Pseudo-data creation system, pseudo-data creation method, and storage medium
The pseudo data creation system addresses the computational challenges of generating training data for underground radar surveys by using geometric transformations and interpolation, allowing for efficient and cost-effective generation of pseudo data on general-purpose PCs.
Patent Information
- Application Number
- PCT/JP2025/005694
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-08-01
- Filing Date
- 2025-02-19
- Publication Date
- 2026-02-05
AI Technical Summary
Existing methods for generating training data for underground radar surveys of buried objects are computationally expensive and require high-performance computing environments, making it difficult to quickly create large amounts of pseudo data for machine learning models.
A pseudo data creation system that utilizes a general-purpose PC to generate pseudo data of underground radar reflections from buried pipes by calculating reflection profile points, reference reflection profile points, and associating pseudo buried pipe signals with probe points using geometric transformations and interpolation methods, reducing computational costs.
Enables the rapid creation of large amounts of pseudo data for underground radar surveys, facilitating efficient training of machine learning models for buried object detection with low computational overhead.
Smart Images

Figure JP2025005694_05022026_PF_FP_ABST
Abstract
Description
Pseudo data creation system, pseudo data creation method, and storage medium Incorporation by Reference
[0001] This application claims priority from Japanese Patent Application No. 2024-125888, filed on August 1, 2024, the contents of which are incorporated herein by reference.
[0002] The present invention relates to a pseudo data generation system, and more particularly to a technique for generating pseudo signals obtained by underground radar surveying of underground pipes.
[0003] Traditionally, construction work and planning involves the use of drawings such as blueprints and construction drawings that compile information on buried objects and surrounding objects in the relevant area. However, because the time of drawing preparation differs from the time of planning and construction, the information on buried objects described in the drawings may differ from the actual situation. Therefore, to resolve the discrepancies between the drawings and the actual situation and reduce rework during planning and construction, surveys of buried objects in the relevant area are conducted. One method of surveying buried objects is ground-penetrating radar, which measures electromagnetic waves irradiated into the ground and reflected by buried objects to obtain information including the location of the buried objects.
[0004] In underground radar exploration for buried object detection, electromagnetic waves of several hundred MHz are typically emitted at a certain measurement point, and the reflected waves are measured for a period of several tens to several hundred nanoseconds from the time of emission, and the delay time until the reflected waves are received and the amplitude of the reflected waves are recorded. By measuring at multiple measurement points, the location of the buried object can be identified. In order to distinguish a long, narrow buried pipe whose direction of installation is unknown from natural objects such as geological layers, it is desirable to measure at multiple measurement points closely spaced in the relevant area.
[0005] Because analyzing ground-penetrating radar surveys requires skilled techniques, in recent years, technology has been developed to automatically detect buried objects from ground-penetrating radar data using machine learning. Generally, the more expressive a model is used to improve the analytical performance of machine learning, the more training data is required. However, it is difficult to collect large amounts of field ground-penetrating radar data with the annotation information required for training, so training machine learning models using simulation data is being considered.
[0006] The following prior art exists as background art in this technical field: Patent Literature 1 (JP 2018-41178 A) discloses a method for generating training data for recognizing the state of a target system, the training data generation method comprising: defining different combinations of the multiple factors representing each of multiple states of the target system in an analytical model defined by multiple factors including at least the spatial occupation mode of one or more objects and the medium surrounding them in the target system and response characteristics to a probe wave that is a longitudinal wave, a shear wave, or a combination of longitudinal and shear waves; and generating the training data for each of the different combinations of the multiple factors based on reflected waves that are virtually received at each of multiple input points when the probe wave is virtually incident on the target system from one or more output points in the analytical model.
[0007] Patent Document 1 illustrates a three-dimensional simulation of buried objects using the FDTD method, and describes how to address the problem of the FDTD method taking a long time to calculate by using a cluster of 64 GPUs to speed up calculations and create a large amount of training data. However, with the method of Patent Document 1, it is difficult to quickly generate a large amount of data on buried objects unless an information processing environment with high computing power is available.
[0008] Therefore, there is a demand for a technology that reduces the calculation cost of creating pseudo data and enables a large amount of pseudo data to be created quickly even on a general-purpose PC.
[0009] The present invention aims to provide a pseudo data creation system, a pseudo data creation method, and a program for creating pseudo underground radar data obtained by reflection from underground pipes at measurement points densely arranged in a target area with low computational cost.
[0010] A representative example of the invention disclosed in the present application is as follows: That is, a pseudo data creation system for creating pseudo data of a pseudo buried pipe simulating data acquired in buried pipe exploration using a ground penetrating radar method, comprising a calculation unit for executing predetermined processing and a computer having a storage device connected to the calculation unit, wherein the calculation unit includes a reflection profile point calculation unit that calculates, based on a profile line representing the shape of the pseudo buried pipe using a straight line or a curve and pseudo exploration points on a pseudo exploration plane, reflection profile points that are feet of perpendicular lines dropped from the pseudo exploration points to the profile line, and a reference reflection profile calculation unit that calculates, based on a reference profile line represented by a straight line with a constant depth and reference exploration points that are on the intersection of the pseudo exploration plane with a plane perpendicular to the reference profile line, and a reference pseudo probe point calculation unit in which the arithmetic device calculates a reference pseudo probe point, which is a point obtained by applying to the pseudo probe point a transformation that matches a minute line segment of the profile line at the reflection profile point with a minute line segment of the reference profile line at the reference reflection profile point, based on the pseudo probe point, the profile line, the reflection profile point, the reference profile line, and the reference reflection profile point; and a buried pipe signal calculation unit in which the arithmetic device calculates a pseudo buried pipe signal at the reference pseudo probe point based on the pseudo probe point, the reference pseudo probe point, the reference probe point, and a reference buried pipe signal obtained at the reference probe point for the buried pipe represented by the reference profile line, and associates the pseudo buried pipe signal with the pseudo probe point to calculate pseudo data.
[0011] According to one aspect of the present invention, pseudo underground radar data obtained by reflection from a pseudo buried pipe, which simulates data acquired in underground pipe detection using the underground radar method, can be generated with low calculation cost. Problems, configurations, and effects other than those described above will become clear from the description of the following embodiments.
[0012] FIG. 1 is a diagram showing an example of functional blocks of a pseudo data creation system according to a first embodiment. FIG. 2 is a diagram showing an example of the configuration of hardware on which the pseudo data creation system according to the first embodiment is implemented. FIG. 3 is a diagram showing an example of pseudo buried pipes and pseudo probe points according to the first embodiment. FIG. 4 is a diagram showing an example of reference buried pipes and reference pseudo probe points according to the first embodiment. FIG. 5 is a diagram showing an example of functional blocks of a pseudo data creation system according to a second embodiment. FIG. 6 is a diagram showing an example of functional blocks of a pseudo data creation system according to a third embodiment. FIG. 7 is a diagram showing an example of functional blocks of a pseudo data creation system according to a fourth embodiment. FIG. 8 is a diagram showing an example of functional blocks of another pseudo data creation system according to the fourth embodiment. FIG. 9 is a diagram showing an example of functional blocks of a pseudo data creation system according to a fifth embodiment. FIG. 10 is a diagram showing an example of functional blocks of a pseudo data creation system according to a sixth embodiment. FIG. 11 is a diagram showing an example of functional blocks of a pseudo data creation system according to a seventh embodiment.
