Physical property information fusion analysis method based on three-dimensional digital technology

Abstract

The invention relates to the technical field of three-dimensional digitization, in particular to a physical property information fusion analysis method based on the three-dimensional digitization technology, and the method comprises the steps: obtaining multi-source physical property data from different sources of geological exploration, rock and soil physical testing and seismic exploration, and carrying out format standardization processing to obtain a preprocessed data set; performing time sequence alignment on the preprocessed data to generate a time-space synchronization data sequence; performing cross-modal fusion and noise reduction processing based on the time-space synchronization data sequence to obtain multi-modal fusion data; and finally, inputting the multi-modal fusion data into a three-dimensional modeling module to generate a three-dimensional physical property information model, and performing visualization processing to obtain a three-dimensional display result. By accurately synchronizing and fusing the data of various sensors, the method overcomes the differences of different modal data in time, space and resolution, realizes seamless integration of physical property information in a three-dimensional space, and has a wide application prospect.

Description

Technical Field

[0001] The present invention relates to three-dimensional digitization technology and its application in physical property information analysis, and particularly to a method for fusing and analyzing physical property information based on three-dimensional digitization technology. Background Art

[0002] In many fields, such as geological exploration, geotechnical physical testing, and seismic surveying, the acquisition depends on a variety of different types of sensors. When the sensors acquire data, they not only face differences in resolution and accuracy, but also, due to the heterogeneity of data sources, there are problems of different timestamps and inconsistent spatial coordinate systems. How to effectively fuse the physical property data from different sensors to ensure the unity of data in time, space, and modality has become a key technical challenge in data processing.

[0003] In the existing technologies, it usually relies on manual calibration, manual alignment, or data processing for a specific type of sensor, but these methods have problems of insufficient accuracy and low efficiency when dealing with complex and multi-modal physical property data. In addition, the existing three-dimensional modeling methods usually cannot handle the precise fusion of multi-modal data, and it is difficult to simultaneously achieve high-precision presentation of physical property information and real-time dynamic adjustment, so they cannot meet the efficient data fusion requirements under the cooperation of multi-sensors.

[0004] Therefore, how to solve the inconsistency of data in space and time through precise data synchronization, cross-modal fusion, and noise reduction processing, and achieve seamless integration and visual display of physical property information in three-dimensional space is a difficult problem that needs to be solved urgently in the current technology. Summary of the Invention

[0005] The present invention provides a method for fusing and analyzing physical property information based on three-dimensional digitization technology to solve the problem of how to effectively fuse the physical property data from different sensors, solve the inconsistency of data in time, space, and modality, and achieve precise integration and visual display of physical property information in three-dimensional space.

[0006] To solve the above technical problems, the present invention provides a method for fusing and analyzing physical property information based on three-dimensional digitization technology, including:

[0007] Obtain multi-source physical property data from different sources such as geological exploration, geotechnical physical testing, and seismic surveying, and perform format standardization processing to obtain a preprocessed data set;

[0008] Perform time series alignment on the preprocessed data set to generate a spatio-temporal synchronization data sequence;

[0009] Based on the spatio-temporal synchronization data sequence, perform cross-modal fusion and noise reduction processing to obtain multi-modal fusion data;

[0010] Input the multi-modal fusion data into a 3D modeling module to generate a 3D physical property information model, and perform visualization processing to obtain a 3D display result.

[0011] Further, the step of performing time series alignment on the preprocessed data set specifically includes:

[0012] Calibrate the preprocessed data set according to timestamps to align the time information from different sensors, and generate a spatio-temporal synchronization data sequence.

[0013] Further, in the time series alignment step, it further includes:

[0014] Adjust the time alignment of each data point according to the spatial coordinate information of the sensor.

[0015] Further, the step of performing cross-modal fusion and noise reduction processing based on the spatio-temporal synchronization data sequence specifically includes:

[0016] Group the physical property data of different modalities, and use a cross-modal fusion algorithm to effectively fuse different types of data to obtain denoised multi-modal fusion data.

[0017] Further, the cross-modal fusion and noise reduction processing step specifically includes:

[0018] Use a filtering algorithm to remove noise in the multi-modal data, and at the same time perform interpolation filling on the missing values in the data to generate a multi-modal fusion data set.

[0019] Further, the step of inputting the multi-modal fusion data into a 3D modeling module specifically includes:

[0020] Input the denoised multi-modal data into a 3D modeling algorithm to generate a physical property information model in 3D space, and the physical property information model includes the structure and material information of different objects.

