DEM data automatic synthesis method based on wavelet analysis

Abstract

The invention discloses a DEM data automatic synthesis method based on wavelet analysis, and the method comprises the steps: processing original data through setting a wavelet function and a decomposition layer number according to the features of landform represented by DEM data of a target region, and obtaining low-frequency data after wavelet decomposition; curvature values of all position target points are calculated based on the low-frequency data, feature points are selected, and new high-frequency coefficients are obtained through an interpolation method by taking the feature points as references; performing filtering operation on the low-frequency data after wavelet decomposition, and calculating a standard deviation; and reconstructing the last layer of the low-frequency data after the filtering operation and the new high-frequency coefficient to obtain reconstructed DEM data, and performing elevation mapping on the reconstructed DEM data to obtain integrated DEM data. According to the method, the DEM data are quickly, efficiently and automatically synthesized, so that the DEM data with different detailed degrees are provided, and the requirements of users on multi-scale landform data are met.

Description

Technical Field

[0001] The invention relates to the technical field of map data integration, and in particular to an automatic DEM data integration method based on wavelet analysis. Background Art

[0002] Geospatial data is an abstract expression of the objective geographical world we live in. Since entering the information age, people have been exploring the objective world more and more frequently, and the means and methods of observing the real world have become more and more diverse, and the data obtained to describe the objective world has become more massive.

[0003] DEM data is an important way to express the topography of the earth's surface. There are many ways to obtain it, including satellite images, synthetic aperture radar, field measurements, and topographic map conversion. Due to the extreme complexity of the objective world, the DEM data obtained by measurement is only a rough description of the objective world. At the same time, the geographical phenomena reflected by DEM data of different levels of detail are also different. Therefore, to ensure that the data can accurately reflect the geographical features without too much redundant information, it is necessary to reasonably optimize the data in a comprehensive manner to meet the needs.

[0004] Therefore, how to comprehensively optimize the acquired DEM data to ensure that it accurately reflects the essential characteristics of the topography has become an urgent problem to be solved. Summary of the invention

[0005] In order to solve the above technical problems, the present invention proposes an automatic integration method of DEM data based on wavelet analysis, which not only meets people's needs for data of different levels of detail to solve different problems, but also can accurately reflect the characteristics of topography.

[0006] To achieve the above object, the present invention provides a DEM data automatic integration method based on wavelet analysis, comprising:

[0007] According to the characteristics of the topography represented by the digital elevation model (DEM) data of the target area, the original data is processed by setting the wavelet function and decomposition layer number of the wavelet decomposition to obtain the low-frequency data after wavelet decomposition;

[0008] Based on the low-frequency data, the curvature values ​​of all position target points are calculated, and feature points are selected, and new high-frequency coefficients are obtained by interpolation method based on the feature points;

[0009] Performing a filtering operation on the low-frequency data after the wavelet decomposition, and calculating a standard deviation, thereby obtaining the low-frequency data after the filtering operation;

[0010] The last layer of the low-frequency data after the filtering operation and the new high-frequency coefficients are reconstructed to obtain reconstructed DEM data, and elevation mapping is performed on the reconstructed DEM data to obtain integrated DEM data.

[0011] Preferably, setting the wavelet function and the number of decomposition layers of wavelet decomposition includes:

[0012] According to the landform characteristics, wavelet functions with different advantages are selected, and the number of decomposition layers is determined; wherein the number of decomposition layers is specifically:

[0013] N = round(log2x)

[0014] Where N is the number of decomposition levels, round represents the rounding operation, and x is the target scale denominator / original scale denominator.

[0015] Preferably, the method for calculating the curvature values ​​of all position target points is:

[0016]

[0017] Where K is the curvature value of each target point, f x 、f y are the rates of change in the horizontal and vertical directions, respectively, and f xx 、f yy are the second-order derivatives in the horizontal and vertical directions, respectively, f xy is the mixed change rate in the horizontal and vertical directions, and slope is the slope.

[0018] Preferably, selecting the feature points includes:

[0019] A curvature threshold is calculated, and a point whose curvature value is greater than the curvature threshold is selected as the feature point; wherein the curvature threshold is calculated by setting a reference curvature threshold.

[0020] Preferably, the method for calculating the curvature threshold is:

[0021]

[0022] Where, T 阈 is the curvature threshold, A 基 is the base curvature threshold, D M is the target scale denominator, D R is the original scale denominator.

[0023] Preferably, the method for calculating the standard deviation is:

[0024]

[0025] In the formula, σ is the standard deviation, S 基 is the benchmark standard deviation, D M is the target scale denominator, D R is the original scale denominator.

[0026] Compared with the prior art, the present invention has the following advantages and technical effects:

[0027] (1) The present invention selects a suitable wavelet decomposition function to process DEM data after accurately analyzing the topography and landforms expressed by the data. There are many wavelet decomposition functions, each of which has its advantages and disadvantages. By analyzing the topography and landforms expressed by DEM data, a reasonable wavelet function is selected for different topography and landforms to ensure that the topographic features are retained to the greatest extent and reduce the influence of the wavelet function on the data;

[0028] (2) The present invention adaptively changes the number of wavelet decomposition layers N, curvature threshold T, and standard deviation σ with different integration levels. When the DEM data is automatically integrated, the integration level of the data also changes accordingly with the change of the scale span. In the integration process, the key parameters N, T, and σ adaptively change with the change of the scale span. The larger the scale span, the higher the degree of automatic integration of the data.

