A system for checking the precision of real scene reproduction by combining a UAV surveying and mapping with 3D printing

Abstract

The application discloses a kind of unmanned plane surveying and mapping and 3D printing combined real scene reproduction precision checking system, for the one-way transmission of multiple module data in the prior art, point cloud fusion precision is insufficient, model mapping exists microcosmic deviation, printing control is insufficient and error analysis is rough, leading to the problem of low reproduction precision, the present application is by integrating unmanned plane data acquisition module, point cloud generation module, three-dimensional model construction module, 3D printing control module, real scene reproduction module, precision comparison module, data storage module and system management module, and utilize high-frequency data bus and time stamp synchronization to realize the closed-loop interaction and data fusion between each module.The system can complete from multi-source data acquisition, high-precision point cloud generation, printable three-dimensional grid model construction, intelligent printing path planning to reproduction body precision comparison and closed-loop feedback correction whole process, realizes the spatial precision closed-loop verification and micron level control of unmanned plane surveying and mapping data and 3D printing reproduction data.

Description

Technical Field

[0001] This invention belongs to the field of real-scene reproduction technology, and in particular relates to a real-scene reproduction accuracy verification system that combines UAV mapping and 3D printing. Background Technology

[0002] Unmanned aerial vehicle (UAV) mapping technology has developed rapidly in recent years, evolving from early acquisition of spatial information from single aerial images to high-precision 3D point cloud data acquisition systems that integrate LiDAR, multispectral sensors, and inertial navigation. Simultaneously, 3D printing technology has expanded from simple model printing to sophisticated industrial-grade reproduction, supporting complex geometries and diverse materials. Against this backdrop, real-world reconstruction technology combining UAV mapping and 3D printing has emerged. By reconstructing 3D models from field-collected data and then physically printing them, it provides intuitive physical reference models for applications such as urban planning, cultural relic preservation, and post-disaster assessment. To improve the accuracy of the reconstruction, systematic data closure, coordinate mapping, and accuracy verification methods are gradually being introduced into this field.

[0003] However, existing real-scene reconstruction technologies still have a series of significant shortcomings in practical applications, limiting further improvements in their accuracy and automation levels. First, data transmission between functional modules in the system, such as data acquisition, processing, and printing control, is mostly unidirectional or loosely coupled, failing to form an effective real-time closed loop. This results in errors not being detected and corrected in a timely manner during the process. Second, in the data fusion stage, the fusion accuracy of multi-source point clouds, including imagery and LiDAR data, is limited by sensor resolution and algorithm processing capabilities, leading to poor compensation for missing areas in the local point cloud and affecting the integrity of the 3D model. Furthermore, in the 3D model construction stage, insufficient optimization of the mesh topology and limited mapping accuracy from the mapping coordinate system to the printing coordinate system often introduce geometric deviations at the microscale, causing distortions in detail between the printed entity and the original scene.

[0004] Furthermore, in the printing control stage, path planning often relies on manual experience or limited optimization algorithms, making it difficult to achieve dynamic and intelligent matching with high-precision surveying data, thus affecting the consistency and reliability of reproduction. Regarding the final accuracy verification, existing error analysis methods are mostly limited to global statistics, lacking the ability for multi-scale, partition-weighted, refined processing, preventing the closed-loop verification system from fully utilizing the spatial distribution characteristics of error data to guide iterative model optimization. Finally, from a data management perspective, the data at each stage of the entire process, including raw data, intermediate point clouds, 3D models, and coordinate information, lacks a unified and rigorous spatiotemporal correlation and traceability mechanism, which poses difficulties for continuous system optimization and problem diagnosis. These shortcomings collectively make it difficult for existing technologies to achieve stable and high-precision real-world reproduction, especially in serious application scenarios requiring micron-level control precision. Therefore, there is an urgent need for a systematic solution that can achieve full-process closed-loop control, high-precision data mapping, and intelligent error verification. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a real-scene reproduction accuracy verification system that combines UAV mapping and 3D printing, thereby resolving the issues present in the existing technologies.

