Three-dimensional abnormal region extraction method and system based on variation direction control

Abstract

The invention relates to the field of earth science, and provides a three-dimensional abnormal region extraction method and system based on variation direction control. The method comprises the following steps: acquiring a geospatial lattice point data set, and determining a variation direction control variable and a neighborhood step variable according to research target requirements and characteristics of the geospatial lattice point data set; establishing a neighborhood grid based on the variation direction control variable and the neighborhood step variable, calculating the variance between the attribute strength of each grid point in the neighborhood grid and the attribute strength of the neighborhood grid point, quantifying the variance of the neighborhood grid point, and constructing a difference coefficient sequence; determining a difference coefficient threshold value, and marking the lattice points with the difference coefficients exceeding the difference coefficient threshold value as difference lattice points; and clustering the difference grid points to construct a three-dimensional abnormal region. According to the method, the three-dimensional characteristics and the local spatial heterogeneity of the spatial data are considered, the abnormal region in the earth system can be identified more accurately, and a reliable data basis is provided for subsequent three-dimensional structure extraction.

Description

Technical Field

[0001] The present invention relates to the field of earth science, and in particular to a three-dimensional abnormal region extraction method and system based on variation direction control. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] In the field of earth sciences, spatial anomalies are common, such as thermoclines and frontal zones with large temperature differences in the ocean. These phenomena are of great significance for understanding ocean dynamics, climate change and ecosystems.

[0004] Traditional two-dimensional analysis methods: These methods usually project three-dimensional data onto a two-dimensional plane for analysis, ignoring the vertical dimension information of the data. For example, the distribution of temperature or salinity is displayed through contour maps, heat maps, etc. Although this method is simple and intuitive, it cannot reveal the complex structure and dynamic changes in three-dimensional space.

[0005] Three-dimensional analysis methods based on fixed models: Some studies use predefined three-dimensional models to fit and analyze spatial grid data. These models may include spherical harmonics, radial basis functions, etc., which can describe the three-dimensional distribution of data. However, this method relies on fixed models or assumptions, is difficult to adapt to data sets of different scales and complexities, and may not be flexible enough when dealing with irregular anomaly areas. Summary of the invention

[0006] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a three-dimensional abnormal area extraction method and system based on variation direction control. The present invention can overcome the limitations of the prior art and accurately extract three-dimensional abnormal areas from a large amount of earth system spatial grid data according to the different formation mechanisms of geographical anomaly phenomena to meet the needs of scientific research and practical applications.

[0007] In order to achieve the above object, the present invention adopts the following technical solution:

[0008] A first aspect of the present invention provides a three-dimensional abnormal region extraction method based on variation direction control.

[0009] A three-dimensional abnormal region extraction method based on variation direction control, comprising:

[0010] Obtain a geospatial grid point dataset, determine the variation direction control variable and neighborhood step length variable according to the research objectives and characteristics of the geospatial grid point dataset; establish a neighborhood grid based on the variation direction control variable and neighborhood step length variable, calculate the variance of the attribute intensity of each grid point in the neighborhood grid and the attribute intensity of its neighborhood grid points, quantify the variance of the neighborhood grid points, and construct a difference coefficient sequence;

[0011] Determine a difference coefficient threshold, and mark the grid points whose difference coefficients exceed the difference coefficient threshold as difference grid points;

[0012] The difference grid points are clustered to construct three-dimensional abnormal regions.

[0013] Furthermore, the variance of the attribute strength of each grid point in the neighborhood grid and the attribute strength of its neighborhood grid points is calculated using the following formula:

[0014]

[0015] Among them, σ ij 2 represents the variance, Represents the grid point V j The strength of the attribute in the lth data point, is the grid point V i The property strength of , and p is the neighborhood grid point V j The number of data points.

[0016] Furthermore, the method for quantifying the variance of the neighborhood grid points includes: performing weighted average calculation on the variance of the neighborhood grid points.

