Seamless steel tube surface defect detection device

Abstract

The invention relates to the field of industrial detection, and discloses a seamless steel tube surface defect detection device, which comprises a support frame, which is used as an equipment foundation chassis and is used for supporting the stable operation of the whole equipment; the electric guide rail is arranged at the top of the supporting frame and used for driving the steel pipe to operate through a stepping motor; the clamping mechanism is arranged at the top of the electric guide rail and used for being driven by a stepping motor of the electric guide rail to drive the steel pipe to move; and the annular guide rail is arranged at the top of the support frame and is used for assisting the high-definition camera to flexibly change the photographing angle. Through a multi-angle and multi-direction full-coverage acquisition mode of a high-definition camera, an annular guide rail and an adjustable LED light source module, dead-corner-free detection on the surface of the steel pipe is realized, and meanwhile, defects such as cracks, recesses and scratches are accurately identified by combining image preprocessing technologies such as Gaussian filtering and histogram equalization and a deep learning algorithm; the detection precision is greatly improved, and missing detection and false detection are effectively avoided.

Description

Technical Field

[0001] The invention relates to the technical field of industrial detection, in particular to a surface defect detection device for a seamless steel pipe. Background Art

[0002] At present, in the field of industrial production, seamless steel pipes are widely used in petroleum, chemical, energy, construction and other industries as an important basic material. The quality of the steel pipe surface directly affects its performance under complex working conditions such as high pressure and corrosion. Therefore, the detection of surface defects of steel pipes has become an important link in the production process. Among the existing seamless steel pipe inspection technologies, traditional methods mainly include manual visual inspection, contact inspection and some non-contact machine vision inspection. Manual visual inspection relies on the operator's experience to observe and judge the surface of the steel pipe. Contact inspection uses a probe or roller to contact the surface of the steel pipe to find defects, while machine vision inspection uses industrial cameras to collect images and then process and analyze them. These methods have their own characteristics. Among them, non-contact machine vision inspection has gradually been used due to its high efficiency and non-destructive advantages, especially in some automated production lines. It plays an important role.

[0003] However, although the existing technology has met the needs of steel pipe surface quality inspection to a certain extent, there are still many shortcomings. Manual visual inspection is inefficient and depends on the skills of the operator. It is easily affected by factors such as fatigue and environment, resulting in false detection or missed detection; contact detection equipment may cause secondary damage when inspecting the surface of steel pipes, and has limited detection capabilities for complex surfaces and large-diameter steel pipes; and non-contact machine vision inspection avoids the above problems, but there is still room for optimization in terms of multi-angle detection coverage, defect recognition accuracy and positioning accuracy. For example, existing visual inspection equipment usually adopts a fixed-angle camera arrangement, resulting in some areas of the steel pipe surface not being effectively covered; some image processing algorithms have weak recognition capabilities for complex or subtle defects, especially when the lighting conditions change or the surface characteristics are more complex. They are easily disturbed; at the same time, it is difficult to accurately locate the three-dimensional spatial position of the defect, which limits the application of the detection data in the subsequent processing links. In addition, the existing detection equipment is not adaptable enough to meet the detection needs of steel pipes of different specifications, materials and surface characteristics, thereby increasing the complexity of equipment adjustment and maintenance. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a seamless steel pipe surface defect detection device, which solves the problems of insufficient surface detection coverage of seamless steel pipes, low defect recognition accuracy, inaccurate positioning and poor adaptability to steel pipes of multiple specifications.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a surface defect detection device for seamless steel pipes, comprising: The support frame, which serves as the basic chassis of the equipment, is used to support the stable operation of the entire equipment; An electric guide rail is arranged on the top of the support frame and is used to drive the steel pipe to run through a stepper motor; A clamping mechanism, which is arranged on the top of the electric guide rail and is used to be driven by the stepping motor of the electric guide rail to drive the steel pipe to move; The annular guide rail is arranged on the top of the support frame to assist the high-definition camera to flexibly change the shooting angle; A high-definition camera is installed inside the circular guide rail to detect the status of the steel pipe; A controller, which is internally arranged on one side of the support frame and is used for integrating detection functions; The clamping mechanism includes a driving block, which is driven by a stepper motor of an electric guide rail, a fixing frame 1 is fixed on the top of the driving block, a rotating shaft is rotated inside the fixing frame 1, a fixing plate is fixed on the outer wall of the rotating shaft, a spring is arranged between the fixing plate and the driving block, one end of the spring is fixed to the bottom of the fixing plate, the other end of the spring is fixed to the top of the driving block, a roller 1 is arranged on the top of the outer wall of the rotating shaft, an electric push rod is arranged inside the driving block, a fixing frame 2 is fixed on the output end of the electric push rod, the fixing frame 2 slides inside the driving block, and a roller 2 rotates inside the fixing frame 2.

