Spatial data local dense anomaly analysis method, device and storage medium

Abstract

The present disclosure provides a method, device and storage medium for analyzing local dense anomalies of spatial data, the method comprising: obtaining the coordinates and observation values ​​of each point in the spatial data; calculating the self-sparseness, mutual sparsity and small-distance variability of each point according to the coordinates and observation values ​​of each point; establishing a three-parameter comprehensive intersection diagram according to the self-sparseness, mutual sparsity and small-distance variability of each point; identifying whether there are local dense anomalies in the spatial data and performing corresponding processing according to the distribution of scattered points and the size of small-distance variability in the three-parameter comprehensive intersection diagram. The present application effectively identifies local dense anomalies in spatial data in the form of a three-parameter comprehensive intersection diagram through the three local dense anomaly judgment parameters of self-sparseness, mutual sparsity and small-distance variability, and performs redundant deletion processing according to data quality on this basis, which can improve the accuracy of spatial interpolation results.

Description

Technical Field

[0001] Embodiments of the present invention relate to the technical field of data processing, and in particular to a method, device, equipment and storage medium for analyzing local dense anomalies in spatial data. Background Art

[0002] Interpolation technology is crucial in spatial analysis. It transforms unevenly distributed discrete data points into continuous and smooth surfaces through mathematical models and algorithms, providing a powerful technology for understanding and explaining complex spatial phenomena. It fills the gaps in data and makes numerical estimates of unknown areas through interpolation technology, supporting prediction, simulation and decision making in spatial analysis. There are many interpolation methods, and commonly used interpolation methods include inverse distance weighted method, spline function interpolation method, radial basis function interpolation method, natural neighbor interpolation method and Kriging interpolation method. Different interpolation methods will produce different results when processing the same data, so it is necessary to select the appropriate interpolation method according to the characteristics of the data and the application scenario.

[0003] Most interpolation methods do not consider the impact of the density distribution of data on the interpolation results. For example, the commonly used inverse distance interpolation method and isotropic kriging interpolation method, under the condition of equal distance, the weight of each data point in the interpolation is equal, that is, equal weight. From the equal weight of interpolation, it can be seen that if the known data points are too concentrated in a certain place in space, the main weight of interpolation comes from the value at that place, and the weights of other places will become very low, forming a local dense anomaly of the data points, which has an adverse effect on the interpolation results. Among them, overly dense points are called dense points, points with large differences in observed values ​​in denser areas are called bad points, and dense points and bad points are collectively referred to as local dense anomaly points.

[0004] At present, in the interpolation or modeling of spatially unevenly distributed data, many commercial software do not specifically consider the impact of local dense anomalies during interpolation, and only a very small number of experts have considered the impact of some local dense anomalies. For example, some experts have proposed to calculate the orientation weight based on the azimuth angle between points or the area of ​​the Voronoi unit, and combine it with the distance weight to obtain the total weight. Some experts also use methods such as data thinning according to certain principles. Among them, the calculation of orientation weight focuses on the angle and area between points, and directly assigns orientation weights to all data; while data thinning according to certain principles is to combine certain methods for calculation, find the data-dense area, and then thin the data.

[0005] However, the above methods focus on spatial orientation weighting such as angles or areas between points, without analyzing and processing local dense anomalies, and without eliminating the influence of bad points in the data; or focus on finding and thinning dense data, which may result in data points that are not dense points or bad points being deleted, while bad points may be retained. These methods have their own shortcomings and limitations, and there is no ideal method to identify and process local dense anomalies so far. Summary of the invention

[0006] The purpose of the present invention is to at least provide a method, device, equipment and storage medium for analyzing local dense anomalies of spatial data, which can at least solve the technical problem of being unable to quickly and intuitively identify local dense anomalies of spatial data, and at least can achieve the goal of quickly and intuitively identifying and processing local dense anomalies of spatial data.

[0007] To solve the above technical problems, at least one embodiment of the present application provides a method for analyzing local dense anomalies of spatial data, including:

[0008] Get the coordinates and observation values ​​of each point in the spatial data;

[0009] According to the coordinates and observations of each point, the self-sparseness, mutual-sparseness and small-distance variability of each point are calculated. The self-sparseness is the ratio of the distance D1 from point Z0 to its nearest point to the average distance D2 from point Z0 to the surrounding interpolation neighborhood points. The mutual-sparseness is the ratio of the distance D1 from point Z0 to its nearest point to the sum of the distances D3 from all points to their respective nearest points. The small-distance variability is the ratio of the difference D4 of the observation value from point Z0 to its nearest point to the distance D1 from point Z0 to its nearest point.

[0010] Establishing a three-parameter comprehensive intersection graph according to the self-sparseness, mutual-sparseness and small-distance variability of each point;

[0011] According to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variability, it is identified whether the spatial data has local dense anomalies.

[0012] At least one embodiment of the present application further provides a spatial data local dense anomaly analysis device, comprising:

[0013] The acquisition module is used to obtain the coordinates and observation values ​​of each point in the spatial data;

[0014] A calculation module is used to calculate the self-sparseness, mutual sparsity and small distance variability of each point according to the coordinates and observation values ​​of each point, wherein the self-sparseness is the ratio of the distance D1 from point Z0 to its nearest point to the average distance D2 from point Z0 to the surrounding interpolation neighborhood points, the mutual sparsity is the ratio of the distance D1 from point Z0 to its nearest point to the sum D3 of the distances of all points to their respective nearest points, and the small distance variability is the ratio of the difference D4 of the observation value from point Z0 to its nearest point to the distance D1 from point Z0 to its nearest point;

[0015] A three-parameter comprehensive intersection diagram creation module, used for establishing a three-parameter comprehensive intersection diagram according to the self-sparseness, mutual sparsity and small-distance variability of each point;

[0016] The identification module is used to identify whether there is a local dense anomaly in the spatial data according to the distribution of scattered points and the size of the small-distance variability in the three-parameter comprehensive intersection diagram.

[0017] At least one embodiment of the present application also provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned local dense anomaly analysis method for spatial data.

[0018] At least one embodiment of the present application further provides a computer-readable storage medium storing a computer program, wherein the computer program implements the above-mentioned spatial data local dense anomaly analysis method when executed by a processor.

[0019] The embodiment of the present application provides a method for analyzing local density anomalies in spatial data, which quantitatively describes the density of data by calculating three parameters: self-sparseness, mutual sparsity and small-distance variability, and intuitively displays the density of data by drawing a three-parameter comprehensive intersection diagram, so as to identify and process whether there are local density anomalies in spatial data, which is fast and intuitive.

[0020] In some optional embodiments, the step of drawing a three-parameter comprehensive cross-plot according to the self-sparseness, mutual-sparseness and small-distance variation of each point includes:

[0021] The logarithm of the self-sparseness of each point is taken as the horizontal axis of the coordinate, the logarithm of the mutual sparsity corresponding to each point is taken as the vertical axis of the coordinate, and the small-distance variation corresponding to each point is taken as the attribute feature to establish a three-parameter comprehensive intersection diagram.

