Intelligent visual detection system and method for casing defects

Abstract

The present invention relates to an intelligent visual detection system and method for casing defects, belonging to the field of defect testing technology. The system comprises an imaging unit, which continuously or at set intervals captures 360-degree high-definition images / video streams of the inner wall of the casing during movement; an illumination unit, which provides uniform illumination for the inner wall of the casing and enables the illumination to be focused on the inner surface of the casing to be imaged; a positioning unit, which records in real time the precise downhole depth or relative position information corresponding to each frame of the image / video segment; a support unit, which is used to install the imaging unit, the illumination unit, and the positioning unit; and a magnetic particle detection device, which is used for supplementary detection after detection by the visual detection device. The present invention provides an intelligent visual detection system for casing defects, which is highly stable during movement within the casing, has good light concentration, and is more adaptable to different casing diameters.

Description

Technical Field

[0001] The present invention relates to an intelligent visual detection system and method for casing defects, belonging to the technical field of defect testing. Background Art

[0002] Oil and gas well casing is a critical component in maintaining the structural integrity of the wellbore. Over long-term service, affected by factors such as formation pressure, fluid erosion, and chemical corrosion, casing may develop defects such as corrosion, deformation, cracks, perforations, scaling, and wear. Failure to promptly identify and assess these defects can lead to reduced casing strength, seal failure, and even downhole accidents, resulting in significant economic losses and safety risks.

[0003] Traditional casing detection methods (such as electromagnetic and ultrasonic) each have their own advantages and disadvantages. Visual inspection is highly intuitive, but manual interpretation is inefficient and highly subjective.

[0004] As disclosed in the Chinese invention patent publication number CN119246532A, the device uses a regular hexagonal prism base for stable support; it is equipped with drive legs and moving wheels, allowing it to flexibly move inside the oil pipe. The built-in adjustment component works in conjunction with the drive mechanism through elastic parts to ensure the precise positioning of the camera and capture the details of the inner wall. The circularly distributed cameras and light sources, combined with the shooting holes on the blocking cover, form a visual coverage without blind spots, allowing direct visualization of the weld status. In addition, the integrated non-destructive testing component uses a lifting frame to carry a flaw detector for in-depth detection of internal defects and precise positioning. During use, the device has poor stability and the internal light cannot be well focused, resulting in limited imaging and low image recognition, which in turn leads to low recognizability of the inner wall surface of the casing. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the existing technology and provide an intelligent visual detection system for casing defects, which has high stability during movement inside the casing, good light concentration, and stronger adaptability to different casing diameters.

[0006] The casing defect intelligent visual detection system of the present invention comprises:

[0007] A pipe moving device for moving within the casing;

[0008] A visual inspection device connected to the front end of the pipeline moving device for imaging the inner wall of the pipeline;

[0009] The visual inspection device includes:

[0010] The support unit comprises a support frame arranged in a prismatic ring shape, wherein the support frame comprises a plurality of support plates evenly arranged in a ring shape;

[0011] An imaging unit, mounted inside the support frame, includes a plurality of high-definition cameras arranged circumferentially, with corresponding imaging holes opened on the support plate. The imaging unit continuously or at set intervals captures 360-degree high-definition images / video streams of the inner wall of the casing during movement;

[0012] At least two groups of lighting units are arranged along the axial direction thereof, each group of lighting units includes a lighting board provided with lighting lamps;

[0013] A driving mechanism that cooperates with the lighting panel;

[0014] The drive mechanism includes an adjustment screw driven by a servo motor, the adjustment screw being threadedly connected to an adjustment plate, the adjustment plate being connected to the lighting plate via an adjustment assembly, the adjustment assembly comprising a triangular plate connected to the lower end of the lighting plate and an obliquely arranged long slotted hole provided in the triangular plate, the adjustment plate being slidably connected to the long slotted hole via a hinged rod; a group of lighting units are respectively mounted on both ends of the support frame along its axial direction, the drive mechanism of which is linked by an adjustment screw arranged along the extension of the support frame, the adjustment screws at both ends rotating in opposite directions;

[0015] Positioning unit, which records the precise downhole depth or relative position information corresponding to each frame of image / video segment in real time;

[0016] It includes a magnetic particle detection device, which is used for supplementary detection after the visual detection device has detected it.

