Satellite-borne laser radar coast water area sounding method based on sliding window

Abstract

The invention discloses a satellite-borne laser radar coast water area sounding method based on a sliding window. The method comprises the following steps: acquiring satellite data of a target water area; obtaining a water surface elevation and a water area height lower limit according to the target water area satellite data; photon data in the target water area satellite data are extracted according to the water surface elevation and the water area height lower limit, and multiple underwater photon data are obtained; obtaining a plurality of underwater terrain photon data according to the plurality of underwater photon data; and calculating the height difference between each underwater terrain photon data and the water surface elevation to obtain the depth of the coast water area. The depth measurement method is high in measurement accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of ocean laser radar, and more specifically, to a coastal water depth measurement method using a satellite-borne laser radar based on a sliding window. Background Art

[0002] ICESat-2 (Ice, Cloud, and Land Elevation Satellite-2) provides on-orbit bathymetry signals. The satellite carries ATLAS (Advanced Topographic Laser Altimeter System), a photon-counting lidar instrument used to measure Earth's surface elevation. Specifically, the satellite payload emits six beams of intense and weak light pulses, with each pulse separated by 90 meters, and pairs of each beam separated by 3.3 kilometers laterally. The beam spot diameter is approximately 11 meters, with a 0.7-meter spacing along the orbital direction. Notably, the ICESat-2 ATLAS laser uses a 532 nm green wavelength at a frequency of 10 kHz, enabling it to penetrate clear, shallow waters. Currently, the instrument is being used to measure the depth of coastal waters and ice-covered lakes. Data generated by this instrument are typically processed using the photon bathymetry signal to calibrate empirical models based on optical imagery. However, the ICESat-2 photon signal is contaminated with significant noise, making direct processing difficult.

[0003] Currently, noise removal algorithms are commonly used for ICESat-2 above-water scenes (such as sea ice and vegetation). The difference, regression, and Gaussian adaptive neighborhood filtering (DRAGAN) algorithm is a representative example and is widely used for processing ICESat-2 ATL08 data. However, these algorithms are ineffective in nearshore shallow waters, particularly for extracting underwater topographic signals. To address the needs of shallow water bathymetry, existing methods have been proposed to adapt to the characteristics of shallow water photon distribution. These methods improve the density-based spatial clustering of applications with noise (DBSCAN) algorithm by adaptively adjusting the neighborhood radius and minimum number of points to effectively extract signals, achieving results superior to high-confidence photon methods. Taking into account the distribution of underwater topographic photons and the decrease in photon density with increasing water depth, an adaptive elliptical filtering method is used to extract water depth photons. This method adjusts the size of the ellipse to accommodate the photon density. On this basis, the rotation direction of the ellipse is adaptively determined to effectively extract the underwater terrain photon signal. In addition, common methods also include grid processing and local statistics based methods to extract the bathymetric photon signal.

[0004] However, existing bathymetric photon extraction algorithms do not perform well when processing seabed photon data with variable density in coastal waters. It is worth noting that photon density varies between different photon datasets and within the same dataset. Specifically, the overall photon density of the ICESat-2 photon dataset varies, which can be divided into low-density and high-density photon datasets. In low-density or high-density photon datasets, photon density can also vary locally, which may lead to a mixture of local high-density noise and low-density signals. Existing water depth extraction methods are usually implemented based on statistics or local density, and cannot effectively separate terrain signals and noise from photon datasets with variable density.

[0005] Prior art discloses an ICESat-2 photon denoising method that takes into account glacier morphology. The method includes the following steps: determining the optimal photon ellipse neighborhood using multi-scale RANSAC; estimating weighted density, including continuity density, similarity density, and mixed density; and determining adaptive thresholds for along-track segments. This method specifically denoises glaciers but does not address coastal waters. Summary of the Invention

[0006] The present invention addresses the drawback of existing technologies that are difficult to cope with changes in photon density in data and provides a sliding window-based coastal water depth measurement method using satellite-borne lidar. This method is characterized by its adaptability to changes in photon density in data.

