Method, device and storage medium for detecting an inner wall of a pipe

Abstract

The application discloses a pipeline inner wall detection method and device and a storage medium, relates to the technical field of pipeline detection, and comprises the following steps: constructing a motion module based on a hollow torque motor and a Mecanum wheel; collecting point cloud data of a pipeline inner wall in real time by driving a laser radar in the axial direction of the motion module; and performing data processing on the point cloud data to generate a three-dimensional model of the pipeline inner wall. The motion module is helpful to improve the environmental adaptability and scene coverage range of the device, break through the precision bottleneck, realize accurate identification of micro-damage, improve the data processing efficiency, and reduce the overall cost of pipeline detection.

Description

Technical Field

[0001] This application relates to the field of pipeline inspection technology, and in particular to a method, equipment and storage medium for inspecting the inner wall of a pipeline. Background Technology

[0002] As a core means to ensure the safety of energy transmission and municipal facilities, pipeline inspection technology has always been developed around the core needs of improving accuracy, enhancing environmental adaptability, and controlling costs.

[0003] Current mainstream technologies mainly include ultrasonic testing, eddy current testing, magnetic flux leakage testing, and electromagnetic ultrasonic testing. These technologies achieve defect identification through contact or short-range sensing principles. Ultrasonic testing relies on medium coupling to transmit sound wave signals, making it sensitive to the properties of the fluid inside the pipeline. It has a significant risk of failure under conditions of gas-liquid mixing or impurity deposition. While magnetic flux leakage testing avoids the medium coupling problem, its effective detection distance is strictly limited to within 1 mm, making it unsuitable for pipeline deformation or coating thickening scenarios. Furthermore, its identification probability and quantification accuracy for small-sized pinhole corrosion (<1 mm) and circumferential weld defects are insufficient, resulting in a deviation rate of 30% to 40% between on-site excavation verification results and inspection reports. Eddy current testing is limited by the principle of electromagnetic induction, resulting in slow detection speed and complex hardware systems, making it difficult to meet the needs of efficient long-distance pipeline inspection. Although electromagnetic ultrasonic testing has the advantages of being non-contact and resistant to medium interference, its complex hardware architecture and algorithm dependence drive up equipment costs, making deployment difficult for small and medium-sized enterprises to afford.

[0004] Therefore, existing technologies suffer from problems such as accuracy bottlenecks, limited applicability, and low efficiency, and cannot meet the integrated pipeline inspection needs for high stability, full adaptability, and low cost under complex working conditions.

[0005] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0006] The main objective of this application is to provide a method, equipment, and storage medium for pipeline inner wall inspection, aiming to solve the technical problems of low stability, poor environmental adaptability, and high cost in pipeline inspection under complex working conditions.

[0007] To achieve the above objectives, this application proposes a method for inspecting the inner wall of a pipeline, the method comprising: A motion module is constructed based on a hollow torque motor and a Mecanum wheel; The motion module drives the lidar axially to collect point cloud data of the inner wall of the pipe in real time. The point cloud data is processed to generate a three-dimensional model of the pipe's inner wall.

[0008] In one embodiment, the step of constructing the motion module based on the hollow torque motor and the Mecanum wheel includes: The rollers of the adjacent Mecanum wheel are installed with the pipe inner wall inspection device; The output shaft of the hollow torque motor is fixedly connected to the center of the inner hub of the Mecanum wheel; The hollow torque motor body is fixed to the main frame of the pipeline inner wall detection device by a rigid connecting rod; The Mecanum wheels are symmetrically distributed on both sides of the device and kept synchronized by rigid connecting rods; The stator of the hollow torque motor is fixed on a rigid connecting rod support, and the rotor of the hollow torque motor drives the wheel of the Mecanum wheel to construct the motion module of the pipeline internal detection device.

[0009] In one embodiment, the step of axially driving the lidar via the motion module to acquire point cloud data of the pipe's inner wall in real time includes: Obtain the pipe diameter; If the pipe diameter is smaller than the preset diameter, the lidar is placed at the front end of the pipe inner wall detection device, offset from the central axis of the pipe. If the pipe diameter is greater than or equal to the preset diameter, the lidar is placed at the position of the central axis of the pipe at the front end of the pipe inner wall detection device.

[0010] In one embodiment, the step of axially driving the lidar via the motion module to acquire point cloud data of the pipe's inner wall in real time includes: The position and forward speed of the pipe inner wall detection device are adjusted in real time by the hollow torque motor so that the lidar maintains a preset scanning distance from the pipe inner wall. The laser beam is emitted by the lidar and the reflected signal from the inner wall of the pipe is received. The reflected signal is transmitted to the terminal in real time via a cable passing through the shaft of the hollow torque motor.

[0011] In one embodiment, the step of processing the point cloud data to generate a three-dimensional model of the pipe inner wall includes: The point cloud data is preprocessed; The preprocessed point cloud data is sampled using a voxel grid to obtain a sparse point cloud. The sparse point cloud is reconstructed using Poisson reconstruction to generate a triangular mesh model; Density filtering is applied to the triangular mesh model, and the Taubin algorithm is used to smooth the density-filtered triangular mesh model. The smoothed triangular mesh model is simplified by a second error to obtain a three-dimensional model of the pipe inner wall.

[0012] In one embodiment, the step of sampling the preprocessed point cloud data using a voxel grid to obtain a sparse point cloud includes: The corresponding voxel dimensions are set based on the pipe diameter; The preprocessed point cloud data is divided into a uniform voxel grid based on the voxel size; Calculate the centroid coordinates of the point cloud within each voxel cell of the voxel mesh; Iterate through each voxel unit to check if there are any curvature points exceeding a preset threshold; If there is a curvature point in the voxel unit that exceeds a preset threshold, then the point with the highest curvature is used to replace all point clouds in the current voxel unit; otherwise, the point cloud corresponding to the centroid coordinates is used to replace all point clouds in the current voxel unit. Output sparse point cloud after voxel mesh sampling.

