Measurement data acquisition method and system of total station

Abstract

The invention relates to the field of data acquisition, in particular to a total station measurement data acquisition method and system, and the method comprises the steps: carrying out the three-dimensional coordinate calculation and arrangement of the original measurement data of a total station, and obtaining an original measurement data set with a time sequence label; carrying out comprehensive analysis on the local neighborhood distance of original measurement data points to obtain the self-adaptive neighborhood radius of each point; obtaining a shared local structure tensor between the point pairs by performing fusion construction on the local structure tensors of the measuring points; performing main axis direction deviation analysis on the shared local structure tensor to obtain symmetric anisotropic distance measurement with self-adaptive forms; through combined judgment of the form adaptive distance and the neighborhood radius, a gross error point identification result is obtained, and abnormal measurement data is eliminated, so that the problem of inaccurate gross error identification caused by non-uniform distribution density of measurement points and data anisotropy is solved.

Description

Technical Field

[0001] The present invention relates to the field of data acquisition, and in particular to a total station measurement data acquisition method and system. Background Art

[0002] Total stations, as high-precision measurement equipment, are widely used in the long-term automated health monitoring of large infrastructure such as bridges, dams, and high-rise buildings. In these applications, total stations perform continuous, high-frequency coordinate measurements of multiple monitoring points on the structure at a preset interval, generating massive amounts of time-series spatial data. To ensure the accuracy of subsequent deformation analysis, it is essential to effectively identify and remove the inevitable gross errors (i.e., anomalous data points caused by factors such as atmospheric turbulence and transient instrument reading errors) in the raw data stream. One existing approach for automated gross error identification is the DBSCAN algorithm. The basic principle of this algorithm is that in coordinate space, normal measurement data points naturally cluster due to their inherent repetitiveness, forming one or more high-density regions; gross errors typically manifest as isolated noise points far from these dense regions. The DBSCAN algorithm automatically classifies data points into core points, boundary points, and noise points using two core parameters: a globally fixed neighborhood radius and a minimum number of points set by the user. Among them, the identified noise points are considered to be gross errors and are eliminated, while the core points and boundary points together constitute normal measurement data clusters and are retained for subsequent analysis. However, when processing actual total station monitoring data, the above-mentioned gross error identification method using the standard DBSCAN algorithm has an unresolved problem between its algorithm mechanism and data characteristics, which limits the accuracy of its identification results. In a typical monitoring network, due to the different physical distances between each monitoring point and the total station, its measurement data will show different distribution densities according to the error propagation law, that is, the data clusters of close-range measurement points are compact and high-density, while the data clusters of distant measurement points are loose and low-density. The standard DBSCAN algorithm requires a globally unified neighborhood radius to measure the neighborhood range of all data points. This setting cannot simultaneously adapt to the various data densities existing in the network. If Eps is set too low to accurately identify high-density clusters, the overly strict scale will lead to a large number of normal low-density data points being misclassified as gross errors. Conversely, if Eps is set too high to accommodate low-density clusters, the overly loose scale will lead to some real gross errors being mistakenly classified as high-density data clusters, resulting in missed detections. This conflict between the global stability of the Eps parameter and the local density heterogeneity of the data makes it difficult for the algorithm to achieve consistent and optimal gross error identification across the entire monitoring network when selecting parameters, and is an urgent problem that needs to be solved in current technology. Summary of the Invention

[0003] In view of this, an embodiment of the present invention provides a total station measurement data acquisition method to solve the problem of inaccurate gross error identification caused by uneven distribution density of measurement points and data anisotropy.

[0004] To achieve the above object, the technical solution of the present invention is achieved as follows: In a first aspect, the present application provides a total station measurement data acquisition method, the method comprising the following steps: Step S1: Obtaining an original measurement data set with time series labels by performing three-dimensional coordinate calculation and sorting on the original measurement data of the total station; Step S2: Obtain the adaptive neighborhood radius of each point by comprehensively analyzing the local neighborhood distances of the original measurement data points; Step S3: Obtain the shared local structure tensor between the point pairs by fusing and constructing the local structure tensors of the measurement points; Step S4: Obtaining a morphologically adaptive symmetric anisotropic distance metric by performing principal axis deviation analysis on the shared local structure tensor; Step S5: By jointly judging the morphological adaptive distance and the neighborhood radius, the gross error point identification result is obtained and abnormal measurement data is eliminated.