[0017] Furthermore, the method for determining the difference coefficient threshold includes: arranging the difference coefficient sequence in order from small to large or from large to small, calculating the cumulative distribution function of the difference coefficient, analyzing the changing trend of the cumulative distribution function, identifying the inflection point of the difference coefficient sequence, and the difference coefficient corresponding to the inflection point is the difference coefficient threshold.

[0018] Furthermore, the difference grid points are clustered to construct a three-dimensional abnormal area. The method includes: using a three-dimensional DBSCAN clustering algorithm, selecting a parameter combination of an optimal neighborhood radius and a minimum number of neighboring points of a core point, clustering the difference grid points, and constructing a three-dimensional abnormal area.

[0019] Furthermore, the neighborhood step variable is the range of neighboring grid points considered when calculating the coefficient of difference.

[0020] Furthermore, the variation direction control variable includes a horizontal direction, a vertical direction or all neighborhood directions.

[0021] A second aspect of the present invention provides a three-dimensional abnormal region extraction system based on variation direction control.

[0022] A three-dimensional abnormal region extraction system based on variation direction control, comprising:

[0023] The difference coefficient calculation module is configured to: obtain a geographic space grid point data set, determine the variation direction control variable and the neighborhood step length variable according to the research target requirements and the characteristics of the geographic space grid point data set; establish a neighborhood grid based on the variation direction control variable and the neighborhood step length variable, calculate the variance of the attribute intensity of each grid point in the neighborhood grid and the attribute intensity of its neighborhood grid points, quantify the variance of the neighborhood grid points, and construct a difference coefficient sequence;

[0024] A difference grid point screening module is configured to: determine a difference coefficient threshold, and mark the grid points whose difference coefficients exceed the difference coefficient threshold as difference grid points;

[0025] The clustering output module is configured to cluster the difference grid points and construct a three-dimensional abnormal area.

[0026] A third aspect of the present invention provides a computer-readable storage medium.

[0027] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the three-dimensional abnormal region extraction method based on variation direction control as described in the first aspect above.

[0028] A fourth aspect of the present invention provides a computer device.

[0029] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the three-dimensional abnormal region extraction method based on variation direction control as described in the first aspect above are implemented.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The present invention aims to solve the technical problems faced by the prior art in processing and analyzing three-dimensional abnormal areas in the spatial grid data of the earth system. The technical solution of the present invention can achieve the following technical effects:

[0032] (1) Make full use of three-dimensional spatial information: Improve the understanding of complex phenomena in the Earth system by analyzing the three-dimensional properties of spatial grid data.

[0033] (2) Improve the flexibility and adaptability of the model: Adopt an algorithm that does not rely on a fixed model, which can change the variation direction variables according to different anomaly formation mechanisms to adapt to data sets of different scales and complexities and accurately capture irregular three-dimensional anomaly areas.

[0034] (3) Improve computational efficiency and accuracy: By optimizing algorithm design, the efficiency of processing large-scale spatial grid data and the accuracy of the results can be improved.

[0035] (4) Realize automation and intelligence of data processing: Develop intelligent algorithms that can automatically identify and extract three-dimensional abnormal areas, reduce human intervention, and improve the level of automation of data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0037] Figure 1 It is a framework diagram of a three-dimensional abnormal region extraction method based on variation direction control shown in the present invention;

[0038] Figure 2 It is a flow chart of a three-dimensional abnormal region extraction method based on variation direction control shown in the present invention;

[0039] Figure 3 is a position relationship diagram of the target grid point and the neighborhood grid points shown in the present invention;

[0040] Figure 4 It is a result diagram of the three-dimensional structure extraction of the horizontal abnormal area (frontal zone) shown in the present invention. DETAILED DESCRIPTION

[0041] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0042] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0043] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0044] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the methods and systems according to various embodiments of the present disclosure. It should be noted that each box in the flowchart or block diagram can represent a module, a program segment, or a part of a code, and the module, program segment, or a part of a code may include one or more executable instructions for implementing the logical functions specified in each embodiment. 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 boxes represented in succession can actually be executed substantially in parallel, or they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the flowchart and / or block diagram, and the combination of boxes in the flowchart and / or block diagram can be implemented using a dedicated hardware-based system that performs a specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0045] Terminology explanation:

[0046] Abnormal area: refers to an area in the Earth system that has significant property differences from the surrounding environment, such as significant frontal zones and thermoclines in the ocean.