[0013] In one embodiment, the step of performing Poisson reconstruction on the sparse point cloud to generate a triangular mesh model includes: Calculate the normal direction of each point cloud and make the normal direction point towards the central axis of the pipe; The octree depth is determined based on the pipe diameter, and the point cloud space is divided into corresponding octrees according to the octree depth. An implicit surface is constructed based on the point cloud coordinates and normal directions stored in the leaf nodes of the octree. Solving the Poisson equation based on the implicit surface generates implicit functions; Isosurfaces are extracted based on the implicit function to generate a triangular mesh model.

[0014] In one embodiment, the step of performing density filtering based on the triangular mesh model and applying the Taubin algorithm to smooth the density-filtered triangular mesh model includes: Calculate the local density of the point cloud; Point clouds with local densities greater than a preset density threshold are selected and retained, while point clouds with local densities less than the preset density threshold are removed. The positions of the grid vertices of the triangular mesh model after filtering the point cloud are adjusted iteratively using the Laplacian operator.

[0015] Furthermore, to achieve the above objectives, this application also proposes a pipe inner wall inspection device, which includes: Motion module, hollow torque motor and Mecanum wheel; Control module, STM32 development board; Detection module, lidar, and data processing terminal.

[0016] In addition, to achieve the above objectives, this application also proposes a pipe inner wall inspection device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the pipe inner wall inspection method described above.

[0017] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the pipe inner wall detection method described above.

[0018] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the pipeline inner wall detection method described above.

[0019] One or more technical solutions proposed in this application have at least the following technical effects: This application constructs a motion module based on a hollow torque motor and a Mecanum wheel; the motion module axially drives a lidar to collect point cloud data of the pipe's inner wall in real time; the point cloud data is processed to generate a 3D model of the pipe's inner wall. This motion module helps improve the device's environmental adaptability and scene coverage, overcomes accuracy bottlenecks, enables precise identification of micro-damage, improves data processing efficiency, and reduces the overall cost of pipe inspection. Attached Figure Description

[0020] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic flowchart of an embodiment of the pipeline inner wall inspection method of this application; Figure 2 A simplified model diagram of the pipeline inner wall inspection device provided in this application; Figure 3-a1 , Figure 3-a2 A schematic diagram of the laser radar position provided for the pipeline inner wall detection device of this application; Figure 3-b1 , Figure 3-b2Another schematic diagram of lidar position provided for the pipeline inner wall detection device of this application; Figure 4 This is a flowchart illustrating Embodiment 2 of the pipeline inner wall inspection method of this application; Figure 5 This is a flowchart illustrating Embodiment 3 of the pipeline inner wall inspection method of this application; Figure 6 This is a flowchart illustrating Embodiment 4 of the pipeline inner wall inspection method of this application; Figure 7a , 7b 7c and 7c are schematic diagrams of the voxel grid before sampling, voxel division, and sampling provided by the pipeline inner wall detection method of this application, respectively. Figure 8 This is a flowchart illustrating Embodiment 5 of the pipeline inner wall inspection method of this application; Figure 9 This is a flowchart illustrating Embodiment Six of the pipeline inner wall inspection method of this application; Figure 10 This is a schematic diagram of the modular structure of the pipeline inner wall detection device according to an embodiment of this application; Figure 11 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the pipeline inner wall detection method in this application embodiment.

[0023] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0024] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0025] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0026] Existing technologies suffer from limitations in accuracy, applicability, and efficiency, making it impossible to meet the demand for integrated pipeline inspection that offers high stability, full adaptability, and low cost under complex working conditions.

[0027] This application provides a solution that constructs a motion module based on a hollow torque motor and a Mecanum wheel; the motion module axially drives a lidar to collect point cloud data of the pipe's inner wall in real time; the point cloud data is processed to generate a 3D model of the pipe's inner wall. This motion module helps improve the device's environmental adaptability and scene coverage, overcomes accuracy bottlenecks, enables precise identification of micro-damage, improves data processing efficiency, and reduces the overall cost of pipe inspection.

[0028] Based on this, this application provides a method for detecting the inner wall of a pipe, referring to... Figure 1 , Figure 1 This is a schematic flowchart of the first embodiment of the pipeline inner wall inspection method of this application.

[0029] In this embodiment, the pipe inner wall detection method includes steps S10~S30: Step S10: Construct a motion module based on a hollow torque motor and a Mecanum wheel; It should be noted that the hollow torque motor refers to a special motor with a through-type hollow structure, which has a large output torque and can directly transmit power; the Mecanum wheel refers to an omnidirectional wheel composed of multiple inclined rollers (usually at a 45° angle), which achieves multi-directional movement through the decomposition of friction between the rollers and the contact surface; the motion module is designed to provide stable and efficient driving capability for the pipeline inspection device, ensuring that the device maintains axial movement and automatic centering in complex pipeline environments (such as those containing oil, uneven welds). The design of directly driving the Mecanum wheel with a hollow torque motor can avoid the transmission losses and failure risks of traditional tracked or linkage structures, improving the environmental adaptability of the device.

[0030] In one possible implementation, the roller configuration is as follows: each Mecanum wheel is equipped with 10 rollers, dispersing the pressure from the pipe wall through multi-point contact; the output shaft of the hollow torque motor is fixedly connected to the center of the Mecanum wheel hub, achieving zero transmission loss (rigid connection); the two wheels rotate in opposite directions to counteract radial force, and the rigid connecting rod maintains synchronous movement; for pipes with steep inclines, wheel sets can be added in pairs to increase driving force. Furthermore, it should be noted that the stator is fixed to the connecting rod support to ensure efficient torque transmission, and the rotor directly drives the wheel rotation, forming a closed power chain. This design solves the problems of "crawling effect" and insufficient stability caused by the complex structure of traditional pipeline robots.