[0005] Preferably, the obtaining of the original measurement data set with time series labels by performing three-dimensional coordinate calculation and sorting on the original measurement data of the total station includes: By setting up multiple monitoring points on the structure or area to be measured, using the total station automated measurement system, periodic three-dimensional coordinate measurement of each monitoring point is performed within a preset measurement period, and the three-dimensional spatial coordinate value of the monitoring point corresponding to each measurement moment and its corresponding time tag are obtained; all the collected measurement results of the monitoring points are sorted in chronological order to form an original measurement data set containing the three-dimensional coordinates and time tags of each monitoring point at multiple measurement moments. The original measurement data set is composed of multiple measurement data points, and each measurement data point contains its three-dimensional coordinate value and acquisition time information.

[0006] Preferably, the step of comprehensively analyzing the local neighborhood distances of the original measurement data points to obtain the adaptive neighborhood radius of each point includes: The number of neighboring data points is set to obtain the set of neighboring data points of any target measurement data point in all measurement data points; for any target measurement data point, the Euclidean distance between the target measurement data point and the farthest data point in its neighboring data point set is taken as the local maximum neighborhood arm length of the target measurement data point; the average Euclidean distance between the target measurement data point and all data points in its neighboring data point set is taken as the local average neighborhood arm length of the target measurement data point; the square root of the sum of the squares of the local maximum neighborhood arm length and the local average neighborhood arm length of the target measurement data point is taken as the adaptive neighborhood radius of the target measurement data point.

[0007] Preferably, the step of fusing and constructing the local structure tensors of the measuring points to obtain a shared local structure tensor between the point pairs includes: For any target measurement data point's neighbor data set, the vector whose geometric center points to any target neighborhood measurement data point in the neighbor data set is used as the position vector of the target measurement data point relative to the target neighborhood measurement data point; the outer product of the position vector of the target measurement data point relative to the target neighborhood measurement data point and itself is used as the first vector outer product of the target measurement data point relative to the target neighborhood measurement data point; for any target measurement data point's neighbor data set, the sum of all first vector outer products in the neighbor data set is used as the local structure tensor of the target measurement data point; for the local structure tensors of any two target measurement data points, the average of the local structure tensors of the two target measurement data points is used as the shared local structure tensor of the two target measurement data points.

[0008] Preferably, the obtaining of a morphologically adaptive symmetric anisotropic distance metric by performing principal axis direction deviation analysis on the shared local structure tensor comprises: The shared principal axis direction set is obtained by performing eigenvector decomposition on the shared local structure tensor. The shared local principal axis deviation factor is obtained by performing projection analysis on the direction of the connection line of the measurement point pair and the principal axis direction set. The morphologically adaptive symmetric anisotropic distance is obtained by weighted fusion of the original Euclidean distance and the principal axis deviation factor.

[0009] Preferably, the step of performing eigenvector decomposition on the shared local structure tensor to obtain the shared principal axis direction set includes: Perform eigenvector decomposition on the shared local structure tensor to obtain eigenvalue-eigenvector pairs corresponding to the number of dimensions of the shared local structure tensor; sort all eigenvalue-eigenvector pairs in descending order according to the size of the eigenvalue, and obtain the eigenvector corresponding to the first-ranked eigenvalue as the principal axis direction of the shared local structure tensor, and the eigenvectors corresponding to the remaining eigenvalues ​​are the secondary axis directions in turn; the set formed by the principal axis direction and the secondary axis direction is taken as the shared principal axis direction set.

[0010] Preferably, the obtaining of the shared local principal axis deviation factor by performing projection analysis on the direction of the line connecting the measurement point pairs and the principal axis direction set includes: For any two target measurement data points, a direction vector between the two target measurement data points is obtained through the coordinates of the two target measurement data points; a dot product calculation result of the direction vector between the two target measurement data points and any eigenvector in the shared principal axis direction set of the two target measurement data points is used as a first vector dot product between the two target measurement data points; The calculation result of adding all the first vector dot products between the two target measurement data points is used as the denominator, the calculation result of adding all the first vector dot products between the two target measurement data points except the principal axis direction is used as the numerator, and the calculation result of the corresponding fraction is used as the shared local principal axis deviation factor of the two target measurement data points.

[0011] Preferably, the step of fusing and weighting the original Euclidean distance and the principal axis deviation factor to obtain the morphologically adaptive symmetric anisotropic distance includes: The result of adding the constant 1 to the shared local principal axis deviation factor is used as the first distance optimization weight, and the result of multiplying the first distance optimization weight of any two target measurement data points by the Euclidean distance between the two target measurement data points is used as the morphologically adaptive symmetric anisotropic distance between the two target measurement data points.

