An efficient and adaptive compression method for visualizing geological-geophysical body data

Abstract

This application provides an efficient and adaptive compression method for geological-geophysical volume data, relating to the field of geophysical exploration. The method includes: acquiring multi-source geological interpretation prior data and geophysical volume data to generate spatially aligned semantic indicators; integrating the display capabilities of a visualization platform to generate a geological semantic mask containing key areas, transition zones, and background areas; assigning compression tolerance to each voxel based on the mask to form a voxel-level compression tolerance field; buffering and smoothing interpolation of key area boundaries, compressing data in blocks, and recording metadata; constructing a spatial-semantic-value multi-level index; and combining view frustum and user interaction intent to stream, decompress, and differentially render data blocks according to priority. This invention, through geological semantic-driven differential compression and multi-level indexing, significantly improves geological interpretation accuracy and visualization efficiency while ensuring high fidelity in key areas such as ore bodies and faults.

Description

Technical Field

[0001] This application relates to the field of geophysical exploration, and in particular to an efficient and visual adaptive compression method for geological-geophysical volume data. Background Technology

[0002] In emerging application scenarios such as digital twin mines, smart geological exploration, and urban underground space management, the scale of 3D geological-geophysical volume data is exploding (often reaching TB levels in a single project). Typical data include high-resolution resistivity volumes, gravity anomaly volumes, and magnetic susceptibility volumes. Traditional compression methods (such as ZFP and OpenVDB) use globally uniform error tolerance or bit rate for lossy compression, which can reduce storage volume, but has two major technical bottlenecks: (1) Insufficient fidelity of key geological features: The same compression strength is used for sensitive areas such as ore body boundaries and fault structures as for homogeneous surrounding rocks, which leads to fuzziness, fractures or artifacts in the former after decompression, seriously affecting the scientificity and credibility of geological interpretation. (2) Poor visualization efficiency and interactive experience: Existing compression schemes do not take into account the terminal rendering requirements and cannot support streaming loading according to semantic priority. When users operate large models on web browsers or mobile devices, they often face problems such as slow loading, low frame rate (<10 FPS) and invisible key targets.

[0003] While general-purpose volumetric data compression tools (such as ZFP) support fixed error patterns, they lack the ability to perceive prior geological semantic knowledge. Existing visualization systems (such as Cesium Ion and Potree), although supporting Level of Detail (LOD), base their hierarchical division on geometric distance, failing to incorporate geological semantics. More importantly, existing technologies lack a closed-loop method that dynamically links "geological interpretation—compression fidelity—streaming loading." For example, even if the user manually delineates the ore body area, traditional methods cannot automatically and smoothly transition compression intensity at the boundaries, resulting in noticeable artifacts after decompression and failing to support client-side priority loading based on "geological semantics." Therefore, there is an urgent need for an intelligent compression method that can integrate geological interpretation knowledge, adaptively allocate compression precision, and support efficient streaming visualization. Summary of the Invention

[0004] The purpose of this invention is to overcome the dual shortcomings of existing geophysical body data compression methods in terms of key geological feature fidelity and terminal visualization efficiency, and to provide an efficient and visual adaptive compression method for geological-geophysical body data.

[0005] The above-mentioned objective of this application is achieved through the following technical solution: S1: Acquire multi-source geological interpretation prior data and geophysical volume data, perform coordinate system unification and rasterization processing, and generate spatially aligned semantic indicator volumes; S2: Based on semantic indicators, and integrating the display capabilities of the target visualization platform, a geological semantic mask containing key areas, transition zones, and background areas is generated; S3: Based on the geological semantic mask, a compression tolerance is assigned to each voxel to form a voxel-level compression tolerance field; S4: Buffer the critical region boundary, smooth the voxel-level compression tolerance field, and compress it in blocks, recording block metadata. S5: Based on block metadata, construct a spatial-semantic-value multi-level index; S6: Based on spatial-semantic-value multi-level indexing, combined with view frustum and user interaction intent, load, decompress and render data blocks in a priority-based streaming manner.

[0006] Optionally, step S1 includes: The geophysical volume data includes: resistivity volume, density volume, and magnetic susceptibility volume; the multi-source geological interpretation prior data includes: borehole data, three-dimensional orebody models, and fault vector data.

[0007] Optionally, step S2 includes: S21: Read the display parameters of the target visualization platform and set the minimum visible geological body size; S22: Based on the minimum visible geological volume size, perform three-dimensional morphological expansion on the semantic indicator volume of the key area to ensure that small-scale geological volumes are visible on the terminal. S23: Construct a transition zone based on Euclidean distance transformation to generate a geological semantic mask with three types of labels: key area, transition zone, and background area.

