JP4105139B2 - Three-dimensional image generation method and program - Google Patents

Description

  The present invention relates to a method and program for generating a three-dimensional image of an underground geological structure in elastic wave exploration.

Conventionally, by performing elastic wave exploration, underground cross section data is measured at multiple points, and pseudo data is generated based on the cross section data, thereby interpolating the pseudo data into areas where no data exists. Therefore, techniques for investigating underground structures are known. Hereinafter, the processing flow of the three-dimensional image generation method used in this conventional technique will be described with reference to the flowchart shown in FIG.
In the conventional three-dimensional image generation method, first, data of a plurality of two-dimensional underground sectional views are measured (step S01), and the structure of the inclined surface existing underground is estimated based on the data of the underground sectional views. The “structural tilt model” has been constructed by performing “horizon interpretation” processing or “tilt reading” processing to detect the corresponding position in the measured cross-sectional view and obtaining the tilt between the two points. (Step S02).

In step S02, when “horizon interpretation” is performed based on the data of the two-dimensional sectional view measured in step S01, an inclined surface extending in the xy plane direction as shown in FIG. Is obtained. The model obtained by the processing of the “horizon interpretation” has an advantage that a planar structural inclination can be obtained with respect to the xy plane direction, but information on the t-axis direction indicating the underground depth direction is sparse. There are disadvantages.
On the other hand, when the “tilt reading” process is performed in step S02, data in which cross-sectional views on the xt plane and the yt plane as shown in FIG. 23B are obtained is obtained. While there is a merit that the information in the t-axis direction indicating the depth direction is dense, there is a demerit that information on the structure inclination can be obtained only on a line where the xt plane and the yt plane intersect.

In the conventional method of constructing a structural tilt model, the data obtained by the “horizon interpretation” process differs from the data obtained by the “tilt reading” process. Both of “inclination reading on the sectional view” could not be handled simultaneously. Therefore, there is a problem that the accuracy of the three-dimensional image to be created cannot be improved.
Returning to FIG. 22, an interpolation method such as a moving average method or an extrapolation method such as a finite difference method is performed based on the structural gradient model constructed in step S <b> 02 by performing the processing of “horizon interpretation” or “tilt reading”. Is used, “wave field interpolation” processing for generating pseudo data is performed on the region other than the region where the data of the two-dimensional sectional view measured in step S01 exists (step S03).
Finally, a “3D migration” process that visualizes (images) a three-dimensional image of a model in which the three-dimensional space is filled with pseudo data is performed by the “wave field interpolation” process in step S03 (step S03). S04).

However, in the conventional three-dimensional image generation method, only one of “horizontal interpretation” and “tilt reading” processing is performed when constructing the structural gradient model, or the wave motion using the moving average algorithm is performed. Because predetermined conditions were not set when performing field interpolation, there was a problem that data could not be completely interpolated from irregular two-dimensional data, a problem that false images were likely to occur during extrapolation, and a large amount of data to be input There is a problem that the resolution is lowered when there is a gap, and further, there is a problem that the image is greatly deteriorated when a three-dimensional image is generated.
As techniques for generating an original three-dimensional image from missing data, those described in Non-Patent Document 1 and Non-Patent Document 2 are known. However, these techniques cannot generate accurate pseudo data and cannot accurately create a three-dimensional image of an underground geological structure.
Jason Chin-Sen Kao, William A. Schneider, Walter Whitman, "Automated interpolation of two-dimensional seismic grids into three-dimensional data volume", GEOPHYSICS. VOL.55, NO.4 (APRIL 1990), p.433-442 . M. Hugh Miller, Jr. William S. French, "Time slices from two-dimensional seismic surveys", GEOPHYSICS, VOL.54, NO.1 (JANUARY 1989), p.13-22.

  The present invention has been made in consideration of the above circumstances, and an object thereof is to provide a three-dimensional image generation method and program capable of generating a highly accurate three-dimensional image from missing data.

  The invention according to claim 1 is a first step of inputting two-dimensional or three-dimensional elastic wave exploration data, and the two-dimensional or three-dimensional elastic wave exploration data input in the first step is a three-dimensional space. A second step in which binning is performed by arranging on the top, a third step in which a structure diagram is created by performing structural interpretation based on the data binned in the second step, and the second step On the basis of the data binned in step (4), the fourth step for performing tilt reading, the structure diagram created in the third step, and the tilt data read in the fourth step, The fifth step of building a hybrid structure tilt model, the data binned in the second step, and the hybrid structure tilt created in the fifth step Based on the model, a sixth step for performing wave field interpolation by generating pseudo data using a moving average algorithm for a region where data is missing, and wave field interpolation in the sixth step And a seventh step of performing migration by converting the data obtained by the three-dimensional imaging into a three-dimensional image.