[0031] In one possible implementation, see Figure 2 It employs two Mecanum wheels with opposite directions of rotation (one left-handed and one right-handed), connected to the motor housing via an axial rigid connecting rod. The output shaft of the hollow torque motor is rigidly connected to the wheel hub, and the motor stator is fixed on the connecting rod support. When the motor drives the two wheels in opposite directions at the same speed, the radial components of the rollers cancel each other out, and the axial components are superimposed to propel the device smoothly along the pipeline axis.

[0032] Step S20: The motion module drives the lidar axially to collect point cloud data of the inner wall of the pipe in real time. It should be noted that the aforementioned lidar refers to an optical detection device that generates a three-dimensional spatial point cloud by emitting a laser beam and receiving reflected signals; the point cloud data refers to a set of spatial coordinates characterizing the geometric features of the inner wall of a pipe. This application, through the stable driving of the motion module, ensures that the lidar acquires pipe inner wall data in the optimal posture while moving forward, thus solving the shortcomings of traditional detection technologies that are sensitive to the medium (such as ultrasonic coupling problems) and limited in distance (such as magnetic leakage requiring a range of 1 mm).

[0033] In one possible implementation, a preset diameter (e.g., 300mm) is used as the dividing point. For pipes with a diameter less than 300mm, the lidar is offset from the central axis (eccentric placement) to expand the scanning radius and avoid near-field blind spots. For pipes with a diameter greater than or equal to 300mm, the lidar is placed on the central axis of the pipe (axial placement) to achieve uniform scanning around the entire circumference. The speed of the hollow torque motor is adjusted by a PID algorithm to maintain a preset scanning distance between the lidar and the pipe wall, achieving dynamic distance control. The lidar automatically reduces speed when encountering pipe bends or diameter changes to ensure data acquisition stability. The lidar cable passes through the hollow motor shaft to avoid entanglement and wear during operation. An IP67-rated or higher sealed housing protects the optical components from oil and water intrusion. Furthermore, it should be noted that the lidar bracket is equipped with silicone shock-absorbing pads to suppress mechanical vibration and ensure that the point cloud is free from jitter and distortion. This application overcomes the bottlenecks of traditional pipe inspection technologies, such as narrow applicability and large accuracy fluctuations.

[0034] In one specific implementation, for small-diameter pipes (such as 200mm in diameter), the control system drives the rotating bracket to offset the lidar from the central axis by 30mm, thereby expanding the lateral scanning range; by adjusting the motor speed through real-time feedback, the distance between the lidar and the pipe wall is adjusted, thereby improving the range and accuracy of the collected point cloud density.

[0035] Step S30: Process the point cloud data to generate a three-dimensional model of the inner wall of the pipe.

[0036] It should be noted that the aforementioned 3D model refers to a lightweight triangular mesh reconstructed through an algorithm, which can accurately represent damage details such as corrosion and welds (≥0.5mm depth). This application achieves the conversion from massive point clouds to high-fidelity models through algorithms, solving the problems of coarseness and low efficiency in traditional pipeline inner wall detection and reconstruction.

[0037] In one possible implementation, the point cloud is classified and converted into a standardized format, and granular noise points are removed; the voxel size is set according to the pipe diameter; the point cloud particles within the voxel are calculated, and high curvature points (such as corrosion edges) are preferentially retained to protect damage features; the normals are uniformly directed towards the central axis of the pipe; implicit functions are generated based on the octree depth (12 levels for small diameters and 10 levels for large diameters) to fill in missing areas such as corrosion pits; noise points of surface density are removed, and the Taubin algorithm is used to achieve iterative smoothing, suppressing jagged noise and preserving features; redundant triangular faces in flat areas are merged to improve the mesh simplification rate.

[0038] Furthermore, the step of "constructing a motion module based on a hollow torque motor and a Mecanum wheel" in step S10 is further refined, including: The rollers of the adjacent Mecanum wheel are installed with the pipe inner wall inspection device; It should be noted that the rollers of a Mecanum wheel refer to cylindrical rolling elements mounted on the rim at a specific angle (typically 45°). Their core function is to decompose the motor torque into axial and radial components through multi-point contact with the inner wall of the pipe. By having adjacent Mecanum wheels rotate in opposite directions, the pipe inner wall detection device moves axially. This application optimizes the roller layout to improve motion stability. The staggered arrangement of adjacent rollers increases the contact point density, disperses the pressure on the inner wall of the pipe, and adapts to complex surfaces such as uneven welds and oil deposits. By configuring a multi-point contact structure with 10 rollers per wheel, the risk of localized wear is effectively reduced, and environmental adaptability is enhanced.

[0039] The output shaft of the hollow torque motor is fixedly connected to the center of the inner hub of the Mecanum wheel; It should be noted that the output shaft of the hollow torque motor refers to the rotating shaft through which the motor transmits power, and its axis is a through-type hollow structure; the internal hub center of the Mecanum wheel refers to the power input interface in the center of the wheel body. This application achieves zero-transmission-loss power transmission through rigid connection, avoiding the energy loss and failure risks of traditional gear or belt drives. The hollow shaft design provides a sealed channel for the lidar cable, preventing the cable from tangling and wearing during operation. In addition, the connection method can preferentially use flange bolt fastening or interference fit to ensure connection reliability under high torque conditions.

[0040] The hollow torque motor body is fixed to the main frame of the pipeline inner wall detection device by a rigid connecting rod; It should be noted that the rigid connecting rod refers to a high-rigidity metal connector (such as aluminum alloy), and the main frame refers to the core load-bearing structure of the pipe inner wall inspection device. This application rigidly connects the motor housing to the frame, forming a stable mechanical support system to prevent power transmission misalignment due to motor vibration. The triangular support structure of the connecting rod enhances torsional stiffness and suppresses device swaying caused by unevenness in the pipe inner wall. In one possible implementation, the connecting rod and frame are connected by a detachable snap-fit ​​mechanism, facilitating quick replacement or expansion of the motor unit.