[0012] Preferably, the step of obtaining a gross error point identification result and eliminating abnormal measurement data by jointly judging the morphological adaptive distance and the neighborhood radius includes: A minimum data point number threshold is set, and DBSCAN clustering is performed on the measurement data points in the original measurement data set according to the adaptive neighborhood radius of the measurement data points and the morphologically adaptive symmetric anisotropic distance between the measurement data points. The clustering process is completed and the core points, boundary points and noise points marked in the clustering process are obtained; the measurement data points marked as noise points are regarded as gross error data points and the gross error data points are deleted from the original data set to obtain the measurement data set without abnormal measurement data.

[0013] In a second aspect, the present application provides a total station measurement data acquisition system, comprising: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a total station measurement data acquisition method is implemented.

[0014] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: This invention effectively addresses the uneven density of point clouds caused by ranging differences by constructing an adaptive neighborhood radius tailored to local density variations of measurement points. This technology combines the local average neighborhood arm length and the maximum arm length to form a composite diffusion radius. This automatically tightens the neighborhood judgment scale in high-density areas and dynamically relaxes the judgment boundary in low-density areas, ensuring that normal measurement points within clusters of different densities are accurately identified and retained, effectively preventing the phenomenon of accidental deletion caused by fixed neighborhood parameters.

[0015] At the same time, the present invention introduces a shared local structure tensor and constructs an anisotropic distance measurement mechanism based on the deviation of the main axis direction, which significantly improves the adaptability to data morphology under environmental disturbances. By analyzing the degree of deviation between the direction of the point-pair connection and the shared main extension direction, the clustering process is guided to extend along the natural stretching direction of the data cluster, avoiding the misidentification of trajectory changes caused by physical disturbances as gross errors. This technical means has demonstrated good directional discrimination and structural recognition capabilities in actual monitoring scenarios, making it possible to maintain high-precision outlier removal even under complex environmental conditions, thereby improving the overall quality of the total station monitoring data and the reliability of subsequent analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0017] Figure 1 The present invention provides a method flow chart of a total station measurement data acquisition method. DETAILED DESCRIPTION

[0018] The embodiments of the present disclosure are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to be used to explain the present disclosure, but should not be understood as limiting the present disclosure.

[0019] In order to illustrate the technical solution of the present invention, specific embodiments are provided below.

[0020] See also Figure 1 , is a method flow chart of a total station measurement data acquisition method provided by the first embodiment of the present invention, such as Figure 1 As shown, the method may include: Step S1 : obtaining an original measurement data set with time series labels by performing three-dimensional coordinate calculation and sorting on the original measurement data of the total station.

[0021] In the total station measurement data acquisition method of the present invention, step S1 is used to obtain a raw measurement data set with time series labels, which serves as the basic data source for subsequent gross error identification and processing. Specifically, multiple total station measurement monitoring points are first arranged in the structure to be measured or in the monitoring area. These monitoring points can be spatially arranged according to key structural parts, deformation-sensitive areas, or differences in ranging conditions.

[0022] By setting up multiple monitoring points on the structure or area to be measured, using the total station automated measurement system, periodic three-dimensional coordinate measurement of each monitoring point is performed within a preset measurement period, and the three-dimensional spatial coordinate value of the monitoring point corresponding to each measurement moment and its corresponding time tag are obtained; all the collected measurement results of the monitoring points are sorted in chronological order to form an original measurement data set containing the three-dimensional coordinates and time tags of each monitoring point at multiple measurement moments. The original measurement data set consists of multiple measurement data points, and each measurement data point contains its three-dimensional coordinate value and acquisition time information.

[0023] During the measurement process, a total station automated monitoring system was used to periodically measure the three-dimensional coordinates of each monitoring point. Each measurement operation recorded the measurement time and result data, including the three-dimensional coordinate values ​​of each monitoring point at that moment, along with the corresponding time stamp. Over time, the system organized the multiple measurement data points obtained in chronological order to form a raw measurement data set, where each data point contains the three-dimensional coordinate value and time stamp information.

[0024] This raw measurement data set comprehensively reflects the spatial distribution and temporal evolution of each measurement point during the monitoring period, providing complete input data for subsequent gross error identification algorithms based on density and structural features. In actual deployment, the sampling period for raw measurement data can be set based on project monitoring requirements. In this implementation, the sampling period is set to 10 minutes to ensure data timeliness and continuity.

[0025] At this point, the three-dimensional coordinate solution and organization of the original measurement data of the total station are completed to obtain the original measurement data set with time series labels.

[0026] Step S2: Obtain the adaptive neighborhood radius of each point by comprehensively analyzing the local neighborhood distances of the original measurement data points.

