A method for extracting and classifying mountain top based on DEM data

Abstract

The application discloses a mountain top point extraction and classification method based on DEM data and relates to the technical field of terrain analysis based on a digital elevation model, and solves the problem that current researches roughly define mountain top points and focus on the feature that a mountain top point is a local highest value, and the feature of a surface shape around a region where the mountain top point is located is not considered, and the technical scheme is as follows: S1, inputting DEM data, determining an analysis window size, and calculating initial mountain top points of an analysis region according to a determined local maximum value; S2, sequentially taking the initial mountain top points generated in the step S1 as circle centers, taking the window size determined in the step S1 as diameters, determining an analysis circle, and determining all grid cells through which a boundary of the analysis circle passes; and the effect is that the application covers two key technologies of mountain top shape index calculation and mountain top point classification considering the shape feature, and the accuracy of extracting mountain top points based on DEM data can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of terrain analysis based on digital elevation model, more particularly, it relates to a mountain top point extraction and classification method based on DEM data. BACKGROUND

[0002] From the perspective of geomorphology, the mountain top is the highest part of a mountain body with a certain spatial area, while in the field of mapping, spatial analysis, etc., the mountain top is abstracted as a point feature. The mountain top point is an important part of the terrain feature elements and is the key feature of the terrain contour skeleton. It is also an important terrain feature that affects hydrological processes, vegetation distribution, etc. Therefore, accurately and quickly extracting the mountain top point of a region is of great significance to the terrain analysis, topographic mapping, etc. of the region.

[0003] From the perspective of data sources, topographic maps and DEMs both contain terrain elevation information, so they are also the main data sources for extracting mountain top points. From the perspective of extraction methods, the existing mountain top point extraction processes can be divided into two strategies: the first is direct extraction, such as using DEM data combined with hydrological analysis to extract mountain top points, and using contour data to extract mountain top points according to the topological relationship between mountain top points and contour lines; the second strategy is to first find the initial mountain top point based on neighborhood statistics, and then remove the pseudo mountain top points that do not meet the definition of mountain top points according to the shape characteristics of the region where the mountain top point is located or the spatial topological relationship with the nearby contour lines.

[0004] Further summary and analysis of the existing literature on mountain top point extraction show that the current research on the definition of mountain top points is relatively rough, focusing on the local highest point of the mountain top, and not considering the shape characteristics of the surrounding terrain where the mountain top point is located. For example, the local highest point formed by the intersection of multiple ridge lines is generally a mountain top point, while on a ridge line, a local highest point may also be formed due to terrain undulations. Whether that point should be defined as a mountain top point? There is no systematic discussion and analysis of this issue, which has led to the fact that there is no consensus on which pseudo mountain top points should be removed and which mountain top points are missing in the accuracy analysis part after the mountain top point extraction, which has also affected the application value of the mountain top point in the field of terrain analysis, topographic mapping, etc.

[0005] To solve the above problems, the application provides a mountain top point extraction and classification method based on DEM data on the basis of existing research. The method firstly obtains initial mountain top points of an analysis region based on neighborhood analysis of local maximum elevation, then designs three parameter indexes from the perspective of differentiating the surface shape of the mountain top point region, and finally removes false mountain top points according to the three parameter indexes calculated by DEM, and classifies the remaining mountain top points into steep mountain top points, flat mountain top points and ridge line mountain top points according to the surface shape features of the mountain top point region. The method provides a new method for fast and accurate extraction of mountain top points, and can realize the shape classification of mountain top points, and has great application value for digital landform mapping and other subject research. SUMMARY

[0006] The application designs three parameter indexes and their calculation methods based on DEM data from the perspective of describing the surface shape features of the mountain top region, and classifies the initial mountain top points extracted based on neighborhood statistics based on the three parameters, so as to remove false mountain top points and classify the mountain top points according to the shape. The method includes two key technologies of mountain top shape index calculation and mountain top point classification considering the shape features, and can improve the accuracy of mountain top point extraction based on DEM data.