[0007] The primary purpose of the present invention is to solve the above technical problems, and the technical solutions of the present invention are as follows: A method for coastal water depth measurement using a spaceborne lidar based on a sliding window, comprising: S1: Acquire satellite data of target water area; S2: Obtaining the water surface elevation and the lower limit of the water area height according to the satellite data of the target water area; S3: extracting photon data from the target water area satellite data according to the water surface elevation and the lower limit of the water area height to obtain a plurality of underwater photon data; S4: obtaining a plurality of underwater terrain photon data according to the plurality of underwater photon data; S5: Calculate the height difference between each underwater terrain photon data and the water surface elevation to obtain the depth of the coastal waters.

[0008] Furthermore, in step S2, the water surface elevation and the lower limit of the water area height are obtained according to the satellite data of the target water area, including: S201: Calculating the photon density at each altitude using kernel density estimation based on the satellite data of the target water area; S202: selecting the maximum value of the photon density as the first photon density; selecting the height corresponding to the first photon density as the first height; selecting a preset magnification of the first photon density as the second photon density; selecting the height corresponding to the second photon density as the second height; S203: Obtain the water surface elevation and the lower limit of the water area height according to the first height and the second height.

[0009] Furthermore, in step S203, the formulas for obtaining the water surface elevation and the lower limit of the water area elevation are as follows based on the first height and the second height:

[0010]

[0011] Furthermore, in step S4, a plurality of underwater terrain photon data are obtained based on the plurality of underwater photon data, including: S401: Determine multiple seed points based on the multiple underwater photon data using a sliding window method and kernel density estimation; S402: Determine whether the distance between each underwater photon data and each seed point meets a preset condition; if the preset condition is met, the underwater photon data is underwater terrain photon data.

[0012] Furthermore, the formula for kernel density estimation is as follows:

[0013] Indicates bandwidth, Indicates height When the density of photons, Indicates the amount of photon data in the target water area, represents the Gaussian kernel function, Indicates the height of the kth photon in the target water area photon data, represents the Gaussian kernel function.

[0014] Furthermore, the formula of the Gaussian kernel function is as follows:

[0015] express arrive The normalized distance.

[0016] Furthermore, in step S402, the preset conditions are as follows:

[0017] represents the photon number, Indicates the seed point number, Indicates the preset first parameter, Indicates the preset second parameter, Indicates the The horizontal coordinate of the photon, Indicates the The horizontal coordinate of the seed point, Indicates the The height of a photon, Indicates the The height value of the seed point.

[0018] Furthermore, after step S5, the method includes: S6: correcting the depth of the coastal waters using a water depth refraction model to obtain a final depth of the coastal waters.

[0019] Furthermore, in step S6, the depth of the coastal waters is corrected using the water depth refraction model, and the formula for obtaining the final depth of the coastal waters is as follows:

[0020] Indicates the depth of coastal waters, represents the final depth of the coastal waters, represents the refractive index of air, Represents the refractive index of water.

[0021] A sliding window-based satellite-borne lidar coastal water depth sounding system, comprising: Satellite data acquisition module: obtain satellite data of target water area; Water area data calculation module: obtains the water surface elevation and the lower limit of the water area height according to the satellite data of the target water area; Photon data classification module: extracting photon data from the target water area satellite data according to the water surface elevation and the lower limit of the water area height to obtain multiple underwater photon data; An underwater terrain inference module is configured to obtain a plurality of underwater terrain photon data according to the plurality of underwater photon data; Water depth calculation module: calculates the height difference between each underwater terrain photon data and the water surface elevation to obtain the depth of the coastal waters.

