An intelligent surveying and mapping method suitable for complex terrain

Abstract

The present invention discloses an intelligent surveying and mapping method suitable for complex terrain, which specifically relates to the field of data processing technology. Target clusters are obtained through surveying and mapping data collection, point cloud registration and cluster analysis, and feature comparison points are set in the target clusters to evaluate the measurement accuracy of the target clusters. Abnormal target clusters are screened based on the relationship between quality accuracy parameters and preset values; after screening abnormal clusters, rescan and verify to ensure data quality. Finally, the updated target clusters are integrated to construct an accurate three-dimensional complex terrain simulation model. By setting the registration base point to reduce the registration resource consumption and time, the clustering algorithm is used to identify the terrain target clusters, and the number of feature comparison reference points in the cluster is dynamically adjusted according to the complexity of the terrain to ensure that the accuracy of each cluster meets the standard; by comparing and analyzing the similarity between the actual and surveying feature parameters, the quality of the target cluster is evaluated, and the abnormal clusters are effectively identified and processed, which effectively improves the surveying and mapping accuracy and data processing efficiency.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically, to an intelligent surveying and mapping method suitable for complex terrain. Background Art

[0002] In the field of surveying and mapping, the surveying and mapping of complex terrain has always been one of the key technical challenges. Traditional surveying and mapping methods often rely on manual measurement and manual data processing, which is not only time-consuming and labor-intensive, but also easily affected by human errors and environmental factors, making it difficult to ensure the accuracy and reliability of surveying and mapping results. In particular, when faced with complex terrains such as mountains, forests, and densely populated urban areas, traditional methods are even more inadequate.

[0003] With the advancement of science and technology, especially the development of modern surveying and mapping technologies such as 3D laser scanning and drone aerial photography, the acquisition of surveying and mapping data has become more efficient and comprehensive. However, how to efficiently process these massive, high-dimensional surveying and mapping data, especially how to accurately align, analyze and model point cloud data of complex terrain, is still a technical problem that needs to be solved urgently.

[0004] Point cloud data, as an important output of modern surveying and mapping technology, contains rich terrain information, but its processing is complex and prone to errors. Especially in complex terrain, due to the large terrain undulations and many obstructions, point cloud data often has serious distortion and noise, which brings great challenges to subsequent data processing and analysis.

[0005] In addition, traditional surveying and mapping methods often lack intelligent means in the data processing process, and cannot effectively identify and eliminate abnormal data, resulting in deviations in the final surveying and mapping results. At the same time, the feedback and application of surveying and mapping data are relatively lagging, making it difficult to meet users' needs for real-time, high-precision surveying and mapping data, and there is a problem that the surveying and mapping quality of complex terrain does not meet the requirements. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides an intelligent surveying and mapping method applicable to complex terrain to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: an intelligent surveying and mapping method suitable for complex terrain, comprising the following steps:

[0008] Step 1: Surveying and mapping data collection: Set the target complex terrain as the area to be detected, evenly set the registration base points in the area to be detected, and collect the surveying and mapping data of the area to be detected through the measuring equipment. The surveying and mapping data includes the point cloud data of the area to be detected and the multi-dimensional attribute parameters of the registration base points;

[0009] Step 2: Point cloud data registration: According to the actual multi-dimensional attribute parameters of the registration base points and the measured multi-dimensional attribute parameters, the point cloud transformation matrix of the area to be detected is obtained, the point cloud data is registered through the point cloud transformation matrix, the registered point cloud data is obtained, and the clustering cluster set is obtained through the clustering algorithm;

[0010] Step 3: Obtain target clusters: Evaluate the acceptable error level of each cluster, use the acceptable error level as a constraint, obtain target clusters corresponding to complex terrain, and number them;

[0011] Step 4: Filter abnormal target clusters, set feature comparison reference points in each target cluster, obtain actual feature parameters of feature comparison reference points, obtain surveying and mapping feature parameters of feature comparison reference points from the registered point cloud data, compare and analyze the similarity between actual feature parameters and surveying and mapping feature parameters, and obtain measurement accuracy parameters of target clusters; filter out abnormal target clusters based on quality accuracy parameter zc;

[0012] Step 5: Update the abnormal target cluster: rescan to obtain the surveying and mapping data of the abnormal target cluster, verify the measurement accuracy parameters of the surveying and mapping data; obtain a target cluster set that meets the quality requirements; register and fuse the target cluster set to obtain a three-dimensional simulation model of the target complex terrain;

[0013] Step 6: Surveying and mapping data feedback: The surveying and mapping data, target cluster information and three-dimensional simulation model are fed back to the user end, and the user queries the surveying and mapping data and three-dimensional simulation model through the port.

