System and method for visually detecting quality of pipe pile

Abstract

The invention discloses a system and method for visually detecting the quality of a pipe pile. The system comprises a bearing module used for bearing the pipe pile, an end detection module and a movable pipe pile body scanning module. The end detection module is used for collecting end face data of the end of the pipe pile and recognizing the defects of the end of the pipe pile according to the end face data of the end of the pipe pile. The movable pipe pile body scanning module moves along a pipe pile and is used for collecting pipe pile body data, and pipe pile body defects are recognized through the pipe pile body data. And summarizing the pipe pile end defects and the pipe pile body defects to generate a defect report. Rapid identification, quantitative analysis and data archiving of pipe pile defects are achieved, and the detection precision and the production efficiency are remarkably improved.

Description

Technical Field

[0001] This application provides a visual inspection system and method for the quality of pipe piles, which relates to the technical field of construction engineering. Background Art

[0002] In the prior art, for the quality inspection of pipe piles, it completely relies on manual inspection and can only conduct spot checks. It can neither ensure the inspection rate nor the inspection accuracy. Therefore, there is an urgent need to introduce an automatic inspection station for the quality of pipe piles in the pipe pile production line. For non-automated production lines, it can replace manual inspection to improve the product inspection rate and inspection accuracy. The core problems of the prior art include: low inspection rate: only spot checks can be carried out, and the quality of each pipe pile cannot be guaranteed; high cost: professional inspectors need to be hired for a long time, and the labor cost remains high; data defect: defective products lack digital records, making it difficult to trace the root cause of quality problems; potential safety hazard: the working environment at the pipe pile production site is harsh (such as high temperature, dust), and manual inspection poses a safety risk. Summary of the Invention

[0003] The technical problem to be solved by this application is the low efficiency of manually inspecting the quality of pipe piles.

[0004] To solve the above technical problem, the technical solution of this application provides a visual inspection system for the quality of pipe piles, including:

[0005] A loading module, an end detection module, and a mobile pipe pile body scanning module for loading pipe piles;

[0006] The end detection module is used to collect the end face data of the pipe pile end and identify the defects at the pipe pile end through the end face data of the pipe pile end; the mobile pipe pile body scanning module moves along the pipe pile to collect the pipe pile body data and identify the defects on the pipe pile body through the pipe pile body data;

[0007] Summarize the defects at the pipe pile end and the defects on the pipe pile body to generate a defect report.

[0008] Preferably, the end detection module includes 3D structured light cameras arranged at both ends of the pipe pile for obtaining three-dimensional information of the pipe pile end face.

[0009] Preferably, the process of identifying the defects at the pipe pile end includes:

[0010] Reconstruct the end three-dimensional model based on the phase measurement profilometry method;

[0011] Compare the end three-dimensional model with the design model to obtain the defects at the pipe pile end.

[0012] Preferably, the defects at the pipe pile end include inclination defects and depression defects;

[0013] Evaluate the quality of the pipe pile end based on the defects at the pipe pile end, and the process includes:

[0014] Classify the end face defects of the pipe pile and evaluate the quality of the end of the pipe pile;

[0015] The inclination defect outputs the actual inclination angle value;

[0016] The depression defect outputs the actual depression depth value;

[0017] If any of the inclination defect and the depression defect fails, the end of the pipe pile is unqualified.

[0018] Preferably, the bearing module includes a bracket for bearing the pipe pile; the bracket is provided with rollers distributed in an array perpendicular to the central axis of the pipe pile, the rollers are in contact with the pipe pile, and the rollers are provided with driving rollers driven by a driving device, and the driving rollers are used to drive the pipe pile to rotate.

[0019] Preferably, the bracket is provided with a slide rail parallel to the central axis of the pipe pile, and the mobile pipe pile body scanning module includes an equipment rack arranged on the slide rail and moving axially along the slide rail, and the equipment rack is provided with a 3D line scanning camera and an image acquisition camera.