[0006] In a first aspect, the present invention also provides a real-scene reproduction accuracy verification system combining UAV mapping and 3D printing, comprising: The data synchronization acquisition module is used to synchronously acquire multi-source mapping data streams with timestamps via UAVs; The point cloud fusion processing module is used to fuse the multi-source mapping data streams to generate a preliminary three-dimensional point cloud. The model building and mapping module is used to build a printable 3D mesh model based on the preliminary 3D point cloud and map the 3D mesh model to the printing coordinate system; The printing execution control module is used to plan the printing path and execute 3D printing based on the mapped 3D mesh model to generate a real-world replica. The accuracy comparison and analysis module is used to collect the three-dimensional shape data of the real scene reproduction, compare it with the three-dimensional mesh model, and generate error information; The closed-loop feedback management module is used to adjust the construction parameters of the 3D mesh model or the planning parameters of the printing path based on the error information.

[0007] Optionally, the data synchronization acquisition module includes: The multi-source sensing unit consists of a multispectral image acquisition unit, a lidar scanning unit, and an inertial navigation unit, and is used to synchronously acquire raw data; A timing control unit is used to coordinate the working timing of each unit in the multi-source sensing unit through a unified timing controller. The timestamp marking unit is used to add timestamps to the data packets collected by each unit using an embedded time synchronization chip, forming a multi-source raw data stream with strict spatiotemporal correspondence.

[0008] Optionally, the point cloud fusion processing module includes: A data preprocessing unit is used to perform coordinate transformation and filtering on the timestamped multi-source raw data stream through a highly parallel data fusion unit; The grid fusion unit is used to perform grid-based fusion of filtered multi-source data using an adaptive spatial resolution algorithm. The density optimization unit is used to perform weighted processing on point cloud density anomalies using a distributed GPU cluster to form a preliminary point cloud.

[0009] Optionally, the model construction mapping module includes: The mesh reconstruction unit is used to reconstruct the preliminary point cloud using a block-based iterative algorithm to generate a triangular mesh. A topology optimization unit is used to adjust the connectivity of the triangular mesh through a topology optimization processor to form a continuous and closed mesh topology. The coordinate mapping unit is used to map the coordinates of the mesh topology to the printing coordinate system using a quaternion rotation algorithm and a homogeneous transformation matrix.

[0010] Optionally, the printing execution control module includes: The layered slicing unit is used to perform adaptive layered slicing based on the mapped 3D mesh model to generate printing layer data; The path planning unit is used to calculate the movement speed and material injection amount of the path points in the printing layer data through a dynamic trajectory optimization algorithm. The instruction generation unit is used to convert the optimized path and parameters into printer control instructions to drive the printing device to generate the real-scene reproduction.

[0011] Optionally, the accuracy comparison and analysis module includes: The topography verification unit is used to acquire verification point cloud data of the real scene reproduction object through a structured light scanning device; The coordinate alignment unit is used to perform coordinate alignment matching between the verification point cloud data and the reference point cloud of the three-dimensional mesh model using an iterative nearest point algorithm; The error calculation unit is used to calculate the spatial coordinate difference between corresponding point pairs after matching, and generate an error matrix containing the offset of each point.

[0012] Optionally, the closed-loop feedback management module includes: The error analysis unit is used to perform local and global analysis on the error matrix using a partitioned grid weighting algorithm and a multi-scale statistical analyzer. The data transmission unit is used to transmit the analysis results to the model building mapping module and the printing execution control module via a high-frequency data bus; The parameter adjustment unit is used to adjust the construction parameters of the three-dimensional mesh model or the planning parameters of the printing path based on the analysis results.

[0013] In a second aspect, the present invention also provides a computer terminal device, comprising: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the real-scene reproduction accuracy verification system combining UAV mapping and 3D printing in the first aspect above.