[0021] Further, the step of generating a 3D physical property information model and performing visualization processing specifically includes:

[0022] Render the generated 3D physical property information model, and generate a corresponding 3D display result according to the set viewing angle and resolution.

[0023] Further, the step of generating a physical property information model in the 3D modeling module specifically includes:

[0024] Based on the spatial coordinates and physical property data in the spatio-temporal synchronization data sequence, construct a 3D geometric model of the object, and assign corresponding physical property attributes to the model.

[0025] Further, the denoising processing step includes:

[0026] Using wavelet transform or Kalman filter algorithm to denoise data of different modalities, eliminating high-frequency noise in the data, and obtaining smoother and more accurate multi-modal fusion data.

[0027] Further, a physical property information fusion analysis system based on three-dimensional digitization technology includes:

[0028] A data acquisition module for obtaining physical property data from different sensors;

[0029] A data preprocessing module for performing format standardization and noise removal processing on the physical property data to obtain a preprocessed data set;

[0030] A spatio-temporal synchronization and alignment module for performing time series calibration and spatial alignment on the preprocessed data to generate a spatio-temporal synchronized data sequence;

[0031] A cross-modal data fusion and denoising module for performing cross-modal fusion and denoising processing on the spatio-temporal synchronized data to obtain multi-modal fusion data;

[0032] A three-dimensional modeling and visualization module for generating a three-dimensional physical property information model based on the multi-modal fusion data and performing visualization processing to obtain a three-dimensional display result.

[0033] The key innovation points of the present invention include:

[0034] (1) Spatio-temporal synchronization processing: By calibrating timestamps and aligning spatial coordinates, it ensures that data from different sensors can be accurately synchronized and unified into the same spatio-temporal reference system, solving the problems of time and space inconsistency in traditional methods.

[0035] (2) Cross-modal data fusion and denoising: Using cross-modal data fusion technology, effectively integrating data from different sensors, and using a denoising algorithm to remove noise in the data, improving the fusion effect and overall accuracy of multi-modal data.

[0036] (3) Three-dimensional modeling and visualization: Inputting the fused data into a three-dimensional modeling module to generate an accurate three-dimensional physical property information model and performing visualization processing, which is convenient for users to intuitively display and analyze complex physical property information, greatly enhancing the operability and application value of the data.

[0037] The following are its main beneficial effects:

[0038] The present invention provides a physical property information fusion analysis method based on three-dimensional digital technology. Through spatio-temporal synchronization, cross-modal fusion, and noise reduction processing of multi-source data, the consistency of physical property data collected by different sensors in terms of time, space, and resolution is achieved, avoiding the processing errors caused by inconsistent multi-source data in traditional methods. Compared with traditional data processing methods, the present invention can more accurately integrate physical property data such as density, magnetic susceptibility, porosity, and elastic modulus from rocks, eliminate interference between modalities, and intuitively display physical property information through three-dimensional modeling and visualization technology, greatly improving the accuracy and usability of data. In addition, the present invention improves the processing speed and accuracy through efficient data fusion algorithms and noise reduction processing, and particularly significantly enhances the real-time data processing ability in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic flowchart of a physical property information fusion analysis method based on three-dimensional digital technology provided by an embodiment of the present application;

[0040] Figure 2 It is a structural block diagram of a physical property information fusion analysis system based on three-dimensional digital technology provided by an embodiment of the present application; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs; the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive inclusion. The terms "first" and "second" in the specification and claims of this application or the above drawings are used to distinguish different objects and not to describe a specific order.

[0042] Reference to "embodiment" herein means that a specific feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of this application. The phrase appears at various positions in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0043] Embodiment 1: Refer to Figure 1 , which is a schematic flowchart of a physical property information fusion analysis method based on three-dimensional digital technology provided by an embodiment of the present invention. The process can at least include steps S100 - S400:

[0044] S100. Obtain multi-source physical property data from different sources such as geological exploration, geotechnical physical testing, and seismic surveys, perform format standardization processing, and obtain a preprocessed data set;

[0045] S200. Align the preprocessed data set in time series to generate a spatio-temporal synchronous data sequence;

[0046] S300. Based on the spatio-temporal synchronous data sequence, perform cross-modal fusion and noise reduction processing to obtain multi-modal fusion data;

[0047] S400. Input the multi-modal fusion data into a 3D modeling module, generate a 3D physical property information model, and perform visualization processing to obtain a 3D display result.

