Tunnel gushing water monitoring and early warning method, device and system based on background difference method
Through the tunnel water surge monitoring and early warning method based on the background difference method, the number of water surge points, water quantity and water quality changes in the tunnel video are identified, and the problem of insufficient real-time and accuracy of monitoring in the existing technology is solved, and the accurate warning of tunnel water surge is achieved and safe construction.
Patent Information
- Application Number
- CN202410974833.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-08
AI Technical Summary
The existing tunnel surge flood disaster monitoring methods have poor monitoring real-time performance, poor identification accuracy, and cannot provide accurate early warnings, and manual monitoring in high-risk environments have safety hazards.
The tunnel water surge monitoring and early warning method is adopted based on the background difference method. By obtaining tunnel video images, establishing background models, conducting real-time video monitoring, identifying the number of water surge points, water quantity and water quality changes, issuing early warning signals, and improving the accuracy and real-time nature of early warnings. .
Accurate early warning of sudden flood disasters in tunnels has been achieved, delayed construction periods and casualties of construction workers, improved the accuracy and safety of monitoring, avoided interruptions in monitoring signal transmission, and ensured 24-hour online monitoring.
Smart Images

Figure CN118911770A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of monitoring and preventing and controlling sudden water inrush in tunnels, and in particular to a method, device and system for monitoring and warning sudden water in tunnels based on a background difference method. Background Art
[0002] During tunnel construction, sudden water disasters caused by fault fracture zones, karst, etc. are often encountered, causing economic losses and delays in construction period, and in serious cases, endangering the lives of construction workers. Effective monitoring and early warning work is an important means to avoid major sudden water disasters. Most of the existing tunnel sudden water disaster monitoring and detection methods are manually arranged sensors for monitoring. This detection method often has some disadvantages, such as the possibility of sudden water in the tunnel at any time, causing harm to the monitoring personnel; manual monitoring in dim and high-risk environments has large errors, and the accuracy is easy to fail to meet actual needs.
[0003] In the above-mentioned method of monitoring sudden water disasters through sensors, the changes in indicators of sudden water disasters can only be observed by the naked eye or identified by instruments, which has the problems of poor real-time monitoring, poor identification accuracy, and inability to accurately warn of sudden water changes.
[0004] To overcome the above problems, machine vision recognition technology has gradually been introduced into tunnel water burst monitoring. The inventors found that the machine vision recognition technology used in the prior art usually searches for water burst points from tunnel monitoring videos and determines the changes in the number of water burst points. This judgment mode based on a single element cannot provide accurate early warning of the development trend of water bursts. In addition, the inventors also found that current machine vision recognition can only obtain blurred videos inside the tunnel, and due to the lack of relevant algorithms, staff cannot accurately find water burst points from the video. Summary of the invention
[0005] In order to overcome the shortcomings of the above-mentioned prior art, the present invention provides a tunnel sudden water burst monitoring and early warning method, device and system based on the background difference method. Based on the three key indicators of tunnel sudden water burst early warning, the number of water burst points, the amount of sudden water burst and the water quality of sudden water burst, the increase in the number of water burst points, the increase in the outline of water burst and the change of water quality from clear to turbid are taken as precursors of disaster. At the same time, a visible light background image model of the tunnel sudden water burst area is established, and real-time video monitoring and early warning of the image are performed based on the background difference method, which improves the accuracy of the early warning.
[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:
[0007] A first aspect of the present invention provides a tunnel water inrush monitoring and early warning method based on a background difference method.
[0008] The tunnel water inrush monitoring and early warning method based on background difference method includes the following steps:
[0009] Obtain the initial image of the tunnel video and establish the background model of the tunnel area;
[0010] Acquire a video image of the tunnel to be monitored, perform background difference processing on the video image of the tunnel to be monitored and the background model of the tunnel area to obtain a binary image;
[0011] Determine whether there is a highlight area in the binary image that conforms to the sudden water flow morphology. If so, further determine the change in the sudden water flow volume based on the change in the number of pixels in the highlight area, determine the change in the sudden water flow quality based on the change in the RGB value of the pixels in the source image corresponding to the highlight area, and determine the change in the number of water flow points based on the contour of the highlight area;
[0012] Based on the changes in the amount of sudden water, the water quality and the number of water gushing points, sudden water burst monitoring and early warning signals are issued.
