Core model generation method

Abstract

The purpose of this project is to provide a core model generation method that uses three-dimensional model data of the front and back of a core storage box containing a columnar core, efficiently joins and shapes the data, removes unnecessary parts, and then constructs a cylindrical three-dimensional model along the depth direction. [Solution] A method for generating a core model using three-dimensional models of the front and back surfaces obtained for a storage box containing a columnar core, comprising: a first step of removing parts other than the columnar core from the three-dimensional models of the front and back surfaces based on a specified area; a second step of aligning and joining the three-dimensional models of the front and back surfaces from which parts other than the columnar core have been removed in the first step based on a reference plane; and a third step of joining the three-dimensional models of multiple columnar cores aligned and joined in the second step along the depth direction.

Description

[Technical Field]

[0001] The present invention relates to a data processing technology for columnar cores collected in geological surveys and soil tests, and in particular to a method for generating a three-dimensional model of a columnar core stored in a core storage box using three-dimensional model data obtained from the core. [Background technology]

[0002] Traditionally, cores collected by drilling surveys have been stored in core storage boxes, and observations of the cores have been recorded as logarithms and photographs. In recent years, systems have been proposed that digitize core images and manage them on a computer. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2004-125672 Summary of the Invention [Problem to be solved by the invention]

[0004] However, conventional techniques have focused on recording and managing data based on two-dimensional images, making it difficult to accurately reproduce the three-dimensional shape and continuity of cores. Even when using three-dimensional scanners and photogrammetry, there are issues such as long processing times, the inclusion of unnecessary parts such as core containers in the data, and the difficulty of joining multiple cores along the depth direction to reconstruct a cylindrical shape.

[0005] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a core model generation method that uses three-dimensional model data of the front and back sides obtained from a core storage box containing a columnar core, efficiently joins and shapes the data, removes unnecessary parts, and then constructs a cylindrical three-dimensional model along the depth direction. [Means for solving the problem]

[0006] The core model generation method of the present invention is a method for generating a core model using three-dimensional models of the front and back surfaces obtained for a storage box containing a columnar core, and is characterized by comprising: a first step of removing parts other than the columnar core from the three-dimensional models of the front and back surfaces based on a specified area; a second step of aligning and joining the three-dimensional models of the front and back surfaces from which parts other than the columnar core have been removed in the first step based on a reference plane; and a third step of joining the three-dimensional models of the columnar cores aligned and joined in the second step along the depth direction.

[0007] In the core model generating method according to the present invention, it is preferable that the removal of the portion other than the columnar core is performed based on a rectangular area specification or a three-dimensional range specification.

[0008] In the core model generating method according to the present invention, it is preferable that the joining along the depth direction is performed by automatically detecting the end face positions of each columnar core. [Effects of the Invention]

[0009] According to the present invention, it is possible to efficiently join and shape the front and back three-dimensional model data obtained for a core container containing a columnar core, remove unnecessary parts, and then automatically construct a cylindrical three-dimensional model along the depth direction.

[0010] As a result, highly accurate three-dimensional models can be obtained in a short time compared to core recording, which was previously limited to two-dimensional images, and three-dimensionalization processing, which required a long processing time. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a perspective view (closed state) from the front side of a dedicated storage box 1 for columnar cores used for observing and photographing columnar cores. [Figure 2]FIG. 2 is a perspective view (closed state) of the housing box 1 for columnar cores as seen from the rear side. [Figure 3] 1 is a perspective view (open state) of a storage box 1 for columnar cores, seen from the front side. [Figure 4] FIG. 1 is a perspective view of a storage box 1 for columnar cores (separated state: with cushioning material). [Figure 5] FIG. 1 is a perspective view of a storage box 1 for columnar cores (separated state: without cushioning material). [Figure 6] FIG. 2 is a perspective view of a cushioning material 2A of a dedicated storage box 1 for columnar cores. [Figure 7] FIG. 2 is a plan view of the cushioning material 2A. [Figure 8] 8 is a cross-sectional view of the cushioning material 2A shown in FIG. 7 taken along the line XX. [Figure 9] FIG. 9 is a partially enlarged view of the cushioning material 2A shown in FIG. [Figure 10] 1 is a perspective view showing a state of columnar cores W11 to W13 housed in a box body 3B of a housing box 1. FIG. [Figure 11] 1 is a perspective view showing a state in which the storage box 1 is separated into box bodies 3A and 3B. [Figure 12] FIG. 10 is a perspective view showing a state of columnar cores W11 to W13 housed in a box body 3B. [Figure 13] FIG. 2 is a perspective view showing a state in which the boxes 3A and 3B are joined together. [Figure 14] FIG. 2 is a perspective view showing a state in which the storage box 1 is closed. [Figure 15] FIG. 2 is a perspective view showing a state in which the storage box 1 is turned upside down. [Figure 16] 1 is a perspective view showing a state of columnar cores W11 to W13 housed in a box body 3A of a housing box 1. FIG. [Figure 17] 10A to 10C are diagrams showing a process of photographing the surface side of a container box and generating a three-dimensional model in the core model generating method of the present invention. [Figure 18] 10A and 10B are diagrams showing a process of turning over the container, photographing the back side, and generating a three-dimensional model of the back side. [Figure 19] FIG. 10 is a diagram showing a process of loading the generated three-dimensional models of the front and back into analysis software. [Figure 20] 10A to 10C are diagrams illustrating a process of removing unnecessary parts from a three-dimensional model and extracting only the columnar core. [Figure 21] 10A and 10B are diagrams illustrating a process of inverting the back model and matching the coordinate system with the front model. [Figure 22] 10A to 10C are diagrams showing a process of aligning and combining the front and back three-dimensional models based on a reference plane. [Figure 23] FIG. 10 is a diagram showing the process of joining multiple columnar cores along the depth direction to construct a single cylindrical model. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following, similar elements in all drawings will be designated by the same reference numerals, and duplicate explanations will be omitted. Furthermore, in the description below, previously described reference numerals will be used as necessary.