[0048] Step S100 at least includes steps S110 - S130:

[0049] S110: Obtain multi-source physical property data from different sources such as geological exploration, geotechnical physical testing, and seismic surveys, and perform preliminary classification according to the data source and format.

[0050] In step S110, the system first obtains geological and physical property data such as the density, magnetic susceptibility, porosity, and elastic modulus of rocks from different sources such as geological exploration, geotechnical physical testing, and seismic surveys. Set the original data sets obtained from different sensors as:

[0051] D = {D CT , D Laser , D US , D MRI}

[0052] Among them, D CT represents the density data of the rock. D Laser represents the physical property data of the magnetic susceptibility. D US represents the physical property data of the porosity. D MRI represents the physical property data of the elastic modulus.

[0053] The obtained data has different formats and units. The system classifies it according to the data source and format by sensor type to form a grouped data set Each data set retains its original accuracy and resolution information for standardization in the next step.

[0054] S120. Perform a format conversion operation on the classified data.

[0055] In step S120, perform format conversion on the classified data set D Group in step S110 to standardize the data of various sensors. Let the physical property data set in the unified standard format be DNorm For each data set (such as ) apply the normalization formula:

[0056]

[0057] where D Norm,i represents the normalized data, μ i and σ i represent the mean and standard deviation of the data set D i G respectively. This process ensures that all sensor data is converted into a unified format D Norm ={D Norm,CT , D Norm,Laser , D Norm,US , D Norm,MRI} for subsequent fusion analysis.

[0058] S130: Preprocess the normalized data to obtain a preliminary cleaned data set.

[0059] Step S130 performs further preprocessing based on the normalized data D Norm generated by S120. First, use a noise filtering algorithm to denoise the data in D Norm . Let the denoised data be D Filt , and use the noise filtering function f Filter to process each data point:

[0060] D Filt,i = f Filter (D Norm,i )

[0061] where D Filt,i is the i-th type of denoised data. For the data set with missing values remaining after denoising, use the linear interpolation method f Interp to fill in the missing data and generate the final preprocessed data set D Pre :

[0062] D Pre,i = f Interp (D Filt,i )

[0063] At this time, the preprocessed D Pre will be used as the input data for spatio-temporal alignment in the next step (S200 module) to ensure that each data point has unified format, noise-free and complete numerical information.

[0064] Step S200 includes at least steps S210 - S230:

[0065] S210: Calibrate the timestamps of the preprocessed dataset, arrange the data from different sensors in time series, and generate preliminary time-aligned data.

[0066] In step S210, based on the preprocessed dataset D Pre ={D Pre,CT , D Pre,Laser , D Pre,US , D Pre,MRI} generated in the previous step S130, calibrate the timestamp information in each dataset. Let t CT , t Laser , t US , t MRI represent the timestamp vectors of the data collected by each sensor respectively. For each sensor dataset D Pre,i (such as D Pre,CT ), adopt the time offset correction formula:

[0067] t calibrated,i =t i +Δt i

[0068] where t calibrated,i is the calibrated timestamp, t i is the original timestamp vector, and Δt i represents the offset of this sensor from the reference time. By calibrating the timestamps of all data, obtain the calibrated preliminary time-aligned dataset:

[0069] D TimeAlign ={D TimeAlign,CT , D TimeAlign,Laser , D TimeAlign,US , D TimeAlign,MRI}.

[0070] S220: Align the positions of the data from different sensors based on the spatial coordinate information, and map each data into the same spatial coordinate system.

[0071] In step S220, based on the time-aligned data D TimeAlign calibrated in S210, use the spatial coordinate information in each dataset to align the positions of the data from different sensors. Set a unified three-dimensional spatial coordinate system (x, y, z), and map the spatial coordinates (x i , y i , z i ) of each dataset into this unified coordinate system. The specific mapping relationship is:

[0072] (x aligned,i , y aligned,i , z aligned,i ) = R i ·(xi , y i , z i ) + T i

[0073] Where: R i is a rotation matrix used to adjust the coordinate system direction of dataset D i ; T i is a translation vector used to adjust the origin of dataset D i to the origin of the unified coordinate system.