[0047] Three-dimensional region extraction: refers to identifying and reconstructing the three-dimensional shape of a region with specific attributes from three-dimensional spatial data.

[0048] Related protocols and standards: Geospatial grid datasets, also known as raster datasets, are a type of data used to represent the earth's surface or spatial distribution. They store information by dividing space into regular grids (grids). Each grid cell (also called a pixel) contains one or more data values ​​at a specific location. This type of dataset is very common in remote sensing science, geographic information systems (GIS), climate change research, and other fields that require spatial data analysis. When processing earth science data, it is necessary to follow relevant data exchange protocols and standards, such as the NetCDF (Network Common Data Format) standard, which is a widely used format for storing and distributing scientific data.

[0049] Embodiment 1

[0050] like Figure 1As shown, this embodiment provides a three-dimensional abnormal area extraction method based on variation direction control. This embodiment uses the method applied to a server as an example. It can be understood that the method can also be applied to a terminal, and can also be applied to a terminal, a server, and a system, and is implemented through the interaction between the terminal and the server. The server can be an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network servers, cloud communications, middleware services, domain name services, security services CDN, and big data and artificial intelligence platforms. The terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited to this. The terminal and the server can be directly or indirectly connected via wired or wireless communication, which is not limited in this application. In this embodiment, the method includes the following steps:

[0051] Obtain a geospatial grid point dataset, determine the variation direction control variable and neighborhood step length variable according to the research objectives and characteristics of the geospatial grid point dataset; establish a neighborhood grid based on the variation direction control variable and neighborhood step length variable, calculate the variance of the attribute intensity of each grid point in the neighborhood grid and the attribute intensity of its neighborhood grid points, quantify the variance of the neighborhood grid points, and construct a difference coefficient sequence;

[0052] Determine a difference coefficient threshold, and mark the grid points whose difference coefficients exceed the difference coefficient threshold as difference grid points;

[0053] The difference grid points are clustered to construct three-dimensional abnormal regions.

[0054] Based on the variation characteristics of the spatial grid point data in the abnormal area, the present invention proposes a method for extracting the three-dimensional structure of the abnormal area based on the spatial grid point difference coefficient. The algorithm first calculates the intensity variance of the spatial neighborhood of each grid point in the spatial data based on the variation direction control variable and the neighborhood step length variable, takes the overall mean as the difference coefficient of the grid point, and then sets a threshold to detect whether it is a difference grid point (which constitutes an abnormal three-dimensional structure), and clusters the grid point set into a volume. Figure 2As shown, the input of this algorithm is a spatial grid data set, a variation direction control variable, a neighborhood step variable, and DBSCAN parameters (Eps and MinPts), and the output is the clustering result of the three-dimensional structure of the abnormal area. The algorithm is divided into three stages: difference coefficient calculation, difference grid point screening, and abnormal area three-dimensional structure construction. The first stage is the difference coefficient calculation. The difference coefficient represents the difference between a single grid point and the surrounding. By inputting the variation direction control variable and the neighborhood step variable, the difference coefficient calculated is used to divide all data points into two categories, namely normal points and difference points; the second stage is the difference grid point screening. This screening process depends on the threshold setting, that is, the set of difference grid points is defined by the appropriate quantile or statistical measure of the difference coefficient; the third stage is the three-dimensional structure construction, which involves the clustering of the difference grid point set. After determining the difference grid point set, the grid point set needs to be clustered based on density to find data point groups that share similar properties. To this end, the present invention adopts the 3D DBSCAN density clustering algorithm, and aggregates the difference grid points into clusters through appropriate MinPts and Eps settings. This cluster is the three-dimensional structure of the abnormal area.