[0013] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The following description and drawings are examples for explaining the present invention, and some omissions and simplifications have been made as appropriate for clarity of explanation. The present invention can be implemented in various other forms.
[0014] To facilitate understanding of the invention, the position, size, shape, range, etc. of each component shown in the drawings may not represent the actual position, size, shape, range, etc. Therefore, the present invention is not necessarily limited to the position, size, shape, range, etc. disclosed in the drawings.
[0015] When there are multiple components with the same or similar functions, they may be described using the same reference numeral with different subscripts. However, when there is no need to distinguish between these multiple components, the subscripts may be omitted.
[0016] Furthermore, in the following description, processing performed by executing a program may be described, but the program is executed by a processor (e.g., a CPU (Central Processing Unit), a GPU (Graphics Processing Unit)) to perform the specified processing while appropriately using storage resources (e.g., memory) and / or interface devices (e.g., communication ports), etc., so the subject of the processing may be the processor. Similarly, the subject of the processing performed by executing a program may be a controller, device, system, computer, or node having a processor. The subject of the processing performed by executing a program may be any computing unit, and may include a dedicated circuit (e.g., an FPGA (Field-Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit)) that performs specific processing.
[0017] A program may be installed on a device such as a computer from a program source. The program source may be, for example, a program distribution server or a computer-readable storage medium. If the program source is a program distribution server, the program distribution server may include a processor and storage resources for storing the program to be distributed, and the processor of the program distribution server may distribute the program to be distributed to other computers. Also, in the following description, two or more programs may be realized as one program, or one program may be realized as two or more programs.
[0018] In addition, in the following explanation, the system may be described as being composed of a single computer, but all or part of the functions of the computer may be distributed across one or more computers, such as a cloud, and similar functions may be realized by communicating with each other via a network.
[0019] First Embodiment FIG. 1 is a diagram showing an example of functional blocks of a pseudo data generation system S1 according to a first embodiment.
[0020] The pseudo data creation system S1 shown in FIG. 1 includes a reflection profile point calculation unit P1, a reference reflection profile point calculation unit P2, a reference pseudo probe point calculation unit P3, and a buried pipe signal calculation unit P4.
[0021] The pseudo data creation system S1 receives the reference buried pipe signal A1, the pseudo probe point T3, the profile line T2, the reference probe point R3, and the reference profile line R2 as inputs, and outputs pseudo data D1 that associates the pseudo probe point T3 with the buried pipe signal.
[0022] The pseudo data creation system S1 may be executed on a server H1 having the hardware configuration shown in Fig. 2. The server H1 is configured as a general-purpose computer having information processing resources such as a processor H2 such as a CPU, a memory H3, a storage H4, and a communication I / F (Interface) H5.
[0023] The processor H2 is a computing device that executes programs stored in the memory H3. The processor H2 executes various programs to realize the functions of each functional unit (such as the reflection profile point calculation unit P1, the reference reflection profile point calculation unit P2, the reference pseudo-probing point calculation unit P3, and the buried pipe signal calculation unit P4) of the pseudo data creation system S1. Note that some of the processing performed by the processor H2 by executing the programs may be executed by another computing device (for example, hardware such as an ASIC or FPGA).
[0024] The memory H3 includes a ROM, which is a non-volatile storage element, and a RAM, which is a volatile storage element. The ROM stores unchanging programs (e.g., BIOS), etc. The RAM is a high-speed, volatile storage element such as a DRAM (Dynamic Random Access Memory), and temporarily stores programs executed by the processor H2 and data used when the programs are executed.
[0025] The storage H4 is, for example, a large-capacity, non-volatile storage device such as a magnetic storage device (HDD) or flash memory (SSD). The storage H4 also stores data used by the processor H2 when executing a program, and the program executed by the processor H2. For example, the storage H4 stores a pseudo-data creation program M1 that causes a computer to execute a pseudo-data creation method, and causes the program to be executed by the information processing resource. That is, the program is read from the storage H4, loaded into the memory H3, and executed by the processor H2, thereby realizing each function of the pseudo-data creation system S1.
[0026] The communication interface H5 is a network interface device that controls communication with other devices in accordance with a predetermined protocol.
[0027] The pseudo data creation system S1 may have an input / output interface. The input / output interface is an interface to which input devices such as a keyboard and a mouse are connected to receive input from an operator, and which is connected to an output device such as a display device and a printer to output the results of program execution in a format visible to the operator. Note that a terminal device connected to the pseudo data creation system S1 may provide the input and output devices. In this case, the pseudo data creation system S1 may have web server functionality, and the terminal device may access the pseudo data creation system S1 using a predetermined protocol (e.g., http).
[0028] The program executed by the processor H2 is provided to the pseudo data creation system S1 from removable media (such as a CD-ROM or flash memory) or via a network, and is stored in non-volatile storage H4, which is a non-transitory storage medium. For this reason, the pseudo data creation system S1 should preferably have an interface for reading data from removable media.
[0029] The pseudo data creation system S1 is a computer system configured on one physical computer or on multiple logically or physically configured computers, and may run on a virtual computer constructed on multiple physical computer resources. For example, multiple programs that realize the functions of the pseudo data creation system S1 may each run on a separate physical or logical computer, or multiple programs may be combined to run on a single physical or logical computer.
[0030] FIG. 3B is a diagram showing an example of a reference buried pipe R1 and a reference pseudo probe point T3', and shows the reference buried pipe R1 and reference probe point R3 that are the target of underground radar probes from which a reference buried pipe signal A1 referenced by the pseudo data creation system S1 was acquired. The reference buried pipe R1 is a buried pipe whose shape is represented by a reference profile line R2, which is a straight line of constant depth shown by a dashed line in FIG. 3B. The reference reflection profile point R4 is a point on the reference profile line R2. The reference slice plane R5 is a plane perpendicular to the reference profile line R2 at the reference reflection profile point R4. The reference intersection line R6 is the intersection line between the reference slice plane R5 and the probe plane C1. The reference probe point R3 is one or more points on the reference intersection line R6. The reference buried pipe signal A1 was acquired by underground radar probes at the reference probe point R3. The buried pipe cross section C2 is a cross section of the reference buried pipe R1 taken along the reference slice plane R5. The reference pseudo-probing point T3' is a point on the reference intersection line R6 corresponding to the pseudo-probing point T3.
[0031] FIG. 3A is a diagram showing an example of a pseudo buried pipe T1 and a pseudo probe point T3, and shows the pseudo buried pipe T1 and the pseudo probe point T3 that are the targets of calculating the pseudo data D1 in the pseudo data creation system S1. The pseudo probe point T3 is one or more points on the probe plane C1. The pseudo buried pipe T1 is a buried pipe whose shape is represented by the profile line T2 shown by the dashed line in FIG. 3A. The reflection profile point T4 is a point at which the electromagnetic wave pseudo-irradiated from the pseudo probe point T3 is perpendicularly incident on the profile line T2. The slice plane T5 is a plane perpendicular to the profile line T2 at the reflection profile point T4. The intersection line T6 is the intersection line between the slice plane T5 and the probe plane C1. In order for the electromagnetic wave irradiated from the pseudo probe point T3 to be received at the pseudo probe point T3, the irradiated electromagnetic wave must be perpendicularly incident on the profile line T2. Therefore, the pseudo probe point T3 is a point on the slice plane T5. Since the pseudo-probing point T3 is also a point on the probing plane C1, the pseudo-probing point T3 exists on the intersection line T6.