[0006] Preferably, the controller comprises: Image acquisition module, used to collect high-definition images of the steel pipe surface from multiple angles; An image recognition module, connected to the image acquisition module, for performing defect recognition on the acquired images; The abnormality positioning module is connected to the image recognition module and is used to calculate the defect location based on the defect recognition result and output the location information; the control unit is connected to the image acquisition module, the image recognition module and the abnormality positioning module and is used to coordinate the work of each module and manage the overall operation of the system.

[0007] Preferably, the image acquisition module includes: A unit for controlling the displacement of the high-definition camera in real time, which is used to set the shooting angles of the high-definition camera in an interlaced manner; Adjustable LED light unit for even illumination to avoid reflections interfering with image quality.

[0008] Preferably, the image recognition module includes: An image preprocessing unit, used to remove noise, enhance contrast and correct distortion of the collected images; A defect feature extraction unit, used to extract the defect edge contour, area and aspect ratio based on an edge detection algorithm; Defect classification unit, used to classify defect types through deep learning algorithms.

[0009] Preferably, the image preprocessing unit denoises the image based on a Gaussian filtering algorithm, and enhances the image contrast through a histogram equalization algorithm.

[0010] Preferably, the defect feature extraction unit adopts an edge detection algorithm to extract the edge contour of the defect by calculating the gradient value of the image grayscale change, and optimizes the edge shape through morphological operations.

[0011] Preferably, the defect classification unit performs defect recognition based on a convolutional neural network, classifies by extracting local features and global features of the image, and outputs a probability distribution of defect types.

[0012] Preferably, the abnormality locating module includes: The defect coordinate calculation unit is used to calculate the spatial coordinates of the defect based on the image acquisition angle of the camera and the real-time position of the light source; the defect position recording unit is used to store the defect position data and transmit it to the control unit for subsequent processing.

[0013] Preferably, the defect coordinate calculation unit calculates the actual spatial coordinates of the defect on the surface of the steel pipe according to the position parameters of the camera and the position pixel points of the defect in the image by triangulation.

[0014] Preferably, the control unit is used for: Adjust the camera angle switch to achieve multi-angle shooting; Control the brightness and angle of the light source system to adapt to the surface characteristics of different light sources; Integrate the data output by the image recognition module and the anomaly location module to generate an inspection report including the defect type and location.

[0015] The present invention provides a seamless steel pipe surface defect detection device, which has the following beneficial effects: 1. The present invention adopts a high-definition camera, a circular guide rail and an adjustable LED light source module, combined with a multi-angle and multi-directional image acquisition method, to perform full coverage and no-dead-angle detection on the surface of the steel pipe. At the same time, the image recognition module uses image preprocessing technologies such as Gaussian filtering and histogram equalization as well as a deep learning algorithm to accurately identify surface defects of the steel pipe, effectively improving the detection accuracy of subtle defects such as cracks, dents, scratches, etc., and avoiding missed detection or false detection.

[0016] 2. The present invention ensures that the steel pipe can be detected in real time while moving and rotating smoothly through the synchronous cooperation of the clamping mechanism and the electric guide rail; the abnormality positioning module accurately calculates the three-dimensional spatial coordinates of the defect through triangulation, and combines the controller to store and process the distribution information of the defect, thereby realizing automatic positioning and real-time recording of the defect. This design simplifies the reliance on manual experience in the traditional detection process, improves the detection efficiency, and provides reliable basic data for subsequent steel pipe grading, repair and quality control.

[0017] 3. The present invention can adapt to the clamping requirements of steel pipes with different diameters through the elastic adjustment design of roller one, roller two and spring; in addition, the camera angle switching device can dynamically adjust the shooting angle according to the size of the steel pipe, and the brightness and lighting angle of the light source system can also be flexibly adjusted to meet the detection requirements of steel pipes with different surface materials and optical properties. This modular and adjustable design enables the device to have strong versatility and adaptability, can meet the needs of various industrial application scenarios, and reduce equipment operation and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A perspective view of the present invention; Figure 2 It is a schematic diagram of the clamping mechanism of the present invention; Figure 3 The internal structure distribution diagram of the clamping mechanism of the present invention; Figure 4 This is a distribution diagram of the internal structure of the controller of the present invention.

