Rockfall early warning methods, systems, equipment and media

Abstract

This application provides a method, system, equipment, and medium for rockfall early warning; it extracts local features from geological point cloud data, aggregates these local features to generate global features, and obtains rockfall characteristics based on the global features; it constructs a three-dimensional model of the rockfall based on these characteristics; it predicts the stability and change trends of the rockfall based on the three-dimensional model, and obtains rockfall early warning information; it can accurately predict change trends using the three-dimensional model of the rockfall, improving the accuracy of rockfall early warning; based on the dynamic expression of the personalized characteristics of rockfall hazards, it meets the uncertainty analysis of rockfall disaster time, improving the reliability and accuracy of rockfall disaster early warning, and providing a data foundation for geological disaster prevention and mitigation.

Description

Technical Field

[0001] The embodiments of this application relate to the field of geological monitoring, and in particular to methods, systems, equipment and media for early warning of dangerous rocks. Background Technology

[0002] Unstable rock formations are prone to collapse, and they are characterized by their concealment and suddenness. To prevent such collapses, early warning and prevention measures are needed, further improving the analysis and monitoring precision of rock mass change trends. This involves identifying rock mass deformation through subtle changes and accelerations. The identification, monitoring, and early warning of unstable rock formations primarily rely on establishing a three-dimensional monitoring and early warning system based on rock mass deformation, cracks, and dislocations.

[0003] However, traditional manual geological hazard investigation methods are inefficient in identifying and analyzing unstable rock masses. They struggle to determine the morphology and deformation patterns of unstable rock masses, especially on steep slopes and in areas inaccessible to investigators. Furthermore, in regions with numerous unstable rock masses in remote locations, it is difficult to establish fixed monitoring stations. For instance, it is challenging to deploy automated monitoring equipment for acoustic emission, microseismic activity, crack detection, and stress analysis on a single unstable rock mass measuring only a few cubic meters. Satellite remote sensing technology can only identify large-scale deformations and cannot accurately identify or analyze the trends of small unstable rock masses. Summary of the Invention

[0004] The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.

[0005] The purpose of this application is to at least partially solve one of the technical problems existing in the related technologies. The embodiments of this application provide a method, system, device and medium for rockfall early warning, which can improve the accuracy of rockfall early warning.

[0006] An embodiment of the first aspect of this application provides a method for early warning of dangerous rocks, comprising:

[0007] Obtain geological point cloud data for the target area;

[0008] Local features are extracted from the geological point cloud data, and the local features are aggregated to generate global features. The dangerous rock features are then obtained based on the global features.

[0009] Construct a three-dimensional model of the dangerous rock based on the described characteristics;

[0010] Based on the three-dimensional model of the dangerous rock, stability and change trends are predicted to obtain early warning information for dangerous rock.

[0011] According to certain embodiments of the first aspect of this application, obtaining geological point cloud data of the target area includes:

[0012] Dense point cloud data is generated based on multi-view images of the target area;

[0013] The dense point cloud data is aligned, registered, and fused to obtain preprocessed point cloud data;

[0014] Based on the rock mass structure characteristics of historical dangerous rock data as the initial cluster centers, the preprocessed point cloud data is clustered to obtain the geological point cloud data of the target area.

[0015] According to certain embodiments of the first aspect of this application, the extraction of local features from the geological point cloud data includes:

[0016] The geological point cloud data is segmented into points, and each point in the geological point cloud data is classified.

[0017] Local information is obtained by sampling the neighborhood of each classified point, and the local information is then aggregated.

[0018] Capture local geometric details based on the local information;

[0019] The local features of the geological point cloud data are obtained by stitching together the low-level local features and high-level local features in the local geometric details information.

[0020] According to certain embodiments of the first aspect of this application, the aggregation of the local features to generate global features includes:

[0021] Obtain the geometric features of the dangerous rock from the local features;

[0022] Calculate the dynamic time-normalization distance of the time series corresponding to the geometric features of dangerous rocks with different fine-grained sizes;

[0023] Minimize the dynamic time warping distance to align the time series corresponding to the geometric features of the dangerous rock to obtain the time series features;

[0024] Spatial characteristics are obtained based on the geometric features of unstable rocks that vary in the region;

[0025] Global features are generated from the temporal features and the spatial features.

