A three-dimensional geological field recording method based on two-dimensional display images of a slope

Abstract

The present application is suitable for the technical field of water conservancy and hydropower engineering geological record, and provides a three-dimensional geological real scene record method based on two-dimensional display image of slope, which comprises the following steps: importing a design slope surface and obtaining a three-dimensional model of a slope real scene, establishing a slope space coordinate system and a plane display coordinate system, and establishing the conversion relationship between the coordinate systems; based on the optical imaging principle, the intersection of the projection ray and the three-dimensional model of the slope real scene is accurately and quickly calculated, the original image is strictly geometrically corrected, and a two-dimensional display image is generated; the large-format two-dimensional display image is tile segmented as a digital base map for geological record, and the three-dimensional model data of the slope real scene are cooperated to realize the spatial calculation of the two-dimensional display image, the automatic calculation of the occurrence, and the generation of a record vector result map. The present application realizes the automatic generation of high-precision and high-resolution two-dimensional display image of the slope and three-dimensional geological record, and significantly improves the precision and efficiency of the slope geological record.

Description

Technical Field

[0001] This invention relates to the field of geological logging technology for water conservancy and hydropower projects, specifically a method for logging three-dimensional geological scenes based on two-dimensional slope images. Background Technology

[0002] Slope geological logging for water conservancy and hydropower projects is a process that uses text, maps, tables, and other forms to systematically and objectively reflect the geological phenomena observed during geological exploration and engineering development, as well as the results of comprehensive research. It is a very important part of the construction of water conservancy and hydropower projects, providing scientific basis for project site selection, engineering design, engineering construction, rock mass stability analysis, and geological disaster early warning.

[0003] Currently, slope geological logging primarily employs the on-site sketching method. This involves observing geological conditions on-site, using a geological compass and measuring tape to measure geometric information such as the orientation of geological structural lines, and then manually drawing sketches. This traditional method suffers from drawbacks such as high labor intensity, low efficiency, low geometric accuracy, delayed project progress, and slow information feedback, making it unsuitable for the rapid response of geological information at construction sites required by modern productivity development and information-based construction. It also poses significant safety hazards, especially when working on unstable, steep slopes. With the development of modern spatial information technology, digital photogrammetry has been introduced due to its advantages such as low cost, high efficiency, flexibility, and WYSIWYG (What You See Is What You Get). Scholars have proposed replacing traditional methods with digital photogrammetric geological surveying and logging. With the maturation of photogrammetric real-scene 3D modeling technology, the current trend is towards real-scene 3D logging, where a real-scene 3D model of the slope serves as a digital base map, enabling geological observation, spatial measurement, and digital drawing on a computer.

[0004] However, existing real-scene 3D logging methods still have many shortcomings in practical applications: First, operating 3D models through 2D screens is inconvenient, especially on mobile web platforms such as smartphones and tablets, affecting logging efficiency; second, 3D logging cannot directly obtain standardized results such as 2D vector graphics like slope display maps and elevation projection maps required by current industry standards; third, real-scene 3D slope models cannot be directly applied to intelligent geological interpretation using technologies such as digital image processing and deep learning, and the spatial resolution of the texture in real-scene 3D slope models is usually lower than that of the original images, making geological visual interpretation relatively difficult. Therefore, in view of the above situation, there is an urgent need to provide a 3D geological real-scene logging method based on 2D slope display images to overcome the shortcomings in current practical applications. Summary of the Invention

[0005] The purpose of this invention is to provide a method for recording three-dimensional geological scenes based on two-dimensional slope images, effectively solving the problems mentioned in the background art.

[0006] This invention is implemented as follows: a method for recording three-dimensional geological scenes based on two-dimensional slope images, comprising the following steps: S1. Import the designed slope and obtain the actual 3D model of the slope, establish the slope spatial coordinate system and the slope plane display coordinate system, and establish the two-way coordinate transformation relationship between the slope spatial coordinate system, the slope plane display coordinate system and the known ground measurement coordinate system; S2. Based on the design slope and coordinate transformation relationship, the original image is projected onto the design slope through geometric correction to generate a two-dimensional display image corresponding to the design slope. During the correction process, spatial geometric correction is performed based on the actual three-dimensional model of the slope to obtain the real ground spatial coordinates of each image point. S3. Using the two-dimensional display image as a digital base map, draw or identify geological structure information, and based on the coordinate transformation relationship and spatial geometric correction, calculate the three-dimensional spatial coordinates and attitude of the geological structure surface, and generate a vector result map that meets the requirements of geological logging specifications.