[0019] Among them, 1. support frame; 2. electric guide rail; 3. clamping mechanism; 4. ring guide rail; 5. high-definition camera; 6. controller; 31. drive block; 32. fixed frame 1; 33. rotating shaft; 34. fixed plate; 35. spring; 36. roller 1; 37. electric push rod; 38. fixed frame 2; 39. roller 2. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] Embodiment 1: Please see attached Figure 1 - Attachment Figure 3 The embodiment of the present invention provides a surface defect detection device for a seamless steel pipe, comprising: Support frame 1, which serves as the basic chassis of the equipment and is used to support the stable operation of the entire equipment; The electric guide rail 2 is arranged on the top of the support frame 1 and is used to drive the steel pipe to run through a stepper motor; The clamping mechanism 3 is arranged on the top of the electric guide rail 2 and is used to be driven by the stepping motor of the electric guide rail 2 to drive the steel pipe to move; The annular guide rail 4 is arranged on the top of the support frame 1 and is used to assist the high-definition camera 5 to flexibly change the photographing angle; A high-definition camera 5, which is arranged inside the annular guide rail 4 and is used to detect the state of the steel pipe; A controller 6, which is internally arranged on one side of the support frame 1 and is used for integrating a detection function; The clamping mechanism 3 includes a driving block 31, which is driven by a stepper motor of the electric guide rail 2. A fixed frame 32 is fixed on the top of the driving block 31, a rotating shaft 33 is rotated inside the fixed frame 32, a fixed plate 34 is fixed to the outer wall of the rotating shaft 33, a spring 35 is arranged between the fixed plate 34 and the driving block 31, one end of the spring 35 is fixed to the bottom of the fixed plate 34, and the other end of the spring 35 is fixed to the top of the driving block 31, a roller 36 is arranged on the top of the outer wall of the rotating shaft 33, an electric push rod 37 is arranged inside the driving block 31, a fixed frame 2 38 is fixed to the output end of the electric push rod 37, the fixed frame 2 38 slides inside the driving block 31, and a roller 2 39 rotates inside the fixed frame 2 38.

[0022] Specifically, the steel pipe is placed on the clamping mechanism 3. At this time, the steel pipe will squeeze the rollers 1 36 on both sides. The rollers 1 36 are in close contact with the surface of the steel pipe through their own rolling, so that they are clamped and fixed between the two rollers 1 36. When the steel pipe squeezes the roller 1 36, its position will squeeze the spring 35 through the fixing plate 34, and the spring 35 will be deformed, so that the clamping position of the roller 1 36 can be flexibly adjusted to adapt to steel pipes of different diameters, ensuring the stability and adaptability of the clamping. At the same time, the electric push rod 37 can drive the fixed frame 2 38 to move along the preset trajectory. The fixed frame 2 38 cooperates with the rollers 1 36 on both sides by pushing the roller 2 39 to jointly complete the efficient clamping of the steel pipe. This design not only enables the device to adapt to steel pipes of various specifications, but also ensures that the steel pipe is evenly stressed during the clamping process to avoid surface damage. After the clamping mechanism 3 firmly fixes the steel pipe, the steel pipe performs a smooth linear motion through the displacement of the driving block 31 under the precise drive of the stepping motor in the electric guide rail 2. While the displacement of the driving block 31 synchronously drives the steel pipe to perform a linear translation, the rollers 36 on both sides of the clamping steel pipe also rotate under the drive of the self-installed motor inside. The rotation of the roller 36 can not only drive the steel pipe to achieve a stable rotational motion, but also make the entire outer surface of the steel pipe uniformly exposed in the detection area, thereby providing comprehensive and blind-angle-free image data support for subsequent detection. When the steel pipe passes the position of the annular guide rail 4, the high-definition camera 5 can take high-definition photos of the steel pipe surface from multiple angles. The rotation of the roller 36 further improves the comprehensiveness and accuracy of image acquisition, ensuring that all the subtle defects on the steel pipe surface can be captured. Finally, the image captured by the high-definition camera 5 will be transmitted to the controller 6 for real-time processing. The controller 6 identifies and analyzes the image of the steel pipe surface based on a preset algorithm, and determines whether there are defects on the steel pipe surface, thereby quickly and accurately determining whether the steel pipe is qualified. This improves detection efficiency while ensuring the accuracy of image recognition, providing strong technical support for subsequent production and quality control.

