Ground movement analysis device and ground movement analysis method

Abstract

This ground movement analysis device comprises an inclination projecting unit (12) which converts LoS movement data acquired for each of a plurality of predetermined unit areas included in a region of interest into inclination movement data in a direction along an inclination direction of the region of interest, an offset correcting unit (13) which performs offset correction of the inclination movement data, and a hotspot extracting unit (14) which uses the offset-corrected inclination movement data to extract, as a movement hotspot, a region in which movement of the ground surface is greater than in a surrounding region, wherein the LoS movement data observed in an area having an undulating topography are converted into inclination movement data in a direction along the inclination direction of the undulations, and noisy data components are removed by means of the offset correction, after which the region having a large ground surface movement is extracted as the movement hotspot, thereby enabling the state of ground movements to be accurately captured, even in areas having an undulating topography.

Description

[Technical Field]

[0001] The present invention relates to a ground movement analysis device and a ground movement analysis method, and is particularly suitable for use in a device and method for analyzing ground movement conditions using observation data measured using a synthetic aperture radar. [Background technology]

[0002] Landslides and other ground failures on slopes are one of the major geological hazards that require monitoring. In recent years, landslides have become more frequent due to extreme weather events and unplanned human activities. Therefore, it is becoming increasingly important to identify and monitor slopes susceptible to landslides in order to prevent or mitigate various landslide-related losses.

[0003] Conventional methods for monitoring ground deformation on slopes have involved frequently repeated field surveys or measurements using ground sensors of the Global Navigation Satellite System (GNSS).However, these methods have the problem that monitoring information can only be obtained within a limited range, such as the site where the field survey was conducted or the location where the ground sensor was installed.

[0004] In response to this, a monitoring method using synthetic aperture radar (SAR) as an earth observation technology has also been proposed (see, for example, Patent Documents 1 and 2). The monitoring method using synthetic aperture radar detects fluctuations in the earth's surface in the direction of the line of sight (LoS) when the earth's surface is viewed from a satellite. It is possible to obtain data showing the LoS variation, and by analyzing this LoS variation data, it is possible to monitor ground deformation over a relatively wide area.

[0005] Patent Document 1: JP 2021-56008 A Patent document 2: Patent No. 3987451

[0006] The landslide area detection device described in Patent Document 1 detects landslide-like ground movement based on a combination of at least landslide characteristics in phase difference data and coherence data obtained by performing interference analysis on ground surface data obtained by radar measurements at different times. Here, the device identifies an area where landslide-like ground movement is not detected based on topographical data corresponding to the area of ​​the ground surface data, and excludes this area from the area of ​​ground movement, thereby reducing unnecessary effort required for the detection process and preventing erroneous detection.

[0007] The method for measuring the amount of ground surface deformation described in Patent Document 2 determines the amount of deformation of the ground surface by performing synthetic aperture interferometry on the ground surface reflection waves of pulsed radio waves emitted from a single synthetic aperture radar mounted on a satellite. Specifically, the reflected waves acquired twice each from an ascending orbit and a descending orbit are subjected to synthetic aperture interferometry to determine the distances from the ascending orbit and the descending orbit to the measurement target point, and by adding a condition for the direction of deformation of the ground surface specified by the ground structure (a condition that landslide deformation occurs along the maximum gradient line of the slope), the actual amount of deformation of the measurement target point from its past position is determined. Summary of the Invention [Problem to be solved by the invention]

[0008] When using synthetic aperture radar, in areas with flat terrain, LoS deformation data can be converted into vertical ground deformation, which helps to meaningfully interpret the ground movement situation. Also, in the case of a slope whose aspect direction is parallel to the LoS direction, LoS deformation data can be interpreted as movement along the slope.

[0009] However, in areas with undulating topography, it is difficult to directly interpret LOS change data or vertical change data converted from LOS change data into the vertical direction as ground movement in that area. For example, mountainous areas typically slope in all directions, and there are undulations even within small areas. Therefore, even if LOS change data and vertical change data are used, it is difficult to accurately grasp the state of ground movement.

[0010] The present invention has been made to solve these problems, and aims to make it possible to accurately grasp the state of ground movement in areas with undulating terrain using observation data obtained by synthetic aperture radar. [Means for solving the problem]

[0011] To solve the above-mentioned problems, the present invention converts LoS variation data acquired for each of a plurality of predetermined unit areas included in a region of interest into tilt variation data in a direction along the tilt direction of the region of interest based on information about the satellite orientation and information about the tilt characteristics of the region of interest, and offsets each of the tilt variation data obtained for each of the plurality of predetermined unit areas using the average value of the tilt variation data obtained for each of the plurality of predetermined unit areas.The offset-corrected tilt variation data is then used to extract areas where the ground surface variation is larger than that of the surrounding area as variation hotspots. [Effects of the Invention]