[0007] As a further aspect of the present invention: in step S2, the spatial resolution of the two-dimensional display image... Based on the average shooting distance of the original images D Camera principal distance f Camera sensor image size d x and image width w o The calculation is determined, and the calculation formula is: .

[0008] As a further aspect of the present invention: In S2, the spatial geometric correction is achieved by constructing a spatial projection ray and intersecting it with the actual three-dimensional model of the slope. The spatial projection ray is constructed with the point obtained by translating the slope point on the design slope surface corresponding to the pixel of the two-dimensional display image outward by a set distance along the slope normal vector direction, and with the opposite direction of the slope normal vector as the ray direction. The set distance is greater than the maximum over- or under-excavation depth of the slope surface relative to the design slope surface.

[0009] As a further aspect of the present invention: in S2, the intersection with the actual 3D model of the slope is achieved using a multi-branch tree index combined with a collision detection model, including: The spatial bounding boxes and triangular mesh information of each node in the actual 3D model of the slope are read in and a multi-branch tree index structure is constructed. The intersection of the spatial projection ray with the bounding boxes of each level of nodes is determined by a collision detection model in order to locate the leaf nodes that intersect with the ray. The triangulation within the located leaf nodes is subjected to intersection calculation, and the effective intersection point closest to the ray origin is selected through visibility analysis.

[0010] As a further aspect of the present invention: in S2, for the two-dimensional display image whose image size exceeds a preset threshold, a block-based strategy is adopted to perform block-by-block geometric correction and independent storage; and before performing geometric correction, a standardized enhancement method is used to perform uniform radiometric correction on the original image.

[0011] As a further aspect of the present invention: In step S3, before extracting geological information using the two-dimensional display image as a digital base map, the two-dimensional display image is first segmented into pyramid tiles, including: The two-dimensional display image is downsampled at multiple resolutions to generate an image pyramid; Each level of the image pyramid is divided into tiles of a preset size and encoded. The tiles are stored independently in a compressed format.

[0012] As a further aspect of the present invention: in step S3, the method for calculating the three-dimensional spatial coordinates of the geological structural surface is as follows: Obtain the node pixel coordinates of structural lines drawn by the user or structural surface patches identified on the two-dimensional display image; The two-dimensional coordinates of each node in the slope plane display coordinate system are calculated based on the node pixel coordinates and the resolution of the two-dimensional display image. The coordinate transformation relationship is used to convert the two-dimensional coordinates into slope spatial coordinates and ground measurement coordinates; Based on the ground measurement coordinates, a spatial projection ray is constructed and intersected with the actual 3D model of the slope to obtain the true 3D spatial coordinates of each node.

[0013] As a further aspect of the present invention: in step S3, the method for calculating the attitude of the geological structural plane is as follows: For three or more non-collinear nodes on the same geological structural surface, a spatial plane equation is fitted based on the true three-dimensional spatial coordinates of the geological structural surface, and the dip and dip angle of the geological structural surface are calculated based on the normal vector of the spatial plane equation.

[0014] As a further aspect of the present invention: in step S3, the vector result map includes at least a three-dimensional geological map, a slope display map, a horizontal projection map, and an elevation projection map; The slope display map is generated based on the slope plane display coordinates, the horizontal projection map is generated based on the plane coordinates in the ground measurement coordinates and the elevation value is set to zero, and the elevation projection map is generated based on the X-axis and Z-axis coordinates in the slope spatial coordinate system.