[0074] After completing the position alignment, an aligned dataset D SpaceAlign = {D SpaceAlign,CT , D SpaceAlign,Laser , D SpaceAlign,US , D SpaceAlign,MRI} is obtained, and the data from different sensors are consistent in space.

[0075] S230: On the basis of time and space alignment, construct a unified spatio-temporal data sequence to form a spatio-temporal synchronization data sequence.

[0076] In step S230, perform spatio-temporal integration on the spatially aligned data D SpaceAlign generated in step S220. Combine the calibration timestamp t calibrated,i of each sensor with the spatial data (x aligned,i , y aligned,i , z aligned,i ) in the unified spatial coordinate system to form the spatio-temporal coordinates of each data point:

[0077] D SpatioTempAlign,i = (t calibrated,i , x aligned,i , y aligned,i , z aligned,i )

[0078] Summarize the data of all sensors to form a unified spatio-temporal data sequence: D SpatioTempAlign

[0079] D SpatioTempAlign = {D SpatioTempAlign,CT , D SpatioTempAlign,Laser , D SpatioTempAlign,US , D SpatioTempAlign,MRI}

[0080] At this time, the obtained D SpatioTempAlign will be used as the input data for the cross-modal fusion in the next step S300 module to ensure complete alignment of the data in time and space for subsequent multi-modal fusion analysis.

[0081] Step S300 includes at least steps S310 - S330:

[0082] S310: Group the data across modalities according to the data modality type based on the spatio-temporal synchronization data sequence.

[0083] In step S310, based on the spatio-temporal synchronization data sequence D generated in step S230 SpatioTempAlign , the system groups the data from different sources according to the modality type. Specifically, let D SpatioTempAlign contain the spatio-temporal synchronization data of density, magnetic susceptibility, porosity, and elastic modulus from rocks. According to the data modality grouping, an image data set D Image and a point cloud data set D PointCloud are generated, where:

[0084] D Image = {D SpatioTempAlign,CT , D SpatioTempAlign,MRI}, representing the image data modality.

[0085] D PointCloud = {D SpatioTempAlign,Laser , D SpatioTempAlign,US}, representing the point cloud data modality.

[0086] S320: Perform multi-scale fusion processing on the cross-modal grouped data, integrate the data with different resolutions and scales in the same modality, and form multi-modal fusion data.

[0087] In step S320, perform multi-scale fusion on the cross-modal data sets D Image and D PointCloud generated in S310.

[0088] Specifically, perform weighted fusion on the data from different sources. Based on algorithms such as weighted average and least squares method, combine the physical property data collected by each sensor to generate a fused multi-modal data set D Fused .

[0089] For different modalities between the data, adopt the fusion method of principal component analysis (PCA) to effectively integrate different physical property data in the same space.

[0090] Further, the implementation method is as follows:

[0091] ① Perform weighted average on the physical property data from different sources. The calculation formula is as follows:

[0092]

[0093] where, D Fused is the finally fused physical property data, D i is the physical property data collected by the i-th sensor (modality), and w iis the weight of the i-th data source, satisfying and n represents the number of data modalities.

[0094] For the physical property data D collected by each sensor i , the data can be fitted by the least squares method to minimize the error and obtain the optimal weighting coefficients. The objective of the least squares method is:

[0095]

[0096] where D i is the data collected by the i-th sensor, is the result after weighted fusion. By this method, the weighting coefficients of each sensor data can be obtained to ensure the minimum error during data fusion.

[0097] ② Dimensionality reduction by principal component analysis (PCA).

[0098] First, standardize the data of each sensor so that the mean of each physical property data is zero and the variance is 1.

[0099] Calculate the covariance matrix C of the standardized data, which describes the correlation between each physical property data. The calculation formula of the covariance matrix is:

[0100]

[0101] where X i is the i-th sample, μ is the mean of the data, and n2 is the number of samples.

[0102] Perform eigenvalue decomposition on the covariance matrix to obtain eigenvalues and eigenvectors. The eigenvectors represent different principal component directions in the data, and the eigenvalues represent the importance of each principal component.

[0103] Select the principal components according to the magnitudes of the eigenvalues. Select the first k principal components with the largest eigenvalues and construct a new data matrix to reduce the data dimension while retaining most of the data variability.

[0104] Project the original data into the selected principal component space to generate a low-dimensional, multi-modal fusion dataset D Fused .

[0105] Through PCA, data from different sources and modalities can be effectively integrated into a unified space, reducing data redundancy while retaining the main features of the data.