[0027] A limitation of the DBSCAN algorithm in the prior art is the global fixedness of its neighborhood radius. This setting conflicts with the uneven density characteristics of total station measurement data, which are dense near and sparse far away due to different distance measurements, resulting in the algorithm being unable to find an optimal neighborhood radius value that can be applied simultaneously to all monitoring points. To solve this problem, the present invention should no longer use a unified radius, but should dynamically evaluate a local neighborhood radius that best suits each data point in the data set based on the data distribution pattern of the area in which it is located. To achieve this goal, it is necessary to assess whether the area in which the data point is located is dense or sparse. The local neighborhood radius of a data point should be uniquely determined by the actual distribution of the data points around it. Specifically, it should simultaneously reflect two pieces of information: the average clustering scale and the maximum dispersion range of the points in its neighborhood. For a high-density data cluster, the local average neighborhood arm length and the local maximum neighborhood arm length of any point within it will be numerically very small. Conversely, for a low-density data cluster, the values ​​of these two indicators will increase significantly. Based on this principle, this step combines these two indicators that can jointly characterize the local scale and calculates an adaptive neighborhood radius that can accurately match the local data density. Specifically, the number of neighboring data points is set. In this embodiment of the present invention, the number of neighboring data points is set. The number of neighboring data points can be adjusted accordingly according to the specific scenario to obtain the set of neighboring data points of any target measurement data point in all measurement data points; for any target measurement data point, the Euclidean distance between the target measurement data point and the farthest data point in its neighboring data point set is taken as the local maximum neighborhood arm length of the target measurement data point; the average Euclidean distance between the target measurement data point and all data points in its neighboring data point set is taken as the local average neighborhood arm length of the target measurement data point; the square root of the sum of the squares of the local maximum neighborhood arm length and the local average neighborhood arm length of the target measurement data point is taken as the adaptive neighborhood radius of the target measurement data point.

[0028] In one embodiment, assuming that The measurement data point and The Euclidean distance between the measured data points is ; The number of neighboring data points is , No. The set of neighboring data points of a data point is , then The calculation expression of the adaptive neighborhood radius of a data point is:

[0029] in, Indicates the Adaptive neighborhood radius of data points; Indicates the The measurement data point and Measuring the Euclidean distance between data points; represents the maximum value function; Indicates the The set of neighboring data points of a data point; Indicates the number of neighboring data points.

[0030] It should be noted that the adaptive neighborhood radius designed in this invention aims to solve the problem caused by the global fixed neighborhood radius parameter in the prior art, and to achieve the adaptation of the local data scale by constructing an adaptive neighborhood radius. In the actual scenario of total station measurement, for a close-range monitoring point, its measurement data is highly concentrated due to the small error, resulting in its local maximum neighborhood arm length. and the local average neighborhood arm length The value of is small, and the adaptive neighborhood radius calculated by the formula is also a small value. This small neighborhood radius can form a fine judgment scale, thereby effectively identifying the small gross errors in the compact data cluster. In contrast, for a long-distance monitoring point, its measurement data is loosely distributed due to error accumulation, and its local maximum neighborhood arm length is and the local average neighborhood arm length The value of will increase significantly, and the calculated adaptive neighborhood radius will be amplified accordingly. This amplified calibrant forms a judgment scale that matches the actual distribution of the loose data cluster, ensuring that all normally distant measurement points are fully included and avoiding misjudgments caused by overly strict sizing. Furthermore, the fusion of the local maximum neighborhood arm length and the local average neighborhood arm length yields better results than using either alone. If only the local maximum neighborhood arm length is used to evaluate the adaptive neighborhood radius, the neighborhood radius is entirely determined by the farthest neighboring point, making it susceptible to excessive influence from the accidental location of that point. By introducing the local average neighborhood arm length, the calculated result integrates the average distribution of all measured data points in the neighborhood, reducing sensitivity to the location of a single farthest point. Furthermore, if only the local average neighborhood arm length is used, the defined radius may be smaller than the actual distance of the farthest neighbor, resulting in a failure to include all neighboring data points. By introducing the local maximum neighborhood arm length, the resulting adaptive neighborhood radius is guaranteed to be greater than or equal to the distance of the farthest neighbor, thus ensuring the integrity of the neighborhood definition.

[0031] At this point, the adaptive neighborhood radius of each point is obtained by comprehensively analyzing the local neighborhood distances of the original measurement data points.

[0032] Step S3: Acquire the shared local structure tensor between the point pairs by fusing and constructing the local structure tensors of the measurement points.

