Search

CN-121998429-A - Intelligent analysis system for ground disaster hidden trouble point risk zone based on live-action three-dimension

CN121998429ACN 121998429 ACN121998429 ACN 121998429ACN-121998429-A

Abstract

The invention relates to the technical field of geographic information systems and discloses an intelligent analysis system for a ground disaster hidden danger point risk zone based on real scene three-dimension, which comprises a data acquisition module, a topological feature extraction module, a risk analysis decision module and an early warning confidence enhancement module, wherein the data acquisition module is used for acquiring a real scene three-dimensional grid model and a digital elevation model of a region to be monitored, the topological feature extraction module is used for extracting grid unit side length statistical distribution features and generating topological strain entropy representing geometric evolution rules, the risk analysis decision module is used for performing gradient projection operation on the topological strain entropy by using elevation field gradient vectors to construct an attribute response deformation field, and determining an evaluation instruction according to space-time cooperative probability between an image optical flow field and the attribute response deformation field.

Inventors

  • WANG XIAOQING
  • TAN LIGUO

Assignees

  • 山西省地质勘查局二一四地质队有限公司

Dates

Publication Date
20260508
Application Date
20260127

Claims (10)

  1. 1. Intelligent analysis system based on ground disaster hidden danger point risk area of live-action three-dimensional, characterized by comprising: The data acquisition module is used for acquiring a live three-dimensional grid model and a digital elevation model of the geographic space region to be monitored; The topological feature extraction module is used for extracting side length statistical distribution features of grid cells in the live-action three-dimensional grid model and generating topological strain entropy reflecting geometric evolution rules of the live-action three-dimensional grid model; The risk analysis decision module comprises Gao Chengchang coupling resolving units and a signal arbitration unit; gao Chengchang the coupling resolving unit is used for retrieving the gradient vector of the elevation field in the digital elevation model, and performing gradient projection operation on the topological strain entropy in the direction of the gradient vector of the elevation field so as to construct an attribute response deformation field after the space modeling noise is eliminated; The signal arbitration unit is used for extracting image textures imaged in the live-action three-dimensional grid model, establishing a causal relation model of geometric deformation and image texture change, combining a quantization judgment rule of attribute response deformation fields and preset risk thresholds, and outputting an evaluation instruction for representing the risk state of a geographic space region to be monitored, wherein the signal arbitration unit constructs an image light flow field according to a displacement vector of the image textures under a space coordinate system, determines the trigger level of the evaluation instruction according to space-time cooperative probability between characteristic jump frequency of the attribute response deformation fields and physical response of the image light flow field, the topological strain entropy is represented by calculating an evolution value of statistical distribution variance of grid side lengths under adjacent time phases, and gradient projection operation is carried out by mapping the evolution value to a motion vector direction meeting gravity constraint by extracting three-dimensional components of gradient vectors of the elevation fields under the geographic coordinate system so as to eliminate isotropic algorithm artifact noise.
  2. 2. The ground disaster hidden danger point risk area intelligent analysis system based on the real scene three-dimension is characterized in that the topological feature extraction module comprises a dynamic resampling unit, the dynamic resampling unit adjusts the subdivision depth of the spatial index according to curvature distribution gradient of the surface of the real scene three-dimensional grid model, the dynamic resampling unit increases grid topological connection density in the area where the curvature change amount is larger than a preset curvature threshold value, and the dynamic resampling unit executes topological thinning processing in the area where the curvature change amount is smaller than or equal to the preset curvature threshold value to merge redundant patches.
  3. 3. The intelligent analysis system based on the real-scene three-dimensional ground disaster hidden danger point risk area is characterized in that when an attribute response deformation field is constructed, the high Cheng Chang coupling resolving unit executes the following anisotropic calibration steps of S1, calculating grid side length statistical distribution variances between two real-scene three-dimensional grid models in different time phases, S2, extracting gravity field vector projections at corresponding positions in a digital elevation model, S3, calculating projection weight coefficients to obtain quantized components of the attribute response deformation field, and a calculation formula is as follows: , wherein, As a projection weight coefficient, delta H is an evolution value of grid side length statistical distribution variance in a preset observation period, and theta is an included angle between an evolution vector of the grid side length statistical distribution variance and gravity field vector projection.
  4. 4. The intelligent analysis system of the ground disaster hidden danger point risk area based on the real scene three dimensions according to claim 1, wherein the signal arbitration unit synchronously calls an image optical flow field to execute image consistency verification when the attribute response deformation field generates characteristic jump, and confirms the signal arbitration unit as a geographic entity displacement signal when the displacement characteristic of the image optical flow field and the evolution characteristic of the attribute response deformation field meet a preset time-space correlation criterion.
  5. 5. The real-scene three-dimensional ground disaster hidden danger point risk zone intelligent analysis system is characterized in that the risk analysis decision module further comprises a medium self-adaptive compensation unit, the medium self-adaptive compensation unit performs fractal dimension analysis on a real-scene three-dimensional grid model and identifies the surface medium type based on the result of the fractal dimension analysis, the medium self-adaptive compensation unit dynamically modulates a preset risk threshold value according to the surface medium type, and the medium self-adaptive compensation unit reduces the preset risk threshold value by 10-20% for the region identified as a loose stack.
  6. 6. The intelligent analysis system for the ground disaster hidden danger point risk area based on the live-action three-dimension is characterized in that the topological feature extraction module is used for identifying topological connection relations of grid units and statistical distribution rules of grid side lengths in the live-action three-dimensional grid model when calculating topological strain entropy, and the topological strain entropy is used for representing the evolution degree of geometric forms of a geographic space area to be monitored from ordered to unordered so as to identify topological singular features caused by geographic entity displacement.
  7. 7. The intelligent analysis system of the ground disaster hidden danger point risk area based on the real scene three dimensions according to claim 1, wherein the data acquisition module further comprises a time phase synchronization subunit, and the time phase synchronization subunit is used for executing space reference alignment and illumination consistency correction on the multi-period real scene three-dimensional grid model acquired by different time nodes, and inputting the corrected grid data into the topological feature extraction module.
  8. 8. The intelligent analysis system for the risk area of the ground disaster hidden trouble point based on the live-action three-dimensional model as claimed in claim 1 is characterized in that the risk analysis decision module is further used for calculating deformation acceleration of an attribute response deformation field, and when the value of the deformation acceleration is continuously larger than 0mm/s 2 in a preset observation period, the risk analysis decision module decides that the geographical space area to be monitored enters an acceleration deformation stage.
  9. 9. The intelligent analysis system for the risk area of the ground disaster hidden danger point based on the live-action three-dimension according to claim 1, wherein the risk analysis decision module further comprises a risk area demarcating unit, and the risk area demarcating unit is used for identifying a geometric closed-loop area affected by deformation based on spatial connectivity analysis of an attribute response deformation field and mapping the geometric closed-loop area into an independent deformation target feature boundary in the geographic space.
  10. 10. The intelligent analysis system for the risk area of the ground disaster hidden trouble point based on the live-action three-dimensional model according to claim 9, wherein the risk area dividing unit is further used for performing potential energy analysis according to the feature of the terrain gradient and the height difference in the feature boundary of the deformation target, calculating the potential influence range after the feature boundary of the deformation target is unstable based on the result of the potential energy analysis, and dynamically correcting the input quantity of the potential energy analysis by the real-time topological displacement quantity provided by the attribute response deformation field.