[0014] Thirdly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the real-scene reproduction accuracy verification system combining UAV mapping and 3D printing described in the first aspect above.

[0015] Fourthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the real-scene reproduction accuracy verification system combining UAV mapping and 3D printing described in the first aspect above.

[0016] Compared with the prior art, the present invention has the following advantages and technical effects: This invention provides a real-scene reproduction accuracy verification system combining UAV mapping and 3D printing. By introducing a high-frequency data bus and timestamp synchronization mechanism, it achieves real-time closed-loop interaction and data fusion between UAV mapping, 3D model construction, 3D printing, and accuracy comparison, effectively overcoming the shortcomings of unidirectional and loosely coupled data transmission between modules in existing technologies. The system employs an adaptive spatial resolution algorithm and distributed processing technology, significantly improving the accuracy and integrity of multi-source point cloud fusion, and utilizes a block-based iterative reconstruction and topology optimization processor to generate a continuous, closed, high-quality 3D mesh. Quaternion rotation and homogeneous matrix mapping ensure accurate conversion from mapping coordinates to printing coordinates. A dynamic trajectory optimization algorithm combined with real-time error feedback enables intelligent matching and fine-tuning of the printing path and mapping data. Simultaneously, an error processing method based on partitioned grid weighting and multi-scale statistical analysis provides a refined closed-loop verification basis for the system. Ultimately, this invention achieves high-precision, adaptive closed-loop control throughout the entire process from data acquisition to entity reproduction, significantly improving the spatial consistency and microscopic geometric fidelity of real-scene reproduction. Attached Figure Description

[0017] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the structure of a real-scene reproduction accuracy verification system combining UAV mapping and 3D printing according to an embodiment of the present invention. Detailed Implementation

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0020] like Figure 1 As shown, this embodiment provides a real-scene reproduction accuracy verification system combining UAV mapping and 3D printing, including: The data synchronization acquisition module is used to synchronously acquire multi-source mapping data streams with timestamps via UAVs; The point cloud fusion processing module is used to fuse the multi-source mapping data streams to generate a preliminary three-dimensional point cloud. The model building and mapping module is used to build a printable 3D mesh model based on the preliminary 3D point cloud and map the 3D mesh model to the printing coordinate system; The printing execution control module is used to plan the printing path and execute 3D printing based on the mapped 3D mesh model to generate a real-world replica. The accuracy comparison and analysis module is used to collect the three-dimensional shape data of the real scene reproduction, compare it with the three-dimensional mesh model, and generate error information; The closed-loop feedback management module is used to adjust the construction parameters of the 3D mesh model or the planning parameters of the printing path based on the error information.

[0021] Furthermore, the data synchronization acquisition module includes: The multi-source sensing unit consists of a multispectral image acquisition unit, a lidar scanning unit, and an inertial navigation unit, and is used to synchronously acquire raw data; A timing control unit is used to coordinate the working timing of each unit in the multi-source sensing unit through a unified timing controller. The timestamp marking unit is used to add timestamps to the data packets collected by each unit using an embedded time synchronization chip, forming a multi-source raw data stream with strict spatiotemporal correspondence.

[0022] Furthermore, the point cloud fusion processing module includes: A data preprocessing unit is used to perform coordinate transformation and filtering on the timestamped multi-source raw data stream through a highly parallel data fusion unit; The grid fusion unit is used to perform grid-based fusion of filtered multi-source data using an adaptive spatial resolution algorithm. The density optimization unit is used to perform weighted processing on point cloud density anomalies using a distributed GPU cluster to form a preliminary point cloud.

[0023] Furthermore, the model construction mapping module includes: The mesh reconstruction unit is used to reconstruct the preliminary point cloud using a block-based iterative algorithm to generate a triangular mesh. A topology optimization unit is used to adjust the connectivity of the triangular mesh through a topology optimization processor to form a continuous and closed mesh topology. The coordinate mapping unit is used to map the coordinates of the mesh topology to the printing coordinate system using a quaternion rotation algorithm and a homogeneous transformation matrix.