[0008] Optionally, step S3 includes: S31: Calculate the global standard deviation σ_global of geophysical volume data; S32: Based on the voxel labels in the geological semantic mask, set the compression tolerance τ and output the voxel-level compression tolerance field as follows: If the label is a critical region, then τ = 0.05 × σ_global; If the label is a transition zone, then τ = 0.2 × σ_global; If the label is in the background area, then τ = 1.0 × σ_global; If the terminal is detected to be a mobile device and the network bandwidth is lower than the preset threshold, the compression tolerance of the background area will be adjusted to τ=2.0×σ_global; Wherein, the compression tolerance τ is the maximum absolute error allowed for lossy compression, and its physical unit is consistent with that of geophysical volume data; the global standard deviation σ_global is used to quantify the overall fluctuation of the volume data and serves as the benchmark scale for compression tolerance configuration.

[0009] Optionally, step S4 includes: The buffering process includes: Perform isotropic expansion on the critical region boundaries to construct a buffer zone; Within the buffer and transition zone, tolerant linear interpolation is performed based on the distance from the voxel to the critical region boundary. The interpolation formula is as follows:

[0010] in, This is the tolerance after interpolation. For background area tolerance, Let be the critical region tolerance, and d be the minimum Euclidean distance from the voxel to the critical region boundary. The width of the buffer; d and The units are all meters; The value is 2 to 4 times the grid spacing, and is dynamically adjusted according to the complexity of the geological structure. The block compression includes: The input data is divided into fixed-size three-dimensional data blocks, and the compression tolerance within a block is taken as the maximum tolerance value within the block; if there is no effective tolerance data within a block, the default tolerance of the background area is used. Call the ZFP compression library to perform lossy compression in precision mode, and record the compressed byte stream and offset address; The block size is determined based on the target terminal's memory usage and network transmission efficiency constraints. The memory usage of a single block after decompression is ≤2MB, and the size of a single block after compression is ≤200KB.

[0011] Optionally, step S5 includes: The spatial-semantic-value multilevel index includes: Octree indexes are used for space-pruning queries; A hash index, using semantic tags and LOD levels as keys, is used for semantic priority loading; B+ tree indexes, using the range of physical property values ​​as keys, are used for filtering queries based on physical properties.

[0012] Optionally, step S6 includes: The priority-based loading includes: Obtain the set of visible blocks based on the view frustum; Obtain the semantic attention block set based on the user's interaction intent; After merging the set of visible blocks and the set of semantically concerned blocks, sort them according to the priority formula:

[0013] in, The loading priority of block b For the detail level of block b, As the dominant semantic tag, For LOD weighting coefficients, These are semantic weight coefficients; and This is a configurable parameter, with a value range of 0.5 to 1.5, and is dynamically adjusted by the user interaction mode. A larger value indicates a higher priority for spatial precision. A larger value indicates a higher priority in semantic concern; when , ,when , ; The differentiated rendering includes: Phong shading is enabled for the critical region, with lighting model parameters including ambient light intensity 0.3, diffuse reflection coefficient 0.6, specular intensity 0.8, and specular index 32. The transition zone uses basic diffuse coloring; The background area uses downsampling rendering and fast coloring; If the cumulative loading and rendering time exceeds the preset threshold, a degraded loading strategy will be automatically triggered.

[0014] An electronic device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to enable the electronic device to perform an efficient and visual adaptive compression method for geological-geophysical body data.

[0015] A computer-readable storage medium storing instructions that, when executed, perform an efficient and visual adaptive compression method for geological-geophysical volume data.

[0016] The beneficial effects of the technical solution provided in this application are: 1. Significantly enhanced fidelity of key geological features: By integrating multi-source geological interpretation priors to generate a geological semantic mask, and allocating extremely low compression tolerances to key areas accordingly, it is ensured that these areas have clear boundaries, no breaks or artifacts after decompression, effectively guaranteeing the scientific nature of geological interpretation.

[0017] 2. Significantly improved overall compression efficiency and visualization performance: By using a high compression ratio in the background area and combining it with a multi-level index of "spatial-semantic-value", the client can stream the most relevant data blocks on demand, avoiding full data loading and achieving a smooth interactive experience of ≥30FPS on both web and mobile devices.

[0018] 3. Visual continuity is guaranteed: By introducing a smooth transition zone between the key area and the background area and linearly interpolating the tolerance of the boundary buffer zone, visual artifacts caused by abrupt changes in compression intensity are completely eliminated, ensuring the visual continuity and authenticity of the geological structure.

[0019] 4. Excellent cross-platform compatibility and practicality: The output of this method is a standard compressed data file and index file, which natively supports HTTP Range Request and progressive loading. It can run stably in mainstream browsers, domestic operating systems and private deployment environments without relying on heavy plugins.