[0007] The technical scheme adopted by the application is as follows:

[0008] A mountain top point extraction and classification method based on DEM data, comprising the following steps:

[0009] S1, input DEM data, determine an analysis window size, and calculate initial mountain top points of an analysis region according to a local maximum value determination method; in the software ArcGIS, input DEM data, the region is different, or the resolution of the DEM used is different, and the parameter may be different, in the example data, the analysis window used is a rectangle of 1200m*1200m;

[0010] S2, sequentially take the initial mountain top points generated in step S1 as the center of a circle, take the window size determined in step S1 as the diameter, determine an analysis circle, and determine all grid cells through the boundary of the analysis circle;

[0011] S3, define the vector from the center of the circle to the center point of the grid cell on the boundary of the analysis circle for all grid cells through the boundary of the analysis circle determined in step S2, then calculate the included angle between the vector and the downward unit vector at the center of the circle, and then calculate the three parameter values: convexity average index, convexity standard deviation index and convexity unevenness index according to the calculated angle values;

[0012] S4, according to the three parameter values calculated in step S3, further processing the hilltop points extracted in step S1, removing false hilltop points, and classifying the hilltop points according to morphological characteristics.

[0013] Further, the specific process of step S3 is as follows:

[0014] S31, defining the center of the circle, the direction being a unit vector vertically downward, denoted as ;

[0015] S32, sequentially reading the three-dimensional coordinates of the center points of the grid cells on the boundary of the analysis circle, and then calculating the vector connecting the center of the circle and the center point of the grid cell on the analysis circle, denoted as ;

[0016] S33, sequentially calculating the cosine value of the angle between the vector and the vector , and storing it in a designated array;

[0017] S34, calculating the average value and the standard deviation of all the calculated cosine values, respectively, as the first and second parameters describing the morphological characteristics of the hilltop points;

[0018] S35, calculating the difference between the elevation value of the center of the circle and the elevation value of the grid cell on the boundary of the analysis circle, selecting the grid cell with the largest difference value, and marking it as C1;

[0019] S36, on the boundary of the analysis circle, for the grid cells other than C1, finding another grid cell such that the angle formed by the center point of the grid cell, the center of the circle, and the center point of the grid cell C1 is the smallest, marking the grid cell as C2, and calculating the cosine value of the angle formed by the center point of the grid cell C1, the center of the circle, and the center point of the grid cell C2, and adding 1 to the calculated cosine value as the third parameter describing the morphological characteristics of the hilltop points.

[0020] Further, the specific process of step S4 is as follows:

[0021] S41, according to the situation of the three parameter values calculated to describe the morphological characteristics of the hilltop points, using the classical Kmeans spatial clustering algorithm to divide the initial hilltop points into three categories;

[0022] S42, using the DEM data of the analysis area to generate a hypsometric map of the analysis area;

[0023] S43, according to the classification result of S41, and referring to the hypsometric tints, the initial mountain top points are classified and marked, wherein the mountain top points on the flat terrain are false mountain top points, the morphological parameter characteristics of which are that the three parameter values are relatively small; the mountain top points on a single ridge line are single-ridge-type mountain top points, the morphological parameter characteristics of which are that the convexity average index and the convexity standard deviation index are relatively large, and the convexity imbalance index is relatively small; and the remaining points are multi-ridge intersection-type mountain top points.

[0024] A mountain top point extraction and classification device based on DEM data, comprising:

[0025] An analysis module for inputting DEM data, determining an analysis window size, and calculating initial mountain top points of an analysis region according to a local maximum value determination method;

[0026] A determination module for sequentially taking the initial mountain top points generated in the analysis module as the center, taking the window size determined by the analysis module as the diameter, determining an analysis circle, and determining all grid cells through which the boundary of the analysis circle passes;

[0027] A calculation module for defining the vector from the center of the circle to the center of the grid cell on the boundary of the analysis circle for all grid cells through which the boundary of the analysis circle passes determined by the determination module, then calculating the included angle between the vector and the downward unit vector at the center of the circle, and then sequentially calculating three parameter values: the convexity average index, the convexity standard deviation index, and the convexity imbalance index according to the calculated angle values;

[0028] A processing module for classifying the initial mountain top points into three categories by using the classical Kmeans spatial clustering algorithm according to the three parameter values calculated by the calculation module, and classifying and marking the mountain top points extracted by the analysis module by means of the hypsometric tints generated from the DEM data of the analysis region, i.e., marking into three categories: false mountain top points, single-ridge-type mountain top points, and multi-ridge intersection-type mountain top points.