  The invention according to claim 2 uses the input data of a specified number or more when calculating the pseudo data in the calculation of the moving average algorithm when performing the wave field interpolation in the sixth step. The three-dimensional image generation method according to claim 1, wherein the three-dimensional image generation method is set as a condition.

  According to a third aspect of the present invention, there is at least one input data within a predetermined distance from the point at which the pseudo data is output in the calculation of the moving average algorithm when performing wave field interpolation in the sixth step. The three-dimensional image generation method according to claim 2, wherein the three-dimensional image generation method is set as a condition.

  The invention according to claim 4 is the three-dimensional image generation method according to claim 3, wherein the moving average algorithm is calculated by setting the predetermined distance to one grid.

  According to a fifth aspect of the present invention, the first step of inputting two-dimensional or three-dimensional elastic wave exploration data, and the two-dimensional or three-dimensional elastic wave exploration data input in the first step are three. A second step of binning by arranging in a dimensional space; a third step of creating a structural diagram by performing structural interpretation based on the data binned in the second step; Based on the fourth step of performing tilt reading based on the data binned in the second step, the structure diagram created in the third step, and the tilt data read in the fourth step The fifth step of building a hybrid structure gradient model, the data binned in the second step, and the hybrid created in the fifth step A sixth step of performing wave field interpolation by generating pseudo data using a moving average algorithm for a region where data is missing based on the gradient model, and in the wave field in the sixth step It is a three-dimensional image generation program for causing a computer to execute a seventh step of performing migration by converting data obtained by insertion into a three-dimensional image.

  The invention according to claim 6 uses the input data of a specified number or more when calculating the pseudo data in the calculation of the moving average algorithm when performing the wave field interpolation in the sixth step. The three-dimensional image generation program according to claim 5, wherein the three-dimensional image generation program is set as a condition.

  According to a seventh aspect of the present invention, there is at least one input data within a predetermined distance of a point at which pseudo data is output in the calculation of the moving average algorithm when performing wave field interpolation in the sixth step. 7. The three-dimensional image generation program according to claim 6, wherein the three-dimensional image generation program is set as a condition.

  The invention according to claim 8 is the three-dimensional image generation program according to claim 7, wherein the moving average algorithm is calculated by setting the predetermined distance to one grid.

According to the first aspect of the present invention, the hybrid structure gradient model is constructed in the fifth step based on the structural diagram created in the third step and the gradient data read in the fourth step.
The data of the structure diagram obtained in the third step and the slope data read in the fourth step have different data properties, so the conventional three-dimensional image generation method could not handle the data simultaneously. By synthesizing these data in the same dimension, the amount of data to be handled is increased, and a more accurate three-dimensional image can be generated.

Further, the invention according to claim 2 performs the calculation based on "use the input data of the specified number or more when calculating the pseudo data" when calculating the moving average algorithm. did.
Thus, when generating pseudo data in the sixth step, it is possible to prevent a large amount of simple copies of the elastic wave exploration data input in the first step from being generated, and to improve the quality of the three-dimensional image. It is possible to prevent the generation of false images and the occurrence of repetitive patterns that cause a decrease in image quality.

In the invention according to claim 3, when performing the calculation of the moving average algorithm, the calculation is performed based on “at least one input data exists within a predetermined distance of the point at which the pseudo data is output”. I did it.
As a result, when generating pseudo data in an area where no data exists, it is possible to prevent a large amount of the same data from being generated by a single calculation of the moving average algorithm, and the quality of the data to fluctuate smoothly. A good three-dimensional image can be generated.

The invention according to claim 4 is to set a condition that “at least one input data exists within one lattice of a point where pseudo data is output” when performing a moving average algorithm calculation. did.
Thereby, the pseudo data can be generated one grid at a time by the calculation of the moving average algorithm, and the same pseudo data can be generated by a single calculation of the moving average algorithm in a plurality of grids. It is possible to prevent a deterioration in the quality of the three-dimensional image.

Hereinafter, a three-dimensional image generation method according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 2 is a block diagram showing a configuration of the three-dimensional image generation apparatus 10 according to the embodiment of the present invention. The three-dimensional image generation apparatus 10 includes a data input unit 1, a binning unit 2, a structure drawing creation unit 3, a gradient reading unit 4, a hybrid structure gradient model construction unit 5, a wave field interpolation unit 6, and a migration unit 7.
The data input means 1 receives a plurality of data of underground sectional views obtained by elastic wave exploration. The data of the underground sectional view input to the data input means 1 is output to the binning means 2.

  The data of the underground sectional view input to the binning means 2 is output to the structural drawing creating means 3 and the inclination reading means 4 after the “binning” process. Here, “binning” refers to a process of arranging data of a plurality of underground sectional views in an area on a three-dimensional space in association with the point where each underground sectional view is measured.