[0014] Preferably, the point cloud transformation matrix is ​​obtained by recording the actual multidimensional attribute parameters of the registration base points as the source point set, and recording the measured multidimensional attribute parameters of the registration base points as the target point set; the point cloud transformation matrix includes rotation, translation, scaling and tilt, and is obtained by minimizing the sum of the squares of the Euclidean distance between the transformed source point set and the target point set.

[0015] Preferably, the number of feature comparison reference points of each target cluster is allocated according to the area, contour perimeter and error tolerance of the target cluster; the number of feature comparison reference points is obtained by setting the number of feature comparison reference points based on the geometric complexity and texture complexity of the target cluster, and being able to dynamically adjust the number of feature comparison reference points according to the terrain complexity, and by formula n d =Δd+INT[(Jk qj +wk qw )*e 1+wc ] Calculate the number of feature comparison reference points n d; Wherein, Δd represents the minimum number of reference points for feature comparison, wk represents texture complexity, Jk represents geometric complexity, qj represents the influence index of geometric complexity, qw represents the influence index of texture complexity, and wc represents the acceptable degree of error; INT(·) represents the rounding function.

[0016] Preferably, the geometric complexity Jk is obtained by: obtaining the area of ​​the target cluster as s0 and the perimeter of the contour as s1, and using the formula The geometric complexity Jk is calculated; the texture complexity wk is obtained by analyzing the remote sensing image corresponding to the target cluster and obtaining the texture features based on the gray-level co-occurrence matrix; and the texture complexity wk is calculated by weighted summation of the texture features.

[0017] Preferably, the mass accuracy parameter of the target cluster is obtained by:

[0018] Assume that each target cluster contains n d feature comparison reference point, i represents the number of the feature comparison reference point;

[0019] Compare and analyze the similarity between the actual feature parameters and the surveyed feature parameters, and calculate the position similarity, elevation similarity, texture similarity and terrain feature cross entropy loss value; the position similarity, elevation similarity, texture similarity and terrain feature cross entropy loss value of the i-th feature comparison reference point are recorded as x1i, x2i, x3i, x4i respectively;

[0020] By formula The quality accuracy parameter zc_j of the j-th target cluster is calculated, where w1, w2, w3, and w4 represent weight coefficients of each item, and w1+w2+w3+w4=1.

[0021] Preferably, the location similarity is obtained by: calculating the Euclidean distance between two points as dxy; defining the location similarity as the reciprocal of the distance or the distance exponential transformation, and using the formula The position similarity is calculated; the elevation similarity is obtained by using the inverse of the square of the difference between the two elevation values ​​to represent the elevation similarity; if the elevation values ​​of the two points are very close, the elevation similarity is high, and the formula The elevation similarity is calculated, where z1 and z2 are the elevation values ​​of the two points respectively, and ∈1 is a very small positive number used to prevent the denominator from being zero; the texture similarity is obtained by using the similarity of the image histogram to approximate the texture similarity; the terrain feature cross entropy loss value is obtained by encoding the terrain feature into a probability distribution and using the cross entropy loss to evaluate the difference between the actual distribution and the predicted distribution.

[0022] Preferably, the screening method of abnormal target clusters is:

[0023] Determine the quality accuracy parameter zc and threshold Y zc If zc≥Y zc , the data quality of the target cluster is judged to be qualified if zc<Y zc , the data quality of the target cluster is judged to be unqualified and marked as an abnormal target cluster.

[0024] Preferably, the intelligent surveying and mapping method applicable to complex terrain also includes: a comprehensive management step of surveying and mapping data quality, which ensures the comprehensive quality of surveying and mapping data by verifying the surveying and mapping data quality coefficient.

