A method for automatically identifying pasture land depressions based on a digital elevation model

Abstract

This invention provides an automatic identification method for pasture depressions and reservoirs based on digital elevation models (DEMs), comprising: acquiring topographic data of a target watershed using UAV-borne LiDAR technology to generate an original DEM of the target watershed; eliminating noise from the topographic data based on the DEM to generate a smoothed DEM_F and raster data of potential depression locations, DEM_EX; generating morphologically reconstructed elevation data based on DEM_F to obtain a residual DEM_F model of the target watershed, CS3; processing CS3 using a MATLAB regional connectivity analysis algorithm to identify regional connectivity and generate a preliminary identification result CS4 of embankments protruding from the surface; obtaining a final identification result CS7 of embankments based on CS4; and spatially overlaying analysis of CS7 and DEM_EX to output the three-dimensional coordinates of the depressions and reservoirs. This invention enables rapid and automatic identification of pasture depressions and reservoirs in semi-arid regions.

Description

Technical Field

[0001] This invention relates to the field of soil and water conservation technology, specifically a method for automatic identification of water storage ponds in pasture depressions based on digital elevation models. Background Technology

[0002] Pastureland is an important land resource in my country, mainly distributed in the semi-arid, arid, and high-altitude cold regions of western and northern my country. It serves as a crucial barrier for maintaining national ecological security and a key node for ecosystem protection and restoration. Simultaneously, pastureland provides basic production and livelihood resources for herders and is the main battleground for developing modern animal husbandry and promoting sustainable regional economic development. However, in some semi-arid and arid areas, insufficient rainfall has led to water shortages and degradation of water conservation functions in some pastureland, resulting in pasture degradation and a decline in its ecosystem services. To fully utilize rainfall resources and provide supplementary water for livestock, based on regional topographical features, herders have spontaneously constructed simple earthen dikes in some low-lying areas within the watershed. The depressions surrounded by these dikes have been transformed into livestock watering ponds (such as...). Figure 2 (As shown). Compared to specialized water storage and irrigation facilities, depression reservoirs are constructed on-site, resulting in lower costs. While storing rainwater resources, they also intercept eroded sediment from the slopes above the depression. Intercepting eroded sediment not only reduces watershed sediment flux but also, due to its higher nutrient content, further enhances the local habitat for depression vegetation, promoting regional ecological function. Therefore, depression reservoirs play a crucial role in the comprehensive utilization of pasture water resources and ecological restoration. However, due to the limited capacity of depressions, their relatively simple overall structure, and the lack of subsequent maintenance and dredging, depression reservoirs and their surrounding earthen embankments are easily damaged by rainfall, runoff, and human and animal disturbances. The sediment accumulated in the surrounding earthen embankments and depressions can easily become new sources of sediment, exacerbating regional soil erosion and ecological degradation (Nichols et al., 2018; Nichols et al., 2021). Therefore, they are a key target for precise prevention and control of soil erosion and subsequent design of soil and water conservation in the new era.

[0003] Accurately identifying the locations of depression-type water reservoirs scattered across vast pastures is a prerequisite for their subsequent management and maintenance, but it is also an extremely challenging task. The difficulties mainly lie in four aspects: First, there were no unified technical standards during the construction of depression-type water reservoirs, resulting in a degree of randomness in the height and shape of the earthen embankments surrounding the depressions; second, due to long-term operation, the height and shape of the earthen embankments surrounding existing depression-type water reservoirs have been further altered by rainfall, runoff, and human and livestock disturbances, exacerbating the difficulty of identifying depression-type water reservoirs through morphological parameters; third, the site selection of depression-type water reservoirs is largely based on herders' subjective judgment of the watershed topography and the water replenishment needs of livestock at different times, thus the location of the reservoirs is highly random; fourth, it is difficult to directly distinguish depression-type water reservoirs from natural depressions within the watershed, and only depressions with artificially constructed earthen embankments are considered depression-type water reservoirs.