[0020] Preferably, the process of identifying the defects of the pipe pile body includes:

[0021] The equipment rack moves axially along the pipe pile. During the movement, the 3D line scanning camera scans the upper surface of the pipe pile to obtain the three-dimensional point cloud data of the upper surface of the pipe pile, and the image acquisition camera takes pictures of the upper surface of the pipe pile to obtain the image data of the upper surface of the pipe pile;

[0022] The driving roller drives the pipe pile to rotate, and turns the lower surface of the pipe pile upwards;

[0023] The equipment rack moves axially along the pipe pile again. During the movement, the 3D line scanning camera scans the lower surface of the pipe pile to obtain the three-dimensional point cloud data of the lower surface of the pipe pile, and the image acquisition camera takes pictures of the lower surface of the pipe pile to obtain the image data of the lower surface of the pipe pile;

[0024] Identify the pile body defects based on the three-dimensional point cloud data and the image data respectively and make a decision fusion.

[0025] Preferably, the process of identifying the pile body defects based on the three-dimensional point cloud data and the image data respectively and making a decision fusion includes:

[0026] Locate the curvature mutation region of the pipe pile body based on the three-dimensional point cloud data;

[0027] Establish a first-branch deep learning neural network for identifying structural defects, input the three-dimensional point cloud data of the curvature mutation region into the first-branch deep learning neural network, and output the structural defects of the pipe pile body;

[0028] A second branch deep learning neural network for identifying surface flaws and defects is established. Image data is input into the deep convolutional neural network, and surface flaws and defects on the pile body of the pipe pile are output.

[0029] Decision fusion is performed to fuse the structural defects of the pile body of the pipe pile and the surface flaws and defects of the pile body of the pipe pile to obtain the defects of the pile body of the pipe pile.

[0030] Preferably, the structural defects of the pile body of the pipe pile include internal broken bars and pile body defects caused by trapped straw ropes; the surface flaws and defects of the pile body of the pipe pile include sticking skin and ash spots.

[0031] The technical solution of the present application also provides a method for visually detecting the quality of pipe piles. The above-mentioned visual detection system for pipe pile quality is used to detect the quality of pipe piles and identify pipe pile defects.

[0032] The visual detection system and method for pipe pile quality provided by the present application solve the problems of low efficiency, high cost and inconsistent standards in manual detection, realize the rapid identification, quantitative analysis and data archiving of pipe pile defects, and significantly improve the detection accuracy and production efficiency. Brief Description of the Drawings

[0033] Figure 1 It is a schematic diagram of the architecture of the visual detection system for pipe pile quality provided by the present application;

[0034] Figure 2 It is a schematic diagram of the bearing module of the present application bearing a pipe pile Figure 1 ;

[0035] Figure 3 It is a schematic diagram of the bearing module of the present application bearing a pipe pile Figure 2 ;

[0036] Figure 4 It is a schematic diagram of the end detection module of the present application for detection;

[0037] Figure 5 It is a process flow chart of the pipe pile end defect detection process;

[0038] Figure 6 It is a schematic diagram of the mobile pipe pile body scanning module for detection;

[0039] Reference numerals: bearing module 100, bracket 101, roller 102, slide rail 103, end detection module 200, 3D structured light camera 201, mobile pipe pile body scanning module 300, equipment rack 301, 3D line scan camera 302. Detailed Description of the Embodiment

[0040] To make the present application more clearly understandable, various exemplary embodiments will be described below. These examples are non-limiting, and it should be understood that they are used to illustrate the broader applications of the devices, systems, and methods. Without departing from the essence and scope of the present application, these embodiments can be variously modified, and equivalents can be substituted. In addition, various modifications can be made to adapt to special circumstances, materials, material components, types of processing, processing actions, or steps to suit the purpose, content, or scope of the present application. All such modifications will be within the scope of protection of the present application.

[0041] Regarding any materials, dimensions, quantities introduced in the overview or detailed description, they are only examples and do not limit the subject matter of the present application. Moreover, the various embodiments of the embodiments described herein will complement each other rather than being purely alternative, unless otherwise stated. In other words, the embodiments from one embodiment can be freely combined with the embodiments from other embodiments, as is easily understood by those of ordinary skill in the art, unless it is stated that these embodiments are only for substitution.