Based on the binned data output from the binning unit 2, the structure diagram creating unit 3 creates a structural diagram representing the structure of the underground geology. The tilt reading means 4 performs “tilt reading” processing based on the binned data output from the binning means 2. Here, “inclination reading” is a process of detecting similar points in a plurality of underground sectional views, connecting the similar points with straight lines, and obtaining the inclination of the straight lines.
The hybrid structure inclination model construction means 5 constructs a hybrid structure inclination model by synthesizing the structure diagram data created by the structure drawing creation means 3 and the slope data obtained by the slope reading means 4. .

The wave field interpolation means 6 performs the process of “wave field interpolation” based on the data of the underground sectional view arranged in the three-dimensional space by the binning means 2 and the model constructed by the hybrid structure tilt model construction means 5. Do. Here, “wave field interpolation” means that the pseudo-data is stored in the three-dimensional space by using the moving average algorithm for the data of the underground cross-sectional view or the area where the pseudo-data generated from the underground cross-sectional view does not exist. It is a process to generate.
The migration unit 7 performs a “migration” process on the data subjected to the wave field interpolation by the wave field interpolation unit 6. Here, “migration” refers to a process of converting the data of the underground sectional view arranged in the three-dimensional space and the pseudo data generated from the data of the underground sectional view into a three-dimensional image.

Next, the process flow of the above-described three-dimensional image generation apparatus 10 will be described with reference to the flowchart shown in FIG.
In the three-dimensional image generation method of the present embodiment, first, two-dimensional or three-dimensional data related to an underground cross-sectional view measured by performing an elastic wave exploration of an underground geological structure is input to the data input unit 1 ( Step S11).
Next, the binning means 2 performs a binning process in which the two-dimensional or three-dimensional data input in step S11 is arranged in an area on the three-dimensional space (step S12).
By performing binning processing in step S12, the underground sectional view of the xt plane shape and the underground sectional view on the yt plane as shown in FIGS. The data arranged in is obtained.

Next, the structural diagram creating means 3 creates a structural diagram by interpreting the structure related to the inclined surface of the underground structure based on the data arranged on the region of the three-dimensional space by binning in step S12 (step S13).
In parallel with the processing in step S13, the tilt reading means 4 performs tilt reading processing based on the data arranged in the region in the three-dimensional space by the binning process in step S12 (step S14).
Thereafter, the hybrid structure inclination model construction means 5 constructs a “hybrid structure inclination model” by synthesizing the structure diagram data created by the structure interpretation in step S13 and the inclination data read in step S14 ( Step S15).

In the three-dimensional image generation method according to the present embodiment, the “hybrid structure tilt” is more accurate than the conventional model by combining the structure diagram data obtained in step S13 and the tilt data obtained in step S14. Model "can be constructed.
A more detailed construction method of the “hybrid structure tilt model” will be described later with reference to FIG.

After that, the wave field interpolation means 6 uses the data arranged in the three-dimensional space by binning in step S12 and the hybrid structure gradient model created in step S15 to the region where no data exists in the three-dimensional space. Then, wave field interpolation is performed by generating pseudo data (step S16). A more detailed processing method of this wave field interpolation will be described later with reference to FIG.
Finally, the migration unit 7 performs migration by converting the data subjected to wave field interpolation in step S16 into a three-dimensional image (step S17).

Next, the process for constructing the “hybrid structure gradient model” in step S15 of FIG. 1 described above will be described in more detail.
FIG. 4 is a flowchart showing the flow of construction of the “hybrid structure gradient model”.
First, two-dimensional or three-dimensional elastic wave exploration data is input (step S21). The process of step S21 corresponds to the process of step S11 in FIG.
Next, binning processing is performed by regularly arranging the two-dimensional or three-dimensional elastic wave exploration data input in step S21 in a three-dimensional space (step S22). The processing in step S22 corresponds to the processing in step S12 in FIG.

Next, the inclination (dt / dx, dt / dy) in the seismic survey sectional view is calculated based on the data binned in step S22 (step S23). The inclination is calculated by detecting similar points between a plurality of binned seismic survey sectional views and obtaining an inclination of a straight line connecting these points.
Next, a structural inclination model is created using the inclination data obtained in step S23 (step S24). Note that the processing in steps S23 and S24 described above corresponds to the processing in step S13 in FIG.

On the other hand, for the data arranged in the three-dimensional space by binning in step S22, a structure diagram is created by interpreting the structure of the underground inclined surface (step S25).
Then, a “normal vector” is obtained from the structure diagram created in step S25, and “slope” is obtained from the “normal vector” (step S26).
Next, a “normal vector model” is created from the “tilt” obtained in step S26 (step S27). Note that the processing in steps S25 to S27 described above corresponds to the processing in step S14 in FIG.