[0041] The Mecanum wheels are symmetrically distributed on both sides of the device and kept synchronized by rigid connecting rods; It should be noted that symmetrical distribution refers to the two Mecanum wheels being arranged mirror images of each other with the device axis as the center. This application uses rigid connecting rods to constrain the relative positions of the two wheels, ensuring that the radial force generated by the rollers during reverse rotation is precisely canceled out, achieving automatic centering of the device (without lateral offset). In one possible implementation, for large-diameter or high-slope environments, a second or more motor-Mecanum wheel units can be symmetrically added, increasing the driving force by 200% through parallel connecting rods.

[0042] The stator of the hollow torque motor is fixed on a rigid connecting rod support, and the rotor of the hollow torque motor drives the wheel of the Mecanum wheel to construct the motion module of the pipeline internal detection device.

[0043] It should be noted that the stator refers to the stationary component of the motor, and the rotor refers to the rotating component; the rigid connecting rod support refers to the extended support structure of the connecting rod described in step A203. This application achieves effective absorption of torque reaction force through stator fixation, while the direct-drive rotor wheel body ensures efficient power conversion. The overall construction of the motion module forms a closed power chain of "motor stator fixation → rotor drive hub → roller contact with pipe wall," solving the problems of "crawling effect" and insufficient stability in traditional pipeline robots. Furthermore, this application enables the pipeline inner wall detection device to travel at a deviation of less than 2mm in oily sewage pipelines, providing a zero-jitter scanning basis for lidar.

[0044] Furthermore, prior to step S20, "real-time acquisition of point cloud data of the inner wall of the pipe by axially driving the lidar through the motion module," the following steps are included: Obtain the pipe diameter; It should be noted that the pipe diameter refers to the inner diameter of the pipe to be inspected, and its value directly affects the scanning coverage and accuracy of the LiDAR. The preset diameter (e.g., 300mm) is a threshold set based on the pipe's geometric characteristics and the LiDAR's performance, used to differentiate between small-diameter and large-diameter pipes for inspection. The actual pipe diameter can be determined through LiDAR pre-scanning or external input (e.g., manual setting).

[0045] If the pipe diameter is smaller than the preset diameter, the lidar is placed at the front end of the pipe inner wall detection device, offset from the central axis of the pipe. It should be noted that, see Figure 3-a1 and Figure 3-a2 At this point, the offset of the lidar is calculated using the pipe diameter and a preset diameter (e.g., offset = pipe diameter / 4). Based on the offset of the lidar, the position of the lidar is adjusted so that the lidar is placed off-center to avoid near-field distortion.

[0046] If the pipe diameter is greater than or equal to the preset diameter, the lidar is placed at the position of the central axis of the pipe at the front end of the pipe inner wall detection device.

[0047] It should be noted that, see Figure 3-b1 and Figure 3-b2 Adjusting the position of the lidar on the central axis of the pipeline ensures that the laser beam evenly covers the entire circumference of the inner wall, and the scanning path spreads in concentric circles, making the point cloud density more uniform and reducing the missed detection rate at the edge of large-diameter pipelines.

[0048] In one possible implementation, the control system (such as an STM32 development board) drives the rotating bracket to adjust the radar position and offset distance based on the real-time pipe diameter data, ensuring that the distance between the lidar and the inner wall of the pipe remains within the optimal scanning range.

[0049] This embodiment overcomes the dependence of traditional detection devices on a single pipe diameter by dynamically adjusting the spatial position of the lidar, ensuring that the laser beam uniformly covers the entire circumference of the inner wall of the pipe, improving the uniformity of point cloud density and reducing the false detection rate.

[0050] Furthermore, referring to Figure 4 The second embodiment of the pipeline inner wall inspection method of this application provides a schematic flowchart, based on the above. Figure 4 The embodiment shown further refines step S20, "real-time acquisition of point cloud data of the inner wall of the pipe by axially driving the lidar through the motion module," including steps A201 to A203: Step A201: Adjust the position and forward speed of the pipe inner wall detection device in real time by means of the hollow torque motor so that the lidar maintains a preset scanning distance with the pipe inner wall; It should be noted that the preset scanning distance refers to the optimal working distance between the lidar and the inner wall of the pipe (typically 10–50 mm). Precise control of this distance directly affects the quality of point cloud data acquisition. The hollow torque motor dynamically adjusts the position and forward speed of the pipe inner wall detection device in real time, executed by the control system (an STM32 development board equipped with a PID algorithm). This aims to solve the problem of inaccurate scanning distance caused by fluctuations in travel speed or uneven pipe walls in traditional detection devices. By feeding back real-time distance data between the lidar and the inner wall of the pipe, the PID algorithm outputs control commands to the hollow torque motor, adjusting the Mecanum wheel speed to achieve smooth changes in the travel speed of the pipe inner wall detection device, ensuring the scanning distance remains stable within the preset range.

[0051] In one possible implementation, for curved sections of the pipeline, the control system automatically reduces the travel speed to 50% of the normal value, while fine-tuning the lidar pitch angle to compensate for distance deviations and avoid point cloud distortion.

[0052] Step A202: Emits a laser beam using a lidar and receives the reflected signal from the inner wall of the pipe; It should be noted that the selection of laser beam wavelength must take into account both penetration and resolution, while the intensity of the reflected signal is related to the incident angle and the pipe wall material. This application can capture geometric details through high-frequency pulse scanning (≥100Hz), and is particularly sensitive to damage features such as weld protrusions and corrosion depressions (≥0.5mm deep can be identified).

[0053] In one possible implementation, the lidar employs multi-echo reception technology to distinguish between the main reflection from the inner wall of the pipe and the scattering signal from the oil layer, thereby improving data reliability in complex environments.

[0054] Step A203: The reflected signal is transmitted to the terminal in real time based on the cable passing through the shaft of the hollow torque motor.

[0055] It should be noted that the through-type structure of the hollow shaft provides a sealed channel for the cable, eliminating the risk of entanglement during the journey that is common with traditional external cabling; the cable adopts a multi-layer shielding design (such as aluminum foil + braided mesh) to suppress the impact of motor electromagnetic interference on signal transmission.