[0022] In this embodiment, by taking the logarithm of the self-sparseness and the logarithm of the mutual sparsity to establish a three-parameter comprehensive intersection diagram, the scope of data display can be effectively expanded, which facilitates the subsequent identification and processing of local dense abnormal data on the three-parameter comprehensive intersection diagram.

[0023] In some optional embodiments, the step of identifying whether there is a local dense anomaly in the spatial data according to the distribution of scattered points and the size of the small-distance variability in the three-parameter comprehensive intersection diagram includes:

[0024] Identify whether there is a gap in the scattered points or a fitting trend of two straight lines in the scattered points in the three-parameter comprehensive intersection diagram, or whether the small-distance variation exceeds a preset threshold;

[0025] When there is no gap in the scattered points and no fitting trend of two straight lines in the scattered points in the three-parameter comprehensive intersection diagram, and the small-distance variability does not exceed the preset threshold, it is determined that there is no local dense anomaly in the spatial data;

[0026] When there are scatter point gaps in the three-parameter comprehensive intersection diagram or there is a fitting trend of two straight lines in the scatter points or the small-distance variability exceeds a preset threshold, it is determined that there is a local dense anomaly in the spatial data.

[0027] In this embodiment, by using whether there are gaps in the scattered points in the three-parameter comprehensive intersection diagram, whether there is a fitting trend of two straight lines in the scattered points, that is, whether there are two fitting straight lines with different slopes, and whether the small-distance variability exceeds a preset threshold as the judgment for identifying whether there are local dense anomalies in the spatial data, the identification of local dense anomalies can be achieved intuitively and quickly.

[0028] In some optional embodiments, the method further includes:

[0029] When it is determined that there are local dense anomalies in the spatial data, the local dense anomaly points or local dense anomaly point groups are determined according to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variation.

[0030] In this embodiment, when it is determined that there are local dense anomalies in the spatial data, it is also necessary to determine the local dense anomaly points or the local dense anomaly point groups, so as to facilitate the subsequent processing operations on the local dense anomaly points or the local dense anomaly point groups to ensure the uniformity of the distribution of the spatial data.

[0031] In some optional embodiments, when it is determined that there are local dense anomalies in the spatial data, the step of determining the local dense anomaly points or local dense anomaly point groups according to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variation includes:

[0032] When there is a scattered point gap in the three-parameter comprehensive cross-plot, the data relatively close to the origin of the three-parameter comprehensive cross-plot is determined as a local dense abnormal point group based on the gap;

[0033] When there are two straight line fitting trends with different slopes for scattered points in the three-parameter comprehensive cross-plot, the data relatively close to the origin of the three-parameter comprehensive cross-plot is determined as a local dense abnormal point group based on the fitting trends of the scattered points with different slopes as the boundary;

[0034] When there is a small-distance variability exceeding a preset threshold in the three-parameter intersection diagram, the point where the small-distance variability exceeds the preset threshold is determined as a local dense outlier point.

[0035] In this embodiment, the location of the local dense outlier point or the local dense outlier point group is further determined according to the distribution of the scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variability.

[0036] In some optional embodiments, the step of determining a local dense outlier point or a local dense outlier point group further includes:

[0037] When a local dense outlier point that exists alone is determined, the local dense outlier point is deleted;

[0038] When a local dense outlier point group is determined, a redundant deletion process is performed on the local dense outlier point group.

[0039] In this embodiment, after the local dense outlier points or the local dense outlier point groups are determined, the local dense outlier points or the local dense outlier point groups need to be processed according to the specific situation to ensure the uniformity of the spatial data distribution, so as to facilitate the subsequent interpolation operation or modeling operation based on the spatial data.

[0040] In some optional embodiments, when determining a local dense outlier point group, the step of performing redundancy removal processing on the local dense outlier point group includes:

[0041] For each of the local dense abnormal point groups, the spatial data quality of the observation value of each point is judged to determine the quality judgment result of each point;

[0042] The points with the best quality judgment results in each of the local dense outlier point groups are retained. In this embodiment, each local dense outlier point group is subjected to redundancy deletion processing, that is, the points in each local dense outlier point group are subjected to spatial data quality judgment, and only the points with the best quality judgment results in each local dense outlier point group are retained, so as to realize the processing of the local dense outlier point group and ensure the uniformity of spatial data distribution. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] One or more embodiments are exemplarily described by the pictures in the corresponding drawings, and these exemplary descriptions do not constitute limitations on the embodiments.

[0044] Figure 1 A schematic diagram of a flow chart of a method for analyzing local dense anomalies of spatial data provided by an embodiment of the present application;

[0045] Figure 2 A schematic flow chart of a method for analyzing and processing local dense anomalies of spatial data provided by another embodiment of the present application;

[0046] Figure 3 A flowchart of a method for analyzing and processing local dense anomalies of spatial data provided by another embodiment of the present application;

[0047] Figure 4 A two-dimensional uniform grid model and a point map of its deformation provided by another embodiment of the present application;

[0048] Figure 5 A relative error scatter plot obtained by performing Kriging interpolation on a two-dimensional uniform grid model and its deformation provided in another embodiment of the present application and calculating the result;

[0049] Figure 6 A three-parameter comprehensive cross-plot of a two-dimensional uniform grid model and its deformation provided by another embodiment of the present application;

[0050] Figure 7 A three-parameter comprehensive cross-plot of a two-dimensional uniform grid model and its deformation after adding Gaussian random noise provided by another embodiment of the present application;

[0051] Figure 8 A point diagram, a three-parameter comprehensive intersection diagram, and a relative error scatter diagram of a random model with Gaussian random noise added provided in another embodiment of the present application;

[0052] Fig. 9 A point map, a three-parameter comprehensive intersection map and a relative error scatter plot after processing a local dense anomaly provided in another embodiment of the present application;

[0053] Fig.10 A point diagram, a three-parameter comprehensive intersection diagram, and a relative error scatter diagram after adding 100 points to a random model containing Gaussian random errors provided in another embodiment of the present application;

[0054] Fig.11 A point map and a three-parameter comprehensive intersection map of original measured data provided by another embodiment of the present application, and a point map and a three-parameter comprehensive intersection map after processing local dense anomalies;

[0055] Fig.12An interpolation result map and a relative error map before and after processing a local dense anomaly provided by another embodiment of the present application;

[0056] Fig.13 A schematic diagram of a spatial data local dense anomaly analysis device provided by another embodiment of the present application;

[0057] Fig.14 A schematic structural diagram of an electronic device provided for another embodiment of the present application. DETAILED DESCRIPTION

[0058] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the embodiments of the present application will be described in detail below in conjunction with the accompanying drawings. However, it will be appreciated by those skilled in the art that in the present application, many technical details are proposed in order to enable the reader to better understand the present application. However, even without these technical details and various changes and modifications based on the following embodiments, the technical scheme claimed in the present application can also be implemented. The division of the following embodiments is for the convenience of description, and the specific implementation of the present application should not be construed as any limitation, and the various embodiments can be combined and referenced with each other under the premise of no contradiction.

[0059] The present invention proposes a method for analyzing and processing local dense anomalies of spatial data. The implementation details of the method for analyzing and processing local dense anomalies of spatial data of this embodiment are specifically described below. The following content is only the implementation details provided for the convenience of understanding and is not necessary for implementing this solution.