[0042] In the description of the present application, it should be noted that the orientation or positional relationship indicated by terms such as "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of this application is usually placed during use. It is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present application. In addition, terms such as "first", "second", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0043] In the description of the present application, it should also be noted that unless otherwise clearly specified and limited, the terms "set", "connect" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected, or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0044] Embodiment

[0045] The embodiment of the present application provides a visual inspection system for the quality of pipe piles, which evaluates the quality of pipe piles by detecting pipe pile defects and can generate defect reports for technicians to view. Among them, pipe pile defects include pipe pile end defects and pipe pile body defects. Pipe pile end defects mainly refer to end face inclination defects and end face depression defects existing on the end face, etc. Pipe pile body defects include pipe pile body structure defects and pipe pile body surface flaw defects. Pipe pile body structure defects mainly refer to structural defects such as pile body depression and protrusion caused by broken steel bars and trapped straw ropes inside the pipe pile. Pipe pile body surface flaw defects mainly refer to surface flaw defects such as sticking skin and ash spots appearing on the surface of the pipe pile body. The visual inspection system for the quality of pipe piles provided by the present application uses vision and 3D scanning technology to obtain pipe pile end face data and pipe pile body data to automatically identify and output pipe pile defects.

[0046] Specifically, referring to Figure 1 , the visual inspection system for the quality of pipe piles provided by the embodiment of the present application includes:

[0047] A bearing module 100, which is used to bear the pipe pile;

[0048] An end detection module 200, which is used to collect pipe pile end face data and identify pipe pile end defects through the pipe pile end face data;

[0049] A mobile pipe pile body scanning module 300, which moves along the pipe pile to collect pipe pile body data and identify pipe pile body defects through the pipe pile body data;

[0050] After identifying the pipe pile end defects and pipe pile body defects, summarize the pipe pile end defects and pipe pile body defects to generate a defect report.

[0051] The visual inspection system for the quality of pipe piles provided by the embodiment of the present application classifies and summarizes the pipe pile end defects and pipe pile body defects, and archives the data for technicians to view and evaluate the quality of pipe piles.

[0052] In one embodiment, referring to Figure 2 、 Figure 3, the bearing module 100 includes a bracket 101 for bearing pipe piles; the bracket is located at the bottom of the entire bearing module. The bracket 101 defines a semi-cylindrical concave position for placing the pipe piles to be detected. The pipe piles are placed on the bracket by a crane. On the bracket 101, rollers 102 perpendicular to the central axis of the pipe pile are arranged in an array. The rollers 102 are in direct contact with the pipe piles. After using the mobile pipe pile body scanning module to perform the first round of scanning to obtain the body data of the upward part (the first part, the upper surface) of the outer surface of the pipe pile, the pipe pile can be rotated 180° by the rollers to turn the downward part (the second part, the lower surface) of the outer surface of the pipe pile upward, and then the mobile pipe pile body scanning module performs the second round of scanning to obtain the body data of the second part of the outer surface of the pipe pile, so as to obtain the complete outer surface body data of the pipe pile.

[0053] Specifically, some of the rollers in the rollers are active rollers driven by a driving device. The active rollers are used to drive the pipe pile to rotate, and the other rollers are driven rollers, which are used to support the pipe pile. The control of the rotation angle of the pipe pile can be realized by installing a positioning block at the end of the pipe pile and cooperating with a laser locator. For example, the initial positioning block is installed at the rightmost position of the end face, and the laser locator and the positioning block are at the same horizontal height. The laser locator emits horizontal laser to identify the distance of the positioning block. After the pipe pile starts to rotate, the positioning block leaves the laser path, and the distance of the positioning block is recognized by the laser locator again after the pipe pile rotates 180°; it can be understood that for the control and detection of the rotation angle, those skilled in the art can also use other implementation methods. Since the control and detection of the rotation angle are conventional prior arts, no application will be made and no further description will be given.

