Unmanned aerial vehicle surveying and mapping and 3D printing combined real scene reproduction precision checking system

Abstract

The invention discloses an unmanned aerial vehicle surveying and mapping and 3D printing combined real scene reproduction precision checking system, and aims to solve the problems of multi-module data one-way transmission, insufficient point cloud fusion precision, microscopic deviation of model mapping, insufficient printing control intelligence and low reproduction precision caused by rough error analysis in the prior art. According to the invention, an unmanned aerial vehicle data acquisition module, a point cloud generation module, a three-dimensional model construction module, a 3D printing control module, a real scene reproduction module, a precision comparison module, a data storage module and a system management module are integrated, and a high-frequency data bus and a timestamp are utilized to synchronously realize closed-loop interaction and data fusion among the modules. According to the system, the whole process 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 can be completed, and spatial precision closed-loop verification and micron-level control of unmanned aerial vehicle surveying and mapping data and 3D printing reproduction data are achieved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of real scene reproduction, and particularly relates to a real scene reproduction precision checking system combining unmanned aerial vehicle surveying and 3D printing. BACKGROUND

[0002] In recent years, unmanned aerial vehicle surveying technology has developed rapidly. It has evolved from early single aerial image acquisition of spatial information to high-precision three-dimensional point cloud data acquisition system integrating laser radar, multi-spectral sensor and inertial navigation. At the same time, three-dimensional printing technology has also expanded from simple model printing to fine industrial-level reproduction, which can support complex geometric structures and diversified materials. Under this background, the real scene reproduction technology combining unmanned aerial vehicle surveying and 3D printing began to appear. Through reconstructing the data collected in the field into a three-dimensional model and printing it into a physical entity, it provides an intuitive physical reference model for application fields such as urban planning, cultural relic protection and post-disaster assessment. In order to improve the accuracy of reproduction, systematic data closed loop, coordinate mapping and precision checking method are gradually introduced into this field.

[0003] However, the existing real scene reproduction technology still has a series of obvious deficiencies in practical application, which limits the further improvement of its precision and automation level. First, the data transmission between each functional module in the system, such as data acquisition, processing and printing control, is mostly one-way or loosely coupled, which cannot form an effective real-time closed loop, so that errors cannot be discovered and corrected in time in the process. Secondly, in the data fusion stage, the fusion precision of multi-source point cloud, including image and laser radar data, is limited by the resolution of the sensor and the processing capacity of the algorithm, and the compensation effect for the local point cloud missing area is not good, which affects the integrity of the three-dimensional model. Thirdly, in the three-dimensional model construction link, the optimization of the grid topology structure is insufficient and the mapping precision from the surveying coordinate system to the printing coordinate system is limited, which often introduces geometric deviation on the micro scale, causing the printed entity to be distorted in detail compared with 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 configured to coordinate the working timing of each of the multiple-source sensing units through a unified timing controller; a timestamp marking unit configured to add a timestamp to the data packets collected by each of the units by using an embedded time synchronization chip, so as to form a multi-source original data stream with strict space-time correspondence.

[0008] Optionally, the point cloud fusion processing module comprises: a data preprocessing unit configured to perform coordinate conversion and filtering on the multi-source original data stream with timestamps by using a high-parallel data fusion unit; a grid fusion unit configured to perform grid fusion on the filtered multi-source data by using an adaptive spatial resolution algorithm; a density optimization unit configured to perform weighted processing on the point cloud density abnormal area by using a distributed GPU cluster, so as to form a preliminary point cloud.

[0009] Optionally, the model construction mapping module comprises: a grid reconstruction unit configured to perform grid reconstruction on the preliminary point cloud by using a block iteration algorithm, so as to generate a triangular mesh; a topology optimization unit configured to adjust the connectivity of the triangular mesh by using a topology optimization processor, so as to form a continuous and closed grid topology structure; a coordinate mapping unit configured to map the coordinates of the grid topology structure to the printing coordinate system by using a quaternion rotation algorithm and a homogeneous transformation matrix.

