Photogrammetry integrated control method and system

Abstract

The application provides a photogrammetric integrated control method and system, relates to the technical field of photogrammetric control, and comprises the following steps: acquiring multi-source original data such as sequence images, initial point clouds and real-time attitude parameters; performing joint preprocessing of timestamp alignment, point cloud denoising and image distortion correction, and storing key logs and quality indicators in a chain through a consensus mechanism; performing bundle adjustment to solve internal and external orientation elements, calling a blockchain history control point verification and dynamic adjustment; generating initial three-dimensional point clouds through dense matching; verifying the accuracy of the point clouds by using high-precision check points, and feeding back correction instructions when the limit is exceeded; gridizing and texture mapping effective point clouds to generate three-dimensional models, and storing the results in blocks in edge nodes and storing the results in a chain; the system comprises multi-source data acquisition, data preprocessing, adjustment solution, point cloud generation, point cloud verification and three-dimensional modeling modules, and the application effectively improves the efficiency, accuracy and reliability of photogrammetry.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photogrammetry control, and in particular to a photogrammetry integrated control method and system. BACKGROUND

[0002] Photogrammetry technology has been widely applied in urban planning, geological disaster monitoring, digital cultural heritage protection, and large-scale engineering construction, etc. by processing image data to obtain object geometric information and texture. However, with the increasing requirements of real-time performance, accuracy and data reliability in application scenarios, the existing technical system faces severe challenges in specific implementation. In the traditional technical process, data acquisition, preprocessing, aerial triangulation, and three-dimensional reconstruction, etc. usually rely on centralized post-processing platforms. For example, in the scene of emergency monitoring of geological disasters, the massive image and point cloud data collected by unmanned aerial vehicles need to be transmitted back to the cloud computing center for processing. Limited by network bandwidth, the data transmission delay can reach several hours, which is difficult to provide real-time support for rescue decision-making. In the regular deformation monitoring of large bridges, tunnels, etc., the quality control of each link of data processing relies on manual intervention and post-checking, and lacks an automatic precision verification and feedback mechanism throughout the entire process, which may lead to error accumulation and make it difficult to meet the requirements of millimeter-level high-precision monitoring. In addition, there is a lack of effective technical means to audit and trace the authenticity and integrity of the final results, and the measurement results are easily questioned in engineering acceptance or judicial identification scenarios. The core problems that the existing technology cannot solve mainly lie in three aspects: first, the data processing timeliness bottleneck, the centralized transmission and processing mode of massive data cannot meet the low-latency requirements of emergency response, real-time monitoring, etc.; second, the quality control links are disconnected, the accuracy control of adjustment solution and three-dimensional reconstruction is relatively independent, and there is a lack of closed-loop verification and dynamic adjustment mechanism based on measured check points, and system errors are difficult to discover and correct in time; third, the data credibility guarantee is missing in the whole process, from raw data, intermediate results to final model, the processing process, quality indicators and version history lack tamper-proof evidence and traceability ability, and it is difficult to meet the compliance requirements. Therefore, there is an urgent need in the market for a new photogrammetry integrated control method and system. SUMMARY

[0003] The present application aims to provide a photogrammetry integrated control method and system to solve the problems of poor timeliness caused by centralized cloud data processing in the prior art, error accumulation difficult to dynamically correct caused by disconnection of quality control in each processing link (such as adjustment solution and three-dimensional reconstruction), and lack of tamper-proof evidence mechanism for data and results in the whole process, etc. The specific technical solutions are as follows:

[0004] The present application provides a photogrammetry integrated control method, comprising:

[0005] S011, synchronously acquiring multi-source original data of a region to be measured by an integrated data acquisition terminal, the multi-source original data including sequence image data, initial point cloud data, and real-time attitude parameter data, jointly preprocessing the multi-source original data to generate preprocessed data, and uploading key logs of the preprocessing process and quality evaluation indexes to a block chain network for notarization by a consensus mechanism;

[0006] S012, based on the preprocessed data, solving initial interior and exterior orientation elements of the sequence image by bundle adjustment at an edge computing node; verifying by calling a historical control point database of the same region stored in the block chain network, if a coordinate deviation value exceeds a preset threshold, dynamically adjusting a weight coefficient of adjustment and re-solving, if the coordinate deviation value is less than or equal to the preset threshold, determining that the currently solved interior and exterior orientation elements are final interior and exterior orientation elements, and uploading a data digest of a result of solving the final interior and exterior orientation elements to the block chain network;

[0007] S013, combining the final interior and exterior orientation elements with the preprocessed data, constructing an associated mapping model of “image pixel-point cloud coordinate” at the edge computing node, performing dense matching, and generating initial three-dimensional point cloud;

[0008] S014, verifying a coordinate error of the initial three-dimensional point cloud by using high-precision checkpoint data measured in the field; if the error is out of limit, generating a precision correction instruction and feeding back to the dense matching step to regenerate the point cloud; after verification, storing a data digest of the effective three-dimensional point cloud and a precision report in the block chain;

[0009] S015, performing grid processing on the effective three-dimensional point cloud, and mapping the preprocessed image data as a texture map to generate a three-dimensional visual model; a final result data package is split and stored in multiple edge nodes, and an index address and an integrity check value thereof are notarized in the block chain network.

[0010] Further, the edge computing node deployed near the data acquisition terminal is used to perform joint preprocessing of timestamp alignment, point cloud denoising, and image distortion correction on the multi-source original data to generate preprocessed data; the preprocessed data includes preprocessed image data and preprocessed point cloud data.