[0024] Furthermore, the printing execution control module includes: The layered slicing unit is used to perform adaptive layered slicing based on the mapped 3D mesh model to generate printing layer data; The path planning unit is used to calculate the movement speed and material injection amount of the path points in the printing layer data through a dynamic trajectory optimization algorithm. The instruction generation unit is used to convert the optimized path and parameters into printer control instructions to drive the printing device to generate the real-scene reproduction.

[0025] Furthermore, the accuracy comparison and analysis module includes: The topography verification unit is used to acquire verification point cloud data of the real scene reproduction object through a structured light scanning device; The coordinate alignment unit is used to perform coordinate alignment matching between the verification point cloud data and the reference point cloud of the three-dimensional mesh model using an iterative nearest point algorithm; The error calculation unit is used to calculate the spatial coordinate difference between corresponding point pairs after matching, and generate an error matrix containing the offset of each point.

[0026] Furthermore, the closed-loop feedback management module includes: The error analysis unit is used to perform local and global analysis on the error matrix using a partitioned grid weighting algorithm and a multi-scale statistical analyzer. The data transmission unit is used to transmit the analysis results to the model building mapping module and the printing execution control module via a high-frequency data bus; The parameter adjustment unit is used to adjust the construction parameters of the three-dimensional mesh model or the planning parameters of the printing path based on the analysis results.

[0027] Specifically, the implementation process of this embodiment includes: In this embodiment, the UAV data acquisition module generates a multi-source raw data stream with timestamps through a multispectral image acquisition unit, a lidar scanning unit, and an inertial navigation unit; the point cloud generation module performs coordinate transformation, filtering, and fusion of the multi-source raw data streams with the multi-source point cloud through a high-parallel data fusion unit to form a preliminary point cloud; the 3D model construction module generates a printable 3D mesh model through a point cloud fusion unit, a mesh reconstruction unit, and a coordinate transformation unit, and maps the printable 3D mesh model to the 3D printing coordinate system; the 3D printing control module receives the mapped printable 3D mesh model and forms printing control commands through a layer slicing unit, a path planning unit, and a printing command generation unit; the real-scene reproduction module forms a closed-loop interaction with the UAV mapping data and the 3D printing process through a high-frequency data bus and timestamp synchronization; the accuracy comparison module compares the reproduced object with the original 3D model through a scanning verification unit, a coordinate alignment unit, and an error analysis unit to generate an error matrix; the data storage module saves the raw data, point cloud, 3D model, and coordinate information of each stage; and the system management module coordinates the data transmission and synchronization of each module through closed-loop feedback logic and an iterative update mechanism.

[0028] This embodiment achieves spatial accuracy closed-loop verification between UAV mapping data and 3D printing reproduction data through high-frequency closed-loop, multimodal data mapping, timestamp synchronization, and coordinate continuity maintenance. The data interface protocol between modules adopts a unified high-frequency real-time data bus standard, supporting at least 10,000 data packet transmissions per second, with a data bus bandwidth of no less than 10Gbps. The multi-source raw data stream output by the UAV data acquisition module is transmitted in real-time to the point cloud generation module via this high-frequency bus. The point cloud generation module has a dual-buffering mechanism to ensure data stream continuity and packet loss-free processing. Timestamp accuracy is achieved through an embedded nanosecond-level synchronization chip, ensuring strict spatial and temporal correspondence between multispectral imagery, LiDAR, and inertial navigation data.

[0029] In this embodiment, the 3D model building module and the 3D printing control module exchange data through a dual-channel interface of shared memory area and distributed file system. The 3D mesh model and its mapped coordinate information are accessed with low latency in shared memory, while the distributed file system provides redundant storage to support iterative access by multiple modules and historical data tracing. The system management module, as the central scheduling unit, periodically polls the data status of each module and iteratively updates it according to the closed-loop feedback logic. It also embeds an anomaly detector to detect and recover from data anomalies, communication delays, or module failures, ensuring the stability and continuity of the entire closed-loop system.