[0013] Optionally, a mixed Gaussian distribution model is used to perform background modeling on the initial image of the tunnel video to obtain a tunnel area background model.
[0014] Optionally, the calculation process of the binary image is:
[0015]
[0016] Among them, E k (x, y) is the pixel value of the binary image coordinate, A k is the kth frame image in the tunnel video image to be monitored, B k is the kth frame image in the tunnel area background model, (x, y) is the coordinate point in the image, and T is the threshold required for binarization processing.
[0017] Optionally, the change in the amount of sudden water inrush is determined based on the change in the number of pixels in the highlight area, specifically:
[0018] Distinguish the background black and highlight areas in the current frame binary image, extract the outline of the highlight area, read the number of pixels in the outline of the highlight area, and use the number of pixels in the outline of the highlight area to represent the amount of sudden water inrush;
[0019] Compare the number of pixels in the outline of the highlighted area identified in the current frame with the number of pixels in the outline of the highlighted area identified in the frame before the set time. If the former is greater than the latter, the water inflow is increasing, otherwise it is decreasing.
[0020] Optionally, the change in water quality of the sudden water is determined based on the change in the RGB value of the pixel point corresponding to the source image in the highlighted area, specifically:
[0021] Divide the source image area corresponding to the highlighted area of the current frame binary image into multiple small areas;
[0022] Select the central pixel of each small area for color recognition, and obtain the RGB value of the central pixel of each small area;
[0023] Take the average RGB value of all the central pixels in the small area to get the average RGB value of the sudden water;
[0024] Compare the average RGB value of the sudden water in the current frame with the average RGB value of the sudden water in the frame before the set time. If the former is smaller than the latter, the sudden water becomes turbid, otherwise it becomes clear.
[0025] Optionally, the change in the number of water inflow points is determined based on the outline of the highlighted area, specifically:
[0026] Distinguish the background black and highlight areas in the current frame binary image, and extract the outline of the highlight area of the current frame;
[0027] Compare the outline of the highlighted area of the current frame with the outline of the highlighted area of the previous frame of the binary image to obtain the area where the outline changes;
[0028] It is determined whether the area with the changed contour is adjacent to the contour of the highlighted area of the previous frame. If not, it is a newly added water gushing point. If adjacent, it is not a newly added water gushing point.
[0029] Optionally, based on the changes in the amount of sudden water inrush, the changes in water quality and the changes in the number of water inrush points, a sudden water inrush monitoring early warning signal is issued, specifically:
[0030] If the number of water inrush points increases, a monitoring warning signal will be issued immediately;
[0031] If the water gushing point remains unchanged or decreases, a monitoring early warning signal will be issued when the amount of sudden water increases and the water quality of the sudden water becomes turbid.
[0032] The second aspect of the present invention provides a tunnel water inrush monitoring and early warning device based on a background difference method.
[0033] The tunnel water inrush monitoring and early warning device based on background difference method includes:
[0034] The background model building module is configured to: obtain the initial image of the tunnel video and build a background model of the tunnel area;
[0035] The background difference processing module is configured to: obtain the video image of the tunnel to be monitored, perform background difference processing on the video image of the tunnel to be monitored and the background model of the tunnel area to obtain a binary image;
[0036] The judgment module is configured to: determine whether there is a highlight area in the binary image that conforms to the sudden water flow morphology, and if so, further determine the change in the sudden water flow volume based on the change in the number of pixels in the highlight area, determine the change in the sudden water flow quality based on the change in the RGB value of the pixels of the source image corresponding to the highlight area, and determine the change in the number of water flow points based on the contour of the highlight area;
[0037] The early warning module is configured to issue a sudden water surge monitoring early warning signal based on changes in the amount of sudden water surge, changes in water quality and changes in the number of water surge points.
[0038] A third aspect of the present invention provides a tunnel water inrush monitoring and early warning system based on a background difference method.
