Non-contact and single contact tunnel face geophone and data processing method

The seismic signal of the palm surface of the tunnel is extracted through high-speed cameras and video vibration analysis methods, and combined with a single contact detector, the problem of difficult to arrange contact detectors in tunnel construction is solved, and a low-cost, safe and efficient high signal-to-noise ratio is achieved. Seismic signal acquisition fills the data loss in the field of tunnel earthquake advance forecasting.

CN118033727BActive Publication Date: 2024-11-01CHONGQING JIAOTONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410202917.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-02-23
Publication Date
2024-11-01
Estimated Expiration
2044-02-23

AI Technical Summary

Technical Problem

In tunnel construction, traditional contact detectors are difficult to be safe and efficiently arranged on the palm surface, resulting in the lack of three-dimensional seismic signals at the palm surface position. The existing video vibration analysis method is mainly used for structural deformation detection and has not been used in tunnel earthquakes. Pre-predictions.

Method used

Design a non-contact tunnel palm surface detector system, using a high-speed camera to capture palm surface vibration information, extract wide-band seismic signals with high signal-to-noise ratio from high-definition video data through video vibration analysis method, and combine with a single contact detector Data is collected under suitable geological conditions and cross-correlation calculations are performed to improve signal-to-noise ratio and bandwidth.

Benefits of technology

It realizes the fast acquisition of high signal-to-noise ratio broadband seismic signals at the palm surface position in a low-cost, safe and efficient contact-free manner, which supplements the data loss in the field of tunnel earthquake advance forecasting and improves the signal-to-noise ratio and bandwidth.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118033727B_ABST
    Figure CN118033727B_ABST
Patent Text Reader

Abstract

This invention discloses a non-contact and single-contact tunnel face detector and a data processing method, including a high-speed camera deployed in front of the tunnel face during excavation to acquire high-speed photographic data V(x,y,t) for video vibration analysis of the tunnel face. The calculation process consists of three steps: First, V(x,y,t) is divided into spatial sub-bands corresponding to different directions θ and scales r; Second, the local vibration signal for each pixel, direction, and scale is calculated, and the vibration signals are combined into a single global vibration signal for the tunnel face through a series of averaging and alignment operations; Third, Gaussian filtering and bandpass filtering methods are applied to remove clutter interference to obtain the recovered vibration signal. This invention deploys detectors at the tunnel face in a low-cost, safe, and efficient manner using both non-contact and single-contact methods, and proposes corresponding data processing methods to quickly acquire broadband seismic signals with high signal-to-noise ratios at the tunnel face location.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of tunnel earthquake advance detection and engineering survey, and in particular to a non-contact and single-contact tunnel face geophone and a data processing method. Background Art

[0002] Due to the limitations of the tunnel construction environment, it is very difficult to deploy contact detectors on the tunnel face. If the detectors can be deployed at the tunnel face in a low-cost, safe, efficient and non-contact manner to directly collect seismic data at the tunnel face, it will promote the breakthrough development of tunnel earthquake advance prediction methods. Video vibration analysis technology is an emerging detection technology with the advantages of full-scale surface, non-contact, no load effect, strong environmental adaptability, and no equipment loss. It is very suitable for tunnel face vibration detection.

[0003] High-speed cameras can capture the vibration information on the surface of objects. At present, video vibration analysis methods are mostly used for structural deformation detection (Liu Min, 2009; Hu Chaohui, 2010; Yuan Xiangrong, 2013). In recent years, researchers have proposed a sparse local key point detection algorithm based on deep learning (DeTone et al., 2018; Dusmanu et al., 2019). After the key points are detected, they are tracked in each frame of the video clip to extract the displacement information of the target object. It has been applied to estimate the displacement of the first two main modes of simple Oberst beams and helicopter blades (Morlier et al., 2010), and has also been extended to traffic signal structures and civil engineering structures. Full-field three-dimensional vibration displacement detection without target objects (Bartilson et al., 2015; Shao et al., 2021). In order to solve the problem that traditional structural measurement cannot extract high-frequency vibration signals, Luo et al. (2022) proposed an improved zero-mean normalized difference sum of squares algorithm.

[0004] Existing video vibration analysis methods are mostly used for structural deformation detection, but there are no related research reports in the field of tunnel earthquake advance prediction. At present, contact geophones are mainly used in the field of tunnel earthquake advance prediction (Wang et al., 2019; Lu et al., 2020; Lu et al., 2022). Due to tunnel construction requirements and safety considerations, a large number of contact geophones cannot be deployed at the tunnel face, resulting in a lack of three-dimensional seismic signals at the tunnel face.