[0011] Further, the joint preprocessing includes:

[0012] Timestamp alignment unifies the timestamps of image data, point cloud data, and attitude data to a global positioning time system based on a hardware synchronization signal;

[0013] Point cloud denoising removes outliers by calculating the local density of point cloud using a statistical filtering algorithm;

[0014] The image distortion correction adopts the Brown-Conrad model, and uses the camera calibration parameters to perform geometric correction.

[0015] Further, in the step S012, the bundle adjustment uses a collinearity condition equation to construct an error equation, uses image point observation values as observation values, uses exterior orientation elements and object point coordinates as unknowns, and uses the Levenberg-Marquardt algorithm to iteratively solve; the historical control point database includes at least three uniformly distributed control points, and the preset threshold includes a plane position deviation threshold and an elevation deviation threshold.

[0016] Further, in the step S013, the construction of the associated mapping model of the image pixels and the point cloud coordinates includes:

[0017] According to the final interior and exterior orientation elements and the preprocessed image data, key point detection and descriptor generation operations are performed through a scale-invariant feature transform algorithm to obtain an image feature point set;

[0018] According to the image feature point set, a mismatch elimination operation is performed through a random sample consensus algorithm to obtain an optimized matching point pair;

[0019] According to the optimized matching point pair and the final interior and exterior orientation elements, a three-dimensional coordinate calculation operation is performed through a forward intersection algorithm to obtain a sparse three-dimensional point cloud;

[0020] According to the sparse three-dimensional point cloud and the preprocessed image data, a pixel-level disparity calculation operation is performed through a semi-global matching algorithm to generate a dense three-dimensional point cloud as an initial three-dimensional point cloud.

[0021] Further, in the step S014, the verification of the coordinate error of the initial three-dimensional point cloud includes:

[0022] According to the initial three-dimensional point cloud and the high-precision check point data measured on site, a point cloud and check point registration operation is performed through a nearest neighbor search algorithm to obtain a registration point pair set;

[0023] According to the registration point pair set, a coordinate system unification operation is performed through a seven-parameter coordinate conversion model to obtain converted check point coordinates;

[0024] According to the converted check point coordinates and the corresponding initial three-dimensional point cloud coordinates, a coordinate residual sequence is calculated through a Euclidean distance;

[0025] According to the coordinate residual sequence, an error distribution analysis is performed through a statistical hypothesis test, and when the error is out of limit, a precision correction instruction including a correction parameter is generated.

[0026] Further, in the step S015, the meshing processing and texture mapping of the effective three-dimensional point cloud include:

[0027] According to the effective three-dimensional point cloud, the point cloud density field calculation and the isosurface extraction operation are carried out through the Poisson surface reconstruction algorithm, and an initial triangular mesh model is obtained;

[0028] According to the initial triangular mesh model, the vertex number optimization operation is carried out through the mesh simplification algorithm based on edge folding, and a simplified mesh model is obtained;

[0029] According to the simplified mesh model, the mesh surface smoothing operation is carried out through the Laplace smoothing algorithm, and an optimized mesh model is obtained;

[0030] According to the optimized mesh model and the preprocessed image data, the image registration and color fusion operation are carried out through the multi-view texture mapping algorithm, and a textured three-dimensional model is generated;

[0031] The distributed storage of the final result data packet in the step S015 includes:

[0032] According to the textured three-dimensional model, the model data segmentation operation is carried out through the spatial block algorithm, and a plurality of model data blocks are generated;

[0033] The plurality of model data blocks are distributedly stored in the edge node, and the integrity check value of each data block is calculated;

[0034] The index address and the integrity check value of the model data block are stored in the blockchain network through the smart contract.

[0035] The application also provides a photogrammetric integrated control system for realizing the photogrammetric integrated control method.

[0036] The multi-source data acquisition module is used for synchronously acquiring multi-source original data of the to-be-measured area through the integrated data acquisition terminal.

[0037] The data preprocessing module is used for jointly preprocessing the multi-source original data by using the edge computing node to generate preprocessed data.

[0038] The adjustment solution module is used for adopting the bundle adjustment to solve the initial inner and outer orientation elements of the sequence image based on the preprocessed data, and performing verification and adjustment.

[0039] The point cloud generation module is used for constructing an associated mapping model by combining the final inner and outer orientation elements with the preprocessed data, and generating an initial three-dimensional point cloud.

[0040] The point cloud verification module is used for verifying the coordinate error of the initial three-dimensional point cloud by using high-precision check point data, and generating precision correction instructions.

[0041] The three-dimensional modeling module is used for grid processing and texture mapping of the effective three-dimensional point cloud, and generating a three-dimensional visual model.

[0042] The application further provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the photogrammetry integrated control method when executing the computer program.

[0043] The application further provides a computer readable storage medium storing a computer program, wherein the computer program implements the steps of the photogrammetry integrated control method when executed by a processor.

[0044] The photogrammetry integrated control method and system provided by the application realize the collaborative processing of data acquisition, preprocessing and adjustment calculation through edge computing nodes, and the distributed storage of key data and precision indicators is combined with blockchain technology, thereby constructing a full-process closed-loop control system from data acquisition to three-dimensional modeling, significantly improving the automation level and reliability of photogrammetry data processing, effectively solving the problems of process fragmentation, precision control lag and low data reliability in traditional methods, and realizing efficient three-dimensional reconstruction with millimeter-level precision and traceable quality management.

