Method and system for monitoring tunnel face of mountain tunnel

Abstract

To provide a method and system for monitoring a tunnel face of a mountain tunnel, which can improve the accuracy of measuring the parallax or the distance to the tunnel face while using existing photographing equipment and can more reliably detect abnormalities such as collapse of the tunnel face, falling of the tunnel face, and falling rocks.SOLUTION: In a method for monitoring a tunnel face of a mountain tunnel by constantly photographing the tunnel face and detecting any abnormalities in the tunnel face from changes in the photographed images, a stereo camera is used to photograph the tunnel face and derive a parallax value, the tunnel face being photographed is divided into areas having a predetermined area, the parallax values are averaged for each area over a predetermined period of time, and any abnormalities in the tunnel face are detected from changes in this averaged parallax value.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a method and system for monitoring the face of a mountain tunnel, which are used to detect dangers such as collapse and skin falls at the face of a mountain tunnel and prevent accidents before they occur. [Background technology]

[0002] At tunnel excavation sites, it is necessary to constantly monitor the tunnel face to prevent accidents due to collapse, etc. Specifically, national guidelines require that the person responsible for monitoring the tunnel face constantly monitor the tunnel face for deformation, the occurrence of cracks, the presence of spring water, the deterioration of the rock, etc., in order to detect any collapse of the tunnel face. When it comes to monitoring by the face monitoring manager, visual monitoring has its limitations, so a method that utilizes a monitoring system using IT technology is also used. The following technologies have been proposed as tunnel face monitoring systems:

[0003] The technology described in Patent Document 1 involves "carrying into a tunnel a photographing unit equipped with a rotatably controlled stereo camera device consisting of multiple cameras fixed at a fixed interval, photographing the internal cross section before concrete is sprayed with the photographing unit, inputting the stereo photographed data into a computer and calculating the internal cross section shape based on the principle of stereo image measurement, and then calculating the planned internal cross section based on the internal cross section shape, and obtaining data on the planned internal cross section shape." This technology is not intended for monitoring the tunnel face, but is intended to grasp the internal cross section shape inside the tunnel using a stereo camera.

[0004] As mentioned above, the technology described in Patent Document 1 uses a stereo camera to photograph the internal cross section and obtain shape data using the principles of stereo image calculation. However, general stereo cameras are prone to measurement errors and insufficient resolution, and even if shape data is calculated based on this, the accuracy is insufficient, particularly from the perspective of monitoring the tunnel face.

[0005] It is possible to improve the accuracy of shape data by improving the hardware performance of the stereo camera, i.e., by improving its precision and resolution, but this would entail problems such as high costs and long development times.

[0006] Patent Document 2 proposes a monitoring device and a monitoring method for inside a tunnel using a TOF (Time Of Flight) camera. In other words, it is "a monitoring device that detects an abnormality that occurs on the surface of a measurement object and issues an alarm, comprising: a distance measuring unit that is installed facing the surface and that measures and outputs the distance between itself and multiple measurement points on the surface non-contact, and a control unit that causes the distance measuring unit to perform measurements at multiple points in time for a measurement object area, which is a specified area on the surface, and determines that there is an abnormality in the measurement object based on changes in the statistical values ​​of the distance at different points in the measurement point area," and this configuration uses a TOF camera as the distance measuring unit.

[0007] Although a certain degree of accuracy can be obtained by using a TOF camera, there is a problem of high cost. Also, TOF cameras may not be able to perform well in extremely bright environments such as under sunlight. Tunnel faces are illuminated with fairly high-intensity lighting to improve monitoring and workability, and there is a problem that TOF cameras may not be able to perform well in this environment. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] Patent Publication No. 2003-013699 [Patent Document 2] Patent Publication 2019-124659 Summary of the Invention [Problem to be solved by the invention]

[0009] Therefore, an object of the present invention is to provide a method and system for monitoring the tunnel face of a mountain tunnel, which can improve the accuracy of measuring the parallax or the distance to the tunnel face while using existing photographing equipment, and can more reliably detect abnormalities such as collapse of the tunnel face, falling skin, and falling rocks. [Means for solving the problem]

[0010] The above object of the present invention can be achieved by the following means.