[0056] In one possible implementation, the water pipeline (DN400, including uneven welds and localized corrosion) is inspected: the control system, based on real-time ranging data from the lidar, detects a distance fluctuation of 18–42 mm in the bend section. A PID algorithm adjusts the output of the hollow torque motor, reducing the travel speed from 0.3 m / s to 0.15 m / s. Simultaneously, the radar elevation angle is fine-tuned via the support to stabilize the scanning distance at 30 ± 2 mm. The lidar emits wavelength pulses to scan the inner wall of the pipeline, receiving strong reflection signals (intensity > 1200 units) from the weld area and weak reflection signals (intensity < 300 units) from the corrosion pits, generating an original point cloud containing spatial coordinates and reflection intensity. The point cloud data is transmitted via a shielded cable at the shaft of the hollow motor and uploaded to the terminal in real time via Ethernet protocol.

[0057] In this embodiment, the control system parses distance data and outputs motor control commands, the detection module independently performs signal reflection and reception, and the physical structure of the hollow torque motor and the data transmission protocol work together to achieve signal transmission to the terminal, forming a technical closed loop of "dynamic ranging to adaptive speed regulation, and then to anti-interference transmission". The detection accuracy of the inner wall of the pipeline is high and the environmental adaptability is strong.

[0058] Furthermore, referring to Figure 5 The third embodiment of the pipeline inner wall inspection method of this application provides a schematic flowchart, based on the above. Figure 5 The embodiment shown further refines step S30, "processing the point cloud data to generate a three-dimensional model of the pipe inner wall," by including steps A301 to A305: Step A301: Preprocess the point cloud data; It should be noted that preprocessing refers to classifying and converting the collected raw point cloud data, converting data from different sources into a standardized format (such as PLY or PCD), removing invalid noise points (such as suspended particle reflection points), and classifying them according to spatial location or reflection intensity. The classification of point cloud types also includes pipe wall body, weld seam, corrosion pit, etc., which can significantly improve reconstruction efficiency and reduce algorithm complexity.

[0059] Step A302: Sample the preprocessed point cloud data using a voxel grid to obtain a sparse point cloud; It should be noted that voxel mesh sampling refers to dividing the three-dimensional space into uniform cubic units (voxels). Each voxel contains several original points, simplifying the point cloud within each voxel into a single representative point, which significantly reduces the amount of data while retaining key geometric features and significantly improves Poisson reconstruction efficiency. The voxel size is dynamically set according to the pipe diameter. For adjusting the voxel size, larger voxels improve processing efficiency but lose details, while smaller voxels retain more original information. The balance is dynamically adjusted according to accuracy requirements. Sparse point cloud refers to a dataset with significantly reduced point cloud density after sampling.

[0060] Step A303: Perform Poisson reconstruction on the sparse point cloud to generate a triangular mesh model; It should be noted that Poisson reconstruction refers to the process of generating a continuous closed triangular mesh using an implicit surface fitting algorithm; the triangular mesh model refers to the geometric representation of the pipe's inner wall surface, consisting of vertices, edges, and faces. It can automatically fill in missing point cloud areas caused by oil contamination or laser occlusion (such as fractures at the edges of corrosion pits) by utilizing the continuity characteristics of the pipe's inner wall. In one possible implementation, a higher reconstruction weight is assigned to the corroded area to force the generation of a continuous surface, thereby increasing the integrity of the damaged area and avoiding the problem of missed detection of minute details in traditional magnetic flux leakage detection.

[0061] Step A304: Perform density filtering on the triangular mesh model and apply the Taubin algorithm to smooth the density-filtered triangular mesh model; It should be noted that density filtering refers to removing noisy triangular patches based on local surface density (number of triangular faces per unit volume); the Taubin algorithm refers to a mesh optimization method that achieves non-shrinking smoothing through iteration using two parameters, λ and μ. Density filtering is used to remove discrete noise points below a threshold (such as reflection points of impurities inside pipes), retaining data in high-density areas and improving model uniformity and realism. The Taubin algorithm is used to suppress high-frequency noise (such as jagged edges caused by scanning jitter) while preventing geometric features (such as weld protrusions and corrosion depressions) from being over-smoothed.

[0062] In one possible implementation, the smoothing intensity is reduced to 30% for high curvature vertices (curvature > 0.5) while maintaining the corrosion edge characteristics, thereby significantly reducing the false detection rate of impurities and surface roughness.

[0063] Step A305: Perform secondary error simplification on the smoothed triangular mesh model to obtain the three-dimensional model of the pipe inner wall.

[0064] It should be noted that the aforementioned quadratic error simplification refers to the process of merging redundant triangular faces based on the quadratic error metric (QEM) algorithm; the 3D model of the pipe inner wall refers to the lightweight and high-fidelity final output mesh. Quadratic error simplification prioritizes deleting triangular faces in flat pipe wall regions (with low error contribution values) and retains faces in high-curvature regions such as corrosion edges and cracks; the goal is to compress the model volume (simplification rate > 80%) while preserving damage details.

[0065] In one possible implementation, a geometric error threshold of 0.1 mm is set, and the model is iteratively simplified to the target number of facets to generate a lightweight 3D model, enabling real-time rendering and facilitating rapid damage analysis on mobile devices.

[0066] Furthermore, referring to Figure 6The fourth embodiment of the pipeline inner wall inspection method of this application provides a schematic flowchart, based on the above. Figure 6 The embodiment shown further refines the step "sampling the preprocessed point cloud data into a voxel grid to obtain a sparse point cloud" by including steps A401 to A406: Step A401: Set the corresponding voxel size based on the pipe diameter; It should be noted that pipe diameter refers to the inner diameter (in mm) of the pipe to be inspected, and its value directly affects the density distribution of point cloud data and the accuracy of damage detection; voxel size refers to the side length (in cubic millimeters) of a cubic mesh unit in three-dimensional space, and this parameter determines the compression degree and feature preservation capability of voxel mesh sampling. Dynamically optimizing the voxel size based on the actual pipe dimensions solves the problems of detail loss in small-diameter pipes or low efficiency in large-diameter pipes caused by traditional fixed voxel sizes.

