Search

CN-121999153-A - Method and system for calculating earth excavation and filling directions based on three-dimensional reconstruction of unmanned aerial vehicle

CN121999153ACN 121999153 ACN121999153 ACN 121999153ACN-121999153-A

Abstract

The invention provides a method and a system for calculating earth excavation and filling based on three-dimensional reconstruction of an unmanned aerial vehicle, wherein the method comprises the following steps of (1) carrying out aerial photography on the unmanned aerial vehicle carrying a laser radar and an RGB camera to obtain multi-angle high-precision point cloud data and aerial photo images, (2) preprocessing the point cloud data by using PDAL tool chains to generate standardized point cloud files, (3) generating a high-precision three-dimensional model containing elevation information by adopting a multi-view stereoscopic vision technology, synchronously constructing a two-dimensional terrain model, (4) carrying out model visualization based on a visual system developed by a NET Framework 4.5 frame, (5) realizing accurate measurement and calculation of engineering earth excavation and filling amount by dynamic grid division and elevation analysis based on the three-dimensional reconstruction model, and (6) outputting three-dimensional reconstruction and earth calculation results by multi-format derivation and intelligent interaction. The invention takes data as a core, opens up a complete chain from acquisition and calculation to application, and provides full-dimension technical support for green construction and intelligent construction sites.

Inventors

  • ZHU RUI
  • WANG SHILONG
  • ZHANG YUXIN
  • Dai Aiping
  • LIU WEISHENG
  • CAI CHUNLEI
  • LUO GUANG
  • CAO BIN

Assignees

  • 中铁建电气化局集团南方工程有限公司

Dates

Publication Date
20260508
Application Date
20260106

Claims (10)