[0060] Embodiment 1

[0061] The spatial data local dense anomaly analysis method of this embodiment can be applied to electronic devices with communication, computing and data storage capabilities. The specific process can be as follows: Figure 1 As shown, including:

[0062] Step 101, obtaining the coordinates and observation values ​​of each point in the spatial data.

[0063] Specifically, a point observation refers to a value actually measured or collected on-site at a specific point (a specific location represented in space or a data set), which usually represents a real measurement result at a certain time and place.

[0064] In some examples, the coordinates and observation values ​​of each point are obtained from an existing spatial database.

[0065] Step 102, according to the coordinates and observation values ​​of each point, calculate the self-sparseness, mutual-sparseness and small-distance variability of each point, wherein the point for which the self-sparseness, mutual-sparseness and small-distance variability are calculated is point , the self-sparseness is the point The distance D1 to its nearest point is The ratio of the average distance D2 to the surrounding interpolation neighborhood points, the mutual sparsity is the ratio of the distance ... The ratio of the distance D1 to its nearest point to the sum of the distances D3 of all points to their respective nearest points. The small distance variance is the ratio of the distance D1 to the nearest point. The difference D4 between the observation value to the nearest point and the point The ratio of the distance D1 to its nearest point.

[0066] Specifically, the distance between points can be determined according to the coordinates of each point, and the self-sparseness, mutual sparsity and small-distance variability of each point can be calculated according to the coordinates and observation values ​​of each point.

[0067] Among them, self-sparseness is used to study the local sparsity of a single point in a data set (i.e., spatial data). This parameter is the ratio of the minimum distance from a single point to nearby points (i.e., the distance from a single point to its nearest point) to the average distance between the single point and the surrounding interpolation neighborhood points. The smaller the self-sparseness, the denser the point.

[0068] Specifically, the formula for self-sparseness is:

[0069]

[0070] in, is the self-sparseness, (i.e. D1) is point The distance to its nearest point, (i.e. D2) is point The average distance to the surrounding interpolation neighborhood points, For point The number of points in the surrounding interpolation neighborhood.

[0071] Mutual sparsity is used to study the global sparsity of a single point in a data set (i.e., spatial data). This parameter is the ratio of the minimum distance between a single point and its nearby points (i.e., the distance from a single point to its nearest point) to the sum of the minimum distances from all points to its nearby points (i.e., the sum of the distances from all points to their respective nearest points). The smaller the mutual sparsity value, the denser the point.

[0072] Specifically, the formula for mutual sparsity is:

[0073]

[0074] in, is the mutual sparsity, For point The number of all surrounding points, (i.e. D1) is point The distance to its nearest point, (ie D3) is the sum of the minimum distances between all points and their surrounding points.

[0075] Small-distance variability is used to find bad points in local dense anomalies. This parameter is the ratio of the difference between the observed value of a single point and its nearest point and the distance between the single point and its nearest point; when the small-distance variability has an abnormally high value, it is considered that the point is likely to be a bad point and needs to be screened.

[0076] Specifically, the formula for small-distance variability is:

[0077]

[0078] in, is the small-distance variability, (i.e. D1) is point The distance to its nearest point, For point The difference from the observed value of its nearest point.

[0079] Step 103, establishing a three-parameter comprehensive intersection diagram in the form of a scatter diagram according to the self-sparseness, mutual sparsity and small-distance variation of each point.

[0080] Specifically, the three-parameter comprehensive cross plot, also known as the XY plot, is a distribution diagram of data points on a rectangular coordinate plane, which shows the general trend of the dependent variable changing with the independent variable. Among them, the three-parameter comprehensive cross plot uses two sets of data to form multiple coordinate points. By observing the distribution of data points on the graph, the correlation between variables can be inferred, and it can clearly show which data points deviate from the distribution range of most points.

[0081] In some examples, a three-parameter comprehensive intersection diagram is established based on the self-sparseness, mutual sparsity and small-distance variability of each point to intuitively display the density of the data and facilitate the identification and determination of local dense outliers.

[0082] Step 104 , identifying whether there is a local dense anomaly in the spatial data according to the distribution of scattered points and the magnitude of small-distance variability in the three-parameter comprehensive intersection diagram.

[0083] In the specific implementation, the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variability are used to identify whether there are local dense anomalies in the spatial data; further, the distribution of data points can be intuitively seen in the three-parameter comprehensive intersection diagram to view the density of the data in space; and the small-distance variability is used to find bad points in local dense anomalies. When the small-distance variability shows an abnormally high value, it is considered that the possibility of this point being a bad point is high. Therefore, by analyzing the distribution of scattered points and the size of the small-distance variability, it is possible to determine whether there are local dense anomalies in the spatial data.

[0084] In this embodiment, the density of the data is quantitatively described by calculating the three parameters of self-sparseness, mutual sparsity and small-distance variability, and the density of the data is intuitively displayed by drawing a three-parameter comprehensive intersection diagram, so as to identify and process whether there are local density anomalies in the spatial data.

[0085] Embodiment 2

[0086] On the basis of the above embodiment, the step of drawing a three-parameter comprehensive cross-plot according to the self-sparseness, mutual sparsity and small-distance variation of each point includes:

[0087] The logarithm of the self-sparseness of each point is taken as the horizontal axis of the coordinate, the logarithm of the mutual sparsity corresponding to each point is taken as the vertical axis of the coordinate, and the small-distance variation corresponding to each point is taken as the attribute feature to establish a three-parameter comprehensive intersection diagram.

[0088] Specifically, the three-parameter comprehensive intersection diagram is established with self-sparseness as the horizontal coordinate and mutual sparsity as the vertical coordinate. Further, the value after taking the logarithm of self-sparseness is used as the horizontal axis of the coordinate, the value after taking the logarithm of mutual sparsity is used as the vertical axis of the coordinate, and the small-distance variability corresponding to each point is used as the attribute feature to establish the three-parameter comprehensive intersection diagram. Among them, the small-distance variability can be used as the color, size or other visual attributes of the point to be submitted in the three-parameter comprehensive intersection diagram.

[0089] In this embodiment, the three-parameter comprehensive intersection diagram is drawn to intuitively display the density of the data, so as to identify and process the local density anomaly of the spatial data.

[0090] Embodiment 3

[0091] On the basis of the above embodiment, the step of detecting whether there is a local dense anomaly in the spatial data according to the distribution of scattered points and the size of the small-distance variability in the three-parameter comprehensive intersection diagram includes:

[0092] Identify whether there is a gap in the scattered points or a fitting trend of two straight lines in the scattered points in the three-parameter comprehensive intersection diagram, or whether the small-distance variation exceeds a preset threshold;

[0093] When there is no gap in the scattered points and no fitting trend of two straight lines in the scattered points in the three-parameter comprehensive intersection diagram, and the small-distance variability does not exceed the preset threshold, it is determined that there is no local dense anomaly in the spatial data;

[0094] When there are scatter point gaps in the three-parameter comprehensive intersection diagram or there is a fitting trend of two straight lines in the scatter points or the small-distance variability exceeds a preset threshold, it is determined that there is a local dense anomaly in the spatial data.