[0055] According to the mechanism of geographical anomaly, the present invention introduces the concept of variation direction control variable, which is used to specify the variation direction of main concern in the analysis. In the analysis of earth system data, the variation direction can be horizontal, vertical, or a combination of the two. For example, in oceanographic research, the change of temperature with depth (vertical direction) or with geographic location (horizontal direction) may be of interest. The variation direction control variable enables the algorithm to adapt to different research objectives and spatial data characteristics, thereby capturing the variation characteristics in a specific dimension.

[0056] The neighborhood step variable defines the range of neighboring grid points considered when calculating the coefficient of difference, thereby determining the spatial scale of interaction between grid points. By adjusting the neighborhood step, the algorithm can adapt to spatial variability at different scales, capturing small-scale local changes to large-scale regional variations. The design of this variable enables the algorithm to not only process uniform spatial data, but also adapt to data sets with complex spatial heterogeneity.

[0057] The following is a detailed description of the three-dimensional abnormal region extraction method based on variation direction control described in this embodiment:

[0058] Step S1: Calculation of coefficient of difference. In the present invention, the concept of coefficient of difference (CD) is introduced to quantify the differences in attribute strength of grid points in their neighborhoods. The calculation of the coefficient of difference is based on the spatial distribution and attribute data of the grid points, and aims to reveal the degree of heterogeneity between each spatial grid point and its neighborhood in specific attributes. The coefficient takes into account the variation direction control variable and the neighborhood step variable respectively during the calculation process. This step is used to select the heterogeneity within the neighborhood range of the main focus direction in the analysis. Assuming that the neighborhood step is 1, the variation direction includes horizontal, vertical, or all neighborhood directions, depending on the research purpose and the characteristics of the spatial data.

[0059] like Figure 3 As shown, for the grid point V i , its coefficient of difference CD(V i ) by calculating all the neighborhood grid points V within the predetermined neighborhood step length S j Specifically, the coefficient of variation is based on the grid point V i Attribute Strength and the neighborhood grid point set N(V i ,S) in each grid point The variance is quantified by ij 2 The calculation is as follows:

[0060]

[0061] in, Represents the grid point V j The strength of the attribute in the lth data point, is the grid point V i The property strength of , and p is the neighborhood grid point V j The number of data points. The coefficient of variation d i By calculating the variance σ of the neighborhood grid points ij 2 Taking the weighted average we get:

[0062]

[0063] Among them, w j is the neighborhood grid point V j The weight of can be determined according to its spatial relationship with the grid point Vi or other correlation measures. The introduction of weights allows the algorithm to take into account the relative importance of neighborhood grid points indexed by different neighborhood step sizes S to the calculation of the difference coefficient.

[0064] Coefficient of variation d iIt provides a quantitative indicator to measure the degree of heterogeneity of a spatial grid point in its neighborhood. When selecting the control variable for the variation direction, researchers will decide whether to focus on horizontal variation, vertical variation, or a combination of the two based on the specific research objectives and the characteristics of the spatial data. This choice will affect the calculation of the coefficient of variation. In the subsequent algorithm process, the coefficient of variation will be used to screen out grid points with significant differences and serve as a key step in constructing the three-dimensional structure of the abnormal area.

[0065] The calculation of the coefficient of difference is based on the attribute strength of the grid point and the attribute strength of its neighborhood grid point set. Guided by the mutation direction control variable and the neighborhood step variable, the algorithm first determines the neighborhood range of each grid point, and then calculates the variance of the attribute strength of all grid points in the neighborhood. These variance values ​​are weighted averaged to obtain the coefficient of difference of a single grid point, which reflects the degree of heterogeneity of the grid point in its neighborhood.

