A method for batch extraction of lidar reference depth control points

Abstract

This paper discloses a method for batch extracting reference depth control points from LiDAR. This method first uses high-resolution remote sensing imagery to rapidly extract shallow-water characteristic photons. It then applies the DBSCAN algorithm to detect underwater photon signals, eliminating some outliers. Finally, it rapidly obtains ICESat-2 shallow-water reference depth control point data. This method can rapidly extract photon signals from shallow-water bottom topography detected by ICESat-2 photons, improving the efficiency of LiDAR echo detection of bottom topography.

Description

Technical Field

[0001] The present invention belongs to the field of laser radar signal processing, and in particular relates to a method for batch extraction of laser radar reference water depth control points. Background Art

[0002] Coastal water mapping provides crucial information for water resource management, governance, and environmental protection. Over the past few years, there has been a growing demand for updated and detailed bathymetric information for shallow waters, requiring higher temporal and spatial coverage and sufficient vertical and horizontal accuracy. However, approximately 70-80% of the world's coastal areas still lack accurate bathymetric maps.

[0003] While traditional depth detection methods and specialized equipment (such as LiDAR and SoNAR) offer high measurement accuracy, they require long measurement times and high investment. Furthermore, data collection in shallow water environments has certain limitations, resulting in relatively limited commercialization of traditional depth detection methods. As an alternative, satellite-based bathymetry has been used since at least the 1970s to estimate water depth in relatively shallow waters. It offers the advantages of low cost, ease of implementation, and wide coverage. Hyperspectral remote sensing, a type of passive remote sensing technology, can retrieve water depth by establishing a relationship between water depth and spectral radiance across various bands. However, empirical bathymetry methods using multispectral imagery typically rely on in situ depths as control points, severely limiting their spatial application. The ICESat-2 / ATLAS data product ATL03 is a large dataset. Based on ICESat-2 / ATLAS data, water depth in coastal areas can be measured directly or indirectly. The large amount of raw photon signals presents significant challenges in effectively extracting seafloor reflection signals. Furthermore, the raw photons in ATL03 contain significant noise. The density-based spatial clustering of noise applications (DBSCAN) algorithm has been proven to be an effective photon signal processing method. However, to ensure extraction accuracy, the traditional reference depth control point extraction method uses a fixed-parameter DBSCAN method for extracting seafloor topography along a single satellite route. This method is inefficient and difficult to apply to large-scale applications, while subsequent depth inversion often requires a large number of depth control points. Therefore, it is necessary to adopt an effective signal processing method to quickly obtain a large number of bottom-reflected photon signals from the original ATL03 photon point cloud as a reference depth control point dataset for subsequent wide-area depth inversion. Summary of the Invention

[0004] In order to remove noise points in the laser radar echo signal for detecting seabed topography, the present invention provides a method for batch extraction of laser radar reference water depth control points.

[0005] The purpose of the present invention is achieved through the following technical solutions:

[0006] S1: Input the original Sentinel-2 image of the island and reef area to be processed, process it, and then perform land and sea separation based on the processed high-resolution Sentinel-2 image of the island and reef to extract the shallow sea contour data of the island and reef;

[0007] S2: Input ICESat-2 ATL03 data, calculate the intersection of the route and the contour, and obtain the ICESat-2 ATL03 shallow sea feature dataset;

[0008] S3: Segment the ICESat-2 ATL03 shallow sea feature dataset and input it into the DBSCAN algorithm to detect seabed echo signals;

[0009] S4: The segmented seafloor echo signal data extracted from S4 is filtered and the data segments are merged to obtain the ICESat-2 reference water depth control point set;

[0010] Furthermore, the method for batch extraction of laser radar reference water depth control points according to claim 1 is characterized in that the specific processing flow for extracting the shallow sea contour data of islands and reefs in S1 is to perform atmospheric correction on the image, crop the image, remove clouds, remove land, separate land and sea based on NDWI, classify the island and reef images and water bodies, remove holes in the island and reef images, and output the shallow sea contour data of the island and reefs in the format of a .shp file.