[0039] A tunnel water inrush monitoring and early warning system based on background difference method comprises a video acquisition module and a tunnel water inrush monitoring and early warning device based on background difference method according to the first aspect, wherein:
[0040] The video acquisition module is used to acquire the video stream of the tunnel side wall and the tunnel face, and send it to the tunnel sudden water inrush monitoring and early warning device based on the background difference method;
[0041] The tunnel water inrush monitoring and early warning device based on the background difference method is used to perform tunnel water inrush monitoring and early warning according to the video image of the tunnel to be monitored in the video stream.
[0042] Optionally, the video acquisition module includes a video surveillance lens, a rotating base and a rotating bearing, a bracket is provided on the rotating bearing, a rotating base is rotatably connected between the brackets, the video surveillance lens is connected to the rotating base, the rotating bearing is used to drive the video surveillance lens to rotate 360° in a vertical section, the rotating base is used to drive the video surveillance lens to rotate 360° in a horizontal section, and a wireless transmission module is provided at the bottom of the rotating bearing, and the wireless transmission module is set on the tunnel surrounding rock.
[0043] One or more of the above technical solutions have the following beneficial effects:
[0044] The present invention is based on three key indicators of tunnel sudden water burst warning, namely the number of water burst points, the amount of sudden water burst and the quality of sudden water burst. The increased number of water burst points, the enlarged water burst outline and the change of water quality from clear to turbid are the precursors of disaster. High-definition video monitoring probes are used to monitor the entire tunnel, and real-time tunnel area images are collected. A visible light background image model of the tunnel sudden water burst area is established, and real-time video monitoring and early warning of the images are performed based on the background difference method to guide safe construction. The background difference method and deep learning algorithm are used to realize real-time dynamic monitoring and accurate early warning of tunnel sudden water burst at a relatively low computing input cost, thereby solving the problems of dangerous personnel patrols and automatic identification in high-risk environments.
[0045] The present invention first determines whether there is a highlighted area in the binary image that conforms to the sudden water flow morphology. If so, it further determines the change in the amount of sudden water based on the change in the number of pixels in the highlighted area, determines the change in the quality of the sudden water based on the change in the RGB value of the pixels of the source image corresponding to the highlighted area, and determines the change in the number of water gushing points based on the contour of the highlighted area. Through the above-mentioned multiple logical judgment process, an early warning is issued for the sudden water gushing disaster in the tunnel, thereby improving the accuracy of the early warning.
[0046] The present invention adopts video monitoring and early warning technology based on background difference method. The analysis result is highly real-time and accurate. It can accurately predict the development trend of sudden water inrush and provide timely early warning prompts, thereby reducing construction delays and casualties of construction personnel caused by untimely discovery and incorrect judgment of sudden water inrush disasters in tunnels, reducing damage to mechanical equipment and preventing the occurrence of sudden water disasters.
[0047] The present invention avoids the interruption of monitoring signal transmission caused by cable damage during construction through wireless transmission technology, and ensures 24-hour online monitoring.
[0048] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0049] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.
[0050] Figure 1 This is a flow chart of the method of the first embodiment.
[0051] Figure 2 This is a layout diagram of the video acquisition module of the third embodiment.
[0052] Figure 3 This is a structural diagram of the video acquisition module of the third embodiment.
[0053] 1: Tunnel water inrush monitoring and early warning system; 2: Tunnel fill light; 3: Tunnel side wall; 4: Working face; 5: Video surveillance camera; 6: Rotating base; 7: Rotating bearing; 8: Wireless transmission module. DETAILED DESCRIPTION
[0054] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.
[0055] It should be noted that the terms used herein are for describing specific embodiments only and are not intended to be limiting of exemplary embodiments according to the present invention.
[0056] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.
[0057] Embodiment 1
[0058] This embodiment discloses a tunnel water inrush monitoring and early warning method based on background difference method.
[0059] like Figure 1 As shown, the tunnel water inrush monitoring and early warning method based on the background difference method includes the following steps:
[0060] Obtain the initial image of the tunnel video and establish the background model of the tunnel area;
[0061] Acquire a video image of the tunnel to be monitored, perform background difference processing on the video image of the tunnel to be monitored and the background model of the tunnel area to obtain a binary image;
[0062] Determine whether there is a highlight area in the binary image that conforms to the sudden water flow morphology. If so, further determine the change in the sudden water flow volume based on the change in the number of pixels in the highlight area, determine the change in the sudden water flow quality based on the change in the RGB value of the pixels in the source image corresponding to the highlight area, and determine the change in the number of water flow points based on the contour of the highlight area;
[0063] Based on the changes in the amount of sudden water, the water quality and the number of water gushing points, sudden water burst monitoring and early warning signals are issued.