[0066] The introduction of mutation direction control variables and neighborhood step size variables enables the algorithm to adapt to spatial datasets of different scales and complexities. This design allows the algorithm to work under different spatial resolutions and attribute variation conditions, thus having wide applicability in various earth science applications.

[0067] In addition, an adaptive neighborhood selection algorithm can be developed that can dynamically adjust the neighborhood step size and mutation direction according to the characteristics of the data. This approach may improve the flexibility and adaptability of the algorithm, but may require more complex algorithm design and optimization.

[0068] Step S2: Quantitative screening of difference grid points. The threshold screening stage aims to accurately determine the threshold through quantitative analysis. The setting of the threshold is the key to distinguishing significant difference grid points from background grid points, and is crucial to ensuring the authenticity and reliability of abnormal area extraction.

[0069] Step S21: Construction and sorting of difference coefficient sequence. First, the difference coefficients of all grid points are aggregated and a sequence D = {d1, d2, ..., d n}, where d i represents the difference coefficient of the i-th grid point. The sequence is then sorted according to the value to form an ascending sequence D sorted .

[0070] Step S22: Analyze the distribution characteristics of the difference coefficient. sorted Mathematical analysis is performed to reveal its distribution characteristics. The cumulative distribution function F(d) of the difference coefficient in the sequence is defined as:

[0071]

[0072] Where I is the indicator function, when d i The value is 1 when ≤d, otherwise it is 0.

[0073] Step S23: Determination of threshold value. By analyzing the changing trend of the cumulative distribution function F(d), the inflection point d in the sequence is identified. threshold These inflection points reflect abrupt changes in the variance distribution and usually correspond to significant differences in the variant regions. threshold The determination of can be achieved by finding the local maximum of the derivative of the cumulative distribution function f(d), that is:

[0074]

[0075] Step S24: Selection of difference grid points. Based on the determined threshold d threshold , select the grid points that are greater than the threshold from the sorted difference coefficient sequence. These grid points are identified as difference grid points, forming the set γ diff = {V i |d i >d threshold ,V i ∈γ}, where γ is the set of all grid points, and the difference point set γ diff It was used in subsequent cluster analysis to identify and construct the three-dimensional structure of abnormal areas.

[0076] Step S3: Construction of the three-dimensional structure of the abnormal area. In the screening process of the difference point set, noise or minor local variations may be included. If these points are directly used to construct the abnormal area, errors may be introduced. In order to accurately extract the three-dimensional structure with geographical significance from the difference point set, the present invention adopts the three-dimensional DBSCAN clustering algorithm. The algorithm identifies and forms clusters by evaluating the local density of data points, thereby effectively distinguishing high-density areas from noise points. The key parameters of the three-dimensional DBSCAN algorithm include the neighborhood radius Eps and the minimum number of neighboring points MinPts of the core point. In order to determine the best parameter combination, the present invention systematically tests a series of parameter configurations through cross-validation and clustering effect evaluation. The goal of parameter selection is to minimize the interference of noise points while maintaining the intrinsic structure of the data. After the parameters are determined, the algorithm is applied to the difference point set to form a series of clusters. Each cluster represents a continuous and internally homogeneous abnormal area in three-dimensional space as the final result of the study.

[0077] The method of the present invention can more accurately identify abnormal areas in the earth system because it takes into account the three-dimensional characteristics of spatial data and local spatial heterogeneity. By accurately calculating the difference coefficient of each grid point, the algorithm can effectively distinguish between normal points and difference points, providing a reliable data basis for subsequent three-dimensional structure extraction.

[0078] In addition, the present invention can also use machine learning methods, such as support vector machines (SVM), random forests or deep learning models, to identify and classify temperature anomaly areas in the ocean. This method can identify the characteristics of the frontal zone through training, but may require a large amount of labeled data and may not be as interpretable as the method of the present invention. Or use multi-scale analysis techniques, such as wavelet transform or scale space theory, to detect ocean temperature anomalies at different scales. This method can reveal the variation characteristics at different spatial resolutions, but may not be as intuitive and accurate as the method of the present invention when processing three-dimensional structures. The method of the present invention can also be combined with a physical model to use the principles of geophysics to guide the identification of abnormal areas and the construction of three-dimensional structures. This method may provide more accurate results, but the computational complexity and implementation difficulty may increase.