[0030] In this embodiment, during the closed-loop interaction between the real-scene reproduction module and the accuracy comparison module, the scanning verification data acquisition frequency per second is synchronized with the 3D printing layer refresh frequency. A timestamp-based synchronization algorithm is used to ensure spatial correspondence. The closed-loop triggering conditions include the completion of the printing layer, the update of the 3D model, or the completion of the error matrix calculation. The accuracy comparison module transmits the generated error matrix back to the 3D model construction module and the 3D printing control module in real time via a high-frequency bus. The system management module triggers iterative adjustments based on the error threshold to ensure that each module maintains a spatial accuracy closed loop under multimodal data mapping and supports continuous verification across multiple scales and regions.

[0031] In this embodiment, the multispectral image acquisition unit adopts an adjustable band sensor array, the lidar scanning unit consists of a high-frequency pulsed laser transmitter and a rotating scanning mirror, and the inertial navigation unit is a combination of a six-axis inertial measurement unit and a GNSS receiver. The multispectral image acquisition unit, lidar scanning unit and inertial navigation unit form a joint data stream through a unified timing controller and a distributed storage bus. The embedded time synchronization chip marks each data packet with a nanosecond-level timestamp to ensure strict spatiotemporal correspondence of multi-source data and realize data closed-loop traceability.

[0032] The multispectral image acquisition unit uses a parallel FPGA processor to achieve real-time data compression and noise reduction. The lidar scanning unit performs dynamic attitude compensation for each lidar point under the action of an embedded real-time coordinate corrector. The six-axis measurement data and GNSS data of the inertial navigation unit are fused with Kalman filtering algorithm to generate a corrected point cloud with three-dimensional attitude information, providing high-precision spatial basic data for the point cloud generation module.

[0033] The rotating scanning mirror of the lidar scanning unit provides real-time feedback on the scanning angle through a high-precision encoder. The scanning pulse interval is automatically adjusted by an embedded pulse controller to ensure the uniformity of the point cloud. The inertial navigation unit outputs the multispectral image data synchronously to form a spatially corresponding point cloud set.

[0034] The point cloud fusion unit performs mesh fusion using an adaptive spatial resolution algorithm, and during the fusion process, it uses a distributed GPU cluster to weight regions with abnormal point cloud density. The mesh reconstruction unit uses a block-based iterative algorithm to generate a printable 3D mesh model and automatically fills in local missing point cloud regions. The coordinate transformation unit uses quaternion rotation and homogeneous matrix mapping to transfer the printable 3D mesh model to the 3D printing coordinate system.

[0035] Specifically, in the point cloud fusion unit, the adaptive spatial resolution algorithm dynamically adjusts the point cloud density in different regions by establishing a three-dimensional voxel grid. Its mathematical model is as follows: ; in, Indicates the current voxel side length. The initial voxel reference length, This is the density adjustment coefficient. This represents the current local point cloud density. The global average point cloud density is used. The GPU cluster adopts a distributed thread scheduling strategy, dividing the entire point cloud space into multiple computing blocks. Each GPU node dynamically allocates thread resources according to the number of points in the block. Local caches achieve fast swapping through shared video memory. During the calculation process, a weighted compensation method is used for areas with abnormal point cloud density. The weight of points in sparse areas is increased to enhance continuity, and dense areas are appropriately downsampled to maintain uniform distribution.

[0036] In the mesh reconstruction unit, the block-based iterative algorithm divides the overall point cloud into several overlapping blocks for independent reconstruction. Each block uses a combination of Delaunay triangulation and Poisson surface reconstruction. The iterative convergence condition is that the boundary error between adjacent blocks is less than a threshold. The iteration count may reach its maximum limit Nmax, where ϵ represents the boundary point coordinate error limit and Nmax is the maximum number of iterations. The missing point compensation logic generates replacement points based on local neighborhood interpolation and spline fitting, while simultaneously updating mesh connectivity to maintain the overall closed structure. The topology optimization processor calculates the adjacency set of each node and executes patch rearrangement rules to ensure that each node connects to at least three valid patches to form a continuous topology; a spatial octree index structure is used to optimize local mesh accuracy, achieving mesh refinement and boundary smoothing.