[0064] This embodiment is different from the previous technical solutions in that this embodiment is based on three key indicators of the development of sudden water bursts in tunnels, namely the number of water bursting points, the amount of sudden water bursts and the quality of sudden water bursts. The increase in the number of water bursting points, the enlargement of the water bursting contour and the change of water quality from clear to turbid are used as precursors of disasters. Through the above-mentioned multiple logical judgment process, early warning of sudden water bursts in tunnels is issued, thereby improving the accuracy of the early warning.
[0065] This embodiment uses a high-definition video surveillance probe to monitor the entire tunnel, collects an initial image of the tunnel video, uses the collected initial image of the tunnel video to establish a background model of the tunnel area, performs background difference between the real-time tunnel video image to be monitored and the background model, and obtains a binary image; determines whether the binary image has a highlighted area that conforms to the morphology of sudden water; if so, there is sudden water in the tunnel image area; if not, there is no sudden water in the area.
[0066] Furthermore, if there is continuous sudden water inrush in the image area, the water volume and water quality can be obtained through machine learning analysis, and then the sudden water inrush situation can be predicted, analyzed and warned.
[0067] Next, the technical solution of this embodiment will be described in detail:
[0068] S1. Collect real-time tunnel images through deployed high-definition video probes and transmit the images through wireless transmission modules;
[0069] S2. Use tunnel real-time images to establish tunnel area background model:
[0070] The Gaussian mixture distribution model (GMM) can model each pixel of the tunnel area image by superimposing several Gaussian distributions. Each Gaussian distribution can represent a tunnel background scene at a specific moment. The mixed use of multiple Gaussian distribution models can simulate multimodal situations in complex scenes. In this embodiment, the background model of the initial image of the tunnel video is performed using a mixed Gaussian distribution model to obtain a tunnel area background model.
[0071] Specifically include:
[0072] The background model of the initial image of the tunnel video is modeled using the mixed Gaussian distribution model. Assume that the pixel value of each pixel in the image at time t is X t , then the Gaussian mixture distribution model P(X t )for:
[0073]
[0074] In the above formula, k is the number of Gaussian probability density functions in the Gaussian mixture model. k is generally 3 to 5, and in this embodiment, it is 5. k,t is the weight of the kth Gaussian probability density function at time t, and the sum of the weights of each Gaussian component is 1, so as to ensure that the sum of the integrals of the probability density function over its scope is 1; η(X t , μ k,t ,Σ k,t ) is the pixel X t The kth Gaussian probability density function at time t, Σ k,t is the covariance matrix of the kth Gaussian probability density function at time t, μ k,t is the sample mean of the kth Gaussian probability density function at time t.
[0075] S3. Process the collected video image of the tunnel to be monitored and the background model using the background difference method:
[0076] Assume E k (x, y) is the pixel value of the binary image coordinate, A k To monitor the kth frame of the image in real time, B kis the k-th frame image in the background model, (x, y) is the coordinate point in the image, T is the threshold required for binarization, and the "binarized" image obtained after background difference is:
[0077]
[0078] S4. Determine whether the processed binary image has a highlight area that conforms to the sudden water flow morphology. If not, return to the previous step to process the next frame of the image; if there is a highlight area that conforms to the sudden water flow morphology, proceed to the next step;
[0079] S5. Continuously process tunnel images in real time, call machine learning algorithms to measure the area of the identified highlight areas of sudden water flow, compare the changes in the area of the highlight areas over a period of time, and obtain the changes in the amount of sudden water flow; compare the colors of the source images of the highlight areas in the images over a period of time through the algorithm to obtain the changes in the quality of the sudden water flow; identify whether there are sudden appearance or disappearance of sudden water flow highlight areas, and obtain the changes in the number of water flow points.