[0012] According to the present invention configured as described above, LoS variation data observed in an area with undulating topography is converted into tilt variation data in the direction along the gradient of the topography, and after noise components of the data are removed by offset correction, areas with large ground surface variations are extracted as deformation hotspots. This makes it possible to accurately capture the state of ground deformation using observation data obtained by synthetic aperture radar, even in areas with undulating topography. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 1 is a diagram illustrating a network configuration of an analysis system according to an embodiment of the present invention. [Figure 2] 1 is a block diagram showing an example of the functional configuration of a ground deformation analysis device according to an embodiment of the present invention; [Figure 3] FIG. 10 is a diagram for explaining the LoS direction. [Figure 4] FIG. 2 is a block diagram showing a specific example of the functional configuration of a hot spot extraction unit according to the present embodiment. [Figure 5] FIG. 10 is a diagram for explaining the processing content of the difference analysis unit according to the present embodiment, showing a state in which the ground surface of a certain area is viewed from above. [Figure 6] FIG. 10 is a diagram for explaining the processing content of the difference analysis unit according to this embodiment, showing the transition (time-series gradient) of tilt fluctuation data in one given unit area. [Figure 7] 10A and 10B are diagrams for explaining the processing content of a nearby area extraction unit according to the present embodiment. [Figure 8] 3 is a flowchart showing an example of the operation of the ground movement analysis device according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0014] An embodiment of the present invention will be described below with reference to the drawings. Figure 1 shows the network configuration of an analysis system 100 according to this embodiment. The analysis system 100 includes a satellite base station 50 that receives satellite data from a satellite S flying in a satellite orbit in space, a database DB that records the satellite data, and a ground deformation analysis device 10 that analyzes the satellite data.

[0015] The satellite S includes a satellite data acquisition unit (including, for example, an optical sensor, a synthetic aperture radar, etc.), and transmits to the satellite base station 50 satellite data including images of the ground captured by the satellite data acquisition unit.

[0016] The satellite base station 50 receives satellite data from the satellite S and records it in a database DB. The database DB is connected to a communication network N such as the Internet. The ground deformation analysis device 10 acquires the satellite data stored in the database DB via the communication network N.

[0017] 2 is a block diagram showing an example of the functional configuration of the ground movement analysis device 10 according to this embodiment. As shown in Fig. 2, the ground movement analysis device 10 of this embodiment includes, as its functional configuration, an LoS movement data acquisition unit 11, an inclination projection unit 12, an offset correction unit 13, and a hotspot extraction unit 14.

[0018] The functional blocks 11 to 14 can be configured by any of hardware, a DSP (Digital Signal Processor), and software. For example, when configured by software, the functional blocks 11 to 14 are configured with a computer CPU, RAM, ROM, etc., and are realized by the operation of an analysis program stored in a storage medium such as RAM, ROM, a hard disk, or a semiconductor memory. The analysis program may be provided by being stored in a storage medium or via a communication network.

[0019] The physical configuration described here is merely an example, and does not necessarily have to be an independent configuration. For example, the ground movement analysis device 10 may be equipped with an LSI (Large-Scale Integration) in which a CPU, RAM, and ROM are integrated. In addition, in this embodiment, the ground movement analysis Although the case where the device 10 is configured by one computer will be described, the ground movement analysis device 10 may also be realized by combining a plurality of computers.

[0020] The LoS variation data acquisition unit 11 analyzes time-series observation data measured using a synthetic aperture radar and acquires LoS variation data, which indicates variations in the line-of-sight direction when viewing the Earth's surface from a satellite, for each of a plurality of predetermined unit areas included in a region of interest. The region of interest refers to an area to be monitored for ground deformation. The observation data used in this embodiment is, for example, time-series SAR images obtained by daily observation using a synthetic aperture radar and stored in a storage medium (not shown).

[0021] A satellite equipped with a synthetic aperture radar observes any point on Earth from a pre-designated satellite orbit in two directions: northbound (ascending) and southbound (descending). As a result, time-series SAR images observed by a satellite traveling in a northward orbit (hereinafter referred to as northward orbit SAR images) and time-series SAR images observed by a satellite traveling in a southward orbit (hereinafter referred to as southward orbit SAR images) are input to the LoS variation data acquisition unit 11.

[0022] The LoS variation data acquisition unit 11 analyzes the northbound orbit SAR images and the southbound orbit SAR images to acquire data indicating variations in the Earth's surface in the line of sight when viewed from a satellite on a northbound orbit (hereinafter referred to as northbound LoS variation data) and data indicating variations in the Earth's surface in the line of sight when viewed from a satellite on a southbound orbit (hereinafter referred to as southbound LoS variation data). Here, the analysis performed by the LoS variation data acquisition unit 11 can be performed using known InSAR analysis.