[0095] In an example, based on the distribution of scattered points and the size of small-distance variability in the three-parameter comprehensive intersection diagram, identifying whether there are local dense anomalies in spatial data means observing in the three-parameter comprehensive intersection diagram whether there are obvious disconnections (i.e. gaps) in the scattered points or a fitting trend of two straight lines (i.e., the appearance of two straight lines with different slopes) in the scattered points, and using this as a dividing criterion to identify whether there are local dense anomalies in the spatial data.

[0096] In this embodiment, it specifically includes identifying whether there is a scatter point gap or a fitting trend of two straight lines of scatter points in the three-parameter comprehensive intersection diagram, or whether the small-distance variability exceeds a preset threshold;

[0097] When there are no gaps in the scattered points and no fitting trend of two straight lines in the three-parameter comprehensive intersection diagram, and the small-distance variability does not exceed the preset threshold, it is determined that there is no local dense anomaly in the spatial data; that is, when there are no gaps and no obvious fitting trend of two straight lines in the three-parameter comprehensive intersection diagram, and the small-distance variability does not have an abnormally high value, it can be determined that there is no local dense anomaly in the spatial data when all the above conditions are met.

[0098] When there are gaps in the scattered points in the three-parameter comprehensive intersection diagram, or there is a fitting trend of two straight lines in the scattered points, or the small-distance variability exceeds a preset threshold, it is determined that there is an anomaly in the spatial data; that is, when there are gaps or obvious fitting trends of two lines in the three-parameter comprehensive intersection diagram, or there is an abnormally high value in the small-distance variability, if any of the above conditions is met, it can be determined that there is a local dense anomaly in the spatial data.

[0099] Embodiment 4

[0100] Based on the above embodiment, the method further includes:

[0101] When it is determined that there are local dense anomalies in the spatial data, the local dense anomaly points or local dense anomaly point groups are determined according to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variation.

[0102] Specifically, when local dense anomalies are identified in spatial data, local dense anomaly points or local dense anomaly point groups need to be determined. Furthermore, local dense anomaly points or local dense anomaly point groups are determined based on the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variability.

[0103] In this embodiment, the three-parameter comprehensive intersection diagram can not only intuitively display the density of the data, but also determine the local dense abnormal point group in the three-parameter comprehensive intersection diagram, which is convenient for subsequent processing of the local dense abnormal point group and ensures the accuracy of the spatial data.

[0104] Embodiment 5

[0105] On the basis of the above embodiment, when it is determined that the spatial data has local dense anomalies, the step of determining the local dense anomaly points or local dense anomaly point groups according to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variation includes:

[0106] When there is a scattered point gap in the three-parameter comprehensive cross-plot, determine that the data relatively close to the origin of the three-parameter comprehensive cross-plot with the gap as the boundary is a local dense abnormal point group;

[0107] When there are two straight line fitting trends with different slopes for scattered points in the three-parameter comprehensive cross-plot, the data relatively close to the origin of the three-parameter comprehensive cross-plot is determined as a local dense abnormal point group based on the fitting trends of the scattered points with different slopes as the boundary;

[0108] When there is a small-distance variability exceeding a preset threshold value in the three-parameter comprehensive intersection map, the point where the small-distance variability exceeds the preset threshold value is determined as a local dense abnormal point.

[0109] Specifically, the steps of determining local dense outliers include:

[0110] When there are scattered point gaps in the three-parameter comprehensive intersection diagram, the gap is used as the boundary to determine the local dense abnormal point group, that is, from the formulas of self-sparseness and mutual sparsity, it can be seen that the self-sparseness and mutual sparsity of the local dense abnormal point are both smaller than the normal values. Therefore, when there are obvious gaps in the upper and lower parts of the scattered points of the three-parameter comprehensive intersection diagram, it can be determined that the self-sparseness and mutual sparsity smaller than the gap position are the local dense abnormal point group. Further, combined with the creation process of the three-parameter comprehensive intersection diagram, it can be known that when there are scattered point gaps in the three-parameter comprehensive intersection diagram, it means that the points with smaller self-sparseness and mutual sparsity than the gap position are located in the lower left area of ​​the three-parameter comprehensive intersection diagram, that is, relatively close to the origin of the three-parameter comprehensive intersection diagram.

[0111] When there is a fitting trend of two straight lines appearing at scattered points in the three-parameter comprehensive intersection diagram, the straight line relatively close to the origin of the three-parameter comprehensive intersection diagram is determined to be the area where the local dense anomaly point group is located; that is, when there is a fitting trend of two straight lines appearing at scattered points in the three-parameter comprehensive intersection diagram, it means that two fitting straight lines with different slopes appear in the three-parameter comprehensive intersection diagram, and the fitting trends with different slopes are determined as the boundary. Combined with the formulas of self-sparseness and mutual sparsity, it can be seen that the self-sparseness and mutual sparsity of the local dense anomaly points are both smaller than the normal values. Therefore, the area relatively close to the origin of the three-parameter comprehensive intersection diagram and bounded by the fitting trends with different slopes shown in the three-parameter comprehensive intersection diagram is the area where the local dense anomaly point group is located.

[0112] When there is a small-distance variability exceeding a preset threshold in the three-parameter comprehensive intersection map, the point whose small-distance variability exceeds the preset threshold is determined to be a local dense outlier point; that is, according to the definition of small-distance variability, when the small-distance variability has an abnormally high value, the possibility that this point is a bad point is extremely high. Therefore, the point whose small-distance variability exceeds the preset threshold can be determined to be a local dense outlier point.

[0113] The value of the preset threshold can be adaptively adjusted according to the distribution of data points of the spatial data and the observed value of each point.

[0114] Embodiment 6

[0115] Based on the above embodiment, the step of determining local dense outlier points or local dense outlier point groups further includes:

[0116] When a local dense outlier point that exists alone is determined, the local dense outlier point is deleted;

[0117] When a local dense outlier point group is determined, a redundancy deletion process is performed on the local dense outlier point group.

[0118] Specifically, after determining the local dense outlier points or the local dense outlier point groups, it is necessary to delete the local dense outlier points or perform redundant deletion processing on the local dense outlier point groups to ensure the uniformity of the spatial distribution of the spatial data and facilitate the subsequent interpolation or modeling operations of the spatial data.

[0119] Embodiment 7

[0120] On the basis of the above embodiment, when determining a local dense outlier point group, the step of performing redundancy removal processing on the local dense outlier point group includes:

[0121] For each of the local dense abnormal point groups, the spatial data quality of the observation value of each point is judged to determine the quality judgment result of each point;

[0122] The point with the best quality determination result in each of the local dense abnormal point groups is retained.

[0123] In one example, determining the spatial data quality of each local dense abnormal point group includes interpolating the observed values ​​of the points of the spatial data using a Kriging interpolation method to obtain a cross-validation result of each point in the spatial data;

[0124] The cross validation results of each point on the local dense outlier point group are compared, the point with the smallest absolute error in the local dense outlier point group is retained, and the remaining points are deleted.