[0080] Specific:
[0081] The area of the highlighted area of the identified sudden water flow is calculated, and the image segmentation method is used to distinguish the background black and the highlighted area in the binary image and extract the outline of the highlighted area. The number of pixels inside the extracted outline is read, and the number of sudden water pixels in the highlighted area outline represents the amount of sudden water flow.
[0082] Identify the color of the source image in the highlighted area and obtain the water quality of the sudden water. Specifically, divide the source image area of the highlighted area of the binary image into multiple smaller areas according to a certain area, select the central pixel point of the small area for color identification to obtain the RGB value of each pixel point, and take the average of the RGB values of all points to represent the average sudden water color.
[0083] By identifying whether there is a sudden appearance or disappearance of a sudden water surge highlight area, the change in the number of water surge points can be obtained. Specifically, the image segmentation method is used to distinguish the background black and highlight areas in the binary image and extract the highlight area contour, which is compared with the highlight area contour of the previous binary image. If the area with the changed contour is not adjacent to the highlight area contour of the previous frame, it is a new water surge point; if they are adjacent, it is not. If the contour area disappears, the water surge points decrease, otherwise there is no change.
[0084] S6. Analyze and predict the changes in the amount, quality and number of sudden water inrush through machine learning algorithms;
[0085] Specifically include:
[0086] Compare the number of pixels in the highlighted area identified in the current frame with the number of pixels identified in the frame ten minutes ago. If the number increases, it indicates an increasing trend in water inflow, otherwise it indicates a decreasing trend.
[0087] Compare the average RGB value of the sudden water in the current frame with the RGB value of the frame identified ten minutes ago. If the RGB value decreases, the water quality of the sudden water will become turbid, otherwise it will become shallower.
[0088] Compare the sudden water gushing contour identified in the current frame with the contour in the previous frame to obtain the increase or decrease of the water gushing points.
[0089] S7. Every ten minutes, a pop-up window will be displayed to remind you of the numerical value and changing trend of the sudden water volume and water quality. If the number of water gushing points changes, a pop-up window will be displayed immediately.
[0090] This embodiment provides a tunnel water inrush monitoring and early warning method based on background difference method, collects real-time monitoring images, establishes a visible light background image model of the tunnel water inrush area, uses background difference method to identify key information of water inrush in real time, predicts and warns of development trends, and guides safe construction. The present invention has the advantage of using background difference method and deep learning algorithm to achieve real-time dynamic monitoring and accurate early warning of tunnel water inrush at a relatively low computing input cost, and solves the problem of personnel patrol danger and automatic identification in high-risk environments.
[0091] Embodiment 2
[0092] This embodiment discloses a tunnel water inrush monitoring and early warning device based on a background difference method.
[0093] The tunnel water inrush monitoring and early warning device based on background difference method includes:
[0094] The background model building module is configured to: obtain the initial image of the tunnel video and build a background model of the tunnel area;
[0095] The background difference processing module is configured to: obtain the video image of the tunnel to be monitored, perform background difference processing on the video image of the tunnel to be monitored and the background model of the tunnel area to obtain a binary image;
[0096] The judgment module is configured to: determine whether there is a highlight area in the binary image that conforms to the sudden water flow morphology, and if so, further determine the change in the sudden water flow volume based on the change in the number of pixels in the highlight area, determine the change in the sudden water flow quality based on the change in the RGB value of the pixels of the source image corresponding to the highlight area, and determine the change in the number of water flow points based on the contour of the highlight area;
[0097] The early warning module is configured to issue a sudden water surge monitoring early warning signal based on changes in the amount of sudden water surge, changes in water quality and changes in the number of water surge points.
[0098] Embodiment 3
[0099] This embodiment provides a tunnel water inrush monitoring and early warning system 1 based on a background difference method.
[0100] The tunnel water inrush monitoring and early warning system 1 based on the background difference method comprises a video acquisition module and the tunnel water inrush monitoring and early warning device based on the background difference method in the second embodiment, wherein:
[0101] The video acquisition module is used to acquire the video stream of the tunnel side wall 3 and the tunnel face 4, and send it to the tunnel sudden water inrush monitoring and early warning device based on the background difference method;
[0102] The tunnel water inrush monitoring and early warning device based on the background difference method is used to perform tunnel water inrush monitoring and early warning according to the video image of the tunnel to be monitored in the video stream.