[0023] Figure 3 is a diagram for explaining the LoS direction, which is the line of sight direction when viewing the Earth's surface from a satellite. The radar wave emitted from the satellite is not irradiated perpendicularly to the Earth's surface, but in the direction of incidence angles λasc and λdesc that are shifted from the perpendicular direction. That is, the northward orbit shown in Figure 3(a) The incidence angles λasc and λdesc are slightly different between the satellite in the southward orbit shown in Figure 3(b) and the satellite in the southward orbit shown in Figure 3(b). As a result, InSAR analysis determines the amount of deformation in the satellite's LoS direction, rather than the amount of deformation in the vertical direction of the Earth's surface.

[0024] The LoS variation data acquisition unit 11 converts the northbound LoS variation data and southbound LoS variation data acquired as described above into variation data for each predetermined unit time and each predetermined unit area. Here, the LoS variation data acquisition unit 11 matches the northbound LoS variation data with the southbound LoS variation data on the time axis and the space axis, and detects the matched northbound LoS variation data and southbound LoS variation data.

[0025] The northbound LoS variation data and southbound LoS variation data may not match in the time axis direction or the space axis direction. Therefore, the LoS variation data acquisition unit 11 sets a threshold in the time axis direction, and if there is northbound LoS variation data and southbound LoS variation data that can be considered identical within the threshold range, it matches and detects them. Similarly, it sets a threshold in the space axis direction, and if there is northbound LoS variation data and southbound LoS variation data that can be considered identical within the threshold range, it matches and detects them. Here, "matching" refers to the process of determining whether there is a point where northbound LoS variation data and southbound LoS variation data can be considered identical within the threshold range set in the time axis direction or the space axis direction.

[0026] In this embodiment, as an example of the time axis threshold and the spatial axis threshold used when performing matching, it is possible to use thresholds generally set for the Sentinel-1 satellite, which is used for measurements that require strictness and precision. Alternatively, the matching conditions may be set more leniently by setting at least one of the time axis threshold and the spatial axis threshold to a value greater than the threshold generally set for the Sentinel-1 satellite. For example, the time axis threshold can be set to 11 days, and the spatial axis threshold can be set to a rectangular area 20 meters square.

[0027] The LoS variation data acquisition unit 11 uses, for example, northbound LoS variation data as master data and southbound LoS variation data as slave data, and detects southbound LoS variation data that can be considered identical to the northbound LoS variation data within a time range of 11 days and a distance difference of 20 m square. Only when a state that can be considered identical is detected is the LoS variation data and southbound LoS variation data adopted as matched northbound LoS variation data for a predetermined unit time (e.g., every 11 days) and a predetermined unit area (e.g., a rectangular area of ​​20 m square). The predetermined unit time and the predetermined unit area in this case are identified based on the northbound LoS variation data used as the master data.

[0028] The tilt projection unit 12 converts the LoS variation data acquired by the LoS variation data acquisition unit 11 into tilt variation data in a direction along the tilt direction of the area of ​​interest, based on information about the satellite's orientation (azimuth angle φ and incidence angle λ) and information about the tilt characteristics of the area of ​​interest (tilt angle S and aspect angle A).

[0029] The LoS variation data to be converted here is the northbound LoS variation data used as the master data. Furthermore, information related to the satellite orientation (azimuth angle φ and incidence angle λ) is used that is related to the northbound LoS variation data. Information related to the tilt characteristics of the region of interest (tilt angle S and aspect angle A) can be obtained using topographical data published by public organizations such as the Geospatial Information Authority of Japan.

[0030] Specifically, the oblique projection unit 12 calculates the LoS fluctuation data V based on the following (Equation 1): LOS Tilt Tilt Deformation Data V slope Convert to. V slope =V LOS / C [mm / predetermined unit time] (Equation 1) where: C=η LOS ×η slope =N LOS ×N slope +E LOS ×E slope +Z LOS×Z slope (N: North direction, E: East direction, Z: Vertical direction perpendicular to North and East) η LOS =[sinφ* sinλ,-cosφ*sinλ,cosλ] η slope =[-cosA*cosS,-sinA*cosS,sinS] A: Aspect angle indicating the direction of the slope of the observation area as seen from the satellite S: Slope angle of the observation area φ: Azimuth angle relative to the north pole (true north) in the orbit λ: Incidence angle from the satellite to the Earth's surface

[0031] The offset correction unit 13 performs offset correction on each of the tilt variation data obtained for each of the plurality of predetermined unit areas by the tilt projection unit 12, using the average value of the tilt variation data obtained for each of the plurality of predetermined unit areas. Specifically, the offset correction unit 13 calculates the average value of the time-series gradients that represent the transition of the tilt variation data for all the matched predetermined unit areas within the observation area, and performs offset correction by subtracting the average value from the time-series gradients of each predetermined unit area.