[0005] The present invention aims to design a non-contact tunnel face video geophone and data processing method for advance prediction of tunnel earthquakes, to deploy video geophones on the tunnel face in a low-cost, safe and efficient non-contact manner, and to propose a corresponding data processing method to quickly obtain broadband seismic signals with a high signal-to-noise ratio at the tunnel face position.

[0006] References

[0007] Bartilson, D. T., Wieghaus, K. T., Hurlebaus, S., 2015. Target-less computer vision for traffic signal structure vibration studies[J]. Mech Syst Sig Process, 60 - 61: 571 - 82。

[0008] DeTone, D., Malisiewicz, T., Rabinovich, A., 2018. Superpoint: Self-supervised interest point detection and description[C]. In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops, 224 - 36。

[0009] Dusmanu, M., Rocco, I., Pajdla, T., et al., 2019. D2-net: A trainable cnn for joint description and detection of local features[C]. In: Proceedings of the IEEE / CVF conference on computer vision and pattern recognition, 8092 - 101。

[0010] Lu, X. L., Hu, X. Q., Xu, Z. Y., Liao, X., Liu, L. H., Fu, Z. H., 2022. Tunnel concealed karst cave joint detection by tunnel seismic and transient electromagnetic[J]. Lithosphere, 1: 2827582。

[0011] Lu,X.L.,Liao,X.,Wang,Y.,Wang,G.M.,Fu,Z.H.,Tai,H.M.,2020.The tunnelseismic advanceprediction method with wide illumination and a high signal-to-noise ratio[J].GeophysicalProspecting,68(8):2444-2458。

[0012] Luo,J.,Liu,B.Y.,Yang,P.,et al.,2022.High-speed vision measurement ofvibration based on animproved ZNSSD template matching algorithm[J].SystemsScience&Control Engineering,

[0013] 10(1):43-54。

[0014] Morlier,J.,Michon,G.,2010.Virtual vibration measurement using KLTmotion trackingalgorithm[J].J Dyn Syst Meas Contr.,132(1):011003。

[0015] Shao,Y.D.,Li,L.,Li,J.,et al.,2021.Computer vision based target-free3D vibration displacementmeasurement of structures[J].Engineering Structures,246:113040。

[0016] Wang,Y.,Fu,N.Y.,Lu,X.L.and Fu,Z.H.,2019.Application of a new geophoneand geometry intunnel seismic detection.Sensors,19,1246。

[0017] Hu Chaohui. 2010. Research on the application of digital image processing technology in bridge structure vibration testing [D]. Guangzhou: Guangzhou University. Liu Min. 2009. Research on the application of digital image processing technology in bridge structure detection [D]. Guangzhou: Guangzhou University.

[0018] Yuan Xiangrong. 2013. Orthogonal polynomial fitting for edge identification and beam deformation detection[J]. Laboratory Research and Exploration, 32(10):11-13. Summary of the invention

[0019] In view of at least one defect of the prior art, the object of the present invention is to provide a non-contact face geophone for advance prediction of tunnel earthquakes, in which geophones are deployed on the tunnel face in a low-cost, safe and efficient non-contact manner to quickly acquire broadband seismic signals with a high signal-to-noise ratio at the tunnel face position.

[0020] In order to achieve the above object, the present invention adopts the following technical solution:

[0021] A non-contact tunnel face detector observation system includes a high-speed camera, which is arranged in front of the tunnel face being excavated and directly faces the tunnel face. The distance between the high-speed camera and the tunnel face is dx. The high-speed photography data observed by the high-speed camera is used for later data analysis.

[0022] dx is the distance from the camera to the palm face, which is generally between 10m and 20m.

[0023] The non-contact tunnel face detector observation system has a high-speed camera recording interval of 2000 frames per second, a resolution of not less than 0.1 mm, and a stable observation time M of the high-speed camera facing the tunnel face.

[0024] M is about 5 minutes to 10 minutes. The high-speed camera includes a camera head, the camera head is connected to a graphic generator, and the graphic generator is connected to a graphic display terminal.

[0025] The data processing method of the non-contact tunnel face geophone observation system is mainly: collecting high-speed photography data for tunnel face video vibration analysis; high-speed photography data is represented by V(x, y, t), where x, y are the horizontal and vertical coordinates of the pixel point respectively, and t is time; the purpose is to recover the signal s(t) from V(x, y, t); the calculation process is divided into three steps:

[0026] Step 1: Decompose V(x,y,t) into spatial subbands corresponding to different directions θ and scales r; each scale r and direction θ is a complex image, which can be expressed by amplitude A and local phase To express;

[0027]

[0028] i represents an imaginary number.