[0029] A computer readable storage medium having a computer program stored thereon, wherein the computer program performs the above method when run by a computer.

[0030] The technical features and beneficial effects of the present application are as follows:

[0031] (1) The mountain top point extraction and classification method proposed by the present application first extracts initial mountain top points according to the principle of maximum local elevation value, and then designs parameter indexes for describing the morphological characteristics of the mountain top point region, so that the false mountain top points on the flat terrain can be removed, and the obtained mountain top points can be ensured to be without omission.

[0032] (2) On the basis of removing false mountain top points, according to the difference of the surface form of the mountain top point area, the method designs relevant parameter indexes to distinguish different mountain top point types, and finally divides the mountain top points into isolated type, single ridge type and multiple ridge intersection type and several other types, so that the classification research of the topographic feature points can be further refined. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 It is a flowchart of the method of the present application.

[0034] Figure 2 It is a schematic diagram of determining the grid cells on the boundary of the analysis circle in the method of the present application.

[0035] Figure 3 It is a schematic diagram of calculating the cosine value of the included angle of two vectors in the method of the present application.

[0036] Figure 4 It is a schematic diagram of calculating the form parameter process in the method of the present application.

[0037] Figure 5 It is a schematic diagram of the DEM data in the experimental area in the embodiment of the present application.

[0038] Figure 6 It is a schematic diagram of the initial mountain top points extracted in the experimental area in the embodiment of the present application.

[0039] Figure 7 It is a schematic diagram of the mountain top points and classification extracted in the experimental area in the embodiment of the present application.

[0040] Figure 8 It is a schematic diagram of the typical false mountain top points extracted in the experimental area in the embodiment of the present application.

[0041] Figure 9 It is a schematic diagram of the typical single ridge type mountain top points extracted in the experimental area in the embodiment of the present application.

[0042] Figure 10 It is a schematic diagram of the typical multiple ridge intersection type mountain top points extracted in the experimental area in the embodiment of the present application.

[0043] Figure 11 It is a module block diagram in the embodiment of the present application. DETAILED DESCRIPTION

[0044] The principles and characteristics of the present application are described below in combination with the drawings, and the examples are only used to explain the present application, and are not used to limit the scope of the present application.

[0045] As shown in the drawings, Figure 1 The specific steps of the mountain top point extraction and classification method based on DEM data are as follows:

[0046] Step 1: input DEM data, determine a window size, and calculate the initial hilltop point according to the determined local maximum value;

[0047] Step 2: sequentially execute steps 3-10 on each initial hilltop point obtained in step 1;

[0048] Step 3: determine an analysis circle with the current selected hilltop point as the center and the window size determined in step S1 as the diameter, and record all grid cells passed by the boundary of the analysis circle Figure 2 );

[0049] Step 4: count the number N of grid cells on the boundary of the analysis circle, and create an array A containing N elements;

[0050] Step 5: for each grid cell on the boundary of the analysis circle, mark its spatial coordinates as , mark the spatial coordinates of the current hilltop point as , define the vector from the center of the circle to the center of the grid cell on the boundary of the analysis circle , define the unit vector passing through the center of the circle and perpendicular to the downward direction , calculate the cosine value of the angle between the two vectors Figure 3 , and the calculation formula is:

[0051]

[0052] Then store the calculated cosine value of the angle in array A;

[0053] Step 6: calculate the average value of the N values obtained in array A, which is the first morphological parameter of the hilltop point, i.e. the average convexity index;

[0054] Step 7: calculate the standard deviation of the N values obtained in array A, which is the second morphological parameter of the hilltop point, i.e. the standard deviation of the convexity index;