[0011] 1. A method for monitoring the tunnel face of a mountain tunnel by constantly photographing the tunnel face and detecting any abnormalities in the face from changes in the photographs. A stereo camera is used to photograph the face and derive the disparity value. Dividing the face of the subject to be photographed into sections having a predetermined area, averaging the parallax values ​​for each section over a predetermined time period, This method for monitoring the tunnel face of a mountain tunnel is characterized by detecting abnormalities at the tunnel face based on changes in this averaged disparity value.

[0012] 2. Calculate the distance to the face from the averaged disparity value. A method for monitoring a mountain tunnel face according to claim 1, characterized in that an abnormality at the face is detected based on a change in the calculated distance.

[0013] 3. The area of ​​the region is adjusted according to the size of the object whose displacement can be detected; A method for monitoring a mountain tunnel face as claimed in claim 1 or 2, characterized in that the area of ​​the region is reduced in order to improve the sensitivity indicating the detectable size of the object.

[0014] 4. A method for monitoring a mountain tunnel face as described in claim 1 or 2, characterized in that the area of ​​the region subject to averaging is increased in order to improve the measurement accuracy of the disparity value or the distance calculated from the disparity value.

[0015] 5. A method for monitoring a mountain tunnel face as described in claim 1 or 2, characterized in that the time period covered by averaging is extended in order to improve the measurement accuracy of the disparity value or the distance calculated from the disparity value.

[0016] 6. A method for monitoring the face of a mountain tunnel as described in claim 1 or 2, characterized in that the parameter of the disparity values ​​for averaging is 13,000 or more.

[0017] 7. A system for monitoring the tunnel face of a mountain tunnel by constantly measuring the distance to the tunnel face and detecting abnormalities at the tunnel face from changes in that distance. taking images of the tunnel face using a stereo camera and obtaining a disparity value; Dividing the face of the subject to be photographed into areas having a predetermined area; averaging the disparity values ​​for each of the regions over a predetermined time period; detecting an abnormality in the face based on a change in the averaged disparity value; A mountain tunnel face monitoring system comprising:

[0018] 8. Calculating the distance to the face from the averaged disparity value; detecting an abnormality in the face based on a change in the calculated distance; A mountain tunnel face monitoring system as claimed in claim 7, characterized in that it comprises:

[0019] 9. A mountain tunnel face monitoring system as described in claim 7 or 8, characterized in that it includes a process for issuing an alarm to the area around the face when an abnormality at the face is detected. Effect of the Invention

[0020] According to the above 1 or 2 inventions, the accuracy can be improved by using a statistical method using an existing stereo camera without improving the accuracy of the hardware. Therefore, the accuracy of the distance measurement to the face can be improved without spending the cost and time required for improving the accuracy of the hardware, and the accuracy of the monitoring can be improved.

[0021] When measuring the distance to the face using an existing stereo camera, the accuracy of the stereo camera affects the distance, resulting in a relatively large variation in the obtained parallax value, making it impossible to grasp the accurate distance. In addition, irregular measurement results may occur due to other machine tools or workers being captured in the image, and these noise components may be an obstacle to grasping the accurate distance.

[0022] The inventions 1 and 2 above utilize statistical methods, so the more data there is to serve as parameters, the more accurate the results can be. As a means of increasing data, the parameters are increased using two parameters, area and time, and then averaged to improve accuracy.

[0023] First, regarding the area, the face, which is the subject of the stereo camera photography, is divided into areas each having a predetermined area, and each area is treated as a unit. For each divided area, the multiple disparity values ​​obtained within this area are averaged. Next, regarding time, a predetermined time is taken as one unit, and a plurality of disparity values ​​obtained in the predetermined time for each of the above-mentioned areas are averaged. That is, a plurality of pieces of disparity value data obtained in a certain area at a predetermined time are averaged to obtain a highly accurate disparity value.

[0024] By averaging multiple disparity values ​​obtained for the two parameters of area and time mentioned above, the accuracy of this disparity value can be improved, and by calculating the distance to the face from this highly accurate disparity value, the accuracy of the calculated distance data to the face is also improved. This method makes it possible to accurately grasp changes in disparity or fluctuations in the distance to the face, that is, even the subtle behavior of the face, so that abnormalities at the face (meaning a situation in which collapse, skin falls, or rock falls occur due to the collapse of soil or rocks from the face, etc.; the same applies hereinafter in this specification) can be detected based on accurate and practical data from which noise and other components have been removed.