[0032] The hotspot extraction unit 14 extracts, as a fluctuation hotspot, an area where the ground surface fluctuation is larger than the surrounding area, using the time-series tilt fluctuation data offset-corrected by the offset correction unit 13. Fig. 4 is a block diagram showing a specific example of the functional configuration of this hotspot extraction unit 14. As shown in Fig. 4, the hotspot extraction unit 14 of this embodiment has, as specific functional configurations, a difference analysis unit 14a, a first filter unit 14b, a second filter unit 14c, and a nearby area extraction unit 14d.

[0033] The difference analysis unit 14a selects one of the specified unit areas as an area of ​​interest, analyzes the difference between the slope fluctuation data of a local area that includes the area of ​​interest and the slope fluctuation data of a control area that includes the local area and is larger than the local area, and extracts an area of ​​interest that satisfies a first condition regarding a large difference as a candidate for a fluctuation hotspot.

[0034] FIG. 5 and FIG. 6 are diagrams for explaining the processing content of the differential analysis unit 14a. FIG. 5 shows a state of overlooking a ground surface of a certain area from above, and each position indicated by a ■ mark shows a plurality of matched predetermined unit areas. Actually, there are many more predetermined unit areas, but they are simplified and illustrated for convenience of explanation. FIG. 6 shows the time-series gradient of the tilt fluctuation data V slope in a certain one of the predetermined unit areas obtained by the processing of the tilt projection unit 12 and the offset correction unit 13, shown for each predetermined unit time (every 11 days), and each dot represents the tilt fluctuation data V obtained every 11 days. slope

[0035] In FIG. 5, the predetermined unit area indicated by the symbol P is the target area. Centered on this target area P, a circular area with a radius r1 (for example, 100 m) is the local area A local . A circular area with a radius r2 (r1 < r2; for example, r2 = 500 m) is the reference area A re gion . In the example of FIG. 5, the target area P and other predetermined unit areas Q local , Q (in the example of FIG. 5, m = 3) exist in the local area A L1 , Q L2 , ···, Q Lm (in the example of FIG. 5, m = 3), and there are further predetermined unit areas Q region , Q R1 , ···, Q R2 , ···, Q Rn (in the example of FIG. 5, n = 9) in the reference area A.

[0036] Here, the shapes of both the local area A local and the reference area A region are circular, but it is not limited thereto. For example, the shape of at least one of the local area A and the reference area A local and the reference area A region may be rectangular. ​​

[0037] In this embodiment, local area A local Tilt change data V local As a local area A local The area of ​​interest P and other predetermined unit areas Q L1 ,Q L2 ,···,Q L m Tilt change data V slope The average value of the control area A is used. region Tilt change data V region As the control area A region The area of ​​interest P and other predetermined unit areas Q L1 ,Q L2 ,···,Q Lm ,Q R1 ,Q R2 ,···,Q Rn Tilt change data V slope The average value of

[0038] Here, the tilt fluctuation data V to be used for calculating the average value slope is the time series gradient shown in Figure 6. A predetermined number of tilt fluctuation data V slope For example, the most recent final Going back from time T1 to time T k k pieces of tilt fluctuation data V up to slope Calculate the average value The value of the sampling number k may be a fixed value or a variable that can be set by the user. Alternatively, when environmental conditions such as the season and the climate of the observation area are set, the value of the sampling number k may be automatically calculated according to the set environmental conditions.

[0039] For example, local area A at time T1 local As the tilt fluctuation data, at time T1 Predetermined unit area P, Q L1 ,Q L2 ,···,Q Lm Tilt change data V slope The average value of Using this, V local-T1 Similarly, at time T k Local area A in local The trend The oblique movement data is from time T k Predetermined unit areas P and Q L1 ,Q L2 ,···,Q Lm Tilt change data V slope This is the average value of V local-Tk It is written as follows.

[0040] Also, control area A at time T1 region As the tilt fluctuation data of the specified unit area P, Q at time T1 L1 ,Q L2 ,···,Q Lm ,Q R1 ,Q R2 ,···,Q Rn Tilt change data V slope Using the average value of V region-T1 Similarly, at time T k Control Area A region The tilt fluctuation data is taken at time T k Predetermined unit areas P and Q L1 ,Q L2 ,···,Q Lm ,Q R2 ,···,Q Rn Tilt change data V slope This is the average value of Re V region-Tk It is written as follows.