[0004] Traditional methods for identifying the location of depression-shaped reservoirs mainly include field surveys and remote sensing visual interpretation. The former involves on-site investigation to record the location, shape, and operational status of the reservoirs; the latter uses remote sensing data to manually delineate existing depressions and reservoirs to determine their location. Both methods require significant manpower and resources, and have long identification cycles, making them unsuitable for effectively identifying and accurately locating scattered depressions in large areas of pasture. Furthermore, even with remote sensing interpretation, it is objectively difficult to determine whether the interpreted depression is naturally occurring or has been artificially manipulated. Therefore, a rapid and automated technology for identifying depression-shaped reservoirs is urgently needed. Summary of the Invention

[0005] The purpose of this invention is to provide a rapid and automatic method for identifying water storage ponds in grassland depressions in semi-arid regions, supporting the sustainable utilization of grassland water resources and watershed management in semi-arid regions.

[0006] The technical solution provided to achieve the above objectives is: an automatic identification method for water storage ponds in pasture depressions based on digital elevation models, comprising the following steps.

[0007] S10. Use UAV-borne LiDAR technology to collect topographic data of the target watershed and generate the original digital elevation model (DEM) of the target watershed.

[0008] S20. Based on the original digital elevation model (DEM) of the target watershed, the terrain data noise is eliminated by the spatial filtering algorithm of the MATLAB platform to generate a smooth digital elevation model (DEM_F). Furthermore, the raster data of potential depression locations in the target watershed (DEM_EX) is generated by the depression filling algorithm of the MATLAB platform.

[0009] S30. Based on the smoothed digital elevation model DEM_F, construct planar disk-shaped structural elements and erode watershed elevation data to generate eroded elevation data. Then, use the watershed elevation data and eroded elevation data to generate morphological reconstruction elevation data, and further obtain the target watershed residual digital elevation model CS3.

[0010] S40. Use the MATLAB regional connectivity analysis algorithm to process the residual digital elevation model CS3 of the target watershed, identify regional connected components, and combine the preset connected component length threshold to screen potential embankment areas, generating the preliminary identification result CS4 of the embankment protruding from the ground surface.

[0011] S50. Based on the preliminary identification result CS4 of the embankment protruding from the ground surface, the final identification result CS7 of the embankment is obtained by using the MATLAB regional connectivity analysis algorithm and setting the image boundary threshold, embankment length threshold and embankment eccentricity threshold.

[0012] S60. Spatial overlay analysis of the final identification result CS7 of the embankment and the raster data DEM_EX of the potential depression location in the area, outputting the three-dimensional coordinates of the depression reservoir.

[0013] Furthermore, step S20 specifically includes:

[0014] S21. In the MATLAB platform, call the filter function to perform convolution operation on the original digital elevation model (DEM) through a 3-cell × 3-cell moving window to eliminate high-frequency noise components in the terrain data and generate target watershed smooth terrain data DEM_F with continuous smooth features.

[0015] S22. Based on the target watershed smoothed terrain data DEM_F, call the fillsinks function to obtain the target watershed filled depression terrain data DEM_FF;

[0016] S23. Subtract the smoothed terrain data DEM_F from the depression-filling terrain data DEM_FF of the target watershed to obtain the raster data DEM_EX of potential depression locations in the target watershed.

[0017] Furthermore, step S30 specifically includes:

[0018] S31. Using the MATLAB platform, store the terrain elevation in the target watershed smooth terrain data DEM_F in matrix form, denoted as CS1.

[0019] S32. Using the strel function in the MATLAB platform, construct a planar disk-shaped structuring element, denoted as SE;

[0020] S33. In the MATLAB platform, call the imerode function to erode the elevation matrix CS1 using the planar disk-shaped structuring element SE to obtain the eroded elevation data CSmaker.

[0021] S34. Using the MATLAB platform, based on the corrosion elevation data CSmaker and the elevation matrix CS1, call the imreconstruct function to construct the morphological reconstruction elevation data CS2;

[0022] S35. Using the MATLAB platform, subtract the morphologically reconstructed elevation data CS2 from the elevation data CS1 to obtain the target watershed residual digital elevation model CS3.

