Tunnel gushing water machine vision unattended monitoring and early warning method, device and system

Abstract

This invention discloses a machine vision-based unattended monitoring and early warning method, device, and system for tunnel water inrush, comprising: calculating the optical flow vector of each pixel in the tunnel face image; determining the location, number, and optical flow vector of the inrush point based on the optical flow vector of each pixel; determining whether the inrush point has disappeared based on the change in the optical flow vector at the inrush point; determining whether a water inrush exists based on the optical flow vector at the inrush point and determining the optical flow vector at the water inrush location; determining the water volume of the water inrush based on the optical flow vector at the water inrush location; and determining whether to issue an early warning based on the water volume. By utilizing the motion characteristics of the water inrush, the optical flow method is used to monitor the range of tunnel water inrush, achieving real-time monitoring of tunnel water inrush.

Description

Technical Field

[0001] This invention relates to the field of tunnel flooding disaster monitoring and prevention technology, and in particular to a machine vision-based unattended monitoring and early warning method, device and system for tunnel flooding. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] During tunnel construction, tunnels often encounter hazards such as sudden water inrushes due to their passage through high mountain canyons, water-rich areas, fault fracture zones, and karst caves. Therefore, monitoring and early warning are essential steps in handling tunnel water inrush disasters.

[0004] Monitoring and early warning of tunnel water inrushes are crucial and a prerequisite for proactive disaster prevention and control. Currently, most existing methods for monitoring and detecting tunnel water inrushes involve manually deploying sensors. This approach often has drawbacks, such as the risk of water inrushes at any time, which could harm monitoring personnel; and the potential for significant errors and insufficient accuracy in dimly lit, high-risk environments.

[0005] To overcome the above problems, machine vision recognition technology has been gradually introduced into the monitoring of sudden water inrush in tunnels. However, at present, machine vision recognition can only obtain blurry videos inside the tunnel, and due to the lack of relevant algorithms, staff cannot accurately find the water inrush point from the video. Summary of the Invention

[0006] To address the aforementioned issues, this invention proposes a machine vision-based unattended monitoring and early warning method, device, and system for tunnel water inrush. Utilizing the motion characteristics of the water flow through the inrush, the invention employs optical flow to monitor the range of water inrush in the tunnel. By calculating the moving optical flow vector in the image, the location, quantity, and volume of the inrush are determined, along with whether an early warning is issued within the tunnel, thus achieving real-time monitoring of tunnel water inrush.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] In a first aspect, the present invention provides a machine vision-based unattended monitoring and early warning method for sudden water inrush in tunnels, comprising:

[0009] Acquire the image of the face to be detected and calculate the optical flow vector of each pixel in the face image;

[0010] The location, number, and optical flow vector of each pixel in the face image are determined based on the optical flow vector of the surge point.

[0011] The number of surge points is updated based on the change in the optical flow vector at the surge point to determine whether the current surge point has disappeared.

[0012] The existence of a sudden water flow is determined by the optical flow vector at the sudden water flow point, and the optical flow vector at the sudden water flow point is determined.

[0013] The volume of the sudden water inrush is determined based on the optical flow vector at the point of the sudden water inrush, and whether an early warning should be issued is determined based on the volume of the sudden water inrush.

[0014] As an alternative implementation, the dense optical flow method is used to solve a two-frame estimation algorithm based on spatial gradient, thereby estimating the optical flow vector of each pixel in the face image.

[0015] As an alternative implementation, the location of the surge point in the face image is determined based on the comparison result between the optical flow vector of each pixel in the face image and the surge point identification threshold, thereby obtaining the location and number of all surge points in the face image, and the optical flow vector at the surge point is determined based on the optical flow vector of the pixel at the surge point.

[0016] As an alternative implementation method, the process of determining whether the current surge point has disappeared includes: determining whether the current surge point has disappeared based on whether the change in the optical flow vector at the surge point is greater than the surge point existence threshold; if the former is less than the latter, the surge point disappears.