[0020] 5. High feasibility of engineering implementation: This method is built on a mature open source technology stack and has been prototyped and verified in typical oil and gas exploration, metal mine exploration, urban underground space management and digital twin mine projects. It has the ability to process petabyte-level geophysical data and can meet the visualization frame rate requirement of ≥30FPS in different scenarios. It can be widely used in domestic energy, mining and other industries. Attached Figure Description

[0021] The present application will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a step diagram of an embodiment of this application; Figure 2 This is a schematic diagram of the electronic device structure in the embodiments of this application. Detailed Implementation

[0022] To provide a clearer understanding of the technical features, objectives, and effects of this application, the specific embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0023] The embodiments of this application provide an efficient and visual adaptive compression method for geological-geophysical body data.

[0024] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating the steps of an efficient and visual adaptive compression method for geological-geophysical body data, as described in this application embodiment, including: S1: Acquire multi-source geological interpretation prior data and geophysical volume data, perform coordinate system unification and rasterization processing, and generate spatially aligned semantic indicator volumes; S2: Based on semantic indicators, and integrating the display capabilities of the target visualization platform, a geological semantic mask containing key areas, transition zones, and background areas is generated; S3: Based on the geological semantic mask, a compression tolerance is assigned to each voxel to form a voxel-level compression tolerance field; S4: Buffer the critical region boundary, smooth the voxel-level compression tolerance field, and compress it in blocks, recording block metadata. S5: Based on block metadata, construct a spatial-semantic-value multi-level index; S6: Based on spatial-semantic-value multi-level indexing, combined with view frustum and user interaction intent, load, decompress and render data blocks in a priority-based streaming manner.

[0025] This application provides an embodiment as follows: The technical solution of this application realizes the automatic fusion of multi-source geological interpretation priors (boreholes, 3D models, vector maps) and geophysical volume data to generate a spatially aligned geological semantic mask for visualization needs; based on the mask, the volume data is subjected to content-aware partitioned compression to ensure high fidelity in key areas such as ore bodies and faults, and high compression in background areas; a key area boundary buffer and tolerance linear interpolation mechanism is introduced to eliminate visual artifacts caused by abrupt changes in compression intensity and ensure the continuity of geological structures; a multi-level index supporting joint queries of spatial, semantic, and physical property values ​​is constructed to enable clients to stream according to geological semantics and interactive intents; in resource-constrained environments such as Web and mobile devices, smooth interactive 3D visualization with ≥30FPS is achieved, balancing scientific accuracy and user experience.

[0026] As one embodiment, this application relates to an adaptive compression method driven by the fusion of heterogeneous geophysical body data (such as resistivity, density, and magnetic susceptibility) collected from multiple platforms such as airborne, ground, and well sites with multi-source geological interpretation priors (such as boreholes, ore body models, and fault vectors). It is particularly suitable for efficient visualization and interactive analysis of large-scale underground space models in resource-constrained web, mobile, and VR / AR environments.

[0027] Step S1 includes: The geophysical volume data includes: resistivity volume, density volume, and magnetic susceptibility volume; the multi-source geological interpretation prior data includes: borehole data, three-dimensional orebody models, and fault vector data.

[0028] As one implementation, this is performed by a multi-source data fusion module deployed on a cloud platform data preprocessing node. This module first accesses heterogeneous data sources through a unified data abstraction layer, then rasterizes all prior geological data (boreholes, ore body models, faults) and aligns them with geophysical data, providing a foundation for the subsequent generation of a geological semantic mask for visualization.

[0029] As one implementation example, the system receives and parses the following inputs through the standardized API interface provided by the unified data abstraction layer: The geophysical data file resistance.vol (format: SEG-Y extended or custom binary) contains metadata including grid size (512×512×256), grid spacing (Δx=Δy=Δz=10m), and physical units (Ω·m). The boreholes.csv table contains metadata including field definitions (x, y, z_top, z_base, lithology, cu_grade). The 3D orebody model orebody.obj (WavefrontOBJ format) represents the triangular mesh on the surface of the copper orebody; Fault data faults.shp (ESRIShapefile format).

[0030] As one example, the system uses the grid extent and resolution of the geophysical volume data resistance.vol as a reference and performs the following operations: (1) Coordinate system one: Project all data to the CGCS2000 / UTMZone50N coordinate system.

[0031] (2) Rasterization of borehole data: Traverse boreholes.csv, first filter invalid data (records where cu_grade is NULL or lithology is undefined); for valid records, if lithology contains 'chalcopyrite' or cu_grade≥0.5, mark the segment as a mineralized segment; call the inverse distance weighted (IDW) interpolator (power exponent=2, search radius=50m) to expand the mineralized segment information into a three-dimensional mineralization indicator; if a grid point has no adjacent valid mineralized segment data (no valid boreholes within the search radius), mark the point as 'unsampled' or 'low confidence', and record the data blank area in the metadata for reference in subsequent interpolation or uncertainty analysis.

