Shallow sea area water depth inversion method based on satellite-borne laser radar

Abstract

The invention discloses a shallow sea area water depth inversion method based on a satellite-borne laser radar, and particularly relates to the technical field of single-photon water depth information extraction. The denoising method comprises the following steps: S1, data preprocessing: reading data, carrying out visual presentation, dividing photons into offshore photons, sea surface photons and undersea photons, and intercepting an effective photon region; s2, coarse denoising: denoising the area intercepted in the step S1 by adopting a DBSCAN (Density Based Spatial Clustering of Applications with Noise) algorithm; s3, fine denoising: screening effective signals of sea surface photons by adopting a Gaussian fitting algorithm, removing sea photons, separating subsea photons from the sea surface photons, performing density analysis on the subsea photons, extracting a density peak point, and constructing a density trend curve of the subsea photons by adopting a cubic spline interpolation method, eliminating residual noise deviating from normal distribution according to the trend curve; and S4, performing water depth inversion: performing refraction correction on the subsea photons subjected to fine denoising, fusing the corrected subsea photons and the sea surface photons subjected to fine denoising, and generating a water depth distribution diagram of the shallow sea area.

Description

Technical Field

[0001] This invention belongs to the field of single-photon depth information technology extraction, specifically involving a method for shallow sea depth inversion based on spaceborne lidar. Background Technology

[0002] With the continuous development of technology, shallow water depth is a core foundational data for coastal zone management, coral reef ecological protection, waterway planning, and water resource development. Its accurate acquisition has irreplaceable value for marine engineering construction, fisheries resource assessment, and global sea-level change research. Traditional water depth measurement methods are limited by the accessibility of complex terrain, while airborne lidar (LiDAR) faces bottlenecks such as high cost and low efficiency for large-scale monitoring, especially in remote island shallow waters and special areas like coral reefs, making it difficult to achieve routine, high-precision detection.

[0003] Breakthroughs in spaceborne photon-counting lidar technology have provided a revolutionary solution for water depth measurement. Among these, the ATL03 product carried by the ICESat-2 satellite, with its extremely high vertical resolution and wide-area coverage, has become a core data source for shallow water depth inversion. Its basic principle is to record the flight time of laser photons from emission to return to the receiver, distinguish between surface and bottom reflection signals, and calculate water depth by combining this with the speed of light in water, effectively overcoming the limitations of traditional measurement methods. However, ICESat-2 ATL03 data still faces significant technical challenges in water depth inversion applications. On the one hand, after atmospheric scattering, solar background radiation, and reflection from suspended particles in the water, the echo signal is easily mixed with a large number of noisy photons, severely submerging surface and bottom signal photons. On the other hand, underwater signal photons exhibit significant spatial heterogeneity: as water depth increases, laser energy continuously attenuates, leading to a gradual decrease in photon density, and the signal distribution shows strong spatiotemporal differences due to topographic undulations, further increasing the difficulty of effective signal identification.

[0004] To address the problem of water depth inversion, existing technologies have developed three types of methods: (1) denoising algorithms based on raster image processing, which achieve noise filtering through spatial gridding, but are prone to losing low-density signal photons; (2) denoising algorithms based on local statistical parameters, which rely on regional statistical features to distinguish between signals and noise, but are difficult to adapt to dynamic changes in photon density; (3) denoising algorithms based on density spatial clustering, represented by DBSCAN and its improved algorithms, which identify signal clusters through density analysis, and are currently the most widely used technical path. Although existing methods have made some progress, they still have obvious defects and are difficult to cope with the dual challenges of uneven underwater signal photon density and complex noise distribution, resulting in the inversion accuracy of the key water depth range of less than 30 meters failing to meet the needs of practical applications.

[0005] Based on the above problems, this invention combines multiple algorithms for the first time to solve the problems of difficulty in identifying near-signal cluster noise and poor parameter adaptability in the prior art, and provides higher-precision data support for shallow water depth inversion. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the present invention aims to provide a shallow sea area water depth inversion method based on spaceborne lidar. This method solves the problems in existing technologies where a single algorithm is difficult to effectively distinguish between near-signal cluster noise and real signals, and parameter settings rely on human experience and have poor adaptability. It enables accurate identification of signal photons and noise suppression in photon point clouds.