[0041] In this case, local area A local Tilt change data V local and control area A region Tilt change data V region The difference ΔV between these is expressed as follows: ΔV=V local -V region ={V local-T1 -V region-T1 ,V local-T2 -V region-T2 ,···,V local-Tk -Vregion-Tk}

[0042] The first condition regarding the magnitude of this difference ΔV is the time series average ΔVavg of the difference ΔV. and the time series standard deviation σΔV of the difference ΔV, it is defined as follows: First condition: ΔVavg>2σΔV Here, the time series average ΔVavg is V local-T1 -V region-T1 ,V local-T2 -V region-T2 , ···,V local-Tk -V region-Tk The time series standard deviation σΔV is the average value of V l ocal-T1 -V region-T1 ,V local-T2 -V region-T2 ,···,V local-Tk -V region-Tk is the standard deviation of

[0043] The time series average ΔVavg calculated as above is the ground deformation in the area of ​​interest P. Control area A around eye area P region In this embodiment, the time series average ΔVavg is greater than twice the time series standard deviation σΔV. If the first condition is met, the area of ​​interest P is extracted as a candidate for a deformation hotspot where deformation of the ground surface is larger than that of the surrounding area.

[0044] The first filter unit 14b narrows down the candidates for fluctuation hotspots by extracting areas of interest P whose magnitude of fluctuation satisfies a second condition based on the slope fluctuation data of the area of ​​interest P extracted by the difference analysis unit 14a as satisfying the first condition and the slope fluctuation data of multiple specified unit areas included in the region of interest.

[0045] For example, the first filter unit 14b may filter the tilt fluctuation data V slope Time series gradient of k tilt fluctuation data V sampled from slope-T1 ~V slope-Tk Calculate the average value of V P avg. The first filter unit 14b also calculates the tilt fluctuation data V slope Sampling from the time series gradient of The k pieces of tilt fluctuation data V slope Calculate the standard deviation of σV slope It is written as follows.

[0046] In this embodiment, the second condition regarding the magnitude of the fluctuation of the attention area P is defined as follows. Second condition: V P avg>σV slope The first filter unit 14b further extracts areas of interest P that satisfy the second condition from the areas of interest P extracted by the difference analysis unit 14a as satisfying the first condition. As a result, only areas of interest P that are experiencing unstable ground movement within the region of interest are extracted as candidates for movement hotspots.

[0047] Here, if the ground is subsiding in the area of ​​interest P, the slope change data V P avg is a negative value. In this case, the second condition is that the standard deviation σV slope than the negative value of , the slope change data V of the area of ​​interest P P On the other hand, if the ground is uplifted in the area of ​​interest P, the slope change data V P avg is a positive value. In this case, the second condition is that the standard deviation σV slope than positive values ​​of Slope change data V for the area of ​​interest P PThe positive value of avg must be larger in absolute value.

[0048] Here, k pieces of slope variation data sampled from the time series gradient of the slope variation data are used to calculate the slope variation data V P avg and the standard deviation σV of the tilt change data for multiple specified unit areas included in the region of interest slope Although an example of calculating For example, V is calculated by using the slope fluctuation data at the latest final time point T1. P ,σV s lope In this case, V P =V slope-T1 is.

[0049] The second filter unit 14c extracts the tilt variation data of the attention area P extracted by the first filter unit 14b as satisfying the second condition from the tilt variation data of the control area A. region The magnitude of the change in the target area P is compared with the slope change data contained in the control area A. region The candidate fluctuation hotspots are further narrowed down by extracting an area of ​​interest P that satisfies a third condition that the magnitude of fluctuation is greater than the magnitude of fluctuation of the area of ​​interest P.

[0050] For example, the second filter unit 14c detects that the magnitude of fluctuation in the attention area P is larger than that in the control area A. regi on A third condition, that is, the magnitude of the fluctuation is greater than the magnitude of the fluctuation of the first area, is defined as follows, and an area of ​​interest P that satisfies this third condition is extracted as a candidate for a fluctuation hotspot. Third condition: V P Sign of avg ≠ sign of ΔVavg

[0051] For example, V P If the sign of avg is positive, it means that the ground is rising in the area of ​​interest P. In this case, if the sign of ΔVavg is positive, local The ground deformation in control area A regionOn the other hand, if the sign of ΔVavg is negative, it means that the ground deformation in the local area A is larger than that in the target area P (the target area P is rising at a faster rate than the surrounding area). local The ground deformation in control area A region This means that the ground movement is smaller than that of the area of ​​interest P (the area of ​​interest P is rising at a slower rate than the surrounding area).

[0052] Also, V P If the sign of avg is negative, it means that the ground is sinking in the area of ​​interest P. In this case, if the sign of ΔVavg is negative, local The ground deformation in control area A region On the other hand, if the sign of ΔVavg is positive, it means that the ground deformation in the local area A local The ground deformation in control area A region This means that the ground movement is smaller than the ground movement of the area of ​​interest P (the area of ​​interest P is sinking at a slower rate than the surrounding area).

[0053] Therefore, the slope variation data V of the target area P P The sign of avg and the local area A local The trend Tilt Deformation Data V local and control area A region Tilt change data V region The average of the difference ΔV The third condition is that the signs of the time series average ΔVavg, which is the average value, are different. By performing the process of 14c, only the area of ​​interest P where the rate of change is faster than the surrounding area is extracted as a candidate for a change hotspot.