[0029] Using the local phase calculated in equation (1) and subtract them from the local phase of the reference frame to t0, which is the first frame of the video, to calculate the phase change;

[0030]

[0031] For small vibrations, these phase changes are roughly proportional to the displacement of the image structure along the corresponding direction and scale;

[0032] Step 2: Calculate the local vibration signal of each pixel, direction and scale, and combine these vibration signals into a single global vibration signal of the tunnel face through a series of averaging and alignment operations;

[0033] For video signals, a “pyramid” can be used to decompose the image into each direction θ and scale r, and the spatial weighted average of the local motion signal is calculated to generate a single motion signal φ j (r,θ,t), the specific calculation formula is:

[0034]

[0035] "Pyramid" means decomposing into different scales and directions step by step from top to bottom; this academic term appears in relevant papers and is a mature technology.

[0036] Calculate the average φ at different scales and directions j Before (r,θ,t), we need to temporarily align the calculation results of each scale and direction; we need to merge the two directions (x and y) from a single spatial scale; if y = -x, then the phase changes in the x and y directions are negatively correlated, but the sum is a constant; the aligned signal is φ(r j ,θ j ,tt j ), the specific calculation formula is:

[0037]

[0038] where j is the index of the scale-direction pair (r, θ), φ0(r0, θ0, t) is the scale and direction of a selected reference; T in formula (4) represents transposition, and the final vibration signal of the tunnel face is

[0039]

[0040] Finally, the vibration signal is classified into [-1, 1].

[0041] Step 3: Use Gaussian filtering and bandpass filtering methods to remove clutter interference and obtain the restored vibration signal.

[0042] The key to the data processing method of the non-contact tunnel face geophone observation system is that the vibration signal extracted from the high-speed photography data needs to be further extracted into a seismic signal of self-excitation and self-reception at the tunnel face; assuming that the total recording time is T, the observation record at a point C(x,y) on the tunnel face is S(x,y,T);

[0043] If the set window length is w, the number of iterations in the correlation calculation is N = T / w, and the calculation formula is:

[0044]

[0045] Where, l=1,2,.....,N, p=w×l,(w+1)×l,.....,(2w-1)×l,2×w×l; ψ c (x, y, t) is the function of multiplying the observation record S(x, y, p) at point C by S(x, y, p), iterating N times and normalizing;

[0046] According to formula (6), the vibration signals of all tunnel faces are calculated in turn.

[0047] A single contact tunnel face geophone observation system including the non-contact tunnel face geophone observation system is characterized in that the high-speed photographic data observed by the high-speed camera is stored in a computer for later data analysis; a single contact geophone G is arranged on the tunnel face, and the single contact geophone is connected to a computer to transmit its data to the computer. The drilling depth of the single contact geophone is 1.5m and the aperture is 50mm.

[0048] A data processing method for a single contact tunnel face geophone observation system including the data processing method for the non-contact tunnel face geophone observation system, the key of which is:

[0049] If the geological conditions of the tunnel face allow, a single contact geophone G is deployed. The contact geophone G can directly collect seismic signals with a high signal-to-noise ratio at the tunnel face. The seismic signal collected by the contact geophone G is G(x, y, p).

[0050] The vibration signal extracted from the high-speed photography data needs to be further extracted into the seismic signal of self-excitation and self-reception at the tunnel face; assuming that the total recording time is T, then the observation record at a point C(x,y) at the tunnel face is S(x,y,T);

[0051] If the set window length is w, the number of iterations in the correlation calculation is N = T / w, and the calculation formula is:

[0052]

[0053] Where, l=1,2,.....,N, p=w×l,(w+1)×l,.....,(2w-1)×l,2×w×l; ψ c (x, y, t) is the function of multiplying the observation record S(x, y, p) at point C by G(x, y, p), iterating N times and normalizing;

[0054] According to formula (7), the vibration signals of all tunnel faces are calculated in turn.