[0033] In actual total station measurement scenarios, especially in monitoring areas across complex terrain or with atmospheric disturbances, such as cross-river observations, long-distance measurements at valley bottoms, or in narrow urban streets, measurement data is often significantly affected by spatial directionality. Due to environmental factors such as water evaporation, surface thermal radiation, or building obstruction, the temperature, humidity, and density distribution of the atmospheric medium in the horizontal or vertical directions are uneven, which in turn causes the laser ranging path to produce varying degrees of lateral or vertical refraction. This atmospheric refraction effect has a clear directional characteristic and usually causes a systematic shift of the measurement point along a specific axis, resulting in the measurement data cluster, which should have been isotropically distributed, to take on an ellipsoidal shape with a significant stretch in the main axis direction.

[0034] In this type of data structure, if the standard Euclidean distance or single-point neighborhood structure is still used for data analysis, it is very easy to ignore the overall structural trend reflected by the directional deformation between the measurement data, thereby affecting the accuracy of gross error identification. To this end, the present invention proposes a method of constructing a shared local structure tensor to accurately describe the common neighborhood structure characteristics of the area where any two measurement data points are located. This structure tensor can not only reflect the geometric morphology and directional consistency of the measurement point neighborhood, but also provide a unified morphological benchmark for the subsequent construction of anisotropic distance metrics, ensuring that the distance calculation has good symmetry and structural adaptability between different point pairs.

[0035] To achieve this structural expression, it is first necessary to establish a mathematical representation of the local structure tensor based on the spatial relative position between each measurement data point and its neighboring data points, and further fuse the respective tensor information between any pair of points to generate a unified expression representing the shared local structure of the point pair. Specifically, for any target measurement data point's neighbor data set, the vector whose geometric center points to any target neighborhood measurement data point in the neighbor data set is used as the position vector of the target measurement data point with respect to the target neighborhood measurement data point; the outer product of the position vector of the target measurement data point with respect to the target neighborhood measurement data point and itself is used as the first vector outer product of the target measurement data point with respect to the target neighborhood measurement data point; for any target measurement data point's neighbor data set, the sum of all first vector outer products in the neighbor data set is used as the local structure tensor of the target measurement data point; for the local structure tensors of any two target measurement data points, the average of the local structure tensors of the two target measurement data points is used as the shared local structure tensor of the two target measurement data points.

[0036] In one embodiment, assuming that The center of the set of neighboring data points of a measured data point points to the The position vector of the data point is ;No. The center of the set of neighboring data points of a measured data point points to the The position vector of the data point is ; then The measurement data point and The calculation expression of the shared local structure tensor of the measured data points is:

[0037] in, Indicates the The measurement data point and The shared local structure tensor of the measured data points; Indicates the The center of the set of neighboring data points of a measured data point points to the The position vector of the data points; Indicates the The center of the set of neighboring data points of a measured data point points to the The position vector of the data points; Indicates the The set of neighboring data points of a data point; Indicates the The set of neighboring data points of a data point; Represents the outer product operation between vectors.

[0038] At this point, the shared local structure tensor between the point pairs is obtained by fusing the local structure tensors of the measurement points.

[0039] Step S4: performing principal axis deviation analysis on the shared local structure tensor to obtain a morphologically adaptive symmetric anisotropic distance metric.

[0040] After completing the construction of the shared local structure tensor between pairs of measurement data points, the present invention further introduces a morphologically adaptive anisotropic distance measurement mechanism to replace the traditional Euclidean distance calculation method, targeting the directional structural characteristics of the measurement data clusters in actual monitoring scenarios. In standard clustering methods, the distance between measurement points is usually calculated using Euclidean distance. This method implicitly assumes that the data clusters present an isotropic structure in space, that is, the distribution characteristics of the point cloud in all directions are basically the same. However, in the actual total station measurement process, especially in scenarios with strong directional disturbances, such as crossing rivers, reservoirs, canyons, or in urban alleys along the street, the measurement data points are often affected by the refractive disturbance caused by atmospheric temperature gradients, water surface evaporation effects, or building obstructions, causing the point cloud to be systematically stretched in a certain main direction, while relatively convergent or unchanged in other directions. Taking river-crossing measurements as an example, as laser signals pass through the atmosphere above the river, they are affected by water evaporation and temperature differences between the two banks, causing horizontal deflection of the optical path. This leads to a directional shift in the plane coordinates, resulting in a measurement point trajectory stretched along the river. The vertical refraction is weaker, making the fluctuation characteristics of elevation changes significantly inconsistent with the horizontal coordinates. Similarly, in urban streets, the alternating presence of hot walls and shadows on either side of buildings at different times can cause temperature unevenness across the street, leading to a shift in the measurement point set along the street. In this asymmetric data structure with a stretched principal axis, continuing to use Euclidean distance to determine the neighborhood between measurement points presents two serious problems: first, the distance calculation ignores the actual morphological shift caused by the principal axis of the point cloud, leading to misjudgment of directional connectivity; second, the independent distance calculations for different measurement points due to local structural differences lead to asymmetric distance metrics, which in turn undermines the stability and repeatability that clustering algorithms should possess. To overcome these problems, the present invention proposes an anisotropic distance metric based on a shared local structure tensor. The core idea of ​​this method is: first, perform eigendecomposition on the shared structural tensor between point pairs to extract the principal axis vector that can represent the main stretching direction of the point cloud; then, combine the direction of the line between the measured point pairs to analyze the degree of deviation from the principal axis direction; finally, construct a directional penalty mechanism based on the degree of deviation to dynamically adjust the original Euclidean distance between the point pairs, thereby forming a morphologically adaptive distance measurement method that is more adaptable to the real structural morphology.