[0032] (3) Rasterization of orebody model: Call the vtkImplicitPolyDataDistance module of VTK library to calculate the signed distance from each resistance.vol grid point to the surface of orebody.obj; if the distance is ≤0, then assign a value of 1 to the corresponding position in the orebody indicator I_ore_from_model.

[0033] (4) Fault data rasterization: Voxelize the fault plane in faults.shp within ±5m along its normal direction to generate a fault indicator I_fault with the same resolution as resistivity.vol.

[0034] As one example, the system integrates multi-source mineralization information and outputs structured results: Aligned resistivity volume V: i.e., resistivity.vol after coordinate system one.

[0035] The semantic indicator set is {"ore":I_ore,"fault":I_fault}, where I_ore = logical_or(I_ore_from_borehole,I_ore_from_model). If the same voxel is marked conflictingly in I_ore_from_borehole and I_ore_from_model (one is 1, the other is 0), then the borehole data (I_ore_from_borehole) takes precedence (bore data is direct exploration data and has higher reliability). All output data have the exact same grid size, spacing, and coordinate system as V.

[0036] Step S2 includes: S21: Read the display parameters of the target visualization platform and set the minimum visible geological body size; S22: Based on the minimum visible geological volume size, perform three-dimensional morphological expansion on the semantic indicator volume of the key area to ensure that small-scale geological volumes are visible on the terminal. S23: Construct a transition zone based on Euclidean distance transformation to generate a geological semantic mask with three types of labels: key area, transition zone, and background area.

[0037] As one embodiment, the semantic mask generator module is executed, taking as input the generated aligned resistivity volume V, semantic indicator I_ore, and I_fault. This step dynamically adjusts the extent of key geological regions based on the characteristics of the target visualization platform to ensure that key targets are visible and have smooth boundaries during terminal rendering. The specific process is as follows: Visualization scene configuration reading: The system reads the user configuration parameter: platform="web". Considering the limited rendering capabilities and network bandwidth of the web interface, the system sets the minimum visible geological body size to d_min=20m.

[0038] Critical region expansion: The system performs the following operations: (1) Merge the indicator body: I_key=logical_or(I_ore,I_fault) to obtain the initial key area.

[0039] (2) Calculate the radius of the structural element: Based on the grid spacing Δx=10m and d_min of the geophysical volume data, calculate the radius r of the spherical structural element r=ceil(d_min / (2*Δx))=ceil(20 / 20)=1 (voxel).

[0040] (3) Perform 3D dilation: Apply a spherical structuring element with a radius of r = 1 to I_key for 3D dilation to obtain the dilated key area I_key_dilated. This operation ensures that any real geological body smaller than d_min occupies at least one voxel in the mask, making it visible on the Web side.

[0041] Transition zone construction: To eliminate visual artifacts caused by subsequent partition compression, the system constructs an initial smooth transition zone (width w = 3Δx = 30m) between the key area and the background area, assigns initial labels to the transition zone based on the Euclidean distance transform, and provides a basic spatial division for subsequent tolerance interpolation: (1) Calculate the Euclidean distance transform D(x, y, z) of I_key_dilated; for the voxels at the boundary of the data domain (located at the grid edge of the resistivity volume V), use the boundary extrapolation method to supplement the virtual neighborhood (the virtual voxel is assigned the background area label 0) to ensure the accuracy of distance calculation; the distance value is reserved to two decimal places.

[0042] (2) Based on the requirements of geological structure continuity and the characteristics of grid resolution, define the transition zone width as w = 3 * Δx = 30m (this coefficient is determined by visual artifact tests under multiple different geological scenarios and can be adjusted within the range of 2 - 4 times Δx according to the actual data resolution and the complexity of the structure). For simple structures (number of faults ≤ 3, regular ore body morphology), use 2 times Δx; for medium structures (number of faults 3 - 8, relatively complex ore body morphology), use 3 times Δx; for complex structures (number of faults > 8, irregular ore body morphology), use 4 times Δx.

[0043] (3) Generate the final geological semantic mask M according to the following rules: If I_key_dilated(x, y, z) == 1, then M(x, y, z) = 1 (key area); If 0 < D(x, y, z) <= w, then M(x, y, z) = 2 (transition zone); Otherwise M(x, y, z) = 0 (background area).

[0044] Output the geological semantic mask: System output: Geological semantic mask M; Semantic definition table: {0: "background", 1: "key_zone", 2: "transition"}.

[0045] Step S3 includes: S31: Calculate the global standard deviation σ_global of the geophysical body data; S32: Set the compression tolerance τ according to the label of the voxel in the geological semantic mask and output the voxel-level compression tolerance field, as follows: If the label is a critical region, then τ = 0.05 × σ_global; If the label is a transition zone, then τ = 0.2 × σ_global; If the label is in the background area, then τ = 1.0 × σ_global; If the terminal is detected to be a mobile device and the network bandwidth is lower than the preset threshold, the compression tolerance of the background area will be adjusted to τ=2.0×σ_global; Wherein, the compression tolerance τ is the maximum absolute error allowed for lossy compression, and its physical unit is consistent with that of geophysical volume data; the global standard deviation σ_global is used to quantify the overall fluctuation of the volume data and serves as the benchmark scale for compression tolerance configuration.