[0017] Furthermore, the support unit includes a support frame, and the support frame includes a plurality of support plates evenly arranged in a ring shape. The support plates can be arranged in a hexagonal prism or an octagonal prism.

[0018] Furthermore, at least two guide rods are correspondingly mounted on the support frame, and the adjustment plate is slidably connected to the guide rods.

[0019] Furthermore, the guide rod is provided with a thread section I and a thread section II, the thread section I has a nominal diameter smaller than the guide rod diameter, and the thread section II has a nominal diameter larger than the guide rod diameter.

[0020] Furthermore, the driving mechanism includes a servo motor installed on the middle support surface of the support frame, the servo motor is connected to the adjustment screw through a bevel gear set, the adjustment screw is rotatably connected to the middle support surface of the support frame, one end of the adjustment screw is threadedly connected to the adjustment plate, and an adjustment component is correspondingly provided on the adjustment plate, and the adjustment component is used to drive the lighting board to swing.

[0021] Furthermore, a second group of lighting units is installed at one end of the support frame away from the lighting unit, with the same structure, wherein the corresponding ends of the adjustment screws are extended, and the adjustment screws located at both ends of the support surface of the support frame rotate in opposite directions.

[0022] Furthermore, the lighting lamp is an array of LED lamps.

[0023] The intelligent visual detection method for casing defects of the present invention comprises:

[0024] Step 1: The pipeline moving device carries the visual inspection device and lowers and positions the visual inspection device to the target casing section to be inspected;

[0025] Step 2: synchronous data acquisition, controlling the pipeline moving device to move axially within the target casing section, and synchronously using the visual inspection device to acquire image data of the casing inner wall and downhole depth information corresponding to the image data;

[0026] Step 3: Image data preprocessing: preprocessing the collected image data to improve image quality and standardize image format;

[0027] Step 4: Intelligent defect identification: The pre-processed image data is input into a pre-trained intelligent algorithm model, which automatically detects, locates, and classifies casing defects in the image.

[0028] Step 5: Result generation and output: Based on the recognition result of the intelligent algorithm model, a detection result including defect type, location, and optional quantitative information is generated and output.

[0029] Furthermore, the magnetic particle detection device is driven by the pipeline moving device to re-inspect the scratches in the casing defects.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The present invention adopts a hexagonal / octagonal prism support frame design, combined with a multi-faceted annularly arranged high-definition camera (with independent imaging holes on each side), to achieve 360° blind-angle imaging of the inner wall of the casing, thereby improving the circumferential detection coverage; the innovative articulated lighting panel and drive mechanism (including a servo motor, bevel gear set, and adjustment screw) can dynamically adjust the illumination angle of the LED light according to the change of the casing diameter, so that the shooting point of the high-definition camera is always in the brightest position, eliminating shadow interference. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a structural diagram of an embodiment of the present invention;

[0033] Figure 2 This is one of the structural diagrams of the visual inspection device according to an embodiment of the present invention;

[0034] Figure 3 This is the second structural diagram of the visual inspection device according to an embodiment of the present invention;

[0035] Figure 4 yes Figure 3 A partial enlarged view of the middle part;

[0036] Figure 5 is a schematic structural diagram of a mounting frame according to an embodiment of the present invention;

[0037] Figure 6 2. It is a structural diagram of a pipeline moving device installed in a magnetic particle detection device according to an embodiment of the present invention;

[0038] Figure 7 2 is a schematic structural diagram of a magnetic particle detection device according to an embodiment of the present invention;

[0039] Figure 8 This is a front view of a magnetic particle detection device according to an embodiment of the present invention;

[0040] Figure 9 yes Figure 8 A partial enlarged view of point B in the middle;

[0041] Figure 10 yes Figure 8 A partial enlarged view of point C in the middle;