[0054] In one embodiment, refer to Figure 4 , the end detection module includes 3D structured light cameras 201 located at both ends of the pipe pile for obtaining three-dimensional information of the pipe pile end faces; the 3D structured light camera is a three-dimensional information acquisition device. By obtaining the three-dimensional information of the target object, a three-dimensional model of the target object can be established. In the embodiment of the present application, refer to Figure 5 , the process of identifying the defects at the end of the pipe pile includes: collecting the spatial information data of the end face of the pipe pile end by the 3D structured light camera, thereby reconstructing the three-dimensional model of the end, and making a difference comparison between the three-dimensional model of the end and the design model to obtain the defects at the end of the pipe pile. Among them, the defect situations at the end of the pipe pile include inclination defects and depression defects.

[0055] Specifically, a difference comparison is made between the end three-dimensional model and the design model. First, the reference points of the end three-dimensional model and the design model are determined. For example, the leftmost points of the two models are used as the reference points. Then, the end three-dimensional model and the design model are aligned at the reference points. The first deflection angle is formed in the X direction due to the different Z component values of the spatial positions of the models, and the second deflection angle is formed in the Y direction due to the different Z component values of the spatial positions of the models. The first deflection angle and the second deflection angle can be used to calculate the inclination of the end face, and the difference between the Z component values of the spatial positions of each sampling point of the end three-dimensional model and the nearby sampling points can be used to calculate the depression of the end face.

[0056] In a further embodiment, the quality of the pile end can be evaluated by the pile end defects of the pipe pile. The process includes: grading the defects of the pipe pile end face and evaluating the quality of the pile end; the inclination defect is graded as qualified when the inclination angle is less than 1%, otherwise it is unqualified; the depression defect is graded as qualified when the maximum depression depth is less than 10 mm, otherwise it is unqualified. In the embodiments of the present application, the quality of the pile end is measured by the inclination defect and the depression defect of the pile end. Only when both the inclination defect and the depression defect are qualified, the quality of the pile end is considered qualified.

[0057] Specifically, the method for the 3D structured light camera to collect the end spatial information data and establish the end three-dimensional model uses the phase measurement profilometry method: First, through the light source projection, projection stripes, dot matrices or coded images are projected to form a specific pattern on the object surface; then image capture is performed to capture the deformation of these light patterns on the object surface; after that, the measurement profilometry algorithm provided by the 3D structured light camera is used to decode and match the image to determine the deformation of the light pattern on the object surface; finally, according to the deformation of the light pattern and the relative position relationship between the light source and the camera, the spatial coordinates of each sampling point on the object surface are calculated through calculation, and these sampling points can be used to construct the three-dimensional model of the object.

[0058] In yet another embodiment, refer to Figure 6, slide rails 103 parallel to the central axis of the pipe pile are provided on both sides of the bracket 101. The mobile pipe pile body scanning module 300 includes an equipment rack 301 arranged on the slide rails and capable of moving axially along the slide rails, and a 3D line scanning camera 302 and an image acquisition camera (not shown in the figure) arranged on the equipment rack. It can be understood that the mobile pipe pile body scanning module of course also includes auxiliary devices such as driving devices, main control devices, and positioning devices. For example, a driving motor for driving the equipment rack to move on the slide rails, a main control device connecting devices such as the driving motor, 3D line scanning camera, and image acquisition camera, and a positioning device for identifying the position where the equipment rack is located; among them, the main control device is used to run a control program preset in the main control device, and the control program controls each device of the mobile pipe pile body scanning module to cooperate to complete the scanning process; for those skilled in the art, different auxiliary devices can be set according to functional requirements, and the auxiliary devices belong to conventional technical means, and those skilled in the art do not need to make creative efforts.

[0059] It can be understood that in addition to the structure form of the equipment rack moving along the slide rails, the mobile pipe pile body scanning module can also adopt a form driven by a robot and a robotic arm, and acquisition devices such as a 3D line scanning camera and an image acquisition camera are arranged at the end of the robotic arm, and the robot and the robotic arm carry the acquisition devices to move to scan the pipe pile body and collect relevant data. In addition, in addition to the 3D line scanning camera, the acquisition device can also be replaced with a multi-view stereo vision camera to reconstruct a three-dimensional model through multi-view images.