[0013] A core model generation method according to one embodiment of the present invention is performed on a columnar core stored in a storage box 1 specifically designed for columnar cores, as shown in Figures 1 to 5. The storage box 1 specifically designed for columnar cores is composed of a set of cushioning materials 2 and a set of rectangular parallelepiped boxes 3 (3A, 3B) that store the cushioning materials 2. This allows the columnar core to be stably held without rolling, and also makes it easy to observe and photograph the columnar core.

[0014] As shown in Figures 1 to 3, the storage box 1 can open and close the box bodies 3A and 3B by opening the fastener K located on the front side. Furthermore, as shown in Figures 4 and 5, the box bodies 3A and 3B can be separated by separating the removable hinge 4 located on the back side. This allows for flexible operation according to the work situation on site, such as opening one side to observe the core or separating both sides to secure work space.

[0015] As shown in Figure 4, cushioning materials 2A and 2B are placed inside each box 3A and 3B, and these are designed to reliably support the columnar cores without them rolling. The box 3 itself is made of a lightweight yet sturdy material, making it easy to transport and install, and also able to hold the columnar cores for a long period of time without damage, even in a humid environment.

[0016] As shown in Figures 6 to 9, the buffer material 2 is formed with a groove 22 that follows the outer shape of the columnar core, allowing the cylindrical columnar core to be accommodated along its semicircular cross section. Furthermore, the cross-sectional shape of the groove is not limited to circular, but can be adjusted to rectangular, polygonal, or other shapes depending on the shape of the core. The groove also has steps or notches, which ensure space for observation around the columnar core, making it easier to fit the side of the core into the field of view when observing visually or photographing with a photographing device.

[0017] 1 and 2, the box 3 has air holes 31 to 36 on its outer periphery, which allow moisture and humidity generated inside the core to escape to the outside. This prevents excessive humidity buildup during long-term storage of the housed columnar cores, thereby suppressing core deterioration.

[0018] As shown in Figures 10 to 16, when observing or photographing the columnar cores W11 to W13, first open one side of the storage box 1 to expose half the circumference of the core and photograph it. Then, turn the storage box 1 over to open the other side and photograph the remaining half. By using the storage box in this way, it is possible to observe and photograph the entire circumference of the columnar cores W11 to W13 without removing them, greatly reducing the risk of damaging the cores.

[0019] Furthermore, by appropriately positioning multiple cameras and lighting devices during photography, it is possible to faithfully record the surface conditions of the columnar cores W11 to W13. In particular, by utilizing the steps and void structure of the buffer material 2, blind spots around the core are reduced, allowing for efficient photography of the entire image. This allows for highly accurate three-dimensional data of the front and back surfaces to be acquired, enabling the subsequent model generation process to proceed smoothly and efficiently.

[0020] Next, the processing steps of a core model generation method according to one embodiment of the present invention will be described with reference to Figures 17 to 23. A feature of the present invention is that both sides of a core are photographed while it is held in a storage box 1, and this data is integrated to construct a single cylindrical three-dimensional model.

[0021] First, as shown in Figure 17, the front side of the storage box 1 containing the columnar cores W11 to W13 is photographed, and a three-dimensional model of the front side is generated based on the photographed data. At this time, a position detection marker 10 is provided at a predetermined position, such as the center of the short side of the storage box 1, making it easy to identify the photographing direction and coordinate system. This position detection marker 10 is one of the important features of the present invention, and functions as a reference point when integrating multiple photographed data into the same coordinate system. The position detection marker is a dedicated jig placed at the center of the short side of the storage box 1, and makes it possible to establish reference coordinates for repeated photographing.

[0022] Next, as shown in Fig. 18, the container 1 is turned over, and the back side is photographed to generate a three-dimensional model of the back side. This makes it possible to obtain three-dimensional data that covers the front and back of the entire columnar cores W11 to W13, enabling full-circumference observation, which was previously difficult.

[0023] Next, as shown in Figure 19, the generated 3D models of the front and back sides are loaded into analysis software and incorporated into the same processing environment. This process allows both to be handled in a unified manner, enabling the subsequent alignment process to be performed with high accuracy. The reference plane of the loaded model is determined by the position detection marker 10, and it is converted into a normalized reference coordinate system even if the container 1 is tilted.