FIG. 5 illustrates the “normal vector model” created in step S27 of FIG. The “normal vector” is shown in a beard shape on the lower surface side of the inclined surface extending in the xy plane direction. This “normal vector” is composed of three components: an x component, a y component, and a t component.
Data obtained by separating the data of the “normal vector” in the “normal vector model” shown in FIG. 5 for each of the x component, the y component, and the t component are shown in FIGS. 6 (a) and 6 (b). FIG. 6 (c) shows the respective drawings.

Next, the “structural gradient model” created in step S24 and the “normal vector model” created in step S27 are combined (step S28).
FIG. 7A illustrates normal vector data on the xt plane obtained by the “normal vector model”. FIG. 7B illustrates data on the xt plane obtained by the “structural gradient model”. The data shown in FIG. 7C is obtained by synthesizing the data of FIG. 7A and FIG. 7B. The data of FIG. 7C is obtained by the “normal vector model” of FIG. 7A, which was not present in the data of FIG. Therefore, it is possible to construct a model with higher accuracy than a conventionally used model.
As described above, a “hybrid structure gradient model” is constructed by combining the “normal vector model” and the “structure gradient model” (step S29 in FIG. 4).
Note that the processing in steps S28 to S29 described above corresponds to the processing in step S15 in FIG.

  Conventionally, the data of “structural gradient model” and “normal vector model” cannot be handled at the same time and have been used separately. However, if the construction method of the “hybrid structure gradient model” shown in FIG. 4 is used, as shown in FIGS. 6A to 6C, the data of the “normal vector” is converted into the x component and the y component. , T component, and using the x component, y component, and t component to calculate the tilt (dt / dx, dt / dy), the tilt (dt / dx, dt) calculated by the structural tilt model / Dy), the “structural gradient model” and the “normal vector model” can be handled simultaneously. This makes it possible to construct a “hybrid structure tilt model” that is more accurate than the conventional “structure tilt model”.

Next, the “wave field interpolation” in step S16 of FIG. 1 described above will be described in more detail with reference to the flowchart showing the flow of the “wave field interpolation” process shown in FIG.
First, two-dimensional or three-dimensional elastic wave exploration data is input (step S31). Note that the processing in step S31 corresponds to the processing in step S11 in FIG.
Next, binning processing is performed by arranging the two-dimensional or three-dimensional elastic wave exploration data input in step S31 in a three-dimensional space (step S32). Note that the processing in step S32 corresponds to the processing in step S12 in FIG.

  Next, pseudo data is generated for an output point that satisfies the following “condition 1” and “condition 2” for an area where no data exists in the three-dimensional space. (Step S33). As “Condition 1”, a condition “use input data of a specified number or more when calculating pseudo data” is set. Here, as the input data, the elastic wave exploration data input in step S31 or the pseudo data generated in step S34 is used.

In addition, as “condition 2”, a condition “at least one input data exists within a predetermined distance from the point where pseudo data is output” is set.
Note that here, a case has been described in which pseudo data is generated by setting both “condition 1” and “condition 2” in step S33, but either “condition 1” or “condition 2” is described. Only one condition may be set to generate pseudo data.
Next, if there is a region that satisfies the above-described “condition 1” and “condition 2” in step S33, pseudo data is created for the output point based on the following formula (1). (Step S34).

In the above equation (1), weighting is performed based on the following equation (2).

  Here, in the formula (2), c1 and c2 are “inclination” data defined by the following formula (3).

Returning to FIG. 8, it is determined whether or not there is no area in which no pseudo data exists due to the generation of the pseudo data in step S34 (step S35).
If there is an area in which no pseudo data is generated among the areas where no data exists, “NO” is determined in step S35, and the process proceeds to step S33 again. Thereafter, the processes in steps S33 to S35 are repeated until the pseudo data is output to all areas where no data exists.

If the pseudo data is generated in all the areas where no data exists, “YES” is determined in the step S35, and the process proceeds to the step S36.
In step S36, the three-dimensional image data filled with the pseudo data in step 35 is output. Note that the processing in steps S33 to S36 described above corresponds to the processing in step S16 in FIG.

Next, the calculation process when calculating the pseudo data in step S33 of FIG. 8 described above will be described in more detail.
Here, for the sake of simplicity of explanation, a case will be described in which pseudo data is generated in a total of 104 lattice areas of 8 lattices in the X direction and 13 lattices in the Y direction as shown in FIG. Further, five data (X, Y) = A (3, 3), B (5, 3), C (4, 4) as initial data corresponding to the elastic wave exploration data in the region shown in FIG. , D (3, 5), E (9, 6) will be described.