[0046] As one embodiment, in a specific implementation of the present invention, step S3 is executed by the compression strategy configuration module, whose inputs are the aligned resistivity volume V generated in step S1 and the generated geological semantic mask M. After input, an association verification is performed first: the mean resistivity values ​​μ_key and μ_background of the key area (M=1) and background area (M=0) are statistically analyzed. If |μ_key-μ_background| / μ_background<0.1 (the difference in physical properties between the key area and the background area is too small), the user is prompted that 'the geological semantic mask has a low matching degree with the physical property data, and it is recommended to re-optimize the mask generation parameters'. After the verification passes, the tolerance field generation is then performed. This step realizes content-aware error allocation to ensure that key geological areas obtain high-fidelity compression. The specific process is as follows: Global standard deviation calculation: The system preprocesses the resistivity volume V for outliers (using the 3σ criterion to filter extreme values: removing voxels with values ​​exceeding [μ-3σ, μ+3σ], where μ is the mean of the original data and σ is the standard deviation of the original data), and then calculates the global standard deviation of the preprocessed data as the benchmark scale for compression tolerance: σ_global=std(V_processed)≈128.7Ω m.

[0047] Tolerance field generation: The system iterates through each voxel (x, y, z): (1) Query its label l=M(x,y,z) in the geological semantic mask M; (2) Based on the requirements for geological interpretation accuracy and visualization fidelity standards, the geological semantic-tolerance mapping rules are preset: the tolerance coefficient is determined by testing the fidelity of key area features (using the hausdorff distance to calculate the boundary deviation, ensuring that the boundary error is ≤2%, and setting the hausdorff distance threshold to 1 / 5 of the grid spacing) and testing the compression ratio of the background area (statistically calculating the ratio of the compressed file volume to the original data volume, with a target compression ratio ≥10:1). If l==1 (critical region), then τ=0.05×σ_global (boundary error ≤2%). If l==2 (transition zone), then τ=0.2×σ_global (to achieve a smooth transition in compressive strength); If l == 0 (background region), then τ = 1.0 × σ_global (compression ratio ≥ 10:1) Network environment adaptation: If the system detects that the client is a mobile device and the current network type is cellular network (4G / 5G) with bandwidth <2Mbps, or Wi-Fi network with bandwidth <1Mbps (Wi-Fi network is more stable, so the threshold can be lower), then the background area tolerance will be further relaxed and reset to τ=2.0×σ_global; at the same time, the network status will be recorded and updated every 5 seconds. If the bandwidth recovers to above the threshold and remains stable for 10 seconds, then the original tolerance setting will be restored.

[0048] Output compression parameter field: The system outputs a complete voxel-level tolerance field τ, which has the exact same dimensionality and spatial alignment as the original data V, and will serve as the direct input for the next step of block compression. The aforementioned τ will be used as the input for step S4.

[0049] Step S4 includes: The buffering process includes: Perform isotropic expansion on the critical region boundaries to construct a buffer zone; Within the buffer and transition zone, tolerant linear interpolation is performed based on the distance from the voxel to the critical region boundary. The interpolation formula is as follows:

[0050] in, This is the tolerance after interpolation. For background area tolerance, Let be the critical region tolerance, and d be the minimum Euclidean distance from the voxel to the critical region boundary. The width of the buffer; d and The units are all meters; The value is 2 to 4 times the grid spacing, and is dynamically adjusted according to the complexity of the geological structure. The block compression includes: The input data is divided into fixed-size three-dimensional data blocks, and the compression tolerance within a block is taken as the maximum tolerance value within the block; if there is no effective tolerance data within a block, the default tolerance of the background area is used. Call the ZFP compression library to perform lossy compression in precision mode, and record the compressed byte stream and offset address; The block size is determined based on the target terminal's memory usage and network transmission efficiency constraints. The memory usage of a single block after decompression is ≤2MB, and the size of a single block after compression is ≤200KB.

[0051] In one specific embodiment of the present invention, step S4 is executed by the adaptive compression engine module, whose inputs are the generated aligned resistivity volume V and the generated voxel-level compression tolerance field τ. The specific process is as follows: Boundary buffering: To eliminate visual artifacts caused by abrupt changes in compressive strength, the system performs isotropic dilation of three voxels (matching the width of the transition zone) on the boundary of the critical area based on the transition zone range of the geological semantic mask M, resulting in a buffer zone B. For voxels within the transition zone and buffer zone, the tolerance τ(x,y,z) is linearly interpolated according to its distance d to the nearest critical area boundary to generate a buffered tolerance field, further optimizing the smoothness of the compressive strength transition.