[0042] Figure 11 is a schematic structural diagram of a rotating frame according to an embodiment of the present invention;

[0043] Figure 12 This is one of the structural diagrams of the mobile unit according to an embodiment of the present invention;

[0044] Figure 13 This is the second structural diagram of the mobile unit according to an embodiment of the present invention;

[0045] Figure 14 Schematic diagram of the working principle of the visual inspection device according to an embodiment of the present invention;

[0046] Figure 15 yes Figure 14 Bottom view of the visual inspection device;

[0047] Figure 16 yes Figure 15 Full cross-section view at GG in the middle;

[0048] Figure 17 yes Figure 15 Full cross-section view at HH in the middle;

[0049] Figure 18 2 is a schematic structural diagram of a guide rod according to an embodiment of the present invention;

[0050] In the picture:

[0051] 1. Pipe moving device; 11. Frame; 12. Moving mechanism; 121. Moving frame; 122. Moving track; 123. Connecting rod; 124. Elastic telescopic rod; 125. Driving motor; 126. Preload element; 1261. Sliding head;

[0052] 2. Visual inspection device; 21. Support frame; 211. Imaging hole; 212. Support surface; 22. High-definition camera; 23. Illumination board; 24. Illumination lamp; 25. Driving mechanism; 251. Bevel gear set; 252. Adjustment screw; 253. Servo motor; 254. Adjustment assembly; 2541. Triangular plate; 2542. Long slotted hole; 255. Adjustment plate; 256. Guide rod; 2561. Threaded segment I; 2562. Threaded segment II; 26. Positioning plate; 27. Connecting plate;

[0053] 3. Magnetic particle detection device; 31. Fixed plate; 311. Brush; 32. Rotating frame; 33. Motor frame; 34. Driving part; 341. Rotating motor; 342. Planetary gear; 343. Sun gear; 35. Magnetic particle flaw detector; 351. Magnetic particle flaw detector seat; 36. Electric cylinder; 361. Compression spring; 37. Accommodating chamber; 38. Water suction pump; 39. Collecting plate; 391. Water diversion trough; 392. Annular water trough. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0055] like Figures 1 to 18 As shown, the casing defect intelligent visual detection system of the present invention includes:

[0056] A pipeline moving device 1, used for moving within the casing;

[0057] The pipeline moving device 1 includes a frame 11, and a plurality of moving mechanisms 12 are installed outside the frame 11;

[0058] The moving mechanisms are evenly distributed on the outer surface of the frame 11. If the frame 11 is a hexagonal prism, six moving mechanisms 12 are installed accordingly. If the frame 11 is an octagonal prism, the moving mechanisms 12 are installed on each surface respectively, so at least eight need to be installed.

[0059] The moving mechanism 12 includes a moving frame 121 fixed to the frame, the moving frame 121 is connected to a moving track 122 through a parallelogram mechanism, the parallelogram mechanism is a connecting rod 123 respectively hingedly connected to the moving frame 121 and the moving track 122, at least two connecting rods 123 are installed in parallel between the moving frame 121 and the moving track 122, and a driving motor 125 is installed at one end of the moving track 122. The driving motor 125 is connected to the moving track 122 through an existing universal joint, thereby driving the driving motor 12 5 is transmitted to the mobile crawler 122, and a pre-tightening member 126 is hingedly installed on the frame 11. The pre-tightening member 126 includes a fixed end and an elastic output end. The elastic output end is slidably installed in the fixed end, and the fixed end is hinged to the frame 11. A compression spring is installed between the elastic output end and the fixed end for pre-tightening the elastic output end. A sliding head 1261 is fixed on the elastic output end. The sliding head 1261 is hinged to an elastic telescopic rod 124. The other end of the elastic telescopic rod 124 is hinged to the side plate of the mobile crawler 122.

[0060] The preload member 126 may also be designed as a cylinder, thereby pushing the sliding head 1261 to move, further adjusting the position of the movable track 122, and completing the fit and locking with the inner wall of the casing.