[0125] Specifically, Kriging interpolation is performed on the observed values ​​of the points of the spatial data, and cross-validation results of each point in the spatial data are obtained; points with large absolute errors in each local dense abnormal point group are eliminated, and only points with the smallest absolute errors in the local dense abnormal point group are retained as valid points.

[0126] In one example, the entire spatial data is interpolated by the Kriging interpolation method, and the interpolation result is compared with the points of the local dense anomaly point group. The local dense anomaly point group is processed according to the comparison result, thereby eliminating the influence of the local dense anomaly on the interpolation and improving the accuracy of the interpolation result.

[0127] In one example, the judgment of the spatial data quality of each local dense abnormal point group includes quality assessment based on the observation value of each point, or judgment through other interpolation methods or quality standards, which are not listed in this article.

[0128] Embodiment 8

[0129] The spatial data local dense anomaly analysis and processing method of this embodiment can be applied to electronic devices with communication, computing and data storage capabilities. The specific process can be as follows: Figures 2 to 3 As shown, including:

[0130] S1: Coordinates and observations in the input dataset.

[0131] Specifically, the coordinates and observation values ​​of the points in the data set are obtained. The observation value of a point refers to the value actually measured or collected on-site at a specific point (a specific location represented in space or in a data set), which usually represents the actual measurement result at a certain time and place.

[0132] S2: Calculate the self-sparseness, mutual sparsity and small-distance variability, and use the self-sparseness after taking the logarithm, the mutual sparsity after taking the logarithm and the small-distance variability as the horizontal axis, vertical axis and attribute features respectively, and draw a three-parameter comprehensive intersection diagram.

[0133] Specifically, the distance between points is determined by the coordinates of the points in the data set; based on the analysis of the distance between points in the data set and the observed values, three parameters, namely, self-sparseness, mutual sparsity, and small-distance variability, are calculated.

[0134] Among them, Self-Sparsity is used to study the local sparsity of a single point in a data set. This parameter is the ratio of the minimum distance between a single point and nearby points to the average distance between the point and the surrounding interpolation neighborhood points. The smaller the value, the denser the point.

[0135] Specifically, the calculation formula of self-sparseness is:

[0136]

[0137] in, is the self-sparseness, For point The distance to its nearest point, For point The average distance to the surrounding interpolation neighborhood points, For point The number of points in the surrounding interpolation neighborhood.

[0138] Mutual Sparsity is used to study the global sparsity of a single point in a data set (i.e., spatial data). This parameter is the ratio of the minimum distance between a single point and its nearby points to the sum of the minimum distances between all points and their nearby points. The smaller the value, the denser the point.

[0139] Specifically, the formula for mutual sparsity is:

[0140]

[0141] in, is the mutual sparsity, For point The number of all surrounding points, For point The distance to its nearest point, is the sum of the minimum distances between all points and their surrounding points.

[0142] Small-Distance Variability is used to find bad points in local dense anomalies. This parameter is the ratio of the difference between the observed value of a single point and its nearest point and the distance between the single point and its nearest point. When an abnormally high value appears, it is considered that the point is likely to be a bad point and needs to be screened.

[0143] Specifically, the formula for small-distance variability is:

[0144]

[0145] in, is the small-distance variability, For point The distance to its nearest point, For point The difference from the observed value of its nearest point.

[0146] After determining the self-sparseness, mutual sparsity and small-distance variability, the logarithm of the self-sparseness is taken as the horizontal axis, the logarithm of the mutual sparsity is taken as the vertical axis, and the small-distance variability is used as the observation value to draw a three-parameter comprehensive intersection diagram; the three-parameter intersection diagram can intuitively display the density of the data, so as to identify and process the local dense anomalies of the data set.

[0147] S3. According to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of small-distance variability, combined with the point map, determine whether there is a local dense anomaly.

[0148] According to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variability, combined with the point map, it is judged whether there is a local dense anomaly, including: looking for obvious disconnection (gaps) in the scattered points in the three-parameter comprehensive intersection diagram or the fitting trend of two straight lines (different slopes) in the scattered points, as well as points with abnormally high small-distance variability.

[0149] When there is no gap or obvious fitting trend of two straight lines in the three-parameter comprehensive intersection diagram, and there is no abnormally high value of small-distance variability, and all conditions are met, it is judged that there is no local dense anomaly in the data set, that is, the exit condition is met. If the exit condition is met, the process is exited.

[0150] When there is an obvious disconnection (gap) in the scattered points in the three-parameter comprehensive intersection diagram, or there is a fitting trend of two straight lines (different slopes) in the scattered points, or there are points with abnormally high values ​​of small-distance variability, if any of the above conditions occurs, it is judged that there is a local dense anomaly in the data set; and when judging that there is a local dense anomaly in the data set, according to the distribution of scattered points in the three-parameter comprehensive intersection diagram, the points with small self-sparseness and mutual sparsity are screened out, and at the same time, the points with abnormally high small-distance variability are found according to the scattered point attribute values ​​of the three-parameter comprehensive intersection diagram, and compared with the point map for analysis, the local dense abnormal point group is found.

[0151] S4. After finding the local dense abnormal point group, perform redundant deletion processing on the local dense abnormal point group so that only one valid point is retained in each point group and the remaining points are deleted.

[0152] Specifically, for each of the local dense abnormal point groups, the spatial data quality of the observation value of each point is determined to determine the quality determination result of each point;

[0153] The point with the best quality determination result in each of the local dense abnormal point groups is retained.

[0154] In one example, determining the spatial data quality of each local dense abnormal point group includes interpolating the observed values ​​of the points of the spatial data using a Kriging interpolation method to obtain a cross-validation result of each point in the spatial data;

[0155] The cross validation results of each point on the local dense outlier point group are compared, the point with the smallest absolute error in the local dense outlier point group is retained, and the remaining points are deleted.

[0156] In this embodiment, interpolation is performed by the Kriging interpolation method to obtain cross-validation results, and the absolute errors in each dense anomaly point group are compared. In each point group of local dense anomalies, all points with large errors are deleted, and only points with the smallest errors are retained as valid points to update the data.

[0157] S5. Calculate the self-sparseness, mutual sparsity, and small-distance variability of the deleted point set again, and draw a three-parameter comprehensive intersection diagram. Use the new three-parameter comprehensive intersection diagram to verify whether the local dense abnormal point group has been completely processed.

[0158] Specifically, the above steps 2 to 4 are repeated until the exit condition is finally met and then the process is exited, that is, the obtained data is data without local dense anomalies.

[0159] In this embodiment, the density of the data is quantitatively described by calculating the three parameters of self-sparseness, mutual sparsity and small-distance variability, and the density of the data is intuitively displayed by drawing a three-parameter comprehensive intersection diagram to identify whether there are anomalies in the spatial data; in addition, local dense abnormal point groups can be found in the three-parameter comprehensive intersection diagram, and the local dense abnormal point groups are reasonably processed in combination with the Kriging interpolation method to improve the data accuracy of the spatial data and facilitate subsequent interpolation operations or modeling operations.

[0160] Among them, see Figures 4 to 11 As shown, this embodiment combines the theoretical model with the measured data to analyze the implementation method of self-sparseness, mutual sparsity and small-distance variability in finding local dense anomalies and the comparative effect of having or not having local dense anomalies.