First, with respect to the case where pseudo data is generated for a lattice in which no data exists in FIG. 9 by performing an operation using the moving average algorithm without setting the above-mentioned “condition 1” and “condition 2”, FIG. A description will be given with reference to FIG.
In this moving average algorithm, since “condition 1” is not set, processing for simply copying data existing in any grid may be performed when calculating pseudo data. In addition, since “condition 2” is not set, a process of generating the same pseudo data on a plurality of grids may be performed by a single moving average algorithm. A case where wave field interpolation is performed using this moving average algorithm will be described below.

When the initial data given in FIG. 9 is calculated once using the moving average algorithm, the “first generation” data shown in FIG. 10A is generated.
The data “a”, “b”, “c”, “d”, and “e” in FIG. 10A are the data “A”, “B”, “C”, “D”, and “E”, respectively. "Is just a copy of". On the other hand, the data indicated by ☆ indicates that the data is generated based on data of two or more different types of lattices.

When the second calculation is performed based on the data shown in FIG. 10A, the “second generation” data shown in FIG. 10B is generated. Thereafter, when the calculation of the moving average algorithm is repeated in the same manner, “third generation” (FIG. 10C), “fourth generation” (FIG. 10D), “fifth generation” (FIG. 11A )), The pseudo data is gradually generated in the lattice where data is missing.
If the moving average algorithm used conventionally is used, as shown in FIG. 11A, in the “fifth generation” in which the calculation is repeated five times, the area of 8 lattices × 13 lattices is all filled with pseudo data.

  However, in the “fifth generation” data, as shown in FIG. 11A, the area around the original data “A” is occupied by “a” which is a copy of the data “A”. Has been. Similarly, data obtained by simply copying the data “B”, “C”, “D”, and “E” in the areas around the initial data “B”, “C”, “D”, and “E”. Are occupied by “b”, “c”, “d”, and “e”, respectively.

  Further, the boundary of the area occupied by the data “a”, “b”, “c”, “d”, “e” is pseudo data generated based on two or more input data. The data marked with ☆ is present. However, the horizontal pattern (X, Y) = (1,4), (2,4), (3,4), the vertical pattern (4,1), (4,2), (4,3), Regular patterns such as diagonal direction patterns (4, 6), (5, 5), (6, 4), (7, 3), (8, 2), (9, 1) are generated. Therefore, it cannot be said that high-quality data is generated.

A large amount of copies “a”, “b”, “c”, “d”, “e” of the above-described initial data “A”, “B”, “C”, “D”, “E” are generated. In order to solve the problem, in the present embodiment, when performing the operation of the moving average algorithm, the condition that “use the specified number or more of input data when calculating the pseudo data” is used as “condition 1”. It was decided to set.
In the following description, the case where the condition “use two or more input data when calculating pseudo data” is set as “condition 1” will be described more specifically.

The case of generating pseudo data for the initial data shown in FIG. 9 will be described with reference to FIGS. 12 (a) to 14 (c).
When the calculation of the moving average algorithm is performed once on the initial data of FIG. 9, the “first generation” data shown in FIG. 12A is generated.
The “first generation” data shown in FIG. 12A can be understood by comparing with the “first generation” data created using the conventional moving average algorithm described in FIG. In addition, there is no “a”, “b”, “c”, “d”, “e” which is a simple copy of the initial data “A”, “B”, “C”, “D”, “E”. .

  Based on the data of the “first generation” shown in FIG. 12A, by repeating the operation of the moving average algorithm that generates pseudo data for the lattice in which no data exists, the “second generation” (FIG. 12 (B)), “3rd generation” (FIG. 12C), “4th generation” (FIG. 12D), “5th generation” (FIG. 13A), “6th generation” ( 13 (b)), “7th generation” (FIG. 13 (c)), “8th generation” (FIG. 13 (d)), “9th generation” (FIG. 14 (a)), “10th generation” ”(FIG. 14B) and“ 11th generation ”(FIG. 14C), pseudo data is gradually generated for an area in which no data exists in an 8 × 13 grid area. To go.

In the “11th generation” (FIG. 14C), a total of 108 lattices of 8 lattices × 13 lattices are completely filled with initial data and pseudo data. As can be seen from comparison of this “11th generation” data (FIG. 14C) with data (FIG. 11A) in which pseudo data is generated using a conventional moving average algorithm, the initial data All grids other than “A”, “B”, “C”, “D”, and “E” are all filled with the data marked with ☆, so that high-quality data with smooth changes can be generated. Show.
However, in the data shown in FIG. 14C, “a”, “b”, “c”, which is a simple copy of the initial data “A”, “B”, “C”, “D”, “E”, Although “d” and “e” do not exist, when the pseudo data is generated by the moving average algorithm, a plurality of the same data may be created.