[0055] Significant effect: The present invention provides a non-contact and single-contact tunnel face geophone and data processing method, deploys geophones on the tunnel face in a low-cost, safe and efficient non-contact and single-contact manner, and proposes a corresponding data processing method to quickly obtain broadband seismic signals with a high signal-to-noise ratio at the tunnel face position. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is the layout diagram of the tunnel face video geophone observation system, G is the contact geophone, and dx is the distance from the camera to the tunnel face;

[0057] Figure 2 are video data and vibration curves: (a) is video data; (b) is vibration curve;

[0058] Figure 3 is the extracted displacement noise signal;

[0059] Figure 4 is the extracted pseudo-tunnel seismic signal;

[0060] Figure 5 The data processing method of the non-contact tunnel face geophone is a flow chart. DETAILED DESCRIPTION

[0061] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] like Figure 1-Figure 5As shown, the present invention relates to a non-contact tunnel face video geophone and signal analysis method for advanced tunnel earthquake prediction, which mainly include the following contents: 1) design of a non-contact tunnel face video geophone observation system; 2) a non-contact tunnel face video vibration analysis method is proposed to extract high-quality displacement signals from high-definition video data; 3) a correlation calculation method is proposed to extract vibration signals from high-quality displacement signals; 4) a single contact geophone on the tunnel face is designed to perform cross-correlation calculation with the entire extracted video vibration signal, and the frequency band of the low-frequency signal is increased by using high-frequency signal constraints, and finally a broadband vibration signal is extracted from the tunnel face high-speed photography data to realize the deployment of three-dimensional geophones on the tunnel face, thereby increasing the amount of three-dimensional seismic data at the tunnel face position.

[0063] The present invention provides a non-contact tunnel face video detector and signal analysis method for advance prediction of tunnel earthquakes. The main advantages are as follows: 1) Compared with traditional contact detectors, the non-contact tunnel face video detector can obtain three-dimensional seismic signals with high signal-to-noise ratio directly at the tunnel face at low cost, safely and quickly, and supplement the missing seismic signals at the tunnel face position; 2) When a single contact detector cannot be deployed on the tunnel face, the signal-to-noise ratio of the vibration signal is improved by increasing the observation time; 3) When the tunnel face allows the deployment of a single contact detector, the vibration signal of the single contact detector is used to perform cross-correlation calculation with the vibration signal of the three-dimensional tunnel face non-contact video detector, and the high-frequency signal is used to constrain the bandwidth of the low-frequency signal to extract the three-dimensional high-signal-to-noise ratio vibration signal of the tunnel face.

[0064] A non-contact tunnel face video detector and signal analysis method for advanced prediction of tunnel earthquakes, including the following contents:

[0065] Content 1): A non-contact tunnel face video detector observation system is proposed;

[0066] Content 2): A non-contact tunnel face video vibration analysis method is proposed;

[0067] Content 3): Propose a relevant calculation method to extract tunnel seismic signals from displacement noise signals;

[0068] Content 4): Single contact geophone signal constraints improve the non-contact video geophone signal, increasing the bandwidth and signal-to-noise ratio of the data.

[0069] The present invention provides a non-contact tunnel face video detector and signal analysis method for advance prediction of tunnel earthquakes. It mainly includes:

[0070] 1. A non-contact tunnel face video detector observation system is proposed. The tunnel face is stably observed for 5 to 10 minutes using the high-resolution characteristics of a high-speed camera to obtain tunnel face vibration information. Under the conditions of ensuring safety and tunnel construction, the present invention proposes to deploy a single contact detector on the tunnel face. The proposed observation system has high deployment efficiency, does not require drilling or only requires a single drilling, and takes up almost no tunnel face construction time. The non-contact tunnel face video detector is within the scope of patent protection.

[0071] 2. A non-contact tunnel face video vibration analysis method is proposed to extract the vibration signal of the tunnel face position from the high-speed photography video signal. The non-contact tunnel face video vibration analysis algorithm falls within the scope of patent protection.

[0072] 3. A related calculation method is proposed to extract the tunnel seismic signal at the tunnel face position from the displacement noise signal, and the vibration signal is converted into the tunnel seismic signal at the tunnel face position based on the video vibration analysis method, and a signal extraction method for improving the non-contact video geophone with a single contact geophone constraint is proposed. The related algorithm for extracting the tunnel seismic signal from the displacement noise signal is within the scope of patent protection.

[0073] The above invention content is currently mainly used in the field of tunnel engineering construction detection.