[0024] Next, as shown in Figure 20, parts other than the columnar cores W11-W13 are removed from the loaded 3D model, i.e., unnecessary areas such as the container box 1 and cushioning material 2. This removal can be performed by specifying a rectangular area or a three-dimensional range, and 3D data that extracts only the pure columnar cores W11-W13 is obtained. Removal of unnecessary parts is achieved by specifying a rectangular area and trimming it three-dimensionally, which allows the texture of the core box and cushioning material to be removed. This process can be performed intuitively in the software using user interface operations such as rectangular area specification.

[0025] Furthermore, as shown in Figure 21, the 3D model of the back side is flipped over to match the coordinate system with the 3D model of the front side. At this time, by aligning the two models based on the central axes of the cores, the data for the front and back sides can be overlaid in the correct positional relationship. When aligning the data, a process is performed to set the central axis.

[0026] Then, as shown in Figure 22, the three-dimensional models of the front and back sides are aligned based on the reference plane, and the two are combined. The reference plane is set based on the horizontal plane defined by the position detection marker 10. If necessary, the user can perform fine adjustments by additionally specifying the edge of the container 1 or any point on the core surface. This allows the coordinate system to be normalized by referring to the virtually generated reference plane, and a three-dimensional model covering the entire circumference of the columnar cores W11 to W13 is constructed. Furthermore, by setting the connection end face, the joining surface can be specified in 1-meter increments, and the end faces can be aligned and connected in the depth direction.

[0027] Furthermore, similar processing is performed on multiple columnar cores W11 to W13, and the end face positions of each are automatically detected and joined along the depth direction, as shown in Figure 23. This joining process allows multiple cores to be continuously reconstructed as a single cylindrical model, and a 3D model corresponding to the drilling depth is obtained.

[0028] Furthermore, the 3D modeling of the present invention reduces the need to store the storage boxes 1 for long periods of time, thereby reducing the costs of warehouse space and reconfirmation work. Conventionally, it was necessary to store a huge number of core boxes and remove and inspect the actual items when necessary, but the present invention makes it possible to refer to digital 3D models, realizing efficient and sustainable core management.

[0029] Furthermore, when acquiring photographic data, by providing reference markers on the outer periphery of the storage box 1 or on the cushioning material 2, it becomes possible to easily integrate data acquired at different angles and different time periods into the same coordinate system. These markers may be special markers that can be identified, such as printed patterns. This allows data acquired at different locations or with different equipment to be integrated with high reproducibility.

[0030] While photography can be performed using commercially available digital cameras and SfM software, the method of the present invention can also be realized using a dedicated large-format multi-lens scanner. In this case, multiple cameras are synchronously positioned to capture images simultaneously, allowing high-precision 3D data to be obtained in a short time, regardless of the photographer's skill or conditions. In particular, the scanner can create 3D models of the front and back of a work in just a few minutes, significantly reducing processing time in practice.

[0031] The process of removing unnecessary parts according to the present invention is not limited to simple rectangular area or three-dimensional area designation, but can be automated using image processing algorithms. For example, the cushioning material 2 has a regular semicircular cross section, so it can be easily identified using a contour extraction algorithm.

[0032] Furthermore, because the core storage box 3 has straight edges, it can be automatically extracted and removed using edge detection processing. Furthermore, by using a machine learning model to distinguish between cores and unnecessary parts, manual editing by the user can be minimized. Specifically, with the top surface image of the core box displayed in a bird's-eye view, it is possible to specify a rectangular area and erase unnecessary parts in three dimensions.

[0033] Furthermore, when joining multiple cores W11-W13 in the depth direction, the continuity of the core surface patterns and color tones can be used as a criterion, rather than simply matching the geometric shapes of the core ends. Specifically, image recognition algorithms can be used to match lithologic boundaries and mineral grain distribution patterns, improving the accuracy of joining adjacent cores. Furthermore, by combining this with drilling depth data during core collection, the reliability of automatic detection of the core end positions can be improved. [Explanation of symbols]

[0034] 1 Storage box, 2, 2A, 2B cushioning material, 3 Core storage box, 3A, 3B Box body, 4 Hinge, 10 Marker, 22 Groove, 31-36 Air holes, W11 Columnar core, W12 Columnar core, W13 Columnar core.

Claims

1. A method for generating a core model using three-dimensional models of a front surface and a back surface obtained for a container that contains a columnar core, comprising: a first step of removing portions other than the columnar core from the three-dimensional models of the front and back surfaces based on designated regions; a second step of aligning and joining the three-dimensional models of the front and back surfaces, from which portions other than the columnar core have been removed in the first step, based on a reference plane; a third step of joining the three-dimensional models of the columnar cores aligned and joined in the second step along a depth direction; A core model generation method comprising:

2. 2. The core model generating method according to claim 1, A core model generating method characterized in that the removal of the portion other than the columnar core is performed based on a rectangular area specification or a three-dimensional range specification.

3. 2. The core model generating method according to claim 1, A core model generation method characterized in that the joining along the depth direction is performed by automatically detecting the end face positions of each columnar core.

Images