[0010] Optionally, the printing execution control module comprises: a layered slicing unit configured to perform adaptive layered slicing on the mapped three-dimensional mesh model, so as to generate printing layer data; a path planning unit configured to calculate the motion speed of the path points and the material ejection amount in the printing layer data by using a dynamic trajectory optimization algorithm; an instruction generation unit configured to convert the optimized path and parameters into a printer control instruction, so as to drive a printing device to generate the real scene reproduction.

[0011] Optionally, the precision comparison and analysis module comprises: a topography review unit configured to obtain review point cloud data of the real scene reproduction by using a structured light scanning device; a coordinate alignment unit configured to perform coordinate alignment and matching of the review point cloud data and the reference point cloud of the three-dimensional mesh model by using an iterative closest point algorithm; an error calculation unit configured to calculate the spatial coordinate difference of the corresponding point pairs after matching, so as to generate an error matrix containing the offset of each point.

[0012] Optionally, the closed-loop feedback management module comprises: an error analysis unit configured to locally and globally analyze the error matrix by using a partitioned grid weighting algorithm and a multi-scale statistical analyzer; a data transmission unit configured to transmit the analysis result to the model construction mapping module and the printing execution control module through a high-frequency data bus; a parameter adjustment unit configured to adjust a construction parameter of the three-dimensional grid model or a planning parameter of the printing path based on the analysis result.

[0013] In a second aspect, the present application further provides a computer terminal device, comprising: one or more processors; a memory coupled to the processor and configured to store 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 unmanned aerial vehicle surveying and mapping and 3D printing combined real scene reproduction precision checking system in the first aspect.

[0014] In a third aspect, the present application further provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the steps of the unmanned aerial vehicle surveying and mapping and 3D printing combined real scene reproduction precision checking system in the first aspect.

[0015] In a fourth aspect, the present application further provides a computer program product comprising a computer program, wherein the computer program is executed by a processor to implement the steps of the unmanned aerial vehicle surveying and mapping and 3D printing combined real scene reproduction precision checking system in the first aspect.

[0016] Compared with the prior art, the present application has the following advantages and technical effects: The unmanned aerial vehicle surveying and 3D printing combined real scene reproduction precision checking system provided by the application realizes real-time closed loop interaction and data fusion between unmanned aerial vehicle surveying, three-dimensional model construction, 3D printing and precision comparison by introducing a high frequency data bus and a time stamp synchronization mechanism, and effectively overcomes the defects of one-way data transmission and loose coupling between modules in the prior art. The system uses an adaptive spatial resolution algorithm and a distributed processing technology to significantly improve the precision and integrity of multi-source point cloud fusion, and generates a continuous closed high-quality three-dimensional grid by using a block iteration reconstruction and a topology optimization processor. Through quaternion rotation and homogeneous matrix mapping, the accurate conversion of surveying coordinates to printing coordinates is ensured. The dynamic trajectory optimization algorithm combined with real-time error feedback realizes intelligent matching and fine tuning of the printing path and the surveying data. At the same time, the error processing method based on partition grid weighting and multi-scale statistical analysis provides a fine closed loop checking basis for the system. Finally, the application realizes full-process, high-precision and adaptive closed loop control from data acquisition to entity reproduction, significantly improving the spatial consistency and micro-geometric fidelity of real scene reproduction. BRIEF DESCRIPTION OF DRAWINGS

[0017] The accompanying drawings, which form a part of this application, are intended to provide further understanding of the application and are incorporated herein in their entirety. The schematic embodiments of the application and their descriptions are used to explain the application and do not constitute an improper limitation on the application. In the drawings: Figure 1 FIG. 1 is a structural schematic diagram of an unmanned aerial vehicle surveying and 3D printing combined real scene reproduction precision checking system according to an embodiment of the application. DETAILED DESCRIPTION

[0018] It should be noted that the embodiments in the application and the features in the embodiments can be combined with each other without conflict. The application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0019] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a group of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown herein.