[0067] In one possible implementation, see Figure 7a , 7b Figures 7 and 7c correspond to schematic diagrams before voxel mesh sampling, after voxel division, and after sampling, respectively. For small-diameter pipes (≤300mm), 1 cubic millimeter voxels are used to retain ≥0.5mm deep corrosion details, while for large-diameter pipes (>500mm), 2 cubic millimeter voxels are used to balance processing speed and accuracy requirements, ensuring effective retention of point cloud features under different pipe diameters.

[0068] Step A402: Divide the preprocessed point cloud data into a uniform voxel grid based on the voxel size; It should be noted that a voxel mesh refers to a collection of 3D point cloud spaces divided into multiple cubic units of equal volume; uniform partitioning means that all cubes are of the same size and are closely packed without overlap. See also Figure 7a , 7b Figures 7 and 7c represent schematic diagrams before, after, and after voxel mesh sampling, respectively. A cubic cell array covering the entire area is generated by equidistantly dividing the point cloud space along the X / Y / Z axes according to voxel size. The aim is to construct a spatial index structure, providing a computational basis for point cloud simplification.

[0069] In one possible implementation, voxel units with incomplete boundaries are expanded and filled to ensure that no point cloud is missed, so that subsequent point cloud replacements can be computed in parallel, thus improving efficiency.

[0070] Step A403: Calculate the centroid coordinates of the point cloud within each voxel cell of the voxel mesh; It should be noted that the centroid coordinates refer to the average of the spatial positions of all point cloud locations within a voxel unit (e.g., a 3D vector (x... c ,y c ,z c)).

[0071] In one possible real-time approach, a point cloud reflection intensity weighted calculation centroid is introduced to strengthen the weight of high-reflectivity areas (such as metal welds); the centroid coordinates retain the overall distribution trend of the voxel point cloud, providing a reasonable simplification scheme for areas with insignificant features (flat pipe walls).

[0072] Step A404: Iterate through each voxel unit to find whether there are curvature points exceeding a preset threshold; It should be noted that curvature points refer to spatial points with high curvature of local surface in the point cloud (such as corrosion edges, weld protrusions, etc.); the preset threshold refers to the curvature critical value (e.g., set to 0.5). Points exceeding this value are designated as high curvature feature points and are used to identify damage-sensitive areas that require special protection.

[0073] In one possible implementation, the KD number is used to accelerate the calculation of the difference of the neighborhood normal vector, thereby improving the efficiency of pendulum traversal and avoiding the removal of key geometric features during the simplification process.

[0074] Step A405: If there is a curvature point in the voxel unit that exceeds the preset threshold, then replace all point clouds in the current voxel unit with the point with the highest curvature; otherwise, replace all point clouds in the current voxel unit with the point cloud corresponding to the centroid coordinates. It should be noted that replacement refers to deleting the original point cloud geometry within a voxel and retaining only a single representative point. For high-curvature elements, the point with the highest curvature is selected as the representative point to ensure the sharpness of the damage edge; for flat elements, the centroid is sufficient to characterize the geometry.

[0075] In one possible implementation, if there are multiple peak points with similar curvature (difference < 5%), the average of their positions is taken as the representative point, which greatly increases the point cloud density of key areas such as corrosion pits and significantly improves the accuracy of subsequent reconstruction.

[0076] Step A406: Output the sparse point cloud after voxel mesh sampling.

[0077] It should be noted that sparse point cloud refers to a dataset in which the number of point clouds is significantly reduced after sampling by voxel grid.

[0078] In this embodiment, the voxel mesh sampling process, which involves dynamic size division, curvature feature protection, and intelligent point cloud replacement, maximizes the preservation of pipeline damage details while compressing the amount of data. This lays a "high efficiency + high fidelity" guarantee for subsequent processes (Poisson reconstruction, lightweight output), solving the problems of insufficient accuracy, low efficiency, and strong environmental interference in traditional detection technologies.

[0079] Furthermore, referring to Figure 8 The fifth embodiment of the pipeline inner wall inspection method of this application provides a schematic flowchart, based on the above. Figure 8 The embodiment shown further refines the step "performing Poisson reconstruction on the sparse point cloud to generate a triangular mesh model" by including steps A501 to A505: Step A501: Calculate the normal direction of each point cloud and make the normal direction point towards the central axis of the pipe; It should be noted that this step is performed by the data processing terminal to ensure the consistency of the direction of the reconstructed surface and to avoid surface distortion or holes caused by disordered normals in Poisson reconstruction.

[0080] Step A502: Determine the octree depth based on the pipe diameter, and divide the point cloud space into corresponding octrees according to the octree depth; It should be noted that the depth of an octree refers to the number of levels in the recursive partitioning of the three-dimensional space; the greater the depth, the higher the spatial resolution. An octree is a data structure that equally divides a cubic space into 8 sub-cubes and recursively partitions them, aiming to provide an adaptive spatial index for implicit surface computation.

[0081] In one possible implementation, the octree depth is dynamically set according to the pipe diameter as follows: small diameter pipe (≤300mm): depth level 12 (resolution 0.5mm) for fine capture of micro-damage; large diameter pipe (>500mm): depth level 10 (resolution 2mm) to improve computational efficiency.

[0082] Step A503: Construct an implicit surface based on the point cloud coordinates and normal directions stored in the leaf nodes of the octree; It should be noted that leaf nodes refer to the terminal cube units at the bottom level of the octree; implicit surfaces refer to surfaces defined by the mathematical function f(p)=0 (where p is a spatial point), used to transform discrete point clouds into continuous surface representations.