  1. 1. The calculation method of the earth excavation and filling direction based on the three-dimensional reconstruction of the unmanned aerial vehicle is characterized by comprising the following steps: (1) Carrying out aerial photography by an unmanned aerial vehicle carrying a laser radar and an RGB camera to obtain multi-angle high-precision point cloud data and aerial photography images; (2) Preprocessing the point cloud data by using PDAL tool chains to generate a standardized point cloud file; (3) Three-dimensional reconstruction modeling, namely generating a high-precision three-dimensional model containing elevation information by adopting a multi-view stereoscopic vision technology, and synchronously constructing a two-dimensional terrain model; (4) Performing model visualization based on a visualization system developed by a NET Framework 4.5 Framework; (5) Based on the three-dimensional reconstruction model, accurate measurement and calculation of the engineering filling amount are realized through dynamic grid division and elevation analysis, and a calculation result and an elevation model are displayed in real time; (6) And the three-dimensional reconstruction and earthwork calculation result is deeply integrated into the whole engineering management flow through multi-format derivation and intelligent interaction, so that a quantitative basis is provided for earthwork balance, mechanical scheduling and construction period optimization.
  2. 2. The earth excavation and filling calculation method based on the three-dimensional reconstruction of the unmanned aerial vehicle is characterized in that in the step (1), a laser radar and a multi-angle camera are carried, an unmanned aerial vehicle is provided with an active detection and passive sensing cooperative system, the laser radar acquires three-dimensional point cloud of the earth surface through pulse scanning, the details of topography fluctuation are accurately captured, the camera multi-view aerial photography supplements rich texture information, meanwhile, aiming at complex environments of mountain areas and mining areas, the unmanned aerial vehicle adopts a dynamic flight strategy, the flying height and heading are adjusted in a self-adaptive mode, topography fluctuation and vegetation shielding are avoided, and a weather monitoring module is combined, redundant route planning is started under severe conditions such as rain and fog, and data integrity is ensured.
  3. 3. The method for calculating the earthwork excavated based on the three-dimensional reconstruction of the unmanned aerial vehicle according to claim 1, wherein the specific process of the step (2) is as follows: (2-1) firstly converting a coordinate system, unifying original data to an engineering coordinate system, eliminating geographic space deviation and ensuring the space consistency of multi-source data; (2-2) expanding the dimension of the data through attribute assignment, adding strength and classification label characteristic information for the point cloud, and enhancing the resolvability of the data; (2-3) optimizing the distribution of the point clouds by adopting a local neighborhood analysis method aiming at common noise interference in a complex terrain scene, so as to effectively inhibit outliers caused by vegetation jitter, meteorological factors or equipment errors and improve the purity of data; (2-4) separating different ground object types of the earth surface, the building and the vegetation through dynamic threshold screening, realizing data layering management and creating conditions for fine modeling.
  4. 4. The method for calculating the earthmoving mass based on the three-dimensional reconstruction of the unmanned aerial vehicle according to claim 1, wherein the three-dimensional reconstruction modeling in the step (3) comprises the steps of: (3-1) firstly, establishing a space topological frame through sparse point cloud matching, and then, utilizing a dense reconstruction algorithm to diffuse point cloud density layer by layer to accurately restore topographic relief and building contour geometric features; (3-2) simultaneously combining with RGB image texture mapping, endowing the model with real colors and surface details, and constructing a three-dimensional entity with geometric precision and visual reality; and (3-3) introducing a layering optimization strategy in a reconstruction process aiming at a complex terrain scene, separating point cloud data of the earth surface, vegetation and artificial structures, and eliminating model holes of a data missing area through surface fitting and topology restoration.
  5. 5. The method for calculating the earthwork excavated based on the three-dimensional reconstruction of the unmanned aerial vehicle according to claim 1, wherein the visualization system developed based on the C# and the WPF framework in the step (4) integrates the functions of user authentication, data management, model reconstruction and calculation analysis, and efficient and stable business logic processing is realized through a modularized design.
  6. 6. The method for calculating the earth excavation and filling based on the three-dimensional reconstruction of the unmanned aerial vehicle according to claim 1, wherein in the step (5), a square grid calculation method is adopted, a measurement area is customized through an interactive interface, a polygonal boundary or a dragging adjustment range is flexibly drawn, requirements of complex terrains and irregular construction areas are adapted, a calculation module automatically extracts elevation data in a selected area, design elevation and current ground surface height are compared, earth excavation and filling difference values are analyzed by grids, total volume results are accumulated and generated, and the topography fluctuation and engineering quantity distribution are intuitively displayed through color gradient mapping.
  7. 7. The method for calculating earthwork based on three-dimensional reconstruction of unmanned aerial vehicle according to claim 6, wherein when used for the earthwork calculation of regular sites, the step (5) comprises: (5-1) topographic data processing Reading a geographic grid file, acquiring an elevation matrix A and space reference information, eliminating terrain noise through 5×5 median filtering, and outputting a smooth AGood matrix to an Excel file to ensure data reliability; (5-2) elevation data pretreatment The 5×5 median filter is used to eliminate the high Cheng Yi constant, and the mathematical expression is: , (5-3) design elevation modeling Taking a site center point (x 0, y 0) as a reference, establishing an inclined design plane, and calculating a design elevation Hn by the following formula: , wherein ix and iy are longitudinal and transverse gradient coefficients, a is grid spacing, gradient gradual change of the central elevation H0 to the periphery is realized, four quadrants of the matrix are covered through four groups of nested circulation partition calculation, and the design elevation of each grid point is ensured to be accurately generated; (5-4) fill-out amount estimation Calculating a construction height h=hn-a, a positive value representing the fill, a negative value representing the dig, planning to calculate the amount of earth by a grid method by dividing the field into (hang-1) × (lie-1) grid cells, estimating the fill volume of each cell using an absolute height c=abs (h), Calculating the unit filling amount by adopting a quadrangular prism method, and adopting a single grid volume formula: where hk=hn-a is the corner construction height, by matrix operation: , 。
  8. 8. The method for calculating the earthwork of the excavated filler based on the three-dimensional reconstruction of the unmanned aerial vehicle according to claim 1, wherein in the calculation of the earthwork based on the CAD, the step (5) comprises: and (3) data acquisition and boundary extraction, namely reading point set data selected by a user through an AutoCAD API, extracting a peripheral boundary by adopting a convex hull algorithm, and connecting adjacent points according to polar angle sequencing by taking the point with the minimum X as a starting point to form a closed polygon so as to ensure that the subsequent triangulation is performed in the boundary. Constructing a triangular network through a vector included angle maximization criterion: Delaunay triangulation, namely constructing a triangular net by adopting a side-by-side expansion method, traversing each boundary edge, searching a third point meeting the maximum angle condition to form a triangle, ensuring an empty circle criterion, calculating and screening an optimal connection point through a vector included angle, dynamically creating a new edge and maintaining a triangle adjacent relation, finally generating a triangular grid conforming to the terrain characteristic, calculating a vector included angle formed by a candidate point and an edge end point for each boundary edge, And selecting the point with the largest included angle to construct a triangle, and meeting the characteristic of a local empty circle, wherein the coordinates of the end points of the edge AB are (xA, yA), (xB, yB), the coordinates of the candidate point P are (xP, yP), and the vector included angle formula is as follows: , Selecting the point with the largest theta to form a triangle, ensuring that the triangle is as close to an equilateral form as possible, Contour interpolation calculation, namely carrying out linear interpolation on triangle edges to obtain contour points, wherein the contour point heights z1 and z2 and the contour distance deltaz are given, and when an integer k exists to ensure that kdeltaz epsilon (z 1 and z 2), interpolation parameters are as follows: , The coordinates of the interpolation points are as follows: , the contour line generation, namely, carrying out the height Cheng Chazhi on the edge of each triangle on the basis of a triangle network, calculating the coordinates of the intersection points according to the linear proportion when the Z values of the adjacent vertexes cross the contour line according to the set contour distance, marking each 5 contour lines as a curve, realizing classification by calculation, associating the generated contour lines with the corresponding triangle indexes, providing topology information for the subsequent smoothing treatment, And CAD integrated visualization, namely drawing the generated boundary, triangle network and contour line into the drawing by utilizing an autoCAD COM interface, so as to realize seamless connection of data and engineering drawing.
  9. 9. The earthmoving computing system based on three-dimensional reconstruction of an unmanned aerial vehicle of claim 1, comprising: (1) The data acquisition module comprises an unmanned aerial vehicle carrying a laser radar and an RGB camera and is used for acquiring multi-angle high-precision point cloud data and aerial images; (2) The point cloud data preprocessing is used for carrying out standardized processing through PDAL tool chains to generate standardized point cloud files; (3) The two-dimensional and three-dimensional modeling module is used for generating a high-precision three-dimensional model containing elevation information through a multi-view stereoscopic vision technology; (4) The visual display module is used for three-dimensional/two-dimensional/point cloud multi-mode display; (5) The earth volume calculating module is used for realizing accurate calculation of the engineering filling volume through dynamic grid division and elevation analysis; (6) And the result output and application module is used for multi-format export and provides data support for engineering budget, construction planning and resource allocation.
  10. 10. The earthwork calculation system based on unmanned aerial vehicle three-dimensional reconstruction according to claim 9, wherein a dynamic signature algorithm based on HMAC-SHA256 is adopted in two-dimensional and three-dimensional reconstruction, unique signatures are generated through a time stamp, a request method, a URI and a load abstract, replay and data tampering are effectively prevented, a temporary STS certificate containing a dynamic ACCESSKEY and Token is obtained through getToken functions, a minimum authority principle is realized, and long-term key leakage risks are avoided.

