Tunnel apparent disease integrated visual acquisition module
By designing an integrated visual acquisition module for tunnel surface defects, using a high-resolution linear array camera and laser supplementary lighting, combined with core control and protective heat dissipation measures, the problems of low integration and poor environmental adaptability of tunnel detection equipment have been solved, realizing high-precision image acquisition and rapid defect detection across the entire cross section.
Patent Information
- Application Number
- CN202410144837.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-02-01
- Publication Date
- 2026-08-25
- Estimated Expiration
- 2044-02-01
AI Technical Summary
Existing tunnel surface defect detection equipment suffers from low integration, low modularity, complex vision system debugging, low acquisition accuracy, and poor environmental adaptability, making it difficult to achieve full coverage of tunnel cross sections and high-precision image acquisition.
Design an integrated visual acquisition module for tunnel surface defects, including a mounting backplate and shell, a machine vision system, an integrated control system, a heat dissipation mechanism, a protection mechanism, and a cable management mechanism. Employ a high-resolution line scan camera, laser illumination, and an adjustment mechanism to achieve full-section coverage and high-precision image acquisition. Power supply, signal distribution, and laser trigger control are handled by a core control motherboard, combined with protective and heat dissipation measures.
It achieves full-section digital imaging of tunnels, enabling rapid, non-contact detection of tunnel defects. It boasts high image acquisition continuity and clarity, adapts to different tunnel profiles, offers high protection levels, has good environmental adaptability, and its modular integration facilitates installation and disassembly.
Smart Images

Figure CN117929388B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to an integrated visual acquisition module for tunnel surface defects, which is applied to tunnel structure inspection. Background Technology
[0002] With the continuous development and improvement of my country's infrastructure, tunnels for highways, railways, and subways are showing a booming trend, with a significant increase in both number and scale. As a large number of tunnels are put into operation, they are affected by various factors such as external loads, ground pressure, and concrete aging, resulting in varying degrees of deformation, cracks, and water seepage. If these defects are not detected and treated in a timely manner, they will endanger the operational safety of the tunnel structure.
[0003] Tunnels are typical long-distance engineering projects, and the number of long and extra-long operational tunnels in China is constantly increasing. Traditional manual inspection is arduous, inefficient, costly, subjective, and prone to errors, and it also has a significant impact on normal traffic operations. It can no longer meet the industry's inspection needs, and there is an urgent need for automated inspection equipment to improve tunnel inspection efficiency.
[0004] With the rapid development of technologies such as machine vision, laser scanning, and artificial intelligence, numerous techniques and methods for detecting surface defects in tunnels have emerged. Patent CN114460090A discloses a tunnel lining surface defect detection system based on an industrial camera, using an industrial camera and deep learning algorithms as the main detection methods, reducing manual labor intensity and human reading errors. Patent CN217385264U discloses a high-definition imaging detection device for railway tunnels. By setting up a specially structured device framework and rationally arranging imaging units, the integration level of the detection device is improved. It uses a high-frequency industrial camera and laser light source to achieve automated detection of surface cracks in tunnels. Patent CN115326819A discloses a tunnel structure surface defect detection device that obtains tunnel structure point cloud data through a laser scanner and monitoring camera, establishes a three-dimensional model of the tunnel and point cloud grayscale and depth image maps to identify and analyze the features of tunnel surface defects. Most existing tunnel surface defect detection technologies and methods cannot achieve full coverage of the tunnel cross section, have low image acquisition accuracy, and are difficult to meet the needs of tunnel crack detection, especially the requirement of detecting durability cracks in subway tunnels with a width of 0.1mm. In addition, the non-fully enclosed design results in low protection level of the mechanism and low adaptability to the working environment.
[0005] In summary, there is an urgent need for an integrated visual acquisition module for tunnel surface defects that supports full coverage of tunnel cross sections, high image acquisition accuracy, high protection level, and good environmental adaptability. Summary of the Invention
[0006] The purpose of this invention is to design an integrated visual acquisition module for tunnel surface defects, mainly to solve the technical problems of low integration, low modularity, complex debugging of vision systems, low acquisition accuracy, and poor environmental adaptability of existing detection equipment.