For example, as shown in FIG. 15, when performing interpolation and extrapolation of pseudo data using the moving average algorithm from the initial data “A1” and “B1” to an area where no data exists, Since the condition “use two or more input data when calculating pseudo data” is set, the data “d1” generated by interpolation and the data “d2” generated by extrapolation are: The initial data “A1” and “B1” are not simply copied. However, since the pseudo data is generated based on the data “B1” and “d2” as the data “d3” and “d4” by one moving average algorithm, the same pseudo data is used for the data “d3” and “d4”. Data is generated.
In the moving average algorithm of FIG. 15, the case where the distance between the pseudo data to be generated and the data used to generate the pseudo data is allowed to be adjacent to three is described.

In the three-dimensional image generation method according to the present embodiment, in addition to “condition 1”, a condition that “at least one input data exists within a predetermined distance from a point where pseudo data is output” is further added as “condition 2”. By adding, it is possible to prevent the same data from being generated by a single calculation of the moving average algorithm, and to generate a more accurate three-dimensional image.
In the following, as more specific conditions, “condition 1” “use two or more input data when calculating pseudo data”, and “at least one input within one grid for outputting pseudo data” A description will be given of the result when the “average 2” “condition 2” is set and the moving average algorithm is calculated.

The case of generating pseudo data for the initial data shown in FIG. 9 will be described with reference to FIGS. 16 (a) to 18 (d).
When the calculation of the moving average algorithm is performed once on the initial data of FIG. 9, the “first generation” data shown in FIG. 16A is obtained.
Based on the data of the “first generation” shown in FIG. 16A, by repeating the calculation of the moving average algorithm that generates pseudo data for the lattice in which no data exists, the “second generation” (FIG. 16) is repeated. (B)), “third generation” (FIG. 16C), “fourth generation” (FIG. 16D), “fifth generation” (FIG. 17A), “sixth generation” ( FIG. 17B), “7th generation” (FIG. 17C), “8th generation” (FIG. 17D), “9th generation” (FIG. 18A), “10th generation” ”(FIG. 18B),“ 11th generation ”(FIG. 18C),“ 12th generation ”(FIG. 18D), the data in the region of 8 lattices × 13 lattices The pseudo data is gradually generated for the non-existing area.

In the “12th generation” (FIG. 18D), the area of 8 lattices × 13 lattices is completely filled with initial data and pseudo data. In the “12th generation” data (FIG. 18D), all the grids other than the initial data “A”, “B”, “C”, “D”, and “E” are filled with the data marked with “☆”. This indicates that the data is changing smoothly.
Further, only when “Condition 1” is set, when the calculation by the moving average algorithm is performed once from the initial data shown in FIG. 9, for example, (X, Y) apart from the initial data “A” by one grid. = (4,3) and (X, Y) = (4,2) apart from the initial data “A” by 2 lattices ((X, Y) = (4,3), (4,2)) ) Pseudo data was generated at once for the grid.
However, by adding “Condition 2”, as shown in FIG. 16A, when the calculation by the moving average algorithm is performed once, (X, Y) = (1) apart from the initial data “A”. 4, 3) pseudo data is generated only in the lattice, and (X, Y) = (2) away from the initial data “A” by two lattices ((X, Y) = (4, 3), (4, 2)). No pseudo data is generated for the grids 4 and 2).

  For example, as shown in FIG. 19, when performing interpolation and extrapolation of pseudo data from the initial data “A2” and “B2” using a moving average algorithm to an area where no data exists, Since the condition “use two or more input data when calculating the pseudo data” is set, the data “d5” generated by the interpolation and the data “d6”, “ “d7” and “d8” are not simply copies of the initial data “A2” and “B2”.

In addition, since the condition “at least one input data exists within one grid for outputting pseudo data” is set as “condition 2”, the data “d6” is calculated for each operation of the moving average algorithm. ”,“ D7 ”, and“ d8 ”, pseudo data is generated for each grid.
In the moving average algorithm of FIG. 19, a case has been described in which the distance between the pseudo data to be generated and the data used to generate the pseudo data is allowed up to three neighbors.
By generating “Condition 1” and “Condition 2” as described above and calculating the moving average algorithm, pseudo data that changes more smoothly by one grid can be generated for an area where no data exists. Therefore, the quality of the three-dimensional image to be created can be improved.