[0079] In the analysis of spatial grid data of the Earth system, accurate identification and extraction of three-dimensional abnormal areas is crucial to understanding the complex interactions of the Earth system. The core innovation of the present invention is to propose a method for calculating the difference coefficient based on the variation direction control variable and the neighborhood step variable. This method can accurately quantify the three-dimensional difference of attribute variables in spatial grid data, thereby providing a new technical means for extracting the three-dimensional structure of abnormal areas.

[0080] The invention is an innovation in the field of earth science and geographic information systems and has significant commercial value. It can provide more accurate data analysis tools for climate analysis, environmental monitoring, resource exploration, etc.

[0081] By introducing the variation direction control variable and the neighborhood step variable, the technical solution of the present invention can accurately capture the abnormal area in three-dimensional space. This accuracy is due to the fact that the algorithm can adjust the analysis direction and neighborhood range according to the characteristics of the data and the research purpose, so as to more accurately identify and quantify the spatial distribution differences of attribute variables. Compared with the simplified model or fixed direction analysis commonly used in the prior art, this can more comprehensively reveal the complex phenomena in the earth system.

[0082] The technical solution of the present invention does not rely on fixed models or assumptions, but adapts to data sets of different scales and complexities by dynamically adjusting the mutation direction and neighborhood step size. This adaptability and flexibility enables the algorithm to handle a variety of data from local small-scale changes to global large-scale changes, solving the limitations of existing technologies in processing diverse and high-resolution data.

[0083] The technical solution of the present invention provides an intuitive visual representation of abnormal phenomena in the Earth system by constructing a three-dimensional structure. This three-dimensional representation not only makes data interpretation easier, but also helps scientists and researchers to deeply understand the dynamic changes of the Earth system. Compared with two-dimensional analysis or simplified three-dimensional models, the present invention provides richer spatial information and deeper scientific insights.

[0084] The technical solution of the present invention is applicable to multiple fields such as earth science, environmental monitoring, resource exploration, etc. due to its high adaptability and accuracy. It provides strong support for interdisciplinary research and promotes cooperation and knowledge exchange between different fields. This can be demonstrated by practical application cases. For example, in marine research, the present invention can help more accurately identify areas with abnormal ocean temperature, such as abnormal horizontal ocean front phenomena, as described below.

[0085] Example: The algorithm of the present invention is applied to the detection of three-dimensional structure of the frontal zone of the ocean level anomaly phenomenon.

[0086] Application Process:

[0087] Step 1: Data collection and preprocessing

[0088] Data collection: Obtain ocean spatial grid datasets. This application is based on the IAP global ocean temperature spatial grid dataset.

[0089] Preprocessing: Clean the collected data, including removing outliers, filling missing values, and data standardization to ensure data quality.

[0090] Step 2: Calculate the coefficient of variation

[0091] Determine the variation direction: For the abnormal phenomenon of ocean level, the variation direction control variable is set to be mainly in the horizontal direction. In this invention, the neighborhood step size is set to 1 according to the spatial resolution of the data and the local variability of ocean temperature. The local changes in ocean temperature are usually closely related to the upper ocean dynamics, such as eddy activity and upwelling. Because the neighborhood step size is set to 1, w j The value of is also 1.

[0092] Calculate the coefficient of variation: Based on the variation direction control variable and the neighborhood step size variable, calculate the intensity variance of each grid point in its horizontal neighborhood to obtain the coefficient of variation.