[0055] Step 8: calculate the difference between the current hilltop point and the elevation value of the grid on the boundary of the analysis circle, select the boundary grid cell corresponding to the maximum difference, and mark it as C1 Figure 4 );

[0056] Step 9: for the grid cells on the boundary of the analysis circle except C1, find another grid cell such that the angle formed by the center of the grid cell, the hilltop point, and the center of grid cell C1 is minimized, and mark the grid cell as C2 Figure 4 );

[0057] Step 10: mark the spatial coordinates of the center of grid cell C1 as , and mark the spatial coordinates of the center of grid cell C2 as , the spatial coordinates of the current mountain peak O are marked as , the cosine value of the included angle formed by the three points is calculated (COS) Figure 4 , and the calculation formula is:

[0058]

[0059] The calculated cosine value is added by 1 as the third form parameter of the mountain peak, i.e. the convexity imbalance index;

[0060] Step 11: According to the calculated three parameter values describing the form of the mountain peak, the classical Kmeans spatial clustering algorithm is used for classification of the mountain peak, and the classification number is specified as 3;

[0061] Step 12: Using the DEM data of the analysis area, a topographic shading map of the analysis area is generated;

[0062] Step 13: According to the classification result, and referring to the topographic shading map, the initial mountain peak is classified and marked. The mountain peak located on the flat terrain is a pseudo mountain peak, and the form parameter characteristics are that the three parameter values are relatively small; the mountain peak located on a single ridge line is a single ridge type mountain peak, and the form parameter characteristics are that the convexity average index and the convexity standard deviation index are relatively large, and the convexity imbalance index is relatively small; the remaining points are multi-ridge intersection type mountain peaks.

[0063] Example 1

[0064] In order to verify the application effect of the method in the extraction and classification of mountain peaks, the example selects a mountainous area located in the Qinghai-Tibet Plateau as the experimental area (Xiaoshan, Yushu, Qinghai, China). Figure 5 The average elevation of the area is close to 4000 meters, the maximum elevation difference is more than 4000 meters, the main terrain is mountainous, and there is also a large area of flat terrain between the mountains, which is an ideal experimental area for verifying the application effect of the method.

[0065] Step 1: Based on the DEM data of the experimental area, the initial mountain peak of the experimental area is extracted by using the method of finding the maximum value of the neighborhood (Xiaoshan, Yushu, Qinghai, China). Figure 6 , wherein the neighborhood size is 41xh, and h is the grid size of the DEM data, which is 30 meters in this example;

[0066] Step 2: Based on the DEM data of the experimental area, the convexity average index of the area is calculated, and the radius of the circular analysis window is specified as 21xh, wherein h is the grid size of the DEM data, which is 30 meters in this example;

[0067] Step 3: Based on the DEM data of the experimental area, the convexity standard deviation index of the area is calculated, and the radius of the circular analysis window is specified as 21xh, wherein h is the grid size of the DEM data, which is 30 meters in this example;

[0068] Step 4: Based on the DEM data of the experimental area, calculate the convexity imbalance index of the area. When calculating, the radius of the circular analysis window is specified as 21×h, where h is the size of the DEM data raster, which is 30 meters in this example.

[0069] Step 5: Assign the average convexity index, standard deviation convexity index, and convexity imbalance index calculated in Steps 2, 3, and 4 to the initial mountain peak extracted in Step 4.

[0070] Step 6: Use the K-means spatial clustering method to perform cluster analysis on the initial mountain peaks based on three parameters: the average convexity index, the standard deviation of convexity index, and the convexity imbalance index. Specify the number of clusters as 3, and obtain the classification results. Figure 7 ).

[0071] Figure 8 , Figure 9 , Figure 10 The classification results show three types of mountain peaks: pseudo-mountain peaks, single-ridge type mountain peaks, and multiple-ridge confluence type mountain peaks, combined with the overall classification result diagram ( Figure 7 It can be seen that the method of the present invention can accurately identify the mountain peaks in the experimental area and classify them according to the morphological characteristics of the mountain peak areas. The results are accurate, indicating that the method of the present invention has good application prospects in the extraction of terrain feature elements and the creation of fine geomorphic element maps.