[0025] According to the third aspect of the present invention, the sensitivity indicating the detectable size of an object (such as earth and sand, falling rocks, etc.) displacing in the face to be monitored can be increased by reducing the area of ​​the zone. In other words, by increasing the area of ​​the zone and decreasing the sensitivity, the objects whose displacement can be detected become larger, and small objects cannot be detected.

[0026] According to the fourth aspect of the present invention, the measurement accuracy of the disparity value or the distance calculated from the disparity value can be adjusted by changing the area of ​​the area to be averaged. That is, the measurement accuracy can be improved by increasing the area of ​​the area. This is because the data parameter (disparity value) for averaging can be increased by increasing the area of ​​the area.

[0027] According to the fifth aspect of the present invention, the measurement accuracy of the disparity value or the distance calculated from the disparity value can be adjusted by changing the time period to be averaged. That is, the measurement accuracy can be improved by lengthening the time period. This is because the data population (disparity value) for averaging can be increased by lengthening the time period. Combined with the area of ​​the four regions above, the number of data points for averaging can be increased, and the larger this number, the higher the measurement accuracy.

[0028] According to the sixth aspect of the present invention, by setting the data parameter of the averaged disparity value to 13,000 or more, it is possible to obtain the measurement accuracy required for detecting an abnormality in the face.

[0029] According to the inventions 7 and 8 above, similarly to the method according to the invention 1 above, it is possible to improve accuracy by using an existing stereo camera and using a statistical method without improving the accuracy of the hardware. Therefore, it is possible to improve the accuracy of the measurement of the parallax value and, in turn, the accuracy of the measurement of the distance to the face, without spending the cost and time required for improving the accuracy of the hardware, and thus to improve the accuracy of the monitoring.

[0030] According to the invention No. 9 above, when an abnormality is detected at the face, an alarm is sounded to personnel around the face, urging them to take action such as evacuation, and even if an accident such as a collapse or a landslide occurs, damage to personnel can be prevented or reduced. [Brief description of the drawings]

[0031] [Figure 1] FIG. 1 is a schematic diagram showing an embodiment of a mountain tunnel face monitoring method according to the present invention; [Diagram 2] Diagram explaining the principle of distance measurement using a stereo camera [Diagram 3] Normal distribution graph showing the variation of disparity values [Figure 4] A schematic diagram showing the state in which the face to be photographed is divided into predetermined areas. [Diagram 5] A simplified normal distribution graph visually showing the averaging of disparity values ​​in the area [Figure 6] A simplified normal distribution graph visually showing the averaging of disparity values ​​over time [Figure 7] A schematic diagram explaining the relationship between the area of ​​area A and the size of object Y. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0032] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will now be described in detail with reference to the accompanying drawings. Fig. 1 is a schematic diagram showing an embodiment of a mountain tunnel face monitoring method or monitoring system according to the present invention. As shown in Fig. 1, the device required to implement the present invention is a stereo camera 1, to which a computer 2 (not shown in Fig. 1, but shown in Fig. 4) is connected.

[0033] A stereo camera is a photographing device that can simultaneously photograph an object from multiple different directions and record information about the depth direction, i.e., the distance to the object. Specifically, it has a configuration in which two lenses are arranged in parallel, and the object can be photographed simultaneously through these lenses. There is no limitation on the configuration of the stereo camera 1 used in the present invention, and any publicly known stereo camera can be used without any special restrictions. In addition, means for simultaneously capturing images using two cameras and measuring distances by reproducing the principle of a stereo camera are also included in the scope of the present invention.

[0034] The principle of the stereo camera will be explained with reference to FIG. 2, when an object X is photographed simultaneously from two lenses 11, a disparity (disparity value) D, which is the difference in the positions at which the object X is photographed, can be obtained. From this obtained disparity value D, the distance Z to the object X can be derived using the principle of triangulation. The baseline length B, which is the distance between the lenses, and the focal length F are known from the specifications of the stereo camera 1. Using these values ​​and the previously obtained value of parallax D, the distance Z to the object X can be calculated using the formula Z=B×F / D.