[0093] Step 3: Difference grid screening

[0094] Setting threshold: By calculating the difference coefficient of the horizontal ocean temperature spatial grid data, construct a sequence D = {d1, d2, ..., d n}, and sort them numerically to get D sorted. And based on the sorted difference coefficient sequence D sorted , calculate the cumulative distribution function F(d). Through mathematical analysis of the cumulative distribution function F(d), special attention is paid to the inflection point of the cumulative distribution function. The inflection point indicates the mutation point in the distribution of the difference coefficient, that is, the local maximum point of the derivative. The inflection point of the cumulative distribution function based on the difference coefficient in the horizontal direction is calculated to be 7.3, so as to determine the threshold d threshold It is 7.3.

[0095] Screening difference grid points: Grid points whose difference coefficient exceeds the threshold are marked as difference grid points, which may be part of the frontal zone.

[0096] Step 4: 3D structure construction

[0097] Application of 3D DBSCAN algorithm: Use the 3D DBSCAN clustering algorithm to cluster the screened difference grid points and construct the three-dimensional structure of the frontal zone.

[0098] Parameter adjustment: According to the characteristics of ocean data, the parameters of 3D DBSCAN, such as neighborhood radius (Eps) and minimum number of samples (MinPts), are adjusted to optimize the clustering results.

[0099] The experimental results are as follows: The significant frontal zones (FZ1, FZ2, FZ3) in the horizontal anomaly area are basically distributed near 40° north and south latitude. The three-dimensional structure is a longitudinal irregular sheet with a large longitudinal span and a maximum depth of 900m. Figure 4 shown.

[0100] Embodiment 2

[0101] This embodiment provides a three-dimensional abnormal region extraction system based on variation direction control.

[0102] A three-dimensional abnormal region extraction system based on variation direction control, comprising:

[0103] The difference coefficient calculation module is configured to: obtain a geographic space grid point data set, determine the variation direction control variable and the neighborhood step length variable according to the research target requirements and the characteristics of the geographic space grid point data set; establish a neighborhood grid based on the variation direction control variable and the neighborhood step length variable, calculate the variance of the attribute intensity of each grid point in the neighborhood grid and the attribute intensity of its neighborhood grid points, quantify the variance of the neighborhood grid points, and construct a difference coefficient sequence;

[0104] A difference grid point screening module is configured to: determine a difference coefficient threshold, and mark the grid points whose difference coefficients exceed the difference coefficient threshold as difference grid points;

[0105] The clustering output module is configured to cluster the difference grid points and construct a three-dimensional abnormal area.

[0106] In one or more implementations, the calculation of the variance of the attribute strength of each grid point in the neighborhood grid and the attribute strength of its neighborhood grid points uses the following formula:

[0107]

[0108] Among them, σ ij 2 represents the variance, Represents the grid point V j The strength of the attribute in the lth data point, is the grid point V i The property strength of , and p is the neighborhood grid point V j The number of data points.

[0109] In one or more embodiments, the method for quantifying the variance of the neighborhood grid points includes: performing weighted average calculation on the variance of the neighborhood grid points.

[0110] In one or more embodiments, the method for determining the difference coefficient threshold includes: arranging the difference coefficient sequence in order from small to large or from large to small, calculating the cumulative distribution function of the difference coefficient, analyzing the changing trend of the cumulative distribution function, identifying the inflection point of the difference coefficient sequence, and the difference coefficient corresponding to the inflection point is the difference coefficient threshold.

[0111] In one or more embodiments, the difference grid points are clustered to construct a three-dimensional abnormal area, and the method includes: using a three-dimensional DBSCAN clustering algorithm, selecting a parameter combination of an optimal neighborhood radius and a minimum number of neighboring points of a core point, clustering the difference grid points, and constructing a three-dimensional abnormal area.

[0112] In one or more embodiments, the neighborhood step variable is the range of neighboring grid points considered when calculating the coefficient of difference.

[0113] In one or more embodiments, the mutation direction control variable includes a horizontal direction, a vertical direction, or all neighborhood directions.