[0025] Preferably, the process of obtaining the surveying and mapping data quality coefficient comprises the following steps:

[0026] Step S11, obtaining the actual registration speed v of the point cloud data ac and preset registration speed v pr ; Obtain the quality accuracy parameters of each target cluster, the theoretical feature data set of the target cluster, the actual feature data set UA of the target cluster and the preset feature data set UB;

[0027] Step S12: by formula The registration velocity loss rate Pz of the three-dimensional simulation model is calculated;

[0028] Step S13: By formula The measurement quality accuracy index Zc of the three-dimensional simulation model is calculated; where k_j represents the weight coefficient of the j-th target cluster, and the weight coefficient is obtained by obtaining the area of ​​each target cluster and recording the area of ​​the j-th target cluster as ms_j; through the formula Calculate the weight coefficient;

[0029] Step S14, using the ratio of the intersection and union of the actual feature data set UA and the preset feature data set UB to represent the integrity parameter Wz of the three-dimensional simulation model of the target surveying and mapping area;

[0030] Step S15, jointly analyzing the registration speed loss rate, measurement quality accuracy index Zc and three-dimensional simulation model integrity parameter Wz of the three-dimensional simulation model; by formula The joint analysis yields the mapping data quality coefficient CZ;

[0031] Step S16: Determine the mapping data quality coefficient and threshold C 阈 If CZ≥C 阈 , indicating that the quality requirements of surveying and mapping data for complex terrain meet the requirements; if CZ<C 阈 , indicating that the quality coefficient of the surveying and mapping data does not meet the requirements, and warns the user to maintain the measuring equipment and re-collect the surveying and mapping data.

[0032] Technical effects and advantages of the present invention:

[0033] (1) The intelligent surveying and mapping method for complex terrain provided by the present invention can improve the processing efficiency and accuracy of complex terrain point cloud data by optimizing the registration base points and dynamically adjusting the feature comparison reference points. The original surveying and mapping data is adjusted using the solved point cloud transformation matrix, saving subsequent registration time, thereby solving the problem that the surveying and mapping quality of complex terrain does not meet the requirements.

[0034] (2) In the intelligent mapping method for complex terrain provided by the present invention, the preset feature data set UB is used to compare with the actual feature data set UA to evaluate the accuracy and completeness of the mapping data. By comparing the differences between the two data sets, the errors, omissions or inconsistencies in the mapping process are identified. At the same time, combined with the registration speed loss rate Pz and the measurement quality accuracy index Zc, the comprehensive quality of the mapping data is comprehensively evaluated to ensure the accuracy and reliability of the mapping results, and solve the problem of difficult to ensure the mapping quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 The figure is a flow chart of the intelligent surveying and mapping method applicable to complex terrain according to the present invention.

[0036] Figure 2 This is a flow chart of the comprehensive management of surveying and mapping data quality of the present invention. DETAILED DESCRIPTION

[0037] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0038] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0039] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present application, its application, or uses.

[0040] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0041] Example 1

[0042] See also Figure 1The present invention provides a flow chart of an intelligent surveying and mapping method suitable for complex terrain. Figure 1 An intelligent mapping method suitable for complex terrain is shown, comprising the following steps:

[0043] Step 1: Surveying and mapping data collection: Set the target complex terrain as the area to be detected, evenly set the registration base points in the area to be detected, and collect the surveying and mapping data of the area to be detected through the measuring equipment. The surveying and mapping data includes the point cloud data of the area to be detected and the multi-dimensional attribute parameters of the registration base points;

[0044] Step 2: Point cloud data registration: According to the actual multi-dimensional attribute parameters of the registration base points and the measured multi-dimensional attribute parameters, the point cloud transformation matrix of the area to be detected is obtained, the point cloud data is registered through the point cloud transformation matrix, the registered point cloud data is obtained, and the clustering cluster set is obtained through the clustering algorithm;

[0045] Step 3: Obtain target clusters: Evaluate the acceptable error level of each cluster, use the acceptable error level as a constraint, obtain target clusters corresponding to complex terrain, and number them;

[0046] Step 4: Filter abnormal target clusters, set feature comparison reference points in each target cluster, obtain actual feature parameters of feature comparison reference points, obtain surveying and mapping feature parameters of feature comparison reference points from the registered point cloud data, compare and analyze the similarity between actual feature parameters and surveying and mapping feature parameters, and obtain measurement accuracy parameters of target clusters; filter out abnormal target clusters based on quality accuracy parameter zc;

[0047] Step 5: Update the abnormal target cluster: rescan to obtain the surveying and mapping data of the abnormal target cluster, verify the measurement accuracy parameters of the surveying and mapping data; obtain a target cluster set that meets the quality requirements; register and fuse the target cluster set to obtain a three-dimensional simulation model of the target complex terrain;

[0048] Step 6: Surveying and mapping data feedback: The surveying and mapping data, target cluster information and three-dimensional simulation model are fed back to the user end, and the user queries the surveying and mapping data and three-dimensional simulation model through the port.