[0052] Blocking and Compression: Based on the memory usage limit of the target terminal (Web / mobile) (memory after single-block decompression ≤ 1MB) and network transmission efficiency (volume after single-block compression ≤ 100KB), combined with the original data grid spacing (Δx=Δy=Δz=10m), the system divides V and τ_buffered into 8×8×8 blocks (single block space range 80m×80m×80m, a total of 262,144 blocks). The block size can be adjusted according to the data resolution in integer multiples of Δx. The adjustment constraints are: memory after single-block decompression ≤ 2MB (maximum block size ≤ 16×16×16 voxels), volume after single-block compression ≤ 200KB (minimum block size ≥ 4×4×4 voxels), ensuring a balance between terminal processing and network transmission efficiency.

[0053] For each block b: First record the block metadata (spatial bounding box, dominant semantic tag) (The label frequency of the geological semantic mask M within the statistical block, taking the highest frequency label), LOD level (based on...) Confirm); then calculate the maximum tolerance within the block; if there is no valid tolerance data within the block (all If the value is NULL or exceeds a reasonable range [0, 10 * σ_global], then the default tolerance of the background area τ = 1.0 * σ_global is used as the compression tolerance for this block; the ZFP library function is called for compression; finally, the compressed byte stream and its offset address in compressed.dat are recorded.

[0054] Block metadata records: For each block b, the system records: Spatial bounding box (e.g., [100,108]×[200,208]×[50,58]); Dominant semantic tags (The M tag that appears most frequently within the block); LOD Level: The LOD level is associated with semantic tags and corresponds to a specific resolution scaling ratio: if If the critical area is 1, then the LOD is 3 (original resolution, no scaling). If the value is 2 (transition zone), then LOD = 2 (resolution scaled to 1 / 2); otherwise (background area) LOD = 1 (resolution scaled to 1 / 4). The scaling method is voxel downsampling averaging.

[0055] Output the compressed data file compressed.dat; a list of block metadata (JSON format).

[0056] Step S5 includes: The spatial-semantic-value multilevel index includes: Octree indexes are used for space-pruning queries; A hash index, using semantic tags and LOD levels as keys, is used for semantic priority loading; B+ tree indexes, using the range of physical property values ​​as keys, are used for filtering queries based on physical properties.

[0057] As one embodiment, the B+ tree index organizes the property value range into a B+ tree structure by using the property value range as the key, supporting efficient range query operations; the hash index and the B+ tree index together constitute a multi-level index system, taking into account the performance requirements of single-point access and range lookup.

[0058] As one implementation example, the octree is constructed as follows: The system initializes the octree root node (covering the entire data domain, i.e., 0–5119m), and recursively inserts the bounding boxes of each block (generated by S4-3) into the tree. To balance query efficiency and memory usage, the maximum capacity of leaf nodes is set to 64 blocks. This index is used to quickly respond to space pruning requests such as query_by_frustum.

[0059] Value range and semantic index construction: Calculate the range of physical property values: During the block division process in step S4, the voxels corresponding to each block b in the original resistivity volume V are traversed synchronously, and the minimum value v_min_b=min(Vinblockb) and the maximum value v_max_b=max(Vinblockb) are calculated and recorded together with the block data; subsequently, based on the recorded range of physical property values, a B+ tree value range index is constructed.

[0060] Construct a value range index: Use [v_min_b,v_max_b] as the key to construct a B+ tree, which supports efficient query of property value range (such as query_by_value_range(100,200)).

[0061] Construct a semantic index: Use (l_b, LOD_b) as the key (where l_b is the dominant geological semantic label and LOD_b is the LOD level), construct a hash table, and use the corresponding block ID list as the value. Support efficient semantic priority loading (such as query_by_semantic_label(l=1, LOD=3)).

[0062] As one example, index serialization: The system serializes the three index structures—octree, B+ tree, and hash table—into binary format according to the Google Protocol Buffers (Protobuf) standard (ensuring cross-platform compatibility) and merges them into a single index file, index.dat, along with the Protobuf data structure definition file, index.proto.

[0063] As one example, the output index file is: the system outputs index.dat and registers it with an object storage service (such as AWS S3 or MinIO) for clients to download and use.

[0064] Step S6 includes: The priority-based loading includes: Obtain the set of visible blocks based on the view frustum; Obtain the semantic attention block set based on the user's interaction intent; After merging the set of visible blocks and the set of semantically concerned blocks, sort them according to the priority formula:

[0065] in, The loading priority of block b For the detail level of block b, As the dominant semantic tag, For LOD weighting coefficients, These are semantic weight coefficients; and This is a configurable parameter, with a value range of 0.5 to 1.5, and is dynamically adjusted by the user interaction mode. A larger value indicates a higher priority for spatial precision. A larger value indicates a higher priority in semantic concern; when , ,when , ; The differentiated rendering includes: Phong shading is enabled for the critical region, with lighting model parameters including ambient light intensity 0.3, diffuse reflection coefficient 0.6, specular intensity 0.8, and specular index 32. The transition zone uses basic diffuse coloring; The background area uses downsampling rendering and fast coloring; If the cumulative loading and rendering time exceeds the preset threshold, a degraded loading strategy will be automatically triggered.