[0011] Furthermore, the method for batch extraction of lidar reference water depth control points according to claim 1 is characterized in that the specific process of obtaining the ICESat-2 ATL03 shallow water feature dataset in S2 is:

[0012] 1) Input ICEsat-2 ATL03 data passing through the island and reef area and the island and reef shallow water contour data obtained in S1;

[0013] 2) Obtain the data segment of the ATL03 passing through the shallow sea contour data of islands and reefs;

[0014] 3) Determine the direction of the island, reef, and land relative to each data segment;

[0015] 4) For the data segments passing through the shallow sea contour data of islands and reefs, is the land length less than 2000m?

[0016] 5) If it is less than 2000m, then along the along-track direction, for each intersection point of ICEsat-2 data and island reef shallow water contour data, the starting point is the intersection point of ICEsat-2 data and island reef shallow water contour data, extending 1000m in the land direction, and the end point is the intersection point of ICEsat-2 data and island reef shallow water contour data, extending 5000m away from the land direction. The ICESat-2 ATL03 data segment between these two points is the ICESat-2 ATL03 shallow water feature data segment;

[0017] 6) If it is greater than 2000m, then along the along-track direction, for the two intersection points of the ICEsat-2 data and the island and reef shallow water contour data in this segment, the starting point is the intersection point A of the ICEsat-2 data and the island and reef shallow water contour data, extending 5000m away from the land direction, and the end point is the intersection point B of the ICEsat-2 data and the island and reef shallow water contour data, extending 5000m away from the land direction. The ICESat-2 ATL03 data segment between these two points is the ICESat-2 ATL03 shallow water feature data segment;

[0018] 7) Traverse each route of ICEsat-2 and merge to obtain the ICESat-2 ATL03 shallow water feature dataset.

[0019] Furthermore, the method for batch extraction of laser radar reference water depth control points according to claim 1 is characterized in that the specific method of data screening in S4 is to calculate three times the mean square error of the data and eliminate data greater than three times the mean square error.

[0020] The beneficial effects of the present invention are: it can overcome the problem that the number of ICESat-2 ATL03 original photons is huge and the shallow sea area has complex terrain. While ensuring detection accuracy, it can quickly collect an effective ICESat-2 reference water depth control point set, improve the efficiency of control point extraction, and thus can quickly and accurately invert the shallow sea area terrain. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flow chart of the batch extraction method of the laser radar reference water depth control points of the present invention.

[0022] Figure 2 It is a flow chart of the S2 part of the present invention.

[0023] Figure 3 This is a diagram of extraction results at reference water depth control points on Mayaguana Island after adopting the method of the present invention.

[0024] Figure 4Figure 1 shows the detection results of the reference water depth control points between 22.327°S and 22.452°S in the ICESat-2 ATL03_20190105043149_01180207_003_01.h5 gt1l route. Figure (a) shows the original signal of the ICESat-2 ATL03_20190105043149_01180207_003_01.h5 gt1l at this location. Figure (b) shows the detection results of the reference water depth control points, where * indicates that a reference water depth control point has been detected.

[0025] Figure 5 These are the detection results of the reference water depth control points between latitudes 22.295°S and 22.354°S in the ICESat-2 ATL03_20210220031315_08881001_004_01.h5 gt3r route. Figure (a) is the original signal of the ICESat-2ATL03_20210220031315_08881001_004_01.h5 gt3r at this location, and Figure (b) is the detection result of the reference water depth control points. DETAILED DESCRIPTION

[0026] The present invention will be described in detail below based on the accompanying drawings and preferred embodiments. The purpose and effects of the present invention will become more apparent. The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention.