Next, the effect when using the three-dimensional image generation method according to the embodiment of the present invention will be described.
FIG.20 (b) is data of sectional drawing (xt plane) of underground geological structure. Here, in order to confirm the effect of the three-dimensional image generation method according to the present embodiment, data obtained by thinning out half of the data in FIG. 20B is shown in FIG. By applying the three-dimensional image generation method according to the present embodiment to the cross-sectional data (FIG. 20A) from which half of this data is thinned, pseudo data is generated for the area where the data is missing. FIG. 20C shows the result of generating and interpolating.

As can be seen by comparing FIG. 20B and FIG. 20C, the three-dimensional data according to the present embodiment is used for the cross-sectional data (FIG. 20A) in which half of the data is thinned out. The data of the cross-sectional view restored by applying the image generation method (FIG. 20C) and the data of the cross-sectional view before the half of the data is thinned (FIG. 20B) match very well. The image can be restored with high accuracy from the image lacking data.
In FIGS. 20A to 20C, only the data of the cross-sectional view at the predetermined point is shown, but actually, the cross-sectional view data also exist in the front and back directions with respect to the paper surface. By using this data, a three-dimensional image can be generated.

On the other hand, FIG.21 (b) is the data of the horizontal surface (xy plane) of an underground geological structure. Here, in order to confirm the effect of the three-dimensional image generation method according to the present embodiment, FIG. 21A shows a result obtained by thinning out half of the data in FIG. By applying the three-dimensional image generation method according to the present embodiment to the horizontal plane data (FIG. 21 (a)) from which half of this data is thinned, pseudo data is generated for the area where the data is missing. The result of generating and interpolating is shown in FIG.
As can be seen by comparing FIG. 21B and FIG. 21C, the three-dimensional image according to the present embodiment is applied to the horizontal plane data (FIG. 21A) from which half of the data is thinned. The horizontal plane data restored by applying the generation method (FIG. 21 (c)) and the horizontal plane data before half of the data is thinned out (FIG. 21 (b)) are in very good agreement. The image can be restored with high accuracy from the missing image.

  According to the above-described three-dimensional image generation method according to the embodiment of the present invention, it is possible to perform more accurate three-dimensional imaging on three-dimensional data having high irregularity and lacking. In addition, since the two-dimensional survey includes a plurality of two-dimensional survey data and is regarded as one irregular three-dimensional data, the three-dimensional image generation method according to the present embodiment can be used. The underground structure can be reproduced with higher accuracy than conventional two-dimensional images.

  In the embodiment of the present invention described above, when “wave field interpolation” in step S33 of FIG. 8 is performed, “condition 1” uses “two or more input data when calculating pseudo data”. However, the present invention is not limited to this. It is also possible to make the condition “use three or more input data when calculating pseudo data”. The quality of the three-dimensional image generated by the three-dimensional image generation method can be improved by using more data when calculating the pseudo data.

Further, in the above-described embodiment of the present invention, when “wave field interpolation” in step S33 of FIG. 8 is performed, “condition 2” indicates that “at least one input data exists within one grid for outputting pseudo data”. However, the present invention is not limited to this. It is also possible to make a condition that “at least one input data exists within n lattices (n is an integer of 2 or more) for outputting pseudo data”. By increasing n, the quality of the generated pseudo data is degraded, but the calculation speed of the moving average algorithm is improved.
In addition, by setting the above-mentioned n to √2, etc., not only the distance from the top, bottom, left and right grids of the grid for generating the pseudo data, but also the distance from the grid in the oblique direction may be considered.

  In the embodiment described above, steps S11 to S17 shown in FIG. 1, steps S21 to S29 shown in FIG. 4, and steps S31 to S36 shown in FIG. 8 are realized by dedicated hardware. The steps S11 to S17 shown in FIG. 1, the steps S21 to S29 shown in FIG. 4, and the steps S31 to S36 shown in FIG. 8 are constituted by a memory and a CPU (central processing unit). The function may be realized by loading a program for realizing these functions into a memory and executing the program.

  Also, a program for realizing the functions of steps S11 to S17 shown in FIG. 1, steps S21 to S29 shown in FIG. 4, and steps S31 to S36 shown in FIG. 8 is recorded on a computer-readable recording medium. The program recorded on the recording medium may be read by a computer system and executed to generate a three-dimensional image. Here, the “computer system” includes an OS and hardware such as peripheral devices.

Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

  The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes design and the like within a scope not departing from the gist of the present invention.

  In order to grasp the underground geological structure in an area where it is difficult to measure a three-dimensional image, the three-dimensional image generation method and program according to the present invention can be used.

