Estimation method of landslide surface shape
The method uses UAV laser scanners to calculate and draw landslide surface shapes quickly and accurately, addressing the inefficiencies of existing technologies by automating the process and improving precision in landslide shape estimation.
Patent Information
- Application Number
- JP2023117952
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-07-20
- Publication Date
- 2025-01-30
- Estimated Expiration
- 2043-07-20
AI Technical Summary
Existing methods for estimating landslide surface shapes are cumbersome, time-consuming, and require specialized software and knowledge, leading to inaccuracies due to reliance on high-order polynomial calculations and manual input of terrain data, making it difficult to achieve high precision and speed in disaster scenarios.
A method that estimates landslide surface shape using UAV-equipped laser scanners to calculate movement vectors, group them by gradient, and draw a two-dimensional surface using NURBS curves, starting from the landslide head to the end point, ensuring accuracy and speed by automating the process.
Enables rapid, accurate estimation of landslide surface shapes, enhancing safety in disaster recovery by reducing the risk of secondary disasters through efficient data processing and automated analysis.
Smart Images

Figure 2025014955000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a method for estimating the shape of a slide surface of a landslide based on displacement measurements of the topography before and after the landslide occurs. [Background technology]
[0002] The shape of a landslide surface varies depending on the topography and geology of the landslide site and the cause of the landslide (e.g. cutting earthworks, or natural phenomena such as earthquakes and heavy rains). In addition, many of the occasions when it is necessary to estimate the shape of a landslide surface occur when a disaster occurs, and by quickly and accurately grasping the shape of a landslide surface and taking measures according to the scale and characteristics of the landslide, it is possible to ensure the safety of disaster recovery activities and avoid or reduce the risk of secondary disasters. In other words, a method that ensures speed, ease, and a certain level of accuracy based on the topography, geology, and cause of the landslide is most desirable. Patent Document 1 discloses a method for estimating the shape of a landslide surface, which involves "dividing the landslide mass into a number of blocks by vertical dividing lines and dividing the mass so that a displacement measuring means is installed on the ground surface of each block. The landslide surface of each block is set as a high-order polynomial. Normal equations are constructed for each block by the least squares method to determine parameters of the high-order polynomial for each block based on at least one known boundary condition. The calculated high-order polynomials for each block are then connected to estimate the shape of the slide surface of the entire landslide mass." [Prior art documents] [Patent documents]
[0003] [Patent Document 1] JP 2008-202994 A [Patent Document 2] Patent No. 7170575 [Non-Patent Document 1] Landslide slide surface shape estimation technology pdf Source: National Research and Development Agency Public Works Research Institute, Soil and Sand Management Research Group, Landslide Team Homepage Landslide Line Shape Estimation Method / Slide Line Estimation Program Copyright (C) National Research and Development Agency Public Works Research Institute [Non-Patent Document 2] Cooperative Research on a Landslide Scale Estimation System Using Ground Surface Displacement Measurement Source: Cooperative Research Report No. 451 Cooperative Research Report of PWRI No. 451 January 2013 Edited and published by: Public Works Research Institute Summary of the Invention [Problem to be solved by the invention]
[0004] However, to estimate the shape of a landslide surface using the technology shown in Patent Document 1, calculations using higher-order polynomials and error equations are required, and it is difficult and time-consuming to perform the calculations without using a computer and software to execute them. Therefore, it is necessary to prepare equipment for analyzing the measured values, special software, and operate the program.
[0005] Furthermore, according to Non-Patent Document 1, which is a document explaining the above-mentioned technology, when estimating the shape of the landslide surface using the above-mentioned technology, it is necessary to input topographical data, coordinates of the measurement points, coordinates of the positions where deformations (cracks and extrusions) are observed (the positions of the head of the landslide and the end of the slide surface), and the ground surface displacement vectors at the measurement points. In addition, in order to improve the accuracy of the analysis results, it is desirable to input the settings of block division lines, underground boundary points, and the gradients of the head scarp and cracks, and it is clear that the accuracy of the estimated results depends on the presence or absence of knowledge and experience. In addition, there is also a problem that the calculated estimated results of the slide surface may deviate from the actual shape of the slide surface depending on the conditions such as topography and geology. Therefore, it is difficult to say that it is easy to run a program to obtain highly accurate results, and it cannot be said that speed is necessarily guaranteed.