[0037] When generating triangular meshes, the mesh reconstruction unit combines the topology optimization processor to adjust the mesh connectivity, automatically generating a continuous and closed mesh topology structure, and enhances the local accuracy of the mesh through a spatial index structure. During the mapping process, the coordinate transformation unit simultaneously performs rotation, scaling and translation matrix calculations to ensure that the printed coordinates correspond to the survey coordinates at the micrometer level, achieving closed-loop collaboration with the error analysis module.

[0038] Specifically, in the coordinate transformation unit, quaternion rotation uses the following formula: ; in, The quaternion representation of the original vertex coordinates. For attitude rotation quaternions, Let it be its conjugate inverse. Let be the coordinates of the rotated vertex. Homogeneous matrix mapping uses the formula: ; in, for Homogeneous transformation matrix, including rotation matrix Scaling matrix With translation vector It comprehensively performs coordinate rotation, scaling adjustment, and translation. Error compensation corrects coordinate drift through least-squares fitting of local feature points, and accuracy verification uses the distance measurement results of standard control points as a reference to ensure that the surveyed coordinates and printed coordinates correspond completely within ±5μm, thereby achieving real-time linkage between coordinate-level micron-level precision closed loop and error analysis module.

[0039] The layer slicing unit generates adaptive printing layer data based on the printable 3D network model. The path planning unit calculates the movement speed and jet volume of the printing path points through a dynamic trajectory optimization algorithm and receives the error matrix from the accuracy comparison module in real time for fine-tuning. The printing instruction generation unit converts the optimized path into printer control instructions, forming a closed-loop iterative printing control process to achieve continuous mapping between printing accuracy and surveying data.

[0040] Specifically, the layered slicing unit generates adaptive printing layer data based on the printable 3D network model. The layered slicing unit employs an adaptive layer thickness adjustment strategy, where the layer thickness... Based on local surface curvature With material shrinkage coefficient calculate: ; in: Base layer thickness; This is the adjustment coefficient; This indicates the surface curvature corresponding to the current slice layer; This represents the linear shrinkage ratio of the printing material.

[0041] Layer slicing units are based on different material parameters (including viscosity). Curing rate Coefficient of thermal expansion The slice layer data is corrected and compared with the error matrix. Closed-loop interaction, through offset correction function Layer thickness is updated in real time.

[0042] The path planning unit calculates the movement speed of the printed path points using a dynamic trajectory optimization algorithm. and injection volume Its input includes path point coordinates Layer thickness Material parameters The output is a path control instruction set. .

[0043] Speed ​​of movement With injection volume Calculate using the following formula: ; in: These are standard speed and injection volume, respectively. This is a dynamic adjustment coefficient; Let be the gradient magnitude of the error matrix, representing the rate of change of the local error; This is the offset of the corresponding print point.

[0044] The real-time error feedback adjustment mechanism is implemented through the control loop. Compare current error Error with target ,when Automatic adjustment and This enables dynamic accuracy compensation and continuous closed-loop control during the printing process.

[0045] The print instruction generation unit will optimize the path Convert to printer control commands This forms a closed-loop iterative printing control process, enabling continuous mapping between printing accuracy and surveying data.

[0046] The scanning and verification unit generates real-world point cloud data through structured light scanning, the coordinate alignment unit uses a high-precision iterative nearest point algorithm for point-by-point matching, and the error analysis unit generates an error matrix containing the offset and distribution matrix of each coordinate point through spatial difference, and stores it in a multi-dimensional array format for the system management module to call, thereby realizing the generation of basic data for accuracy closed loop.