[0041] First, the shared local structure tensor is decomposed by eigenvectors to obtain a set of shared principal axis directions. Specifically, the shared local structure tensor is decomposed by eigenvectors to obtain eigenvalue-eigenvector pairs corresponding to the number of dimensions of the shared local structure tensor; all eigenvalue-eigenvector pairs are arranged in descending order according to the size of the eigenvalues, and the eigenvector corresponding to the first-ranked eigenvalue is obtained as the principal axis direction of the shared local structure tensor, and the eigenvectors corresponding to the remaining eigenvalues ​​are the secondary axis directions in turn; the set formed by the primary axis direction and the secondary axis direction is taken as the shared principal axis direction set.

[0042] Afterwards, a shared local principal axis deviation factor is obtained by performing projection analysis on the direction of the line connecting the measurement point pairs and the principal axis direction set. Specifically, for any two target measurement data points, the direction vector between the two target measurement data points is obtained through the coordinates of the two target measurement data points; the dot product calculation result of the direction vector between the two target measurement data points and any eigenvector in the shared principal axis direction set of the two target measurement data points is used as the first vector dot product between the two target measurement data points; the calculation result of the sum of all the first vector dot products between the two target measurement data points is used as the denominator, the calculation result of the sum of all the first vector dot products between the two target measurement data points except the principal axis direction is used as the numerator, and the calculation result of the corresponding fraction is used as the shared local principal axis deviation factor of the two target measurement data points.

[0043] In one embodiment, assuming that The measurement data point points to the The direction vector of the measured data point is ;No. The measurement data point and The first part of the shared local structure tensor of the measured data points The eigenvectors are ; The data latitude of the measurement data is , then The measurement data point and The calculation expression of the shared local principal axis deviation factor of the measured data points is:

[0044] in, Indicates the The measurement data point and The shared local principal axis deviation factor of the measured data points; Indicates the The measurement data point points to the The direction vector of each measurement data point; Indicates the The measurement data point and The first part of the shared local structure tensor of the measured data points feature vectors; Indicates the data latitude of the measurement data.

[0045] It should be noted that the symmetric anisotropic distance metric designed in this invention aims to solve the problem of accurately identifying anisotropic data clusters caused by physical factors such as instrument axis errors or atmospheric lateral refraction. Its core lies in the introduction of the shared local structure tensor. This concept. and The respective local structure tensors are averaged and fused. It no longer takes a biased view to any one side, but objectively reflects the geometric shape of the transition area on the ellipsoid trajectory where the pair of points are located. Shared principal axes obtained by eigendecomposition Become an unbiased reference coordinate system that can accurately describe the current extension direction of the trajectory. Based on this shared benchmark that can accurately reflect the local data form, the shared local principal axis deviation factor The calculation also has mathematical symmetry. Quantitatively measures the line connecting two points direction, and the natural extension direction of the data track where it is located (i.e., sharing the main extension axis ) are consistent. If the direction of the line connecting the two points is exactly the extension direction of the data track, The value of ; If the direction of the line connecting the two points is perpendicular to the extension direction of the trajectory, The value of Due to the calculation When using the same shared tensor A vector with opposite directions but collinear , the sum of squares of its projections on each principal axis is are exactly the same, so there are , ensuring the symmetry of distance.

[0046] Finally, the morphologically adaptive symmetric anisotropic distance is obtained by fusing and weighting the original Euclidean distance and the principal axis deviation factor. Specifically, the result of adding the constant 1 to the shared local principal axis deviation factor is used as the first distance optimization weight, and the result of multiplying the first distance optimization weight of any two target measurement data points by the Euclidean distance between the two target measurement data points is used as the morphologically adaptive symmetric anisotropic distance between the two target measurement data points.