[0026] According to certain embodiments of the first aspect of this application, constructing a three-dimensional model of the dangerous rock based on the characteristics of the dangerous rock includes:

[0027] Based on the characteristics of the dangerous rock, a mesh model is obtained through Poisson reconstruction, surface reconstruction using spherical wave transform, voxelization, and voxel fusion.

[0028] The geometric morphology and topological structure of the dangerous rock are generated based on the characteristics of the dangerous rock.

[0029] By combining the geometry and topology of the unstable rock with the mesh model, a three-dimensional model of the unstable rock is constructed.

[0030] According to certain embodiments of the first aspect of this application, the step of predicting the stability and change trend based on the three-dimensional model of the dangerous rock to obtain dangerous rock early warning information includes:

[0031] Monitoring was conducted on the unstable rocks in the target area to obtain monitoring data;

[0032] The monitoring data is input into the three-dimensional model of the dangerous rock, and the spatiotemporal features of the dangerous rock at the current timestamp are extracted.

[0033] The geological monitoring indicators for the next time point are updated based on the dynamic changes in the characteristics of the dangerous rock. The spatiotemporal characteristics of the dangerous rock at the next time point are obtained based on the spatiotemporal characteristics of the dangerous rock at the current time point and the geological monitoring indicators for the next time point.

[0034] Stability and trend prediction are made based on the spatiotemporal characteristics of the dangerous rock at the current time stamp and the spatiotemporal characteristics of the dangerous rock at the next time stamp, and dangerous rock warning information is obtained.

[0035] According to certain embodiments of the first aspect of this application, after obtaining early warning information for a dangerous rock mass by predicting its stability and change trends based on the three-dimensional model of the dangerous rock mass, the method includes:

[0036] Sensitivity analysis is performed based on the aforementioned rockfall warning information to determine the hazard level and corresponding response facilities;

[0037] The dangerous rock warning information is combined with geographic information to generate a risk zoning map;

[0038] By importing monitoring data, rockfall early warning information, and risk zoning maps into a geographic information system, we can obtain the displacement vector field of the rockfall, crack propagation information, rockfall collapse movement path, and threat area.

[0039] A second aspect of this application provides a rockfall early warning system, comprising:

[0040] The data collection unit is used to acquire geological point cloud data of the target area;

[0041] The feature extraction unit is used to extract local features from the geological point cloud data, aggregate the local features to generate global features, and obtain dangerous rock features based on the global features.

[0042] The model building unit is used to construct a three-dimensional model of the dangerous rock based on the characteristics of the dangerous rock.

[0043] The prediction unit is used to predict the stability and change trend of the dangerous rock based on the three-dimensional model of the dangerous rock, and obtain the early warning information of the dangerous rock.

[0044] According to a third aspect of this application, an electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the rockfall warning method as described in the first aspect of this application.

[0045] According to a fourth aspect of this application, a computer storage medium stores computer-executable instructions for performing the rockfall early warning method as described in the first aspect of this application.

[0046] The above-mentioned scheme has at least the following beneficial effects: by extracting local features from geological point cloud data, aggregating local features to generate global features, and obtaining dangerous rock features based on global features; constructing a three-dimensional model of dangerous rocks based on dangerous rock features; predicting stability and change trends based on the three-dimensional model of dangerous rocks to obtain dangerous rock warning information; accurately predicting change trends using the three-dimensional model of dangerous rocks, improving the accuracy of dangerous rock warnings; and meeting the uncertainty analysis of dangerous rock disaster time based on the dynamic expression of personalized characteristics of dangerous rock hazards, thus improving the reliability and accuracy of dangerous rock disaster warnings and providing a data foundation for geological disaster prevention and mitigation work. Attached Figure Description

[0047] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0048] Figure 1 This is a flowchart illustrating the steps of a rockfall early warning method.