[0066] As one example, the rendering strategy is linked to the LOD level; the higher the LOD level, the more detailed the lighting model. The LOD level is associated with semantic tags: the critical area corresponds to LOD=3 (original resolution), the transition zone corresponds to LOD=2 (resolution scaled to 1 / 2), and the background area corresponds to LOD=1 (resolution scaled to 1 / 4). If the cumulative loading and rendering time exceeds a preset threshold, a downgrade loading strategy is automatically triggered, using a lower LOD version for subsequent data blocks to load and render, prioritizing the interactive frame rate. The preset threshold is 2 seconds, and the target interactive frame rate is ≥30 FPS.

[0067] As one example, the client downloads index.dat via HTTPS (time = 620ms). After downloading, the file integrity is verified (based on MD5 checksum comparison, the server pre-generates and stores index.dat.md5). If the verification passes, the file is parsed into memory to build a runtime query structure of an octree, B+ tree, and hash table. If the verification fails, the download request is re-initiated (maximum of 3 retries). If the retries fail, the user is prompted with 'Index file loading failed, please check network connection'.

[0068] Block filtering and sorting: Spatial clipping: Call query_by_frustum(camera_matrix) to obtain the set of visible blocks B_visible within the current view frustum using the octree index.

[0069] Semantic priority loading: Because the user has enabled "mineral body focus mode" (an interactive option), query_by_semantic_label(l=1,LOD=3) is called additionally to obtain the high-fidelity key block set B_focus.

[0070] Merge and sort: Merge B_candidate = B_visible ∪ B_focus, and sort according to the following priority formula: Streaming decompression and rendering: Clients execute according to priority order: Request data: Initiate an HTTP Range Request to the server to request the byte stream of compressed block b.

[0071] Local decompression: Call the ZFPWASM module to decompress the byte stream into raw floating-point data and upload it to the WebGL3D texture.

[0072] Differentiated rendering and LOD level linkage: critical area ( ==1) If LOD=3, enable Phong shading (ambient light intensity 0.3, diffuse index 0.6, specular intensity 0.8, specular index 32); if LOD=2, enable simplified Phong shading (disable specular reflection, ambient light intensity 0.3, diffuse index 0.7); if LOD=1 (1 / 4 resolution), use basic diffuse shading; transition band ( ==2) Regardless of LOD level, basic diffuse shading is used; background area ( ==0) Uses downsampling rendering + fast shading to ensure that rendering performance matches the detail display.

[0073] Time budget control: If the cumulative loading and rendering time exceeds a preset threshold (e.g., 2 seconds), a degraded loading strategy will be automatically triggered: For subsequent data blocks, regardless of their original priority, the lower-level LOD version will be used for loading and rendering to prioritize the smoothness of system interaction.

[0074] Output visualization results: The client successfully built and maintained an interactive 3D visualization scene with a stable frame rate of ≥32FPS. Users can perform operations such as rotation, scaling, and switching display modes in real time.

[0075] This application also discloses an electronic device. (See reference...) Figure 2 , Figure 2 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. The electronic device 500 may include: at least one processor 501, at least one network interface 504, a user interface 503, a memory 505, and at least one communication bus 502.

[0076] The communication bus 502 is used to enable communication between these components.

[0077] The user interface 503 may include a display screen, and optionally, the user interface 503 may also include a standard wired interface or a wireless interface.

[0078] The network interface 504 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0079] This application also discloses a computer-readable storage medium storing multiple instructions adapted for loading by a processor to execute the above-described efficient and visual adaptive compression method for geological-geophysical volume data.

[0080] The above are merely exemplary embodiments of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure.

[0081] This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. An efficient and visual adaptive compression method for geological-geophysical body data, characterized in that, The method includes the following steps: S1: Acquire multi-source geological interpretation prior data and geophysical volume data, perform coordinate system unification and rasterization processing, and generate spatially aligned semantic indicator volumes; S2: Based on semantic indicators, and integrating the display capabilities of the target visualization platform, a geological semantic mask containing key areas, transition zones, and background areas is generated; S3: Based on the geological semantic mask, a compression tolerance is assigned to each voxel to form a voxel-level compression tolerance field; S4: Buffer the critical region boundary, smooth the voxel-level compression tolerance field, and compress it in blocks, recording block metadata. S5: Based on block metadata, construct a spatial-semantic-value multi-level index; S6: Based on spatial-semantic-value multi-level indexing, combined with view frustum and user interaction intent, load, decompress and render data blocks in a priority-based streaming manner.