[0017] As an alternative implementation, the process of determining the volume of the sudden inrush water includes: calculating the volume of the sudden inrush water based on the optical flow vector at the location of the sudden inrush water, the number of video frames acquired, the distance between the video acquisition device and the working face, and the range and angle of the working face image acquired by the video acquisition device.

[0018] As an alternative implementation method, the process of determining whether to issue an early warning includes: if the change in water volume at the sudden water inrush point is greater than the sudden water inrush threshold, then an early warning for the sudden water inrush change is issued; otherwise, no early warning is issued.

[0019] Secondly, the present invention provides a machine vision-based unattended monitoring and early warning device for sudden water inrush in tunnels, comprising:

[0020] The data acquisition module is configured to acquire the face image to be detected and calculate the optical flow vector of each pixel in the face image;

[0021] The surge point identification module is configured to determine the location, number, and optical flow vector of the surge point based on the optical flow vector of each pixel in the face image.

[0022] The surge point update module is configured to determine whether the current surge point has disappeared based on the change in the optical flow vector at the surge point, and thus update the number of surge points;

[0023] The sudden water flow identification module is configured to determine whether a sudden water flow exists based on the optical flow vector at the sudden point, and to determine the optical flow vector at the sudden water flow location;

[0024] The water volume and early warning module is configured to determine the water volume of the sudden water flow based on the optical flow vector at the sudden water flow point, and to determine whether to issue an early warning based on the water volume of the sudden water flow.

[0025] Thirdly, the present invention provides a machine vision unattended monitoring and early warning system for sudden water inrush in tunnels, comprising: a video acquisition module and the machine vision unattended monitoring and early warning device for sudden water inrush in tunnels as described in the second aspect.

[0026] The video acquisition module is used to acquire video streams of the tunnel face and sidewalls and send them to the tunnel water inrush machine vision unattended monitoring and early warning device.

[0027] The tunnel water inrush machine vision unattended monitoring and early warning device is used to monitor and warn of tunnel water inrush in real time based on the tunnel face images in the video stream.

[0028] As an alternative implementation, the video acquisition module includes: a video monitoring lens, a rotary bearing, and a rotary base; the video monitoring lens is fixed on the rotary bearing, enabling the video acquisition module to rotate 110° in the vertical section; the rotary bearing is connected to the rotary base, enabling the video acquisition module to rotate 360° in the horizontal section.

[0029] As an alternative implementation, the video acquisition module is mounted on the surrounding rock of the tunnel via a bracket, the height of which is adjustable.

[0030] The video acquisition module is equipped with a fill light to increase the light intensity.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0032] This invention proposes a machine vision-based unattended monitoring and early warning method, device, and system for tunnel water inrush. Based on optical flow and video surveillance, it identifies and monitors tunnel water inrush. Since water inrush in tunnels is in motion, it often persists for extended periods. This motion generates a large optical flow vector. Therefore, this invention obtains real-time monitoring images of potential water inrush locations such as the tunnel excavation face and sidewalls. Utilizing the motion characteristics of the water flow, it identifies the water inrush during its evolution using optical flow. By calculating the moving optical flow vector in the image, it determines the location, quantity, and volume of the water inrush and decides whether to issue an early warning within the tunnel. This achieves real-time monitoring of tunnel water inrush, predicts its evolution, and provides assurance for safe tunnel construction.

[0033] This invention improves the real-time performance of video transmission through wireless transmission technology, further ensuring the real-time performance of monitoring and early warning, while reducing the need to lay transmission cables at the construction face, thus avoiding the impact of monitoring signal interruption caused by cable damage.

[0034] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0035] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0036] Figure 1 This is a flowchart of the unattended monitoring and early warning method for tunnel water inrush using machine vision, provided in Embodiment 1 of the present invention.