[0103] like Figure 2 and Figure 3 As shown, further, the video acquisition module includes a video surveillance lens 5, a rotating base 6 and a rotating bearing 7, a bracket is arranged on the rotating bearing 7, a rotating base 6 is rotatably connected between the brackets, the video surveillance lens 5 is connected to the rotating base 6, the rotating bearing 7 is used to drive the video surveillance lens 5 to rotate 360° in the vertical section, the rotating base 6 is used to drive the video surveillance lens 5 to rotate 360° in the horizontal section, and a wireless transmission module 8 is arranged at the bottom of the rotating bearing 7, and the wireless transmission module 8 is arranged on the tunnel surrounding rock.
[0104] More specifically, the tunnel face 4 and the side walls are as follows Figure 2 As shown, the tunnel sudden water inrush monitoring and early warning system 1 based on the background difference method is used to monitor and transmit tunnel images in real time, and is generally installed near the tunnel side wall 3 on one side close to the tunnel face 4, so that the equipment can monitor the entire tunnel face 4 and the tunnel side wall 3;
[0105] The high-brightness tunnel fill light 2 is installed near the tunnel side wall 3 on the other side close to the tunnel face 4 to fill in the light of the tunnel environment, increase the light intensity in the tunnel, improve the shooting quality of the video stream, and improve the video acquisition effect.
[0106] like Figure 3 As shown, the video surveillance lens 5 is used to obtain tunnel images in real time;
[0107] 360° rotating base 6, which can realize 360° rotation of the integrated monitoring equipment in the horizontal section;
[0108] 360° rotating bearing 7, which can realize 360° rotation of the integrated monitoring equipment in the vertical section;
[0109] The wireless transmission module 8 can avoid interruption of monitoring signal transmission due to cable damage during construction, thus achieving real-time monitoring.
[0110] The steps involved in the apparatus of the above embodiment 2 correspond to the method embodiment 1. For the specific implementation, please refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method of the present invention.
[0111] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.
[0112] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it 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 on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.
Claims
1. Tunnel water inrush monitoring and early warning method based on background difference method, It is characterized in that The following steps are involved: Obtain the initial image of the tunnel video and establish the background model of the tunnel area; Acquire a video image of the tunnel to be monitored, perform background difference processing on the video image of the tunnel to be monitored and the background model of the tunnel area to obtain a binary image; Determine whether there is a highlight area in the binary image that conforms to the sudden water flow morphology. If so, further determine the change in the sudden water flow volume based on the change in the number of pixels in the highlight area, determine the change in the sudden water flow quality based on the change in the RGB value of the pixels in the source image corresponding to the highlight area, and determine the change in the number of water flow points based on the contour of the highlight area; Based on the changes in the amount of sudden water, the water quality and the number of water gushing points, sudden water burst monitoring and early warning signals are issued.
2. The tunnel water inrush monitoring and early warning method based on background difference method as claimed in claim 1, It is characterized in that The mixed Gaussian distribution model is used to model the background of the initial image of the tunnel video and obtain the tunnel area background model.
3. The tunnel water inrush monitoring and early warning method based on background difference method as claimed in claim 1, It is characterized in that The calculation process of the binary image is: Among them, E k (x, y) is the pixel value of the binary image coordinate, A k is the kth frame image in the tunnel video image to be monitored, B k is the kth frame image in the tunnel area background model, (x, y) is the coordinate point in the image, and T is the threshold required for binarization processing.
4. The tunnel water inrush monitoring and early warning method based on background difference method as claimed in claim 1, It is characterized in that The change in the amount of sudden water inrush is determined based on the change in the number of pixels in the highlighted area, specifically: Distinguish the background black and highlight areas in the current frame binary image, extract the outline of the highlight area, read the number of pixels in the outline of the highlight area, and use the number of pixels in the outline of the highlight area to represent the amount of sudden water inrush; Compare the number of pixels in the outline of the highlighted area identified in the current frame with the number of pixels in the outline of the highlighted area identified in the frame before the set time. If the former is greater than the latter, the water inflow is increasing, otherwise it is decreasing.