[0161] The theoretical model mainly explains the impact of local dense anomalies on interpolation accuracy and verifies whether it can find and process local dense anomaly points.

[0162] First, the first set of theoretical models is constructed. A two-dimensional uniform grid theoretical model is constructed, with a latitude and longitude range of [-5°, 5°], an interval of 1°, and a total of 121 grid points. Since this embodiment focuses on the analysis and processing of local dense anomalies, only a simple isotropic model is established, and the Kriging interpolation method is used as the experimental carrier for research. The calculation formula of its synthetic data is:

[0163]

[0164] in is the synthetic data value, is the longitude, is the latitude.

[0165] like Figure 4 As shown in the figure, a model with local dense anomalies is formed by moving scattered points in the two-dimensional uniform grid theoretical model to form dense points or by directly adding dense points. Figure 4 Middle (A) is a two-dimensional uniform grid model; Figure 4 Middle (B) is the formation of local dense anomaly by moving 3 points; Figure 4 In the middle (C), a local dense anomaly is formed by adding 100 points. Kriging interpolation processing with the same parameters is performed on the model containing local dense anomalies and the uniform grid model. The longitude and latitude range of the interpolation grid is selected as [-5°, 5°], and the longitude and latitude spacing of the interpolation grid is 0.4°. A total of 676 interpolation points are obtained. Taking the synthetic data value as the standard, the relative error of the interpolation grid can be calculated to analyze the impact of local dense anomalies on interpolation accuracy. During interpolation, the variation function of the uniform grid is selected to ensure that the change in interpolation accuracy is mainly caused by dense points, so as to achieve the effect of controlling variables. The effect of Kriging interpolation is shown as follows: Figure 5 As shown, Figure 5 (A) is the interpolation result of the two-dimensional uniform grid model; Figure 5 Middle (B) is the interpolation result of forming a local dense anomaly by moving 3 points; Figure 5 Middle (C) is the interpolation result of forming a local dense anomaly by adding 100 points.

[0166] from Figure 5 It can be seen that local dense anomalies will affect the accuracy of interpolation. By calculating the self-sparseness, mutual sparseness, and small-distance variation and plotting Figure 6 The three-parameter comprehensive cross-plot shown in Figure 1 shows the three-parameter comprehensive cross-plot, where Figure 6 Middle (A) is the three-parameter comprehensive crossplot of the two-dimensional uniform grid model; Figure 6 Middle (B) is a three-parameter comprehensive cross-plot of local dense anomalies formed by moving three points; Figure 6Middle (C) is a three-parameter comprehensive intersection diagram of local dense anomalies formed by adding 100 points. It can be seen that in the two-dimensional uniform grid theoretical model, the scatter point fitting of the three-parameter comprehensive intersection diagram is a straight line, and there is no gap or different slope phenomenon; but when the local dense anomaly model is formed by moving the points, the upper and lower parts of the scatter points of the three-parameter comprehensive intersection diagram have obvious gaps. By extracting the points with abnormally small mutual sparsity and self-sparseness for comparison and analysis, the local dense anomaly formed after the moving points is confirmed; when the local dense anomaly model is formed by adding points, the scatter point fitting of the upper and lower parts of the three-parameter comprehensive intersection diagram can find two lines with obvious different slopes. By extracting the points with small self-sparseness and mutual sparsity in the lower half for comparison and analysis, the local dense anomaly formed after adding points is confirmed. In addition, since no random noise is added to the attribute values ​​of the models in these three models, we can see that there are no points with abnormally high small-distance variability, that is, there are no small-distance variability abnormal points in these three models. The above shows that when we calculate the self-sparseness, mutual-sparseness, and small-distance variability of the data and draw a three-parameter comprehensive intersection diagram, we can clearly find out whether the data contains local dense anomalies.

[0167] In order to verify the role of small-distance variability in the theoretical model, based on the three models tested previously (the two-dimensional uniform grid model, the model of moving some points to form local dense anomalies, and the model of adding some points to form local dense anomalies), Gaussian noise is added with a standard deviation of 5% of each synthetic data value. Then the three parameters of self-sparseness, mutual sparsity and small-distance variability are calculated and the three-parameter comprehensive intersection diagram is drawn, as shown in Figure 7 As shown, Figure 7 (A) is the three-parameter comprehensive cross plot of the two-dimensional uniform grid model with noise. Figure 7 The middle (B) is a three-parameter comprehensive cross-plot containing noise and moving three points to form a local dense anomaly. Figure 7 (C) is a three-parameter comprehensive cross-plot containing noise and adding 100 points to form a local dense anomaly. Figure 7 It can be seen that after adding noise, in the three-parameter comprehensive intersection diagram, not only can we see points with abnormally small self-sparseness and mutual sparsity, but also we can find dense points based on the gap phenomenon and the different fitting slopes, and there is also an abnormally high small-distance variability phenomenon. These points are extracted and compared with the actual point map, which verifies that these abnormal points are the corresponding local dense abnormal points.

[0168] Next, the second set of theoretical models is constructed. A randomly generated theoretical model (referred to as the random model) is constructed, whose longitude and latitude are all between [-50°, 50°]. 200 data points are randomly generated according to the formula of the synthetic data value of the uniform grid mentioned above, and the addition of noise is consistent with the above uniform grid model. For the random model, we compared two ways of Kriging interpolation: the first one does not perform dense anomaly processing and directly performs Kriging interpolation; the second one performs dense anomaly processing and then performs Kriging interpolation operation. The specific operation process is to first identify the local dense anomaly point group based on the three-parameter comprehensive intersection diagram of self-sparseness, mutual sparsity and small-distance variability, and then use the cross-validation technology of Kriging interpolation to compare and analyze each dense point, delete the points with large errors, and form a data point set without local dense anomalies, and then perform Kriging interpolation operation.

[0169] The point map of the initial 200 data, the relative error scatter plot calculated after the interpolation grid is obtained by Kriging interpolation, and the corresponding three-parameter comprehensive intersection map of self-sparseness, mutual sparsity and small-distance variability are plotted, as shown in Figure 2. Figure 8 As shown, Figure 8 (A) is the point map of the random model. Figure 8 Middle (B) is a scatter plot of relative errors of each interpolation point calculated by the random model; Figure 8 (C) is a three-parameter comprehensive cross plot with a random model. The longitude and latitude range of the interpolation grid selected is [-50°, 50°], and the longitude and latitude spacing of the interpolation grid is 4°, resulting in a total of 676 interpolation points. Figure 8 From the three-parameter comprehensive intersection diagram (C), it can be seen that there is a more obvious gap phenomenon in the upper and lower parts. The points with abnormally small self-sparseness and mutual sparsity are extracted for comparison, and they are confirmed to be local dense anomalies. The local dense anomalies are identified and processed together with the small-distance variation anomalies. For each local dense anomaly point group, the errors are compared according to the cross-validation results of Kriging interpolation. Each local dense anomaly point group is processed, and only the points with the smallest error are retained, and the remaining points are deleted. After deleting the three-parameter comprehensive intersection diagram of the random model and the cross-validation results, 191 points remain. The three-parameter comprehensive intersection diagram drawn after the three parameters are calculated for the remaining points, as well as the error between the grid obtained by Kriging interpolation using the remaining points and the true value result are shown in the figure. Fig. 9 As shown, Fig. 9 (A) is the point map of the random model after processing the complete local dense anomaly. Fig. 9 (B) is the relative error scatter plot after processing the completely local dense anomaly. Fig. 9 (C) is the three-parameter comprehensive cross-plot after processing the complete local dense anomaly. Fig. 9 As can be seen in (C), the local dense anomaly has been completely removed. Fig. 9 As can be seen in (B), the interpolation accuracy is improved overall, and the accuracy is significantly improved near dense anomalies.