[0083] Step A504: Solve the Poisson equation based on the implicit surface to generate implicit functions; It should be noted that the Poisson equation refers to a partial differential equation. 2 x= *( (x); an implicit function refers to a scalar function x(p), whose isosurface x=τ defines the inner wall surface of the pipe; the gradient field is transformed into a continuous geometric expression through numerical solution. For example, the spatial gradient field is discretized to octagonal leaf nodes; 2 x= *( x) is transformed into a linear system of equations Ax=b, which is then solved iteratively using the conjugate gradient method; the x value at each spatial location is output.

[0084] Step A505: Extract isosurfaces based on the implicit function to generate a triangular mesh model.

[0085] It should be noted that isosurface extraction refers to extracting a surface from an implicit function x with a specific threshold τ (usually 0.5); for example, the implicit function is converted into an explicit triangular mesh using the Marching Cubes algorithm; in dense point cloud regions, the triangular mesh closely fits the original data; in missing point cloud regions, the mesh extends according to the direction inferred from the gradient field.

[0086] In this embodiment, the Poisson reconstruction steps of normal constraint, dynamic subdivision, implicit repair and closed output are used to overcome the accuracy bottleneck, intelligently fill fracture areas such as corrosion pits, and output a high-fidelity mesh model.

[0087] Furthermore, referring to Figure 9 The sixth embodiment of the pipeline inner wall inspection method of this application provides a schematic flowchart, based on the above. Figure 9 The embodiment shown further refines the step "perform density filtering based on the triangular mesh model, and apply the Taubin algorithm to smooth the density-filtered triangular mesh model", including steps A601 to A603: Step A601: Calculate the local density of the point cloud; It should be noted that the local density of point cloud refers to the number of triangular mesh patches within a unit volume of space. It can be used to distinguish between the solid area of ​​the inner wall of a pipe and the area of ​​environmental noise, and to quantify the reliability of the geometric information of different areas in the mesh model.

[0088] Step A602: Filter and retain point clouds with local densities greater than a preset density threshold, and remove point clouds with local densities less than a preset density threshold. It should be noted that the preset density threshold refers to the critical value (unit: faces / cubic centimeter) for distinguishing effective pipe wall structure from noise. This threshold is used to remove false triangular faces caused by environmental interference, improving model realism. For example, the screening strategy includes dynamic threshold setting and execution deletion. In the dynamic threshold setting, the threshold for regular pipes is set to 15 faces / cubic centimeter; the threshold for oil-contaminated pipes is set to 10 faces / cubic centimeter (to avoid false deletion due to weak signal damage). Execution deletion permanently removes triangular faces with local densities below the threshold.

[0089] Step A603: Iteratively adjust the positions of the mesh vertices of the triangular mesh model after filtering the point cloud using the Laplacian operator.

[0090] It should be noted that the Laplacian operator refers to the mathematical tool used in mesh smoothing to calculate the average position of the vertex neighborhood; iterative adjustment refers to suppressing surface noise through multiple displacement optimizations; and the Taubin algorithm eliminates jagged imperfections caused by scan jitter while avoiding over-smoothing that damages geometric features. Specifically, the Taubin two-parameter iteration consists of a shrinkage step (λ): vertex displacement towards the neighborhood centroid (e.g., λ = -0.53) to suppress high-frequency noise; and an expansion step (μ): vertex reverse displacement compensation (e.g., μ = 0.51) to prevent model volume shrinkage; the vertex curvature C is calculated, and if C > 0.5 (eroded edge), the λ parameter value is reduced to protect geometric features.

[0091] In one possible implementation, the number of iterations is fixed at 3 (balancing efficiency and effectiveness), which reduces surface roughness by 40% while the dimensional errors of weld height and corrosion depth are less than 1%.

[0092] In this embodiment, density filtering is used to adapt to complex working conditions such as oil stains and bubbles based on a dynamic threshold mechanism. The Taubin smoothing algorithm is also used to eliminate jagged defects caused by scanning jitter through λ / μ dual-parameter iteration, thereby reducing surface roughness, generating a lightweight model, improving model realism, and solving the problems of insufficient accuracy, poor environmental adaptability, and low efficiency of traditional detection technologies.

[0093] This application also provides a pipe inner wall inspection device, please refer to... Figure 10 The pipe inner wall detection device includes: 10 motion modules, hollow torque motor and Mecanum wheel; 20 control modules, STM32 development board; using PID control algorithm to achieve autonomous movement of the pipeline inner wall detection device; The system consists of 30 detection modules, a lidar, and a data processing terminal; the lidar collects point cloud data, which is then transmitted to the data processing terminal for 3D reconstruction.

[0094] The pipeline inner wall inspection device provided in this application, employing the pipeline inner wall inspection method described in the above embodiments, can solve the technical problems of low stability, poor environmental adaptability, and high cost in pipeline inspection under complex working conditions. Compared with the prior art, the beneficial effects of the pipeline inner wall inspection device provided in this application are the same as those of the pipeline inner wall inspection method provided in the above embodiments, and other technical features of the pipeline inner wall inspection device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0095] This application provides a pipe inner wall inspection device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the pipe inner wall inspection method in the above embodiment 1.

[0096] The following is for reference. Figure 11 The diagram illustrates a structural schematic suitable for implementing the pipe inner wall inspection device of the embodiments of this application. The pipe inner wall inspection device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), vehicle terminals (e.g., vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 11 The pipe inner wall inspection device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0097] like Figure 11 As shown, the pipe inner wall inspection device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory 1002 or a program loaded from a storage device 1003 into a random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the pipe inner wall inspection device. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, a touch screen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows the pipe wall inspection equipment to communicate wirelessly or wiredly with other equipment to exchange data. Although the figure shows pipe wall inspection equipment with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented alternatively.

[0098] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0099] The pipeline inner wall inspection equipment provided in this application, employing the pipeline inner wall inspection method described in the above embodiments, can solve the technical problems of low stability, poor environmental adaptability, and high cost in pipeline inspection under complex working conditions. Compared with the prior art, the beneficial effects of the pipeline inner wall inspection equipment provided in this application are the same as those of the pipeline inner wall inspection method provided in the above embodiments, and other technical features of this pipeline inner wall inspection equipment are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0100] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0101] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0102] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the pipe inner wall detection method in the above embodiments.