5. The tunnel water inrush monitoring and early warning method based on background difference method as claimed in claim 4, It is characterized in that The change in the quality of the sudden water is determined based on the change in the RGB value of the pixel points of the source image corresponding to the highlighted area, specifically: Divide the source image area corresponding to the highlighted area of the current frame binary image into multiple small areas; Select the central pixel of each small area for color recognition, and obtain the RGB value of the central pixel of each small area; Take the average RGB value of all the central pixels in the small area to get the average RGB value of the sudden water; Compare the average RGB value of the sudden water in the current frame with the average RGB value of the sudden water in the frame before the set time. If the former is smaller than the latter, the sudden water becomes turbid, otherwise it becomes clear.
6. The tunnel water inrush monitoring and early warning method based on background difference method as claimed in claim 5, It is characterized in that The change in the number of water inflow points is determined based on the outline of the highlighted area, specifically: Distinguish the background black and highlight areas in the current frame binary image, and extract the outline of the highlight area of the current frame; Compare the outline of the highlighted area of the current frame with the outline of the highlighted area of the previous frame of the binary image to obtain the area where the outline changes; It is determined whether the area with the changed contour is adjacent to the contour of the highlighted area of the previous frame. If not, it is a newly added water gushing point. If adjacent, it is not a newly added water gushing point.
7. The tunnel water inrush monitoring and early warning method based on background difference method as claimed in claim 6, It is characterized in that Based on the changes in the amount of sudden water inrush, the changes in water quality and the changes in the number of water inrush points, sudden water inrush monitoring and early warning signals are issued, specifically: If the number of water inrush points increases, a monitoring warning signal will be issued immediately; If the water gushing point remains unchanged or decreases, a monitoring early warning signal will be issued when the amount of sudden water increases and the water quality of the sudden water becomes turbid.
8. Tunnel water inrush monitoring and early warning device based on background difference method, It is characterized in that include: The background model building module is configured to: obtain the initial image of the tunnel video and build a background model of the tunnel area; The background difference processing module is configured to: obtain the video image of the tunnel to be monitored, perform background difference processing on the video image of the tunnel to be monitored and the background model of the tunnel area to obtain a binary image; The judgment module is configured to: determine whether there is a highlight area in the binary image that conforms to the sudden water flow morphology, and if so, further determine the change in the sudden water flow volume based on the change in the number of pixels in the highlight area, determine the change in the sudden water flow quality based on the change in the RGB value of the pixels of the source image corresponding to the highlight area, and determine the change in the number of water flow points based on the contour of the highlight area; The early warning module is configured to issue a sudden water surge monitoring early warning signal based on changes in the amount of sudden water surge, changes in water quality and changes in the number of water surge points.
9. Tunnel water inrush monitoring and early warning system based on background difference method, It is characterized in that The device comprises a video acquisition module and a tunnel water inrush monitoring and early warning device based on the background difference method as claimed in claim 8, wherein: The video acquisition module is used to acquire the video stream of the tunnel side wall and the tunnel face, and send it to the tunnel sudden water inrush monitoring and early warning device based on the background difference method; The tunnel water inrush monitoring and early warning device based on the background difference method is used to perform tunnel water inrush monitoring and early warning according to the video image of the tunnel to be monitored in the video stream.
10. The tunnel water inrush monitoring and early warning system based on background difference method as claimed in claim 9, It is characterized in that The video acquisition module includes a video surveillance lens, a rotating base and a rotating bearing, wherein a bracket is arranged on the rotating bearing, a rotating base is rotatably connected between the brackets, the video surveillance lens is connected to the rotating base, the rotating bearing is used to drive the video surveillance lens to rotate 360° in a vertical section, and the rotating base is used to drive the video surveillance lens to rotate 360° in a horizontal section, and a wireless transmission module is arranged at the bottom of the rotating bearing, and the wireless transmission module is arranged on the tunnel surrounding rock.
Citation Information
Cited By
Periscopic vault monitoring system
CN119964088A
Tunnel inrush water pressure monitoring method and system based on dual-acquisition multi-modal data
CN121113176A