[0061] When the pipeline moving device 1 is placed in the casing, the pre-tightening member 126 pushes the sliding head 1261 to slide toward the front end, the sliding head 1261 drives the elastic telescopic rod 124 to move, and the elastic telescopic rod 124 pushes the moving track 122 to move, so that the moving track 122 fits the inner wall of the casing.

[0062] The visual inspection device 2 is connected to the front end of the pipeline moving device 1 and is used to image the inner wall of the pipeline;

[0063] The visual inspection device 2 includes:

[0064] An imaging unit, which continuously or at set intervals captures 360-degree high-definition images / video streams of the inner wall of the casing during movement;

[0065] The lighting unit provides uniform illumination to the inner wall of the casing, and can focus the illumination on the inner surface of the casing where imaging is required;

[0066] The positioning unit records the precise downhole depth or relative position information corresponding to each frame of image / video segment in real time. The positioning unit uses inertial navigation or a depth encoder installed in the pipeline moving device 1 to achieve real-time mapping between the image and the downhole position, support three-dimensional coordinate reconstruction of defects, and greatly improve positioning accuracy;

[0067] A supporting unit, used for mounting an imaging unit, an illumination unit, and a positioning unit;

[0068] It includes a magnetic particle detection device 3, which is used for supplementary detection after the visual detection device 2 is detected.

[0069] After the visual inspection device 2 completes the inspection, it is sometimes difficult to distinguish between scratches and cracks in the picture. Scratches do not affect subsequent use and do not require repair. However, cracks are more serious and require welding repairs to prevent them from expanding. At this time, re-inspection is required through the magnetic particle inspection device 3 to re-inspect specific points and ensure the accuracy of identification.

[0070] The magnetic particle inspection device 3 includes a fixed disk 31 mounted on the front end of the pipeline moving device 1. One end of the fixed disk 31 is rotatably connected to a rotating frame 32. A brush 311 is installed between the rotating frame 32 and the fixed disk 31 for providing power to the electronic components on the rotating frame 32. Driven by the motor inside the pipeline moving device 1, a rotatable magnetic particle flaw detector 35 is mounted on the rotating frame 32. A driving member 34 is mounted on the rotating frame 32, which drives the magnetic particle flaw detector 35 to rotate, thereby performing axial or circumferential inspection on the inner wall of the casing.

[0071] A collecting tray 39 is installed at one end of the rotating frame 32 away from the pipeline moving device 1. Part of the collecting tray 39 is made of rubber material, which is convenient for moving in sleeves of different diameters.

[0072] A accommodating chamber 37 is provided in the middle of the collecting tray 39, and an annular water trough 392 is reserved between the upper end of the accommodating chamber 37 and the collecting tray 39. A water diversion trough 391 is provided on the collecting tray 39. A water suction pump 38 is installed at the upper end of the accommodating chamber 37, and the water suction end of the water suction pump 38 extends into the annular water trough 392. The water suction pump 38 is used to suck water in the annular water trough 392 into the accommodating chamber 37. The water suction pump 38 provides a certain pressure inside the accommodating chamber 37. A water spray pipe is connected to the accommodating chamber 37, and a corresponding magnetic particle flaw detector 35 is installed at one end with a water spray head for spraying the magnetic ink in the accommodating chamber 37 onto the inner wall of the pipe to detect cracks.

[0073] The driving member 34 includes a rotating motor 341 mounted on the rotating frame 32. The rotating motor 341 is mounted on the rotating frame 32 through the motor frame 33. The output end of the rotating motor 341 is installed with a planetary gear 342. The corresponding rotating frame 32 is installed with a sun gear 343. The sun gear 343 is connected to a magnetic particle flaw detection seat 351. The magnetic particle flaw detection seat 351 is slidably connected to the magnetic particle flaw detector 35. The magnetic particle flaw detection seat 351 is installed with an electric cylinder 36. The output end of the electric cylinder 36 is connected to a compression spring 361. The output end of the electric cylinder 36 is connected to the magnetic particle flaw detector 35 through the compression spring 361, so that the magnetic particle flaw detector 35 fits against the inner wall of the casing.