[0032] The pseudo data D1 at the pseudo probe point T3 can be calculated by referring to the reference buried pipe signal A1. When the system shown in FIG. 3A is moved to the system shown in FIG. 3B so that the reflection profile point T4 and the reference reflection profile point R4 coincide and the slice plane T5 and the reference slice plane R5 coincide, a reference pseudo probe point T3', which is the point after the pseudo probe point T3 is moved, is calculated, and the pseudo buried pipe signal at the reference pseudo probe point T3' is calculated from the value of the reference buried pipe signal A1 at the reference probe point R3. The calculated value of the pseudo buried pipe signal at the reference pseudo probe point T3' is used as the pseudo data D1 at the pseudo probe point T3. Calculation of the pseudo data D1 at the pseudo probe point T3 using the reference buried pipe signal A1 of the reference buried pipe R1 having the buried pipe cross section C2 is equivalent to calculation of the pseudo probe result when the cross section of the pseudo buried pipe T1 according to the slice plane T5 is the buried pipe cross section C2.
[0033] The reference profile line R2 is a straight line with a constant depth. If the depth of the reference profile line R2 is not constant, the reference slice plane R5 may not be perpendicular to the detection plane C1, and the acquired reference buried pipe signal A1 in this case varies depending on the position of the reference detection point R3 in the direction of the reference profile line R2 and is not uniquely determined. If the reference profile line R2 is not a straight line, for example, using the foot of a perpendicular line dropped from the reference reflection profile point R4 on the reference slice plane R5 to the detection plane C1 as a reference, the signal value acquired at the reference detection point R3 on the side where the center of the curvature circle of the reference profile line R2 is located will be different from the signal value acquired at the reference detection point R3 on the side where the center of the curvature circle is not located. Electromagnetic waves emitted from the reference probe point R3 on the side where the center of the circle of curvature is not present are reflected only by one reference reflection profile point R4 on the reference slice plane R5 and received at the reference probe point R3, whereas electromagnetic waves emitted from the reference probe point R3 on the side where the center of the circle of curvature is present may be reflected not only by the reference reflection profile point R4 on the reference slice plane R5 but also by points on the reference profile line R2 that are not on the reference slice plane R5 and received at the reference probe point R3. In other words, the reference buried pipe signal A1 is not a bilaterally symmetrical hyperbolic signal, but is an asymmetric signal depending on the direction of the curve. In the pseudo data creation system S1 of this embodiment, the buried pipe signal is calculated by utilizing the fact that the electromagnetic waves emitted from the pseudo probe point T3 are reflected by one reflection profile point T4 on one slice plane T5 and received at the pseudo probe point T3. Therefore, using a reference buried pipe signal A1 that may be reflected from multiple points will result in inconsistent data. For these reasons, it is preferable to use a straight line with a constant depth as the reference profile line R2.
[0034] In FIG. 1 , the reflection profile point calculation unit P1 receives a profile line T2 and a pseudo-detection point T3 as input. The profile line T2 is a piecewise smooth straight line or curve in three-dimensional space, such as a line, line segment, polygonal line, arc, or sigmoid curve. The profile line T2 may be interrupted midway and may include self-intersections. The profile line T2 can be expressed using a parameter t, as shown in equation (1). In equation (1), p(t) represents the position vector of a point on the profile line T2, and f(t), g(t), and h(t) are functions that calculate the x-component, y-component, and z-component of the point on the profile line T2, respectively. Note that the vector p(t) will be written in the specification without the vector symbol (→).
[0035]
[0036] The pseudo probe point T3 includes position information in three-dimensional space for each included point, and can be expressed as, for example, a position vector q.
[0037] The reflection profile point calculation unit P1 first extracts one of the pseudo exploration points T3, and calculates a reflection profile point T4 for the extracted pseudo exploration point T3. When the profile line T2 is expressed as in equation (1), the normal vector n(t) of the slice plane T5 at parameter t is parallel to the direction vector of the profile line T2, and is calculated by equation (2) in which each component is differentiated by t. Note that the vector n(t) will be written in the specification without the vector sign (→).
[0038]
[0039] Since the pseudo exploration point T3 and the reflection profile point T4 are both points on the slice plane T5, the position vector of the pseudo exploration point T3 is defined as q, and the position vector q-p(t) of the pseudo exploration point T3 relative to the reflection profile point T4 is perpendicular to the normal vector of the slice plane T5. Therefore, the parameters of the reflection profile point T4 can be obtained by solving equation (3) for t. Note that the vector q will be written in the specification without the vector sign (→).
[0040]
[0041] If there are multiple solutions to equation (3), there are multiple reflection profile points T4 for the pseudo probe point T3. Also, if there is no solution to equation (3), there is no reflection profile point T4 for the extracted pseudo probe point T3. In this case, the electromagnetic waves pseudo-transmitted from the pseudo probe point T3 are not received at the pseudo probe point T3.
[0042] All solutions to equation (3) are stored in association with the pseudo exploration point T3. The solutions to equation (3) may be stored in list form. However, if there is no solution to equation (3), an empty list is created. The reflection profile point calculation unit P1 creates an array of solutions to equation (3) corresponding to all pseudo exploration points T3 given as input, and sets this array as the reflection profile point T4. In other words, the reflection profile point T4 is calculated so that the dot product of the position vector of the reflection profile point T4, starting from the position of the pseudo exploration point T3, and the normal vector of the slice plane T5 perpendicular to the profile line T2 at the reflection profile point T4, is zero.
[0043] In addition to the analytical method described above, the reflection profile point T4 may be determined by, for example, emitting a large number of virtual straight-line rays from the pseudo-detection point T3 as by ray tracing, and determining the portions where they intersect perpendicularly with the profile line T2, or by calculating a slice plane T5 while gradually shifting points on the profile line T2, and recording the reflection profile point T4 when the pseudo-detection point T3 is on the slice plane T5.
[0044] The reference reflection profile point calculation unit P2 executes the same processing as the reflection profile point calculation unit P1 on the reference profile line R2 and reference detection point R3, creates an array of solutions to equation (3) corresponding to all reference detection points R3 given as input, and outputs the created array as the reference reflection profile point R4. In other words, the reference reflection profile point R4 is calculated so that the dot product of the position vector of the reference reflection profile point R4, starting from the position of the reference detection point R3, and the normal vector of the reference slice plane R5, which is perpendicular to the reference profile line R2 at the reference reflection profile point R4, is zero.
[0045] The reference pseudo exploration point calculation unit P3 applies a geometric transformation to the pseudo exploration point T3 by parallel translation and rotation so that the positional relationship between the reflection profile point T4 and the pseudo exploration point T3 on the profile line T2 matches the positional relationship between the reference reflection profile point R4 and the reference exploration point R3 on the reference profile line R2, calculates the position of the reference pseudo exploration point T3', and creates a list storing the reference pseudo exploration point T3' corresponding to each pseudo exploration point T3.