[0022] Compared with the prior art, the present invention has the following beneficial effects: The present invention extracts photon data from satellite data of target water areas according to the water surface elevation and the lower limit of the water area height to obtain multiple underwater photon data; further obtains multiple underwater terrain photon data based on the multiple underwater photon data, and finally calculates the height difference between each underwater terrain photon data and the water surface elevation to obtain the depth of the coastal water area; this method can effectively extract, separate and process underwater photon signals, avoid the influence of photon density on the calculation of underwater terrain, and thus improve the accuracy of water area depth measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A flowchart of a sliding window-based satellite-borne lidar coastal water depth measurement method provided in Example 1.

[0024] Figure 2 Flowchart of the method for calculating the water surface elevation and the water surface height range provided in Example 1.

[0025] Figure 3 This is a flow chart of the method for calculating multiple underwater terrain photon data provided in Example 1.

[0026] Figure 4 Schematic diagram of the principle of the sliding window-based satellite-borne lidar coastal water depth measurement method provided in Example 1.

[0027] Figure 5 This is a bar chart of the MAE values of each algorithm provided in Example 1.

[0028] Figure 6 This is a bar chart of the RMSE values of each algorithm provided in Example 1.

[0029] Figure 7 This is a diagram showing the effect of the Median filtering algorithm provided in Example 1 for a data set with low photon density.

[0030] Figure 8 This is a diagram showing the effect of the DBSCAN-AE algorithm provided in Example 1 for a data set with low photon density.

[0031] Figure 9 This is a diagram showing the effect of the C-SHELPh algorithm provided in Example 1 for a data set with low photon density.

[0032] Figure 10 This is a rendering of the algorithm for the sliding window-based satellite-borne lidar coastal water depth sounding method for a data set with low photon density provided in Example 1.

[0033] Figure 11 This is a diagram showing the effect of the Median filtering algorithm for a data set with medium photon density provided in Example 1.

[0034] Figure 12 This is a diagram showing the effect of the DBSCAN-AE algorithm for a data set with medium photon density provided in Example 1.

[0035] Figure 13 This is a diagram showing the effect of the C-SHELPh algorithm for a data set with medium photon density provided in Example 1.

[0036] Figure 14 This is a rendering of the sliding window-based satellite-borne lidar coastal water bathymetry method for a medium photon density data set provided in Example 1.

[0037] Figure 15 This is a diagram showing the effect of the Median filtering algorithm provided in Example 1 for a data set with high photon density.

[0038] Figure 16 This is a diagram showing the effect of the DBSCAN-AE algorithm provided in Example 1 for a data set with high photon density.

[0039] Figure 17 This is a diagram showing the effect of the C-SHELPh algorithm provided in Example 1 for a data set with high photon density.

[0040] Figure 18 This is a rendering of the sliding window-based satellite-borne lidar coastal water bathymetry method for a data set with high photon density provided in Example 1. DETAILED DESCRIPTION

[0041] The accompanying drawings are for illustrative purposes only and are not to be construed as limiting this patent; In order to better illustrate this embodiment, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product size; It is understandable to those skilled in the art that some well-known structures and descriptions thereof may be omitted in the drawings.

[0042] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0043] Example 1 like Figure 1 As shown, a method for measuring coastal water depth using a spaceborne lidar based on a sliding window includes: S1: Acquire satellite data of target water area; S2: Obtaining the water surface elevation and the lower limit of the water area height according to the satellite data of the target water area; S3: extracting photon data from the target water area satellite data according to the water surface elevation and the lower limit of the water area height to obtain a plurality of underwater photon data; S4: obtaining a plurality of underwater terrain photon data according to the plurality of underwater photon data; S5: Calculate the height difference between each underwater terrain photon data and the water surface elevation to obtain the depth of the coastal waters.

[0044] In a specific embodiment, the target water area satellite data is the photon dataset ALT03 from the Ice, Cloud and Land Elevation Satellite-2 (ICESat-2). This method can effectively extract high-precision underwater terrain photon data with topographic continuity for datasets with different photon densities.