Description

Method and system for calculating earth excavation and filling directions based on three-dimensional reconstruction of unmanned aerial vehicle Technical Field The invention relates to the technical field of digital construction of civil engineering, in particular to a method and a system for calculating earth excavation and filling based on three-dimensional reconstruction of an unmanned aerial vehicle, which are suitable for measuring and calculating the earth works and monitoring the construction in the fields of building engineering, road construction, mining and the like. Background With the development of the age, the conventional two-dimensional image cannot meet the development requirement at present, and the three-dimensional reconstruction technology is widely focused. In the accelerating development of urban construction in China, when the data acquisition of the geographical information of the construction sites is carried out, the data are mainly drawn and marked on the sites by manpower, and the traditional mode consumes a lot of manpower, material resources and financial resources, and the drawn geographical data information has larger deviation, so that the requirements on the accuracy and the high efficiency of the geographical information are difficult to meet. The traditional operation mode relies on optical instruments such as a total station and a level to conduct manual point-by-point mapping, systematic defects are exposed in actual engineering, firstly, measurement efficiency is limited by the complexity of terrain, the conventional operation needs to be put into a professional team for continuous field data acquisition, special geological units often need to be checked for multiple times, secondly, data precision is doubly limited by human operation errors and instrument performance, capturing capability of fine terrain changes and complex curved surface features is insufficient, thirdly, space-time cutting exists between measurement results and construction processes, instant dynamic analysis capability of filling and excavating amount is lacking, fourthly, human input and safety risk are obviously increased along with the severe degree of an operation environment, and the application of the traditional mapping mode in a dangerous area is obviously limited. The prior art complementary solution has not yet formed a complete closed loop. Satellite remote sensing has wide area monitoring advantages, but is limited by meteorological conditions and data updating frequency, so that data timeliness required by engineering is difficult to ensure, and a vehicle-mounted mobile measurement system is limited by road traffic capacity and is limited in application in unstructured terrains. Although the unmanned aerial vehicle mapping breaks through the bottleneck of the traditional operation efficiency, the problems of lengthy data processing flow, insufficient adaptability of an earthwork calculation model, weak interaction function of a visual platform and the like generally exist, and the comprehensive requirements of the engineering site on high-precision dynamic monitoring, intelligent analysis decision and multi-dimensional data fusion are difficult to support, so that the improvement of the modern engineering management level is restricted. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a method and a system for calculating earth excavation and filling based on unmanned aerial vehicle three-dimensional reconstruction, which are used for constructing a complete processing flow covering links such as coordinate system conversion, noise filtration, attribute enhancement and the like by fusing laser radar point cloud processing and multi-view geometric three-dimensional reconstruction technology, remarkably optimizing the quality of point cloud data and model precision, fully considering the adaptability of complex terrains and severe environments, stably obtaining high-precision data in scenes such as vegetation dense, topography fluctuation and the like, and providing a reliable basis for engineering planning. The technical scheme provided by the invention is that the earth excavation and filling calculation method based on three-dimensional reconstruction of the unmanned aerial vehicle comprises the following steps: (1) Carrying out aerial photography by an unmanned aerial vehicle carrying a laser radar and an RGB camera to obtain multi-angle high-precision point cloud data and aerial photography images; (2) Preprocessing the point cloud data by using PDAL tool chains to generate a standardized point cloud file; (3) Three-dimensional reconstruction modeling, namely generating a high-precision three-dimensional model containing elevation information by adopting a multi-view stereoscopic vision technology, and synchronously constructing a two-dimensional terrain model; (4) Performing model visualization based on a visualization system developed by a NET