[0045] The technical solutions of the application will be further described in detail below with reference to the accompanying drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0046] The accompanying drawings are used to provide a further understanding of the application, and constitute a part of the specification, together with the embodiments of the application, to explain the application, and do not constitute a limitation on the application. In the drawings:

[0047] Figure 1 is a step schematic diagram of a photogrammetry integrated control method of the application;

[0048] Figure 2 is a structural schematic diagram of a photogrammetry integrated control system of the application. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the application. Obviously, the described embodiments are only a part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0050] It should be noted that all the direction indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present application are only used to explain the relative position relationship, motion condition, etc. between components in a certain specific posture (as shown in the drawings), and if the specific posture changes, the direction indications will also change accordingly.

[0051] It should also be noted that when an element is referred to as being "fixed" or "disposed" on another element, it can be directly on the other element or can have a middle element. When an element is referred to as being "connected" to another element, it can be directly connected to the other element or can have a middle element.

[0052] In addition, the descriptions involving "first", "second", etc. in the present application are only for the purpose of description, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first", "second" can explicitly or implicitly include at least one of the features. In addition, the technical solutions of various embodiments can be combined with each other, but it must be based on the fact that a person skilled in the art can realize it, and when the combination of technical solutions contradicts each other or cannot be realized, it should be considered that the combination of technical solutions does not exist, nor is it within the protection scope required by the present application.

[0053] In the embodiments of the present application, a photogrammetric integrated control method is provided, please refer to Figure 1 , the method comprises the following steps:

[0054] S011, synchronously acquiring multi-source original data of a to-be-measured region by an integrated data acquisition terminal, the multi-source original data comprising sequence image data, initial point cloud data and real-time attitude parameter data; performing joint preprocessing of timestamp alignment, point cloud denoising and image distortion correction on the multi-source original data by an edge computing node deployed in the near end of the data acquisition terminal, to generate preprocessed data, the preprocessed data comprising preprocessed image data and preprocessed point cloud data; the edge computing node uploads key logs and quality evaluation indexes of the preprocessing process to a block chain network for notarization through a consensus mechanism.

[0055] In particular, the integrated data acquisition terminal includes a high-resolution digital aerial camera (e.g., pixel number ≥ 24 million), a laser scanner (e.g., scanning frequency ≥ 100 kHz), and an inertial measurement unit (IMU) / global positioning system (GPS) integrated navigation system (e.g., positioning accuracy ≤ 2 cm). During data acquisition, the terminal flies along the planned flight line at a preset flight height (e.g., 100 meters) and heading overlap (e.g., 80%), synchronously triggers the camera to acquire sequential images (e.g., 2 frames per second), the laser scanner to acquire initial point clouds, and the IMU / GPS to record real-time attitude parameters (including three-dimensional position, roll angle, pitch angle, and yaw angle). The edge computing node uses a high-performance embedded processor (such as ARM Cortex-A77 architecture) and is deployed near the data acquisition terminal (e.g., connected through a 5G low-latency link); it is used to complete data preprocessing, real-time adjustment calculation, and lightweight point cloud generation, etc., which have high time efficiency requirements, and for computationally intensive fine three-dimensional reconstruction and modeling tasks, the network conditions can be selected to complete the calculation in the edge node or return to the cloud computing center. Joint preprocessing first performs timestamp alignment: based on the hardware synchronization signal (such as PPS pulse), the timestamps of image data, point cloud data, and attitude data are unified to the global positioning time system (such as GPS T), with a deviation controlled within milliseconds (e.g., <10 ms). Point cloud denoising uses a statistical filtering algorithm: calculate the local density of the point cloud (e.g., a spherical neighborhood with a radius of 0.5 meters), and remove outliers (e.g., points with a density less than twice the standard deviation of the average density). Image distortion correction uses the Brown-Conrads model, using camera calibration parameters (such as focal length, principal point, radial distortion coefficients k1, k2, and tangential distortion coefficients p1, p2) for geometric correction, eliminating lens distortion (e.g., distortion residual ≤ 0.1 pixels). Key logs include preprocessing timestamps, data size, and processing status; quality evaluation indicators include image signal-to-noise ratio (SNR ≥ 30 dB), point cloud integrity (missing rate ≤ 5%), and geometric accuracy (reprojection root mean square error ≤ 0.1 pixels). The consensus mechanism uses the practical Byzantine fault tolerance (PBFT) algorithm, uploads the hash values (such as SHA-256 digest) of the logs and indicators to the blockchain network (such as a private chain based on Ethereum), ensuring data tamper resistance.

[0056] S012, based on the preprocessed data, the initial interior and exterior orientation elements of the sequential images are calculated in the edge computing node; the same area historical control point database stored in the blockchain network is called for verification, if the coordinate deviation value exceeds the preset threshold, the adjustment weight coefficient is dynamically adjusted and recalculated, if the coordinate deviation value is less than or equal to the preset threshold, the current calculated interior and exterior orientation elements are determined as the final interior and exterior orientation elements, and the data digest (such as SHA-256 hash value) of the calculation result of the final interior and exterior orientation elements is uploaded to the blockchain network.