[0046] The buried pipe signal calculation unit P4 calculates the value of the pseudo buried pipe signal at the reference pseudo probe point T3' from the value of the reference probe point R3 and stores the value in a list associated with each pseudo probe point T3. The value of the pseudo buried pipe signal may be calculated by nearest neighbor interpolation, which assigns the reference buried pipe signal at the closest reference probe point R3, or by linear interpolation from surrounding reference probe points R3. This process is repeated for the number of reflection profile points T4 corresponding to the pseudo probe point T3, and the processing results are stored in a list. Once processing for all pseudo probe points T3 is complete, the values of the pseudo buried pipe signals in the list stored for each pseudo probe point T3 are added together to calculate the pseudo data D1 for the pseudo probe point T3. Note that for pseudo probe points T3 for which no reflection profile point T4 exists and for which no value exists in the list, an array in which all zeros are stored as pseudo data D1 is output.
[0047] In this embodiment, the reference buried pipe signal A1 may be generated using a simulator, or a buried pipe signal extracted from on-site survey data may be used. When using a simulator, the method described in Patent Document 1 requires a three-dimensional simulation each time the orientation, length, or other configuration of the buried pipe changes, resulting in high computational costs. In this embodiment, the reference buried pipe signal A1 can be obtained using a two-dimensional simulation. Furthermore, multiple data sets with different configurations can be prepared from a single reference buried pipe signal A1, allowing the user to obtain a large amount of data at low computational costs.
[0048] Example 1 is related to claims 1 to 5, and the elements of claims 1 to 5 are numbered as follows:
[0049] (Claim 1) A pseudo data creation system (S1) for creating pseudo data (D1) of a pseudo buried pipe (T1) simulating data acquired in buried pipe exploration using a ground penetrating radar method, the system being configured by a computer having an arithmetic unit for executing predetermined processing and a storage device connected to the arithmetic unit, the arithmetic unit comprising: a reflection profile point calculation unit (P1) for calculating a reflection profile point (T4) which is the foot of a perpendicular line dropped from the pseudo exploration point (T3) to the profile line (T2) based on a profile line (T2) representing the shape of the pseudo buried pipe (T1) using a straight line or a curve and a pseudo exploration point (T3) on a pseudo exploration plane (C1); The calculation device includes a reference reflection profile point calculation unit (P2) that calculates a reference reflection profile point (R4) that is the foot of a perpendicular line dropped from the reference profile point (T3) to the reference profile line (R2) based on a reference profile line (R2) represented by a straight line with a constant depth and a reference exploration point (R3) that is on an intersection (R6) between a plane (R5) perpendicular to the reference profile line (R2) and the pseudo exploration plane (C1); a reference pseudo probe point calculation unit (P3) configured to calculate, based on the pseudo probe point (T3), the profile line (T2), the reflection profile point (T4), the reference profile line (T2), and the reference reflection profile point (R4), a reference pseudo probe point (T3') obtained by applying a transformation to the pseudo probe point (T3) that matches a minute line segment of the profile line (T2) at the reflection profile point (T4) with a minute line segment of the reference profile line (R2) at the reference reflection profile point (R4); The pseudo data creation system (S1) is characterized in that the calculation device includes a buried pipe signal calculation unit (P4) that calculates a pseudo buried pipe signal at the reference pseudo exploration point (T3') based on the pseudo exploration point (T3), the reference pseudo exploration point (T3'), the reference exploration point (R3), and a reference buried pipe signal (A1) acquired at the reference exploration point (R3) for the buried pipe (R1) represented by the reference profile line (R2), and associates the pseudo buried pipe signal with the pseudo exploration point (T3) to calculate pseudo data (D1).
[0050] (Claim 2) A pseudo data creation system (S1) according to claim 1, wherein the reflection profile point calculation unit (P1) calculates the reflection profile point (T4) so that the dot product of a position vector of the reflection profile point (T4) starting from the position of the pseudo exploration point (T1) and a normal vector of a plane (T5) perpendicular to the profile line (T2) at the reflection profile point (T4) is zero.
[0051] (Claim 3) A pseudo data creation system (S1) according to claim 1, characterized in that the reference reflection profile point calculation unit (P2) calculates the reference reflection profile point (R4) so that the dot product of the position vector of the reference reflection profile point (R4) starting from the position of the reference inspection point (R3) and the normal vector of a plane (R5) perpendicular to the reference profile line at the reference reflection profile point (R4) is zero.
[0052] (Claim 4) A pseudo data creation system (S1) as described in claim 1, characterized in that the reference pseudo exploration point calculation unit (P3) calculates a reference pseudo exploration point (T3') by applying a geometric transformation to the pseudo exploration point (T3) using parallel translation and rotation so that the positional relationship between the reflection profile point (T4) and the pseudo exploration point (T3) coincides with the positional relationship between the reference reflection profile point (R4) and the reference exploration point (R3).
[0053] (Claim 5) A pseudo data creation system (S1) as described in claim 1, characterized in that the buried pipe signal calculation unit (P4) calculates the pseudo data (D1) at the reference pseudo exploration point (T3') by interpolation using the value of a reference exploration point (R3) near the reference pseudo exploration point (T3').
[0054] 4 is a diagram showing an example of functional blocks of a pseudo data creation system S2 according to Example 2. In Example 2, differences from Example 1 will be mainly described, and the same functions as those in Example 1 will be assigned the same reference numerals, and descriptions thereof will be omitted.
[0055] The pseudo data creation system S2 shown in FIG. 4 includes a reflection profile point calculation unit P1, a reference reflection profile point calculation unit P2, a reference pseudo probe point calculation unit P3, a buried pipe signal calculation unit P4, and a reference buried pipe signal calculation unit P5.
[0056] The pseudo data creation system S2 of the second embodiment differs from the pseudo data creation system S1 of the first embodiment in that one or more reference buried pipe signals A1 and identification information A2 that associates the reference buried pipe signal A1 with a point on the profile line T2 are added as input to the pseudo data creation system S2 of the second embodiment, and that a calculated corresponding reference buried pipe signal A1' is used as input to the buried pipe signal calculation unit P4 instead of the reference buried pipe signal A1. The other parts are the same as those of the pseudo data creation system S1 of the first embodiment.
[0057] For example, when profile line T2 is described by parameter t as in equation (1), identification information A2 can be expressed as a function that returns a corresponding reference buried pipe signal A1 depending on the value of t. The domain of t for which identification information A2 can return reference buried pipe signal A1 may be the entire range of values of t that profile line T2 can take, or a subset thereof.
[0058] The reference buried pipe signal calculation unit P5 determines whether identification information A2 exists on the profile line T2. If identification information A2 exists, it extracts a reference buried pipe signal A1 that matches the identification information A2 from the input and outputs the extracted corresponding reference buried pipe signal A1'. If identification information A2 does not exist, it searches for a point on the profile line T2 near the reflection profile point T4 where identification information A2 exists. If such a point is found near the reflection profile point T4, it calculates the corresponding reference buried pipe signal A1' at the reflection profile point T4 by interpolation.
[0059] As an example of interpolation processing, for example, the buried pipe reflection signal can be translated and scaled. When the size of the buried pipe can be ignored and the ground penetrating radar can be approximated as a point reflection, it is known that the buried pipe signal present at a depth h from the detection surface will be a hyperbola expressed by equation (4). In equation (4), c represents the speed of light in a vacuum, and εr represents the relative dielectric constant of the ground.