[0037] Figure 2 This is a schematic diagram of the tunnel water inrush machine vision unattended monitoring and early warning system provided in Embodiment 3 of the present invention;

[0038] Figure 3 This is a schematic diagram of the video acquisition module provided in Embodiment 3 of the present invention;

[0039] The components include: 1. Video acquisition module; 2. Fill light; 3. Bracket; 4. Tunnel surrounding rock; 5. Tunnel face; 6. Video monitoring lens; 7. 110° rotating bearing; 8. 360° rotating base; 9. Wireless transmission module; and 10. Audio receiving module. Detailed Implementation

[0040] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0041] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0042] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. Furthermore, it should be understood that the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0043] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0044] Example 1

[0045] This embodiment provides a machine vision-based unattended monitoring and early warning method for sudden water inrush in tunnels, such as... Figure 1 As shown, it includes:

[0046] Acquire the image of the face to be detected and calculate the optical flow vector of each pixel in the face image;

[0047] The location, number, and optical flow vector of each pixel in the face image are determined based on the optical flow vector of the surge point.

[0048] The number of surge points is updated based on the change in the optical flow vector at the surge point to determine whether the current surge point has disappeared.

[0049] The existence of a sudden water flow is determined by the optical flow vector at the sudden water flow point, and the optical flow vector at the sudden water flow point is determined.

[0050] The volume of the sudden water inrush is determined based on the optical flow vector at the point of the sudden water inrush, and whether an early warning should be issued is determined based on the volume of the sudden water inrush.

[0051] In this embodiment, the optical flow vector of each pixel in the face image is calculated using the optical flow method. The specific process includes:

[0052] In acquiring a video stream, the motion of the target object between adjacent video frames is called optical flow, as the target is not stationary. All optical flows on the image form an optical flow field, representing the movement of each point from the first frame to the second. Since the video stream contains two-dimensional image information, the velocity vector of a moving object in three-dimensional space is projected as a two-dimensional instantaneous velocity vector through the optical flow field. By judging the two-dimensional instantaneous velocity vector field, the state of sudden water inrush in the tunnel can be identified.

[0053] Specifically: a two-frame estimation algorithm based on spatial gradient is used to solve the problem using dense optical flow, specifically the classic Farneback optical flow algorithm; the face image is converted into a grayscale image and treated as a function of a two-dimensional signal, then its dependent variable is the two-dimensional coordinate position X = (x, y). T ;

[0054] First, the neighborhood information of each pixel is approximated using local polynomial approximation. Then, the polynomial expansion coefficients of pixels in two consecutive frames are analyzed, and the optical flow vector is estimated based on this. The main steps are as follows:

[0055] By approximating the neighborhood signal using a polynomial f(x) in the local coordinate space, we can obtain:

[0056] f(x)~x T Ax+b T x+c (1)

[0057] Where A is a symmetric matrix, b is a vector, and c is a scalar.

[0058] The symmetric matrix A is obtained by least-squares weighted fitting of the neighborhood information of the pixels, and the weight coefficients are related to the size and position of the neighboring pixels.

[0059] Let d denote the optical flow of displacement between two frames. Then, the images of the two frames can be represented as follows:

[0060]

[0061]

[0062] Since the appearance information of pixels in the obtained video frames remains unchanged between frames, the corresponding coefficients are the same. If the optical flow change A1 is not singular, then:

[0063]

[0064] Now, we calculate the spread coefficients for two adjacent video frames, A1(x), b1(x), c1(x) and A2(x), b2(x), c2(x), respectively. A1 and A2 are not actually equal, so an approximation is needed as follows:

[0065]

[0066] At this point, a solution cannot be obtained due to insufficient conditions. Therefore, the following equation is added as a condition:

[0067]

[0068] Based on the above conditions, the constraints are as follows:

[0069] A(x)d(x)=Δb(x) (7)

[0070] Images often contain a lot of noise, so we can establish the least squares form of equation (7) within the neighborhood of the feature pixels, and obtain:

[0071]

[0072] Where ω(Δx) is the weight function of pixel x in neighborhood I.