[0114] Embodiment 3

[0115] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the three-dimensional abnormal region extraction method based on variation direction control as described in the first embodiment above are implemented.

[0116] Embodiment 4

[0117] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the three-dimensional abnormal region extraction method based on variation direction control as described in the first embodiment above are implemented.

[0118] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0119] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0120] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0121] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0122] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0123] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A three-dimensional abnormal region extraction method based on variation direction control, characterized in that: include: Obtain a geospatial grid data set, and determine the variation direction control variable and neighborhood step variable based on the research objectives and the characteristics of the geospatial grid data set; A neighborhood grid is established based on the mutation direction control variable and the neighborhood step size variable, the variance of the attribute strength of each grid point in the neighborhood grid and the attribute strength of its neighborhood grid points is calculated, the variance of the neighborhood grid points is quantified, and a difference coefficient sequence is constructed; Determine a difference coefficient threshold, and mark the grid points whose difference coefficients exceed the difference coefficient threshold as difference grid points; The difference grid points are clustered to construct three-dimensional abnormal regions.

2. The three-dimensional abnormal region extraction method based on variation direction control according to claim 1 is characterized in that: The calculation of the variance of the attribute strength of each grid point in the neighborhood grid and the attribute strength of its neighboring grid points adopts the following formula: in, represents the variance, Represents the grid point V j The strength of the attribute in the lth data point, is the grid point V i The property strength of , and p is the neighborhood grid point V j The number of data points.

3. The three-dimensional abnormal region extraction method based on variation direction control according to claim 1 is characterized in that: The method for quantifying the variance of the neighborhood grid points includes: performing weighted average calculation on the variance of the neighborhood grid points.

4. The three-dimensional abnormal region extraction method based on variation direction control according to claim 1 is characterized in that: The method for determining the difference coefficient threshold includes: arranging the difference coefficient sequence in an ascending order or a descending order, calculating the cumulative distribution function of the difference coefficient, analyzing the change trend of the cumulative distribution function, identifying the inflection point of the difference coefficient sequence, and the difference coefficient corresponding to the inflection point is the difference coefficient threshold.

5. The three-dimensional abnormal region extraction method based on variation direction control according to claim 1 is characterized in that: The difference grid points are clustered to construct a three-dimensional abnormal area. The method includes: using a three-dimensional DBSCAN clustering algorithm, selecting a parameter combination of an optimal neighborhood radius and a minimum number of neighboring points of a core point, clustering the difference grid points, and constructing a three-dimensional abnormal area.

6. The three-dimensional abnormal region extraction method based on variation direction control according to claim 1 is characterized in that: The neighborhood step variable is the range of neighboring grid points considered when calculating the coefficient of difference.

7. The three-dimensional abnormal region extraction method based on variation direction control according to claim 1 is characterized in that: The mutation direction control variable includes a horizontal direction, a vertical direction or all neighborhood directions.

8. A three-dimensional abnormal region extraction system based on variation direction control, characterized in that: include: A coefficient of variation calculation module is configured to: obtain a geographic space grid point data set, and determine a variation direction control variable and a neighborhood step length variable according to research target requirements and characteristics of the geographic space grid point data set; A neighborhood grid is established based on the mutation direction control variable and the neighborhood step size variable, the variance of the attribute strength of each grid point in the neighborhood grid and the attribute strength of its neighborhood grid points is calculated, the variance of the neighborhood grid points is quantified, and a difference coefficient sequence is constructed; A difference grid point screening module is configured to: determine a difference coefficient threshold, and mark the grid points whose difference coefficients exceed the difference coefficient threshold as difference grid points; The clustering output module is configured to cluster the difference grid points and construct a three-dimensional abnormal area.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the three-dimensional abnormal area extraction method based on variation direction control as described in any one of claims 1 to 7 are implemented.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the three-dimensional abnormal area extraction method based on variation direction control as described in any one of claims 1 to 7 are implemented.

Images