[0023] Please see attached Figure 4 , the controller 6 comprises: Image acquisition module, used to collect high-definition images of the steel pipe surface from multiple angles; An image recognition module, connected to the image acquisition module, for performing defect recognition on the acquired images; The abnormality location module is connected to the image recognition module and is used to calculate the defect location based on the defect recognition result and output the location information; the control unit is connected to the image acquisition module, the image recognition module and the abnormality location module and is used to coordinate the work of each module and manage the overall operation of the system; The image acquisition module includes: A unit for controlling the displacement of the high-definition camera 5 in real time, which is used to stagger the shooting angles of the high-definition camera 5; an adjustable LED light unit, which is used to provide uniform lighting to avoid reflections interfering with the image quality; The image recognition module includes: An image preprocessing unit, used to remove noise, enhance contrast and correct distortion of the collected images; A defect feature extraction unit, used to extract the defect edge contour, area and aspect ratio based on an edge detection algorithm; Defect classification unit, used to classify defect types through deep learning algorithms; The image preprocessing unit denoises the image based on the Gaussian filtering algorithm and enhances the image contrast through the histogram equalization algorithm; The defect feature extraction unit uses edge detection algorithm to extract the edge contour of the defect by calculating the gradient value of the image grayscale change, and optimizes the edge shape through morphological operations; The defect classification unit performs defect recognition based on a convolutional neural network, extracts local and global features of the image for classification, and outputs the probability distribution of the defect type; The abnormal location module includes: A defect coordinate calculation unit, used to calculate the spatial coordinates of the defect based on the image acquisition angle of the camera and the real-time position of the light source; a defect position recording unit, used to store the defect position data and transmit it to the control unit for subsequent processing; The defect coordinate calculation unit calculates the actual spatial coordinates of the defect on the steel pipe surface according to the position parameters of the camera and the position pixel points of the defect in the image through triangulation; The control unit is used to: Adjust the camera angle switch to achieve multi-angle shooting; Control the brightness and angle of the light source system to adapt to the surface characteristics of different light sources; Integrate the data output by the image recognition module and the anomaly location module to generate an inspection report including the defect type and location.

[0024] Specifically, in this embodiment, the image acquisition module includes a high-definition camera, a camera angle adjustment device, and an adjustable LED light unit. The high-definition camera is used to acquire high-resolution images of the steel pipe surface, the camera angle adjustment device can flexibly adjust the camera shooting angle, and the LED light unit provides a uniform, non-interference light source for acquisition to avoid image distortion caused by lighting problems.

[0025] Specifically, the high-definition camera 5 uses an industrial-grade camera with a resolution of not less than 4K, which is distributed around the steel pipe. Generally, the cameras are evenly distributed inside the annular guide rail 4 at an angle of 120° and fixed on a movable camera bracket. As an option, the angle of the camera can be adjusted in real time by an electric adjustment mechanism to adapt to the appearance characteristics of steel pipes of different diameters. For example, for large-diameter steel pipes, the camera can be adjusted to a higher tilt angle to cover the side surface and upper area of ​​the steel pipe; for small-diameter steel pipes, the camera can be adjusted to a position closer to parallel to accurately capture its surface features.

[0026] In one possible implementation, the camera in the image acquisition module cooperates with the moving part of the annular guide rail 4. When the steel pipe moves linearly under the drive of the electric guide rail 2 and rotates through the clamping mechanism 3, the dynamic adjustment of the camera angle can ensure that every surface area of ​​the steel pipe is completely captured. In order to further improve the imaging quality, the camera can be equipped with an optical distortion correction function, which can correct the distortion caused by the optical characteristics of the lens in real time by calibrating the intrinsic parameter matrix and distortion coefficient of the camera. The specific formula for optical correction is as follows: x 校正 =x(1+k1r 2 +k2r 4 +k3r 6 ) y 校正 =y(1+k1r 2 +k2r 4 +k3r 6 ) Among them, x and y are the pixel coordinates before distortion, x 校正 and 校正 is the corrected pixel coordinate, r 2 =x 2 +y 2 It represents the radial distance from the pixel to the optical center. k1, k2, k3 are the radial distortion coefficients of the camera.