[0047] Specifically, the scanning verification unit generates cloud data of the real-world scenic spot through structured light scanning, wherein the structured light scanning light source frequency... Scan angle resolution Point cloud sampling employs a hierarchical adaptive sampling strategy, increasing the sampling density in high curvature regions. Maintain basic density in low curvature areas .

[0048] The coordinate alignment unit uses a high-precision Iterative Closest Point (ICP) algorithm for point-by-point matching. Its iteration termination condition is when the average matching error or number of iterations Stop when the time comes.

[0049] Matching error matrix Calculate using the following formula: ; in: : The coordinates of the real-world scenic spot cloud; : Corresponding points of the surveying benchmark point cloud; : Number of matching point pairs; Error covariance matrix.

[0050] The error analysis unit generates a value containing the offset of each coordinate point by using spatial interpolation. and distribution matrix The error matrix is ​​generated and stored in a multidimensional array format for the system management module to access, enabling the generation of basic data for closed-loop accuracy. The error analysis unit uses a partitioned grid weighting algorithm and a multi-scale statistical analyzer to perform local and global analysis on the error matrix. A high-frequency data bus transmits the analysis results to the 3D model building module and the 3D printing control module. The system management module iteratively updates the data of all modules through timestamp synchronization and closed-loop feedback logic, realizing the spatial accuracy closed-loop verification and continuous mapping of multimodal data between UAV mapping data and 3D printing reproduction data.

[0051] Specifically, the error analysis unit utilizes a partitioned grid weighted algorithm and a multi-scale statistical analyzer to perform local and global analysis of the error matrix. The partitioning strategy divides the complex reality into sections based on the three-dimensional spatial coordinate axes. Each sub-grid cell has a weight. Based on local point density and error variance calculate: ; in For smoothing terms. The multi-scale statistical analyzer applies this to the global error matrix. Perform multi-scale decomposition and calculate the mean square error at each scale. and through the error transmission protocol The analysis results are mapped to each module. The high-frequency data bus transmits the analysis results to the 3D model building module and the 3D printing control module. The system management module iteratively updates the data of all modules through timestamp synchronization and closed-loop feedback logic, realizing the spatial accuracy closed-loop verification of UAV mapping data and 3D printing reproduction data and continuous mapping of multimodal data.

[0052] In this embodiment, a computer terminal device is provided, including: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the above-described real-scene reproduction accuracy verification system that combines UAV mapping and 3D printing.

[0053] In this embodiment, a computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the above-described real-scene reproduction accuracy verification system that combines UAV mapping and 3D printing.

[0054] In this embodiment, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of the above-described real-scene reproduction accuracy verification system combining UAV mapping and 3D printing.

[0055] The aforementioned program can run on a processor or be stored in memory (or a computer-readable medium). Computer-readable media includes both permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random-access memory (SRAM), dynamic random-access memory (DRAM), other types of random-access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0056] These computer programs may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes can be implemented by different modules for different steps.

[0057] This invention provides a real-scene reproduction accuracy verification system combining UAV mapping and 3D printing. By introducing a high-frequency data bus and timestamp synchronization mechanism, it achieves real-time closed-loop interaction and data fusion between UAV mapping, 3D model construction, 3D printing, and accuracy comparison, effectively overcoming the shortcomings of unidirectional and loosely coupled data transmission between modules in existing technologies. The system employs an adaptive spatial resolution algorithm and distributed processing technology, significantly improving the accuracy and integrity of multi-source point cloud fusion, and utilizes a block-based iterative reconstruction and topology optimization processor to generate a continuous, closed, high-quality 3D mesh. Quaternion rotation and homogeneous matrix mapping ensure accurate conversion from mapping coordinates to printing coordinates. A dynamic trajectory optimization algorithm combined with real-time error feedback enables intelligent matching and fine-tuning of the printing path and mapping data. Simultaneously, an error processing method based on partitioned grid weighting and multi-scale statistical analysis provides a refined closed-loop verification basis for the system. Ultimately, this invention achieves high-precision, adaptive closed-loop control throughout the entire process from data acquisition to entity reproduction, significantly improving the spatial consistency and microscopic geometric fidelity of real-scene reproduction.