[0170] In order to better demonstrate the impact of local dense anomalies on data interpolation accuracy, and to highlight the importance of self-sparseness, mutual sparsity and small-distance variability in finding local dense anomalies, 100 points are added to the random model near (25°, 25°) to form local dense anomalies. After calculating the parameters, a three-parameter comprehensive intersection diagram is drawn, and Kriging interpolation is performed to obtain the relative error scatter diagram of the corresponding grid. The point map, three-parameter comprehensive intersection diagram and relative error scatter diagram are shown in Figure 1. Fig.10 It can be seen that, Fig.10 (A) is the point map after adding 100 points to a certain position inside the random model after processing the local dense anomaly. Fig.10 (B) is a scatter plot of relative errors with 100 points added. Fig.10 (C) is the three-parameter comprehensive cross-plot with the random model after adding 100 points. Fig. 9 Compared with (C), based on Fig.10 In (C), we can clearly see the newly added local dense anomalies and bad pixel data in the dataset. Figure 8 , 9 , 10. Comparison of corresponding types of graphs shows that: from the three-parameter comprehensive intersection graph, it can be seen that local dense anomalies can be found by calculating self-sparseness, mutual sparsity and small-distance variability. After processing the local dense anomaly point group, there is no local dense anomaly in the three-parameter comprehensive intersection graph. From the point map, it can be seen that the three-parameter comprehensive intersection graph drawn by calculating the three parameters can quickly find the corresponding local dense anomaly position, and the local dense anomaly no longer exists after processing. From the error comparison between the interpolation grid and the true value, it can be seen that after screening and processing the local dense anomaly point group, the interpolation accuracy is improved, and the accuracy is significantly improved near the local dense anomaly.

[0171] The results of measured data are used to verify the role of the three parameters of self-sparseness, mutual sparseness and small-distance variability in identifying local dense anomalies, and to process local dense anomalies. Fig.11 As shown in (A), there are 527 measured data distribution points, and the observed data value is a continuously changing physical quantity. First, calculate the three parameters and draw Fig.11 In the three-parameter comprehensive cross-plot shown in (B), local dense abnormal point groups are identified based on the obvious slope differences and abnormally high values ​​of small-distance variability in the three-parameter comprehensive cross-plot, and dense abnormal point groups are extracted; then the Kriging method is used for interpolation, and based on the cross-validation results of each dense point group, the point with the smallest error is retained as the valid point, and the rest are deleted (in Fig.11In the figure (C), 16 points were deleted. The three parameters were calculated again and the three-parameter comprehensive intersection diagram was drawn. Fig.11 As shown in (D), it can be found that the local dense anomaly has been processed.

[0172] Since the extrapolation results are not data-constrained and the results are unreliable, we only analyze the interpolation results at the interpolation positions in the data set. Fig.12 As shown in Figure 2, the Kriging interpolation results before and after local dense anomaly processing are compared. Fig.12 (A) is the interpolation result of the measured data. Fig.12 (B) is the interpolation result after processing local dense anomalies (the white dots in the figure represent dense anomalies). It is found that the overall interpolation result is not obvious. This is mainly because each point group in the dense anomaly point group contains only two dense anomaly points, there are fewer dense points and bad points, and the positions between point groups are also more scattered, so the impact on the interpolation result is not significant. However, combined with the relative error formula of the measured data, the relative error formula is calculated and plotted. Fig.12 In (C) The relative error diagram (black points are dense anomaly points, and gray points are measured data point distribution) shows that the interpolation results of the measured data before and after processing local dense anomalies are different in some locations, and there are local dense anomalies near their locations, which indicates that local dense anomalies have an impact on interpolation. According to the analysis of the previous synthetic data results, the accuracy of the interpolation results should be higher after processing local dense anomalies.

[0173]

[0174] in, is the relative error of the measured data, is the interpolation grid result of the original measured data, This is the interpolation grid result after processing local dense anomalies.

[0175] Therefore, self-sparseness, mutual-sparseness, small-distance variability and the three-parameter comprehensive cross-plot can effectively identify and process local dense anomalies in the measured data set. Combined with the cross-validation results of Kriging interpolation, the interpolation results can be more reliable.

[0176] Embodiment 9

[0177] Another embodiment of the present application relates to a device for analyzing local dense anomalies of spatial data. The implementation details of the device for analyzing local dense anomalies of spatial data of this embodiment are specifically described below. The following content is only for the convenience of understanding the implementation details provided, and is not necessary for implementing this solution. The schematic diagram of the device for analyzing local dense anomalies of spatial data of this embodiment can be as follows: Fig.13 As shown, it includes an acquisition module 801 , a calculation module 802 , a three-parameter comprehensive intersection diagram creation module 803 and an identification module 804 .

[0178] The acquisition module 801 is used to acquire the coordinates and observation values ​​of each point in the spatial data.

[0179] The calculation module 802 is used to calculate the self-sparseness, mutual sparseness and small-distance variation of each point according to the coordinates and observation values ​​of each point, wherein the self-sparseness is the value of the point The distance D1 to its nearest point is The ratio of the average distance D2 to the surrounding interpolation neighborhood points, the mutual sparsity is the ratio of the distance ... The ratio of the distance D1 to its nearest point to the sum of the distances D3 of all points to their respective nearest points. The small distance variance is the ratio of the distance D1 to the nearest point. The difference D4 between the observation value to the nearest point and the point The ratio of the distance D1 to its nearest point.

[0180] The three-parameter comprehensive intersection graph creation module 803 is used to establish a three-parameter comprehensive intersection graph according to the self-sparseness, mutual sparsity and small-distance variability of each point.

[0181] The identification module 804 is used to identify whether there is a local dense anomaly in the spatial data and perform processing according to the distribution of scattered points and the size of small-distance variability in the three-parameter comprehensive intersection diagram.

[0182] It is worth mentioning that all modules involved in this embodiment are logic modules. In practical applications, a logic unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, in order to highlight the innovative part of this application, this embodiment does not introduce units that are not closely related to solving the technical problems proposed by this application, but this does not mean that there are no other units in this embodiment.

[0183] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the above-mentioned module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0184] Embodiment 10

[0185] Another embodiment of the present application relates to an electronic device, such as Fig.14 As shown, it includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can implement the above method steps.