[0054] In addition, the magnitude of the fluctuation in the target area P is region A third condition, which is greater than the magnitude of the fluctuation of , may be defined as follows: Third condition: V P avg>V region avg where V region avg is the control area A region Tilt change data V regionof the time series gradient of It is the average value and is expressed as follows: V region avg=V region-T1 +V region-T2 +···+V region-Tk / k

[0055] The nearby area extraction unit 14d extracts areas of interest P that are located within a predetermined distance from each other from among multiple areas of interest P that have been extracted as having larger fluctuations on the ground surface than the surrounding area (multiple areas of interest P extracted by the processing of the difference analysis unit 14a, the first filter unit 14b, and the second filter unit 14c), and extracts an area that includes a collection of the extracted areas of interest P as a fluctuation hotspot.

[0056] The value of the predetermined distance can vary depending on the SAR image used as input for the InSAR analysis. For example, when performing Sentinel-1-based InSAR analysis using the IPTA (Interferometric Point Target) method, which includes both single-look and multi-look processing, the predetermined distance can be set to 44 m.

[0057] Fig. 7 is a diagram for explaining the processing content of the nearby area extraction unit 14d. Fig. 7 shows a state in which a certain area of ​​the ground surface is viewed from above, and the individual positions indicated by squares indicate multiple areas of interest P1 to P5 extracted as candidates for fluctuation hotspots through processing by the difference analysis unit 14a, the first filter unit 14b, and the second filter unit 14c. In reality, there may be many more areas of interest P that are candidates for fluctuation hotspots, but Fig. 7 shows a simplified illustration for ease of explanation.

[0058] Of the multiple areas of interest P1 to P5 shown in FIG. 7, three areas of interest P1 to P3 are located close to each other, with an interval within a predetermined distance. On the other hand, the remaining two areas of interest P4 and P5 are not located close to any of the others. In this case, the nearby area extraction unit 14d extracts an area including the set of three areas of interest P1 to P3 as a fluctuation hotspot HS. The area including the set of three areas of interest P1 to P3 is, for example, a rectangular area inscribed around the three areas of interest P1 to P3, as shown in FIG. 7. The remaining two areas of interest P4, P5 is not extracted as a fluctuation hotspot HS. Areas of interest P4 and P5, which are unlikely to have a spatial relationship with area P, are excluded from the fluctuation hotspot HS.

[0059] Here, the nearby area extraction unit 14d may extract an area including a set of multiple attention areas P as a variable hotspot HS only if the number of attention areas P that exist in nearby positions within a predetermined distance from each other is a predetermined number (e.g., five) or more.

[0060] 7 is an example, and the shape of the area of ​​the varying hotspot HS is not limited to this. For example, the shape may be a rectangle, a circle, an ellipse, or another polygon with the minimum area that includes the areas of interest P1 to P3.

[0061] As described above, by performing the processing of the difference analysis unit 14a, the first filter unit 14b, the second filter unit 14c, and the adjacent area extraction unit 14d, it becomes possible to detect fluctuation hotspots HS with a higher level of reliability.

[0062] 8 is a flowchart showing an example of operation (an example of a processing procedure for a ground movement analysis method) of the ground movement analysis device 10 according to this embodiment. First, the LoS movement data acquisition unit 11 analyzes time-series observation data measured using a synthetic aperture radar, and acquires time-series LoS movement data for each of a plurality of predetermined unit areas and a plurality of predetermined unit times included in a region of interest (step S1).

[0063] Next, the tilt projection unit 12 converts the LoS variation data acquired by the LoS variation data acquisition unit 11 into tilt variation data in accordance with the above-mentioned (Equation 1) based on information about the satellite's orientation (azimuth angle φ and incidence angle λ) and information about the tilt characteristics of the region of interest (tilt angle S and aspect angle A) (step S2).

[0064] Next, the offset correction unit 13 performs offset correction on each of the tilt fluctuation data obtained for each of the plurality of predetermined unit areas by the tilt projection unit 12 using the average value of the tilt fluctuation data obtained for each of the plurality of predetermined unit areas (step S3).

[0065] Next, the difference analysis unit 14a analyzes the local area A including the area of ​​interest P. local Tilt change data and local area A local Wider control area A region The difference between the tilt change data and An area of ​​interest P in which the magnitude of the difference satisfies the first condition is extracted as a candidate for a fluctuating hotspot (step S4).

[0066] Next, the first filter unit 14b extracts, for the area of ​​interest P extracted by the difference analysis unit 14a as satisfying the first condition, an area of ​​interest P whose magnitude of fluctuation satisfies a second condition based on the slope fluctuation data of the area of ​​interest P and the slope fluctuation data of multiple specified unit areas included in the region of interest (step S5).