[0074] The sun gear 343 is respectively mounted on both sides of the rotating frame 32, thereby driving the magnetic particle inspection seat 351 to rotate in different directions, thereby preventing the connecting pipe of the water spray head from being entangled. Because when the rotating frame 32 rotates in one direction, the water spray heads next to the magnetic particle inspection devices 35 on both sides are mounted in the same direction. The magnetic particle inspection devices 35 on both sides can work simultaneously. When rotating, they can also perform axial inspections at the same time, which greatly improves the efficiency of inspection.

[0075] The support unit includes a support frame 21 , which includes a plurality of support plates evenly arranged in a ring shape. The support plates can be arranged in a hexagonal prism or an octagonal prism. An imaging unit is correspondingly installed inside the support frame 21 .

[0076] The imaging unit includes several high-definition cameras 22, which are respectively installed on the inner side of the support plate. The support plate is provided with corresponding imaging holes 211. These high-definition cameras 22 are preferably industrial-grade high-definition cameras with high resolution and high frame rate to capture clear details. The circular and uniform arrangement of the imaging holes 211 is intended to achieve 360-degree panoramic imaging of the inner wall of the casing without blind spots.

[0077] One end of the support frame 21 is connected to the lighting unit, which includes a lighting board 23 hinged to the corresponding support plate. The lighting board 23 is provided with a lighting lamp 24. The support frame 21 is equipped with a driving mechanism 25 for driving the lighting board 23 to swing along the hinge point.

[0078] The lighting lamp 24 is an array of LED lamps.

[0079] The driving mechanism 25 includes a servo motor 253 installed on the middle support surface 212 of the support frame 21. The servo motor 253 is connected to the adjustment screw 252 through the bevel gear set 251. The adjustment screw 252 is rotatably connected to the middle support surface 212 of the support frame 21. One end of the adjustment screw 252 is threadedly connected to the adjustment plate 255. The adjustment plate 255 is correspondingly provided with an adjustment component 254. The adjustment component 254 is used to drive the lighting panel 23 to swing.

[0080] like Figure 3 、 5 As shown in Figure 14, the two ends of the adjusting screw rod 252 are rotatably connected to the positioning plate 26 and the connecting plate 27 respectively. The two ends of the adjusting screw rod 252 are rotatably connected to the positioning plate 26 and the connecting plate 27 through the ball screw support seat, providing support for the adjusting screw rod 252, thereby making the adjusting screw rod 252 rotate more smoothly.

[0081] A second group of lighting units is installed at one end of the support frame 21 away from the lighting unit, with the same structure, wherein the corresponding ends of the adjustment screw rods 252 are extended, and the adjustment screw rods 252 located at both ends of the support surface 212 on the support frame 21 rotate in opposite directions, thereby controlling the swing angles of the lighting units on both sides, so that the light can be better irradiated onto the inner wall of the sleeve, thereby forming a clear picture during imaging.

[0082] The adjustment assembly 254 includes a triangular plate 2541 connected to the lower end of the lighting panel 23. The triangular plate 2541 is provided with a long slot hole 2542 which is arranged obliquely relative to the lighting panel 23. The adjustment plate 255 is hinged to the long slot hole 2542. When the adjustment plate 255 moves, the hinge rod of the adjustment plate 255 moves along the long slot hole 2542, thereby causing the lighting panel 23 to swing through the oblique long slot hole 2542.

[0083] At least two guide rods 256 are correspondingly mounted on the support frame 21 , and the adjustment plate 255 is slidably connected to the guide rods 256 . The arrangement of the guide rods 256 allows the adjustment plate 255 to run more smoothly.

[0084] like Figure 18 As described above, the guide rod 256 is respectively provided with a threaded section I 2561 and a threaded section II 2562. The nominal diameter of the threaded section I 2561 is smaller than the diameter of the guide rod 256, which facilitates the installation of the adjustment plate 255 on one side. The nominal diameter of the threaded section II 2562 is larger than the diameter of the guide rod 256, which facilitates installation inside the support frame 21.