[0074] 1. Non-contact face video detector observation system

[0075] The present invention designs a non-contact face video detector. Figure 1 The figure shows the designed tunnel face video detector observation system. Symbol G is a single contact detector, and dx is the distance from the camera to the tunnel face, which is generally between 10m and 20m. The high-speed camera is arranged in front of the tunnel face being excavated, and stable observation is carried out for about 5 to 10 minutes. During the observation period, the tunnel face cannot be operated, and the lighting remains unchanged. The recording interval of the high-speed camera is 2000 frames / second, and the resolution is required to be not less than 0.1mm. The observed high-speed photography data is stored in a laptop or a memory card built into the high-speed camera for later data analysis. At the same time, in order to further improve the signal-to-noise ratio of the data, under the conditions of ensuring safety and tunnel construction permission, the present invention proposes to arrange a single contact detector on the tunnel face, with a drilling depth of 1.5m and an aperture of 50mm.

[0076] 2. Non-contact tunnel face video vibration analysis method

[0077] High-speed photography data is collected for non-contact face video vibration analysis. The data of the high-speed camera is V(x,y,t), and the frequency of the high-speed camera is 1kHz to 2kHz. x,y are the horizontal and vertical coordinates of the high-speed photography data pixels, and t is time. The goal is to recover the signal S(t) from V. The calculation process is mainly divided into three steps:

[0078] Step 1: We decompose the input video V into spatial subbands corresponding to different directions θ and scales r. Each scale r and direction θ is a complex image, which can be expressed as amplitude A and phase To express;

[0079]

[0080] We use the local phase calculated in this equation and subtract them from the local phase of the reference frame at t0 (usually the first frame of the video) to calculate the phase change;

[0081]

[0082] For small vibrations, these phase changes are roughly proportional to the displacement of the image structure along the corresponding direction and scale.

[0083] Step 2: Calculate the local vibration signal for each pixel, direction, and scale, and combine these vibration signals into a single global vibration signal of the tunnel face through a series of averaging and alignment operations.

[0084] For video signals, a pyramid decomposition can be used to decompose the image into each direction θ and scale r. We calculate the spatial weighted average of the local motion signals to generate a single motion signal φ j (r,θ,t), the specific calculation formula is:

[0085]

[0086] Calculate the average φ at different scales and directions j Before we can calculate the phase shift in the x and y directions, we need to temporarily align the results of the calculations at each scale and direction. To combine two directions (x and y) from a single spatial scale, we need to temporarily align the results of the calculations at each scale and direction. If y = -x, then the phase shifts in the x and y directions are negatively correlated, but the sum is a constant. The aligned signal is φ(r j ,θ j ,tt j ), the specific calculation formula is:

[0087]

[0088] where j is the index of a scale-direction pair (r,θ) and φ0(r0,θ0,t) is the scale and direction of a selected reference.

[0089] The final vibration signal of the tunnel face is

[0090]

[0091] Finally, the vibration signal is classified into [-1, 1].

[0092] Step 3: Use Gaussian filtering, bandpass filtering and other methods to remove clutter interference and obtain the restored vibration signal.

[0093] 3. Propose a related calculation method to extract tunnel seismic signals from displacement noise signals

[0094] The vibration signal extracted from the high-speed photography data is a disordered noise signal and cannot be directly used for tunnel seismic advance detection imaging. It needs to be further extracted as a seismic signal that is self-excited and self-received at the tunnel face. The present invention proposes a related calculation method to realize the seismic signal extraction of the tunnel face. Assuming that the total recording time is T, the observation record at a point C(x,y) on the tunnel face is S(x,y,T).

[0095] If the set window length is w, the number of iterations in the correlation calculation is N = T / w, and the calculation formula is:

[0096]

[0097] Among them, l=1,2,...,N, p=w*l, (w+1)*l,..., (2w-1)*l, 2*w*l. ψ c (x, y, t) is the function obtained by multiplying the observation record S(x, y, p) at point C by S(x, y, p), iterating N times and normalizing it.

[0098] According to formula (6), the vibration signals of all tunnel face positions are calculated in sequence.

[0099] 4. Single contact geophone signal constraints improve the non-contact video geophone signal, improve the bandwidth and signal-to-noise ratio of the data, mainly including the following:

[0100] If the geological conditions of the tunnel face allow the deployment of a single contact geophone, the seismic signal with a high signal-to-noise ratio at the tunnel face can be directly collected. Formula (6) can be changed to:

[0101]

[0102] G(x,y,p) is the seismic signal collected by the contact geophone.