[0015] As a further aspect of the present invention: in step S3, the method for extracting the geological information includes manual drawing with human-computer interaction and automatic recognition based on deep learning.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: This method enables the automatic generation of high-precision, high-resolution two-dimensional slope images and three-dimensional geological logging of 2D images. It achieves accurate measurement of true three-dimensional coordinates based on two-dimensional images and can directly generate vector graphics that meet the specifications, which can significantly improve the accuracy and efficiency of slope geological logging. This method can significantly shorten the fieldwork time of geological professionals, greatly reduce the labor intensity of fieldwork, achieve high-quality, high-efficiency and information-based management of geological logging results, improve work quality, product quality and economic benefits, and achieve the goal of safety and efficiency. Compared with the logging of 3D models of slope real scenes, this method is more convenient and flexible for 2D image operation. Through pyramid tile segmentation and dynamic scheduling, it occupies less computer hard disk and memory resources and can realize rapid geological logging on ordinary computers or Web terminals. It provides key technical support for the development of high-performance geological logging software for ordinary PCs and mobile Web terminals. This method can change the traditional manual operation method and move towards the development and comprehensive utilization of high technology, effectively meet the needs of modern water conservancy and hydropower engineering construction, and help the development of smart water conservancy and the digital transformation of the industry. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the method of the present invention.

[0019] Figure 2 This is a schematic diagram of the slope coordinate system provided in an embodiment of the present invention.

[0020] Figure 3 This is a schematic diagram of the two-dimensional display image recording results of the slope provided in the embodiments of the present invention. Figure 1 .

[0021] Figure 4 This is a schematic diagram of the two-dimensional display image recording results of the slope provided in the embodiments of the present invention. Figure 2 . Detailed Implementation

[0022] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] The present invention will be further explained below with reference to specific embodiments.

[0024] Example 1: Please refer to Figures 1-4 The present invention provides a method for recording three-dimensional geological scenes based on two-dimensional slope images, comprising: S1. Establishing the coordinate system and transformation model: Import the designed slope surface, establish a spatial coordinate system and a planar display coordinate system for the slope, and use the slope surface as the projection reference surface, i.e., the display surface, for the geological logging results map. For example... Figure 2 As shown, coordinate system T - ENH Given a known ground measurement coordinate system, O - XYZ For the slope spatial coordinate system, O - xy A coordinate system is displayed for the slope plane. The origin is shown below. O To design the lower left corner point of the slope; in the coordinate system T - ENH middle ， Elevation is H O ( O The intersection of the horizontal plane (point elevation) and the design slope is determined by the intersection line pointing to the right. X axis; Z The axis is vertically upward, aligned with the ground measurement coordinate system. H The axes are parallel; Y The axis points inwards from the slope and is in line with... X , Z Orthogonal axes; slope plane display coordinate system x shaft and X Axis alignment; y The axis is located on the designed slope, upward along the slope and parallel to... x The axes are orthogonal.

[0025] Let in coordinate system T - ENH In the design slope equation as shown in equation (1), the direction and dip angle of the design slope are calculated according to equations (2) and (3), respectively.

[0026] (1) In the formula, ( E , N ,H ( ) represents the three-dimensional coordinates of points on the design slope surface. A, B, C ) is the slope normal vector, and is a unit vector.

[0027] (2) In the formula The slope direction is represented by a value ranging from 0 to 360°.

[0028] (3) In the formula The slope angle is designed with a range of 0 to 90°.

[0029] Then coordinate system T - ENH With O- XYZ The transformation relationship is shown in equation (4).

[0030] (4) In the formula ( )for O In coordinate system T - ENH The three-dimensional coordinates in the middle; To design a point on a slope in the coordinate system T - ENH The three-dimensional coordinates in ( ) represents the point in coordinate system O- XYZ The three-dimensional coordinates in the image.

[0031] The inverse transformation formula of equation (4) is shown in equation (5).

[0032] + (5) The transformation relationship between the slope display coordinates and the slope spatial coordinates is shown in equation (6). (6) The inverse transformation formula of equation (6) is shown in equation (7).

[0033] (7) S2. Generate a two-dimensional display image: (1) Generation of blank two-dimensional display images Based on the actual extent of the slope and the set spatial resolution, a blank two-dimensional display image is generated, specifically as follows: Based on the shooting distance, camera focal length, camera sensor size, and image size, the spatial resolution of the original slope image is estimated, and then a blank two-dimensional display image is generated based on the actual extent of the slope.