[0023] Furthermore, step S40 specifically includes:

[0024] S41. In the MATLAB platform, call the bwconncomp function to identify the connected region in the residual digital elevation model CS3 of the target watershed, denoted as CSC1;

[0025] S42. In the MATLAB platform, call the `regionprops` function to count the major axis of the ellipse of each connected component in the connected region CSC1. Number the connected components as 1, 2, 3, ... N, and denote their corresponding major axes as L1, L2, L3, ... L N Data storage, denoted as CSstats;

[0026] S43. Using the MATLAB platform, set the threshold of the major axis of the ellipse of the earthen embankment, call the ismember function to exclude connected components in the connected region CSC1 whose major axis of the ellipse is smaller than the set threshold, store the remaining connected components, and obtain the preliminary identification image CS4 of the earthen embankment protruding from the ground surface.

[0027] Furthermore, step S50 specifically includes:

[0028] S51. Using the MATLAB platform, set the image boundary threshold, exclude the connected components located around the image in the preliminary identification result CS4 of the embankment protruding from the ground surface, store the position information of the remaining connected components, and obtain the image CS5 of the embankment protruding from the ground surface.

[0029] S52. Using the MATLAB platform, based on the image CS5 of the earthen embankment protruding from the ground surface, call the imfill function to fill the connected voids in the earthen embankment image CS5 to obtain the image CS6 of the earthen embankment protruding from the ground surface.

[0030] S53. Using the MATLAB platform, based on the image CS6 of the earthen embankment protruding from the ground surface, call bwconncomp to calculate the connected region CSC2 in the image CS6 of the earthen embankment protruding from the ground surface.

[0031] S54. In the MATLAB platform, call the `regionprops` function to calculate the major axis and eccentricity of the ellipse for each connected component in the connected region CSC2. Number the connected components as 1, 2, 3, ... N, and denote their corresponding major axes as L1, L2, L3, ... L N The corresponding eccentricities of the ellipse are F1, F2, F3, ... F N The data is stored and denoted as CSstats2;

[0032] S55. Using the MATLAB platform, set the threshold for the major axis of the ellipse and the threshold for the eccentricity of the ellipse. Call the ismember function to exclude connected components in the connected region CSC2 whose major axis of the ellipse is less than the set threshold and whose eccentricity of the ellipse is less than the set threshold. That is, exclude the embankments that protrude from the ground in a straight line shape. Store the remaining connected components to obtain the final image CS7 of the embankments that protrude from the ground.

[0033] Furthermore, step S60 specifically includes:

[0034] S61. In the MATLAB platform, call the bwconncomp function to identify the connected region in the image CS7 of the earthen embankment protruding from the ground surface, and denote it as CSC3;

[0035] S62. In the MATLAB platform, call the labelmatrix function to number each independent connected body in the connected region CSC3 as 1, 2, 3, ... N, and record its corresponding elevation as 1, 2, 3, ... N. Store the data as CS7_label.

[0036] S63. Replace the elevation matrix in the original digital elevation model (DEM) with CS7_lable to obtain the newly created digital elevation model (DEM_new);

[0037] S64. Based on the newly created digital elevation model DEM_new, call the GRIDobj2polygon function to obtain the boundary coordinates S of the embankment;

[0038] S65. Spatial overlay analysis of the boundary coordinates S of the earthen embankment and the raster data DEM_EX of the potential depression location in the region, outputting the three-dimensional coordinates of the depression reservoir.

[0039] The advantage of this invention lies in its approach: compared to traditional methods of manually locating water-retaining depressions, this invention utilizes the MATLAB platform to focus on the key characteristic of the elliptical shape of the surrounding earthen embankments. First, it identifies easily identifiable embankments protruding from the ground. Then, by determining whether the embankment's shape is linear or elliptical, it determines whether the surrounding area is a water-retaining depression. By determining the presence of an embankment and whether it exhibits an elliptical shape, it can directly locate artificially induced water-retaining depressions, avoiding the identification of natural depressions rather than artificially disturbed ones. Attached Figure Description