[0020] As shown in FIG. 1, the embodiment provides an unmanned aerial vehicle surveying and 3D printing combined real scene reproduction precision checking system, which comprises: Figure 1 A data synchronization acquisition module for synchronously acquiring multi-source surveying data streams with time stamps by an unmanned aerial vehicle; A point cloud fusion processing module for fusion processing of the multi-source surveying data streams to generate a preliminary three-dimensional point cloud; A three-dimensional model construction module for constructing a three-dimensional model based on the preliminary three-dimensional point cloud; The model construction and mapping module is configured to construct a printable three-dimensional grid model based on the preliminary three-dimensional point cloud and map the three-dimensional grid model to a printing coordinate system; The printing execution control module is configured to plan a printing path according to the mapped three-dimensional grid model and perform 3D printing to generate a real scene reproduction body. The precision comparison and analysis module is configured to collect three-dimensional topographic data of the real scene reproduction body, compare the three-dimensional topographic data with the three-dimensional grid model, and generate error information. The closed-loop feedback management module is configured to feed back and adjust the construction parameters of the three-dimensional grid model or the planning parameters of the printing path according to the error information.

[0021] Further, the data synchronous acquisition module comprises: The multi-source sensing unit is composed of a multi-spectral image acquisition unit, a laser radar scanning unit, and an inertial navigation unit, and is configured to synchronously acquire original data. The timing control unit is configured to coordinate the working timing of each unit in the multi-source sensing unit through a unified timing controller. The timestamp marking unit is configured to add timestamp markers to the data packets collected by each unit using an embedded time synchronization chip to form a multi-source original data stream that strictly corresponds in space and time.

[0022] Further, the point cloud fusion processing module comprises: The data preprocessing unit is configured to perform coordinate conversion and filtering on the multi-source original data stream with timestamps through a high-parallel data fusion unit. The grid fusion unit is configured to perform grid fusion on the filtered multi-source data using an adaptive spatial resolution algorithm. The density optimization unit is configured to perform weighted processing on point cloud density abnormal regions using a distributed GPU cluster to form a preliminary point cloud.

[0023] Further, the model construction and mapping module comprises: The grid reconstruction unit is configured to perform grid reconstruction on the preliminary point cloud using a block iteration algorithm to generate a triangular grid. The topology optimization unit is configured to adjust the connectivity of the triangular grid through a topology optimization processor to form a continuous and closed grid topology structure. The coordinate mapping unit is configured to map the coordinates of the grid topology structure to the printing coordinate system using a quaternion rotation algorithm and a homogeneous transformation matrix.

[0024] Further, the printing execution control module comprises: The layered slicing unit is configured to perform adaptive layered slicing according to the mapped three-dimensional grid model to generate printing layer data. a path planning unit configured to calculate a movement speed and a material ejection amount of a path point in the print layer data by a dynamic trajectory optimization algorithm; an instruction generation unit configured to convert the optimized path and parameters into a printer control instruction to drive the printing device to generate the real scene reproduction.

[0025] Further, the precision comparison and analysis module comprises: a topography review unit configured to acquire review point cloud data of the real scene reproduction by a structured light scanning device; a coordinate alignment unit configured to perform coordinate alignment matching of the review point cloud data and the reference point cloud of the three-dimensional grid model by using an iterative closest point algorithm; an error calculation unit configured to calculate a spatial coordinate difference value of a corresponding point pair after matching to generate an error matrix containing a point offset amount.

[0026] Further, the closed-loop feedback management module comprises: an error analysis unit configured to perform local and global analysis of the error matrix by using a partition grid weighting algorithm and a multi-scale statistical analyzer; a data transmission unit configured to transmit the analysis result to the model construction mapping module and the printing execution control module through a high-frequency data bus; a parameter adjustment unit configured to adjust a construction parameter of the three-dimensional grid model or a planning parameter of the printing path based on the analysis result.