[0049] What needs to be further explained in the embodiments of the present invention is that the point cloud transformation matrix is ​​obtained in the following manner: the actual multi-dimensional attribute parameters of the registration base points are recorded as the source point set, and the measured multi-dimensional attribute parameters of the registration base points are recorded as the target point set; the point cloud transformation matrix includes rotation, translation, scaling and tilt, and is obtained by minimizing the sum of the squares of the Euclidean distances between the transformed source point set and the target point set.

[0050] Specifically, the actual multidimensional attribute parameters of the registration base points (i.e., the real multidimensional attribute parameters of the registration base points in the actual terrain, such as real coordinates and attributes) are defined as the source point set; the measured multidimensional attribute parameters of the registration base points obtained during the measurement process (which may be affected by measurement errors, equipment limitations, or terrain complexity) are defined as the target point set; these two point sets will be used to solve the subsequent point cloud transformation matrix; in the solution process, the four transformation parameters of rotation, translation, scaling, and tilt need to be considered simultaneously. Rotation and translation can be preliminarily solved by methods such as singular value decomposition, but scaling and tilt usually require a more complex optimization process; scaling parameters can be estimated by comparing the overall size of the point cloud before and after the transformation, and the tilt parameters can be solved by defining a specific tilt model and using it as part of the optimization problem.

[0051] What needs to be further explained in the embodiments of the present invention is that the number of feature comparison reference points is allocated to each target cluster according to the area, contour circumference and acceptable error level of the target cluster; the number of feature comparison reference points is obtained in the following way: the number of feature comparison reference points is set based on the geometric complexity and texture complexity of the target cluster, and the number of feature comparison reference points can be dynamically adjusted according to the complexity of the terrain.

[0052] It is necessary to further explain in the embodiment of the present invention that, by the formula n d =Δd+INT[(Jk qj +wk qw )*e 1+wc ] Calculate the number of feature comparison reference points n d ; Wherein, Δd represents the minimum number of reference points for feature comparison, wk represents texture complexity, Jk represents geometric complexity, qj represents the influence index of geometric complexity, qw represents the influence index of texture complexity, and wc represents the acceptable degree of error; INT(·) represents the rounding function.

[0053] It is necessary to further explain in the embodiment of the present invention that the geometric complexity Jk is obtained by: obtaining the area of ​​the target cluster as s0 and the perimeter of the contour as s1, and using the formula The geometric complexity Jk is calculated; the texture complexity wk is obtained by analyzing the remote sensing image corresponding to the target cluster and obtaining the texture features based on the grayscale co-occurrence matrix; the texture features in the embodiment of the present invention include: image contrast, image uniformity, image entropy value and correlation; the texture complexity wk is calculated by weighted summation of multiple texture features (such as information entropy, edge ratio, contrast factor, correlation and energy, etc.).

[0054] It needs to be further explained in the embodiment of the present invention that the mass accuracy parameter of the target cluster is obtained in the following manner:

[0055] Assume that each target cluster contains n d feature comparison reference point, i represents the number of the feature comparison reference point;

[0056] Compare and analyze the similarity between the actual feature parameters and the surveyed feature parameters, and calculate the position similarity, elevation similarity, texture similarity and terrain feature cross entropy loss value; the position similarity, elevation similarity, texture similarity and terrain feature cross entropy loss value of the i-th feature comparison reference point are recorded as x1i, x2i, x3i, x4i respectively;

[0057] By formula The quality accuracy parameter zc_j of the j-th target cluster is calculated, where w1, w2, w3, and w4 represent weight coefficients of each item, and w1+w2+w3+w4=1.