[0027] The adjustable LED light unit is another important component of the image acquisition module, which is used to provide uniform lighting conditions and reduce imaging errors caused by insufficient lighting or radiation interference. Generally, the LED light unit adopts a high-brightness array structure, and its light forms a uniform light field after passing through the diffuser. As an optimized design, the LED light unit can be linked with the camera bracket to adjust the angle and brightness of the light source to adapt to different light source surface reflection characteristics. For example, for the smooth polished steel pipe surface, the density of the light source diffuser can be appropriately increased to reduce the interference of strong reflected light on the image; for steel pipes with rough surfaces, a higher light source angle can be used to achieve full coverage lighting.

[0028] In some embodiments, the image acquisition module supports a dynamic exposure adjustment function to cope with environments with large changes in light intensity. Specifically, when the surface of the steel pipe is overexposed or underexposed due to changes in material or angle, the camera can automatically adjust the shutter speed and ISO value to balance the brightness and clarity of the image. This automatic adjustment is based on the histogram calculation of the real-time image collected by the camera. The specific adjustment formula is as follows: Among them, I 曝光 is the median value of exposure, L is the grayscale level of the image, and P(i) is the pixel distribution frequency of grayscale level iii. By adjusting the shutter and ISO of the camera, the median value of the histogram is close to the set target value, thereby achieving exposure optimization.

[0029] In order to improve acquisition efficiency, the image acquisition module also supports multi-frame stitching technology. For example, when the diameter of the steel pipe is large and a single shot cannot cover its entire surface, the camera can continuously shoot multiple images of adjacent areas, and then use feature point matching and stitching algorithms to synthesize a complete steel pipe surface image. The feature point extraction algorithms commonly used in this stitching process include the SIFT algorithm (scale-invariant feature transform) and the ORB algorithm (rapid feature point extraction), and the specific matching process is completed using the KNN (K nearest neighbor) algorithm.

[0030] The image preprocessing unit is used to perform preliminary processing on the high-definition image provided by the image acquisition module to remove noise, enhance contrast, and correct distortion. These operations can effectively improve image quality and reduce the complexity of subsequent analysis. Generally, the image preprocessing unit will first use the Gaussian filter algorithm to denoise the image. The formula of Gaussian filtering is as follows: Where G(x,y) represents the weight of the filter, σ is the standard deviation of the Gaussian distribution, and x,yx,yx,y are the coordinates of the pixel points. The algorithm removes noise by taking a weighted average of the pixel values ​​while retaining the edge information of the image.

[0031] In addition, in order to enhance the overall contrast of the steel pipe surface image, the preprocessing unit uses a histogram equalization algorithm to adjust the image. The core goal of histogram equalization is to improve the detail performance of low-contrast areas by redistributing gray values. The adjustment formula is: Among them, H new (i) is the new value of gray value i, L is the number of gray levels, MN is the image size, and P(j) is the cumulative probability distribution of gray level jjj.

[0032] In a possible implementation, the preprocessing unit also includes a distortion correction function. This function corrects the geometric distortion caused by the lens characteristics by calibrating the camera parameters. The specific correction formula is: x′=x(1+k1r 2 +k2r 4 +k3r 6 ) y′=y(1+k1r 2 +k2r 4 +k3r 6 ) Among them, x, y are the coordinates before distortion, and x′, y′ are the coordinates after correction.

[0033] The defect feature extraction unit is one of the core functions of the image recognition module, which is used to extract the geometric features and texture features of the defect from the preprocessed image. Generally, this unit uses an edge detection algorithm to extract the edge contour of the defect. Commonly used algorithms include the Sobel operator and the Canny operator. In some embodiments, the edge information is extracted by calculating the gradient value of the image grayscale. The calculation formula is: in, is the gradient of the image, and are the gradient values ​​of the image in the x and y directions respectively.

[0034] The defect classification unit is used to classify the extracted features and determine the type of surface defects of the steel pipe. In one implementation, the unit performs classification based on a convolutional neural network (CNN). The CNN model includes multiple convolutional layers, pooling layers, and fully connected layers, wherein the convolutional layer extracts local features of the image through a convolution kernel, and its calculation formula is: Among them, y(i,j) is the pixel value of the output feature map, x(i+m,j+n) is the pixel value of the input image, and w(m,n) is the weight of the convolution kernel.