[0027] Specifically, the implementation process of the embodiment comprises: In the embodiment, the unmanned aerial vehicle data acquisition module generates a time-stamped multi-source original data stream through a multi-spectral image acquisition unit, a laser radar scanning unit and an inertial navigation unit; the point cloud generation module performs coordinate conversion, filtering and multi-source point cloud fusion on the multi-source original data stream through a high-parallel data fusion unit to form a preliminary point cloud; the three-dimensional model construction module generates a printable three-dimensional grid model through a point cloud fusion unit, a grid reconstruction unit and a coordinate transformation unit, and maps the printable three-dimensional grid model to a 3D printing coordinate system; the 3D printing control module receives the mapped printable three-dimensional grid model and forms a printing control instruction through a layer slicing unit, a path planning unit and a printing instruction generation unit, and the real scene reproduction module forms a closed-loop interaction with the unmanned aerial vehicle surveying and mapping data and the 3D printing process through a high-frequency data bus and a time stamp; the precision comparison module generates an error matrix by comparing the reproduced object and the original three-dimensional model through a scanning review unit, a coordinate alignment unit and an error analysis unit, the data storage module saves the original data, point cloud, three-dimensional model and coordinate information at each stage, and the system management module coordinates the data transmission and synchronization of each module through a closed-loop feedback logic and an iterative update mechanism.

[0028] The embodiment realizes the spatial accuracy closed loop check of the unmanned aerial vehicle surveying and mapping data and the 3D printing reproduction data by high frequency closed loop, multi-modal data mapping, timestamp synchronization and coordinate continuity maintenance. The data interface protocol between the modules adopts a unified high frequency real-time data bus standard, supports at least 10,000 times of data packet transmission per second, and the data bus bandwidth is not less than 10 Gbps. The multi-source original data stream output by the unmanned aerial vehicle data acquisition module is transmitted in real time to the point cloud generation module through the high frequency bus, the point cloud generation module is provided with a double buffer caching mechanism to ensure data stream continuity and no packet loss processing, the timestamp accuracy is realized through an embedded nanosecond level synchronization chip, and the spatial and temporal correspondence of the multi-spectral image, laser radar and inertial navigation data is ensured.

[0029] In the embodiment, the three-dimensional model construction module and the 3D printing control module exchange data through a shared memory area and a distributed file system dual-channel interface, the three-dimensional grid model and its mapping coordinate information realize low delay access in the shared memory, and the distributed file system provides redundant storage to support multi-module iterative access and historical data tracing. The system management module periodically polls the data state of each module, iteratively updates according to the closed loop feedback logic, and embeds an abnormality detector to detect and recover data abnormalities, communication delays or module failures, thereby ensuring the stability and continuity of the entire closed loop system.

[0030] In the closed loop interaction of the real scene reproduction module and the precision comparison module in the embodiment, the scanning review 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, and the closed loop trigger conditions include the completion of the printing layer, the update of the three-dimensional model or the completion of the error matrix calculation. The precision comparison module generates an error matrix and returns it to the three-dimensional model construction module and the 3D printing control module in real time through the high frequency bus, the system management module triggers iterative adjustment according to the error threshold, ensures that each module maintains spatial accuracy closed loop under multi-modal data mapping, and supports multi-scale and multi-region continuous calibration.

[0031] In the embodiment, the multi-spectral image acquisition unit adopts an adjustable waveband sensor array, the laser radar scanning unit is composed of a high frequency pulse laser emitter and a rotating scanning mirror, the inertial navigation unit is composed of a six-axis inertial measurement unit and a GNSS receiver, the multi-spectral image acquisition unit, the laser radar scanning unit and the inertial navigation unit form a joint data stream through a unified timing controller and a distributed storage bus, an embedded time synchronization chip marks each data packet with a nanosecond level timestamp to ensure the strict correspondence of multi-source data in time and space, and realizes data closed loop traceability.

[0032] The multi-spectrum image acquisition unit realizes real-time data compression and denoising through a parallel FPGA processor, the laser radar scanning unit dynamically compensates the posture of each laser point under the action of an embedded real-time coordinate corrector, the six-axis measurement data of the inertial navigation unit and the GNSS data are fused through a Kalman filtering algorithm to generate a corrected point cloud with three-dimensional posture information, and the point cloud generation module is provided with high-precision spatial basic data.

[0033] The rotating scanning mirror of the laser radar scanning unit feeds back the scanning angle in real time 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, and the inertial navigation unit and the multi-spectrum image data are synchronously output to form a spatial corresponding point cloud set.