[0106] ③ Combine the accuracy and reliability of each sensor data and use a weighted algorithm to adjust the contribution of each data.

[0107] Signal-to-Noise Ratio (SNR) Weighting: Calculate the signal-to-noise ratio (SNR) for the data of each sensor, and then assign higher weights to the data with higher SNR. The formula for calculating the SNR is:

[0108]

[0109] where μ i is the mean of the data of the i-th sensor, and σ i is the standard deviation of this data. A higher SNR indicates higher data quality, and a larger weight is assigned to it. Determine the weights based on the calibration accuracy of the sensors. Data sources with higher accuracy are assigned larger weights, and data sources with lower accuracy have smaller weights.

[0110] These weights can be applied to data fusion through the weighted average method to finally generate the precision-adjusted multi-modal fusion dataset D Fused .

[0111] For the image data modality D Image , let the resolution parameters of the image data be R CT and R MRI . First, perform resolution matching to make the two reach the unified resolution R Image , specifically:

[0112] R Image = max(R CT , R MRI )

[0113] After unifying the image data to R Image , merge the image data to form the fused image modality data D Image,fused . Similarly, for the point cloud data modality D PointCloud , match the resolutions of the laser scanning and ultrasonic data to the unified resolution R PointCloud to obtain the fused point cloud data modality D PointCloud,fused .

[0114] Through multi-scale fusion, form the final multi-modal fusion data D Fused ={D Image,fused , D PointCloud,fused}, providing a basis for the next noise reduction process.

[0115] S330: Perform multi-dimensional noise reduction processing on the multi-modal fusion data to eliminate interference between modalities and obtain the denoised multi-modal fusion data.

[0116] In step S330, perform multi-dimensional noise reduction processing based on the multi-modal fusion data D Fused generated in step S320. Let the noise reduction algorithm be f denoise , for the fused image data DImage,fused and the point cloud data D PointCloud,fused Independently perform a noise reduction operation:

[0117] D Image,denoised = f denoise (D Image,fused )

[0118] D PointCloud,denoised = f denoise (D PointCloud,fused )

[0119] where D Image,denoised and D PointCloud,denoised are the denoised image data and point cloud data, completing the elimination of interference between modalities. Finally, generate the denoised multi-modal fusion data D denoised = {D Image,denoised , D PointCloud,denoised}, and this dataset will be used as the input for 3D modeling and visualization of the S400 module to ensure the clarity and consistency of the data.

[0120] Step S400 includes at least steps S410 - S430:

[0121] S410: Input the denoised multi-modal fusion data into the 3D modeling module to perform comprehensive 3D modeling of physical property information.

[0122] In step S410, use the denoised multi-modal fusion data D denoised = {D Image,denoised , D PointCloud,denoised} generated in step S330 to perform 3D modeling processing on physical property information. Specifically, jointly model the image data and point cloud data to construct a 3D model with a unified coordinate system. Assume the coordinates of the image data in 3D space are x Image , y Image , z Image , and the coordinates of the point cloud data are x PointCloud , y PointCloud , z PointCloud . Perform interpolation processing on the physical property information of each data point to generate continuous 3D model data M 3D :

[0123] M 3D = f interpolate (D Image,denoised , D PointCloud,denoised )

[0124] where f interpolate is the interpolation function that fuses discrete data of different modalities into the continuous 3D model M 3D , and this model contains unified physical property information and spatial distribution data.

[0125] S420: Based on the constructed 3D model, render the model according to the set viewing angle and resolution to generate a visual 3D physical property information model.

[0126] In step S420, perform a rendering process on the 3D model M generated in step S410. 3D Set the rendering viewing angle parameter θ and the resolution parameter M Render , and apply the rendering function f render to process the data in the 3D model:

[0127] M Render = f render (M 3D , θ, R Render )

[0128] where R Render represents the rendered 3D model. The rendering process converts 3D space information into a 2D image or multi-angle views, and by setting different viewing angles θ and resolutions R Render , generates 3D physical property model images with different clarity and levels of detail.

[0129] Use the constructed 3D physical property information model for data analysis, including but not limited to spatial distribution analysis of physical property data, trend analysis of changes, and analysis of differences in physical property characteristics in different geological regions, etc.

[0130] ① Spatial distribution analysis. Label the physical property data in the constructed 3D physical property information model M 3D according to the 3D coordinates (x, y, z).