[0073] Minimizing the above expression yields:

[0074] d(x)=(∑ωA T A) -1 ∑ωA T Δb (9)

[0075] e(x)=(∑ωb T b)-d T (x)∑ωA T Δb (10)

[0076] The above inferences are based on the assumption that the local polynomials of two signals with the same coordinates are the same. Therefore, this assumption does not hold when the optical flow displacement is not zero. To avoid errors when the displacement is too large, a priori displacement d′(x) is introduced, and equations (5) and (6) become:

[0077]

[0078]

[0079] Where x and x′ satisfy:

[0080] x′=x+d′(x) (13)

[0081] By combining equations (9) and (11) to (13), the estimated optical flow displacement value Δb(x) can be obtained through iterative calculation, which is the optical flow vector of each pixel.

[0082] In this embodiment, the location of the surge point in the face image is determined by comparing the optical flow vector of each pixel in the face image with the preset surge point identification threshold, thereby obtaining the location and number of all surge points in the face image, and the optical flow vector at the surge point is determined by the optical flow vector of the pixel at the surge point.

[0083] Specifically, the optical flow vectors of the pixels identified as the surge points in the face image are saved into a set and called the optical flow vectors at the surge points. This means that the values ​​of the optical flow vectors at these pixels are extracted and assigned the meaning of the optical flow vectors at the surge points.

[0084] This is understandable, because the tunnel face image changes in real time. Therefore, based on the optical flow vector of each pixel in the real-time tunnel face image, it is determined whether a new surge point has appeared.

[0085] In this embodiment, the change value of the optical flow vector at the surge point is calculated according to the optical flow method. It is then determined whether the change value of the optical flow vector at the surge point is greater than a preset surge point existence threshold, thereby determining whether the current surge point has disappeared. If the former is greater than or equal to the latter, the surge point exists; otherwise, the surge point disappears.

[0086] In this embodiment, the existence of a sudden water flow is determined based on the optical flow vector at the sudden water flow point, and the optical flow vector at the sudden water flow point is determined.

[0087] Specifically: After processing the face image using the Farneback algorithm, the optical flow displacement value of each pixel can be obtained, that is, the optical flow vector of each pixel. The magnitude of the optical flow vector is compared with the preset value to determine the location of the pixel where there is a sudden surge of water flow. The optical flow vector value of this pixel is extracted and determined as the optical flow vector at the sudden surge of water flow.

[0088] In this embodiment, determining the volume of the gushing water based on the optical flow vector at the gushing water location includes: calculating the volume of the gushing water based on the optical flow vector at the gushing water location, the number of video frames acquired, the distance between the video equipment and the working face, and the range and angle of the image acquired by the video monitoring lens;

[0089] Specifically:

[0090] The optical flow vector is considered as the instantaneous velocity of the pixel's movement. By multiplying the flow velocity by time and then performing a proportional conversion, the surge water volume can be obtained. The specific formula is as follows:

[0091]

[0092] Where: Q is the volume of the sudden inrush water, w is the correlation coefficient, Δb(x) is the optical flow vector, f is the number of video frames, is the distance between the video equipment and the working face, and α is the range and angle of the image acquired by the video monitoring lens.

[0093] In this embodiment, determining whether to issue an early warning based on the volume of the sudden water flow includes: if the change in water volume at the sudden water flow is greater than the sudden water flow threshold, then an early warning for the change in sudden water flow is issued; otherwise, no early warning is issued.

[0094] Example 2

[0095] This embodiment provides a machine vision-based unattended monitoring and early warning device for sudden water inrush in tunnels, characterized in that it includes:

[0096] The data acquisition module is configured to acquire the face image to be detected and calculate the optical flow vector of each pixel in the face image;

[0097] The surge point identification module is configured to determine the location, number, and optical flow vector of the surge point based on the optical flow vector of each pixel in the face image.

[0098] The surge point update module is configured to determine whether the current surge point has disappeared based on the change in the optical flow vector at the surge point, and thus update the number of surge points;

[0099] The sudden water flow identification module is configured to determine whether a sudden water flow exists based on the optical flow vector at the sudden point, and to determine the optical flow vector at the sudden water flow location;

[0100] The water volume and early warning module is configured to determine the water volume of the sudden water flow based on the optical flow vector at the sudden water flow point, and to determine whether to issue an early warning based on the water volume of the sudden water flow.