[0057] Specifically, bundle adjustment uses collinearity condition equations to construct error equations, uses image point observations (such as feature point coordinates) as observations, uses exterior orientation elements (6 parameters for each image: three-dimensional position and three attitude angles) and object point coordinates as unknowns, uses the Levenberg-Marquardt algorithm for iterative solution (for example, the maximum number of iterations is 100 times, and the convergence threshold is ≤1e-6), the initial value is provided by IMU / GPS data (such as position error ≤5cm), the historical control point database is retrieved from the blockchain network, and contains at least three uniformly distributed control points (for example, high-precision coordinates obtained by previous measurement, planar accuracy ≤5cm). When verifying, the calculated object point coordinates are compared with the historical control point coordinates, the deviation value is calculated (such as the planar position deviation using the Euclidean distance, and the height deviation using the absolute difference), and the preset threshold is set according to the measurement accuracy requirement (for example, the planar position deviation threshold is ≤10cm); if the deviation is out of limit, the adjustment weight coefficient is dynamically adjusted: according to the deviation, the weight of the control point is increased (for example, the weight coefficient is adjusted from 1.0 to 2.0) or additional parameters (such as lens distortion correction terms) are introduced, and the calculation is restarted until the deviation meets the requirements. The final interior and exterior orientation elements include camera interior orientation elements (focal length, principal point coordinates) and exterior orientation elements (X, Y, Z, ω, 、κ for each image), the calculation results and their hash values are uploaded to the blockchain through a smart contract (such as Solidity programming), ensuring traceability of the data.

[0058] S013, combined with the final interior and exterior orientation elements and the preprocessed data, an edge computing node is constructed to build a correlation mapping model of “image pixel-point cloud coordinates”, perform dense matching, and generate an initial three-dimensional point cloud.

[0059] In a specific embodiment, the building of the correlation mapping model of “image pixel-point cloud coordinates”, the dense matching, and the generation of the initial three-dimensional point cloud include:

[0060] According to the final interior and exterior orientation elements and the preprocessed image data, a scale-invariant feature transform algorithm is used for key point detection and descriptor generation operation to obtain an image feature point set;

[0061] According to the image feature point set, a random sample consensus algorithm is used for mismatch elimination operation to obtain an optimized matching point pair;

[0062] According to the optimized matching point pair and the final interior and exterior orientation elements, a forward intersection algorithm is used for three-dimensional coordinate calculation operation to obtain a sparse three-dimensional point cloud;

[0063] According to the sparse three-dimensional point cloud and the preprocessed image data, a semi-global matching algorithm is used for pixel-level disparity calculation operation to generate a dense three-dimensional point cloud as an initial three-dimensional point cloud.

[0064] Specifically, first, the pre-processed image data is processed using a scale-invariant feature transform (SIFT) algorithm: taking each image as input, a Gaussian pyramid scale space is constructed (for example, the number of scale layers is set to 4, and the Gaussian blur coefficient of each layer is increased by a base of 1.6), and local extreme points are detected as candidate key points. The positioning accuracy of the key points is fitted to a three-dimensional quadratic function to determine the sub-pixel position (with a tolerance of less than 0.5 pixels), and low-contrast points (with a contrast threshold dynamically set according to the image signal-to-noise ratio, for example, when the signal-to-noise ratio is greater than or equal to 30 dB, the threshold is set to 0.03) and edge response points (with a principal curvature ratio threshold set to 10) are removed. An 128-dimensional descriptor is generated for each key point: taking a 16x16 pixel region centered on the key point, an 8-direction gradient histogram is calculated, and normalized to a unit length to enhance the invariance to illumination. The final output includes a set of image feature points containing key point coordinates, scales, directions, and descriptors. Then, false match removal is performed: based on the set of image feature points, the k-nearest neighbor algorithm (k=2) is used for initial matching, the Euclidean distance ratio of the descriptors (the ratio of the nearest neighbor distance to the second nearest neighbor distance) is calculated, and if the ratio is less than a pre-set threshold (determined to be 0.7 through a matching reliability experiment), the candidate matching point pair is retained. Subsequently, the random sample consensus (RANSAC) algorithm is applied for optimization: a minimum sample set (for example, 4 matching point pairs) is randomly extracted to calculate the fundamental matrix, and the number of iterations is adaptively adjusted according to the inlier ratio (for example, the confidence is set to 0.99, and the initial estimate of the inlier ratio is 0.5). In each iteration, the distance of the matching point pair to the epipolar line is calculated, and if the distance is less than a threshold (for example, 1 pixel), it is marked as an inlier. Finally, the matching point pair corresponding to the model with the highest inlier ratio is retained, and the optimized matching point pair is output.

[0065] Based on the optimized matching point pairs and the final internal and external orientation elements, forward intersection calculation is performed: for each matching point pair, a collinearity condition equation is constructed to relate the image point coordinates (x, y) and the object coordinates (X, Y, Z). The least squares method is used to solve the over-determined equation set (at least 2 matching points for each image, for example), and the object point coordinates are iteratively solved (such as the Levenberg-Marquardt algorithm), with the convergence condition set to a coordinate update of less than 1e-6 meters. At the same time, the re-projection error is calculated, and if the error exceeds the threshold (for example, 2 pixels), the point is removed. The sparse three-dimensional point cloud is output, and the point cloud density is determined by the number of matching points (for example, an average of 500 feature points are extracted from each image, generating about 10^4 sparse points). Finally, the semi-global matching (SGM) algorithm is executed to generate a dense point cloud: the sparse three-dimensional point cloud is used as the initial constraint, and the pre-processed image data is used to construct a disparity space image. The matching cost calculation uses Census transformation and gradient information fusion: Census transformation calculates the local binary pattern with a 9x7 pixel window, and gradient information is extracted by the Sobel operator (with a gradient weight of 0.5). The cost aggregation is performed along 8 paths (0°, 45°, 90°, 135° and the opposite direction), and the penalty coefficients P1 and P2 are set according to the texture richness of the image (for example, P1=5 and P2=20 for rich texture areas, and P1=2 and P2=10 for smooth areas). The winner-takes-all strategy is used for disparity calculation, and sub-pixel accuracy is optimized by quadratic interpolation. The depth range is dynamically set according to the depth prior provided by the sparse point cloud (for example, a depth range of ±10% is used as the disparity search interval). The disparity map is converted to three-dimensional coordinates to generate a dense three-dimensional point cloud (with a point spacing of 1-2 times the ground sampling distance of the image), which is used as the initial three-dimensional point cloud.