Claims

1. A method for extracting and classifying mountain top points based on DEM data, characterized in that, The method comprises the following steps: S1, inputting DEM data, determining a window size, and calculating initial mountain peak points in the analysis area according to a method for determining local maximum values; S2, determining an analysis circle with the initial mountain peak points as the center and the window size as the diameter, and determining all grid cells through the boundary of the analysis circle; S3, defining a vector from the center of the analysis circle to the center of a grid cell on the boundary of the analysis circle, calculating the included angle between the vector and a unit vector downward from the center of the analysis circle, and then calculating three parameter values, namely a convexity average index, a convexity standard deviation index and a convexity imbalance index, according to the calculated angle values; S4, further processing the mountain peak points extracted in step S1 according to the three parameter values calculated in step S3, removing false mountain peak points, and classifying the mountain peak points according to morphological characteristics; The specific process of step S3 is as follows: S31, define the center of the circle, the direction is the vertical downward unit vector, denoted as ; S32, read the three-dimensional coordinates of the center points of the grid cells on the analysis circle boundary in sequence, and then calculate the vectors connecting the circle center and the center points of the grid cells on the analysis circle, denoted as ; S33, sequentially calculate the vector cosine of the angle with the vector cosine of the angle with the vector S34, calculating the average value and the standard deviation of all calculated cosine values as the first and second parameters for describing the morphological characteristics of the mountain peak points; S35, calculating the difference between the elevation value of the center of the analysis circle and the elevation value of each grid cell on the boundary of the analysis circle, selecting a grid cell with the maximum difference value as C1; S36, for the grid cells on the boundary of the analysis circle except C1, finding another grid cell such that the included angle formed by the center of the grid cell, the center of the grid cell C1 and the center of the analysis circle is the smallest, marking the grid cell as C2, and calculating the cosine value of the included angle formed by the center of the grid cell C1, the center of the analysis circle and the center of the grid cell C2, and adding 1 to the calculated cosine value as the third parameter for describing the morphological characteristics of the mountain peak points; The specific process of step S4 is as follows: S41, using a classical Kmeans spatial clustering algorithm to divide the initial mountain peak points into three categories according to the three parameter values for describing the morphological characteristics of the mountain peak points; S42, generating a hypsometric map of the analysis area by using the DEM data of the analysis area; S43, classifying and marking the initial mountain peak points according to the classification results of S41 and referring to the hypsometric map, wherein the mountain peak points located on flat terrain are false mountain peak points, the mountain peak points located on a single ridge line are single-ridge-type mountain peak points, and the remaining points are multi-ridge-intersection-type mountain peak points.

2. A device for extracting and classifying summit points based on DEM data, for implementing the method according to claim 1, characterized in that, The method comprises the following steps: An analysis module for inputting DEM data, determining a window size, and calculating initial mountain peak points in the analysis area according to a method for determining local maximum values; A determination module for determining an analysis circle with the initial mountain peak points as the center and the window size as the diameter, and determining all grid cells through the boundary of the analysis circle; A calculating module is configured to define a vector from the center of the analysis circle to the center of the grid cell on the boundary of the analysis circle for all grid cells passed by the boundary of the analysis circle determined in the determining module, calculate an angle between the vector and a unit vector downward at the center of the analysis circle, and then calculate three parameter values, namely, a convexity average index, a convexity standard deviation index, and a convexity unevenness index, in sequence according to the calculated angle value; A processing module is configured to divide the initial mountain top points into three categories by using a classical Kmeans spatial clustering algorithm according to the three parameter values calculated by the step calculating module, and to mark the mountain top points extracted by the analyzing module by means of a hypsometric tint map generated from the DEM data of the analysis region, namely, to mark the mountain top points into three categories, namely, pseudo mountain top points, single-ridge-type mountain top points, and multi-ridge-intersection-type mountain top points.

3. A computer-readable storage medium, characterized in that, A computer program is stored thereon, which, when executed by a computer, performs the method as claimed in claim 1.

Images