[0058] It is necessary to further explain in the embodiment of the present invention that the position similarity is obtained by calculating the spatial distance between two points, namely the actual feature parameter point and the surveying feature parameter point, assuming that the coordinates of the two points are (x1, y1) and (x2, y2), respectively, and calculating the Euclidean distance between the two points as dxy; defining the position similarity as the reciprocal of the distance or the distance exponential transformation, through the formula Calculate the location similarity;

[0059] The elevation similarity is obtained by comparing the elevation values ​​of two points and using the inverse of the square of the difference between the two elevation values ​​to represent the elevation similarity; if the elevation values ​​of the two points are very close, the elevation similarity is high, and the formula is The elevation similarity is calculated, where z1 and z2 are the elevation values ​​of the two points, and ∈1 is a small positive number to prevent the denominator from being zero;

[0060] The texture similarity is obtained by comparing the texture features of two regions (such as gray level co-occurrence matrix, local binary pattern, etc.) to calculate, and using the similarity of image histograms (such as Bhattacharyya distance, histogram intersection, etc.) to approximate the texture similarity;

[0061] The terrain feature cross entropy loss value is obtained by encoding the terrain feature into a probability distribution and using the cross entropy loss to evaluate the difference between the actual distribution and the predicted distribution.

[0062] It needs to be further explained in the embodiment of the present invention that the screening method of the abnormal target cluster is:

[0063] Determine the quality accuracy parameter zc and threshold Y zc If zc≥Y zc , the data quality of the target cluster is judged to be qualified if zc<Y zc, the data quality of the target cluster is judged to be unqualified and marked as an abnormal target cluster.

[0064] In summary: The embodiments of the present invention can save the registration resources of point cloud data and reduce the registration time by setting the registration base points; based on the point cloud data after the registration adjustment, a clustering algorithm is used to obtain a target cluster of complex terrain; based on the characteristics of the target cluster, the number of feature comparison reference points in the target cluster is set, and the feature comparison reference points in each target cluster are dynamically adjusted to achieve the setting according to the complexity of the terrain. By increasing the feature comparison reference points, the accuracy of the target cluster can be increased; the feature comparison reference points are allocated according to the characteristics of the target cluster to ensure that the accuracy of each target cluster meets the requirements; based on the comparative analysis of the similarity between the actual feature parameters and the surveying and mapping feature parameters, the similarity parameters are jointly analyzed to obtain the quality accuracy parameters of each target cluster, and the abnormal target clusters are screened out, thereby effectively improving the surveying and mapping quality of complex terrain.

[0065] Example 2

[0066] See also Figure 2 The flow chart of comprehensive management of surveying and mapping data quality is shown in the embodiment of the present invention. The difference between the embodiment of the present invention and embodiment 1 is that the intelligent surveying and mapping method suitable for complex terrain also includes: a comprehensive management step of surveying and mapping data quality, which ensures the comprehensive quality of surveying and mapping data by verifying the surveying and mapping data quality coefficient. The process of obtaining the surveying and mapping data quality coefficient includes the following steps:

[0067] Step S11, obtaining the actual registration speed v of the point cloud data ac and preset registration speed v pr ; Obtain the quality accuracy parameters of each target cluster, the theoretical feature data set of the target cluster, the actual feature data set UA of the target cluster and the preset feature data set UB;

[0068] Step S12: by formula The registration velocity loss rate Pz of the three-dimensional simulation model is calculated;

[0069] Step S13: By formula The measurement quality accuracy index Zc of the three-dimensional simulation model is calculated; where k_j represents the weight coefficient of the j-th target cluster, and the weight coefficient is obtained by obtaining the area of ​​each target cluster and recording the area of ​​the j-th target cluster as ms_j; through the formula Calculate the weight coefficient;

[0070] Step S14, using the ratio of the intersection and union of the actual feature data set UA and the preset feature data set UB to represent the integrity parameter Wz of the three-dimensional simulation model of the target surveying and mapping area;

[0071] Step S15, jointly analyzing the registration speed loss rate, measurement quality accuracy index Zc and three-dimensional simulation model integrity parameter Wz of the three-dimensional simulation model; by formula The joint analysis yields the mapping data quality coefficient CZ;

[0072] Step S16: Determine the mapping data quality coefficient and threshold C 阈 relationship; if CZ ≥ C 阈 , indicating that the quality requirements of surveying and mapping data for complex terrain meet the requirements; if CZ<C 阈 , indicating that the quality coefficient of the surveying and mapping data does not meet the requirements, and warns the user to maintain the measuring equipment and re-collect the surveying and mapping data.