[0034] existO - xy In the coordinate system, let the slope be along x The actual length of the direction is L ,along y The actual width of the direction is W The spatial resolution of the two-dimensional display image is The image size of the two-dimensional display image is calculated according to formula (8).

[0035] (8) In the formula, L, W They are respectively the slope along x The actual length of the direction and along y Actual width in the direction, in meters; Spatial resolution of a two-dimensional image, unit: m / pixel; l、w These represent the image length and width of a two-dimensional display image, respectively, in pixels.

[0036] To ensure that the two-dimensional display image has the same clarity as the original image, the average resolution of the original image is estimated according to formula (9) and used as the spatial resolution of the two-dimensional display image.

[0037] (9) In the formula, d x , f Here, represents the camera sensor image width and represents the camera main distance, in mm; D Average shooting distance, unit: m; w o The width of the original image, in pixels.

[0038] (2) Calculate the spatial coordinates of the slope pixel by pixel Based on the spatial resolution of the two-dimensional display image and the row and column numbers of each pixel, and based on the slope coordinate system transformation model established in step two, the slope spatial coordinates and ground measurement coordinates are calculated pixel by pixel. For a blank two-dimensional display image, the corresponding display coordinates are calculated pixel by pixel according to formula (10). x, y Then calculate its slope spatial coordinates according to formula (6). X, Y, Z ), and then, its ground measurement coordinates are calculated according to formula (5). E, N, H ).

[0039] (10) In the formula, r , c These are the row and column numbers of the pixel in the 2D display image, in pixels.

[0040] (3) Establishment of spatial projection rays By displaying the slope points corresponding to the pixels in the two-dimensional image, and extending a distance greater than the maximum estimated over- or under-excavation along the slope normal vector direction, a spatial ray opposite to the slope normal vector direction is established as the starting point in the coordinate system. T - ENH In the equation, the projection ray is shown in equation (11).

[0041] (11) In the formula, ( X ray , Y ray , Z ray Let be the three-dimensional coordinates of any point on the ray, and ( E i , N i , H i (pixels) i The coordinates of the corresponding slope point, i.e., the coordinates of the intersection of the ray and the slope; (>0) is the proportionality coefficient; X 0, Y 0, Z 0) represents the coordinates of the ray's starting point; (>0) is a set constant.

[0042] The above formula extends the slope point by a distance along the direction of the slope normal vector. The ray then serves as the starting point, and its extension direction is opposite to the direction of the slope normal. To ensure that the ray intersects with the actual 3D model of the slope, a constant is used. Set to any value greater than the maximum over- or under-excavation depth of the slope (e.g., 10m).

[0043] (4) Calculation of pixel-based real ground coordinates This invention employs a multi-branch tree index and collision detection model to achieve accurate and rapid intersection finding between spatial rays and a real-world 3D slope model. Through visibility analysis, the 3D coordinates of the actual ground points corresponding to the pixels in the 2D display image are obtained. Given the massive number of spatial triangles in the real-world 3D slope model, conventional spatial intersection algorithms require repeatedly reading the model tile data file and traversing all triangles, resulting in extremely low efficiency. Therefore, this invention uses a multi-branch tree index and collision detection model to quickly locate slope real-world 3D model tiles that intersect with the projected ray, avoiding repeated access to hard drive files and full traversal of the model's massive spatial triangles, thus achieving rapid intersection finding. The algorithm flow is as follows: The slope real scene 3D model tile data file is traversed layer by layer, and the spatial information of each node of the slope real scene 3D model, such as the bounding box of the cuboid, vertex coordinates, and spatial triangulation, is stored in a multi-way tree index data structure at one time; A collision detection model is used to quickly determine whether the projected ray intersects with the cuboid bounding box of the model's root node; If there is an intersection, calculate whether the projected ray intersects with the bounding box of the child nodes of that node; If a leaf node is reached during traversal, meaning that the node has no child nodes, then the ray intersects with the spatial triangulation in that node, and the coordinates of the intersection point are stored in the stack. Once the multi-branch tree traversal is complete, the number of intersection points is counted. If multiple intersection point coordinates exist, a visibility analysis is performed, and the intersection point closest to the ray origin is taken as the valid intersection point. The three-dimensional coordinates of the intersection point are the real ground coordinates corresponding to the two-dimensional display image pixels.