[0067] Furthermore, the second filter unit 14c extracts the tilt variation data of the attention area P extracted by the first filter unit 14b as satisfying the second condition from the tilt variation data of the control area A. region The slope change data included in the target area P is compared with that of the control area A. region An area of ​​interest P that satisfies a third condition, that is, the magnitude of the fluctuation of the area P is greater than the magnitude of the fluctuation of the area P (step S6).

[0068] Finally, the nearby area extraction unit 14d extracts areas of interest P that are located at positions spaced apart from each other by a predetermined distance from each other among the multiple areas of interest P extracted by the second filter unit 14c, and extracts an area including a set of the extracted areas of interest P as a fluctuation hotspot (step S7). This completes the processing of the flowchart shown in FIG.

[0069] As described above in detail, in this embodiment, based on information about the satellite orientation and information about the tilt characteristics of the region of interest, LoS variation data acquired for each of a plurality of predetermined unit areas included in the region of interest is converted into tilt variation data in a direction along the tilt direction of the region of interest, and each of the tilt variation data obtained for each of the plurality of predetermined unit areas is offset-corrected using the average value of the tilt variation data obtained for each of the plurality of predetermined unit areas.The offset-corrected tilt variation data is then used to extract areas where the ground surface variation is larger than that of the surrounding area as variation hotspots.

[0070] According to this embodiment, LoS variation data observed in an area with undulating terrain is converted into tilt variation data in the direction along the slope of the terrain, and then noise components are removed by offset correction, after which deformation hotspots are extracted from areas with large ground surface deformation. This makes it possible to accurately capture ground deformation conditions using observation data obtained by synthetic aperture radar, even in areas with undulating terrain.

[0071] In the above embodiment, the hotspot extraction unit 14 performs processing using the differential analysis unit 14a, the first filter unit 14b, and the second filter unit 14c. This makes it possible to extract, as candidates for fluctuation hotspots, not just areas where ground movement is greater than in the surrounding area, but also areas experiencing unstable ground movement and where the rate of movement is greater than in the surrounding area. Furthermore, the hotspot extraction unit 14 also performs processing using the adjacent area extraction unit 14d, making it possible to exclude areas that are unlikely to have a spatial relationship from candidates for fluctuation hotspots. This allows for detection of fluctuation hotspots with a higher level of reliability.

[0072] Note that the processing by the hotspot extraction unit 14 does not necessarily require the processing of all of the difference analysis unit 14a, the first filter unit 14b, the second filter unit 14c, and the adjacent area extraction unit 14d. For example, only the processing by the difference analysis unit 14a may be performed. In this case, the hotspot extraction unit 14 extracts, as a fluctuating hotspot HS, an area including a set of attention areas P whose difference magnitude satisfies a first condition. In this case, since the area has not been processed by the adjacent area extraction unit 14d, there is a possibility that the area may exist in multiple dispersed locations. To prevent the extraction locations of the fluctuating hotspots HS from being dispersed, the processing by the difference analysis unit 14a and the processing by the adjacent area extraction unit 14d may be performed in combination.

[0073] Alternatively, the processing by the difference analysis unit 14a may be performed in combination with the processing by any one or two of the first filter unit 14b, the second filter unit 14c, and the nearby area extraction unit 14d. For example, the processing by the difference analysis unit 14a and the processing by the first filter unit 14b may be performed together. In this case, the hotspot extraction unit 14 extracts an area including a set of attention areas P extracted by the first filter unit 14b as a fluctuation hotspot HS. In this case, the area may also be dispersed across multiple locations. To prevent the extraction locations of the fluctuation hotspots HS from being dispersed, the processing by the difference analysis unit 14a, the processing by the first filter unit 14b, and the processing by the nearby area extraction unit 14d may be performed in combination.

[0074] Alternatively, the process of the difference analysis unit 14a and the process of the second filter unit 14c may be performed. In this case, for the attention area P extracted by the difference analysis unit 14a as satisfying the first condition, the second filter unit 14c performs a process of filtering the tilt variation data of the attention area P and the tilt variation data of the control area A. region The slope change data of the target area P is compared with that of the control area A. region The hotspot extraction unit 14 extracts areas of interest P that satisfy a third condition, where the magnitude of the fluctuation is greater than the magnitude of the fluctuation in the area P. The hotspot extraction unit 14 extracts an area that includes a set of areas of interest P extracted by the second filter unit 14c as a fluctuation hotspot HS. In this case, the areas may also be dispersed across multiple locations. To prevent the extracted locations of the fluctuation hotspots HS from being dispersed, the processing of the difference analysis unit 14a, the processing of the second filter unit 14c, and the processing of the adjacent area extraction unit 14d may be performed in combination.

[0075] Alternatively, the processing by the nearby area extraction unit 14d may be omitted, and only the processing by the difference analysis unit 14a, the processing by the first filter unit 14b, and the processing by the second filter unit 14c may be performed. In this case, the hotspot extraction unit 14 extracts, as a changing hotspot HS, an area including a set of attention areas P extracted by the second filter unit 14c.