[0186] Among them, the memory and the processor are connected in a bus manner, and the bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors and memories together. The bus can also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be one element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices on a transmission medium. The data processed by the processor is transmitted on a wireless medium via an antenna, and further, the antenna also receives data and transmits the data to the processor.

[0187] The processor is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory can be used to store data used by the processor when performing operations.

[0188] The processor may include, but is not limited to, one or more processors or microprocessors, etc. Each processor may be an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a controller, a microcontroller, a microprocessor or other electronic components to execute the methods in the above embodiments.

[0189] Embodiment 11

[0190] Another embodiment of the present application relates to a computer-readable storage medium storing a computer program, which implements the above method steps when executed by a processor.

[0191] That is, those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including a number of instructions to enable a device (which can be a single-chip microcomputer, chip, etc.) or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as: ROM), random access memory (Random Access Memory, referred to as: RAM), disk or optical disk and other media that can store program codes.

[0192] Those skilled in the art will appreciate that the above embodiments are specific embodiments for implementing the present application, and in actual applications, various changes may be made thereto in form and detail without departing from the spirit and scope of the present application.

Claims

1. An interpolation preprocessing method for local dense anomaly analysis of spatial data, characterized in that: Applied to an electronic device having communication, computing and data storage capabilities, the method comprises: Obtain the coordinates and observation values ​​of each point from the existing spatial data set. The observation value of a point refers to the value actually measured or collected on-site at a specific point. According to the coordinates and observations of each point, the self-sparseness, mutual-sparseness and small-distance variability of each point are calculated, wherein the self-sparseness is the ratio of the distance D1 from point Z0 to its nearest point to the average distance D2 from point Z0 to the surrounding interpolation neighborhood points, the mutual-sparseness is the ratio of the distance D1 from point Z0 to its nearest point to the sum of the distances D3 from all points to their respective nearest points, and the small-distance variability is the ratio of the difference D4 of the observation value from point Z0 to its nearest point to the distance D1 from point Z0 to its nearest point; Establishing a three-parameter comprehensive intersection graph according to the self-sparseness, mutual-sparseness and small-distance variability of each point; According to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variability, identifying whether there is a local dense anomaly in the spatial data where the points near a certain known point are too densely distributed; When it is determined that there are local dense anomalies in the spatial data, the local dense anomaly points or local dense anomaly point groups are determined according to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variation; When it is determined that a single point with a small distance variation exceeding a preset threshold is deleted; When determining a local dense outlier point group, for each of the local dense outlier point groups, the observation values ​​of the spatial data points are interpolated to obtain cross-validation results of each point. Based on the comparison of the cross-validation results of each point, the point with the smallest absolute error in the local dense outlier point group is retained as a valid point, and the remaining points are deleted, and the interpolation operation is performed on the spatial data set after the above processing.

2. The interpolation preprocessing method for local dense anomaly analysis of spatial data according to claim 1 is characterized in that: The step of drawing a three-parameter comprehensive cross-plot according to the self-sparseness, mutual-sparseness and small-distance variation of each point comprises: The logarithm of the self-sparseness of each point is taken as the horizontal axis of the coordinate, the logarithm of the mutual sparsity corresponding to each point is taken as the vertical axis of the coordinate, and the small-distance variation corresponding to each point is taken as the attribute feature to establish a three-parameter comprehensive intersection diagram.

3. The interpolation preprocessing method for local dense anomaly analysis of spatial data according to claim 1 is characterized in that: According to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variability, the step of identifying whether the spatial data has local dense anomalies includes: Identify whether there is a gap in the scattered points or a fitting trend of two straight lines in the scattered points in the three-parameter comprehensive intersection diagram, or whether the small-distance variation exceeds a preset threshold; When there is no gap in the scattered points and no fitting trend of two straight lines in the scattered points in the three-parameter comprehensive intersection diagram, and the small-distance variability does not exceed the preset threshold, it is determined that there is no local dense anomaly in the spatial data; When there are scatter point gaps in the three-parameter comprehensive intersection diagram or there is a fitting trend of two straight lines in the scatter points or the small-distance variability exceeds a preset threshold, it is determined that there is a local dense anomaly in the spatial data.

4. The interpolation preprocessing method for local dense anomaly analysis of spatial data according to claim 1 is characterized in that: When it is determined that there are local dense anomalies in the spatial data, the step of determining the local dense anomaly points or local dense anomaly point groups according to the distribution of scattered points in the three-parameter comprehensive intersection diagram and the size of the small-distance variation includes: When there is a scattered point gap in the three-parameter comprehensive cross-plot, the data relatively close to the origin of the three-parameter comprehensive cross-plot is determined as a local dense abnormal point group based on the gap; When there are two straight line fitting trends with different slopes for scattered points in the three-parameter comprehensive cross-plot, the data relatively close to the origin of the three-parameter comprehensive cross-plot is determined as a local dense abnormal point group based on the fitting trends of the scattered points with different slopes as the boundary; When there is a small-distance variability exceeding a preset threshold value in the three-parameter comprehensive intersection map, the point where the small-distance variability exceeds the preset threshold value is determined as a local dense abnormal point.

5. An interpolation preprocessing device for local dense anomaly analysis of spatial data, characterized in that: Applicable to electronic devices with communication, computing and data storage capabilities, the device comprises: The acquisition module is used to obtain the coordinates and observation values ​​of each point from the existing spatial data set. The observation value of a point refers to the value actually measured or collected on-site at a specific point; A calculation module, for calculating the self-sparseness, mutual sparseness and small distance variation of each point according to the coordinates and observation values ​​of each point, wherein the self-sparseness is the ratio of the distance D1 from point Z0 to its nearest point to the average distance D2 from point Z0 to the surrounding interpolation neighborhood points, the mutual sparseness is the ratio of the distance D1 from point Z0 to its nearest point to the sum D3 of the distances of all points to their respective nearest points, and the small distance variation is the ratio of the difference D4 of the observation value from point Z0 to its nearest point to the distance D1 from point Z0 to its nearest point; A three-parameter comprehensive intersection diagram creation module, used for establishing a three-parameter comprehensive intersection diagram according to the self-sparseness, mutual-sparseness and small-distance variability of each point; The interpolation preprocessing module is used to identify whether there are local dense anomalies in the spatial data according to the distribution of scattered points and the size of small-distance variability in the three-parameter comprehensive intersection diagram; when it is determined that there are local dense anomalies in the spatial data, the local dense anomaly point or local dense anomaly point group is determined according to the distribution of scattered points and the size of small-distance variability in the three-parameter comprehensive intersection diagram; when it is determined that a point whose small-distance variability exceeds a preset threshold exists alone, the point is deleted; when a local dense anomaly point group is determined, for each of the local dense anomaly point groups, the observation value of the spatial data point is interpolated to obtain a cross-validation result of each point, and based on the comparison of the cross-validation results of each point, the point with the smallest absolute error in the local dense anomaly point group is retained as a valid point, and the remaining points are deleted, and the interpolation operation is performed on the spatial data set after the above processing.

6. An electronic device, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the interpolation preprocessing method for local dense anomaly analysis of spatial data as described in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the interpolation preprocessing method for local dense anomaly analysis of spatial data according to any one of claims 1 to 4 is implemented.

Images