It is a flowchart which shows the flow of a process of the three-dimensional image generation method by embodiment of this invention. It is a block diagram which shows the structure of the three-dimensional image generation apparatus 10 by this embodiment. It is the figure which showed the data obtained by the binning by this embodiment. It is a flowchart which shows the flow of preparation of the "hybrid structure inclination model" by this embodiment. It is a figure which shows the structure of the "normal vector model" by this embodiment. This is data illustrating the x component, y component, and t component of the “normal vector”. It is a figure which shows the process of adding the data of a normal vector to a "structural gradient model". It is a flowchart which shows the flow of the process of the "wave field interpolation" by this embodiment. This is data indicating a region where “wave field interpolation” is performed and initial data. This is data indicating a result when “wave field interpolation” is performed without setting “condition 1” and “condition 2”. This is data indicating a result when “wave field interpolation” is performed without setting “condition 1” and “condition 2”. This is data indicating a result when “condition 1” is set and “wave field interpolation” is performed. This is data indicating a result when “condition 1” is set and “wave field interpolation” is performed. This is data indicating a result when “condition 1” is set and “wave field interpolation” is performed. It is a figure which shows the moving average algorithm in the case of setting only "condition 1". This is data indicating a result when “condition 1” and “condition 2” are set and “wave field interpolation” is performed. This is data indicating a result when “condition 1” and “condition 2” are set and “wave field interpolation” is performed. This is data indicating a result when “condition 1” and “condition 2” are set and “wave field interpolation” is performed. It is a figure which shows the moving average algorithm in the case of setting "Condition 1" and "Condition 2". It is a figure which shows the effect at the time of using the three-dimensional image generation method by this embodiment. It is a figure which shows the effect at the time of using the three-dimensional image generation method by this embodiment. It is a flowchart which shows the flow of a process of the three-dimensional image generation method used conventionally. It is a figure which shows the structure inclination model conventionally used when performing three-dimensional imaging.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Data input means 2 ... Binning means 3 ... Structure drawing preparation means 4 Inclination reading means 5 ... Hybrid structure inclination model construction means 6 ... Wave field interpolation means 7 ... Migration means 10 ... 3D image generation devices S01 to S04 ... Processes S11 to S17 performed when generating a conventionally used 3D image S21 to S29 performed when generating a 3D image Processing performed when creating a structural gradient model S31 to S36: Processing performed during wave field insertion processing

Claims (8)

A first step of inputting 2D or 3D elastic wave exploration data;
A second step of performing binning by arranging the two-dimensional or three-dimensional elastic wave exploration data input in the first step in a three-dimensional space;
A third step of creating a structural diagram by performing structural interpretation based on the data binned in the second step;
A fourth step of performing tilt reading based on the data binned in the second step;
A fifth step of constructing a hybrid structure tilt model based on the structure diagram created in the third step and the tilt data read in the fourth step;
Based on the data binned in the second step and the hybrid structure gradient model created in the fifth step, pseudo data is generated using a moving average algorithm for the area where data is missing A sixth step of performing wave field interpolation by
A seventh step of performing migration by converting the data obtained by performing wave field interpolation in the sixth step into a three-dimensional image;
A three-dimensional image generation method characterized by comprising:   The calculation of the moving average algorithm when performing wave field interpolation in the sixth step is set on the condition that at least the specified number of input data is used when calculating pseudo data. The three-dimensional image generation method according to 1.   In the calculation of the moving average algorithm at the time of wave field interpolation in the sixth step, it is set on condition that at least one input data exists within a predetermined distance from the point where pseudo data is output. The three-dimensional image generation method according to claim 2.   The three-dimensional image generation method according to claim 3, wherein the moving average algorithm is calculated by setting the predetermined distance to one grid. A first step of inputting 2D or 3D elastic wave exploration data;
A second step of performing binning by arranging the two-dimensional or three-dimensional elastic wave exploration data input in the first step in a three-dimensional space;
A third step of creating a structural diagram by performing structural interpretation based on the data binned in the second step;
A fourth step of performing tilt reading based on the data binned in the second step;
A fifth step of constructing a hybrid structure tilt model based on the structure diagram created in the third step and the tilt data read in the fourth step;
Based on the data binned in the second step and the hybrid structure gradient model created in the fifth step, pseudo data is generated using a moving average algorithm for the area where data is missing A sixth step of performing wave field interpolation by
A seventh step of performing migration by converting the data obtained by performing wave field interpolation in the sixth step into a three-dimensional image;
A three-dimensional image generation program for causing a computer to execute.   The calculation of the moving average algorithm when performing wave field interpolation in the sixth step is set on the condition that at least the specified number of input data is used when calculating pseudo data. 5. The three-dimensional image generation program according to 5.   In the calculation of the moving average algorithm at the time of wave field interpolation in the sixth step, it is set on condition that at least one input data exists within a predetermined distance from the point where pseudo data is output. The three-dimensional image generation program according to claim 6. The three-dimensional image generation program according to claim 7, wherein the moving average algorithm is calculated by setting the predetermined distance to one grid.

Images