[0035] The above-mentioned disparity value D has measurement error (variation) depending on the performance and specifications of the stereo camera. The reason for this variation is as follows.

[0036] Most stereo cameras currently on the market are digital cameras, so the above-mentioned disparity D is calculated by multiplying the number of pixels N that are displaced (changed) in the left and right images by the pixel size P. The pixel size P depends on the performance of the CMOS sensor mounted on the stereo camera. For example, when using a stereo camera with baseline length B: 100 mm, focal length F: 4 mm, and pixel size P: 3.75 μm, if a distance Z: 10 m is measured, Z:10m=B:100mm×F:4mm / D(N×P:3.75μm) The number of pixels changed, N, is calculated as 10.67 pixels using the formula above. Since the number of pixels N must be an integer, it can be assumed that when measuring a distance Z of 10 m, the number of pixels N is processed as 10 or 11 pixels.

[0037] So, when calculating the distance Z when the number of pixels N is 10 pixels and when it is 11 pixels, first, when the number of pixels N is 10 pixels, Z=B×F / D=100mm×4mm / 10×3.75μm=10.666m Next, for 11 pixels, Z=B×F / D=100mm×4mm / 11×3.75μm=9.696m The distance Z calculated by these two methods has an error of 9.696 to 10.666 m per pixel, which is 3.3% to 6.6%, and it can be seen that the difference is about 5% on average.

[0038] For example, if a stereo camera with a baseline length B of 100 mm, focal length F of 4 mm, and pixel size P of 3.75 μm is used, and a distance Z of 7 m is measured, Z:7m=B:100mm×F:4mm / D(N×P:3.75μm) The number of pixels changed, N, is calculated as 15.23 pixels using the formula. Similarly, since the number of pixels N must be an integer, it can be assumed that when the distance Z is measured at 7 m, the number of pixels N is processed as 15 or 16 pixels.

[0039] So, when calculating the distance Z when the number of pixels N is 15 pixels and when it is 16 pixels, first, when the number of pixels N is 10 pixels, Z=B×F / D=100mm×4mm / 15×3.75μm=7.111m Next, for 11 pixels, Z=B×F / D=100mm×4mm / 16×3.75μm=6.666m The difference between the distances Z calculated by the two methods is 6.666 to 7.111 m per pixel, which is 1.6% to 4.8%, and on average it is found that the difference is about 3.2%.

[0040] Using the example above, if the distance to object X (face K) is 7 m and the distance accuracy is 3%, then when measuring the distance to object X at 7 m, there may be a variation in the parallax value D of 7 m (7000 mm) × 3% = 210 mm (see Figure 3).

[0041] In cases where it is necessary to detect even minute changes in monitoring the tunnel face, as in the present invention, the above-mentioned 210 mm variation has a large impact and is not sufficient to detect the behavior or abnormalities of the tunnel face. It is possible to minimize this variation by improving the hardware accuracy of the stereo camera used, but in this case, it is necessary to develop a new high-precision stereo camera, and specifically, it is necessary to install an ultra-multiple pixel sensor camera to increase the resolution and to make improvements such as extending the baseline length, which will require large development costs and a long development period.

[0042] Therefore, an object and purpose of the present invention is to improve the accuracy of depth distance by a means other than the improvement of accuracy of a stereo camera.

[0043] The present invention is characterized by its configuration that minimizes the variation in the obtained disparity value D and enables accurate measurement of the distance Z to the object X by utilizing a statistical method while using a general-purpose stereo camera. Because statistical methods are used, the more data (parameters), the higher the accuracy. In other words, the more disparity values ​​D are obtained, the more consistent the disparity values ​​used to calculate distance Z will be, and the higher the accuracy of measuring the distance to object X.

[0044] In order to increase the data parameter, a certain range is given to the area and time, and the disparity values ​​D obtained between them are averaged to obtain an accurate disparity value D' with little variation. It is also possible to remove noise components. In this specification, the disparity value before averaging (the disparity value as the data parameter) is represented by the symbol "D", and the disparity value after averaging is represented by the symbol "D'".