[0085] Working process:

[0086] like Figures 14 to 17 As shown, the servo motor 253 is controlled to rotate by an external controller, and the servo motor 253 transmits power to the adjustment screw 252 through the bevel gear set 251. The adjustment screw 252 rotates, and the nut on the adjustment plate 255 that cooperates with the adjustment screw 252 causes the adjustment plate 255 to move linearly along the guide rod 256, thereby realizing the reciprocating movement of the adjustment plate 255 by rotating the adjustment screw 252; during the movement of the adjustment plate 255, the hinge column on the adjustment plate 255 that is slidingly hinged to the long slot hole 2542 moves in the long slot hole 2542, thereby pushing the triangular plate 2541 to move. Since the triangular plate 2541 is welded to the lower end of the lighting plate 23, the triangular plate 2541 drives the lighting plate 23 to swing during the movement, thereby adjusting the lighting angle of the lighting plate 23 (as shown in FIG. Figure 14 shown), Figure 14 The middle arrow F represents the irradiation angle of the light, and the vertical line E represents the inner wall of the sleeve. Through the swing of the lighting plate 23, the angle of the arrow F on one side becomes smaller, thereby adapting to the sleeve with a small diameter. Through adjustment, the lighting lamp 24 can be better focused on the inner wall of the sleeve.

[0087] The intelligent visual detection method for casing defects includes:

[0088] Step 1: The pipeline moving device 1 is lowered and positioned with the visual inspection device 2, and the visual inspection device 2 is lowered to the target casing section to be inspected;

[0089] The imaging unit (such as CMOS / CCD sensor) of the visual detection device 2 is calibrated for white balance, color correction and distortion correction.

[0090] Test and adjust the light intensity uniformity of the lighting unit to ensure uniform illumination under the target casing inner diameter.

[0091] Perform zero point calibration and accuracy verification on the positioning unit (high-precision depth encoder or IMU).

[0092] Confirm that there is sufficient data storage space or the wireless / wired transmission link bandwidth meets the requirements.

[0093] Exposure time: Adjust according to light intensity and tool speed to avoid motion blur.

[0094] Light intensity: Set the value to ensure that the image is neither overexposed nor underexposed.

[0095] Depth Sampling Interval: Sets the frequency at which the encoder or IMU records depth.

[0096] Downhole operation: Connect the calibrated and parameter-set device through a suitable transportation method (electrical cable, optical cable, continuous tubing, tractor, etc.), install necessary stabilizers to ensure that the tool is centered, and lower it stably to a certain distance below the starting depth D of the target detection well section.

[0097] Step 2: synchronous data acquisition, controlling the pipeline moving device 1 to move axially within the target casing section, and synchronously acquiring image data of the casing inner wall and downhole depth information corresponding to the image data by the visual inspection device 2;

[0098] Start synchronous acquisition: At depth D, initialize depth recording and start the illumination, imaging and positioning units.

[0099] Uniform speed lifting (or lowering) collection: Control the winch or conveyor to move the device at a preset uniform speed V (e.g. 3-10 m / min).

[0100] Step 3: Image data preprocessing: preprocessing the collected image data to improve image quality and standardize image format;

[0101] Data unpacking and verification: Receive or read data and perform integrity verification.

[0102] Image distortion correction: Apply the camera intrinsic parameters and distortion coefficients obtained in step 1 to correct each frame of image.

[0103] Image stitching and unfolding:

[0104] Stitching: For multi-camera systems, use feature matching (such as SIFT, SURF) or pre-calibrated geometric relationships to stitch images together to generate a 360° panorama.

[0105] Step 4: Intelligent defect identification: The pre-processed image data is input into a pre-trained intelligent algorithm model, which automatically detects, locates, and classifies casing defects in the image.

[0106] The preprocessed (usually expanded) image is fed into a pre-trained defect detection model (such as YOLOv8, Faster R-CNN, DETR, etc.). The model needs to be trained on a large dataset of images annotated with different types of casing defects (corrosion pits, cracks, deformations, perforations, scratches, etc.).