[0103] Engineering Experiment

[0104] In order to verify the accuracy of the signal analysis method of the non-contact face video detector in the present invention, a high-speed photography acquisition experiment was carried out with a light rail bridge in Chongqing as the target. The acquisition frequency was 1000Hz, and a total of 0.621s long records were collected. The original images collected and the extracted vibration records are shown in Figure 2. Figure 2 The non-contact tunnel face video vibration analysis method proposed in this invention is used to extract the displacement noise signal. The results are shown in Figure 3 As shown in Figure 1, the amplitude of the noise signal is between [-1,1]. On this basis, the related calculation method is used to extract the self-excited and self-received pseudo-seismic signal. The results are shown in Figure 1. Figure 4 As shown, the effective signal is concentrated between 0.02s and 0.1s, which can be used for imaging processing of later data.

[0105] Finally, it should be noted that the above examples are only specific implementation examples of the present invention. Of course, those skilled in the art can make changes and modifications to the present invention. If these modifications and variations fall within the scope of the claims of the present invention and their equivalent technologies, they should be considered to be within the scope of protection of the present invention.

Claims

1. A data processing method for a non-contact tunnel face geophone observation system, wherein the non-contact tunnel face geophone observation system comprises a high-speed camera, which is arranged in front of a tunnel face being excavated and directly faces the tunnel face, and the distance between the high-speed camera and the tunnel face is dx, and the high-speed photography data observed by the high-speed camera is used for later data analysis; Features: High-speed photography data is collected for tunnel face video vibration analysis; high-speed photography data is represented by V(x, y, t), where x and y are the horizontal and vertical coordinates of the pixel point, respectively, and t is time; the purpose is to recover the signal s(t) from V(x, y, t); the calculation process is divided into three steps: Step 1: Decompose V(x,y,t) into spatial subbands corresponding to different directions θ and scales r; each scale r and direction θ is a complex image, which can be expressed by amplitude A and local phase To express; Using the local phase calculated in equation (1) and subtract them from the local phase of the reference frame to t 0 , t 0 is the first frame of the video to calculate the phase change; Step 2: Calculate the local vibration signal of each pixel, direction and scale, and combine these vibration signals into a single global vibration signal of the tunnel face through a series of averaging and alignment operations; For video signals, a "pyramid" can be used to decompose them into images of each direction θ and scale r, and the spatial weighted average of the local motion signal is calculated to generate a single motion signal φ j (r,θ,t), the specific calculation formula is: Calculate the average φ at different scales and directions j Before (r, θ, t), we need to temporarily align the calculation results of each scale and direction; we need to merge two directions from a single spatial scale, namely x and y; if y = -x, then the phase changes in the x and y directions are negatively correlated, but the sum is a constant; the aligned signal is φ(r j ,θ j ,tt j ), the specific calculation formula is where j is the index of a scale-direction pair (r,θ), φ 0 (r 0 ,θ 0 , t) is the scale and direction of a selected reference; the final vibration signal of the tunnel face is Finally, the vibration signal is classified between [-1,1]; Step 3: Use Gaussian filtering and bandpass filtering methods to remove clutter interference and obtain the restored vibration signal; The vibration signal extracted from the high-speed photography data needs to be further extracted into the seismic signal of self-excitation and self-reception at the tunnel face; assuming that the total recording time is T, then the observation record at a point C(x,y) at the tunnel face is S(x,y,T); If the set window length is w, the number of iterations in the correlation calculation is N = T / w, and the calculation formula is: Where, l=1,2,.....,N, p=w×l,(w+1)×l,.....,(2w-1)×l,2×w×l; ψ c (x, y, t) is the function obtained by multiplying the observation record S(x, y, p) at point C by S(x, y, p), iterating N times and normalizing. According to formula (6), the vibration signals of all tunnel faces are calculated in turn.

2. A data processing method for a single contact tunnel face geophone observation system comprising the data processing method of claim 1, Features: If the geological conditions of the tunnel face allow, a single contact geophone G is deployed. The contact geophone G can directly collect seismic signals with a high signal-to-noise ratio at the tunnel face. The seismic signal collected by the contact geophone G is G(x, y, p). The vibration signal extracted from the high-speed photography data needs to be further extracted into the seismic signal of self-excitation and self-reception at the tunnel face; assuming that the total recording time is T, then the observation record at a point C(x,y) at the tunnel face is S(x,y,T); If the set window length is w, the number of iterations in the correlation calculation is N = T / w, and the calculation formula is: Where, l=1,2,.....,N, p=w×l,(w+1)×l,.....,(2w-1)×l,2×w×l; ψ c (x, y, t) is the function obtained by multiplying the observation record S(x, y, p) at point C by G(x, y, p), iterating N times and normalizing. According to formula (7), the vibration signals of all tunnel faces are calculated in turn.