[0060]
[0061] Here, in order to convert the hyperbola with vertex (u0, v0+b) expressed by equation (5) into a hyperbola with vertex (u1, v1+b') expressed by equation (6), a linear transformation shown in equation (7) can be performed. In other words, a hyperbola whose axes are parallel to the coordinate axes can be converted into any hyperbola whose axes are parallel to the coordinate axes by translation and scaling.
[0062]
[0063] Therefore, when there are two or more reference buried pipe signals A1, the parameters a, b, u0, and v0 of the most matching hyperbola are calculated for the reference buried pipe signal A1 near the reflection profile point T4, and the parameters of the hyperbola at the reflection profile point T4 of interest are calculated by linear interpolation from the position of the reflection profile point T4 having the reference buried pipe signal A1 whose parameters were calculated and the position of the reflection profile point T4 of interest.The reference buried pipe signal A1 whose parameters were calculated is then translated and scaled so that it has a shape that matches the parameters at the reflection profile point T4 of interest, and the corresponding reference buried pipe signal A1' for the reflection profile point T4 of interest can be calculated by interpolation.
[0064] The interpolation method is not limited to linear interpolation, and other methods may be used. For example, nearest neighbor interpolation may be used, which outputs the reference buried pipe signal A1 at the point on the profile line T2 nearest to the reference buried pipe signal A1. Alternatively, optical flow may be used to calculate changes in the feature points of the reference buried pipe R1 for multiple reference buried object signals near the reflection profile point T4, and the image may be interpolated assuming that the same changes have been made to nearby pixels. Alternatively, the values of the reference buried pipe signal A1 may be continuously interpolated using a machine learning model trained on the values of the reference buried pipe signal A1.
[0065] The pseudo data creation system S2 of the second embodiment has the advantage that, for example, when the diameter of a buried pipe changes from one position to another, or when the material of the buried pipe or the composition of the substances in the buried pipe changes from one position to another, the user can create appropriate buried pipe signal data by inputting multiple corresponding reference buried pipe signals A1. Furthermore, for example, when the reference buried pipe signal A1 continuously changes due to changes in the depth or diameter of the buried pipe, the pseudo data creation system S2 has the advantage that, even for areas where there is no reference buried object signal, a corresponding reference buried pipe signal A1' can be calculated from several reference buried pipe signals A1.
[0066] The second embodiment is related to claims 6 and 7, and the elements of claims 6 and 7 are given the following reference numerals.
[0067] (Claim 6) A pseudo data creation system (S1) according to claim 1, wherein the arithmetic device is provided with a reference buried pipe signal calculation unit (P5) that calculates a corresponding reference buried pipe signal (A1') corresponding to the reflection profile point (T4) based on the reflection profile point (T4), one or more of the reference buried pipe signals (A1), and identification information (A2) that associates the reference buried pipe signal (A1) with a point on the profile line (T2), and outputs the calculated corresponding reference buried pipe signal (A1') to the buried pipe signal calculation unit (P4), and the reference buried pipe signal calculation unit (P4) receives as input the calculated corresponding reference buried pipe signal (A1') instead of the reference buried pipe signal (A1).
[0068] (Claim 7) A pseudo data creation system (S2) as described in claim 6, characterized in that the reference buried pipe signal calculation unit (P5) calculates the corresponding reference buried pipe signal (A1') by data processing of at least one of parallel translation, enlargement, and reduction so that the hyperbola representing the corresponding reference buried pipe signal (A1') coincides.
[0069] 5 is a diagram showing an example of functional blocks of a pseudo data creation system S3 according to Example 3. In Example 3, differences from Example 1 described above will be mainly explained, and the same functions as those in Example 1 will be assigned the same reference numerals, and explanations thereof will be omitted.
[0070] The pseudo data creation system S3 shown in Figure 5 has a reflection profile point calculation unit P1, a reference reflection profile point calculation unit P2, a reference pseudo exploration point calculation unit P3, a buried pipe signal calculation unit P4, a reference buried pipe signal calculation unit P5, and a background synthesis unit P6.
[0071] The pseudo data creation system S3 of the third embodiment differs from the pseudo data creation system S1 of the first embodiment in that a background A3 that does not include the reference buried pipe signal A1 is added as an input to the pseudo data creation system S3 of the third embodiment. Other parts are the same as the pseudo data creation system S1 of the first embodiment.
[0072] The background A3 may be generated using a simulator or may be acquired on-site. When an on-site acquired background is used, it is preferable to use background data acquired at a location where the absence of buried pipes has been confirmed by test drilling or the like so as not to include the reference buried pipe signal A1.
[0073] The background synthesis unit P6 synthesizes the background A3 and the pseudo data D1 to generate synthetic data D2 and outputs the generated synthetic data D2. For example, the background synthesis unit P6 may add the normalized pseudo data D1 and the normalized background A3 together, or may replace the background A3 in the portion where the pseudo buried pipe signal data exists with the pseudo buried pipe signal data.
[0074] The pseudo data creation system S3 of the third embodiment has the effect of enabling the user to create data that is close to the results of an actual underground radar survey by combining background data with a buried pipe signal.
[0075] The third embodiment is related to claim 8, and the elements of claim 8 are given the following reference numerals.
[0076] (Claim 8) A pseudo data creation system (S1) as described in claim 1, characterized in that the calculation device is provided with a background synthesis unit (P6) that calculates synthesized data (D2) by synthesizing the pseudo data (D1) and the background (A1) based on a background (A1) that does not include buried pipe signals and the pseudo data (D1).
[0077] <Embodiment 4> Fig. 6 is a diagram showing an example of functional blocks of a pseudo data creation system S4 of embodiment 4. In the example shown in Fig. 6, differences from the above-mentioned embodiment 3 will be mainly explained, and the same functions as those in embodiment 3 will be assigned the same reference numerals, and explanations thereof will be omitted.
[0078] The pseudo data creation system S4 shown in FIG. 6 includes a reflection profile point calculation unit P1, a reference reflection profile point calculation unit P2, a reference pseudo inspection point calculation unit P3, a buried pipe signal calculation unit P4, a background synthesis unit P6, and a background expansion unit P7.
[0079] The pseudo data creation system S4 of the fourth embodiment differs from the pseudo data creation system S3 of the third embodiment in that a background extension parameter A4 is added as an input to the pseudo data creation system S4 of the fourth embodiment. Other parts are the same as the pseudo data creation system S1 of the first embodiment and the pseudo data creation system S3 of the third embodiment.
[0080] The background extension unit P7 receives the background A3 and the background extension parameter A4, calculates an extended background D3, and outputs the calculated extended background D3 to the background synthesis unit P6. For example, the background extension unit P7 performs processes such as left-right inversion and contrast adjustment according to the background extension parameter A4.
[0081] Fig. 7 is a diagram showing an example of a functional block of another pseudo data creation system S5 according to Example 4. In the example shown in Fig. 7, differences from Example 1 described above will be mainly explained, and the same functions as those in Example 1 will be assigned the same reference numerals, and explanations thereof will be omitted.
[0082] The pseudo data creation system S5 shown in Figure 7 includes a reflection profile point calculation unit P1, a reference reflection profile point calculation unit P2, a reference pseudo probe point calculation unit P3, a buried pipe signal calculation unit P4, and a reference buried pipe signal expansion unit P8. It differs from the pseudo data creation system S1 in that a reference buried pipe signal expansion parameter A5 is added as an input to the pseudo data creation system S5 shown in Figure 7. The other parts are the same as the pseudo data creation system S1 of the first embodiment.