[0007] The objective of this invention can be achieved through the following technical solution: an integrated visual acquisition module for tunnel surface defects, comprising a mounting backplate and housing, a machine vision system, an integrated control system, a heat dissipation mechanism, a protective mechanism, and a cable management mechanism; the machine vision system is encapsulated inside the mounting backplate and housing, the heat dissipation mechanism is installed on the back of the mounting backplate, the protective mechanism is placed between the components that make up the visual acquisition module, and the integrated control system and the cable management mechanism are installed at the center of the front of the mounting backplate.
[0008] The mounting backplate and housing are polygonal in structure, with an aesthetically pleasing and compact design. On the front of the mounting backplate, multiple mounting positions for the integrated camera / laser assembly, laser, and laser driver circuit board are located on each side of the polygon. The central area of the mounting backplate houses the core control motherboard, fiber optic bracket, and waterproof connector mounting positions. Two handles are located on the back of the mounting backplate, allowing for easy transport of the visual acquisition module after it has been detached from a specific acquisition platform. Mounting fasteners are located around the central cable hole area for mounting to the mounting bracket. A pre-drilled square hole of the corresponding size is provided on the back of the mounting backplate for installing a heat dissipation device. Each side of the polygonal housing is fitted with a camera / laser window, which is attached to the inside of the housing using optical glass via a window clamping plate. The central area of the mounting backplate has multiple power / signal cable holes, and multiple cooling fan cable holes are distributed at the square holes of the heat dissipation devices along the edges of the mounting backplate. Furthermore, the visual acquisition module mounting backplate is installed and fixed to the mounting bracket using a sliding groove installation method. Flexible shock-absorbing pads are set on the contact surface between the mounting fasteners of the visual acquisition module and the sliding groove of the mounting bracket to ensure the shock absorption and stability of the visual acquisition module. Limit pins are set in the sliding groove to achieve self-locking of the mounting fasteners and ensure that the visual acquisition module will not shake.
[0009] The machine vision system consists of an imaging system, a supplementary lighting system, and an adjustment mechanism. The imaging system comprises multiple high-resolution line scan cameras and corresponding camera lenses. After the line scan cameras and camera lenses are assembled, they are installed in the camera mounting position of the camera / laser integrated assembly and rigidly connected by bolts. The supplementary lighting system consists of a laser and a laser lens. The laser is used for electro-optical conversion to provide a laser light source, and the laser lens shapes the laser point light source into a line light source. The adjustment mechanism drives the laser lens to achieve rotation and deflection adjustment. Furthermore, since the object of detection is a tunnel, the data collected by the machine vision system for each imaging is a cross-section of one ring of the tunnel. In order to ensure that the machine vision system can adapt to tunnels with different cross-sections, the machine vision system needs to be installed at the center point of the tunnel cross-section, and the corresponding equipment of the machine vision system is distributed around the center point of the vision acquisition module to ensure that the amount of change that each set of camera / laser integrated components adapts to is consistent when the tunnel outline changes. Furthermore, multiple line scan cameras are paired with corresponding large-area line scan lenses, and the shooting ranges of adjacent cameras overlap, which can achieve full cross-sectional coverage of various contour tunnels. Furthermore, the laser emitted by the laser is connected to the laser lens through an optical fiber. The cylindrical mirror inside the laser lens shapes the laser point source into a linear source, and the prism emits it out at the corresponding fan angle, forming a linear uniform light band on the shooting surface. The light band covers the shooting area of the camera, providing high-intensity supplementary lighting for the camera to shoot. Furthermore, the adjustment mechanism is a linkage mechanism of adjusting worm gear and meshing gear. Rotating the adjusting worm gear drives the meshing gear to move, achieving stable step-by-step adjustment of the adjustment mechanism. The adjusting worm gear is engraved with scale marks, which can precisely control the adjustment step size of the adjusting worm gear, with an accuracy of up to 0.0428°. The adjustment mechanism is divided into two degrees of freedom: lens rotation adjustment and lens deflection adjustment. The lens rotation adjustment worm gear is located below the laser lens mounting position, and there are four fixing bolts on the back of the lens. After the rotation adjustment is completed, the bolts are tightened. The lens deflection adjustment worm gear is located above the laser lens mounting position, and there is a fastening bolt on the front of the lens for tightening after the deflection adjustment is completed. The camera lens is fixed and serves as a positioning reference. The fastening part of the adjustment mechanism is separate from the adjustment part and does not interfere with each other, so that the adjustment mechanism will not be offset when the bolts are tightened.