[0073] Specific explanation: The actual feature dataset UA refers to the target cluster obtained through the actual surveying and mapping process, such as using LiDAR and UAV photogrammetry. The actual feature dataset UA is a collection of feature data of a specific area or object cluster on the ground, reflecting the real state of the target cluster in the real world;

[0074] The actual feature dataset UA is the basis for evaluating the quality of surveying and mapping data, and provides direct evidence about the true state of the target cluster. The actual feature dataset UA includes at least: geometric features: such as the shape, size, position, height, etc. of the target cluster; physical properties: such as material, reflectivity, color, etc. (if the surveying and mapping equipment can obtain this information); it may also include texture information, spectral information, etc., depending on the surveying and mapping technology and equipment used.

[0075] The preset feature dataset UB is a feature dataset of a target cluster that is predefined or assumed based on known geographic information, historical data, design drawings or other reference materials before the surveying and mapping task begins; it represents the goal or expected result of the surveying and mapping task.

[0076] Summary: The preset feature dataset UB is used to compare with the actual feature dataset UA to evaluate the accuracy and completeness of the surveying and mapping data. By comparing the differences between the two datasets, errors, omissions or inconsistencies in the surveying and mapping process can be identified. At the same time, combined with the registration speed loss rate Pz and the measurement quality accuracy index Zc, the comprehensive quality of the surveying and mapping data is comprehensively evaluated to ensure the accuracy and reliability of the surveying and mapping results.

[0077] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent mapping method suitable for complex terrain, characterized in that: The following steps are involved: Step 1: Surveying and mapping data collection: Set the target complex terrain as the area to be detected, evenly set the registration base points in the area to be detected, and collect the surveying and mapping data of the area to be detected through the measuring equipment. The surveying and mapping data includes the point cloud data of the area to be detected and the multi-dimensional attribute parameters of the registration base points; Step 2: Point cloud data registration: According to the actual multi-dimensional attribute parameters of the registration base points and the measured multi-dimensional attribute parameters, the point cloud transformation matrix of the area to be detected is obtained, the point cloud data is registered through the point cloud transformation matrix, the registered point cloud data is obtained, and the clustering cluster set is obtained through the clustering algorithm; Step 3: Obtain target clusters: Evaluate the acceptable error level of each cluster, use the acceptable error level as a constraint, obtain target clusters corresponding to complex terrain, and number them; Step 4: Filter abnormal target clusters, set feature comparison reference points in each target cluster, obtain actual feature parameters of feature comparison reference points, obtain surveying and mapping feature parameters of feature comparison reference points from the registered point cloud data, compare and analyze the similarity between actual feature parameters and surveying and mapping feature parameters, and obtain measurement accuracy parameters of target clusters; filter out abnormal target clusters based on quality accuracy parameter zc; Step 5: Update the abnormal target cluster: rescan to obtain the surveying and mapping data of the abnormal target cluster, verify the measurement accuracy parameters of the surveying and mapping data; obtain a target cluster set that meets the quality requirements; register and fuse the target cluster set to obtain a three-dimensional simulation model of the target complex terrain; Step 6: Surveying and mapping data feedback: The surveying and mapping data, target cluster information and three-dimensional simulation model are fed back to the user end, and the user queries the surveying and mapping data and three-dimensional simulation model through the port.

2. The intelligent surveying and mapping method applicable to complex terrain according to claim 1, characterized in that: The point cloud transformation matrix is ​​obtained by recording the actual multidimensional attribute parameters of the registration base points as the source point set and recording the measured multidimensional attribute parameters of the registration base points as the target point set; the point cloud transformation matrix includes rotation, translation, scaling and tilt, and is obtained by minimizing the sum of the squares of the Euclidean distances between the transformed source point set and the target point set.

3. The intelligent mapping method for complex terrain according to claim 1, characterized in that: According to the area, contour perimeter and error acceptability of the target cluster, the number of feature comparison reference points is allocated to each target cluster; the number of feature comparison reference points is obtained by setting the number of feature comparison reference points based on the geometric complexity and texture complexity of the target cluster, and the number of feature comparison reference points can be dynamically adjusted according to the terrain complexity, and the formula n is used to obtain the number of feature comparison reference points. d =Δd+INT[(Jk qj +wk qw )*e 1+wc ] Calculate the number of feature comparison reference points n d ; Wherein, Δd represents the minimum number of reference points for feature comparison, wk represents texture complexity, Jk represents geometric complexity, qj represents the influence index of geometric complexity, qw represents the influence index of texture complexity, and wc represents the acceptable degree of error; INT(·) represents the rounding function.