[0047] In one embodiment, the The measurement data point and The calculation expression of the morphologically adaptive symmetric anisotropic distance between measurement data points is:

[0048] in, Indicates the The measurement data point and The morphologically adaptive symmetric anisotropic distance between the measured data points; Indicates the The measurement data point and The shared local principal axis deviation factor of the measured data points; Indicates the The measurement data point and Measures the Euclidean distance between data points.

[0049] It should be noted that the shared local principal axis deviation factor is applied to the calculation of the symmetric anisotropic distance, thereby realizing a mechanism that can intelligently identify and adapt to the direction-sensitive distance penalty. Explore its neighbors When, if The subsequent points on the same trajectory are also located due to the continuous change of the lateral refraction effect, so the direction of their connection is consistent with the extension direction of the data cluster. will approach , morphologically adaptive symmetric anisotropic distance This allows the algorithm to extend along the trajectory formed by the physical process, thereby completely identifying the entire elongated ellipsoidal data cluster as a whole. On the contrary, if is a gross error outside the trajectory caused by other accidental factors, then arrive The line connecting the two will inevitably deviate from the main axis of the trajectory. will approach , its effective distance This distance penalty will effectively prevent the algorithm from jumping out of the normal trajectory to connect an irrelevant gross error.

[0050] At this point, the morphologically adaptive symmetric anisotropic distance metric is obtained by analyzing the principal axis deviation of the shared local structure tensor.

[0051] Step S5: By jointly judging the morphological adaptive distance and the neighborhood radius, a gross error point recognition result is obtained and abnormal measurement data is eliminated.

[0052] After obtaining the adaptive neighborhood radius and the morphologically adaptive symmetric anisotropic distance metric of the target measurement data point, a minimum data point number threshold can be set, and the measurement data points in the original measurement data set can be clustered according to the adaptive neighborhood radius of the measurement data point and the morphologically adaptive symmetric anisotropic distance between the measurement data points. It should be noted that in the embodiment of the present invention, the minimum data point number threshold is set to 15. This threshold setting can be set differently according to specific circumstances and is not required; in the clustering process, when the algorithm needs to determine the measurement data points, Whether data points are being measured When within the neighborhood of The adaptive neighborhood radius and the measured data points and A morphologically adaptive symmetric anisotropic distance metric between two data points. and The morphologically adaptive symmetric anisotropic distance metric between the measured data points is less than or equal to The measurement data point is determined only when the adaptive neighborhood radius is Located at the measurement data point within the neighborhood.

[0053] Complete the DBSCAN clustering process and obtain the core points, boundary points and noise points marked in the clustering process; treat the measurement data points marked as noise points as gross error data points and delete the gross error data points from the original data set to obtain the optimized measurement data set.

[0054] At this point, the morphological adaptive distance and the neighborhood radius are jointly judged to obtain the gross error point identification results and eliminate abnormal measurement data.

[0055] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A total station measurement data acquisition method, characterized in that: The total station measurement data acquisition method comprises: Step S1: Obtaining an original measurement data set with time series labels by performing three-dimensional coordinate calculation and sorting on the original measurement data of the total station; Step S2: Obtain the adaptive neighborhood radius of each point by comprehensively analyzing the local neighborhood distances of the original measurement data points; Step S3: Obtain the shared local structure tensor between the point pairs by fusing and constructing the local structure tensors of the measurement points; Step S4: Obtaining a morphologically adaptive symmetric anisotropic distance metric by performing principal axis deviation analysis on the shared local structure tensor; Step S5: By jointly judging the morphological adaptive distance and the neighborhood radius, the gross error point identification result is obtained and abnormal measurement data is eliminated.

2. A total station measurement data acquisition method according to claim 1, characterized in that: The three-dimensional coordinates of the original measurement data of the total station are calculated and sorted to obtain the original measurement data set with time series labels, including: By setting up multiple monitoring points on the structure or area to be measured, using the total station automated measurement system, periodic three-dimensional coordinate measurement of each monitoring point is performed within a preset measurement period, and the three-dimensional spatial coordinate value of the monitoring point corresponding to each measurement moment and its corresponding time tag are obtained; all the collected measurement results of the monitoring points are sorted in chronological order to form an original measurement data set containing the three-dimensional coordinates and time tags of each monitoring point at multiple measurement moments. The original measurement data set is composed of multiple measurement data points, and each measurement data point contains its three-dimensional coordinate value and acquisition time information.