[0107] For each detected BBox, input it into a pre-trained instance segmentation model (such as Mask R-CNN, U-Net variants) or obtain a pixel-level mask (Mask) M for the detection model itself (such as some YOLO variants). M is a binary image that marks the exact pixels occupied by the defect.

[0108] Extract features of the segmented defect area (which can be deep features extracted by CNN; or manual features, such as shape descriptors, texture features LBP / GLCM, etc.).

[0109] Step 5: Result generation and output: Based on the recognition results of the intelligent algorithm model, a detection result including defect type, location, and optional quantitative information is generated and output.

[0110] Shading or shape-based shading: Less precise, but provides relative depth information.

[0111] Based on industry standards (such as API 5CT / 5B) or custom rules, combined with defect type, size, estimated depth / loss rate, quantity, and density, a scoring model or rule base is established to automatically assess the risk level of individual defects and entire well sections (such as Grade 1-5 or Low / Medium / High Risk).

[0112] For scratches in casing defects, the pipeline moving device 1 drives the magnetic particle detection device 3 to perform re-inspection.

[0113] To detect scratches that cannot be identified, magnetic ink is sprayed and then viewed through the camera of the magnetic particle detector 35 to see the molding state of the magnetic ink and thus determine the scratches or cracks.

[0114] The description of the direction and relative position relationship of the structure in the present invention, such as the description of front, back, left, right, up and down, does not constitute a limitation of the present invention and is only for the convenience of description.

Claims

1. An intelligent visual detection system for casing defects, comprising: A pipeline moving device (1) for moving within the casing; A visual detection device (2) is connected to the front end of the pipeline moving device (1) and is used to image the inner wall of the pipeline; It is characterized in that The visual inspection device (2) comprises: The support unit comprises a support frame (21) arranged in a prismatic annular shape, wherein the support frame (21) comprises a plurality of support plates evenly arranged in an annular shape; An imaging unit is installed inside the support frame (21), comprising a plurality of high-definition cameras (22) arranged circumferentially, and corresponding imaging holes (211) are opened on the support plate. The imaging unit continuously or at set intervals captures 360-degree high-definition images / video streams of the inner wall of the casing during movement; The lighting assembly includes a first lighting unit and a second lighting unit symmetrically arranged along the axis of the support frame, each lighting unit including: a lighting panel provided with lighting lamps; A driving mechanism (25) matched with the lighting panel (23) for adjusting the illumination angle of the lighting panel; The driving mechanism (25) includes an adjusting screw (252) driven by a servo motor (253), the adjusting screw (252) is threadedly connected to an adjusting plate (255), the adjusting plate (255) is connected to the lighting plate (23) through an adjusting assembly (254), the adjusting assembly (254) includes a triangular plate (2541) connected to the lower end of the lighting plate (23) and an obliquely arranged long slot (2542) opened on the triangular plate (2541), and the adjusting plate (255) is slidably connected to the long slot (2542) through a hinge rod; the support frame (21) is respectively equipped with a group of lighting units along its two axial ends, and its driving mechanism (25) is linked by the adjusting screw (252) extended along the support frame (21), and the adjusting screws (252) at both ends rotate in opposite directions; Positioning unit, which records the precise downhole depth or relative position information corresponding to each frame of image / video segment in real time; It includes a magnetic particle detection device (3) for supplementary detection after detection by the visual detection device (2); The magnetic particle detection device (3) includes a fixed disk (31) mounted on the front end of the pipeline moving device (1), one end of the fixed disk (31) is rotatably connected to a rotating frame (32), and a brush (311) is installed between the rotating frame (32) and the fixed disk (31) for providing electric energy to electronic components on the rotating frame (32). The rotating frame (32) is driven by a motor inside the pipeline moving device (1), and a rotatable magnetic particle flaw detector (35) is installed on the rotating frame (32). A driving member (34) is installed on the rotating frame (32), and the driving member (34) drives the magnetic particle flaw detector (35) to rotate, thereby performing axial or circumferential detection on the inner wall of the casing; A collecting plate (39) is installed at one end of the rotating frame (32) away from the pipeline moving device (1). Part of the collecting plate (39) is made of rubber material, which is convenient for moving in sleeves of different diameters. A receiving chamber (37) is provided in the middle of the collecting tray (39), an annular water trough (392) is reserved between the upper end of the receiving chamber (37) and the collecting tray (39), a water guide trough (391) is provided on the collecting tray (39), a water suction pump (38) is installed on the upper end of the receiving chamber (37), the water suction end of the water suction pump (38) extends into the annular water trough (392), the water suction pump (38) is used to suck water in the annular water trough (392) into the receiving chamber (37), the water suction pump (38) provides a certain pressure inside the receiving chamber (37), a water spray pipe is connected to the receiving chamber (37), and a corresponding magnetic particle flaw detector (35) is installed at one end with a water spray head for spraying the magnetic ink in the receiving chamber (37) onto the inner wall of the pipe to detect cracks.