[0006] Therefore, an object of the present invention is to provide a method for estimating the shape of a landslide surface quickly, easily, and with a certain degree of accuracy, based on displacement measurements of the topography before and after the landslide occurs. [Means for solving the problem]
[0007] In order to solve the above-mentioned problems, the invention described in claim 1 provides a method for estimating a landslide surface shape in a two-dimensional vertical cross section from displacement measurement values on the ground surface, and drawing a landslide surface from a landslide head to an end point which is the end point of the landslide, the method comprising the steps of: calculating a movement vector of an observation point from the displacement measurement values; grouping the movement vectors based on the gradient of the movement vector; determining a representative movement vector for each group; dividing the group at the start point by a vertical dividing line; drawing a virtual line connecting the start point of the group to the end point of the landslide head with the gradient of the representative movement vector of the group to an intersection with the dividing line of the next group; drawing a virtual line connecting the start point of the next group to the end point of the landslide head with the gradient of the representative movement vector of the next group to an intersection with the dividing line of the next group; drawing the landslide surface for each group to the end point of the landslide head of the group; and drawing the landslide surface in descending order from the start point of the landslide head to the end point of the landslide head to estimate the landslide surface shape. Effect of the Invention
[0008] The present invention provides a method for estimating the shape of a landslide surface quickly, easily, and with a certain degree of accuracy, based on displacement measurements of the topography before and after the landslide occurs. [Brief description of the drawings]
[0009] [Figure 1] An explanatory diagram of topographical surveys conducted by UAV laser surveying at two periods, before and after the landslide occurred. [Diagram 2] A tabular diagram listing the topographical survey results for two periods, before and after the landslide occurred. [Diagram 3]An explanatory diagram of the topography before and after the landslide and the initial slide surface (the landslide surface drawn based on the head, the gradient of the movement vector around the scarp, and the end). [Figure 4] FIG. 1 is an explanatory diagram of landslide surface shape drawing 1 according to the present invention. [Diagram 5] FIG. 2 is an explanatory diagram of landslide surface shape drawing 2 according to the present invention. [Figure 6] FIG. 3 is an explanatory diagram of landslide surface shape drawing 3 according to the present invention. [Figure 7] FIG. 4 is an explanatory diagram of landslide surface shape drawing 4 according to the present invention. [Figure 8] FIG. 5 is an explanatory diagram of a landslide surface shape drawing 5 according to the present invention. [Figure 9] FIG. 6 is an explanatory diagram of a landslide surface shape drawing 6 according to the present invention. [Figure 10] FIG. 1 is a comparative explanatory diagram of a slide surface drawn using the method for estimating the landslide surface shape according to the present invention and an actual slide surface that reflects boring results. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0010] The following describes an embodiment of the present invention with reference to the drawings. However, unless otherwise specified, the devices, shapes, and the like described in the embodiment are merely illustrative examples and do not limit the scope of the present invention.
[0011] (Method of determining topographical displacement) In carrying out the present invention, in order to obtain displacement measurements of the topography 01 before the landslide and the topography 04 after the landslide as shown in FIG. 1, topographical survey results at least at two times are required. Conventionally, landslide surface shapes have been estimated by core sampling by boring, borehole testing, and landslide observation using boreholes, but drilling in a manner that ensures safety for landslide observation requires advanced technology and costs. However, with the recent spread of topographical measurements using UAV-mounted laser scanners 02, it is possible to perform surveys of the topography before the landslide and after the landslide without surveying. Surveying methods using airborne laser surveying and UAV-mounted laser scanners 02 as shown in FIG. 1 can calculate the movement vectors of multiple observation points simultaneously, and are therefore desirable in terms of ensuring safety and speed during observation. However, there is no limit to the use of moving piles, extensometers, ground surface inclinometers, etc. during observation and surveying to calculate the movement vectors.