[0029] (3) The present invention obtains the final DEM data by fusing the low-frequency coefficients of the last layer after the filtering operation with the high-frequency coefficients obtained by interpolation; this not only completely preserves the landform features expressed by the original data, but also eliminates the redundant information of the original data, so as to facilitate more accurate transmission of effective information. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0031] Figure 1 A flow chart of an automatic DEM data integration method based on wavelet analysis according to an embodiment of the present invention;

[0032] Figure 2 It is a DEM data display diagram of an embodiment of the present invention;

[0033] Figure 3 A display diagram of geomorphic feature points of DEM data according to an embodiment of the present invention;

[0034] Figure 4 It is a display diagram of DEM data after integration according to an embodiment of the present invention;

[0035] Figure 5This is a comparison diagram of the contour lines of the original DEM data and the integrated DEM data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0037] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0038] To solve the problem of how to comprehensively optimize the acquired DEM data, not only solve the problem that DEM data expresses a lot of information and cannot quickly make an accurate judgment on the general situation of the terrain, but also meet the problem that the degree of detail of DEM data required for different needs is different. Therefore, the present invention proposes an automatic synthesis method of DEM data based on wavelet analysis, such as Figure 1 ,include:

[0039] According to the characteristics of the topography represented by the DEM data of the target area, the original data is processed by setting the wavelet function and decomposition layer number of the wavelet decomposition to obtain the low-frequency data after wavelet decomposition;

[0040] Based on the low-frequency data, the curvature values ​​of all position target points are calculated, and feature points are selected, and new high-frequency coefficients are obtained by interpolation method based on the feature points;

[0041] Performing a filtering operation on the low-frequency data after the wavelet decomposition, and calculating a standard deviation, thereby obtaining the low-frequency data after the filtering operation;

[0042] The last layer of the low-frequency data after the filtering operation and the new high-frequency coefficients are reconstructed to obtain reconstructed DEM data, and elevation mapping is performed on the reconstructed DEM data to obtain integrated DEM data.

[0043] This embodiment uses wavelet function to decompose DEM data, and then processes low-frequency coefficients, first calculates the curvature of the low-frequency coefficients, then selects feature points through the curvature threshold, and then uses the selected feature points as a reference to interpolate new high-frequency coefficients using the natural neighbor interpolation method, and then performs filtering operations on the original low-frequency coefficients to obtain new low-frequency coefficients, and then reconstructs the new high-frequency coefficients and the new low-frequency coefficients, and finally maps the original DEM data to the reconstructed DEM data based on the elevation value of the original DEM data, thereby obtaining the integrated DEM data. This method not only performs reasonable comprehensive optimization on the DEM data, but also retains the original topographic information to the greatest extent.

[0044] Furthermore, by analyzing the surface information displayed by the DEM data, we can understand its main topographic and geomorphic features, and select wavelet functions with different advantages according to its geomorphic features to ensure that the DEM data can retain its main topographic and geomorphic features as much as possible during wavelet decomposition. Figure 2 This is a graph showing the DEM data.

[0045] When determining the wavelet function, it is also necessary to determine its decomposition level. The number of wavelet decomposition levels is related to the scale span to ensure its adaptability and avoid the influence of a specific decomposition level on the data.

[0046] The specific number of decomposition layers is:

[0047] N = round(log2x)

[0048] Where N is the number of decomposition levels, round represents the rounding operation, and x is the target scale denominator / original scale denominator.

[0049] In this embodiment, according to the characteristics of the topography represented by the DEM data (such as Figure 3 ), select the biorthogonal wavelet function f = "bior3.9" with good symmetry and regularity; at the same time, according to the scale span of 4 times, determine the decomposition level N = 2 to ensure its adaptability and avoid the influence of a specific decomposition level on the data.

[0050] Furthermore, the method for calculating the curvature values ​​of all the position target points is:

[0051]

[0052] Where K is the curvature value of each target point, f x 、f y are the rates of change in the horizontal and vertical directions, respectively, and f xx 、f yy are the second-order derivatives in the horizontal and vertical directions, respectively, f xyis the mixed change rate in the horizontal and vertical directions, and slope is the slope.

[0053] Specifically, based on low-frequency data, specifically the low-frequency coefficients of 8 positions around the determined data, the curvature values ​​K of all position target points are calculated by the comprehensive curvature calculation formula, and then points greater than the curvature threshold T are selected as feature points according to the calculated curvature threshold T (T changes with the change of the scale span). In this embodiment, the calculated curvature threshold T=18 (T changes with the change of the scale span, and the final result retains one decimal place), and the feature points are selected.