[0049] Figure 2 This is a step-by-step diagram illustrating the process of acquiring geological point cloud data for a target area;

[0050] Figure 3 This is a step-by-step diagram illustrating the extraction of local features from geological point cloud data;

[0051] Figure 4 This is a diagram illustrating the steps involved in aggregating local features to generate global features.

[0052] Figure 5 This is a step-by-step diagram of constructing a 3D model of a dangerous rock based on its characteristics;

[0053] Figure 6 This diagram illustrates the steps involved in predicting the stability and changing trends of a dangerous rock formation to obtain early warning information.

[0054] Figure 7 This is a step diagram of the response process. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0056] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, or the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0057] The embodiments of this application will be further described below with reference to the accompanying drawings.

[0058] The embodiments of this application provide a method for early warning of dangerous rocks.

[0059] Reference Figure 1 The method for early warning of dangerous rocks includes the following steps:

[0060] Step S100: Obtain geological point cloud data for the target area;

[0061] Step S200: Extract local features from geological point cloud data, aggregate local features to generate global features, and obtain dangerous rock features based on global features;

[0062] Step S300: Construct a three-dimensional model of the dangerous rock based on its characteristics;

[0063] Step S400: Based on the three-dimensional model of the dangerous rock, predict its stability and change trend to obtain early warning information for the dangerous rock.

[0064] Reference Figure 2 For step S100, acquiring geological point cloud data of the target area includes the following steps:

[0065] Step S110: Generate dense point cloud data based on multi-view images of the target area;

[0066] Step S120: Align, register and fuse the dense point cloud data to obtain preprocessed point cloud data;

[0067] Step S130: Based on the rock mass structure characteristics of historical dangerous rock data as the initial cluster centers, the preprocessed point cloud data is clustered to obtain the geological point cloud data of the target area.

[0068] For example, multi-view images of the target area are acquired, and dense point cloud data is generated based on these images. Radius filtering is used to denoise the dense point cloud, removing interference from vegetation, artificial structures, etc. An iterative nearest-point algorithm suitable for surface registration is then used to align, register, and fuse the multi-point point cloud data, through its objective function. To minimize the mean square error between corresponding points, statistical methods are used to remove mismatched point pairs. Weights are assigned to different point pairs to reduce the impact of noise. Low-resolution point clouds are first coarsely registered and then gradually refined. Geological environment data of the target area are obtained from multi-view images of large-scale complex terrain as preprocessed point cloud data.

[0069] The preprocessed multi-source point cloud data is pre-classified; based on existing historical dangerous rock data as a sample dataset, the sample dataset is clustered according to the rock mass structure characteristics of the disaster samples. The rock mass structure characteristics of the target area are obtained by clustering the rock mass structure features using the geometric skeleton, color information, and normal vectors of the dangerous rock point cloud data.

[0070] Rock mass structural features are used as initial cluster centers in clustering. The point cloud data of dangerous rock samples in the sample dataset are assigned to the nearest initial cluster center. The clusters corresponding to the initial cluster centers are obtained, the cluster centers are updated, and the final clustering results are obtained through iterative clustering. The geological point cloud data of the target area is obtained, and the dangerous rock point cloud dataset is generated.

[0071] This process acquires the rock mass structural features of the target area, using point cloud data with multiple attributes to describe complex geometric shapes and physical properties in three-dimensional space, providing a rich information foundation for subsequent processing and analysis. Millimeter-level precision point cloud data is obtained through lidar scanning. Preprocessing of the surface and surrounding topographic data of the unstable rock mass is performed, and an iterative nearest-point algorithm is used to progressively reduce the distance between the source and target point clouds. Local features of the point clouds are extracted to improve the robustness of registration, and neural networks are used to learn the intrinsic features of the point clouds to improve registration accuracy. The process then proceeds to register the rock mass structural features in the target area. Based on the location and characteristics of the unstable rocks, the preprocessed cloud data is pre-classified to screen for potential hazard locations within the target area.