[0012] (The displacement vector is calculated from the displacement measurements at the observation point.) Figure 2 shows an example of calculating a movement vector from the displacement measurement values at the observation point. The observation point 06 on the terrain before the landslide occurs is set as the start point, and the observation point 07 on the terrain after the landslide occurs is set as the end point, and their positions are specified by X and Y coordinates. Then, by connecting the start point and the end point with a line, the movement distance and movement direction of the observation point (gradient 08 or angle of the movement vector of the observation point) can be calculated. Next, observation points with a larger start point X value than the point corresponding to the cliff 05 at the time the landslide occurred (hereinafter referred to as the head start point 10 of the landslide surface) and observation points with a smaller start point X value than the end point 11 of the landslide surface are excluded from the list, and a list is created by arranging the start point X in descending order so that the head start point 10 of the landslide surface is number 1 and the last number is the end point 11 of the landslide surface.
[0013] (Group the movement vectors by the gradient of the vector.) Conventionally, as shown in Fig. 3, when drawing the shape of a landslide surface, it was common to draw a curve connecting the position of the head start point 10 of the landslide surface and the position of the end point 11 of the landslide surface based on the gradient of the movement vector around the slide scarp 05. (The landslide surface estimated by this conventional drawing method is hereinafter referred to as the initial slide surface 12.) However, the slide surface from the head start point 10 of the landslide surface to the position of the end point 11 of the landslide surface may have not only a downward gradient, which is often seen in the gradient of the movement vector around the slide scarp 05, but also a horizontal or upward gradient in some places, because of differences in shape such as unevenness in the topography, differences in geology, and differences in layer thickness. Therefore, by creating a list such as that shown in Fig. 2, it is possible to quickly find groups with similar gradients of the movement vector and observation points showing abnormal values, and therefore it is easy to perform grouping based on the gradient of the movement vector, which is necessary for solving the problem of the present invention, which is to "draw the shape of the slide surface for each group." When plotting the landslide surface shape described later, each group is divided by a vertical dividing line at the observation point that is the starting point of each group.
[0014] (Calculating the gradient by group) Next, the gradient 09 of the movement vector as a group is calculated from the multiple observation points of these groups, and the median value of the gradient of the movement vector of each observation point is used. The reason for using the median value is that there is a risk of outliers at the observation point due to the accidental presence of a large rock, etc., and if the average value is used, the value will deviate significantly from the center. In addition, when a survey is performed using a UAV-mounted laser scanner 02, it is expected that there will be observation points where sufficient survey data cannot be obtained due to the density of vegetation on a slope where vegetation work has been performed. In such cases, the movement vector of the observation point is excluded from the calculation of the median value, and the gradient 09 of the movement vector of each group is calculated.
[0015] (Advantages of creating groups based on the gradient of the movement vector, Patent Document 1) The grouping of the movement vectors and the calculation of the gradient 09 of the movement vectors of the groups before drawing the landslide surface shape are means for solving the problems of the present invention. Therefore, if the geology of the landslide slope and the cause of the landslide are known in advance, the grouping of the movement vectors and the calculation of the gradient 09 of the movement vectors can be performed based on various conditions, such as whether the landslide is a collapse, sliding, or flow, whether it is a clayey soil slide or a rock slide, a circular slide or a planar slide, etc., and the landslide surface shape can be estimated with irregular elements removed in advance. On the other hand, the estimation method of Patent Document 1 requires the input of the ground surface line, the positions of the ground surface measurement points, the positions of the ground surface boundary points, the measurement data (movement vectors), and the selection of analysis conditions (polygon parallel method, polygon rotation method, polynomial method) for the analysis during estimation, and further, when it is intended to reflect information such as the geology and the cause of the landslide, the validity of the estimation result is improved by changing the analysis conditions. In other words, since this is a method in which after the analysis results are obtained, a person analyzes, adjusts, and then performs analysis again using software, it takes time to obtain highly valid estimation results. Therefore, the estimation method of the present invention does not require multiple analyses, and can be said to easily increase the speed at which the landslide surface shape can be drawn.