[0044] (5) Original image object space transformation According to the collinear equation of the center projection of the camera optical imaging, as shown in Equation (12), the intersection point of the ray and the model (denoted as A) is projected into each original image, the theoretical value of the image plane coordinate of the projection point is calculated, the optimal projection point that is closest to the image principal point within the image range is found, and the corresponding original image is used as the reference original image.

[0045] (12) In the formula, ( x A , y A () represents the theoretical coordinates of point A in the original image; x 0, y 0) represents the coordinates of the principal point; f The principal distance of the camera; Let A be the coordinates of the intersection point A between the ray and the model; X S , Y S , Z S For the outer orientation line elements of the image; exterior azimuth elements of the image The rotation matrix is ​​determined and the calculation formula is shown in equation (13).

[0046] (13) Calculate the theoretical coordinates of point A's projection point in each original image, find the projection point closest to the principal image point within the image frame, and use it as the optimal projection point. Use the original image corresponding to the optimal projection point as the reference original image.

[0047] (6) Image point coordinate distortion correction Given that ordinary digital images have objective lens distortion, the actual image point coordinates are calculated using the objective lens distortion correction model shown in Equation (13) based on the theoretical values ​​of the reference image coordinates. This model takes into account both radial and tangential distortion differences and retains higher-order distortion coefficients, achieving a high accuracy of less than 0.1 pixels for simulating distortion differences of various lenses.

[0048] (14) In the formula, The actual image coordinates of point A; , where is the squared distance from the image point to the principal point; Radial distortion coefficient; denoted as the tangential distortion coefficient.

[0049] (7) Pixel grayscale value resampling Based on the actual image point coordinates, bilinear interpolation is used to resample the RGB value of the pixel from the original image, thus assigning it to the pixel in the two-dimensional display image. i .

[0050] Following the steps (1) to (7) above, assign RGB values ​​to each pixel of the two-dimensional display image to finally obtain a complete two-dimensional slope two-dimensional display image.

[0051] If the image size of the 2D display is too large, exceeding the set threshold size (e.g., 4096×4096 pixels), the 2D display image will be divided into blocks, and geometric correction will be performed on each block, and each block will be saved.

[0052] To avoid noticeable seam lines in the two-dimensional display image due to differences in brightness values ​​among the photos, a standardized enhancement method was used to perform uniform radiometric correction on the original images before correction, so that the grayscale histogram of each image remained basically consistent.

[0053] S3. Geological logging based on two-dimensional imagery: (1) Two-dimensional display image pyramid tile segmentation To address the issue of insufficient computer memory preventing the normal opening of ultra-large format images, a pyramid tile segmentation method is used for large-format 2D display images. First, the 2D display image is spatially downsampled to a size smaller than the screen size using multi-resolution downsampling, generating a multi-resolution pyramid image. Then, the 2D display image at each resolution level of the pyramid is segmented into 256×256 pixel tiles, and each tile is encoded and stored independently using JPG compression format. This allows for rapid opening and scheduling of 2D display images with minimal computer memory consumption, and enables efficient scaling and roaming of large-format images.

[0054] (2) Two-dimensional display image spatial measurement Using 2D imagery as the base map for geological logging, and through collaboration between the 2D imagery and 3D slope model data, 3D spatial measurements of the slope are computationally achieved. This allows the high-intensity, high-risk geological logging work to be completed indoors. The automatic 3D coordinate measurement algorithm for 2D imagery is as follows: For any pixel in a two-dimensional display image q Based on its pixel row and column numbers and the resolution of the two-dimensional display image, its coordinate system is calculated according to formula (10). O-xy Two-dimensional display coordinates ( x q , y q ).

[0055] The transformation formula between coordinates and slope space coordinates is shown in equation (6), which converts the two-dimensional coordinates into slope space coordinates. X q , Y q , Z q ); Then, according to the conversion formula between slope spatial coordinates and ground measurement coordinates in equation (5), the slope spatial coordinates are converted into ground measurement coordinates. ); Then, establish a path through the point according to equation (11). q A spatial ray perpendicular to the slope; Intersecting the spatial ray with the actual 3D model of the slope, we obtain the image points. q The actual three-dimensional coordinates of the ground.

