, these continuously collected point clouds are stitched together to obtain the contour point cloud of the pipe's cross-section.

[0060] This invention allows for flexible switching between the two contour point cloud acquisition methods based on specific needs and scenarios. For example, the single-frame contour point cloud acquisition method is suitable for scenarios where the pipe rotation does not exceed the field of view of the visual sensing device (if it does, the visual sensing device needs to be adjusted to cover the pipe again). In scenarios where the pipe rotation exceeds the field of view of the visual sensing device, the scanning acquisition method can be used to obtain the complete contour of the pipe through rotational scanning. As another example, the scanning acquisition method is suitable for scenarios where the pipe surface has no obvious depressions (if there are obvious depressions, the depressed parts may not be scanned). In scenarios where the pipe surface has obvious depressions, the single-frame contour point cloud acquisition method can be used to obtain the pipe contour through multi-angle single-frame stitching.

[0061] Furthermore, the above description of the scanning and acquisition of contour point clouds uses a line laser sensor as an example of a visual sensing device. However, as mentioned earlier, the visual sensing device can also be a laser rangefinder or a 3D camera. If the visual sensing device uses a 3D camera, such as a structured light camera, the pipe cross-section point cloud can be acquired in one go by installing the structured light camera in a suitable position, without the need for line laser rotation scanning. Similarly, if the visual sensing device uses a laser rangefinder (which can be mounted on a processing head, for example, a point laser rangefinder), the pipe cross-section point cloud can be acquired without line laser rotation scanning in processing scenarios where the processing head does not need to rotate (only the pipe needs to be rotated).

[0062] According to an embodiment of the present invention, a method for processing pipes is also provided.

[0063] Figure 5 A flowchart of a pipe processing method 500 according to an embodiment of the present invention is shown. Figure 5 As shown, the method 500 may include steps 510 to 530.

[0064] In step 510, the pipe feeding direction identification method of the present invention (e.g.) can be used. Figure 1 The embodiment method 100 shown obtains the actual directional deviation angle of the pipe after feeding relative to its corresponding drawing.

[0065] The “actual deviation angle” here may be the actual deviation angle determined in step 140 of method 100, or it may be the actual deviation angle determined in step 150 of method 100 (which can be understood as the point cloud deviation angle (deviation angle θ) obtained by mirroring the point cloud of the pipe or the point cloud of the drawing after loading and then executing step 120 again after considering the possibility that the pipe may be installed backwards, and then superimposing 180 degrees).

[0066] Next, in step 520, the direction of the pipe can be corrected according to the obtained actual deviation angle, the pipe can be re-loaded in the reverse direction, or the drawing corresponding to the pipe can be mirrored.

[0067] If the actual deviation angle is determined in step 140 above, it means that the pipe is not installed backwards. Therefore, the direction of the pipe can be corrected according to the obtained actual deviation angle.

[0068] If the actual deviation angle is determined in step 150 above, it indicates that the pipe is installed backwards. Therefore, it is necessary to re-load the pipe in reverse order to correct the reverse installation. Then, steps 110 and 120 of method 100 can be executed again according to the actual situation to obtain the deviation angle β of the re-loaded pipe relative to the drawing direction. Step 130 is then used to verify whether the deviation angle β meets the requirements. In fact, since the re-loaded pipe is no longer installed backwards, and the dimensions of the pipe are confirmed to match the drawing after the previous operation of method 100, the deviation angle β of the re-loaded pipe relative to the drawing direction obtained in step 120 should also meet the requirements. Step 130 is not a necessary step. Therefore, the direction of the re-loaded pipe can be directly corrected based on the deviation angle β obtained from steps 110 and 120.

[0069] Alternatively, if the actual deviation angle was determined in step 150 above, although this indicates that the pipe is installed backwards, it is not necessary to reinstall the pipe; instead, the drawing corresponding to the pipe can be mirrored. In this case, the orientation of the pipe only needs to be corrected based on the deviation angle θ obtained in step 150.

[0070] Returning to method 500, in step 530, the pipe with the corrected orientation or the pipe with the corrected orientation after reverse reloading can be processed based on the drawing, or the pipe with the corrected orientation can be processed based on the mirrored drawing.

[0071] If the actual deviation angle is determined in step 140 above, then in step 530 the pipe whose direction was corrected in step 520 can be processed based on the drawing.

[0072] If the actual deviation angle is determined in step 150 above, then in step 530 the pipe whose orientation has been corrected (based on deviation angle β) after being reversed and reloaded in step 520 can be processed based on the drawing, or the pipe whose orientation has been corrected (based on deviation angle θ) in step 520 can be processed based on the mirrored drawing.

[0073] The above-mentioned pipe processing method of the present invention utilizes the pipe feeding direction identification method of the present invention to accurately identify the deviation of the pipe after feeding from the theoretical drawing in direction, and effectively correct the deviation, thereby ensuring that the pipe after feeding is processed according to the desired processing trajectory, which greatly improves the processing accuracy.

[0074] According to embodiments of the present invention, a computer-readable storage medium is also provided, on which encoded instructions are recorded, which, when executed, implement the method of the present invention for identifying the feeding direction of pipes or the pipe processing method. The computer-readable storage medium may include hard disk drives, floppy disk drives, optical disc read / write (CD-R / W) drives, digital universal disk (DVD) drives, flash memory drives, and / or solid-state storage devices, etc.

[0075] According to embodiments of the present invention, a computer program product is also provided, comprising a computer program that, when executed, implements the method of the present invention for identifying the feeding direction of pipes or the pipe processing method described above. The computer program product can be implemented using various programming languages, such as C, C++, Java, Python, JavaScript, etc., to adapt to different development environments and platform requirements.