[0076] Furthermore, the above-described embodiments are merely examples of specific embodiments for carrying out the present invention, and the technical scope of the present invention should not be construed as being limited thereby. In other words, the present invention can be carried out in various forms without departing from the gist or main characteristics thereof. [Explanation of symbols]

[0077] 10. Ground deformation analysis device 11 LoS fluctuation data acquisition section 12 Tilt projection section 13 Offset correction section 14 Hotspot Extraction 14a Difference analysis part 14b First filter section 14c Second filter section 14d Proximal area extraction section

Claims

1. an LoS variation data acquisition unit that analyzes time-series observation data measured using a synthetic aperture radar and acquires LoS variation data, which is data indicating variations in the line of sight direction when viewing the Earth's surface from a satellite, for each of a plurality of predetermined unit areas included in the region of interest; a tilt projection unit that converts the LoS variation data acquired by the LoS variation data acquisition unit into tilt variation data in a direction along the tilt direction of the region of interest, based on information about the orientation of the satellite and information about tilt characteristics of the region of interest; an offset correction unit that performs offset correction on each of the tilt fluctuation data obtained for each of the plurality of predetermined unit areas using an average value of the tilt fluctuation data obtained for each of the plurality of predetermined unit areas; a hotspot extraction unit that uses the tilt variation data that has been offset-corrected by the offset correction unit to extract, as a variation hotspot, an area where the ground surface variation is larger than that of the surrounding area. A ground deformation analysis device characterized by:

2. The hot spot extraction unit a difference analysis unit that sets one of the predetermined unit areas as an area of ​​interest, and analyzes a difference between the slope variation data of a local area including the area of ​​interest and the slope variation data of a control area that includes the local area and is wider than the local area, A region including a set of the areas of interest where the magnitude of the difference satisfies a first condition is extracted as the fluctuation hotspot.

2. The ground deformation analysis device according to claim 1.

3. The hot spot extraction unit a first filter unit that extracts, from the area of ​​interest extracted by the difference analysis unit as satisfying the first condition, the area of ​​interest that satisfies a second condition that the magnitude of fluctuation in the area of ​​interest is large, based on the slope fluctuation data of the area of ​​interest and the slope fluctuation data of a plurality of predetermined unit areas included in the region of interest; extracting an area including the set of areas of interest extracted by the first filter unit as the fluctuation hotspot; 3. The ground deformation analysis device according to claim 2.

4. The hot spot extraction unit a second filter unit that compares the slope change data of the area of ​​interest extracted by the difference analysis unit as satisfying the first condition with the slope change data included in the comparison area, and extracts the area of ​​interest that satisfies a third condition that the magnitude of the change in the area of ​​interest is greater than the magnitude of the change in the comparison area; An area including the set of areas of interest extracted by the second filter unit is extracted as the fluctuation hotspot.

3. The ground deformation analysis device according to claim 2.

5. The hot spot extraction unit a second filter unit that compares the slope change data of the area of ​​interest extracted by the first filter unit as satisfying the second condition with the slope change data included in the comparison area, and extracts the area of ​​interest that satisfies a third condition that the magnitude of the change in the area of ​​interest is greater than the magnitude of the change in the comparison area; An area including the set of areas of interest extracted by the second filter unit is extracted as the fluctuation hotspot.

4. The ground deformation analysis device according to claim 3.

6. The hot spot extraction unit The method further includes a nearby area extraction unit that extracts areas of interest that are located at positions spaced apart from each other by a predetermined distance or less from among the plurality of areas of interest that have been extracted as having larger variations in the ground surface than the surrounding areas, An area including the set of areas of interest extracted by the adjacent area extraction unit is extracted as the fluctuation hotspot. The ground deformation analysis device according to any one of claims 2 to 5.

7. a first step in which an LoS variation data acquisition unit of the ground deformation analysis device analyzes time-series observation data measured using a synthetic aperture radar and acquires LoS variation data, which is data indicating variations in the line of sight when viewing the ground surface from a satellite, for each of a plurality of predetermined unit areas included in a region of interest; a second step in which a tilt projection unit of the ground deformation analysis device converts the LoS variation data acquired by the LoS variation data acquisition unit into tilt variation data in a direction along the tilt direction of the area of ​​interest based on information about the orientation of the satellite and information about the tilt characteristics of the area of ​​interest; a third step in which an offset correction unit of the ground movement analysis device performs offset correction on each of the tilt change data obtained for each of the plurality of predetermined unit areas using an average value of the tilt change data obtained for each of the plurality of predetermined unit areas; and a fourth step in which a hotspot extraction unit of the ground movement analysis device uses the tilt change data offset-corrected by the offset correction unit to extract, as a change hotspot, an area where the ground surface has changed more than the surrounding area. A ground deformation analysis method characterized by:

Images