[0066] This step ensures that the point cloud covers the entire measurement area while maintaining geometric accuracy, providing basic data for subsequent accuracy verification. For example, the sparse point cloud is used to stabilize the model framework, and the dense point cloud is used to refine the surface morphology, which improves the reliability and integrity of the three-dimensional reconstruction.

[0067] S014, using high-precision checkpoint data measured in the field, to verify the coordinate error of the initial three-dimensional point cloud; if the error is out of limit, generate precision correction instructions and feed back to the dense matching step to regenerate the point cloud; after verification, store the data summary (such as SHA-256 hash value) of the effective three-dimensional point cloud and the precision report in the blockchain.

[0068] In a specific embodiment, the use of high-precision checkpoint data measured in the field to verify the coordinate error of the initial three-dimensional point cloud comprises:

[0069] According to the initial three-dimensional point cloud and the high-precision checkpoint data measured in the field, a point cloud-checkpoint registration operation is performed by a nearest neighbor search algorithm to obtain a set of registered point pairs.

[0070] According to the set of registration points, a coordinate system unification operation is performed through a seven-parameter coordinate conversion model to obtain converted checkpoint coordinates;

[0071] According to the converted checkpoint coordinates and the corresponding initial three-dimensional point cloud coordinates, a coordinate residual sequence is calculated through Euclidean distance calculation;

[0072] According to the coordinate residual sequence, error distribution analysis is performed through statistical hypothesis testing, and when the error is out of limit, an accuracy correction instruction containing correction parameters is generated;

[0073] According to the accuracy correction instruction, the matching cost function weight of the dense matching algorithm is adjusted, and the point cloud is regenerated;

[0074] After verification, the effective three-dimensional point cloud and its accuracy report are stored in the blockchain network through the smart contract.

[0075] Specifically, first, point cloud-checkpoint registration is performed: using the k-d tree nearest neighbor search algorithm, taking the high-precision checkpoint coordinates (obtained by total station or RTK measurement, plane accuracy ≤2 cm, height accuracy ≤3 cm) as the query point, searching for the nearest neighbor point in the initial three-dimensional point cloud. The search radius is set according to the point cloud density (for example, 3 times the average point spacing), ensuring that each checkpoint can find the corresponding point cloud point. For the matched point pairs, the distance between the two points is calculated, and if the distance exceeds the tolerance threshold (for example, 5 times the point cloud accuracy), it is determined as invalid matching and excluded. Finally, the registration point pair set containing the checkpoint coordinates and the corresponding point cloud point coordinates is output. Then, coordinate system unification is performed: using the seven-parameter Helmert coordinate conversion model (containing 3 translation parameters, 3 rotation parameters and 1 scale parameter), taking the registration point pair set as the reference, the conversion parameters are solved by least squares method. The conversion accuracy is evaluated by root mean square error, and if the root mean square error exceeds the threshold (for example, 1.5 times the checkpoint accuracy), the iterative weighted least squares method is used, and the point pair weight is dynamically adjusted according to the residual size (the weight of the point pair with large residual is reduced). After completing the coordinate system unification, the converted checkpoint coordinates (consistent with the point cloud coordinate system) are output. Then, the coordinate residual sequence is calculated: for each registration point pair, the Euclidean distance between the converted checkpoint coordinates and the corresponding point cloud point coordinates is calculated to generate the coordinate residual sequence. The residual calculation covers the plane position residual and the height residual, and the average value, maximum value and standard deviation are calculated respectively.

[0076] Error distribution analysis is performed based on the coordinate residual sequence: the Kolmogorov-Smirnov test is used to determine whether the residuals follow a normal distribution. If the test passes (p-value > 0.05), the error range of the 95% confidence interval is calculated as the accuracy index; if the test fails, the Chebyshev inequality is used to determine the error boundary. The calculated error range is compared with a preset accuracy threshold (for example, in topographic surveying, the horizontal position error threshold is set to 0.1m, and the elevation error threshold is set to 0.15m). If the error in any direction exceeds the limit, a precision correction instruction is generated based on the spatial distribution characteristics and statistical properties of the coordinate residual sequence. The precision correction instruction is used to guide the point cloud generation module to adjust the parameter configuration of the dense matching algorithm (such as the weight of the matching cost function, the disparity search range, etc.) or trigger the re-acquisition and processing of data in a specific area. Specifically, if the error in any direction exceeds the limit, a precision correction instruction is generated, and the instruction content includes: the image area to be matched (located according to the residual distribution), the adjustment ratio of the gray-level similarity weight and the geometric constraint weight in the matching cost function (for example, the geometric constraint weight is increased by 20% in areas with large residuals), and the correction value of the disparity search range (dynamically shrinking or expanding according to the error size).