[0072] Point clouds are characterized by their disorder, sparsity, and unstructured nature. The core objective of point cloud segmentation is to classify points in a 3D point cloud into different semantic categories or instances. The role of a Multilayer Perceptron (MLP) in point cloud processing is feature extraction and local feature learning. The MLP maps the original point coordinates or low-dimensional features to a high-dimensional latent space through fully connected layers and nonlinear activation functions, capturing local geometric patterns. By using MLP layers with shared weights to process all points in the neighborhood independently, and combining pooling operations (such as max pooling), robust features of local regions are extracted. Point cloud segmentation techniques and MLPs are used to extract local features from the point cloud, and then max pooling is applied to aggregate the features of all points, generating a global feature vector that forms the "dangerous rock" feature.

[0073] Reference Figure 3 For step S200, extracting local features from the geological point cloud data includes the following steps:

[0074] Step S211: Perform point cloud segmentation on the geological point cloud data and classify each point in the geological point cloud data;

[0075] Step S212: Perform neighborhood sampling on each classified point to obtain local information, and aggregate the local information;

[0076] Step S213: Capture local geometric detail information based on local information;

[0077] Step S214: The low-level local features and high-level local features in the local geometric detail information are stitched together to obtain the local features of the geological point cloud data.

[0078] For example, the location of potential dangerous rocks in the target area is obtained, and the corresponding point cloud data is extracted. Point cloud segmentation is performed on the geological point cloud data, classifying each point in the geological point cloud data into different object types. The neighborhood of each point is sampled to form a local point set, obtaining local information. A shared-weighted Multi-Level Processing (MLP) is used to process each point and its neighborhood, and then the local information is aggregated. Multiple stacked MLP modules and pooling modules are used to capture local geometric details. Low-level local features are concatenated with high-level features to obtain local features, improving detail preservation. Graph optimization is used to refine the segmentation results.

[0079] Reference Figure 4 The process of aggregating local features to generate global features includes the following steps:

[0080] Step S221: Obtain the geometric features of the dangerous rock from local features;

[0081] Step S222: Calculate the dynamic time regularization distance of the time series corresponding to the geometric features of dangerous rocks with different fine-grained characteristics;

[0082] Step S223: Minimize the dynamic time warping distance to align the time series corresponding to the geometric features of the dangerous rock to obtain the time series features;

[0083] Step S224: Obtain spatial characteristics based on the geometric features of the unstable rock in the region;

[0084] Step S225: Generate global features from temporal and spatial features.

[0085] By analyzing curvature and normal vectors, steep surfaces and uneven regions (potential instability surfaces) of dangerous rocks are identified. Based on region growing and clustering (such as DBSCAN), the fracture point cloud is segmented to obtain the geometric features of the dangerous rocks. Parameters such as fracture length, width, dip angle, and connectivity are extracted to obtain the fracture parameters of the dangerous rocks as their geometric features.

[0086] The dynamic time warping algorithm is used to calculate the dynamic time warping distance of the time series corresponding to the geometric features of dangerous rocks with different fine granularities. The different time series deformation sequences are aligned by minimizing the dynamic time warping distance, and spatial features are obtained based on the geometric features of regional changes. A global feature vector is generated from the time series features and spatial features.

[0087] Characteristics of potential rockfall hazards are obtained based on temporal and spatial features.

[0088] Reference Figure 5 For step S300, constructing a three-dimensional model of the dangerous rock based on its characteristics includes the following steps:

[0089] Step S310: Based on the characteristics of the dangerous rock, a mesh model is obtained through Poisson reconstruction, surface reconstruction by spherical wave transform, voxelization, and voxel fusion.

[0090] Step S320: Generate the geometric shape and topological structure of the dangerous rock based on its characteristics;

[0091] Step S330: Combine the geometry and topology of the unstable rock with the mesh model to construct a three-dimensional model of the unstable rock.