[0131] Through the spatial positions in the 3D coordinate system, conduct a spatial distribution analysis of the physical property data to identify the distribution trends of the physical property data in different geological regions.

[0132] Use the spatial interpolation algorithm, Kriging interpolation method, to expand the sparse physical property data points into continuous data for the entire geological volume and generate corresponding physical property distribution maps in 3D space.

[0133] According to the distribution of the physical property data in 3D space, draw isosurfaces or contour maps to display the spatial distribution of physical property characteristics.

[0134] ② Trend analysis of changes. For each physical property data point (x i , y i , z i ) in the model, extract the time series data. Use time series analysis methods, such as the moving average method or the exponential smoothing method, to smooth the physical property data and reduce the impact of noise on data analysis.

[0135] Analyze the changing trends of physical property data over different time periods by fitting curves or trend lines.

[0136] Mark the regions where physical property data changes within each time period in the 3D model, and display the trend of physical property data changing over time through a dynamic view.

[0137] ③ Analysis of differences in physical property characteristics of different geological regions. According to the spatial coordinates in the 3D physical property information model, divide different geological regions (e.g., bedrock area, sedimentary layer area, fault area, etc.). Conduct statistical analysis on the physical property data within each geological region, and calculate indicators such as the mean, variance, and standard deviation of the physical property data in each region to describe the physical property characteristics of different regions.

[0138] Use the clustering analysis method, the K-means algorithm, to classify geological regions and identify the differences in physical properties of different regions.

[0139] Compare the distribution of physical property data in different geological regions, and use heat maps or 3D visualization techniques to display the differences in physical property characteristics between regions.

[0140] Combine the physical property data and analyze the influencing factors of physical property changes in different geological regions, such as rock type, porosity, groundwater, etc.

[0141] S430: Display and dynamically adjust the 3D physical property model to obtain the multi-angle and multi-level 3D display results required by the user.

[0142] In step S430, based on the rendering model M generated in S420 Render , further perform display and dynamic adjustment. Specifically, set the dynamic viewing angle parameter θ and the scaling parameter S t (used to adjust the display scale), and realize the multi-angle display of the 3D model through the dynamic adjustment function f adjust :

[0143] M Display = f adjust (M Render , θ t , S t )

[0144] where M Display represents the finally displayed 3D model. At this time, the visualization model can perform displays at different angles and levels according to user needs, ensuring the clear presentation of 3D physical property information and meeting the display requirements in different analysis scenarios.

[0145] The key innovation points of the present invention include:

[0146] (1) Spatiotemporal synchronization processing: By calibrating timestamps and aligning spatial coordinates, it ensures that data from different sensors can be accurately synchronized and unified into the same spatiotemporal reference system, solving the problems of time and space inconsistency in traditional methods.

[0147] (2) Cross-modal data fusion and noise reduction: Using cross-modal data fusion technology, it effectively integrates data from different sensors and uses noise reduction algorithms to remove noise from the data, improving the fusion effect and overall accuracy of multi-modal data.

[0148] (3) 3D modeling and visualization: Inputting the fused data into a 3D modeling module to generate an accurate 3D physical property information model and performing visualization processing, which is convenient for users to intuitively display and analyze complex physical property information, greatly enhancing the operability and application value of the data.

[0149] The following are its main beneficial effects:

[0150] The present invention provides a method for fusing and analyzing physical property information based on 3D digital technology. Through spatiotemporal synchronization, cross-modal fusion, and noise reduction processing of multi-source data, it realizes the consistency of physical property data collected by different sensors in terms of time, space, and resolution, avoiding processing errors caused by inconsistent multi-source data in traditional methods. Compared with traditional data processing methods, the present invention can more precisely integrate physical property data such as density, magnetic susceptibility, porosity, and elastic modulus from rocks, eliminate interference between modalities, and intuitively display physical property information through 3D modeling and visualization technology, greatly improving the accuracy and usability of the data. In addition, the present invention improves the processing speed and accuracy through efficient data fusion algorithms and denoising processing, and especially significantly enhances the real-time data processing ability in complex environments.

[0151] Example 2: Figure 2 It is a structural block diagram of a system for fusing and analyzing physical property information based on 3D digital technology provided by an embodiment of the present invention. As Figure 2 shown, the system may include the following modules:

[0152] The data acquisition module 10 is used to obtain physical property data from multiple sensors. Through high-precision sensors, the system real-time collects various data related to the physical property characteristics of an object, including internal structure, material properties, and stress distribution. The data collected by various sensors have different data forms, resolutions, and precisions. The system ensures the extensiveness and effectiveness of the data and real-time collects the data stream to generate an original sensor data set.