[0077] The dense matching process is adjusted according to the accuracy correction instructions: the cost function of the semi-global matching algorithm is modified, for example, by increasing the weight coefficient of geometric continuity constraints in the error-exceeding region (adjusting it from the default value of 1.0 to 1.2), while simultaneously narrowing the disparity search range (determining the search interval offset based on the error vector direction). Dense matching is re-executed to generate the corrected 3D point cloud. After successful verification, the valid 3D point cloud and its accuracy report (including indicators such as plane mean square error, elevation mean square error, and maximum residual) are stored as evidence via a smart contract. The smart contract automatically generates a data hash (using the SHA-256 algorithm) and binds the hash value to a storage address (such as an IPFS content identifier) ​​before writing it to the blockchain. A multi-node verification mechanism (such as PBFT consensus) is triggered during evidence storage to ensure the data is immutable.

[0078] This step achieves closed-loop control of point cloud accuracy through statistical testing and dynamic parameter adjustment. For example, adaptive weight adjustment based on residual distribution effectively suppresses the accumulation of systematic errors, ensuring that the 3D model meets engineering accuracy requirements. The generated accuracy report provides a reliable basis for subsequent applications, such as determining the sensitivity threshold for structural deformation monitoring in civil engineering based on point cloud accuracy.

[0079] S015, the effective 3D point cloud is meshed, and the preprocessed image data is mapped as a texture map to generate a 3D visualization model; the final result data package is split and stored in multiple edge nodes, and its index address and integrity verification value are stored in the blockchain network to realize distributed management and tamper-proof traceability of the results.

[0080] In a specific embodiment, the grid processing and texture mapping of the effective three-dimensional point cloud comprises:

[0081] According to the effective three-dimensional point cloud, the point cloud density field calculation and the isosurface extraction operation are performed through a Poisson surface reconstruction algorithm to obtain an initial triangular mesh model;

[0082] According to the initial triangular mesh model, the vertex number optimization operation is performed through an edge collapse-based mesh simplification algorithm to obtain a simplified mesh model;

[0083] According to the simplified mesh model, the mesh surface smoothing operation is performed through a Laplace smoothing algorithm to obtain an optimized mesh model;

[0084] According to the optimized mesh model and the preprocessed image data, the image registration and color fusion operation are performed through a multi-view texture mapping algorithm to generate a textured three-dimensional model;

[0085] According to the textured three-dimensional model, the model data segmentation operation is performed through a spatial block algorithm to generate a plurality of model data blocks;

[0086] The plurality of model data blocks are distributedly stored in edge nodes, and the integrity check value of each data block is calculated;

[0087] The index address and the integrity check value of the model data block are stored in the blockchain network through a smart contract.

[0088] In detail, first, Poisson surface reconstruction is performed based on the effective three-dimensional point cloud: an octree data structure is used to hierarchically divide the point cloud space, and the depth is adaptively set according to the point cloud density (for example, when the average point distance is 0.1 m, the depth is set to 8). The point cloud density field is calculated, and the scalar field is obtained by solving the Poisson equation (such as using the conjugate gradient method for iterative solution), and the isosurface threshold is set according to the point cloud noise level (for example, determined by multiplying the point cloud standard deviation by a coefficient of 0.5). The isosurface is extracted to generate an initial triangular mesh model, and the surface closure is ensured (such as no hole defects). The mesh resolution is controlled by the size of the octree leaf node, for example, set to 1.5 times the average point distance of the point cloud. Then, the initial triangular mesh model is simplified: an edge folding-based algorithm is used to calculate the folding cost of each edge (such as based on quadratic error measurement), and the edge with the smallest folding cost is preferentially folded. The simplification target is controlled by the threshold of the number of mesh facets (for example, when the number of original mesh facets exceeds 1 million, simplify to within 500,000 facets). Feature edges (such as edges with a curvature change greater than 0.1) are preserved during the folding process to avoid loss of important geometric details. The output simplified mesh model maintains the geometric error within the allowable range (such as the Hausdorff distance is less than 2 times the point cloud accuracy). Then, mesh smoothing is performed: the Laplace smoothing algorithm is applied, and each vertex position is updated iteratively (such as 10 iterations). The smoothing weight is assigned according to the reciprocal of the distance between neighboring vertices, while introducing a feature protection constraint (such as a normal vector change threshold of 15°), preventing excessive smoothing from causing sharp edges to be blurred. The output optimized mesh model has significantly reduced surface roughness (such as a reduction of more than 30% in average curvature variance). Based on the optimized mesh model, texture mapping is performed: first, the preprocessed image data is aligned with the mesh model through multi-view image registration. A feature point-based registration method (such as SIFT feature matching) is used to calculate the mapping relationship between the image and the mesh UV coordinates (such as using a perspective transformation model). For each triangular facet, select the image with the most nearly vertical viewing angle as the texture source (such as a viewing angle less than 30°). If multiple images cover the same facet, perform color fusion (such as weighted averaging, with weights calculated according to image resolution and lighting consistency). The texture resolution is set according to the mesh size (for example, 512x512 pixels per meter), and a textured three-dimensional model is generated. Subsequently, the textured three-dimensional model is divided into data blocks: a spatial block algorithm (such as octree partitioning) is used to divide the model into uniformly sized data blocks (such as each block covering a 10m x 10m x 10m space). The block size is dynamically adjusted according to the edge node storage capacity (for example, when the node capacity is 1GB, the block size is set to 100MB). Each data block contains geometric data (vertex coordinates, facet indices) and texture data (image map), and block metadata (such as boundary box coordinates, data size) is generated. The divided model data blocks are distributed and stored in multiple edge nodes (such as based on a consistent hashing algorithm to allocate storage nodes).The integrity check value (such as SHA-256 hash value) of each data block is calculated, and the index address (such as IPFS content identifier) of the storage node is recorded. The index address and the hash value are packaged into a transaction by the smart contract and submitted to the blockchain network (such as using the PBFT consensus mechanism). The multi-node verification is triggered during the evidence storage to ensure that the data cannot be tampered with.