[0010] The integrated control system consists of a core control motherboard and multiple laser driver circuit boards. The core control motherboard can perform power distribution, signal distribution, laser trigger control, and imaging system signal acquisition control for the entire machine vision module. The laser driver circuit boards control the trigger signal, power adjustment, and temperature protection for each laser. Furthermore, to ensure the high integration and low coupling of the vision acquisition module, the power input is a single-channel input, which is connected to the core control motherboard via an external power cable from the vision acquisition module. The core control motherboard regulates, adjusts, and distributes the input power, and outputs it to various devices in the imaging system, the supplementary lighting system, and the heat dissipation mechanism. Furthermore, the core control motherboard integrates an encoder pulse distributor. Based on the pulse signal triggered by the encoder at a fixed distance, a multi-channel time synchronization and nanosecond-level pulse signal parallel excitation system for a high-precision time and frequency reference source is constructed to realize high-speed synchronous imaging and transmission of the linear array camera array. Under high-speed dynamic detection conditions, the maximum misalignment of the tunnel longitudinal imaging is less than 1mm. Using the encoder pulse triggering method, the loading platform of the detection vision acquisition system collects fixed data when it moves forward and does not collect data when it stops or moves backward. Furthermore, the core control motherboard integrates a laser trigger signal distributor. The operation of the laser is controlled by the laser trigger signal distributor, which can precisely control the lighting time and duration of the laser. When multiple laser driver circuit boards receive the laser trigger signal distributed by the core control motherboard, they control the corresponding laser to perform electro-optic conversion and output a laser light source.
[0011] The heat dissipation mechanism consists of multiple sets of heat dissipation fins and corresponding external cooling fans, which are symmetrically fixed to each side of the polygonal mounting back plate with bolts. When the heat dissipation fins are installed at the bottom, they pass directly into the interior of the vision acquisition module through the reserved square holes. The exhaust surface of the cooling fan faces the heat dissipation fins for cooling. Furthermore, the main heat-generating component inside the vision acquisition module—the laser—is directly fixed to the bottom of the heat sink fins. Thermally conductive adhesive is applied between its mounting surface and the bottom mounting surface of the heat sink fins. This ensures that the heat can be effectively conducted to the heat sink fins on the outside of the vision acquisition module while maintaining a highly modular and integrated package.
[0012] The protective mechanism primarily provides dustproof, waterproof, and anti-disassembly protection for the vision acquisition module. It includes a dustproof and waterproof mechanism and an anti-disassembly mechanism. The dustproof and waterproof mechanism consists of: a waterproof strip affixed to the connection between the vision acquisition module's outer shell and the mounting backplate; a waterproof strip between the heat sink fins and the mounting backplate during installation; waterproof adhesive used to secure the edges of the camera / laser window's optical glass; waterproof connector bases used for each cable hole on the mounting backplate, with corresponding cables using waterproof aviation plugs, achieving IP67 dustproof and waterproof ratings, and the entire vision acquisition module achieving IP65 dustproof and waterproof ratings. The anti-disassembly mechanism comprises a rotating baffle, a spring, and a pin. The pin is fixed to the mounting backplate; rotating the baffle around the pin compresses the outer shell, preventing non-technical personnel from removing it, thus preventing the outer shell from being removed even after removing the mounting bolts. This achieves anti-disassembly protection, safeguarding the internal security of the vision acquisition module and maintaining technical confidentiality.
[0013] The cable management mechanism is a ring-shaped fiber optic bracket in the central area of the front of the mounting backplate. The fiber optic bracket consists of two concentric rings of different diameters, with a connecting rod provided between the two rings at certain angles. The optical fiber is fixed on the fiber optic bracket in a coiled manner.
[0014] Based on the visual acquisition module, the tunnel structure appearance images can be acquired quickly, intelligently, and digitally, enabling the detection of surface defects such as cracks, water leakage, spalling, and segment damage in the tunnel structure, as well as defects in ancillary facilities such as signaling mechanisms, optical cables, and communication cables, through methods such as image deep learning, subpixel edge segmentation, and experimental correction of image geometric feature values.