[0040] Figure 1 This is a flowchart of an automatic identification method for water storage ponds in pasture depressions based on digital elevation models according to an embodiment of the present invention;

[0041] Figure 2 This is a schematic diagram of the water storage depressions to be identified within the watershed according to an embodiment of the present invention;

[0042] Figure 3 This is a schematic diagram of the smoothed digital elevation model DEM_F according to an embodiment of the present invention;

[0043] Figure 4 This is a schematic diagram of the target watershed residual digital elevation model CS3 according to an embodiment of the present invention;

[0044] Figure 5 This is a schematic diagram of the preliminary identification result CS4 of an embodiment of the present invention;

[0045] Figure 6 This is a schematic diagram of the final identification result CS7 of the earthen embankment in an embodiment of the present invention;

[0046] Figure 7 This is a schematic diagram of the output of the identified depression coordinates in an embodiment of the present invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] This invention utilizes the MATLAB platform and, based on terrain data, can batch, quickly, accurately, and automatically identify water-retaining depressions in pastureland, outputting their coordinate locations. Furthermore, the identification process eliminates the need to save intermediate calculation data, thus saving hard drive storage space. The key to this invention lies in the presence of simple earthen embankments surrounding the water-retaining depressions, with these embankments exhibiting an elliptical shape. If an identified depression in a region is surrounded by an earthen embankment, it is definitely a water-retaining depression that has undergone human intervention, rather than a natural depression. The eccentricity of an object is the ratio of its focal length to its major axis when the object is approximated as an ellipse. Eccentricity generally falls between 0 and 1; an object with an eccentricity of 0 can be considered a circle, while an object with an eccentricity of 1 can be considered a straight line segment. Therefore, if the eccentricity of an earthen embankment found in a certain region is less than a certain threshold, the embankment exhibits an elliptical shape, and the area surrounded by the embankment is a water-retaining depression. In other words, instead of directly searching for water-retaining depressions, we first look for more easily identifiable earthen embankments that protrude from the ground, and then determine whether the area around the embankment is a water-retaining depression by judging whether the embankment is straight or elliptical.

[0049] like Figure 1 As shown, this embodiment of the invention takes a specific target small watershed as an example. Figure 2 This paper provides an automatic identification method for water-retaining depressions in pastureland based on a digital elevation model, comprising the following steps:

[0050] S10. Use UAV-borne LiDAR technology to acquire raw digital elevation model (DEM) data. The data acquisition in step S10 uses UAV LiDAR technology, including flight path design, point cloud data acquisition, trajectory calculation, point cloud fusion and coordinate transformation, point cloud filtering and construction of digital elevation model (DEM).

[0051] S20. Based on the original digital elevation model (DEM) of the target watershed, the terrain data noise is eliminated by the spatial filtering algorithm of the MATLAB platform to generate a smooth digital elevation model (DEM_F). Furthermore, the potential depression location raster data (DEM_EX) of the target watershed is generated by the depression filling algorithm of the MATLAB platform.

[0052] Step S20 specifically includes:

[0053] S21. In the MATLAB platform, call the filter function to perform convolution operations on the original digital elevation model (DEM) using a 3-cell × 3-cell moving window to eliminate high-frequency noise components in the terrain data and generate target watershed smoothed terrain data DEM_F with continuous smoothness characteristics. Figure 3 ).

[0054] S22. Based on the target watershed smoothed terrain data DEM_F, call the fillsinks function to obtain the target watershed filled depression terrain data DEM_FF;

[0055] S23. Subtract the smoothed terrain data DEM_F from the depression-filling terrain data DEM_FF of the target watershed to obtain the raster data DEM_EX of potential depression locations in the target watershed.

[0056] S31. Using the MATLAB platform, store the terrain elevation in the target watershed smooth terrain data DEM_F in matrix form, denoted as CS1.

[0057] S32. In the MATLAB platform, call the strel function to construct a planar disk-shaped structuring element, denoted as SE;

[0058] S33. In the MATLAB platform, call the imerode function to erode the elevation matrix CS1 using the planar disk-shaped structuring element SE to obtain the eroded elevation data CSmaker.