[0083] The reference buried pipe signal expansion unit P8 receives the reference buried pipe signal A1 and the reference buried pipe signal expansion parameter A5, modifies the reference buried pipe signal A1, calculates an expanded reference buried pipe signal D4, and outputs the calculated expanded reference buried pipe signal D4 to the buried pipe signal calculation unit P4. The buried pipe signal changes its hyperbolic width and depth-wise attenuation rate depending on the magnitude of the soil's dielectric constant. Furthermore, the buried pipe signal is distorted by surface irregularities and localized substances such as cavities and pebbles. To reproduce such phenomena, the reference buried pipe signal expansion unit P8 expands or reduces the reference buried pipe signal A1, changes its depth-wise attenuation rate, or adds local distortions according to the reference buried pipe signal expansion parameter A5.
[0084] The pseudo data creation system S4 and pseudo data creation system S5 of Example 4 can create pseudo data D1 and synthetic data D2 having a wide variety of extended reference buried pipe signals D4 and extended backgrounds D3 based on a small number of reference buried pipe signals A1 and backgrounds A3, thereby achieving the effect of improving the accuracy of machine learning using the created data.
[0085] The fourth embodiment is related to claims 9 and 10, and the elements of claims 9 and 10 are given the following reference numerals.
[0086] (Claim 9) A pseudo data creation system (S3) as described in claim 8, characterized in that it has a background extension unit (P7) that performs at least one of data processing of left-right inversion and contrast adjustment on an input background (A3) based on background extension parameters (A4) for extending the background (A3), calculates an extended background (D3), and outputs the calculated extended background (D3) to the background synthesis unit (P6).
[0087] (Claim 10) The pseudo data creation system (S1) according to claim 1, further comprising a reference buried pipe signal expansion unit (P8) that calculates an expanded reference buried pipe signal (D4) that is an expanded reference buried pipe signal by performing at least one of data processing of changing the attenuation rate, expanding, contracting, and applying distortion to the input reference buried pipe signal (A1) based on a reference buried pipe signal expansion parameter (A5) for expanding the reference buried pipe signal (A1), and outputs the calculated expanded reference buried pipe signal (D4) to the buried pipe signal calculation unit (P4). The buried pipe signal calculation unit (P4) calculates a pseudo buried pipe signal at the reference pseudo exploration point (T3') based on the extended reference buried pipe signal (D4) instead of the reference buried pipe signal (A1), the pseudo exploration point (T3), the reference pseudo exploration point (T3'), and the reference exploration point (R3), and associates the pseudo buried pipe signal with the pseudo exploration point (T3) to calculate pseudo data (D1).
[0088] 8 is a diagram showing an example of functional blocks of a pseudo data creation system S1′ according to Example 5. In Example 5, differences from Example 1 described above will be mainly explained, and the same functions as those in Example 1 will be assigned the same reference numerals, and explanations thereof will be omitted.
[0089] The pseudo data creation system S1' shown in Figure 8 includes a reflection profile point calculation unit P1, a reference reflection profile point calculation unit P2, a reference pseudo probe point calculation unit P3, and a buried pipe signal calculation unit P4'. The pseudo data creation system S1' of Example 5 differs from the pseudo data creation system S1 of Example 1 in that a profile line T2 and a reflection profile point T4 are added as inputs to the buried pipe signal calculation unit P4'. The other parts are the same as the pseudo data creation system S1 of Example 1.
[0090] The buried pipe signal calculation unit P4' calculates pseudo data D1 at the pseudo inspection point T3 based on the shape of the profile line T2 at the reflection profile point T4. For example, the buried pipe signal calculation unit P4' calculates a rectangular parallelepiped having sides parallel to the profile line T2 in the area near the reflection profile point T4, and determines the addition coefficient by dividing the length of the portion of the profile line T2 included in the rectangular parallelepiped by the length of one side parallel to the profile line T2. Furthermore, if the addition coefficient is a value greater than 1, the addition coefficient is set to 1. The addition coefficients of all reflection profile points T4 that exist for the pseudo inspection point T3 are calculated. The pseudo buried pipe signal at the pseudo inspection point T3 is determined as the sum of values obtained by multiplying the buried pipe signals calculated by interpolation from the reference buried pipe signal A1 by the addition coefficients.
[0091] The pseudo data creation system S1' of the fifth embodiment has the effect of allowing the user to calculate pseudo data D1 that is closer to actual data obtained by the underground radar method by taking into account the shape of the profile line T2 around the reflection profile point T4.
[0092] The fifth embodiment is related to claim 11, and the elements of claim 11 are given the following reference numerals.
[0093] (Claim 11) A pseudo data creation system (S1) according to claim 1, characterized in that the buried pipe signal calculation unit (P4) calculates an additive coefficient from the shape of the profile line (T2) in the vicinity of the reflection profile point (T4) based on the pseudo exploration point (T3), the reference pseudo exploration point (T3'), the reference exploration point (R3), the reference buried pipe signal (A1), the profile line (T2), and the reflection profile point (T4), and adds a value obtained by multiplying the pseudo buried pipe signal at the reference pseudo exploration point (T3') by the calculated additive coefficient to calculate pseudo data (D1) at the pseudo exploration point (T3).
[0094] 9 is a diagram showing an example of functional blocks of a pseudo data creation system S6 according to Example 6. In Example 6, differences from Example 1 described above will be mainly explained, and the same functions as those in Example 1 will be assigned the same reference numerals, and explanations thereof will be omitted.
[0095] 9 includes a reflection profile point calculation unit P1, a reference reflection profile point calculation unit P2, a reference pseudo probe point calculation unit P3, a buried pipe signal calculation unit P4, a pseudo probe point extraction unit P9, and a buried pipe signal duplication unit P11. The input to the pseudo data creation system S6 of Example 6 differs from that of the pseudo data creation system S1 in that pseudo probe point group information A8 is added.
[0096] The pseudo probe point extraction unit P9 takes as input the pseudo probe point T3 and the pseudo probe point group information A8, extracts one pseudo probe point T3 for each group, and outputs the extracted pseudo probe point T3 as the extracted pseudo probe point T3". The reflection profile point P1, the reference pseudo probe point calculation unit P3, and the buried pipe signal calculation unit P4 take as input the extracted pseudo probe point T3" instead of the pseudo probe point T3. The buried pipe signal duplicating unit P11 takes as input the pseudo probe point T3, the pseudo probe point group information A8, and the pseudo data D1, and for those pseudo probe points T3 that do not have a pseudo buried pipe signal associated with them in the pseudo data D1, it calculates and outputs the overall pseudo data D1' based on the pseudo probe point group information by duplicating and associating the pseudo buried pipe signals of pseudo probe points that belong to the same group and have a pseudo buried pipe signal in the pseudo data D1. The other parts are the same as those of the pseudo data creation system S1 of the first embodiment.
[0097] The pseudo buried pipe signals are approximated as being the same and grouped in the pseudo probe point group information A8 when, for example, the profile line T2 includes a portion that can be approximated as a straight line with a constant depth, and the pseudo probe point T3 is arranged almost parallel to that portion. In this case, the positional relationship between the pseudo probe point T3 and the reflection profile point T4 is almost the same, so the pseudo buried pipe signals can be approximated as being the same.