[0089] This step ensures the geometric accuracy of the model by Poisson reconstruction and grid optimization, enhances the visual authenticity by multi-view texture mapping, and improves the data reliability by distributed storage combined with blockchain evidence storage. For example, the grid simplification rate is controlled within 50% to ensure the precision requirements of engineering applications, the texture fusion eliminates the image splicing gaps, the block storage supports fast loading of large-scale models, the integrity check value prevents data transmission damage, and the whole process realizes efficient and reliable three-dimensional result management.

[0090] Correspondingly, as shown in Figure 2 The present application embodiment provides a photogrammetry integrated control system based on a photogrammetry integrated control method, which realizes the photogrammetry integrated control method of the present application embodiment, and the system comprises:

[0091] A multi-source data acquisition module is configured to synchronously acquire multi-source original data of a region to be measured through an integrated data acquisition terminal, wherein the multi-source original data includes sequence image data, initial point cloud data, and real-time attitude parameter data.

[0092] A data preprocessing module is configured to perform joint preprocessing of timestamp alignment, point cloud denoising, and image distortion correction on the multi-source original data by using an edge computing node deployed in the vicinity of the data acquisition terminal, to generate preprocessed data, and to upload key logs and quality evaluation indexes of the preprocessing process to a blockchain network for evidence storage through a consensus mechanism.

[0093] An adjustment and solution module is configured to perform bundle adjustment on the preprocessed data to solve initial interior and exterior orientation elements of the sequence image based on the preprocessed data; to verify a historical control point database of the same region stored in the blockchain network, and if a coordinate deviation value exceeds a preset threshold, to dynamically adjust an adjustment weight coefficient and re-solve, and if the coordinate deviation value is less than or equal to the preset threshold, to determine that the currently solved interior and exterior orientation elements are final interior and exterior orientation elements, and to upload a solution result of the final interior and exterior orientation elements and a hash value thereof to the blockchain network.

[0094] A point cloud generation module is configured to construct an associated mapping model of "image pixel-point cloud coordinates" based on the final interior and exterior orientation elements and the preprocessed data, to perform dense matching, and to generate initial three-dimensional point clouds.

[0095] A point cloud verification module is configured to verify the coordinate error of the initial three-dimensional point cloud by using the high-precision check point data measured on site, and if the error exceeds the limit, generate a precision correction instruction and feed back to the point cloud generation module to regenerate the point cloud; after verification, the effective three-dimensional point cloud and its precision report are stored in the blockchain network;

[0096] A three-dimensional modeling module is configured to grid the effective three-dimensional point cloud, and map the preprocessed image data as a texture map to generate a three-dimensional visual model; the final achievement data packet is split and stored in multiple edge nodes, and the index address and integrity check value thereof are stored in the blockchain network to realize distributed management and tamper-proofing of the achievement. It should be noted that the photogrammetric integrated control system provided by the embodiment of the present application is used to execute all process steps of the photogrammetric integrated control method provided by the above embodiment, and the working principles and beneficial effects of the two are one-to-one corresponding, so they will not be described again. The embodiment of the present application further provides an electronic device. The electronic device comprises a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a photogrammetric integrated control program. The processor executes the computer program to implement the steps in each of the above photogrammetric integrated control method embodiments, such as steps S011 to S015 shown in the above embodiment. Alternatively, the processor executes the computer program to implement the functions of each module / unit in each of the above device embodiments, such as a photogrammetric integrated control module. Figure 1

[0097] For example, the computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the electronic device.

[0098] The electronic device can be a desktop computer, a notebook, a palm computer, and a smart tablet, etc. The electronic device can include, but is not limited to, a processor, a memory. Those skilled in the art can understand that the above components are only examples of the electronic device and do not constitute a limitation on the electronic device, and can include more or fewer components than the above, or combine certain components, or different components, for example, the electronic device can also include an input / output device, a network access device, a bus, etc.

[0099] ​The processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The processor is a control center of the electronic device, and connects various parts of the electronic device through various interfaces and lines.

[0100] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer program and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), etc.; and the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a nonvolatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory device.

[0101] The modules / units integrated in the electronic device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0102] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. A photogrammetric integrated control method, characterized in that, include: S011, multi-source raw data of the area to be measured is synchronously acquired through an integrated data acquisition terminal. The multi-source raw data includes sequential image data, initial point cloud data and real-time attitude parameter data. The multi-source raw data is jointly preprocessed to generate preprocessed data. The edge computing node uploads the key logs and quality evaluation indicators of the preprocessing process to the blockchain network for evidence storage through a consensus mechanism. S012, based on the preprocessed data, the initial interior and exterior orientation elements of the sequence image are solved by bundle adjustment at the edge computing node; the database of historical control points in the same area stored in the blockchain network is called for verification. If the coordinate deviation value exceeds the preset threshold, the adjustment weight coefficient is dynamically adjusted and recalculated. If the coordinate deviation value is less than or equal to the preset threshold, the interior and exterior orientation elements currently solved are determined to be the final interior and exterior orientation elements. The data summary of the solution results of the final interior and exterior orientation elements is uploaded to the blockchain network. S013, combining the final interior and exterior orientation elements with the preprocessed data, constructs an association mapping model of "image pixel - point cloud coordinates" at the edge computing node, performs dense matching, and generates the initial three-dimensional point cloud; S014: Using high-precision checkpoint data measured in the field, verify the coordinate error of the initial 3D point cloud; if the error exceeds the limit, generate an accuracy correction instruction and feed it back to the dense matching step to regenerate the point cloud; after verification, store the data summary and accuracy report of the effective 3D point cloud on the blockchain. S015. The effective 3D point cloud is meshed, and the preprocessed image data is mapped as a texture map to generate a 3D visualization model; the final result data package is split and stored in multiple edge nodes, and its index address and integrity verification value are stored in the blockchain network.