[0092] For example, point cloud data from multiple perspectives are registered and stitched together to form an initial model. Surface reconstruction is performed using Poisson reconstruction and spherical wave transform, and the point cloud data is divided into a series of small cubes through voxelization. A mesh model is obtained by voxel fusion and refinement. Multiple point cloud data are aligned using ICP, and triangulation, surface fitting, or deep learning generation are performed using various algorithms (such as PointNet, NPGA, and 3DGaussian Splatting). By combining the geometric shape and topology of the unstable rock with the mesh model, a three-dimensional model of the unstable rock is constructed.

[0093] Reference Figure 6For step S400, the stability and change trend prediction of the dangerous rock are used to obtain the early warning information of the dangerous rock, which includes the following steps:

[0094] Step S410: Monitor the unstable rocks in the target area and obtain monitoring data;

[0095] Step S420: Input the monitoring data into the three-dimensional model of the dangerous rock and extract the spatiotemporal features of the dangerous rock at the current timestamp;

[0096] Step S430: Update the geological monitoring indicators for the next time stamp based on the dynamic changes in the characteristics of the dangerous rock. Obtain the spatiotemporal characteristics of the dangerous rock at the next time stamp based on the spatiotemporal characteristics of the dangerous rock at the current time stamp and the geological monitoring indicators for the next time stamp.

[0097] Step S440: Based on the spatiotemporal characteristics of the dangerous rock at the current time stamp and the spatiotemporal characteristics of the dangerous rock at the next time stamp, the stability and change trend are predicted to obtain dangerous rock warning information.

[0098] For example, dynamic monitoring of unstable rocks in a target area yields monitoring data. Dynamic monitoring methods can include UAV-borne LiDAR, 3D laser scanning, etc., employing periodic data acquisition and fixed acquisition parameters. The dynamic monitoring data undergoes noise reduction, filtering, and interpolation preprocessing, followed by index calculations such as displacement rate, acceleration, and crack propagation rate to obtain the final monitoring data.

[0099] The monitoring data is input into the 3D model of the unstable rock mass, and the spatiotemporal features of the unstable rock mass at the current timestamp are extracted. The spatial attributes corresponding to the unstable rock mass are extracted from the monitoring data, and the obtained time series of index features with spatial attributes are imported into a gated loop unit to obtain the spatiotemporal features of the unstable rock mass at the current timestamp.

[0100] The geological monitoring indicators for the next monitoring time stamp are updated based on the dynamic changes in the characteristics of the potential rock hazard. The spatiotemporal characteristics of the rock hazard at the next monitoring time stamp are obtained using a three-dimensional model of the rock hazard. Based on the spatiotemporal characteristics of the rock hazard locations at adjacent time stamps, the three-dimensional model is imported into mathematical analysis models, such as time series analysis and regression analysis, to predict the changing trend of the rock hazard. Through mechanical models such as the limit equilibrium method and deformation threshold, combined with geological environmental factors, stress distribution and instability conditions are simulated to conduct sensitivity analysis and rock hazard stability evaluation.

[0101] It should be noted that using lidar to generate high-precision 3D point clouds is suitable for large-scale or complex surface objects; using laser rangefinders for single-point or multi-point ranging is suitable for small targets or low-cost scenarios. Each data acquisition is marked with a timestamp and a spatial coordinate system (such as WGS84 or a local coordinate system) to ensure data alignment and synchronization. For environmental control, fixed reference points are set near the object to assist in data alignment and error correction.

[0102] It should be noted that point-by-point coordinate difference calculations are performed on the registered point cloud to generate a displacement vector field. Simultaneously, the centroid, volume, or surface curvature changes of the point cloud are calculated using a 3D model to achieve stability evaluation through quantitative difference analysis of unstable rocks. By increasing the acquisition frequency or using a phase-detection lidar (such as FMCW LiDAR) to improve ranging resolution, small displacement detection is enhanced. A displacement threshold (e.g., 5 mm) is set; values ​​exceeding the threshold are considered object movement.