[0045] In the target water area satellite data, each photon's data includes time, coordinates, altitude, and along-track distance. In step S1 of the present invention, the target water area satellite data has already filtered the input photon's time. This algorithm only processes photon data at specific start and end times.

[0046] It should be noted that the satellite data of the target water area is two-dimensional data, and the location information only includes the horizontal coordinate and altitude of the photon.

[0047] Further, if Figure 2 As shown, in step S2, the water surface elevation and the lower limit of the water area height are obtained according to the satellite data of the target water area, including: S201: Calculating the photon density at each altitude using kernel density estimation based on the satellite data of the target water area; S202: selecting the maximum value of the photon density as the first photon density; selecting the height corresponding to the first photon density as the first height; selecting a preset magnification of the first photon density as the second photon density; selecting the height corresponding to the second photon density as the second height; S203: Obtain the water surface elevation and the lower limit of the water area height according to the first height and the second height.

[0048] In a specific embodiment, the preset magnification is 0.1 times.

[0049] It should be noted that for clear, shallow coasts, underwater topographic photon signals are typically strong and continuous. Specifically, underwater topographic photons in coastal waters are typically unaffected by structures like large-scale vegetation and buildings on land, and they typically vary continuously rather than in isolation. Therefore, sliding window-based kernel density estimation (KDE), which uses the detected local photon density maximum as the signal center and takes into account spatial autocorrelation, is an effective method for extracting topographic signals from variable-density photon datasets in coastal waters. This, in turn, allows for the more efficient and robust construction of coastal bathymetry methods that adapt to variable-density photon data.

[0050] In a specific embodiment, the bandwidth in the kernel density estimation is determined to be 0.01 meters, and this value can be adjusted according to the span of the photon trajectories involved in the calculation.

[0051] Furthermore, in step S203, the formulas for obtaining the water surface elevation and the lower limit of the water area elevation are as follows based on the first height and the second height:

[0052]

[0053]

[0054] It should be noted that the distribution of photons on water conforms to the Gaussian distribution, so The photon at the second height always exists, and the photon at the second height is unique.

[0055] It should be noted that the photon data between the upper limit of the water area height and the water surface elevation is above-water photon data, and the photon data from the water area elevation to the lower limit of the water area height is underwater photon data.

[0056] Further, if Figure 3 As shown, in step S4, a plurality of underwater terrain photon data are obtained according to the plurality of underwater photon data, including: S401: Determine multiple seed points based on the multiple underwater photon data using a sliding window method and kernel density estimation; S402: Determine whether the distance between each underwater photon data and each seed point meets a preset condition; if the preset condition is met, the underwater photon data is underwater terrain photon data.

[0057] It should be noted that it is necessary to judge whether the distance between each underwater photon data and each seed point meets the preset conditions; if it meets the preset conditions, the underwater photon data is underwater terrain photon data, which takes into account the spatial correlation between underwater terrain photons and avoids the influence of local noise in multiple underwater photon data, thereby improving the continuity and integrity of the underwater terrain photon data, and further improving the accuracy of the water depth sounding method.

[0058] It should be noted that the sliding window width and step size are 500 meters and 5 meters, respectively. The step size can be reduced in areas with high terrain volume, and increased for datasets with low signal-to-noise ratios. The bandwidth of the sliding window kernel density estimate is determined using the Silverman bandwidth method.

[0059] Furthermore, the formula for kernel density estimation is as follows:

[0060] Indicates bandwidth, Indicates height When the density of photons, Indicates the amount of photon data in the target water area, represents the Gaussian kernel function, Indicates the height of the kth photon in the target water area photon data, represents the Gaussian kernel function.

[0061] It should be noted that In, right The distance is normalized and the scaling ratio is .

[0062] Furthermore, the formula of the Gaussian kernel function is as follows:

[0063] express arrive The normalized distance.