[0015] The beneficial effects of this invention are: (1) This invention can realize digital imaging of the entire tunnel section and complete the rapid, non-contact detection of surface defects such as tunnel lining cracks, water leakage, and peeling, as well as defects of ancillary facilities such as signal mechanisms, optical cables, and communication cables. (2) The present invention uses a line scan camera, paired with a laser for supplementary lighting, and is equipped with an adjustment mechanism in which the fastening part and the adjustment part do not interfere with each other, which can realize the stable step-by-step adjustment of the visual acquisition module, ensuring the continuity, integrity and clarity of image acquisition. The layout of the line scan camera takes into account various tunnels and is adapted to different tunnel profiles. (3) The present invention realizes a highly modular integrated packaging of the visual acquisition module. The visual acquisition module can be integrated into different detection platforms as a single detection unit. Users do not need to care about the specific operation, assembly and adjustment process of each imaging component inside. The anti-disassembly mechanism configured on the shell can effectively prevent non-professional technicians from disassembling it, further ensuring the internal security of the visual acquisition module and the confidentiality of the implementation technology. (4) The overall protection level of this invention is IP65. It achieves good heat dissipation performance while being highly modularly packaged. It is easy to disassemble and has good environmental adaptability. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the structural topology of the present invention; Figure 2 This is a layout diagram of the present invention; Figure 3 This is a schematic diagram of the internal cavity structure of the present invention; Figure 4 This is a schematic diagram of the back structure of the present invention; Figure 5 This is a schematic diagram of the camera / laser integrated component of the present invention.
[0017] In the diagram: 1-Outer shell; 2-Fiber optic bracket; 3-Optical glass; 4-Window pressure plate; 5-Rotating baffle; 6-Spring; 7-Pin; 8-Cooling fan; 9-Heat dissipation fins; 10-Laser; 11-Core control motherboard; 12-Camera / laser integrated assembly; 13-Laser driver circuit board; 14-Mounting bracket; 15-Network connector; 16-Handle; 17-Flexible shock-absorbing pad; 18-Mounting backplate; 19-Waterproof connector; 20-Waterproof vent valve; 21-Mounting fastener; 22-Slide groove; 23-Tightening bolt; 24-Limit pin; 25-Signal connector; 26-Power connector; 27-Laser lens; 28-Camera lens; 29-Camera / laser integrated assembly mounting base; 30-Line scan camera; 31-Deflection adjustment mechanism; 32-Laser fixing adjustment ring; 33-Rotation adjustment mechanism; 34-Tunnel. Detailed Implementation
[0018] The present invention will now be described in detail with reference to the accompanying drawings.
[0019] like Figures 1-5 As shown, the layout of the visual acquisition module should consider symmetry and aesthetics, such as a quadrilateral or hexagonal layout, which is symmetrical in both the horizontal and vertical directions. This is not only aesthetically pleasing but also helps to ensure that the accuracy of the images acquired by each component imaging system is consistent. Taking a circular shield tunnel 34 with an inner diameter of 5.5m as an example, the visual acquisition module is designed with a hexagonal symmetrical layout according to the design acquisition accuracy. An integrated visual acquisition module for tunnel surface defects consists of a mounting backplate and shell, a machine vision system, an integrated control system, a heat dissipation mechanism, a protective mechanism, and a cable management mechanism. The machine vision system is encapsulated inside the mounting backplate and shell, the heat dissipation mechanism is installed on the back of the backplate, the protective mechanism is placed between the components of the visual acquisition module, and the integrated control system and cable management mechanism are installed at the center of the front of the backplate.
[0020] The mounting backplate 18 and the outer shell 1 adopt a hexagonal symmetrical design, which is aesthetically pleasing and compact in structure. On the front of the mounting backplate 18, six sets of mounting positions for camera / laser integrated components 12, laser 10, and laser drive circuit board 13 are symmetrically arranged on each side of the hexagon. The central area of the backplate has mounting positions for the core control motherboard 11, fiber optic bracket 2, and waterproof connector 19. The back of the mounting backplate 18 is designed with two handles 16. After the visual acquisition module is detached from a platform, it can be carried and transported by hand through the handles 16. The area around the central cable hole is provided with mounting fasteners 21 for mounting to the mounting bracket 14. The back of the laser 10 mounting position has a pre-drilled square hole of the corresponding size for the installation of heat dissipation fins 9. Each side of the hexagonal outer shell 1 is equipped with a camera / laser window, which is attached to the inside of the outer shell 1 by optical glass 3 through the window pressure plate 4. The central area of the mounting backplate 18 has 9 power / signal cable holes, and the edge heat dissipation fins 9 of the mounting backplate 18 have 6 cooling fan 8 cable holes symmetrically arranged. Furthermore, the visual acquisition module is installed and fixed to the mounting bracket 14 via the mounting back plate 18 and the slide groove 22. The mounting back plate 18 is equipped with a waterproof and breathable valve 20. Flexible shock-absorbing pads 17 are arranged on the contact surface between the mounting fastener 21 of the visual acquisition module and the slide groove 22 of the mounting bracket 14 to ensure the shock absorption and stability of the visual acquisition module. After the mounting fastener 21 of the visual acquisition module is slid into the slide groove 22 of the mounting bracket 14 and reaches the designated position, the limit pin 24 in the slide groove 22 will automatically lock the mounting fastener 21 to prevent it from sliding out. Tightening the top bolt 23 can ensure that the visual acquisition module will not shake.