Description

Intelligent analysis system for ground disaster hidden trouble point risk zone based on live-action three-dimension Technical Field The invention relates to an intelligent analysis system for a ground disaster hidden trouble point risk zone based on live-action three-dimension, and belongs to the technical field of geographic information systems. Background Currently, generating a three-dimensional grid model by oblique photography to represent geographic entity topographic features has become a mainstream mode, and deducing deformation trend of hidden danger points by comparing geometric changes of the grid model generated by different time phases, however, when the three-dimensional grid model is applied in a complex geological environment, the dilemma of decoupling geometric representation precision and physical mechanics reliability is faced. In the prior art, a three-dimensional grid model is regarded as a static geometric envelope, recognition logics depend on absolute displacement contrast of grid nodes with different time phases, but in an actual engineering scene, unmanned aerial vehicle navigation is influenced by environmental illumination change, shadow deviation or vegetation swing, so that a modeling algorithm generates non-physical topology jitter in a grid reconstruction process, the topology artifact generated by the modeling mechanism has high overlapping performance on geometric characteristics with early stage rock-soil body creep, so that a system is difficult to distinguish spontaneous grid torsion and actual physical displacement, if the problem is solved by adding an external sensor or improving full-field scanning precision, the contradiction of large calculation cost rise and insufficient edge end processing efficiency is caused, the prior art tries to adopt a filtering algorithm or curvature analysis to remove model noise, but the method ignores mechanical logics contained in a grid topological structure, geometrical smoothing is easy to lead to key deformation characteristics loss, and cannot establish physical association between grid evolution and gravity driving instability, the traditional analysis logics are limited to data association and probability calculation, and cannot touch the essence of evolution physics, for example, the invention patent with an authorization number of CN118095866B discloses a method based on artificial intelligent carrier body collapse and actual physical displacement, when the system is used for evaluating the situation that the three-dimensional model is difficult to estimate the quality factor by using the precision of the geometrical model has the precision, and the problem that the three-dimensional deformation coefficient is difficult to be estimated by using the geometrical model has the precision, and the depth coefficient is difficult to be estimated by evaluating the precision of the point-down, and the graph is difficult to be estimated by using the analysis, and the geometrical model noise is difficult to be estimated based on the accuracy, and the quality-down. Therefore, how to realize decoupling judgment of deformation signals and modeling noise by excavating topological evolution rules generated in the three-dimensional grid model and combining with gravitational field physical constraints, and improve the confidence level of intelligent analysis of risk areas becomes the technical problem to be solved by the invention. Disclosure of Invention In order to solve the problems in the background technology, the technical scheme of the invention is as follows, a real-scene three-dimensional-based intelligent analysis system for risk areas of ground disaster hidden danger points comprises: The data acquisition module is used for acquiring a live three-dimensional grid model and a digital elevation model of the geographic space region to be monitored; The topological feature extraction module is used for extracting side length statistical distribution features of grid cells in the live-action three-dimensional grid model and generating topological strain entropy reflecting geometric evolution rules of the live-action three-dimensional grid model; The risk analysis decision module comprises Gao Chengchang coupling resolving units and a signal arbitration unit; gao Chengchang the coupling resolving unit is used for retrieving the gradient vector of the elevation field in the digital elevation model, and performing gradient projection operation on the topological strain entropy in the direction of the gradient vector of the elevation field so as to construct an attribute response deformation field after the space modeling noise is eliminated; The signal arbitration unit is used for extracting image textures imaged in the live-action three-dimensional grid model, establishing a causal relation model of geometric deformation and image texture change, combining a quantization judgment rule of attribute response deformation fields and preset ris