[0059] S34. Using the MATLAB platform, based on the corrosion elevation data CSmaker and the elevation matrix CS1, call the imreconstruct function to construct the morphological reconstruction elevation data CS2;

[0060] S35. Using the MATLAB platform, subtract the morphologically reconstructed elevation data CS2 from the elevation data CS1 to obtain the target watershed residual digital elevation model CS3. Figure 4 ).

[0061] S40. Using the MATLAB regional connectivity analysis algorithm to process the residual digital elevation model CS3 of the target watershed, identify regional connected components, and combine with the preset connected component length threshold to screen potential embankment areas, generating preliminary identification results of embankments protruding from the ground surface CS4.

[0062] Step S40 specifically includes:

[0063] S41. In the MATLAB platform, call the bwconncomp function to identify the connected region in the residual digital elevation model CS3 of the target watershed, denoted as CSC1;

[0064] S42. In the MATLAB platform, call the `regionprops` function to count the major axis of the ellipse of each connected component in the connected region CSC1. Number the connected components as 1, 2, 3, ... N, and denote their corresponding major axes as L1, L2, L3, ... L N Data storage, denoted as CSstats;

[0065] S43. Using the MATLAB platform, set the threshold for the major axis of the ellipse of the embankment, call the ismember function to exclude connected components in the connected region CSC1 whose major axis is smaller than the set threshold, store the remaining connected components, and obtain the preliminary identification image CS4 of the embankment protruding from the ground surface. Figure 5 ).

[0066] S50. Based on the preliminary identification result CS4 of the embankment protruding from the ground surface, the final identification result CS7 of the embankment is obtained by using the MATLAB regional connectivity analysis algorithm and setting the image boundary threshold, embankment length threshold and embankment eccentricity threshold.

[0067] Step S50 specifically includes:

[0068] S51. Using the MATLAB platform, set the image boundary threshold, exclude the connected components located around the image in the preliminary identification result CS4 of the embankment protruding from the ground surface, store the position information of the remaining connected components, and obtain the image CS5 of the embankment protruding from the ground surface.

[0069] S52. Using the MATLAB platform, based on the image CS5 of the earthen embankment protruding from the ground surface, call the imfill function to fill the connected voids in the earthen embankment image CS5 to obtain the image CS6 of the earthen embankment protruding from the ground surface.

[0070] S53. Using the MATLAB platform, based on the image CS6 of the earthen embankment protruding from the ground surface, call bwconncomp to calculate the connected region CSC2 in the image CS6 of the earthen embankment protruding from the ground surface.

[0071] S54. In the MATLAB platform, call the `regionprops` function to calculate the major axis and eccentricity of the ellipse for each connected component in the connected region CSC2. Number the connected components as 1, 2, 3, ... N, and denote their corresponding major axes as L1, L2, L3, ... L N The corresponding eccentricities of the ellipse are F1, F2, F3, ... F N The data is stored and denoted as CSstats2;

[0072] S55. Using the MATLAB platform, set the ellipse major axis threshold and the ellipse eccentricity threshold, and call the ismember function to exclude connected components in the connected region CSC2 whose ellipse major axis is smaller than the set major axis threshold and whose ellipse eccentricity is smaller than the set eccentricity threshold. That is, exclude the embankments that protrude from the ground surface in a straight line shape. Store the remaining connected components to obtain the final image CS7 of the identified embankments protruding from the ground surface. Figure 6 ).

[0073] S60. Spatial overlay analysis of the final identification result CS7 of the embankment and the raster data DEM_EX of the potential depression location in the area, outputting the three-dimensional coordinates of the depression reservoir.

[0074] Step S60 specifically includes:

[0075] S61. In the MATLAB platform, call the bwconncomp function to identify the connected region in the image CS7 of the earthen embankment protruding from the ground surface, and denote it as CSC3;

[0076] S62. In the MATLAB platform, call the labelmatrix function to number each independent connected body in the connected region CSC3 as 1, 2, 3, ... N, and record its corresponding elevation as 1, 2, 3, ... N. Store the data as CS7_label.