[0101] It should be noted that the above modules correspond to the steps described in Embodiment 1, and the examples and application scenarios implemented by the above modules and the corresponding steps are the same, but are not limited to the content disclosed in Embodiment 1. It should also be noted that the above modules, as part of the system, can be executed in a computer system such as a set of computer-executable instructions.

[0102] Example 3

[0103] like Figure 2 and Figure 3 As shown, this embodiment provides a tunnel water inrush machine vision unattended monitoring and early warning system, including: a video acquisition module and the tunnel water inrush machine vision unattended monitoring and early warning device described in Embodiment 2;

[0104] The video acquisition module is used to acquire video streams of the tunnel face and sidewalls and send them to the tunnel water inrush machine vision unattended monitoring and early warning device.

[0105] The tunnel water inrush machine vision unattended monitoring and early warning device is used to monitor and warn of tunnel water inrush in real time based on the tunnel face images in the video stream.

[0106] In this embodiment, the video acquisition module 1 specifically includes: a video monitoring lens 6, a 110° rotating bearing 7, a 360° rotating base 8, a wireless transmission module 9, and a radio module 10;

[0107] Among them, the video surveillance lens 6 is fixed on the 110° rotating bearing 7, which enables the video acquisition module to rotate 110° in the vertical section.

[0108] The 110° rotating bearing 7 is connected to the 360° rotating base 8, which enables the video acquisition module to rotate 360° in the horizontal plane.

[0109] The 360° rotating base 8 is connected to the wireless transmission module 9 and the radio module 10;

[0110] The wireless transmission module 9 is used to transmit video streams, enabling real-time acquisition of video from the tunnel site, avoiding monitoring interruptions caused by damage to transmission cables during construction, and improving data transmission efficiency.

[0111] The audio receiving module 10 is used to collect audio information to assist in judgment.

[0112] As an alternative implementation method, the video surveillance lens 6 is used to acquire real-time image information of sudden water inrush. It adopts a high-definition wide-angle lens, which helps to improve the image acquisition accuracy, increase the shooting range, and improve the identification efficiency of sudden water inrush in the tunnel.

[0113] In this embodiment, the video acquisition module 1 is mounted on the tunnel surrounding rock 4 via a bracket 3. The installation position of the video acquisition module 1 needs to be able to monitor the entire tunnel face 5 and sidewalls. The bracket 3 is height adjustable for easy shooting.

[0114] As an alternative implementation method, supplementary lighting 2 is also provided to increase the light intensity inside the tunnel and improve the video acquisition effect.

[0115] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A machine vision-based unattended monitoring and early warning method for sudden water inrush in tunnels, characterized in that: include: Acquire the image of the face to be detected and calculate the optical flow vector of each pixel in the face image; The location, number, and optical flow vector of the surge point are determined based on the optical flow vector of each pixel in the face image. The location of the surge point in the face image is determined based on the comparison between the optical flow vector of each pixel in the face image and the surge point identification threshold. Thus, the location and number of all surge points in the face image are obtained. The optical flow vector at the surge point is determined based on the optical flow vector of the pixel at the surge point. The process of determining whether a surge point has disappeared is based on the change in the optical flow vector at the surge point, thereby updating the number of surge points. The process of determining whether a surge point has disappeared includes: determining whether a surge point has disappeared based on whether the change in the optical flow vector at the surge point is greater than the surge point existence threshold. If the former is less than the latter, the surge point disappears. The existence of a sudden water flow is determined by the optical flow vector at the sudden water flow point, and the optical flow vector at the sudden water flow point is determined. The volume of the sudden water inrush is determined based on the optical flow vector at the point of sudden water inrush, and whether an early warning should be issued is determined based on the volume of the sudden water inrush. The process of determining the volume of the sudden water inrush includes: calculating the volume of the sudden water inrush based on the optical flow vector at the location of the inrush, the number of video frames acquired, the distance between the video acquisition device and the working face, and the angle of the range of the working face image acquired by the video acquisition device. Specifically, the optical flow vector is regarded as the instantaneous velocity of the pixel movement. The volume of the sudden water inrush is obtained by multiplying the flow velocity by time and then performing a proportional conversion. The specific formula is as follows: in: For the volume of the sudden inrush water, The correlation coefficient, For optical flow vector, For video frame rate, The distance between the video equipment and the working face. To acquire the range and angle of images from video surveillance cameras.