2. The casing defect intelligent visual detection system according to claim 1, characterized in that: The support unit comprises a support frame (21), and the support frame (21) comprises a plurality of support plates evenly arranged in an annular shape, and the support plates can be arranged in a hexagonal prism or an octagonal prism.

3. The casing defect intelligent visual detection system according to claim 1, characterized in that: At least two guide rods (256) are correspondingly mounted on the support frame (21), and the adjustment plate (255) is slidably connected to the guide rods (256).

4. The casing defect intelligent visual detection system according to claim 3, characterized in that: The guide rod (256) is provided with a threaded section I (2561) and a threaded section II (2562), respectively. The nominal diameter of the threaded section I (2561) is smaller than the diameter of the guide rod (256), and the nominal diameter of the threaded section II (2562) is larger than the diameter of the guide rod (256).

5. The casing defect intelligent visual detection system according to claim 1, characterized in that: The driving mechanism (25) includes a servo motor (253) mounted on the middle support surface (212) of the support frame (21). The servo motor (253) is connected to the adjustment screw (252) via the bevel gear set (251). The adjustment screw (252) is rotatably connected to the middle support surface (212) of the support frame (21). One end of the adjustment screw (252) is threadedly connected to the adjustment plate (255). The adjustment plate (255) is correspondingly provided with an adjustment component (254). The adjustment component (254) is used to drive the lighting panel (23) to swing.

6. The casing defect intelligent visual detection system according to claim 1, characterized in that: A second group of lighting units is mounted on one end of the support frame (21) away from the lighting units, and has the same structure, wherein the corresponding ends of the adjustment screw rods (252) are extended, and the adjustment screw rods (252) located at both ends of the support surface (212) on the support frame (21) rotate in opposite directions.

7. The casing defect intelligent visual detection system according to claim 1, characterized in that: The lighting lamp (24) is an array of LED lamps.

8. A casing defect intelligent visual detection method, based on the casing defect intelligent visual detection system according to any one of claims 1 to 7, characterized in that: include: Step 1: The pipeline moving device (1) is lowered and positioned with the visual inspection device (2), and the visual inspection device (2) is lowered to the target casing section to be inspected; Step 2: synchronous data acquisition, controlling the pipeline moving device (1) to move axially within the target casing section, and synchronously acquiring image data of the casing inner wall and downhole depth information corresponding to the image data by the visual detection device (2); Step 3: Image data preprocessing: preprocessing the collected image data to improve image quality and standardize image format; Step 4: Intelligent defect identification: The pre-processed image data is input into a pre-trained intelligent algorithm model, which automatically detects, locates, and classifies casing defects in the image. Step 5: Result generation and output: Based on the recognition result of the intelligent algorithm model, a detection result including defect type, location, and optional quantitative information is generated and output.

9. The intelligent visual detection method for casing defects according to claim 8, characterized in that: For scratches in casing defects, the pipeline moving device (1) drives the magnetic particle detection device (3) to perform re-inspection.

Images