[0035] In order to improve the accuracy of classification, the classification unit can also perform probability calculation on the defect classification results, specifically using the Softmax function, and its calculation formula is: Among them, P i is the probability of defect category i, z i is the feature value output by the classifier, and C is the total number of defect categories.

[0036] In this embodiment, the abnormality positioning module mainly includes a defect coordinate calculation unit and a defect position recording unit. The defect coordinate calculation unit is responsible for calculating the actual position of the defect in three-dimensional space based on the camera imaging geometric parameters and the motion state of the light source. The defect position recording unit is used to store the position information of the defect and connect with the control unit for further analysis or triggering related operations.

[0037] Specifically, the positioning principle of the defect coordinate calculation unit is based on triangulation. In general, the imaging model of the camera can be described by a perspective projection model, and the spatial coordinates of any point on the steel pipe surface can be calculated by the following formula: in: X, Y, and Z are the three-dimensional coordinates of the defect point in space; v is the pixel coordinate of the defect point on the image plane; c x ,c y is the coordinate of the camera optical center on the image plane; f x ,f y is the focal length of the camera (in pixels); d is the known distance between the steel pipe surface and the camera.

[0038] As an option, when there is a non-ideal motion trajectory on the steel pipe surface (such as position deviation caused by steel pipe spin or guide rail error), the defect coordinate calculation unit can dynamically correct the positioning result by obtaining the motion parameters of the steel pipe (such as translation speed and rotation angle) in real time. In this case, the compensated coordinates can be expressed as: X 修正 =X+ΔX,Y 修正 =Y+ΔY,Z 修正 =Z Among them, ΔX and ΔY are the correction amounts based on the motion model, and the calculation formula is: ΔX=R·cos(θ), ΔY=R·sin(θ) Among them, R is the radius of the steel tube's rotation; θ is the steel tube's rotation angle.

[0039] The function of the defect location recording unit is to store the spatial coordinates of each defect and transmit the results to the control unit through the data interface. In general, the recorded content includes the three-dimensional coordinates of the defect, the defect type, the defect characteristic parameters (such as area and aspect ratio), and the camera number and other information. In some embodiments, in order to facilitate subsequent analysis and processing, the recording unit will mark the detection time point of each defect in the form of a timestamp.

[0040] In a possible implementation, the abnormality location module can also perform statistical analysis on the distribution of defects, such as statistical distribution density of defects on the surface of the steel pipe, concentration of defects in a specific area, etc. The calculation formula for defect density is: in: ρ is the defect density; N 缺陷 is the number of defects in a specific area; A 区域 is the surface area of ​​the region.

[0041] Specifically, the anomaly location module can also generate a visual distribution map of defects, which can intuitively display the distribution of defects on the surface of the steel pipe by mapping the defect coordinates onto the expanded view of the steel pipe. This distribution map can provide a reference for subsequent repair or quality grading of the steel pipe.

[0042] In some embodiments, in order to improve positioning accuracy, the abnormal positioning module can also combine multi-sensor data for fusion positioning. For example, when using an infrared thermal imager or a laser ranging sensor to assist in detection, the positioning results of different sensors can be fused by weighted averaging, and the weights can be dynamically adjusted according to the positioning accuracy of each sensor.

[0043] Working principle: When the steel pipe is placed on the clamping mechanism 3, the steel pipe will squeeze the rollers 36 on both sides, and then be clamped between the two rollers 36 by the rolling of the rollers 36. When the position of the roller 36 is squeezed, the spring 35 will be squeezed through the fixing plate 34, so that the spring 35 will be deformed, and then the position of the roller 36 can be adjusted to adapt to steel pipes of different diameters. At the same time, the electric push rod 37 can drive the fixing frame 38 to move, so that the fixing frame 38 pushes the roller 39 to cooperate with the two sides. The steel pipe is clamped by roller 36. At this time, the steel pipe is driven by the stepper motor in the electric guide rail 2, and then the driving block 31 is displaced, thereby driving the synchronous displacement. In this way, the corresponding steel pipe can be driven to translate and can also be driven to rotate by the motors provided in the rollers 36 on both sides, so that the high-definition camera 5 can better recognize the image of the steel pipe when it passes through the position of the annular guide rail 4. The recognized image will finally be processed by the controller 6 to determine whether the steel pipe is normal.