[0045] First, as for the area, as shown in an example in Fig. 4, a face K, which is an object (object X) to be photographed by a stereo camera 1, is divided into areas A each having a predetermined area, and each divided area is regarded as one unit. For each divided area A, a plurality of disparity values ​​D obtained within the area A are averaged. In other words, the effect of measurement error, noise, etc. is minimized by averaging the disparity values ​​on an area-by-area basis.

[0046] Next, regarding time, a predetermined time is taken as one unit, and a plurality of disparity values ​​D obtained in the predetermined time for each of the above-mentioned areas A are averaged. That is, a configuration is adopted in which a plurality of pieces of disparity value D data obtained in a certain area over a certain time period are averaged to obtain a highly accurate disparity value D'. In other words, the configuration is one in which the influence of noise is minimized by averaging the disparity values ​​over a certain time period in area units. The frame rate is known due to the specifications of the stereo camera used. For example, if the frame rate is 60 fps (frames per second), 60 frames can be captured per second. Therefore, the number of disparity values ​​D, which is the parameter for averaging, can be calculated in relation to the disparity value obtained in area A per frame.

[0047] By averaging multiple disparity values ​​D obtained for the two parameters of area and time described above, the accuracy of the disparity value D' can be improved, and by calculating the distance Z to the face (=B×F / D') from this highly accurate disparity value D', the accuracy of the calculated distance data to the face can also be improved. This method makes it possible to accurately grasp the fluctuation of the distance Z to the face K, i.e., the behavior of the face, and therefore makes it possible to detect abnormalities at the face based on accurate and practical data from which noise and other components have been removed.

[0048] As mentioned above, the greater the data parameter for averaging, the more the measurement accuracy can be improved. Therefore, by setting the area of ​​area A to be large and the specified time to be long, this parameter can be increased and the measurement accuracy can be improved. It is preferable to adjust the area of ​​area A and the specified time so that the data parameter is 13,000 or more. This is believed to enable highly accurate detection of face abnormalities.

[0049] Graphs visually showing the above-mentioned averaging of the disparity values ​​with respect to the area and time in the normal distribution graph showing the variation in the disparity values ​​shown in FIG. 3 are shown in FIGS. As shown in Fig. 5, values ​​far from the median in the normal distribution graph are excluded from the averaging parameter, which minimizes the effect of noise components. In addition, values ​​close to the median (average value) that are not excluded are averaged as the parameter, so the disparity value D' obtained by this averaging is a highly accurate value. The range of values ​​far from the median that are excluded can be set arbitrarily based on the degree of variation (standard deviation) from the average value. In the present invention, the distance K between the face K and the stereo camera 1 is known to some extent, so the range where the distance is clearly different may be excluded.

[0050] 6 is a graph visually showing the averaging over a predetermined time, and shows how the average disparity value in a certain area A is further averaged over time. The longer the predetermined time, the more disparity values ​​D can be obtained, and therefore a more accurate disparity value D' can be obtained by averaging.

[0051] In order to accurately detect abnormalities in the face K, it is necessary to adjust the data parameter for averaging based on the area of ​​the above-mentioned area A and a predetermined time, and also to set the size of the object to be detected (detectable object). This object will be represented by the symbol Y. The object Y to be detected is soil, rocks, etc. that may be displaced inside the working face K that is the subject of photography, and is an object that may be displaced by peeling off or falling off due to natural or man-made factors. The larger the size of the object Y to be detected, the easier it is to detect, and the smaller it is, the more difficult it is to detect. In other words, the larger the size of the object Y, the more it can be detected even with a small amount of movement (synonymous with amount of displacement; the same applies below), and the smaller the size of the object Y, the more it cannot be detected without a large amount of movement.

[0052] Here, the detectable size of object Y displacing in the face, which is the object to be monitored, is defined as sensitivity. The smaller the detectable size of object Y, the higher the sensitivity, and conversely, the larger the size, the lower the sensitivity. In other words, in order to increase this sensitivity, it is necessary to set the area of ​​area A to be small.

[0053] Regarding the area of ​​area A, the area needs to be large in order to increase the averaging parameter (to improve the accuracy of measuring disparity and distance), but the area needs to be small in order to set a high sensitivity.