[0064] Furthermore, in step S402, the preset conditions are as follows:

[0065] represents the photon number, Indicates the seed point number, Indicates the preset first parameter, Indicates the preset second parameter, Indicates the The horizontal coordinate of the photon, Indicates the The horizontal coordinate of the seed point, Indicates the The height of a photon, Indicates the The height value of the seed point.

[0066] It should be noted that The tables represent the major and minor axes of the ellipse respectively; their relationship is as follows:

[0067] Represents the average point density of multiple underwater photon data.

[0068] Furthermore, after step S5, the method includes: S6: correcting the depth of the coastal waters using a water depth refraction model to obtain a final depth of the coastal waters.

[0069] Furthermore, in step S6, the depth of the coastal waters is corrected using the water depth refraction model, and the formula for obtaining the final depth of the coastal waters is as follows:

[0070] Indicates the depth of coastal waters, represents the final depth of the coastal waters, represents the refractive index of air, Represents the refractive index of water.

[0071] In a specific embodiment, the following formula can be used to evaluate the effect of water depth measurement:

[0072]

[0073] M is the number of selected verification points, represents the depth of the coastal waters at the jth point, Indicates the actual depth of the j-th point.

[0074] In a specific embodiment, airborne laser depth sounding or shipborne sonar depth sounding is used as the actual depth for verification.

[0075] like Figure 4 As shown, a sliding window-based satellite-borne laser radar coastal water depth sounding system includes: Satellite data acquisition module: obtain satellite data of target water area; Water area data calculation module: obtains the water surface elevation and the lower limit of the water area height according to the satellite data of the target water area; Photon data classification module: extracting photon data from the target water area satellite data according to the water surface elevation and the lower limit of the water area height to obtain multiple underwater photon data; An underwater terrain inference module is configured to obtain a plurality of underwater terrain photon data according to the plurality of underwater photon data; Water depth calculation module: calculates the height difference between each underwater terrain photon data and the water surface elevation to obtain the depth of the coastal waters.

[0076] This paper proposes a sliding window-based coastal bathymetric method (SWCBM-Ph) for processing ICESat-2 single-photon data with varying photon densities. The method demonstrates good performance in tests conducted over six island coastal areas, effectively addressing the effects of variable-density noise.

[0077] Six island and reef coasts around the world were selected, including Oahu Island, Dongsha Island, Saipan Island, Carter Island, Barbuda Island, and Cyprus. 15 ICESat-2 photon track ATL03 data were obtained. The results obtained after various algorithms were processed are as follows: Figures 4 to 18 shown.

[0078] from Figure 5~Figure 6 It can be seen from the results that the MAE and RMSE of the method (SWCBW-Ph) of the present invention are better than those of other algorithms.

[0079] from Figures 7 to 10 It can be seen from the figure that for data with low photon density, the method of the present invention (SWCBW-Ph) reduces the number of missing or wrong marks, and its effect is better.

[0080] from Figures 11 to 14 It can be seen that for data with medium photon density, the underwater terrain photon data (Underwater Terrain Photons) annotated by the method of the present invention (SWCBW-Ph) is more accurate.

[0081] from Figures 15 to 18 It can be seen that for data with high photon density, only the method of the present invention (SWCBW-Ph) successfully annotates the underwater terrain photon data (Underwater Terrain Photons).

[0082] The same or similar reference numerals correspond to the same or similar components; The terms used in the drawings to describe positional relationships are for illustrative purposes only and should not be construed as limiting this patent; Obviously, the above embodiments of the present invention are merely examples for the purpose of clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Those skilled in the art will appreciate that other variations or modifications can be made based on the above description. It is not necessary and impossible to enumerate all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the claims of the present invention.

Claims

1. A method for coastal water depth measurement using a satellite-borne lidar based on a sliding window, characterized in that: include: S1: Acquire satellite data of target water area; S2: Obtaining the water surface elevation and the lower limit of the water area height according to the satellite data of the target water area; S3: extracting photon data from the target water area satellite data according to the water surface elevation and the lower limit of the water area height to obtain a plurality of underwater photon data; S4: obtaining a plurality of underwater terrain photon data according to the plurality of underwater photon data; S5: Calculate the height difference between each underwater terrain photon data and the water surface elevation to obtain the depth of the coastal waters.