[0098] The pseudo data creation system S6 of Example 6 calculates a pseudo buried pipe signal for one pseudo exploration point T3 extracted for each group for pseudo exploration points where the pseudo buried pipe signal can be approximated to have the same value, thereby eliminating the need to calculate reflection profile points T4 and pseudo buried pipe signals for all pseudo exploration points T3, thereby having the effect of being able to create data faster than the pseudo data creation system S1 of Example 1.
[0099] The sixth embodiment is related to claim 12, and the elements of claim 12 are given the following reference numerals.
[0100] (Claim 12) A pseudo data creation system (S1) according to claim 1, comprising: a pseudo probe point extraction unit (P9) that receives as input pseudo probe point group information (A8) that groups pseudo probe points that can be approximated as having the same value as the pseudo probe point (T3), extracts one pseudo probe point for each group, and outputs the extracted pseudo probe point (T3''); and a reflection profile point calculation unit (P1), a reference pseudo probe point calculation unit (P3), and a buried pipe signal calculation unit (P4) that receive as input the extracted pseudo probe point (T3'') instead of the pseudo probe point (T3). A pseudo data creation system (S6) characterized by comprising a buried pipe signal duplication unit (P11) that receives as input the pseudo exploration points (T3), the pseudo exploration point group information (A8), and the pseudo data (D1), and calculates and outputs the entire pseudo data (D1') by duplicating and associating the pseudo buried pipe signal of a pseudo exploration point (T3) belonging to the same group with the pseudo buried pipe signal of the pseudo exploration point (T3) that does not have the pseudo buried pipe signal associated with it.
[0101] 10 is a diagram showing an example of functional blocks of a pseudo data creation system S7 according to Example 7. In Example 7, differences from Example 1 described above will be mainly explained, and the same functions as those in Example 1 will be assigned the same reference numerals, and explanations thereof will be omitted.
[0102] The pseudo data creation system S7 shown in FIG. 10 includes a reflection profile point calculation unit P1, a reference reflection profile point calculation unit P2, a reference pseudo probe point calculation unit P3, a buried pipe signal calculation unit P4, and a learning unit P10.
[0103] The pseudo data creation system S7 of Example 7 differs from the pseudo data creation system S1 in that annotation information A6 and a model A7 to be assigned to the training data are added as inputs to the pseudo data creation system S7 of Example 7. A learning unit P10 learns the model A7 using the pseudo data D1, annotation information A6, and the model A7 as inputs. The other parts are the same as the pseudo data creation system S1 of Example 1.
[0104] The annotation information A6 includes information to be output to the model A7 and auxiliary input information to be provided to the model A7 together with the pseudo data D1. For example, the annotation information A6 includes information on the profile line T2, the pipe diameter, the pipe type, the degree of deterioration of the pipe, information on the substances contained in the pipe, the amount, depth, length, laying direction, and dielectric constant of the soil.
[0105] Model A7 may be a model capable of supervised learning, such as a neural network or support vector machine that outputs annotation information A6 based on pseudo data D1, or it may be a model capable of unsupervised learning, such as a generative adversarial network or variational autoencoder that does not use annotation information A6 as the correct answer.
[0106] The pseudo data creation system S7 of the seventh embodiment has the effect of improving the performance of the learning model by quickly generating learning data for machine learning.
[0107] The seventh embodiment is related to claim 13, and the elements of claim 13 are given the following reference numerals.
[0108] (Claim 13) A pseudo data creation system (S1) as described in claim 1, characterized in that it includes a learning unit (P10) that performs machine learning of a model (A7) using learning data that associates the pseudo data (D1) with annotation information (A6).
[0109] The present invention is not limited to the above-described embodiments, and includes various modifications and equivalent configurations within the spirit and scope of the appended claims. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to configurations including all of the described configurations. Furthermore, part of the configuration of one embodiment may be replaced with the configuration of another embodiment. Furthermore, the configuration of another embodiment may be added to the configuration of one embodiment. Furthermore, part of the configuration of each embodiment may be added, deleted, or replaced with other configurations.
[0110] Furthermore, the aforementioned configurations, functions, processing units, processing means, etc. may be realized in part or in whole in hardware, for example by designing them as integrated circuits, or may be realized in software by a processor interpreting and executing a program that realizes each function.
[0111] Information such as programs, tables, and files that realize each function can be stored in a storage device such as a memory, hard disk, or SSD (Solid State Drive), or in a recording medium such as an IC card, SD card, or DVD.
[0112] In addition, the control lines and information lines shown are those that are considered necessary for the explanation, and do not necessarily show all the control lines and information lines that are necessary for implementation. In reality, it can be considered that almost all components are interconnected.
Claims
1. A pseudo data creation system for creating pseudo data of a pseudo buried pipe that imitates data obtained in buried pipe exploration using underground radar, comprising a computer having an arithmetic unit for executing predetermined processes and a storage device connected to the arithmetic unit, wherein the arithmetic unit calculates reflection profile points that are the feet of perpendicular lines dropped from pseudo exploration points to a profile line that represents the shape of the pseudo buried pipe using straight or curved lines and pseudo exploration points on a pseudo exploration plane; and a reference reflection profile point calculation unit calculates reference reflection profile points that are the feet of perpendicular lines dropped from reference exploration points to a reference profile line that represents a reference profile line that is represented by a straight line with a constant depth and reference exploration points that are on the intersection of a plane perpendicular to the reference profile line and the pseudo exploration plane. a reference pseudo probe point calculation unit that calculates a reference pseudo probe point, which is a point obtained by applying a transformation to the pseudo probe point that matches a minute line segment of the profile line at the reflection profile point with a minute line segment of the reference profile line at the reference reflection profile point, based on the pseudo probe point, the profile line, the reflection profile point, the reference profile line, and the reference reflection profile point; and a buried pipe signal calculation unit that calculates a pseudo buried pipe signal at the reference pseudo probe point based on the pseudo probe point, the reference pseudo probe point, the reference probe point, and a reference buried pipe signal obtained at the reference probe point for the buried pipe represented by the reference profile line, and associates the pseudo buried pipe signal with the pseudo probe point to calculate pseudo data.
2. A pseudo data creation system as described in claim 1, characterized in that the reflection profile point calculation unit calculates the reflection profile point so that the dot product of the position vector of the reflection profile point, starting from the position of the pseudo exploration point, and the normal vector of a plane perpendicular to the profile line at the reflection profile point, is zero.
3. A pseudo data creation system as described in claim 1, characterized in that the reference reflection profile point calculation unit calculates the reference reflection profile point so that the dot product of the position vector of the reference reflection profile point, starting from the position of the reference inspection point, and the normal vector of a plane perpendicular to the reference profile line at the reference reflection profile point, is zero.
4. A pseudo data creation system as described in claim 1, characterized in that the reference pseudo exploration point calculation unit calculates the reference pseudo exploration point by applying a geometric transformation to the pseudo exploration point by parallel translation and rotation so that the positional relationship between the reflection profile point and the pseudo exploration point matches the positional relationship between the reference reflection profile point and the reference exploration point.
5. A pseudo data creation system as described in claim 1, characterized in that the buried pipe signal calculation unit calculates pseudo data at the reference pseudo exploration point by interpolation using values of reference exploration points in the vicinity of the reference pseudo exploration point.