[0060] Specifically, the process of the mobile pipe pile body scanning module executing the scanning process to identify the defects of the pipe pile body includes:

[0061] The equipment rack moves axially along the pipe pile. During the movement, the 3D line scanning camera scans the upper surface of the pipe pile to obtain three-dimensional point cloud data of the upper surface of the pipe pile, and the image acquisition camera takes pictures of the upper surface of the pipe pile to obtain image data of the upper surface of the pipe pile, and establishes a spatial position correspondence relationship between the three-dimensional point cloud data and the image data; the active rollers drive the pipe pile to rotate, turning the downward part of the outer surface of the pipe pile upwards; the equipment rack moves axially along the pipe pile again. During the movement, the 3D line scanning camera scans the second part of the outer surface of the pipe pile to obtain three-dimensional point cloud data of the lower surface of the pipe pile, and the image acquisition camera takes pictures of the lower surface of the pipe pile to obtain image data of the lower surface of the pipe pile; identify the pile body defects based on the three-dimensional point cloud data and the image data respectively and make a decision fusion.

[0062] In a further embodiment, refer to Figure 6 , both the 3D line scanning camera 302 and the image acquisition camera are arranged in pairs and distributed on the left and right sides of the pipe pile to ensure that the acquisition range can cover the entire outer surface of the upward part of the pipe pile.

[0063] In a further embodiment, the process of respectively identifying the pile body defects based on the three-dimensional point cloud data and the image data and performing weighted fusion includes:

[0064] First, locate the regions of sudden change in the curvature of the pipe pile body based on the three-dimensional point cloud data, such as concave regions or convex regions; locate the regions of sudden change in the curvature of the pipe pile body based on the three-dimensional point cloud data. One implementation is to compare the deviation between the actually collected three-dimensional point cloud data of the pipe pile and the design model. The places where the deviation is greater than the threshold are considered to have a large deformation of the pipe pile body, forming regions of sudden change in curvature. Another implementation is to calculate the change in the surface curvature of adjacent sampling points, and the regions where the change amplitude is greater than the threshold are identified as regions of sudden change in the curvature of the pipe pile body. Another implementation is to perform voxel noise reduction and normal vector calculation on the point cloud data to extract the regions of sudden change in curvature. Another implementation can also locate the regions of sudden change in the curvature of the pipe pile body through image data. For example, based on traditional vision preprocessing, the concave regions are located through morphological operations.

[0065] Establish a trained first-branch deep learning neural network for identifying structural defects, and input the three-dimensional point cloud data of the region of sudden change in curvature into the first-branch deep learning neural network. The deep learning neural network outputs the structural defects of the pipe pile body.

[0066] Establish a trained second-branch deep learning neural network for identifying surface flaw defects, and input the image data into the first-branch deep learning neural network, which outputs the surface flaw defects of the pipe pile body. The image data can also be pre-processed through image processing to enhance the texture to improve the recognition speed and accuracy of surface flaw defects.

[0067] Decision fusion: fuse the structural defects of the pipe pile body and the surface flaw defects of the pipe pile body output by the first-branch deep learning neural network and the second-branch deep learning neural network to obtain the defects of the pipe pile body.

[0068] Among them, the types of structural defects of the pipe pile body include internal broken steel bars and trapped straw ropes; the surface flaw defects of the pipe pile body include sticking skin and ash spots.

[0069] It can be understood that the first-branch deep learning neural network can also be replaced by network models such as PointNet++ network and Dynamic Graph Convolutional Network (DGCNN); the second-branch deep learning neural network can be replaced by network models such as ResNet-50 network or EfficientNet series network. For example, when inputting the two-dimensional unfolded image with enhanced texture, the ResNet-50 network is used to classify surface flaw defects such as sticking skin and ash spots.

[0070] In a further implementation, the input of the first-branch deep learning neural network is the three-dimensional point cloud data of the region of sudden change in the curvature of the pipe pile body, and the output is the type of structural defects of the pipe pile body. The learning and training of the first-branch deep learning neural network need to be trained by accumulating historical data:

[0071] First, collect and accumulate historical data for the analysis of the structural defects of the pipe pile body, including the three-dimensional point cloud data of the outer contour of the structural defects of the pipe pile body and the corresponding types of structural defects of the pipe pile body. The number of collected samples is not less than 500, and through operations such as rotating, mirroring, and stretching deformation on the three-dimensional point cloud data of the outer contour, the number of training samples is increased to 5000. Each sample includes the three-dimensional point cloud data of the outer contour and the type of structural defect of the pipe pile body with a corresponding relationship.