[0054] From the above, in order to detect abnormalities at the tunnel face with high accuracy, it is necessary to optimize three points: the area of ​​the averaging area A, the averaging time, and the size (sensitivity) of the object Y to be detected.

[0055] The relationship between the area of ​​the area A to be averaged and the size of the object Y to be detected can be expressed as follows: Here, the elements used in the formula are shown below, and are illustrated in FIG.

[0056] Z: Distance from stereo camera 1 to face K Z': Distance from stereo camera 1 to object Y Z): Average distance from stereo camera 1 to the entire area A D: Average disparity value for face K D': Average disparity value for object Y D): Average disparity value for the entire area A P: Number of pixels in area A P': Number of pixels of object Y S: Area of ​​area A S': Area of ​​object Y

[0057] In the above, the average disparity value D'' in a specific area A is calculated by the following formula.

[0058]

number

[0059] Next, the average distance Z'' from the stereo camera 1 to the entire area A can be calculated from Z''=B×f / D'' as described above, and can be calculated by the following formula.

[0060]

number

[0061] From the above formula, the distance Z' from the stereo camera 1 to the object Y can be calculated by the following formula.

[0062]

number

[0063] Therefore, the amount of movement of object Y that results in an average distance of 1 mm across the entire area A can be calculated using the following formula:

[0064]

number

[0065] It is believed that if a 1mm change in the behavior of the tunnel face can be detected, it will be possible to detect an abnormality in the tunnel face. If even minute abnormalities can be detected, it will be possible to identify the possibility of a serious accident such as a collapse before it occurs. The relationship between the size of object Y at which it can be determined that a specific area A has moved 1mm and the amount of movement of object Y, derived using the above formula, is as follows: In the table, the division setting refers to the number of divisions into which the subject to be photographed (photographable area) is divided; the greater the number of divisions, the smaller the area of ​​area A to be averaged, and the smaller the number of divisions, the larger the area of ​​area A to be averaged. For example, in the table, a division of 40 horizontal x 40 vertical means that the subject to be photographed is finely divided, i.e., the area of ​​area A is small, and a division of 15 horizontal x 15 vertical means that the subject to be photographed is roughly divided, i.e., the area of ​​area A is large.

[0066] [Table 1]

[0067] In Table 1 above, the areas shown in light grey indicate that the size (area) S' of object Y exceeds the area S of area A. Under the conditions shown in light grey, it can be determined that area A has moved 1 mm based on the relationship between the area of ​​area A and the area of ​​object Y. For example, when the division setting is 40 horizontal x 40 vertical, and the distance between the face K and the stereo camera 1 is 7 m, if the area S' exceeds the area S of area A by 1 mm, the area S of object Y by 1 mm is 400 cm. 2 If object Y with an area of ​​(20cm x 20cm) is displaced 0.5mm, it can be determined that area A has moved 1mm.

[0068] Next, the installation position of the stereo camera 1 will be described. It is preferable that the distance from the stereo camera 1 to the face K is about 6 to 7 m. This is because the general angle of view of a general-purpose stereo camera is about 50 degrees, and this angle of view and the size (area) of the entire face must be taken into consideration in order to capture the entire face. In addition, as shown in Fig. 1, work is being carried out near the face K by work vehicles C such as excavators (drill jumbos) and concrete sprayers, so it is necessary to position the stereo camera 1 so as not to interfere with this work, and for these reasons, it is preferable that the distance from the stereo camera 1 to the face K is 6 to 7 m.

[0069] Furthermore, if the imaging range of the stereo camera 1 cannot cover the entire face, it is preferable to use two or more stereo cameras 1 to widen the imaging range. In many cases, the imaging range in the height direction can be covered by one stereo camera 1 by setting the distance from the stereo camera 1 to the face K at 6 to 7 m, but in some cases, the imaging range in the width direction cannot be covered by one stereo camera 1 due to the size of the face K. Therefore, it is preferable to use two or more stereo cameras 1 to widen the imaging range, especially in the width direction.

[0070] In the explanation so far, the present invention has been mainly described as a method invention, but it can also be implemented as a mountain tunnel face monitoring system that executes the above-mentioned method using the stereo camera 1 and computer 2. In Fig. 4, the stereo camera 1, computer 2 and alarm device 3 are displayed to show one embodiment of a mountain tunnel face monitoring system.