[0056] (3) Calculation and mapping of geological occurrence When cataloging two-dimensional display images, a human-computer assisted method is used to draw geological structure lines on the images, or image intelligent recognition algorithms such as deep learning are used to automatically extract structural surface patches. Then, according to the above-mentioned two-dimensional display image spatial measurement algorithm, the spatial coordinates of more than three nodes of the structural line or patch outline are automatically calculated. Then, through least squares spatial plane fitting, the structural surface equation shown in equation (1) is obtained. Finally, according to equations (2) and (3), the dip azimuth and dip angle of the structural surface are calculated, that is, the geological attitude, such as Figure 3 and Figure 4 As shown.

[0057] After the logging is completed, the logging result vector maps required by the geological logging specifications will be automatically generated. The generation methods for each vector map are as follows: 3D vector map: Calculating the 3D ground measurement coordinates of graphic nodes ( E, N, H) , to generate an image; Slope display diagram: Calculate the planar display coordinates of each graphic node ( (x, y) Directly generate an image; Planar projection: Take the coordinates of the graphic nodes on the ground ( E, N, H) Planar coordinates in E, N ), elevation value H Set to 0 to generate the image; Elevation projection: Calculate the slope spatial coordinates of each graphic node ( (X, Y, Z) ,Pick( X, Z The coordinates are used as the two-dimensional plane coordinates of the elevation projection drawing to form the drawing.

[0058] Example 2: Example 2 provides a method for compiling three-dimensional geological real-scene data based on two-dimensional slope imagery. This construction method can be used as a preliminary step for the method described in Example 1, providing three-dimensional slope real-scene model data for S2 and S3 in Example 1. The specific implementation steps are as follows: S0, Construction of a realistic 3D model of the slope: The spatial location and excavation morphology data of the engineering excavation slope design were collected, control points were set up, and equipment such as drones were used to collect slope sequence images to ensure the overlap of slope image acquisition. Furthermore, based on the images, the internal and external orientation elements and sparse point clouds of the images were calculated using the motion reconstruction structure method. Then, through dense reconstruction and grid construction, a real-world 3D model of the slope was generated.

[0059] The slope real-scene 3D model construction method in this embodiment is not limited to the geological logging method described in embodiment 1, but can also be applied independently to other scenarios that require the construction of slope real-scene 3D models.

[0060] In summary, this invention fundamentally eliminates positional errors caused by traditional image projection distortion by constructing a dedicated spatial coordinate system based on the designed slope and achieving bidirectional precise transformation between it and the ground measurement coordinate system. This gives the generated two-dimensional slope display image a rigorous mathematical foundation, allowing it to be directly used for high-precision three-dimensional spatial measurements (such as structural surface attitude, trace length, etc.) and recording, significantly improving measurement accuracy and reliability. By employing strategies such as "spatial ray intersection + multi-branch tree index collision detection," pixel-by-pixel high-precision mapping between the 3D model of the slope and the 2D display image was achieved. This technology can effectively identify and remove occluded areas, thereby solving the problems of image splicing misalignment, texture stretching, or overlap on slopes with uneven surfaces, and generating a visually realistic and geometrically distortion-free geological logging base map. By performing spatial multi-resolution downsampling and encoding on large-format 2D display images, and combining this with a self-established coordinate transformation system, the system can not only output high-precision slope display maps, but also automatically and in batches generate a series of standardized geological maps such as planar projection maps and elevation projection maps. This changes the cumbersome process of traditional manual drawing or software processing, realizing one-click automated map generation from "3D reality" to "2D multi-view," greatly improving the efficiency of data entry and the standardization of results.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for recording three-dimensional geological scenes based on two-dimensional slope images, characterized in that, Includes the following steps: S1. Import the designed slope and obtain the actual 3D model of the slope, establish the slope spatial coordinate system and the slope plane display coordinate system, and establish the two-way coordinate transformation relationship between the slope spatial coordinate system, the slope plane display coordinate system and the known ground measurement coordinate system; S2. Based on the design slope and coordinate transformation relationship, the original image is projected onto the design slope through geometric correction to generate a two-dimensional display image corresponding to the design slope. During the correction process, spatial geometric correction is performed based on the actual three-dimensional model of the slope to obtain the real ground spatial coordinates of each image point. S3. Using the two-dimensional display image as a digital base map, draw or identify geological structure information, and based on the coordinate transformation relationship and spatial geometric correction, calculate the three-dimensional spatial coordinates and attitude of the geological structure surface, and generate a vector result map that meets the requirements of geological logging specifications.