[0021] The machine vision system consists of an imaging system, a supplementary lighting system, and an adjustment mechanism. The imaging system comprises six high-resolution line scan cameras 30 and six corresponding camera lenses 28. After the line scan cameras 30 and camera lenses 28 are assembled, they are installed on the camera mounting position of the camera / laser integrated component mounting base 29 and rigidly connected by bolts. The supplementary lighting system consists of a laser 10 and a laser lens 27. The laser 10 is used for electro-optical conversion to provide a laser light source. The laser lens 27 shapes and outputs the laser light source provided by the laser 10. The laser lens 27 is fixed on the laser fixing adjustment ring 32 and is integrally installed with the laser fixing adjustment ring 32 onto the adjustment mechanism. The adjustment mechanism drives the laser lens 27 to achieve rotation and deflection adjustment. Furthermore, the vision acquisition module is first installed on the center line of the tunnel track. The height adjustment device of the mounting bracket 14 is used to adjust the vision acquisition module to the height of the tunnel center point to ensure that the center point of the vision acquisition module is located at the center point of the tunnel. Each camera / laser integrated component 12 of the machine vision system is installed symmetrically around the center point of the vision acquisition module. When the track contour changes, the amount of change that each set of camera / laser integrated components 12 adapts to is consistent, and the accuracy of the acquired image remains consistent. Furthermore, the six line scan cameras 30 are equipped with corresponding large target surface line scanning lenses, and the overlap of the shooting range of each adjacent camera is 130mm-230mm, which can achieve full cross-sectional coverage of various contour tunnels; the line scan cameras 30, equipped with line laser light source 7, can meet the detection speed of 20km / h, and the image acquisition accuracy can be better than 0.25mm / Pixel. Furthermore, the laser emitted by the laser 10 is connected to the laser lens 27 via an optical fiber. The cylindrical mirror inside the laser lens 27 shapes the laser point source into a line source, and the prism emits the laser at a fixed fan angle, forming a linear uniform light band with a width of no more than 20mm on the shooting surface. The light band covers the shooting area of the camera, providing high-intensity supplementary lighting for the shooting of the line scan camera 30. Furthermore, the adjustment mechanism is a linkage mechanism of adjusting worm gear and meshing gear. Rotating the adjusting worm gear drives the meshing gear to move, achieving stable step-by-step adjustment of the adjustment mechanism. The adjusting worm gear is engraved with scale marks, which can precisely control the adjustment step size of the adjusting worm gear, with an accuracy of up to 0.0428°. The adjustment mechanism is divided into two degrees of freedom: lens rotation adjustment and lens deflection adjustment. The lens rotation adjustment mechanism 33 is located below the mounting position of the laser lens 27, and there are 4 fixing bolts on the back of the lens. After the rotation adjustment is completed, the bolts are tightened. The lens deflection adjustment mechanism 31 is located above the mounting position of the laser lens 27, and there is 1 fixing bolt on the front of the lens, which is used to tighten after the deflection adjustment is completed. The camera lens 28 is fixed and serves as a positioning reference. The tightening part and the adjustment part in the adjustment mechanism are separated and do not interfere with each other, so that the adjustment mechanism will not be offset when the bolt is tightened.