[0027] like Figure 1 As shown, the method for batch extraction of laser radar reference water depth control points of the present invention includes the following steps:

[0028] Step 1: Obtain shallow sea contour data of islands and reefs from the original high-resolution passive satellite remote sensing image of the island and reef area;

[0029] As one of the implementation methods, atmospheric correction can be performed on the original high-resolution passive satellite remote sensing images of the island and reef areas in sequence, and then the images can be cropped to obtain the precise impact of the island and reef areas. Cloud cover and land parts can then be removed, and then land and sea separation can be performed to obtain a complete island and reef image. The island and reef images can be classified from water bodies, and the holes inside the islands and reefs in the island and reef images can be removed to finally obtain the shallow sea contour data of the islands and reefs.

[0030] Step 2: Calculate the intersection of the ICESat-2 ATL03 data flight path and the shallow water contour data of the islands and reefs obtained in step 1 to obtain the ICESat-2 ATL03 shallow water feature dataset; Figure 2 As shown, it specifically includes the following sub-steps:

[0031] (1) Based on the ICESat-2 ATL03 shallow water feature data and the shallow water contour data of islands and reefs in S1, the data segment of the ICESat-2 ATL03 shallow water feature data passing through the shallow water contour data of islands and reefs is obtained;

[0032] (2) Determine the direction of the island, reef, and land relative to each data segment;

[0033] (3) For the data segment passing through the shallow sea contour data of islands and reefs, is the land length less than 2000m?

[0034] If it is greater than 2000m, then along the along-track direction, for the two intersection points of the ICEsat-2 data and the island-reef shallow-water contour data, the starting point is the intersection point of the ICEsat-2 data and the island-reef shallow-water contour data, extending 1000m inland, and the end point is the intersection point of the ICEsat-2 data and the island-reef shallow-water contour data, extending 5000m away from the land. The ICESat-2ATL03 data segment between the starting point and the end point is the ICESat-2 ATL03 shallow-water feature data segment;

[0035] If it is less than 2000m, then along the along-track direction, for the two intersection points of ICEsat-2 data and island and reef shallow water contour data in this segment, the starting point is the first intersection point of ICEsat-2 data and island and reef shallow water contour data, extending 5000m away from the land direction, and the end point is the other intersection point of ICEsat-2 data and island and reef shallow water contour data, extending 5000m away from the land direction. The ICESat-2 ATL03 data segment between the starting point and the end point is the ICESat-2 ATL03 shallow water feature data segment;

[0036] (4) Traverse each route of ICEsat-2 and merge to obtain the ICESat-2 ATL03 shallow sea feature dataset.

[0037] Step 3: Segment the ICESat-2 ATL03 shallow water feature dataset and input it into the DBSCAN algorithm to detect the seabed echo signal. This can avoid processing the entire lidar track, reduce the amount of calculation and improve detection efficiency.

[0038] Step 4: The segmented seabed echo signals extracted in step 3 are screened, and after removing data anomalies, all data are merged into one data set to obtain the ICESat-2 bathymetric control point set.

[0039] The specific method of removing data outliers in step 4 is:

[0040] Calculate the three times mean error of the data and eliminate the data that is larger than three times the mean error.

[0041] The effect of the method of the present invention is described below with reference to a specific embodiment.

[0042] This embodiment is based on ICESat-2 ATL03_20190105043149_01180207_003_01.h5, ATL03_20210220031315_08881001_004_01.h5, ATL03_20210423001702_04461101_004_01.h5, ATL03_20210521225306_08881101_004_01.h5 original data. Figure 3 This is a schematic diagram of the results of extracting reference water depth control points in the Mayaguana Island area using the method described in the present invention. Figure 3 This is an image of the Mayaguana Island basemap. The horizontal axis is longitude and the vertical axis is latitude. The points in the image passing through Mayaguana Island are detected ICESat-2 reference water depth control points. The color of the points corresponds to the water depth at that point. The bar chart below the image shows the water depth of the reference water depth control points shown in the image. Figure 4 The ICESat-2 ATL03_20190105043149_01180207_003_01.h5 gt1l route shows the reference water depth control point detection results from 22.327°S to 22.452°S. Figure 4 The left picture is the original ICESat-2ATL03_20190105043149_01180207_003_01.h5 gt1l original signal at this location. Figure 4 The right picture shows the detection result of the reference water depth control point, and * indicates that the reference water depth control point has been detected. Figure 5 The ICESat-2ATL03_20210220031315_08881001_004_01.h5 gt3r route shows the reference water depth control point detection results from 22.295°S to 22.354°S. Figure 5 The left picture is the original ICESat-2ATL03_20210220031315_08881001_004_01.h5 gt3r original signal at this location. Figure 5 The right picture shows the detection result of the reference water depth control point. * indicates that the reference water depth control point is detected. It can be seen that this method can detect the seabed underwater terrain echo signal very well and has good detection accuracy.