2. The method according to claim 1, characterized in that, In S2, the spatial resolution of the two-dimensional display image Based on the average shooting distance of the original images D Camera principal distance f Camera sensor image size d x and image width w o The calculation is determined, and the calculation formula is: 。 3. The method according to claim 1, characterized in that, In S2, the spatial geometric correction is achieved by constructing a spatial projection ray and intersecting it with the actual three-dimensional model of the slope. The spatial projection ray is constructed with the point obtained by translating the slope point on the design slope surface corresponding to the pixel of the two-dimensional display image outward by a set distance along the slope normal vector direction, and with the opposite direction of the slope normal vector as the ray direction. The set distance is greater than the maximum over- or under-excavation depth of the slope surface relative to the design slope surface.

4. The method according to claim 3, characterized in that, In S2, the intersection with the actual 3D model of the slope is achieved using a multi-way tree index combined with a collision detection model, including: The spatial bounding boxes and triangular mesh information of each node in the actual 3D model of the slope are read in and a multi-branch tree index structure is constructed. The intersection of the spatial projection ray with the bounding boxes of each level of nodes is determined by a collision detection model in order to locate the leaf nodes that intersect with the ray. The triangulation within the located leaf nodes is subjected to intersection calculation, and the effective intersection point closest to the ray origin is selected through visibility analysis.

5. The method according to claim 1, characterized in that, In step S2, for the two-dimensional display image whose image size exceeds a preset threshold, a block-based strategy is adopted to perform block-by-block geometric correction and independent storage; and before performing geometric correction, a standardized enhancement method is used to perform uniform radiometric correction on the original image.

6. The method according to claim 1, characterized in that, In step S3, before extracting geological information using the two-dimensional display image as a digital base map, the two-dimensional display image is first segmented into pyramid tiles, including: The two-dimensional display image is downsampled at multiple resolutions to generate an image pyramid; Each level of the image pyramid is divided into tiles of a preset size and encoded. The tiles are stored independently in a compressed format.

7. The method according to claim 1, characterized in that, In S3, the method for calculating the three-dimensional spatial coordinates of the geological structural surface is as follows: Obtain the node pixel coordinates of structural lines drawn by the user or structural surface patches identified on the two-dimensional display image; The two-dimensional coordinates of each node in the slope plane display coordinate system are calculated based on the node pixel coordinates and the resolution of the two-dimensional display image. The coordinate transformation relationship is used to convert the two-dimensional coordinates into slope spatial coordinates and ground measurement coordinates; Based on the ground measurement coordinates, a spatial projection ray is constructed and intersected with the actual 3D model of the slope to obtain the true 3D spatial coordinates of each node.

8. The method according to claim 1, characterized in that, In S3, the method for calculating the attitude of geological structural planes is as follows: For three or more non-collinear nodes on the same geological structural surface, a spatial plane equation is fitted based on the true three-dimensional spatial coordinates of the geological structural surface, and the dip and dip angle of the geological structural surface are calculated based on the normal vector of the spatial plane equation.

9. The method according to claim 1, characterized in that, In S3, the vector result map includes at least a three-dimensional geological map, a slope display map, a horizontal projection map, and an elevation projection map; The slope display map is generated based on the slope plane display coordinates, the horizontal projection map is generated based on the plane coordinates in the ground measurement coordinates and the elevation value is set to zero, and the elevation projection map is generated based on the X-axis and Z-axis coordinates in the slope spatial coordinate system.

10. The method according to any one of claims 1 to 9, characterized in that, In S3, the methods for extracting geological information include interactive manual drawing and automatic recognition based on deep learning.

Images