[0072] Use the three-dimensional point cloud data of each training sample as the input of the deep learning neural network to be trained, and the corresponding type of structural defect of the pipe pile body as the output of the deep learning neural network for the training of the deep learning neural network and perform supervised learning; perform training with no less than 5000 training samples, and measure the accuracy through a number of samples. Set the accuracy target above 95%. If the accuracy of the deep learning neural network does not meet the standard, continue to accumulate and increase the number of training samples until the accuracy meets the standard, marking the completion of the training of the deep learning neural network, which can be used by the visual inspection pipe pile quality system provided in this application to identify the structural defects of the pipe pile body.

[0073] In a further embodiment, the input of the second-branch deep learning neural network is the image data of the pipe pile body, and the output is the type of surface flaw defect of the pipe pile body. The learning and training of the second-branch deep learning neural network need to be trained by accumulating historical data:

[0074] First, collect and accumulate historical data for the analysis of the surface flaw defects of the pipe pile body, including the outer surface image data of the surface flaw defects of the pipe pile body and the corresponding types of surface flaw defects of the pipe pile body. The technical personnel mark and accumulate the collected samples. The number of collected samples is not less than 500, and through operations such as rotating, mirroring, adjusting the color level, flipping, and stretching deformation on the outer surface image data, the number of training samples is increased to 5000. Each sample includes the outer surface image data and the type of surface flaw defect of the pipe pile body with a corresponding relationship.

[0075] Use the outer surface image data of each training sample as the input of the deep learning neural network to be trained, and the corresponding type of surface flaw defect of the pipe pile body as the output of the deep learning neural network for the training of the deep learning neural network and perform supervised learning; perform training with no less than 5000 training samples, and measure the accuracy through a number of samples. Set the accuracy target above 95%. If the accuracy of the deep learning neural network does not meet the standard, continue to accumulate and increase the number of training samples until the accuracy meets the standard, marking the completion of the training of the deep learning neural network, which can be used by the visual inspection pipe pile quality system provided in this application to identify the surface flaw defects of the pipe pile body.

[0076] It is understandable that the operation and function implementation of the visual inspection pile quality system provided in the embodiments of the present application rely on computer programs to allocate or instruct corresponding hardware devices to complete. After those of ordinary skill in the art read and understand all or part of the implementation manners provided in the embodiments of the present application, it is easy to implement all or part of the process through computer programs, and there are no technical obstacles and no creative labor is required for those of ordinary skill in the art.

[0077] The embodiments of the present application further provide a method for visual inspection of pile quality. Based on visual inspection for identifying pile defects, the foregoing visual inspection pile quality system is used to detect the quality of piles and identify pile defects.

[0078] The embodiments of the present application further provide a computer program product or a computer program. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the electronic device reads the computer instructions from the computer-readable storage medium, and when the processor executes the computer instructions, it controls the visual inspection pile quality system provided in the embodiments of the present application to implement pile quality detection.

[0079] As mentioned above, only the preferred implementation manner of the present application is described, and there is no limitation in any form and substance to the present application. It should be noted that for those of ordinary skill in the technical field, several improvements and supplements can still be made without departing from the present application, and these improvements and supplements should also be regarded as the protection scope of the present application. For those who are familiar with the professional technology, any equivalent changes such as slight modifications, decorations and evolutions made without departing from the content and scope of the present application shall be equivalent embodiments of the present application; at the same time, any equivalent changes, modifications and evolutions made to the above-mentioned implementation manners based on the substantial technology of the present application shall still fall within the scope of the technical solutions of the present application.