2. The method for coastal water depth measurement based on a sliding window spaceborne lidar according to claim 1, characterized in that: In step S2, the water surface elevation and the lower limit of the water area height are obtained according to the satellite data of the target water area, including: S201: Calculating the photon density at each altitude using kernel density estimation based on the satellite data of the target water area; S202: selecting the maximum value of the photon density as the first photon density; selecting the height corresponding to the first photon density as the first height; selecting a preset magnification of the first photon density as the second photon density; selecting the height corresponding to the second photon density as the second height; S203: Obtain the water surface elevation and the lower limit of the water area height according to the first height and the second height.

3. The method for coastal water depth measurement based on a sliding window satellite-borne lidar according to claim 2, characterized in that: In step S203, the formulas for obtaining the water surface elevation and the lower limit of the water area elevation are as follows based on the first height and the second height: 。 4. The method for coastal water depth measurement based on a sliding window spaceborne lidar according to claim 1, characterized in that: In step S4, a plurality of underwater terrain photon data are obtained according to the plurality of underwater photon data, including: S401: Determine multiple seed points based on the multiple underwater photon data using a sliding window method and kernel density estimation; S402: Determine whether the distance between each underwater photon data and each seed point meets a preset condition; if the preset condition is met, the underwater photon data is underwater terrain photon data.

5. The method for coastal water depth measurement based on a sliding window spaceborne lidar according to claim 2 or 4, characterized in that: The formula for kernel density estimation is as follows: Indicates bandwidth, Indicates height When the density of photons, Indicates the amount of photon data in the target water area, represents the Gaussian kernel function, Indicates the height of the kth photon in the target water area photon data, represents the Gaussian kernel function.

6. The method for coastal water depth measurement based on a sliding window spaceborne lidar according to claim 5, characterized in that: The formula of the Gaussian kernel function is as follows express arrive The normalized distance.

7. The method for coastal water depth measurement based on a sliding window satellite-borne laser radar according to claim 4, characterized in that: In step S402, the preset conditions are as follows: represents the photon number, Indicates the seed point number, Indicates the preset first parameter, Indicates the preset second parameter, Indicates the The horizontal coordinate of the photon, Indicates the The horizontal coordinate of the seed point, Indicates the The height of a photon, Indicates the The height value of the seed point.

8. The method for coastal water depth measurement based on a sliding window spaceborne lidar according to claim 1, characterized in that: After step S5, the method includes: S6: correcting the depth of the coastal waters using a water depth refraction model to obtain a final depth of the coastal waters.

9. The method for coastal water depth measurement based on a sliding window satellite-borne laser radar according to claim 8, characterized in that: In step S6, the depth of the coastal waters is corrected using the water depth refraction model, and the formula for obtaining the final depth of the coastal waters is as follows: Indicates the depth of coastal waters, represents the final depth of the coastal waters, represents the refractive index of air, Represents the refractive index of water.

10. A sliding window-based satellite-borne laser radar coastal water depth measurement system, applied to the depth measurement method according to any one of claims 1 to 9, characterized in that: include: Satellite data acquisition module: obtain satellite data of target water area; Water area data calculation module: obtains the water surface elevation and the lower limit of the water area height according to the satellite data of the target water area; Photon data classification module: classifying the photon data in the target water area satellite data into a plurality of water surface photon data and a plurality of underwater photon data according to the water surface elevation and the lower limit of the water area height; An underwater terrain inference module is configured to obtain a plurality of underwater terrain photon data according to the plurality of underwater photon data; Water depth calculation module: calculates the height difference between each underwater terrain photon data and the water surface elevation to obtain the depth of the coastal waters.

Images