[0022] The integrated control system consists of a core control motherboard 11 and six laser drive circuit boards 13. The core control motherboard 11 is located in the central area of the mounting backplate 18 and can perform power distribution, signal distribution, laser 10 trigger control, and imaging system signal acquisition control for the entire machine vision module. The laser drive circuit boards 13 are located on each side of the hexagonal mounting backplate 18 and respectively perform trigger signal, power adjustment, and temperature protection control for each laser 10. Furthermore, to ensure the high integration and low coupling of the vision acquisition module, the power supply line is a single input, which is connected to the core control motherboard 11 by the external power cable of the vision acquisition module. After the core control motherboard 11 stabilizes, regulates and distributes the input power, the power output line is 18, which are respectively output to 6 line scan cameras 30, 6 laser drive circuit boards 13 and 6 cooling fans 8. Furthermore, the core control motherboard 11 integrates an encoder pulse distributor. Based on the pulse signal triggered by the encoder at a fixed distance, a multi-channel time synchronization and nanosecond-level pulse signal parallel excitation system for a high-precision time and frequency reference source is constructed, realizing high-speed synchronous imaging and transmission of the 30-array linear scan camera. At a speed of 20km / h, the maximum misalignment of the longitudinal imaging in the tunnel is less than 1mm. Using the encoder pulse triggering method, fixed data is collected when the loading platform of the vision acquisition module moves forward, and no data is collected when it stops or moves backward. Furthermore, the core control motherboard 11 integrates a laser trigger signal distributor. The operation of the laser 10 is controlled by the laser trigger signal distributor, which can precisely control the lighting time and duration of the laser 10. When the six laser driver circuit boards 13 receive the laser trigger signal distributed by the core control motherboard 11, they control the corresponding laser 10 to perform electro-optic conversion and output a laser light source.
[0023] The heat dissipation mechanism consists of 6 sets of heat dissipation fins 9 and corresponding external cooling fans 8, which are symmetrically fixed to each side of the hexagonal mounting backplate 18 with bolts. The bottom of the heat dissipation fins 9 passes directly into the interior of the vision acquisition module through the reserved square holes of the mounting backplate 18. The exhaust surface of the cooling fan 8 faces the heat dissipation fins 9, and the heat dissipation fins 9 are cooled by the principle of air cooling. Furthermore, the main heat-generating component inside the vision acquisition module—the laser 10—is directly fixed to the bottom of the heat sink 9. Thermally conductive adhesive is applied between its mounting surface and the bottom mounting surface of the heat sink 9, allowing heat to be effectively transferred to the heat sink 9 outside the vision module and further cooled by the cooling fan 8.
[0024] The protective mechanism provides dustproof, waterproof, and anti-disassembly protection for the visual acquisition module. The protective mechanism includes a dustproof and waterproof mechanism and an anti-disassembly mechanism. The dustproof and waterproof mechanism is as follows: the outer shell 1 of the visual acquisition module and the mounting backplate 18 are sealed together with waterproof adhesive strips; when the heat sink fins 9 are installed, waterproof adhesive strips are placed between them and the mounting backplate 18; the edges of the optical glass 3 of the camera / laser window are glued with waterproof adhesive, providing both fixation and protection for the window area; each cable hole on the mounting backplate 18 is fixed with a waterproof connector 19 base, and the corresponding cables are equipped with waterproof plugs. The cable connections achieve IP67-level dustproof and waterproof protection, and the entire visual acquisition module achieves IP65-level dustproof and waterproof protection. Furthermore, the anti-tamper mechanism consists of a rotating baffle 5, a spring 6, and a pin 7, which is fixed to the mounting back plate 18 via the pin 7. After the anti-tamper device is installed, in its natural state, the spring 6 ensures that the rotating baffle 5 does not affect the installation of other components. After the other components are installed, the rotating baffle 5 is rotated around the pin 7 and restricted, pressing the outer shell 1 to complete the anti-tamper setting. Only when the technician removes the anti-tamper setting, the spring 6 stores energy to restore the natural state, completing the disassembly.
[0025] The cable management mechanism is a ring-shaped fiber optic bracket 2 in the central area of the front of the mounting backplate 18. The cable is fixed to the fiber optic bracket 2 in a coiled manner. Six connecting brackets are symmetrically arranged between the two concentric rings of the fiber optic bracket 2. The fiber optic cable is coiled into a ring and placed on the fiber optic bracket 2, and fixed with cable ties at each connecting bracket. It is fixed in six places in the circumferential direction, and when it reaches the designated exit position, it is connected to the corresponding component in six ways.