[0103] For the rate of change of the unstable rock mass, a critical displacement rate (e.g., 10 mm / day) is set according to the standard displacement rate threshold method. Acceleration analysis is used to determine the accelerated deformation stage using the second derivative of displacement. The displacement trend of the unstable rock mass is predicted using the statistical model grey system theory (GM(1,1)). Historical data is trained using the machine learning model LSTM neural network and random forest regression to predict the future deformation of the unstable rock mass. Simultaneously, a rainfall infiltration model (e.g., GreenAmpt) and seismic response analysis are introduced to simulate the impact of external factors on stability.

[0104] After predicting the stability and trend of a dangerous rock mass based on a three-dimensional model and obtaining early warning information, the method includes response steps. (Refer to...) Figure 7 The response steps include the following:

[0105] Step S510: Perform sensitivity analysis based on the rockfall warning information to determine the hazard level and corresponding response facilities;

[0106] Step S520: Combine the rockfall early warning information with geographic information to generate a risk zoning map;

[0107] Step S530: Import the monitoring data, rockfall early warning information, and risk zoning map into the geographic information system to obtain the displacement vector field of the rockfall, crack propagation information, rockfall collapse movement path, and threat area.

[0108] For example, based on the stability assessment and changing trends of unstable rock masses, sensitivity analysis or probability assessment is conducted to determine the hazard level and plan emergency response plans and evacuation routes. Based on geological disaster prevention, the early warning system sets yellow, orange, and red warning levels and corresponding response measures. Using a geographic information system (GIS) platform, 3D models, and dynamic charts, monitoring data is combined with geographic information to generate heat maps or risk zoning maps. Based on indicators such as deformation, acceleration, and crack density, the fuzzy comprehensive evaluation method is used to classify the hazard level of the target unstable rock mass, such as high, medium, and low hazard levels. Monitoring data, model prediction results, and risk zoning maps are imported into the GIS platform to construct a 3D model of the unstable rock mass. Topographic data is overlaid on the GIS platform to display the displacement vector field and crack propagation animation in real time, simulating the rockfall path, calculating the threat area, displaying the risk level distribution on a heat map, and overlaying population density and infrastructure layers to construct a 3D visualization platform. Based on the threat range of the unstable rock mass, different levels of early warning are generated according to the disaster range. Regional evacuation decisions are made based on the connection of relevant knowledge graphs to the early warning information. The visualization interface supports real-time querying of monitoring data, early warning information, and the emergency response plan database.

[0109] The methods for early warning of dangerous rocks include data acquisition, identification of potential dangerous rocks, dynamic monitoring, early warning models, and early warning processing.

[0110] By acquiring multi-source point cloud data of the target area, including key attributes such as spatial coordinates, color information, and normal vectors, the iterative nearest point algorithm and FINet algorithm are used to address issues such as data noise and local optima in point cloud registration. Data fusion analysis and feature extraction are performed on the preprocessed multi-source point cloud data to identify and screen the locations of potential rockfall hazards within the target area. Based on the point cloud data of potential rockfall hazard locations, a 3D model of the rockfall is established to obtain the existing deformation and crack characteristics, and potential slip surfaces and separation bodies are analyzed based on block theory. According to the characteristics of the potential rockfall hazard, multi-dimensional monitoring is conducted regularly to obtain multi-dimensional, multi-period point cloud difference detection data on crack expansion or displacement changes. Based on the multi-dimensional monitoring data, hierarchical feature fusion and feature extraction are performed to construct a rockfall early warning model, dynamically judging the change trend of the rock mass, and conducting hazard analysis and avoidance decisions based on the change trend. The rockfall hazard analysis and avoidance decisions are visualized, and interactive information of hazard analysis and avoidance decisions is obtained. The rockfall hazard early warning model is dynamically corrected based on the interactive information.

[0111] By dynamically expressing the personalized characteristics of potential rockfall hazards, the system can meet the uncertainty analysis of rockfall disaster timing, thereby improving the reliability and accuracy of rockfall disaster early warning and providing a data foundation for geological disaster prevention and mitigation.

[0112] An embodiment of this application provides a rockfall early warning system.