[0054] Furthermore, after obtaining the low-frequency data after wavelet decomposition of the original data, the curvature values ​​K of all the target points are calculated using the comprehensive curvature calculation formula according to the coefficients of the selected 8 points.

[0055] Calculate the curvature value of each position in the low-frequency coefficient, calculate the curvature threshold T, select the point greater than the curvature threshold as the feature point, and first set the base curvature threshold A 基 =4.5, then calculate the curvature threshold T, and use the standardized formula shown below to calculate the curvature threshold:

[0056]

[0057] Where, T 阈 is the curvature threshold, A 基 is the base curvature threshold, D M is the target scale denominator, D R is the original scale denominator.

[0058] Furthermore, based on the feature points, new high-frequency coefficients are obtained through interpolation. Multiple interpolation methods are used, and their results are compared. The best interpolation method is selected as the final interpolation method. The specific operations are as follows:

[0059] Based on the selected feature points, interpolation processing is performed using a variety of interpolation methods, including Kriging interpolation, IDW interpolation, spline interpolation and natural neighbor interpolation in this embodiment. Several interpolation results are compared, and the natural neighbor interpolation method with the best effect is finally selected as the final interpolation method.

[0060] Furthermore, the method for calculating the standard deviation is:

[0061]

[0062] In the formula, σ is the standard deviation, S 基 is the benchmark standard deviation, D M is the target scale denominator, D R is the original scale denominator.

[0063] Specifically, in this embodiment, the standard deviation σ=0.60 (the scale span is 4 times), and then the low-frequency data is filtered to obtain new low-frequency data. The specific operation is as follows:

[0064] Set the base standard deviation S 基 = = 0.30, target scale denominator D M ==100000, original scale denominator D R =25000, the scale span is 4 times, and then the standard deviation σ is calculated to be 0.60 (the final result is rounded to two decimal places). The standard deviation is used to control the strength of the filter. The larger the standard deviation value, the stronger the processing effect of the filter. Conversely, the smaller the value, the weaker the processing effect.

[0065] The last layer of the low-frequency data after the filtering operation and the new high-frequency coefficients are reconstructed to obtain the reconstructed DEM data, and the reconstructed DEM data is subjected to elevation mapping to obtain the integrated DEM data. Figure 4 To display the DEM data in the future, Figure 5 This is a comparison chart of the original DEM data and the integrated DEM data contour stacking. In the experiment, the low-frequency data of each layer was selected for reconstruction. The experimental results show that when reconstructing, the low-frequency coefficient of the last layer is used, which has a better effect, so the last layer is selected.

[0066] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A DEM data automatic integration method based on wavelet analysis, characterized in that: include: According to the characteristics of the topography represented by the digital elevation model (DEM) data of the target area, the original data is processed by setting the wavelet function and decomposition layer number of the wavelet decomposition to obtain the low-frequency data after wavelet decomposition; Based on the low-frequency data, the curvature values ​​of all target points are calculated, and feature points are selected, and new high-frequency coefficients are obtained by interpolation method based on the feature points; Performing a filtering operation on the low-frequency data after the wavelet decomposition, and calculating a standard deviation, thereby obtaining the low-frequency data after the filtering operation; The last layer of the low-frequency data after the filtering operation and the new high-frequency coefficients are reconstructed to obtain reconstructed DEM data, and elevation mapping is performed on the reconstructed DEM data to obtain integrated DEM data.

2. The automatic DEM data integration method based on wavelet analysis according to claim 1 is characterized in that: Set the wavelet function and decomposition level of wavelet decomposition, including: According to the landform characteristics, wavelet functions with different advantages are selected, and the number of decomposition layers is determined; wherein the number of decomposition layers is specifically: N = round(log2x) Where N is the number of decomposition levels, round represents the rounding operation, and x is the target scale denominator / original scale denominator.

3. The automatic DEM data integration method based on wavelet analysis according to claim 1 is characterized in that: The method for calculating the curvature values ​​of all the position target points is: Where K is the curvature value of each target point, f x 、f y are the rates of change in the horizontal and vertical directions, respectively, and f xx 、f yy are the second-order derivatives in the horizontal and vertical directions, respectively, f xy is the mixed change rate in the horizontal and vertical directions, and slope is the slope.

4. The automatic DEM data integration method based on wavelet analysis according to claim 3 is characterized in that: Selecting the feature points includes: A curvature threshold is calculated, and a point whose curvature value is greater than the curvature threshold is selected as the feature point; wherein the curvature threshold is calculated by setting a reference curvature threshold.

5. The automatic DEM data integration method based on wavelet analysis according to claim 4 is characterized in that: The method for calculating the curvature threshold is: Where, T 阈 is the curvature threshold, A 基 is the base curvature threshold, D M is the target scale denominator, D R is the original scale denominator.

6. The automatic DEM data integration method based on wavelet analysis according to claim 1 is characterized in that: The method for calculating the standard deviation is: In the formula, σ is the standard deviation, S 基 is the benchmark standard deviation, D M is the target scale denominator, D R is the original scale denominator.

Images