[0026] In practical applications, the visual acquisition module uses a non-contact method to detect surface defects in tunnels. Specifically: (1) The power supply is input to the vision acquisition module through the power connector 26, and after being distributed by the core control motherboard 11, it supplies power to all devices inside the machine vision module, and all devices enter the power-on working state. (2) The main control system sends a start operation command to the image acquisition system through the network port connector 15 and the signal connector 25. At the same time, the laser trigger pulse signal is input to the vision acquisition module. The laser trigger pulse signal is then input to the six laser drive circuit boards 13 through the laser trigger pulse signal distributor. After receiving the signal, the laser drive circuit board 13 controls the laser 10 to output laser. The laser enters the laser lens 27 through the optical fiber and is shaped to form a linear light strip, which provides a supplementary light source for the shooting area of the line scan camera 30. (3) After receiving the start operation command, the image acquisition system controls the six line scan cameras 30 to enter the image acquisition state. The line scan cameras 30 need to receive a trigger signal before they can acquire images according to the signal. (4) The pulse signal of the encoder is input to the vision acquisition module. The pulse signal is input to the six line scan cameras 30 through the encoder pulse distributor. The line scan cameras 30 start to acquire images upon receiving the pulse signal.
[0027] Based on the tunnel structure appearance images acquired by the module, and through methods such as image deep learning, subpixel edge segmentation, and experimental correction of image geometric feature values, it is possible to achieve rapid, non-contact detection of apparent defects such as tunnel lining cracks, water leakage, and spalling, as well as defects of ancillary facilities such as signal mechanisms, optical cables, and communication cables. Users can judge the condition of the tunnel based on the analysis results and use this as a basis to guide tunnel maintenance.
[0028] The preferred embodiments of the present invention have been described above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection claimed by the claims of the present invention.
Claims
1. An integrated visual acquisition module for tunnel surface defects, characterized in that: It consists of a mounting backplate and housing, a machine vision system, an integrated control system, a heat dissipation mechanism, a protective mechanism, and a cable management mechanism. The machine vision system is encapsulated inside the mounting backplate and housing, the heat dissipation mechanism is installed on the back of the mounting backplate, the protective mechanism is placed between the components that make up the vision acquisition module, and the integrated control system and the cable management mechanism are installed in the center of the front of the mounting backplate. The mounting backplate and housing are polygonal structures; on the front of the mounting backplate, a set of camera / laser integrated component mounting positions, laser mounting positions, and laser driver circuit board mounting positions are provided on each side of the polygon. The front center area of the mounting backplate has a core control motherboard mounting position, a fiber optic bracket, and a waterproof connector mounting position; each side of the polygonal shell is equipped with a camera / laser window, which is attached to the inside of the shell with optical glass through a window clamping plate; the center area of the mounting backplate has multiple power / signal cable through holes, and the square holes of the heat dissipation device at the edge of the mounting backplate have multiple fan cable through holes. The machine vision system comprises an imaging system, a supplementary lighting system, and an adjustment mechanism. The imaging system consists of several line-scan cameras and their corresponding lenses. The line-scan cameras are mounted on the camera mounting position of the camera / laser integrated assembly and rigidly connected with bolts. The supplementary lighting system consists of a laser and a laser lens. The laser performs electro-optical conversion, providing a laser light source. The laser lens is fixed to a laser fixing and adjustment ring, linearly converting the laser light source provided by the laser into a linear light band output from a point light source. The adjustment mechanism acts on the laser lens, driving it to achieve rotation and deflection adjustments. The laser lens, its fixing and adjustment mechanism, the line-scan cameras, and the camera lens are mounted on the backplate mounting position as part of the camera / laser integrated assembly. The adjustment mechanism is a linkage mechanism of an adjusting worm and meshing teeth. Rotating the adjusting worm drives the meshing teeth to move, achieving stable, step-by-step adjustment of the adjustment mechanism. The adjusting worm is engraved with scale markings. The lens rotation adjustment mechanism is located below the laser lens and is fixed after adjustment using multiple fixing bolts on the back. The lens deflection adjustment mechanism is located above the laser lens and is fixed after adjustment using a fixing bolt on the front. The camera lens is fixed in place, and the fastening part and the adjustment part in the adjustment mechanism