[0113] The rockfall early warning system includes: a data collection unit, a feature extraction unit, a model building unit, and a prediction unit.

[0114] The system includes a data collection unit for acquiring geological point cloud data of the target area; a feature extraction unit for extracting local features from the geological point cloud data, aggregating the local features to generate global features, and obtaining dangerous rock features based on the global features; a model building unit for constructing a three-dimensional model of dangerous rock based on the dangerous rock features; and a prediction unit for predicting the stability and change trend of dangerous rock based on the three-dimensional model of dangerous rock to obtain dangerous rock warning information.

[0115] It is understood that the rockfall warning system in this embodiment adopts the rockfall warning method described above. Each unit of the rockfall warning system corresponds to each step of the rockfall warning method. The rockfall warning system and the rockfall warning method have the same technical solution, solve the same technical problem, and have the same technical effect.

[0116] An embodiment of this application provides an electronic device. The electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the rockfall early warning method described above.

[0117] This electronic device can be any smart terminal, including computers.

[0118] In general, for the hardware structure of electronic devices, the processor can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits to execute relevant programs and implement the technical solutions provided in the embodiments of this application.

[0119] The memory can be implemented in the form of read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory and is called and executed by the processor.

[0120] Input / output interfaces are used to implement information input and output.

[0121] The communication interface is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0122] The bus transmits information between various components of a device, such as the processor, memory, input / output interfaces, and communication interfaces. The processor, memory, input / output interfaces, and communication interfaces communicate with each other within the device via the bus.

[0123] Embodiments of this application provide a computer storage medium. The computer storage medium stores computer-executable instructions for executing the rockfall early warning method described above.

[0124] Those skilled in the art will appreciate that all or some of the steps and systems in the method disclosed above can be implemented as software, firmware, hardware, and appropriate combinations thereof. Some physical components or all physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, and the computer-readable medium can include computer storage media (or non-transitory media) and communication media (or temporary media). As known to those skilled in the art, the term computer storage media is included in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data) and is volatile and non-volatile, removable, and non-removable. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory, or other memory technology, CD-ROM, digital versatile disks (DVD), or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage, or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium. In the foregoing description of this specification, references to terms such as "one embodiment," "another embodiment," or "some embodiments," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of this application. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0125] Those skilled in the art will appreciate that all or some of the steps in the methods, systems, and functional modules / units in the devices disclosed above may be implemented as software, firmware, hardware, or appropriate combinations thereof.

[0126] The units described above as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0127] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0128] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store programs.

[0129] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of the above units is merely a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings or direct couplings or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms. Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of this application. The scope of this application is defined by the claims and their equivalents. The above is a detailed description of preferred embodiments of this application, but this application is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application, and these equivalent modifications or substitutions are all included within the scope defined by the claims of this application.

Claims

1. A method for early warning of dangerous rocks, characterized in that, include: Obtain geological point cloud data for the target area; Local features are extracted from the geological point cloud data, and the local features are aggregated to generate global features. The dangerous rock features are then obtained based on the global features. Construct a three-dimensional model of the dangerous rock based on the described characteristics; Based on the three-dimensional model of the dangerous rock, stability and change trends are predicted to obtain early warning information for the dangerous rock. The extraction of local features from the geological point cloud data includes: The geological point cloud data is segmented into points, and each point in the geological point cloud data is classified. Local information is obtained by sampling the neighborhood of each classified point, and the local information is then aggregated. Capture local geometric details based on the local information; The low-level local features and high-level local features in the local geometric detail information are concatenated to obtain the local features of the geological point cloud data. The aggregation of the local features to generate global features includes: Obtain the geometric features of the dangerous rock from the local features; Calculate the dynamic time-normalization distance of the time series corresponding to the geometric features of dangerous rocks with different fine-grained sizes; Minimize the dynamic time warping distance to align the time series corresponding to the geometric features of the dangerous rock to obtain the time series features; Spatial characteristics are obtained based on the geometric features of unstable rocks that vary in the region; Global features are generated from the temporal features and the spatial features.