[0153] The data preprocessing module 20 is used to perform preliminary processing on the collected multi-source physical property data. Specifically, the module performs the following functions:

[0154] Unify and format the data from different sensors and convert it into a standardized dataset.

[0155] Filter the noise data using advanced noise removal algorithms to obtain a denoised dataset.

[0156] Fill in the missing values in the data through interpolation methods to generate a complete and consistent preliminary dataset.

[0157] The spatio-temporal synchronization and alignment module 30 performs spatio-temporal alignment of the data based on the preprocessed data. This module executes the following functions:

[0158] Time synchronization: By calibrating the timestamps of the data from each sensor, the data from different sensors can be aligned according to a unified time standard to obtain a time-aligned dataset.

[0159] Spatial alignment: Use spatial coordinate transformation algorithms to map the spatial data of each sensor to a unified three-dimensional coordinate system to generate a spatially aligned dataset.

[0160] Spatio-temporal synchronization: Combine the time- and space-aligned data to form a spatio-temporal synchronized data sequence, providing a basis for subsequent fusion processing.

[0161] The cross-modal data fusion and noise reduction module 40 is responsible for performing fusion processing on the spatio-temporally synchronized data. Specifically, the module includes:

[0162] Cross-modal data grouping: Group the data according to the data modality type to generate different modality data sets.

[0163] Multi-scale fusion: Perform multi-scale fusion on the multi-resolution data under different modalities to ensure effective integration of the data within the same modality and generate a multi-modal fusion dataset.

[0164] Noise reduction processing: Perform multi-dimensional noise reduction on the fused multi-modal data to remove the interference between modalities and obtain a denoised dataset, providing accurate data for subsequent modeling and visualization.

[0165] The 3D modeling and visualization module 50 is responsible for performing 3D modeling and visualization processing of physical property information based on the denoised multi-modal data. The module includes:

[0166] 3D modeling: Input the denoised data into the 3D modeling module and perform modeling of physical property information according to the unified three-dimensional space coordinates to generate a 3D physical property information model.

[0167] Model rendering: Render the generated 3D model according to the set viewing angle and resolution to obtain a visualized 3D physical property information model, and achieve different levels of visualization effects by adjusting the viewing angle and resolution of the rendering.

[0168] Display and dynamic adjustment: Provide users with three-dimensional model displays from multiple perspectives and levels, support dynamic adjustment and interactive operations, generate the final display results, and help users intuitively understand physical property information.

[0169] The user interaction and visualization module 60 provides an intuitive user interface to display the prediction results and analysis data of the system. By supporting multi-scenario simulations and interactive operations, this module enables users to:

[0170] View three-dimensional display results: Display the three-dimensional physical property information model from the three-dimensional modeling and visualization module, support rotation and scaling operations, and help users view different angles and details of the model.

[0171] Data interaction operations: Through interactive operations, users can dynamically adjust views, data parameters, or select parts of interest for in-depth analysis according to their needs.

[0172] The system monitoring and adaptive module 70 continuously monitors the operating status of the system to ensure stable and efficient operation of the system in various complex environments. This module performs the following functions:

[0173] Real-time monitoring: Real-time monitor each module of the system, collect operation data, and ensure the accuracy and efficiency of data processing.

[0174] Adaptive optimization: Based on real-time feedback, automatically adjust and optimize the parameters of each module to improve the system processing ability and computing efficiency.

[0175] Beneficial effects of the embodiments

[0176] The physical property information fusion analysis method system based on three-dimensional digital technology provided by the present invention ensures the unity of different modal data in time and space through the fusion and precise alignment of multi-source data, thus greatly improving the analysis accuracy of physical property information. Through cross-modal data fusion and noise reduction processing, the interference and noise between data are eliminated, further enhancing the reliability and accuracy of the data. The three-dimensional modeling and visualization module enables complex physical property information to be intuitively displayed in the form of a three-dimensional model, facilitating comprehensive analysis and decision-making by users. The real-time monitoring and adaptive functions of the system ensure the stable operation of the system in complex environments, with strong adaptability and scalability. Through this system, the problems of time synchronization, spatial alignment, and multi-modal data processing existing in traditional data analysis methods can be effectively solved, providing an efficient and accurate solution for the physical property information analysis in geological exploration, geotechnical physical testing, and seismic exploration.