6. A pseudo data creation system as described in claim 1, characterized in that the arithmetic device is provided with a reference buried pipe signal calculation unit that calculates a corresponding reference buried pipe signal corresponding to the reflection profile point based on the reflection profile point, one or more of the reference buried pipe signals, and identification information that associates the reference buried pipe signal with a point on the profile line, and outputs the calculated corresponding reference buried pipe signal to the buried pipe signal calculation unit, and the reference buried pipe signal calculation unit receives the calculated corresponding reference buried pipe signal as input instead of the reference buried pipe signal.
7. A pseudo data creation system as described in claim 6, characterized in that the reference buried pipe signal calculation unit calculates the corresponding reference buried pipe signal by performing at least one of data processing including parallel translation, expansion, and contraction so that the hyperbola representing the corresponding reference buried pipe signal coincides.
8. A pseudo data creation system as described in claim 1, characterized in that the arithmetic device is provided with a background synthesis unit that calculates synthesized data by synthesizing the pseudo data and the background based on a background that does not include buried pipe signals and the pseudo data.
9. A pseudo data creation system as described in claim 8, characterized in that it has a background extension unit that performs at least one of data processing, such as left-right inversion and contrast adjustment, on an input background based on background extension parameters for extending the background, to calculate an extended background, and outputs the calculated extended background to the background synthesis unit.
10. A pseudo data creation system as claimed in claim 1, comprising a reference buried pipe signal expansion section which calculates an expanded reference buried pipe signal, which is an expanded reference buried pipe signal, by performing at least one of data processing operations of changing the attenuation rate, expanding, contracting and distorting the input reference buried pipe signal based on a reference buried pipe signal expansion parameter for expanding the reference buried pipe signal, and outputs the calculated expanded reference buried pipe signal to the buried pipe signal calculation section, wherein the buried pipe signal calculation section calculates a pseudo buried pipe signal at the reference pseudo exploration point based on the expanded reference buried pipe signal in place of the reference buried pipe signal, the pseudo exploration point, the reference pseudo exploration point and the reference exploration point, and calculates pseudo data by correlating the pseudo buried pipe signal with the pseudo exploration point.
11. A pseudo data creation system as described in claim 1, wherein the buried pipe signal calculation unit calculates an additive coefficient from the shape of the profile line in the vicinity of the reflection profile point based on the pseudo exploration point, the reference pseudo exploration point, the reference exploration point, the reference buried pipe signal, the profile line, and the reflection profile point, and adds a value obtained by multiplying the pseudo buried pipe signal at the reference pseudo exploration point by the calculated additive coefficient, thereby calculating pseudo data at the pseudo exploration point.
12. A pseudo data creation system as claimed in claim 1, comprising: a pseudo probe point extraction unit that receives as input pseudo probe point group information that groups pseudo probe points that can be approximated as having the same value as the pseudo buried pipe signal, extracts one pseudo probe point for each group, and outputs it as an extracted pseudo probe point; the reflection profile point calculation unit, reference pseudo probe point calculation unit, and buried pipe signal calculation unit that receive as input the extracted pseudo probe points instead of the pseudo probe points; and a buried pipe signal duplication unit that receives as input the pseudo probe points, the pseudo probe point group information, and the pseudo data, and, for those pseudo probe points to which no pseudo buried pipe signal is associated, duplicates and associates the pseudo buried pipe signal with the pseudo buried pipe signal of a pseudo probe point belonging to the same group, thereby calculating and outputting the entire pseudo data.
13. A pseudo data creation system as described in claim 1, characterized in that it comprises a learning unit that performs machine learning of a model using learning data that associates the pseudo data with annotation information.
14. A pseudo data creation method in which a pseudo data creation system creates pseudo data of a pseudo buried pipe simulating data acquired in buried pipe exploration using a ground penetrating radar method, the pseudo data creation system being configured by a computer having an arithmetic unit that executes predetermined processing and a storage device connected to the arithmetic unit, the pseudo data creation method comprising: a reflection profile point calculation procedure in which the arithmetic unit calculates, based on a profile line that represents the shape of the pseudo buried pipe using a straight line or a curve and pseudo exploration points on a pseudo exploration plane, reflection profile points that are the feet of perpendicular lines that are dropped from the pseudo exploration points to the profile line; and a reference reflection profile point calculation procedure in which the arithmetic unit calculates, based on a reference profile line that is represented by a straight line with a constant depth and reference exploration points that are on the intersection of the pseudo exploration plane with a plane perpendicular to the reference profile line, reference reflection profile points that are the feet of perpendicular lines that are dropped from the reference exploration points to the reference profile line. a reference pseudo probe point calculation procedure in which the arithmetic device calculates a reference pseudo probe point, which is a point obtained by applying a transformation to the pseudo probe point that matches a minute line segment of the profile line at the reflection profile point with a minute line segment of the reference profile line at the reference reflection profile point, based on the pseudo probe point, the profile line, the reflection profile point, the reference profile line, and the reference reflection profile point; and a buried pipe signal calculation procedure in which the arithmetic device calculates a pseudo buried pipe signal at the reference pseudo probe point based on the pseudo probe point, the reference pseudo probe point, the reference probe point, and a reference buried pipe signal obtained at the reference probe point for the buried pipe represented by the reference profile line, and associates the pseudo buried pipe signal with the pseudo probe point to calculate pseudo data.
15. A program for causing a pseudo data creation system to create pseudo data of a pseudo buried pipe simulating data obtained in buried pipe exploration using a ground penetrating radar method, the pseudo data creation system being composed of a computer having an arithmetic unit that executes predetermined processing and a storage device connected to the arithmetic unit, the program comprising: a reflection profile point calculation procedure for calculating, based on a profile line that represents the shape of the pseudo buried pipe using a straight or curved line and pseudo exploration points on a pseudo exploration plane, reflection profile points that are the feet of perpendicular lines that are dropped from the pseudo exploration points to the profile line; and a reference reflection profile point calculation procedure for calculating, based on a reference profile line that is represented by a straight line with a constant depth and reference exploration points that are on the intersection of the pseudo exploration plane with a plane perpendicular to the reference profile line, reference reflection profile points that are the feet of perpendicular lines that are dropped from the reference exploration points to the reference profile line. a reference pseudo probe point calculation procedure for calculating a reference pseudo probe point, which is a point obtained by applying a transformation to the pseudo probe point that matches a minute line segment of the profile line at the reflection profile point with a minute line segment of the reference profile line at the reference reflection profile point, based on the pseudo probe point, the profile line, the reflection profile point, the reference profile line, and the reference reflection profile point; and a buried pipe signal calculation procedure for calculating a pseudo buried pipe signal at the reference pseudo probe point based on the pseudo probe point, the reference pseudo probe point, the reference probe point, and a reference buried pipe signal obtained at the reference probe point for a buried pipe represented by the reference profile line, and associating the pseudo buried pipe signal with the pseudo probe point to calculate pseudo data.
Citation Information
Patent Citations
Measurement method and underground rader device
JP2016211897A
Learning data generation method and target space state recognition method employing the same
JP2018041178A
Data processing method for underground radar, data processing program, and underground radar device
JP2021124290A
Underground structure detection apparatus and underground structure detection method
JP2023134305A
Ground penetrating radar and deep learning-based underground pipeline detection method and system
US20230108634A1