[0103] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0104] The aforementioned computer-readable storage medium may be included in the pipeline inner wall inspection equipment; or it may exist independently and not assembled into the pipeline inner wall inspection equipment.

[0105] The aforementioned computer-readable storage medium carries one or more programs. When the aforementioned one or more programs are executed by the pipeline inner wall detection device, the pipeline inner wall detection device: constructs a motion module based on a hollow torque motor and a Mecanum wheel; drives the laser radar axially through the motion module to collect point cloud data of the pipeline inner wall in real time; processes the point cloud data to generate a three-dimensional model of the pipeline inner wall.

[0106] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0107] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0108] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0109] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described pipeline inner wall inspection method. This solves the technical problems of low stability, poor environmental adaptability, and high cost in pipeline inspection under complex working conditions. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the pipeline inner wall inspection method provided in the above embodiments, and will not be repeated here.

[0110] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the pipe inner wall detection method described above.

[0111] The computer program product provided in this application can solve the technical problems of low stability, poor environmental adaptability, and high cost in pipeline inspection under complex working conditions. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the pipeline inner wall inspection method provided in the above embodiments, and will not be repeated here.

[0112] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for inspecting the inner wall of a pipe, characterized in that, The method for detecting the inner wall of the pipeline includes: A motion module is constructed based on a hollow torque motor and a Mecanum wheel; The motion module drives the lidar axially to collect point cloud data of the inner wall of the pipe in real time. The point cloud data is processed to generate a three-dimensional model of the pipe's inner wall.

2. The pipe inner wall inspection method as described in claim 1, characterized in that, The steps for constructing the motion module based on the hollow torque motor and Mecanum wheel include: The rollers of the adjacent Mecanum wheel are installed with the pipe inner wall inspection device; The output shaft of the hollow torque motor is fixedly connected to the center of the inner hub of the Mecanum wheel; The hollow torque motor body is fixed to the main frame of the pipeline inner wall detection device by a rigid connecting rod; The Mecanum wheels are symmetrically distributed on both sides of the device and kept synchronized by rigid connecting rods; The stator of the hollow torque motor is fixed on a rigid connecting rod support, and the rotor of the hollow torque motor drives the wheel of the Mecanum wheel to construct the motion module of the pipeline internal detection device.

3. The pipe inner wall inspection method as described in claim 2, characterized in that, Prior to the step of axially driving the lidar via the motion module to acquire point cloud data of the inner wall of the pipe in real time, the following steps are included: Obtain the pipe diameter; If the pipe diameter is smaller than the preset diameter, the lidar is placed at the front end of the pipe inner wall detection device, offset from the central axis of the pipe. If the pipe diameter is greater than or equal to the preset diameter, the lidar is placed at the position of the central axis of the pipe at the front end of the pipe inner wall detection device.

4. The pipe inner wall inspection method as described in claim 3, characterized in that, The step of using the motion module to axially drive the lidar to acquire point cloud data of the pipe's inner wall in real time includes: The position and forward speed of the pipe inner wall detection device are adjusted in real time by the hollow torque motor so that the lidar maintains a preset scanning distance from the pipe inner wall. The laser beam is emitted by the lidar and the reflected signal from the inner wall of the pipe is received. The reflected signal is transmitted to the terminal in real time via a cable passing through the shaft of the hollow torque motor.

5. The pipe inner wall inspection method as described in claim 4, characterized in that, The step of processing the point cloud data to generate a three-dimensional model of the pipe inner wall includes: The point cloud data is preprocessed; The preprocessed point cloud data is sampled using a voxel grid to obtain a sparse point cloud. The sparse point cloud is reconstructed using Poisson reconstruction to generate a triangular mesh model; Density filtering is applied to the triangular mesh model, and the Taubin algorithm is used to smooth the density-filtered triangular mesh model. The smoothed triangular mesh model is simplified by a second error to obtain a three-dimensional model of the pipe inner wall.

6. The pipe inner wall inspection method as described in claim 5, characterized in that, The step of sampling the preprocessed point cloud data into a voxel grid to obtain a sparse point cloud includes: The corresponding voxel dimensions are set based on the pipe diameter; The preprocessed point cloud data is divided into a uniform voxel grid based on the voxel size; Calculate the centroid coordinates of the point cloud within each voxel cell of the voxel mesh; Iterate through each voxel unit to check if there are any curvature points exceeding a preset threshold; If there is a curvature point in the voxel unit that exceeds a preset threshold, then the point with the highest curvature is used to replace all point clouds in the current voxel unit; otherwise, the point cloud corresponding to the centroid coordinates is used to replace all point clouds in the current voxel unit. Output sparse point cloud after voxel mesh sampling.

7. The pipe inner wall inspection method as described in claim 6, characterized in that, The step of performing Poisson reconstruction on the sparse point cloud to generate a triangular mesh model includes: Calculate the normal direction of each point cloud and make the normal direction point towards the central axis of the pipe; The octree depth is determined based on the pipe diameter, and the point cloud space is divided into corresponding octrees according to the octree depth. An implicit surface is constructed based on the point cloud coordinates and normal directions stored in the leaf nodes of the octree. Solving the Poisson equation based on the implicit surface generates implicit functions; Isosurfaces are extracted based on the implicit function to generate a triangular mesh model.

8. The pipe inner wall inspection method as described in claim 7, characterized in that, The steps of performing density filtering based on the triangular mesh model and smoothing the density-filtered triangular mesh model using the Taubin algorithm include: Calculate the local density of the point cloud; Point clouds with local densities greater than a preset density threshold are selected and retained, while point clouds with local densities less than the preset density threshold are removed. The positions of the grid vertices of the triangular mesh model after filtering the point cloud are adjusted iteratively using the Laplacian operator.

9. A pipe inner wall inspection device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the pipe inner wall inspection method as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the pipe inner wall detection method as described in any one of claims 1 to 8.

Images