4. The intelligent mapping method for complex terrain according to claim 3, characterized in that: The geometric complexity Jk is obtained by: obtaining the area of ​​the target cluster as s0 and the perimeter of the contour as s1, and using the formula The geometric complexity is calculated; the texture complexity wk is obtained by analyzing the remote sensing image corresponding to the target cluster and obtaining the texture features based on the gray level co-occurrence matrix; Texture complexity is calculated by taking a weighted sum of texture features.

5. The intelligent mapping method for complex terrain according to claim 4, characterized in that: The mass accuracy parameter of the target cluster is obtained as follows: Assume that each target cluster contains n d feature comparison reference point, i represents the number of the feature comparison reference point; Compare and analyze the similarity between the actual feature parameters and the surveyed feature parameters, and calculate the position similarity, elevation similarity, texture similarity and terrain feature cross entropy loss value; the position similarity, elevation similarity, texture similarity and terrain feature cross entropy loss value of the i-th feature comparison reference point are recorded as x1i, x2i, x3i, x4i respectively; By formula The quality accuracy parameter zc_j of the j-th target cluster is calculated, where w1, w2, w3, and w4 represent weight coefficients of each item, and w1+w2+w3+w4=1.

6. The intelligent mapping method for complex terrain according to claim 1, characterized in that: The screening method for abnormal target clusters is: Determine the quality accuracy parameter zc and threshold Y zc If zc≥Y zc , the data quality of the target cluster is judged to be qualified if zc<Y zc , the data quality of the target cluster is judged to be unqualified and marked as an abnormal target cluster.

7. The intelligent mapping method for complex terrain according to claim 5, characterized in that: The method for obtaining the position similarity is: calculate the Euclidean distance between two points as dxy, and use the formula The position similarity is calculated; the elevation similarity is obtained by using the inverse of the square of the difference between the two elevation values ​​to represent the elevation similarity; if the elevation values ​​of the two points are very close, the elevation similarity is high, and the formula The elevation similarity is calculated, where z1 and z2 are the elevation values ​​of the two points respectively, and ∈1 is a very small positive number used to prevent the denominator from being zero; the texture similarity is obtained by using the similarity of the image histogram to approximate the texture similarity; the terrain feature cross entropy loss value is obtained by encoding the terrain feature into a probability distribution and using the cross entropy loss to evaluate the difference between the actual distribution and the predicted distribution.

8. The intelligent surveying and mapping method applicable to complex terrain according to claim 5, characterized in that: The intelligent surveying and mapping method applicable to complex terrain also includes: a comprehensive management step of surveying and mapping data quality, which ensures the comprehensive quality of surveying and mapping data by verifying the surveying and mapping data quality coefficient.

9. The intelligent mapping method for complex terrain according to claim 8, characterized in that: The process of obtaining the surveying and mapping data quality coefficient comprises the following steps: Step S11, obtaining the actual registration speed v of the point cloud data ac and preset registration speed v pr ; Obtain the quality accuracy parameters of each target cluster, the theoretical feature data set of the target cluster, the actual feature data set UA of the target cluster and the preset feature data set UB; Step S12: by formula The registration velocity loss rate Pz of the three-dimensional simulation model is calculated; Step S13: By formula The measurement quality accuracy index Zc of the three-dimensional simulation model is calculated; where k_j represents the weight coefficient of the j-th target cluster, and the weight coefficient is obtained by obtaining the area of ​​each target cluster and recording the area of ​​the j-th target cluster as ms_j; through the formula Calculate the weight coefficient; Step S14, using the ratio of the intersection and union of the actual feature data set UA and the preset feature data set UB to represent the integrity parameter Wz of the three-dimensional simulation model of the target surveying and mapping area; Step S15, jointly analyzing the registration speed loss rate, measurement quality accuracy index Zc and three-dimensional simulation model integrity parameter Wz of the three-dimensional simulation model; by formula The joint analysis yields the mapping data quality coefficient CZ; Step S16: Determine the mapping data quality coefficient and threshold C 阈 relationship; if CZ ≥ C 阈 , indicating that the quality requirements of surveying and mapping data for complex terrain meet the requirements; if CZ<C 阈 , indicating that the quality coefficient of the surveying and mapping data does not meet the requirements, and warns the user to maintain the measuring equipment and re-collect the surveying and mapping data.

Images