[0077] S63. Replace the elevation matrix in the original digital elevation model (DEM) with CS7_lable to obtain the newly created digital elevation model (DEM_new);

[0078] S64. Based on the newly created digital elevation model DEM_new, call the GRIDobj2polygon function to obtain the boundary coordinates S of the embankment;

[0079] S65. Spatial overlay analysis of the earthen embankment boundary coordinates S and the raster data DEM_EX of the potential depression location in the region, outputting the three-dimensional coordinates of the water storage depression. Figure 7 ).

[0080] This invention can identify water-retaining depressions in pastureland in a batch, quickly, accurately, and automatically, and output their coordinate locations. Taking the above example, this invention can accurately locate three water-retaining depressions within a 2.4km × 4.2km area in 15.4 seconds, with 100% accuracy. If these three water-retaining depressions were identified manually using ArcGIS, the time required to output their coordinates would be approximately 25 minutes. Furthermore, the identification process does not require saving intermediate calculation data, thus saving hard drive storage space. In the above example, if manual interpretation were used, the location of the depressions would first need to be located, then a vector file of the depression locations would need to be constructed, edited, and then saved. This invention, however, can directly output the depression locations in the final step, without saving any intermediate process data. The identification results can support the sustainable utilization of water resources in pastureland and watershed management in semi-arid regions.

[0081] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for automatic identification of water storage ponds in pasture depressions based on digital elevation models, characterized in that, Includes the following steps: S10. Use UAV-borne LiDAR technology to collect topographic data of the target watershed and generate the original digital elevation model (DEM) of the target watershed. S20. Based on the original digital elevation model (DEM) of the target watershed, the terrain data noise is eliminated by the spatial filtering algorithm of the MATLAB platform to generate a smooth digital elevation model (DEM_F). Furthermore, the raster data of potential depression locations in the target watershed (DEM_EX) is generated by the depression filling algorithm of the MATLAB platform. S30. Based on the smoothed digital elevation model DEM_F, construct planar disk-shaped structural elements and erode watershed elevation data to generate eroded elevation data. Then, use the watershed elevation data and eroded elevation data to generate morphological reconstruction elevation data, and further obtain the target watershed residual digital elevation model CS3. S40. Using the MATLAB regional connectivity analysis algorithm to process the residual digital elevation model CS3 of the target watershed, identify regional connected components, and combine with the preset connected component length threshold to screen potential embankment areas, generating preliminary identification results CS4 for embankments protruding from the ground surface; S50. Based on the preliminary identification result CS4 of the embankment protruding from the ground surface, the final identification result CS7 of the embankment is obtained by using the MATLAB regional connectivity analysis algorithm and setting the image boundary threshold, embankment length threshold and embankment eccentricity threshold. S60. Using the MATLAB platform, spatially overlay analysis is performed on the final identification result CS7 of the earthen embankment and the raster data DEM_EX of the potential depression location in the area, and the three-dimensional coordinates of the water storage depression are output.

2. The automatic identification method for water storage ponds in pasture depressions based on digital elevation models as described in claim 1, characterized in that: Step S20 specifically includes: S21. In the MATLAB platform, call the filter function to perform convolution operation on the original digital elevation model (DEM) through a 3-cell × 3-cell moving window to eliminate high-frequency noise components in the terrain data and generate target watershed smooth terrain data DEM_F with continuous smooth features. S22. Based on the smoothed terrain data DEM_F of the target watershed, call the fillsinks function to obtain the depression-filled terrain data DEM_FF of the target watershed; S23. Subtract the smoothed terrain data DEM_F from the depression-filling terrain data DEM_FF of the target watershed to obtain the raster data DEM_EX of potential depression locations in the target watershed.