3. The total station measurement data acquisition method according to claim 1, characterized in that: The adaptive neighborhood radius of each point is obtained by comprehensively analyzing the local neighborhood distance of the original measurement data points, including: The number of neighboring data points is set to obtain the set of neighboring data points of any target measurement data point in all measurement data points; for any target measurement data point, the Euclidean distance between the target measurement data point and the farthest data point in its neighboring data point set is taken as the local maximum neighborhood arm length of the target measurement data point; the average Euclidean distance between the target measurement data point and all data points in its neighboring data point set is taken as the local average neighborhood arm length of the target measurement data point; the square root of the sum of the squares of the local maximum neighborhood arm length and the local average neighborhood arm length of the target measurement data point is taken as the adaptive neighborhood radius of the target measurement data point.

4. The total station measurement data acquisition method according to claim 1, characterized in that: The method of fusing and constructing the local structure tensors of the measuring points to obtain the shared local structure tensors between the point pairs includes: For any target measurement data point's neighbor data set, the vector whose geometric center points to any target neighborhood measurement data point in the neighbor data set is used as the position vector of the target measurement data point relative to the target neighborhood measurement data point; the outer product of the position vector of the target measurement data point relative to the target neighborhood measurement data point and itself is used as the first vector outer product of the target measurement data point relative to the target neighborhood measurement data point; for any target measurement data point's neighbor data set, the sum of all first vector outer products in the neighbor data set is used as the local structure tensor of the target measurement data point; for the local structure tensors of any two target measurement data points, the average of the local structure tensors of the two target measurement data points is used as the shared local structure tensor of the two target measurement data points.

5. The total station measurement data acquisition method according to claim 1, characterized in that: The method of obtaining a morphologically adaptive symmetric anisotropic distance metric by performing principal axis deviation analysis on the shared local structure tensor includes: The shared principal axis direction set is obtained by performing eigenvector decomposition on the shared local structure tensor. The shared local principal axis deviation factor is obtained by performing projection analysis on the direction of the connection line of the measurement point pair and the principal axis direction set. The morphologically adaptive symmetric anisotropic distance is obtained by weighted fusion of the original Euclidean distance and the principal axis deviation factor.

6. The total station measurement data acquisition method according to claim 1, characterized in that: The step of performing eigenvector decomposition on the shared local structure tensor to obtain a shared principal axis direction set includes: Perform eigenvector decomposition on the shared local structure tensor to obtain eigenvalue-eigenvector pairs corresponding to the number of dimensions of the shared local structure tensor; sort all eigenvalue-eigenvector pairs in descending order according to the size of the eigenvalue, and obtain the eigenvector corresponding to the first-ranked eigenvalue as the principal axis direction of the shared local structure tensor, and the eigenvectors corresponding to the remaining eigenvalues ​​are the secondary axis directions in turn; the set formed by the principal axis direction and the secondary axis direction is taken as the shared principal axis direction set.

7. A total station measurement data acquisition method according to claim 2, characterized in that: The method of obtaining the shared local principal axis deviation factor by performing projection analysis on the direction of the line connecting the measurement point pairs and the principal axis direction set includes: For any two target measurement data points, a direction vector between the two target measurement data points is obtained through the coordinates of the two target measurement data points; a dot product calculation result of the direction vector between the two target measurement data points and any eigenvector in the shared principal axis direction set of the two target measurement data points is used as a first vector dot product between the two target measurement data points; The calculation result of adding all the first vector dot products between the two target measurement data points is used as the denominator, the calculation result of adding all the first vector dot products between the two target measurement data points except the principal axis direction is used as the numerator, and the calculation result of the corresponding fraction is used as the shared local principal axis deviation factor of the two target measurement data points.

8. The total station measurement data acquisition method according to claim 1, characterized in that: The method of obtaining the morphologically adaptive symmetric anisotropic distance by fusing and weighting the original Euclidean distance and the principal axis deviation factor includes: The result of adding the constant 1 to the shared local principal axis deviation factor is used as the first distance optimization weight, and the result of multiplying the first distance optimization weight of any two target measurement data points by the Euclidean distance between the two target measurement data points is used as the morphologically adaptive symmetric anisotropic distance between the two target measurement data points.

9. The total station measurement data acquisition method according to claim 1, characterized in that: The method of obtaining gross error point identification results and eliminating abnormal measurement data by jointly judging the morphological adaptive distance and the neighborhood radius includes: A minimum data point number threshold is set, and DBSCAN clustering is performed on the measurement data points in the original measurement data set according to the adaptive neighborhood radius of the measurement data points and the morphologically adaptive symmetric anisotropic distance between the measurement data points. The clustering process is completed and the core points, boundary points and noise points marked in the clustering process are obtained; the measurement data points marked as noise points are regarded as gross error data points and the gross error data points are deleted from the original data set to obtain the measurement data set without abnormal measurement data.

10. A total station measurement data acquisition system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a total station measurement data acquisition method according to any one of claims 1 to 9 is implemented.

Images