[0076] According to embodiments of the present invention, a laser processing machine tool is also provided accordingly.

[0077] refer to Figure 6 The image shows a laser processing machine tool 600 according to an embodiment of the present invention, which may include a laser processing head 620, a chuck 640 and a control unit 660.

[0078] A vision sensing device 622 can be installed on the laser processing head 620. The vision sensing device 622 can be used to acquire the contour point cloud of the pipe after it has been loaded.

[0079] The chuck 640 can be used to clamp the loaded pipe.

[0080] The control unit 660 can be used to control the laser processing head 620, the vision sensing device 622 and the chuck 640, and can be configured to perform the method for identifying the pipe feeding direction and the pipe processing method according to the above embodiments of the present invention.

[0081] Optionally, the visual sensing device 622 can be a line laser sensor, a laser rangefinder, or a 3D camera.

[0082] The aforementioned laser processing machine tool can realize the method for identifying the feeding direction of pipes and the pipe processing method according to the present invention, as described above. Many of the design concepts and details applicable to the method for identifying the feeding direction of pipes and the pipe processing method of the present invention are also applicable to the aforementioned laser processing machine tool and can achieve the same beneficial technical effects, which will not be repeated here.

[0083] The various aspects of the present invention have been described above through exemplary embodiments. However, it should be understood that various modifications can be made to the above exemplary embodiments without departing from the spirit and scope of the invention. For example, if suitable results can be achieved if the described techniques are performed in a different order and / or if components in the described system, architecture, device, or circuit are combined in different ways and / or replaced or supplemented by other components or their equivalents, then correspondingly, these modified other embodiments also fall within the scope of protection of the claims.

Claims

1. A method for identifying the feeding direction of pipes, comprising the following steps: a) Collect contour point cloud data of the pipe after it has been loaded using a visual sensing device; b) Perform registration based on the outline point cloud and the drawing point cloud corresponding to the pipe to obtain the point cloud deviation angle of the outline point cloud relative to the drawing point cloud. c) Determine whether the point cloud deviation angle meets the threshold deviation requirement; d) If the point cloud deviation angle meets the threshold deviation requirement, then the point cloud deviation angle is determined as the actual directional deviation angle of the pipe after loading relative to its corresponding drawing; and e) If the point cloud deviation angle does not meet the threshold deviation requirement, then mirroring is performed on the contour point cloud or the drawing point cloud, and steps b) and c) are executed again. When the point cloud deviation angle meets the threshold deviation requirement, the point cloud deviation angle is determined as the actual deviation angle in the direction of the pipe after it is installed in reverse after loading, relative to its corresponding drawing.

2. The method as described in claim 1, characterized in that, Step b) includes: b1) Set N initial rotation angles, where N is an integer ≥ 4; b2) Rotate the acquired contour point cloud or the drawing point cloud by each of the N initial rotation angles, and after each rotation, perform registration based on the rotated contour point cloud and the drawing point cloud to obtain the rotational point cloud deviation angle of the rotated contour point cloud relative to the drawing point cloud; or after each rotation, perform registration based on the rotated drawing point cloud and the acquired contour point cloud to obtain the rotational point cloud deviation angle of the acquired contour point cloud relative to the rotated drawing point cloud; and b3) Select the one with the smallest error from all N obtained rotation point cloud deviation angles as the optimal rotation point cloud deviation angle, and output the point cloud deviation angle by superimposing the corresponding initial rotation angle on the optimal rotation point cloud deviation angle.

3. The method as described in claim 1 or 2, characterized in that, Step a) includes: a11) Rotate the loaded pipe M times, each rotation being 360 / M degrees, and pause after each rotation, then collect a point cloud image of the pipe using the visual sensing device, where M is an integer ≥ 3; and a12) The point cloud frames acquired after the M rotations are stitched together to form the outline point cloud of the pipe.

4. The method as described in claim 1 or 2, characterized in that, Step a) includes: a21) The pipe, after being loaded, is rotated continuously for one revolution, and during this rotation, point clouds of the pipe are continuously collected by the visual sensing device; and a22) The continuously collected point clouds are stitched together to form the outline point cloud of the pipe.

5. The method as described in claim 1 or 2, characterized in that, Step e) further includes: When the point cloud deviation angle does not meet the threshold deviation requirement, a message is displayed indicating that the pipe material does not match its corresponding drawing.

6. A method for processing pipes, comprising: The actual directional deviation angle of the pipe after loading relative to its corresponding drawing is obtained by the method described in any one of claims 1-5. Correct the direction of the pipe according to the obtained actual deviation angle, reverse and reload the pipe, or mirror the drawing corresponding to the pipe; as well as The pipe with its orientation corrected is processed based on the drawing or the pipe with its orientation corrected after being re-loaded in the reverse direction, or the pipe with its orientation corrected is processed based on the mirrored drawing.

7. A laser processing machine tool, comprising: A laser processing head, on which a vision sensing device is installed, is used to collect the contour point cloud of the pipe after it has been loaded. A chuck is used to hold the loaded pipe. as well as A control unit is configured to control the laser processing head, the vision sensing device, and the chuck, and to perform the method as described in any one of claims 1-6.

8. The laser processing machine tool as described in claim 7, characterized in that, The visual sensing device is a line laser sensor, a laser rangefinder, or a 3D camera.

9. A computer-readable storage medium having encoded instructions recorded thereon, which, when executed, implement the method as described in any one of claims 1-6.

10. A computer program product comprising a computer program that, when executed, implements the method as described in any one of claims 1-6.

Images