2. The control method as described in claim 1, characterized in that, The multi-source raw data is subjected to joint preprocessing, including timestamp alignment, point cloud denoising, and image distortion correction, using edge computing nodes deployed near the data acquisition terminal to generate preprocessed data. The preprocessed data includes preprocessed image data and preprocessed point cloud data.

3. The control method as described in claim 2, characterized in that, The joint preprocessing includes: Timestamp alignment is based on hardware synchronization signals, which unifies the timestamps of image data, point cloud data and attitude data to the Global Positioning System. Point cloud denoising employs a statistical filtering algorithm to calculate the local density of the point cloud and remove outliers. Image distortion correction employs the Brown-Conrad model, using camera calibration parameters for geometric correction.

4. The control method as described in claim 1, characterized in that, In step S012, the bundle adjustment uses the collinearity condition equation to construct the error equation, with the image point observation value as the observation value and the exterior orientation element and object point coordinates as the unknowns. The Levenburg-Marquardt algorithm is used for iterative solution. The historical control point database contains at least three uniformly distributed control points, and the preset thresholds include the plane position deviation threshold and the elevation deviation threshold.

5. The control method as described in claim 1, characterized in that, Step S013 involves constructing the association mapping model between image pixels and point cloud coordinates, including: Based on the final interior and exterior orientation elements and the preprocessed image data, key point detection and descriptor generation operations are performed using the scale-invariant feature transformation algorithm to obtain the image feature point set. Based on the image feature point set, a random sampling consensus algorithm is used to remove mismatches and obtain optimized matching point pairs. Based on the optimized matching point pairs and the final interior and exterior orientation elements, the three-dimensional coordinate calculation operation is performed through the forward intersection algorithm to obtain a sparse three-dimensional point cloud; Based on sparse 3D point cloud and preprocessed image data, a pixel-level disparity calculation operation is performed using a semi-global matching algorithm to generate a dense 3D point cloud as the initial 3D point cloud.

6. The control method as described in claim 1, characterized in that, The verification of the coordinate error of the initial 3D point cloud in step S014 includes: Based on the initial 3D point cloud and the high-precision checkpoint data measured in the field, the point cloud and checkpoints are registered using the nearest neighbor search algorithm to obtain a set of registered point pairs. Based on the set of registration point pairs, coordinate system one operation is performed through a seven-parameter coordinate transformation model to obtain the transformed checkpoint coordinates; Based on the transformed checkpoint coordinates and the corresponding initial 3D point cloud coordinates, the coordinate residual sequence is calculated using Euclidean distance. Based on the coordinate residual sequence, error distribution analysis is performed through statistical hypothesis testing. When the error exceeds the limit, a precision correction command containing correction parameters is generated.

7. The control method as described in claim 1, characterized in that, Step S015 involves meshing and texture mapping of the effective 3D point cloud, including: Based on the effective 3D point cloud, the point cloud density field is calculated and the isosurface is extracted using the Poisson surface reconstruction algorithm to obtain the initial triangular mesh model. Based on the initial triangular mesh model, a simplified mesh model is obtained by optimizing the number of vertices through a mesh simplification algorithm based on edge folding. Based on the simplified mesh model, the mesh surface is smoothed using the Laplacian smoothing algorithm to obtain the optimized mesh model; Based on the optimized mesh model and preprocessed image data, image registration and color fusion operations are performed using a multi-view texture mapping algorithm to generate a textured 3D model. The distributed storage of the final result data package in step S015 includes: Based on the textured 3D model, the model data is segmented using a spatial block algorithm to generate multiple model data blocks. Multiple model data blocks are distributed and stored on edge nodes, and the integrity check value of each data block is calculated. The index address and integrity verification value of the model data block are stored on the blockchain network through a smart contract.

8. A photogrammetric integrated control system for implementing the photogrammetric integrated control method according to any one of claims 1-7, characterized in that, The system includes: The multi-source data acquisition module is used to synchronously acquire multi-source raw data of the area to be measured through an integrated data acquisition terminal; The data preprocessing module is used to perform joint preprocessing on multi-source raw data using edge computing nodes to generate preprocessed data; The adjustment module is used to perform initial interior and exterior orientation elements of the image sequence using bundle adjustment based on the preprocessed data, and to verify and adjust them. The point cloud generation module is used to combine the final interior and exterior orientation elements with the preprocessed data to build an association mapping model and generate an initial 3D point cloud. The point cloud verification module is used to verify the coordinate errors of the initial 3D point cloud using high-precision checkpoint data and generate accuracy correction instructions. The 3D modeling module is used to perform meshing and texture mapping on effective 3D point clouds to generate 3D visualization models.

9. An electronic device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the integrated photogrammetric control method as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the integrated photogrammetric control method as described in any one of claims 1-7.

Images