[0177] Obviously, the embodiments described above are only a part of the embodiments of this application, rather than all of them. The preferred embodiments of this application are shown in the drawings, but they do not limit the patent scope of this application. This application can be implemented in many different forms. On the contrary, the purpose of providing the described embodiments is to make the understanding of the disclosed content of this application more thorough and comprehensive. Although this application has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions described in the foregoing specific embodiments, or perform equivalent replacements on some of the technical features. Any equivalent structure that makes use of the content of this application's specification and drawings and is directly or indirectly applied in other related technical fields shall be within the scope of the patent protection of this application by the same token.

Claims

1. A physical property information fusion analysis method based on three-dimensional digital technology, characterized in that Including: Obtain multi-source physical property data from different sources such as geological exploration, geotechnical physical testing, and seismic surveys, perform format standardization processing to obtain a preprocessed data set; Perform time series alignment on the preprocessed data set to generate a spatio-temporal synchronization data sequence; Based on the spatio-temporal synchronization data sequence, perform cross-modal fusion and noise reduction processing to obtain multi-modal fusion data; Input the multi-modal fusion data into a 3D modeling module to generate a 3D physical property information model and perform visualization processing to obtain a 3D display result.

2. The physical property information fusion analysis method based on three-dimensional digitization technology according to claim 1, characterized in that The step of performing time series alignment on the preprocessed data set specifically includes: Calibrate the preprocessed data set according to timestamps to align the time information from different sensors and generate a spatio-temporal synchronization data sequence.

3. The physical property information fusion analysis method based on 3D digital technology according to claim 2, wherein In the time series alignment step, it further includes: Adjust the time alignment of each data point according to the spatial coordinate information of the sensor.

4. The physical property information fusion analysis method based on 3D digitization technology according to claim 1, wherein The step of performing cross-modal fusion and noise reduction processing based on the spatio-temporal synchronization data sequence specifically includes: Group the physical property data of different modalities and effectively fuse different types of data through a cross-modal fusion algorithm to obtain denoised multi-modal fusion data.

5. The physical property information fusion analysis method based on three-dimensional digitization technology according to claim 4, characterized in that The cross-modal fusion and noise reduction processing step specifically includes: Use a filtering algorithm to remove noise in the multi-modal data and interpolate and fill in the missing values in the data to generate a multi-modal fusion data set.

6. The physical property information fusion analysis method based on three-dimensional digitization technology according to claim 1, wherein The step of inputting the multi-modal fusion data into the 3D modeling module specifically includes: Input the denoised multi-modal data into a 3D modeling algorithm to generate a physical property information model in 3D space, and the physical property information model contains the structural and material information of different objects.

7. The physical property information fusion analysis method based on three-dimensional digitization technology according to claim 1, characterized in that The step of generating the 3D physical property information model and performing visualization processing specifically includes: Render the generated 3D physical property information model and generate a corresponding 3D display result according to the set viewing angle and resolution.

8. The physical property information fusion analysis method based on 3D digital technology according to claim 1, wherein The step of generating the physical property information model in the 3D modeling module specifically includes: Based on the spatial coordinates and physical property data in the spatio-temporal synchronization data sequence, construct a 3D geometric model of the object and assign corresponding physical property attributes to the model.

9. The physical property information fusion analysis method based on three-dimensional digitization technology according to claim 1, characterized in that The denoising processing step includes: Adopt wavelet transform or Kalman filtering algorithm to denoise the data of different modalities, eliminate the high-frequency noise in the data, and obtain smoother and more accurate multi-modal fusion data.

10. A physical property information fusion analysis system based on three-dimensional digital technology, characterized in that, Including: A data acquisition module for obtaining physical property data from different sensors; A data preprocessing module for performing format standardization and noise removal processing on the physical property data to obtain a preprocessed data set; A spatio-temporal synchronization and alignment module for performing time series calibration and spatial alignment on the preprocessed data to generate a spatio-temporal synchronization data sequence; A cross-modal data fusion and noise reduction module for performing cross-modal fusion and noise reduction processing on the spatio-temporal synchronization data to obtain multi-modal fusion data; A 3D modeling and visualization module for generating a 3D physical property information model according to the multi-modal fusion data and performing visualization processing to obtain a 3D display result.

Images