[0044] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A surface defect detection device for seamless steel pipe, characterized in that: include: A support frame (1), which serves as the basic chassis of the equipment and is used to support the stable operation of the entire equipment; An electric guide rail (2) is arranged on the top of the support frame (1) and is used to drive the steel pipe to move through a stepping motor; A clamping mechanism (3) is arranged on the top of the electric guide rail (2) and is used to be driven by the stepping motor of the electric guide rail (2) to drive the steel pipe to move; An annular guide rail (4) is arranged on the top of the support frame (1) and is used to assist the high-definition camera (5) in flexibly changing the photographic angle; A high-definition camera (5), which is arranged inside the annular guide rail (4) and is used to detect the state of the steel pipe; A controller (6), which is internally arranged on one side of the support frame (1) and is used for integrating a detection function; The clamping mechanism (3) comprises a driving block (31), wherein the driving block (31) is driven by a stepper motor of an electric guide rail (2), a fixing frame 1 (32) is fixed on the top of the driving block (31), a rotating shaft (33) is rotatably arranged inside the fixing frame 1 (32), a fixing plate (34) is fixed on the outer wall of the rotating shaft (33), a spring (35) is arranged between the fixing plate (34) and the driving block (31), one end of the spring (35) is fixed to the bottom of the fixing plate (34), and the other end of the spring (35) is fixed to the top of the driving block (31), a roller 1 (36) is arranged on the top of the outer wall of the rotating shaft (33), an electric push rod (37) is arranged inside the driving block (31), a fixing frame 2 (38) is fixed on the output end of the electric push rod (37), the fixing frame 2 (38) slides inside the driving block (31), and a roller 2 (39) is rotatably arranged inside the fixing frame 2 (38).

2. A seamless steel pipe surface defect detection device according to claim 1, characterized in that: The controller (6) comprises: Image acquisition module, used to collect high-definition images of the steel pipe surface from multiple angles; An image recognition module, connected to the image acquisition module, for performing defect recognition on the acquired images; The abnormality location module is connected to the image recognition module and is used to calculate the defect location based on the defect recognition result and output the location information; The control unit is connected to the image acquisition module, the image recognition module and the abnormality positioning module to coordinate the work of each module and manage the overall operation of the system.

3. A seamless steel pipe surface defect detection device according to claim 2, characterized in that: The image acquisition module comprises: A unit for controlling the displacement of the high-definition camera (5) in real time, which is used to set the shooting angles of the high-definition camera (5) in an interlaced manner; Adjustable LED light unit for even illumination to avoid reflections interfering with image quality.

4. A seamless steel pipe surface defect detection device according to claim 2, characterized in that: The image recognition module comprises: An image preprocessing unit, used to remove noise, enhance contrast and correct distortion of the collected images; A defect feature extraction unit, used to extract the defect edge contour, area and aspect ratio based on an edge detection algorithm; Defect classification unit, used to classify defect types through deep learning algorithms.

5. A seamless steel pipe surface defect detection device according to claim 4, characterized in that: The image preprocessing unit performs denoising on the image based on a Gaussian filtering algorithm and enhances the image contrast through a histogram equalization algorithm.

6. A seamless steel pipe surface defect detection device according to claim 4, characterized in that: The defect feature extraction unit adopts an edge detection algorithm to extract the edge contour of the defect by calculating the gradient value of the image grayscale change, and optimizes the edge shape through morphological operations.

7. A seamless steel pipe surface defect detection device according to claim 4, characterized in that: The defect classification unit performs defect recognition based on a convolutional neural network, classifies by extracting local features and global features of the image, and outputs a probability distribution of defect types.

8. The surface defect detection device for seamless steel pipe according to claim 2, characterized in that: The abnormality positioning module includes: A defect coordinate calculation unit, used to calculate the spatial coordinates of the defect based on the image acquisition angle of the camera and the real-time position of the light source; The defect location recording unit is used to store the defect location data and transmit it to the control unit for subsequent processing.

9. A seamless steel pipe surface defect detection device according to claim 8, characterized in that: The defect coordinate calculation unit calculates the actual spatial coordinates of the defect on the surface of the steel pipe according to the position parameters of the camera and the position pixel points of the defect in the image through triangulation.

10. The surface defect detection device for seamless steel pipe according to claim 2, characterized in that: The control unit is used for: Adjust the camera angle switch to achieve multi-angle shooting; Control the brightness and angle of the light source system to adapt to the surface characteristics of different light sources; Integrate the data output by the image recognition module and the anomaly location module to generate an inspection report including the defect type and location.

Images