Claims

1. A visual inspection system for the quality of pipe piles, characterized in that, Including: A bearing module for bearing pipe piles, an end detection module, and a mobile pipe pile body scanning module; The end detection module is used to collect the end face data of the pipe pile end, and identify the defects of the pipe pile end through the end face data of the pipe pile end; the mobile pipe pile body scanning module moves along the pipe pile to collect the pipe pile body data, and identify the defects of the pipe pile body through the pipe pile body data; Summarize the defects of the pipe pile end and the pipe pile body to generate a defect report.

2. The visual inspection pile quality system according to claim 1, characterized in that The end detection module includes 3D structured light cameras arranged at both ends of the pipe pile for obtaining three-dimensional information of the pipe pile end face.

3. The visual inspection pipe pile quality system according to claim 2, wherein The process of identifying the defects of the pipe pile end includes: Reconstruct the three-dimensional model of the end based on the phase measurement profilometry method; Compare the three-dimensional model of the end with the design model to obtain the defects of the pipe pile end.

4. A visual inspection system for pipe pile quality according to claim 3, characterized in that, The defects of the pipe pile end include inclination defects and depression defects; Based on the defects of the pipe pile end, evaluate the quality of the pipe pile end. The process includes: Classify the defects of the pipe pile end face and evaluate the quality of the pipe pile end; The inclination defect outputs the actual inclination angle value; The depression defect outputs the actual depression depth value; If any of the inclination defect and the depression defect is unqualified, the pipe pile end is unqualified.

5. The visual inspection system for pipe pile quality according to claim 1, wherein, The bearing module includes a bracket for bearing the pipe pile; an array of rollers perpendicular to the central axis of the pipe pile is provided on the bracket, the rollers are in contact with the pipe pile, and a driving device is provided in the rollers to drive the active rollers, and the active rollers are used to drive the pipe pile to rotate.

6. The visual inspection system for pipe pile quality according to claim 5, wherein, A slide rail parallel to the central axis of the pipe pile is provided on the bracket. The mobile pipe pile body scanning module includes an equipment rack arranged on the slide rail and moving axially along the slide rail. A 3D line scan camera and an image acquisition camera are provided on the equipment rack.

7. The visual inspection system for pipe pile quality according to claim 6, wherein The process of identifying the defects of the pipe pile body includes: The equipment rack moves axially along the pipe pile. During the movement, the 3D line scan camera scans the upper surface of the pipe pile to obtain the three-dimensional point cloud data of the upper surface of the pipe pile, and the image acquisition camera takes a picture of the upper surface of the pipe pile to obtain the image data of the upper surface of the pipe pile; the active roller drives the pipe pile to rotate, and turns the lower surface of the pipe pile upwards; The equipment rack moves axially along the pipe pile again. During the movement, the 3D line scan camera scans the lower surface of the pipe pile to obtain the three-dimensional point cloud data of the lower surface of the pipe pile, and the image acquisition camera takes a picture of the lower surface of the pipe pile to obtain the image data of the lower surface of the pipe pile; Identify the defects of the pile body based on the three-dimensional point cloud data and the image data respectively and make a decision fusion.

8. The visual inspection system for pipe pile quality according to claim 7, characterized in that, The process of identifying the defects of the pipe pile body based on the three-dimensional point cloud data and the image data respectively and making a decision fusion includes: Locate the curvature mutation region of the pipe pile body based on the three-dimensional point cloud data; Establish a first branch deep learning neural network for identifying structural defects, input the three-dimensional point cloud data of the curvature mutation region into the first branch deep learning neural network, and output the structural defects of the pipe pile body; Establish a second branch deep learning neural network for identifying surface flaw defects, input the image data into the deep convolutional neural network, and output the surface flaw defects of the pipe pile body; Make a decision fusion, and fuse the structural defects of the pipe pile body and the surface flaw defects of the pipe pile body to obtain the defects of the pipe pile body.

9. The visual inspection system for pipe pile quality according to claim 8, wherein The structural defects of the pipe pile body include internal broken reinforcement bars and pile body defects caused by trapped straw ropes; the surface flaw defects of the pipe pile body include sticking skin and ash spots.

10. A method for visually inspecting the quality of pipe piles, characterized in that, Use the visual inspection pipe pile quality system according to any one of claims 1-9 to detect the quality of the pipe pile and identify the pipe pile defects.

Images