[0016] (Advantages of creating groups based on the gradient of the movement vector, Patent Document 2) In addition, Patent Document 2 shows a method of calculating the difference vector for each mesh, which can be said to be a more precise estimation method than the estimation method of the present invention. However, in order to improve the validity of the estimation result, it is described that "the subsequent mesh is selected according to the direction of the difference vector," and there is a problem that it cannot be automated. Therefore, when using the estimation method of Patent Document 2, which captures the landslide shape with meshes rather than cross sections, it is presumed that the number of adjustments that cannot be automated increases as described above compared to the estimation method of the present invention, which estimates with cross sections, and it can be said that there is a problem in estimating the landslide surface shape quickly during disasters, etc. Therefore, improving the speed of the method of estimating the landslide surface shape with cross sections as in the present invention brings about a different benefit compared to the invention of Patent Document 2.
[0017] (Draw the landslide surface for each group) The method for drawing the shape of a landslide surface will now be described. In this method, the shape of a landslide surface in a two-dimensional vertical cross section is estimated from displacement measurement values on the ground surface, and a landslide surface is drawn from a landslide head point to an end point, which is a landslide end point. The method calculates a movement vector of an observation point from the displacement measurement values, groups the movement vectors based on the gradient of the movement vectors, obtains a representative movement vector for each group, divides the group by a vertical division line at the start point of the group, and draws a landslide surface from the start point of the group. Then, a virtual line connecting the end point of the landslide with the gradient of the representative movement vector of the same group is drawn to the intersection with the dividing line of the next group, and then a virtual line connecting the end point of the landslide with the gradient of the representative movement vector of the next group is drawn from the intersection of the next group to the intersection with the dividing line of the next group.Then, a landslide surface is drawn for each group up to the end point of the landslide end of the group, and the landslide surface is drawn in descending order from the starting point of the landslide head to the end point of the landslide end, thereby estimating the shape of the landslide surface.
[0018] Specifically, as shown in Fig. 4, the drawing of the landslide surface shape begins from the head starting point 10 of the landslide surface. Then, based on the gradient of the movement vector of group A which has the head starting point 10 of the landslide surface, a virtual line connecting the head starting point 10 of the landslide surface to the end point 11 of the landslide surface (slide surface 12a estimated by the gradient of the movement vector of group A) is drawn using a NURBS curve from the head starting point 10 of the landslide surface to the position of a vertical dividing line dropped from the observation point which is the boundary between group A and group B, as shown in Fig. 5. Next, as shown in Figure 6, the point where the vertical dividing line drawn from the observation point that is the boundary between Group A and Group B intersects with the drawing of the landslide surface shape is designated as intersection point (ab) 14a, and based on the gradient of the movement vector of Group B, an imaginary line connecting intersection point (ab) 14a to the end point 11 of the landslide surface (slide surface 12b estimated from the gradient of the movement vector of Group B) is drawn using a NURBS curve from intersection point (ab) 14a to the position of the vertical dividing line drawn from the observation point that is the boundary between Group B and Group C. As a result, a new intersection point (bc) 14b is generated, and based on the gradient of the movement vector of the group starting from the new intersection point, a virtual line connecting the new intersection point to the end point 11 of the landslide surface is drawn using a NURBS curve from the new intersection point to the position of the vertical dividing line drawn from the observation point that is the boundary with the next group, and the estimated landslide surface shape is drawn from the start point 10 of the head of the landslide surface to the end point 11 of the landslide surface, as shown in Fig. 8. Furthermore, if there is an observation point from which sufficient survey data cannot be obtained in drawing the landslide surface of the above group, a virtual line connecting the intersection point (de) 14d to the end point 11 of the landslide surface (slide surface 12e estimated by the gradient of the movement vector of group E) is drawn using a NURBS curve from the intersection point (de) 14d to the intersection point (ef) 14e that is the position of the vertical dividing line drawn from the observation point that is the starting point of group F, as shown in the range from group E to group F in Fig. 7. Regarding drawing using NURBS curves, it is assumed that drawing will be done using spline curves in general CAD software, but this is not limited to this.