[0034] The point cloud fusion unit performs grid fusion through an adaptive spatial resolution algorithm, and performs weighted processing on the point cloud density abnormal area in the fusion process by using a distributed GPU cluster, the grid reconstruction unit generates a printable three-dimensional grid model by using a block iteration algorithm and automatically fills the local point cloud missing area, and the coordinate conversion unit maps the printable three-dimensional grid model to a 3D printing coordinate system by using a quaternion rotation and a homogeneous matrix.

[0035] Specifically, in the point cloud fusion unit, the adaptive spatial resolution algorithm dynamically adjusts the point cloud density of different areas by establishing a three-dimensional voxel grid (VoxelGrid), and the mathematical model is as follows: ; Among them, represents the current voxel edge length, is an initial voxel reference length, is a density adjustment coefficient, is the current local point cloud density, is the global average point cloud density. The GPU cluster adopts a distributed thread scheduling strategy, divides the entire point cloud space into multiple calculation blocks, and each GPU node dynamically allocates thread resources according to the number of point clouds in the block; the local cache realizes fast exchange through shared memory, and in the calculation process, the weighted compensation method is used for the point cloud density abnormal area, the point weight in the sparse area is increased to enhance the continuity, and the dense area is moderately down-sampled to maintain uniform distribution.

[0036] In the grid reconstruction unit, the block iteration algorithm divides the whole point cloud into several overlapping blocks for independent reconstruction, each block uses Delaunay triangulation combined with Poisson surface reconstruction, and the iteration convergence condition is that the boundary error of adjacent blocks is less than a threshold or the iteration number reaches an upper limit Nmax, where e represents a boundary point coordinate error limit value, and Nmax is a maximum iteration step number. The missing point compensation logic generates compensation points based on local neighborhood interpolation and spline fitting, and simultaneously updates the mesh connection relationship to maintain the overall closed structure. The topology optimization processor ensures that each node is connected to at least three valid patches to form a continuous topology by calculating the adjacency set of each node and performing patch rearrangement rules; and the local mesh precision is optimized by using an Octree index structure to realize mesh refinement and boundary smoothing.

[0037] The mesh reconstruction unit adjusts the mesh connectivity when generating a triangular mesh in combination with the topology optimization processor, automatically generates a continuous and closed mesh topology structure, and enhances the local precision of the mesh through a spatial index structure. The coordinate transformation unit simultaneously performs rotation, scaling, and translation matrix calculation in the mapping process to ensure that the printing coordinates and the surveying coordinates correspond at a micron level of precision, and realizes closed-loop cooperation with the error analysis module.

[0038] Specifically, in the coordinate transformation unit, the quaternion rotation uses the formula: ; wherein, is the quaternion representation of the original vertex coordinates, is the attitude rotation quaternion, is the conjugate inverse thereof, is the vertex coordinate after rotation. The homogeneous matrix mapping uses the formula: ; wherein, is the homogeneous transformation matrix, which includes a rotation matrix , a scaling matrix , and a translation vector , and comprehensively performs coordinate rotation, scaling adjustment, and translation. Error compensation corrects coordinate drift through least squares fitting of local feature points, and precision verification uses standard control point distance measurement results as a control to ensure that the surveying coordinates and the printing coordinates completely correspond within a range of ±5 pm, thereby realizing real-time linkage of the coordinate level micron precision closed loop and the error analysis module.

[0039] The layered slicing unit generates adaptive printing layer data according to the printable three-dimensional network model, the path planning unit calculates the movement speed and ejection amount of the printing path points through a dynamic trajectory optimization algorithm, and receives the error matrix of the precision comparison module in real time for fine tuning, the printing instruction generation unit converts the optimized path into a printer control instruction, forms a closed-loop iterative printing control process, and realizes continuous mapping of printing precision 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.

2. The system according to claim 1, characterized in that, 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.

3. The system according to claim 2, characterized in that, 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.

4. The system according to claim 3, characterized in that, 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.

5. The system according to claim 1, characterized in that, 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.

6. The system according to claim 1, characterized in that, 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.

7. The system according to claim 6, characterized in that, 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.

8. 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 described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the system as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the system according to any one of claims 1-7.

Images