are separate; The integrated control system consists of a core control motherboard and multiple laser driver circuit boards. The core control motherboard can perform power distribution, signal distribution, laser trigger control, and imaging system signal acquisition control for the entire machine vision module. The laser driver circuit boards control the trigger signal, power adjustment, and temperature protection for each laser. The entire vision acquisition module is powered by a single external power cable. The encoder pulse distributor is integrated into the core control motherboard, enabling high-speed synchronous imaging and transmission of the line scan camera array; it uses encoder pulse triggering to detect when the loading platform of the vision acquisition system moves forward and collects fixed data, and does not collect data when it stops or moves backward. The heat dissipation mechanism consists of multiple sets of heat dissipation fins and corresponding external cooling fans, which are fixed to each side of the polygonal mounting backplate with bolts; the bottom of the heat dissipation fins penetrates into the visual acquisition module through the pre-reserved square holes in the mounting backplate, and the exhaust surface of the cooling fan faces the heat dissipation fins. The protective mechanism includes a dustproof and waterproof mechanism and an anti-disassembly mechanism. The dustproof and waterproof mechanism is as follows: a waterproof strip is pasted at the connection between the outer shell of the vision acquisition module and the mounting back plate; a waterproof strip is placed between the heat dissipation fins and the mounting back plate; the edge of the optical glass of the camera / laser window is fixed and protected with waterproof adhesive; each cable hole of the mounting back plate is fixed with a waterproof connector base, and the cable is connected to a waterproof plug. The anti-disassembly mechanism consists of a rotating baffle, a spring, and a pin. The pin is fixed to the mounting back plate, and the rotating baffle is moved to rotate around the pin and press the outer shell tightly. The cable management mechanism is a ring-shaped fiber optic bracket in the central area of the front of the mounting backplate. The ring-shaped fiber optic bracket consists of two concentric rings of different diameters. A connecting rod is provided between the two rings at certain angles. The optical fiber is fixed on the fiber optic bracket in a coiled manner, and the cable is fixed on the fiber optic bracket in a coiled manner. When they reach the designated outlet position, they are connected to the corresponding devices respectively.
2. The integrated visual acquisition module for tunnel surface defects according to claim 1, characterized in that, The mounting backplate has a pre-drilled square hole of the corresponding size for installing a heat dissipation device; the backplate has two handles and mounting fasteners are designed around the central wire hole area for mounting to the mounting bracket.
3. The integrated visual acquisition module for tunnel surface defects according to claim 1, characterized in that, The visual acquisition module is fixed by a sliding groove, and the contact surface between the mounting fastener of the visual acquisition module and the sliding groove of the mounting bracket is provided with a flexible shock-absorbing pad.
4. The integrated visual acquisition module for tunnel surface defects according to claim 1, characterized in that, The visual acquisition module is installed at the center point of the tunnel cross section. The camera / laser integrated component, laser, camera / laser window, and heat dissipation mechanism are distributed around the center point of the visual acquisition module, and each component is distributed in a circumferential manner around the concentric point of the visual acquisition module.
5. The integrated visual acquisition module for tunnel surface defects according to claim 1, characterized in that, The laser emitted by the laser is connected to the laser lens through an optical fiber. The cylindrical mirror inside the laser lens shapes the laser point source into a linear source, and the prism emits it out at a certain fan angle, forming a linear and uniform light band on the imaging surface.
6. The integrated visual acquisition module for tunnel surface defects according to claim 1, characterized in that, The laser trigger signal distributor is integrated into the core control motherboard. The operation of the laser is controlled by the laser trigger signal distributor, which can precisely control the lighting time and duration of the laser. When the laser driver circuit board receives the laser trigger signal distributed by the core control motherboard, it controls the corresponding laser to perform electro-optic conversion and output a laser light source.
7. The integrated visual acquisition module for tunnel surface defects according to claim 1, characterized in that, The laser is directly fixed to the bottom of the heat sink fins, and thermally conductive adhesive is applied between its mounting surface and the bottom mounting surface of the heat sink fins.
Citation Information
Patent Citations
Tunnel lining apparent disease detection system based on industrial camera
CN114460090A
Rail transit tunnel structure apparent disease detection device
CN115326819A
Adjustable three-dimensional detection system for tunnel diseases
CN111855664A
Railway tunnel high-definition imaging detection device
CN217385264U