[0019] (Drawing the landslide surface shape in descending order) Here, the drawing of the estimated landslide surface shape in the present invention is performed in descending order starting from the group with the landslide head. As shown in Fig. 9, it is possible to predict not only the slide surface of the drawn group but also the slide surface of the next group, which not only helps to quickly grasp the overall picture but also increases the amount of information to be used when reflecting the above-mentioned information on geology and occurrence cause.
[0020] (Comparison and evaluation of the drawn slip surface and the actual slip surface) Figure 10 shows an example of comparing a slide surface 15 drawn by the method for estimating the landslide surface shape according to the present invention with an actual slide surface 16 reflecting the boring results. Compared with the original slide surface 12 drawn by a curve connecting the start point 10 of the head of the landslide surface and the end point 11 of the tail of the landslide surface based on the gradient of the movement vector around the conventional slide scarp 05, the slide surface 15 drawn by the present invention and the actual slide surface 16 show differences in slide surface depth, but no significant differences in slide surface shape. As such, the present invention can estimate the landslide surface shape with a certain degree of accuracy, so once the actual depth of a slide surface at one location is determined by boring work, it is easy to infer from the landslide surface shape whether the next boring point is shallower or deeper than the previous boring point, and misjudgment of the depth during boring can be prevented.
[0021] (Conclusion) In view of the above, among methods for estimating landslide surfaces based on displacement measurements of the topography before and after a landslide occurs, the present invention is particularly excellent in terms of speed and ease and also satisfies a certain level of accuracy, and is therefore useful for ensuring safety in disaster recovery activities and quickly avoiding and reducing the risk of secondary disasters. [Explanation of symbols]
[0022] 01 Topography before the landslide 02 UAV-mounted laser scanner 03 Crack 04 Landform after the landslide 05 Sliding cliff 06 Observation point before the landslide 07 Observation point after the landslide 08 Gradient of the observation point's movement vector 09 Gradient of group's movement vector 10 Head start point (of a landslide surface) 11 End of landslide surface 12 Initial slide surface (landslide surface drawn based on the head, gradient of the movement vector around the scarp, and end) 12a Slip surface estimated by the gradient of the movement vector of Group A 12b Slip surface estimated by the gradient of the movement vector of Group B 12e Slip surface estimated by the gradient of the displacement vector of group E 13 Movement Vector 14 Intersection of the vertical division line from the observation point and the drawn slide surface 14a Intersection (ab) 14b Intersection (bc) 14d intersection (de) 14e Intersection (ef) 15 Drawn slip surface 16 Slide surface from boring survey
Claims
[Claim 1] The shape of the landslide surface in a two-dimensional vertical cross section is estimated from the displacement measurements on the ground surface. In this method, a landslide surface is drawn from the head of the landslide to the end point, which is the end of the landslide. Calculating a movement vector of the observation point from the displacement measurement value; Grouping the motion vectors based on a gradient of the motion vectors; A representative motion vector is obtained for each of the groups; dividing the group at its origin by a vertical dividing line; A virtual line is drawn from the starting point of the group to the end point of the landslide to the gradient of the representative movement vector of the group to the intersection with the dividing line of the next group; A virtual line is drawn from the intersection of the next group to the end point of the landslide to the slope of the representative movement vector of the next group, and then to the intersection with the division line of the next group. Draw a landslide surface for each group up to the end position of the landslide toe of the group; A method for estimating a landslide surface shape, comprising drawing a line from the start point of the landslide head to the end point of the landslide tow point in descending order.
Citation Information
Patent Citations
A landslide displacement similarity matching method and system based on motion angle difference
CN109344867A
Method, device and program of estimating landslide surface shape
JP2008202994A
Method, device and program for measurement of displacement
JP2011058875A
Target detecting, editing and rebuilding method and system by 3D image
US20090161944A1
JP202994A
Cited By
Method for estimating three-dimensional landslide surface shape
JP7688800B1