[0071] The computer 2 is used to perform averaging processing and the like on information such as the parallax value D obtained by the stereo camera 1. There is no limitation on the computer 2 used in the present invention, and any known and publicly used computer can be used without any special restrictions. The computer in the present invention includes not only general desktop and notebook personal computers, but also highly portable computers such as tablet PCs and smartphones. The stereo camera 1 may also be an integrated type having a computer function.

[0072] The computer 2 is connected to the stereo camera 1 so as to be able to communicate with it, but there is no limitation on the communication means, and any publicly known and commonly used communication means, such as wired or wireless LAN, wired communication such as USB, wireless communication such as Bluetooth, etc., can be used without any special restrictions.

[0073] This computer 2 is used to average the disparity values ​​D obtained by the stereo camera using the above-mentioned means, and calculate the distance Z from the stereo camera 1 to the face K based on the disparity value D' obtained by a statistical method. It is also used to convert the face K into a fractal pattern and display it based on the obtained distance Z. As a result, it is also used to detect changes and abnormalities in the face K and to determine whether there is a risk of collapse or surface falls.

[0074] If the computer 2 detects an abnormality in the face K and determines that there is a risk of collapse, etc., the computer 2 may be configured to issue an alarm to personnel working around the face K using an alarm function provided in the computer 2 or an alarm device 3 connected to the computer 2. [Explanation of symbols]

[0075] 1 Stereo camera 11 Lens B Baseline length F focal length D Parallax (parallax value) 2. Computer 3 Alarm device K Face (face) W Excavation surface or spray surface C Work vehicles (concrete sprayers, drill jumbos, etc.) Area A (1 unit) Z Distance from stereo camera to face X Object Y Object (earth, sand, rocks, etc., that may displace the face being photographed)

Claims

1. A method for monitoring the tunnel face of a mountain tunnel by constantly photographing the tunnel face and detecting any abnormalities in the face from changes in the photographs. A stereo camera is used to photograph the face and derive the disparity value. Dividing the face of the subject to be photographed into sections having a predetermined area, averaging the parallax values ​​for each section over a predetermined time period, This method for monitoring the tunnel face of a mountain tunnel is characterized by detecting abnormalities at the tunnel face based on changes in this averaged disparity value.

2. The distance to the face is calculated from the averaged disparity value.

2. A method for monitoring a tunnel face in a mountain tunnel according to claim 1, characterized in that an abnormality in the tunnel face is detected based on a change in the calculated distance.

3. The area of ​​the region is adjusted according to the size of the object capable of detecting the displacement, 3. A method for monitoring a mountain tunnel face according to claim 1 or 2, characterized in that the area of ​​the region is reduced in order to improve the sensitivity indicating the detectable size of the object.

4. A method for monitoring the face of a mountain tunnel as described in claim 1 or 2, characterized in that the area of ​​the region subject to averaging is increased in order to improve the measurement accuracy of the disparity value or the distance calculated from the disparity value.

5. A method for monitoring a mountain tunnel face as described in claim 1 or 2, characterized in that the time period for averaging is extended in order to improve the measurement accuracy of the disparity value or the distance calculated from the disparity value.

6. 3. A method for monitoring a mountain tunnel face according to claim 1 or 2, characterized in that a parameter of the disparity values ​​for averaging is 13,000 or more.

7. This system constantly measures the distance to the tunnel face and detects abnormalities at the tunnel face from changes in that distance, thereby monitoring the tunnel face. taking images of the tunnel face using a stereo camera and obtaining a disparity value; Dividing the face of the subject to be photographed into areas having a predetermined area; averaging the disparity values ​​for each of the regions over a predetermined time period; detecting an abnormality in the face based on a change in the averaged disparity value; A mountain tunnel face monitoring system comprising:

8. Calculating a distance to the face from the averaged disparity value; detecting an abnormality in the face based on a change in the calculated distance; A mountain tunnel face monitoring system as claimed in claim 7, further comprising:

9. 9. A mountain tunnel face monitoring system as claimed in claim 7 or 8, characterized in that it comprises a step of issuing an alarm to the vicinity of the face when an abnormality at the face is detected.

Images