3. The automatic identification method for water storage ponds in pasture depressions based on digital elevation models as described in claim 1, characterized in that: Step S30 specifically includes: S31. Using the MATLAB platform, store the terrain elevations in the smoothed terrain data DEM_F of the target watershed in matrix form, denoted as CS1; S32. Using the strel function in the MATLAB platform, construct a planar disk-shaped structuring element, denoted as SE; S33. In the MATLAB platform, call the imerode function to erode the elevation matrix CS1 using the planar disk-shaped structuring element SE to obtain the eroded elevation data CSmaker. S34. Using the MATLAB platform, based on the corrosion elevation data CSmaker and the elevation matrix CS1, call the imreconstruct function to construct the morphological reconstruction elevation data CS2; S35. Using the MATLAB platform, subtract the morphologically reconstructed elevation data CS2 from the elevation matrix CS1 to obtain the target watershed residual digital elevation model CS3.

4. The method for automatic identification of water storage ponds in pasture depressions based on digital elevation models as described in claim 1, characterized in that: Step S40 specifically includes: S41. In the MATLAB platform, call the bwconncomp function to identify the connected region in the residual digital elevation model CS3 of the target watershed, denoted as CSC1; S42. In the MATLAB platform, call the `regionprops` function to count the major axis of the ellipse of each connected component in the connected region CSC1. Number the connected components as 1, 2, 3, ... N, and denote their corresponding major axes as L1, L2, L3, ... L N Data storage, denoted as CSstats; S43. Using the MATLAB platform, set the threshold of the major axis of the ellipse of the earthen embankment, call the ismember function to exclude connected components in the connected region CSC1 whose major axis of the ellipse is smaller than the set threshold, store the remaining connected components, and obtain the preliminary identification image CS4 of the earthen embankment protruding from the ground surface.

5. The automatic identification method for water storage ponds in pasture depressions based on digital elevation models as described in claim 1, characterized in that: Step S50 specifically includes: S51. Using the MATLAB platform, set the image boundary threshold, exclude the connected components located around the image in the preliminary identification result CS4 of the embankment protruding from the ground surface, store the position information of the remaining connected components, and obtain the image CS5 of the embankment protruding from the ground surface. S52. Using the MATLAB platform, based on the image CS5 of the earthen embankment protruding from the ground surface, call the imfill function to fill the connected voids in the earthen embankment image CS5 to obtain the image CS6 of the earthen embankment protruding from the ground surface. S53. Using the MATLAB platform, based on the image CS6 of the earthen embankment protruding from the ground surface, call bwconncomp to calculate the connected region CSC2 in the image CS6 of the earthen embankment protruding from the ground surface. S54. In the MATLAB platform, call the `regionprops` function to calculate the major axis and eccentricity of the ellipse for each connected component in the connected region CSC2. Number the connected components as 1, 2, 3, ... N, and denote their corresponding major axes as L1, L2, L3, ... L N The corresponding eccentricities of the ellipse are F1, F2, F3, ... F N Data storage, denoted as CSstats2; S55. Using the MATLAB platform, set the threshold for the major axis of the ellipse and the threshold for the eccentricity of the ellipse. Call the ismember function to exclude connected components in the connected region CSC2 whose major axis of the ellipse is less than the set threshold and whose eccentricity of the ellipse is less than the set threshold. That is, exclude the embankments that protrude from the ground in a straight line shape. Store the remaining connected components to obtain the final image CS7 of the embankments that protrude from the ground.

6. The method for automatic identification of water storage ponds in pasture depressions based on digital elevation models as described in claim 1, characterized in that: Step S60 specifically includes: S61. Call the bwconncomp function in the MATLAB platform to identify the connected region in the final identification result CS7 of the embankment, and denote it as CSC3; S62. In the MATLAB platform, call the labelmatrix function to number each independent connected body in the connected region CSC3 as 1, 2, 3, ... N, and record its corresponding elevation as 1, 2, 3, ... N. Store the data as CS7_label. S63. Replace the elevation matrix in the original digital elevation model (DEM) with CS7_lable to obtain the newly created digital elevation model (DEM_new); S64. Based on the newly created digital elevation model DEM_new, call the GRIDobj2polygon function to obtain the boundary coordinates S of the embankment; S65. Spatial overlay analysis of the earthen embankment coordinate boundary S and the raster data DEM_EX of the potential depression location in the region, outputting the three-dimensional coordinates of the depression reservoir.

Images