2. The method for early warning of dangerous rocks according to claim 1, characterized in that, The acquisition of geological point cloud data for the target area includes: Dense point cloud data is generated based on multi-view images of the target area; The dense point cloud data is aligned, registered, and fused to obtain preprocessed point cloud data; Based on the rock mass structure characteristics of historical dangerous rock data as the initial cluster centers, the preprocessed point cloud data is clustered to obtain the geological point cloud data of the target area.

3. The method for early warning of dangerous rocks according to claim 1, characterized in that, The construction of a three-dimensional model of the dangerous rock based on the characteristics of the dangerous rock includes: Based on the characteristics of the dangerous rock, a mesh model is obtained through Poisson reconstruction, surface reconstruction using spherical wave transform, voxelization, and voxel fusion. The geometric morphology and topological structure of the dangerous rock are generated based on the characteristics of the dangerous rock. By combining the geometry and topology of the unstable rock with the mesh model, a three-dimensional model of the unstable rock is constructed.

4. The method for early warning of dangerous rocks according to claim 1, characterized in that, The step of predicting the stability and change trend of the dangerous rock based on the three-dimensional model of the dangerous rock to obtain dangerous rock early warning information includes: Monitoring was conducted on the unstable rocks in the target area to obtain monitoring data; The monitoring data is input into the three-dimensional model of the dangerous rock, and the spatiotemporal features of the dangerous rock at the current timestamp are extracted. The geological monitoring indicators for the next time point are updated based on the dynamic changes in the characteristics of the dangerous rock. The spatiotemporal characteristics of the dangerous rock at the next time point are obtained based on the spatiotemporal characteristics of the dangerous rock at the current time point and the geological monitoring indicators for the next time point. Stability and trend prediction are made based on the spatiotemporal characteristics of the dangerous rock at the current time stamp and the spatiotemporal characteristics of the dangerous rock at the next time stamp, and dangerous rock warning information is obtained.

5. The method for early warning of dangerous rocks according to claim 4, characterized in that, After predicting the stability and change trend based on the three-dimensional model of the dangerous rock to obtain early warning information about the dangerous rock, the method includes: Sensitivity analysis is performed based on the aforementioned rockfall warning information to determine the hazard level and corresponding response facilities; The dangerous rock warning information is combined with geographic information to generate a risk zoning map; By importing monitoring data, rockfall early warning information, and risk zoning maps into a geographic information system, we can obtain the displacement vector field of the rockfall, crack propagation information, rockfall collapse movement path, and threat area.

6. A rockfall early warning system, characterized in that, include: The data collection unit is used to acquire geological point cloud data of the target area; The feature extraction unit is used to extract local features from the geological point cloud data, aggregate the local features to generate global features, and obtain dangerous rock features based on the global features. The model building unit is used to construct a three-dimensional model of the dangerous rock based on the characteristics of the dangerous rock. The prediction unit is used to predict the stability and change trend based on the three-dimensional model of the dangerous rock, and obtain early warning information of the dangerous rock. The extraction of local features from the geological point cloud data includes: The geological point cloud data is segmented into points, and each point in the geological point cloud data is classified. Local information is obtained by sampling the neighborhood of each classified point, and the local information is then aggregated. Capture local geometric details based on the local information; The low-level local features and high-level local features in the local geometric detail information are concatenated to obtain the local features of the geological point cloud data. The aggregation of the local features to generate global features includes: Obtain the geometric features of the dangerous rock from the local features; Calculate the dynamic time-normalization distance of the time series corresponding to the geometric features of dangerous rocks with different fine-grained sizes; Minimize the dynamic time warping distance to align the time series corresponding to the geometric features of the dangerous rock to obtain the time series features; Spatial characteristics are obtained based on the geometric features of unstable rocks that vary in the region; Global features are generated from the temporal features and the spatial features.

7. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the rockfall early warning method as described in any one of claims 1 to 5.

8. A computer storage medium, characterized in that, The device stores computer-executable instructions for performing the rockfall early warning method as described in any one of claims 1 to 5.

Images