2. The unattended monitoring and early warning method for tunnel water inrush using machine vision as described in claim 1, characterized in that, The dense optical flow method is used to solve a two-frame estimation algorithm based on spatial gradient, thereby estimating the optical flow vector of each pixel in the face image.

3. The unattended monitoring and early warning method for tunnel water inrush using machine vision as described in claim 1, characterized in that, The process of determining whether to issue an early warning includes: if the change in water volume at the sudden water inrush point is greater than the sudden water inrush threshold, an early warning for the sudden water inrush change is issued; otherwise, no early warning is issued.

4. A machine vision-based unattended monitoring and early warning device for sudden water inrush in tunnels, characterized in that: include: The data acquisition module is configured to acquire the face image to be detected and calculate the optical flow vector of each pixel in the face image; The surge point identification module is configured to determine the location, number, and optical flow vector of the surge point based on the optical flow vector of each pixel in the face image. Based on the comparison between the optical flow vector of each pixel in the face image and the surge point identification threshold, the module determines the location of the surge point in the face image, thereby obtaining the location and number of all surge points in the face image. Finally, based on the optical flow vector of the pixel at the surge point, the module determines the optical flow vector at the surge point. The surge point update module is configured to determine whether the current surge point has disappeared based on the change in the optical flow vector at the surge point, and then update the number of surge points. The process of determining whether the current surge point has disappeared includes: determining whether the current surge point has disappeared based on whether the change in the optical flow vector at the surge point is greater than the surge point existence threshold. If the former is less than the latter, the surge point disappears. The sudden water flow identification module is configured to determine whether a sudden water flow exists based on the optical flow vector at the sudden point, and to determine the optical flow vector at the sudden water flow location; The water volume and early warning module is configured to determine the water volume of the sudden water flow based on the optical flow vector at the sudden water flow point, and to determine whether to issue an early warning based on the water volume of the sudden water flow. The process of determining the volume of the sudden water inrush includes: calculating the volume of the sudden water inrush based on the optical flow vector at the location of the inrush, the number of video frames acquired, the distance between the video acquisition device and the working face, and the angle of the range of the working face image acquired by the video acquisition device. Specifically, the optical flow vector is regarded as the instantaneous velocity of the pixel movement. The volume of the sudden water inrush is obtained by multiplying the flow velocity by time and then performing a proportional conversion. The specific formula is as follows: in: For the volume of the sudden inrush water, The correlation coefficient, For optical flow vector, For video frame rate, The distance between the video equipment and the working face. To acquire the range and angle of images from video surveillance cameras.

5. A machine vision-based unattended monitoring and early warning system for sudden water inrush in tunnels, characterized in that it comprises: a video acquisition module and the machine vision-based unattended monitoring and early warning device for sudden water inrush in tunnels as described in claim 4; The video acquisition module is used to acquire video streams of the tunnel face and sidewalls and send them to the tunnel water inrush machine vision unattended monitoring and early warning device. The tunnel water inrush machine vision unattended monitoring and early warning device is used to monitor and warn of tunnel water inrush in real time based on the tunnel face images in the video stream.

6. The tunnel sudden water inrush machine vision unattended monitoring and early warning system as described in claim 5, characterized in that, The video acquisition module includes: a video monitoring lens, a rotary bearing, and a rotary base; the video monitoring lens is fixed on the rotary bearing, enabling the video acquisition module to rotate 110° in the vertical plane; the rotary bearing is connected to the rotary base, enabling the video acquisition module to rotate 360° in the horizontal plane.

7. The tunnel sudden water inrush machine vision unattended monitoring and early warning system as described in claim 5, characterized in that, The video acquisition module is mounted on the surrounding rock of the tunnel via a bracket, the height of which is adjustable. The video acquisition module is equipped with a fill light to increase the light intensity.

Images