[0058] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A real-scene reproduction accuracy verification system combining UAV mapping and 3D printing, characterized in that, Includes the following steps: The data synchronization acquisition module is used to synchronously acquire multi-source mapping data streams with timestamps via UAVs; The point cloud fusion processing module is used to fuse the multi-source mapping data streams to generate a preliminary three-dimensional point cloud. The model building and mapping module is used to build a printable 3D mesh model based on the preliminary 3D point cloud and map the 3D mesh model to the printing coordinate system; The printing execution control module is used to plan the printing path and execute 3D printing based on the mapped 3D mesh model to generate a real-world replica. The accuracy comparison and analysis module is used to collect the three-dimensional shape data of the real scene reproduction, compare it with the three-dimensional mesh model, and generate error information; The closed-loop feedback management module is used to adjust the construction parameters of the 3D mesh model or the planning parameters of the printing path based on the error information. The data synchronization acquisition module includes: The multi-source sensing unit consists of a multispectral image acquisition unit, a lidar scanning unit, and an inertial navigation unit, and is used to synchronously acquire raw data; A timing control unit is used to coordinate the working timing of each unit in the multi-source sensing unit through a unified timing controller. The timestamp marking unit is used to add timestamps to the data packets collected by each unit using an embedded time synchronization chip, forming a multi-source raw data stream with strict spatiotemporal correspondence. The point cloud fusion processing module includes: A data preprocessing unit is used to perform coordinate transformation and filtering on the timestamped multi-source raw data stream through a highly parallel data fusion unit; The grid fusion unit is used to perform grid-based fusion of filtered multi-source data using an adaptive spatial resolution algorithm. The density optimization unit is used to weight regions with abnormal point cloud density using a distributed GPU cluster to form a preliminary point cloud. The model construction mapping module includes: The mesh reconstruction unit is used to reconstruct the preliminary point cloud using a block-based iterative algorithm to generate a triangular mesh. A topology optimization unit is used to adjust the connectivity of the triangular mesh through a topology optimization processor to form a continuous and closed mesh topology. The coordinate mapping unit is used to map the coordinates of the mesh topology to the printing coordinate system using a quaternion rotation algorithm and a homogeneous transformation matrix; The printing execution control module includes: The layered slicing unit is used to perform adaptive layered slicing based on the mapped 3D mesh model to generate printing layer data; The path planning unit is used to calculate the movement speed and material injection amount of the path points in the printing layer data through a dynamic trajectory optimization algorithm. The instruction generation unit is used to convert the optimized path and parameters into printer control instructions to drive the printing device to generate the real-scene reproduction. The accuracy comparison and analysis module includes: The topography verification unit is used to acquire verification point cloud data of the real scene reproduction object through a structured light scanning device; The coordinate alignment unit is used to perform coordinate alignment matching between the verification point cloud data and the reference point cloud of the three-dimensional mesh model using an iterative nearest point algorithm; The error calculation unit is used to calculate the spatial coordinate difference of corresponding point pairs after matching and generate an error matrix containing the offset of each point. The closed-loop feedback management module includes: The error analysis unit is used to perform local and global analysis on the error matrix using a partitioned grid weighting algorithm and a multi-scale statistical analyzer. The data transmission unit is used to transmit the analysis results to the model building mapping module and the printing execution control module via a high-frequency data bus; The parameter adjustment unit is used to adjust the construction parameters of the three-dimensional mesh model or the planning parameters of the printing path based on the analysis results.

2. A computer terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors perform the steps of the system as claimed in claim 1.

3. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the system as described in claim 1.

4. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the system of claim 1.

Images