2. The efficient and visual adaptive compression method for geological-geophysical body data as described in claim 1, characterized in that, Step S1 includes: The geophysical volume data includes: resistivity volume, density volume, and magnetic susceptibility volume; the multi-source geological interpretation prior data includes: borehole data, three-dimensional orebody models, and fault vector data.

3. The efficient and visual adaptive compression method for geological-geophysical body data as described in claim 1, characterized in that, Step S2 includes: S21: Read the display parameters of the target visualization platform and set the minimum visible geological body size; S22: Based on the minimum visible geological volume size, perform three-dimensional morphological expansion on the semantic indicator volume of the key area to ensure that small-scale geological volumes are visible on the terminal. S23: Construct a transition zone based on Euclidean distance transformation to generate a geological semantic mask with three types of labels: key area, transition zone, and background area.

4. The efficient and visual adaptive compression method for geological-geophysical body data as described in claim 1, characterized in that, Step S3 includes: S31: Calculate the global standard deviation σ_global of geophysical volume data; S32: Based on the voxel labels in the geological semantic mask, set the compression tolerance τ and output the voxel-level compression tolerance field as follows: If the label is a critical region, then τ = 0.05 × σ_global; If the label is a transition zone, then τ = 0.2 × σ_global; If the label is in the background area, then τ = 1.0 × σ_global; If the terminal is detected to be a mobile device and the network bandwidth is lower than the preset threshold, the compression tolerance of the background area will be adjusted to τ=2.0×σ_global; Wherein, the compression tolerance τ is the maximum absolute error allowed for lossy compression, and its physical unit is consistent with that of geophysical volume data; the global standard deviation σ_global is used to quantify the overall fluctuation of the volume data and serves as the benchmark scale for compression tolerance configuration.

5. The efficient and visual adaptive compression method for geological-geophysical body data as described in claim 1, characterized in that, Step S4 includes: The buffering process includes: Perform isotropic expansion on the critical region boundaries to construct a buffer zone; Within the buffer and transition zone, tolerant linear interpolation is performed based on the distance from the voxel to the critical region boundary. The interpolation formula is as follows: in, This is the tolerance after interpolation. For background area tolerance, Let be the critical region tolerance, and d be the minimum Euclidean distance from the voxel to the critical region boundary. The width of the buffer; d and The units are all meters; The value is 2 to 4 times the grid spacing, and is dynamically adjusted according to the complexity of the geological structure. The block compression includes: The input data is divided into fixed-size three-dimensional data blocks, and the compression tolerance within a block is taken as the maximum tolerance value within the block; if there is no effective tolerance data within a block, the default tolerance of the background area is used. Call the ZFP compression library to perform lossy compression in precision mode, and record the compressed byte stream and offset address; The block size is determined based on the target terminal's memory usage and network transmission efficiency constraints. The memory usage of a single block after decompression is ≤2MB, and the size of a single block after compression is ≤200KB.

6. The efficient and visual adaptive compression method for geological-geophysical body data as described in claim 1, characterized in that, Step S5 includes: The spatial-semantic-value multilevel index includes: Octree indexes are used for space-pruning queries; A hash index, using semantic tags and LOD levels as keys, is used for semantic priority loading; B+ tree indexes use the range of physical property values ​​as keys to filter queries based on physical properties.

7. The efficient and visual adaptive compression method for geological-geophysical body data as described in claim 1, characterized in that, Step S6 includes: The priority-based loading includes: Obtain the set of visible blocks based on the view frustum; Obtain the semantic attention block set based on the user's interaction intent; After merging the set of visible blocks and the set of semantically concerned blocks, sort them according to the priority formula: in, The loading priority of block b For the detail level of block b, As the dominant semantic tag, This is the LOD weighting coefficient. These are semantic weight coefficients; and This is a configurable parameter, with a value range of 0.5 to 1.5, and is dynamically adjusted by the user interaction mode. A larger value indicates a higher priority for spatial precision. A larger value indicates a higher priority in semantic concern; when , ,when , ; The differentiated rendering includes: Phong shading is enabled for the critical region, with lighting model parameters including ambient light intensity 0.3, diffuse reflection coefficient 0.6, specular intensity 0.8, and specular index 32. The transition zone uses basic diffuse coloring; The background area uses downsampling rendering and fast coloring; If the cumulative loading and rendering time exceeds the preset threshold, a degraded loading strategy will be automatically triggered.

8. An electronic device, characterized in that, The device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to enable the electronic device to perform the efficient and visual adaptive compression method for geological-geophysical body data as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed by a computer, perform the efficient visualization adaptive compression method for geological-geophysical body data as described in any one of claims 1-7.

Images