[0043] Those skilled in the art will understand that the foregoing descriptions are merely preferred embodiments of the invention and are not intended to limit the invention. Although the invention has been described in detail with reference to the foregoing examples, those skilled in the art will still be able to modify the technical solutions described in the foregoing examples or substitute equivalents for some of the technical features therein. Any modifications, equivalent substitutions, etc. made within the spirit and principles of the invention shall be included within the scope of protection of the invention.

Claims

1. A method for batch extraction of laser radar reference water depth control points, characterized in that: The method comprises the following steps: S1: Obtain shallow sea contour data of islands and reefs by performing atmospheric correction, image cropping, cloud removal, land removal, land-sea separation, classification of island and reef images and water bodies, and removal of holes in island and reef images on the original high-resolution passive satellite remote sensing images of the island and reef area. S2: Calculate the intersection of the ICESat-2 ATL03 data flight path and the shallow water contour data of the islands and reefs obtained in step 1 to obtain the ICESat-2 ATL03 shallow water feature dataset. The specific process is as follows: (1) Based on the ICESat-2 ATL03 shallow water feature data and the shallow water contour data of islands and reefs in S1, the data segment of the ICESat-2 ATL03 shallow water feature data passing through the shallow water contour data of islands and reefs is obtained; (2) Determine the direction of the island, reef, and land relative to each data segment; (3) For the data segment passing through the shallow sea contour data of islands and reefs, is the land length less than 2000m? If it is greater than 2000m, then along the along-track direction, for the two intersection points of the ICEsat-2 data and the island-reef shallow-water contour data, the starting point is the intersection point of the ICEsat-2 data and the island-reef shallow-water contour data, extending 1000m in the land direction, and the end point is the intersection point of the ICEsat-2 data and the island-reef shallow-water contour data, extending 5000m away from the land direction. The ICESat-2 ATL03 data segment between the starting point and the end point is the ICESat-2 ATL03 shallow-water feature data segment; If it is less than 2000m, then along the along-track direction, for the two intersection points of the ICEsat-2 data and the island and reef shallow water contour data in this data segment, the starting point is the first intersection point of the ICEsat-2 data and the island and reef shallow water contour data, extending 5000m away from the land direction, and the end point is the other intersection point of the ICEsat-2 data and the island and reef shallow water contour data, extending 5000m away from the land direction. The ICESat-2 ATL03 data segment between the starting point and the end point is the ICESat-2 ATL03 shallow water feature data segment; (4) Traverse each route of ICEsat-2 and merge to obtain the ICESat-2 ATL03 shallow sea feature dataset; S3: Segment the ICESat-2 ATL03 shallow sea feature dataset and input it into the DBSCAN algorithm to detect the seabed echo signal to obtain the seabed echo signal; S4: The segmented seabed echo signals extracted in S3 are screened, outliers are removed, and all data are merged into one data set to obtain the ICESat-2 reference water depth control point set.

2. The method for batch extraction of laser radar reference water depth control points according to claim 1, characterized in that: The specific method of removing data outliers in S4 is: Calculate the three times mean error of the data and eliminate the data that is larger than three times the mean error.

Images