[0007] The objective of this invention can be achieved through the following technical solutions: A method for shallow sea area water depth inversion based on spaceborne lidar includes the following steps: Step S1, Data Preprocessing: Read the photon point cloud data of the spaceborne lidar and project it into a two-dimensional space for visualization; based on the spatial distribution characteristics and signal properties of photons, divide all photons into three categories: marine photons, sea surface photons, and underwater photons; according to the detection range of shallow sea targets, extract the effective photon area as the core object for subsequent noise reduction processing. Step S2, coarse denoising: The DBSCAN density clustering algorithm is used to process the denoising region determined in step S1. By setting the neighborhood radius and minimum number of points, discrete noise photons are initially removed to obtain the photon point cloud dataset after coarse denoising. Step S3, Fine Denoising: For sea surface photons in the point cloud data, a Gaussian fitting algorithm is used to further filter effective signals to achieve fine denoising of sea surface photons; after fine denoising, sea photons are removed from the data, leaving only sea surface photons and underwater photons. Based on the differences in spatial location and signal intensity, underwater photons are separated from sea surface photons. Density analysis is performed on the separated underwater photons to extract their density peak points. A density trend curve of underwater photons is constructed using cubic spline interpolation. Based on the trend curve, residual noise that deviates from the normal distribution is removed to complete the overall fine denoising process. Step S4, Water Depth Inversion: Laser signal refraction correction is performed on the finely denoised underwater photon data to eliminate errors caused by seawater refraction effects; the corrected underwater photon data is spatially matched and information fused with the finely denoised sea surface photon data to generate a water depth distribution map of the shallow sea area.

[0008] Furthermore, step S1 includes the following sub-steps: Step S1-1: Read the photon point cloud data of the ATL03 product and project it into a two-dimensional space for visualization. Step S1-2: Based on the spatial distribution characteristics and signal properties of photons, all photons are divided into three categories: sea photons (green), sea surface photons (blue), and underwater photons (red); Steps S1-3: Based on the detection range of shallow sea targets, extract the effective photon region as the range for subsequent noise reduction processing.

[0009] Furthermore, step S2 includes the following sub-steps: Step S2-1: Load the photon region captured in step S1; Step S2-2: Set the neighborhood radius and minimum number of points parameters for the DBSCAN density clustering algorithm; Step S2-3: Perform coarse noise reduction on the captured photon region according to the set parameters.

[0010] Furthermore, step S3 includes the following sub-steps: Step S3-1: Perform fine noise reduction on sea surface photons using a Gaussian fitting algorithm; Step S3-2: Remove the sea surface photons and separate the remaining sea surface photons from the underwater photons; Step S3-3: Perform density analysis on the separated underwater photons and extract the density peak points; Step S3-4: Construct the density trend curve of underwater photons using cubic spline interpolation, and remove noise that deviates from the normal distribution based on the trend curve to complete the fine denoising of underwater photons.

[0011] Step S3-1-1: Extract points from the data that are close to the initial sea level, construct a height histogram, and obtain the sea level height by fitting a one-dimensional Gaussian function to the histogram. The formula for the one-dimensional Gaussian function is: in, Indicates amplitude; This represents the mean, i.e., the fitted sea level height; Standard deviation indicates the degree of dispersion. Step S3-1-2: Based on the three-standard-deviation principle of the Gaussian distribution curve, the sea surface height range is limited to three standard deviations. The formula for the sea surface height range is: in, Indicates the lower boundary of the range of sea surface height; Indicates the upper boundary of the range of sea surface height.

[0012] Step S3-2-1: Remove sea photons based on the sea surface height range determined in step S3-1.

[0013] Step S3-2-2: Separate the denoised sea surface photons from the underwater photons.

[0014] Step S3-3-1: Set the underwater photons as the original data point set, divide the window by latitude, and calculate using the following formula: in, Represents the original set of data points; Indicates latitude; Indicates elevation; Indicates the first The start or end boundary of a window; Represents the smallest of all original points. coordinate; Indicates the window width, 0.0005°; Indicates the first The center of each window coordinate; Indicates the first The left boundary of the window; Indicates the first The right boundary of the window; Step S3-3-2: Extract the elevation direction density peak value within each window, construct a histogram of density peak values, and obtain the elevation corresponding to the density peak value. The calculation formula is as follows: in, Indicates the first The set of all elevations within a window; Indicates the first The range of a window; Indicates the first The probability density values ​​of each bin; Indicates falling into the first Points for each bin; This indicates the width of the elevation bin within the current window; Indicates the first Peak density elevation of each window; Indicates the left boundary of the bin with the highest density; Indicates the right boundary of the bin with the highest density; The bin number representing the bin with the highest density; The bin index represents the bin with the highest density value.

[0015] Step S3-4-1: Sort the peak points in ascending order. The calculation formula is as follows: in, An array representing the center coordinates of all active windows; Represents an index sequence sorted in ascending order; Step S3-4-2: Construct the curve using the cubic spline interpolation function. The calculation formula is as follows: in, The function representing the constructed trend curve (interpolation function); Indicates the sorted order of the first... The center of each window (horizontal coordinate); This represents the corresponding peak density elevation (vertical axis). This indicates that the curve is second-order continuously differentiable, ensuring a smooth curve. Step S3-4-3: Predict the trend elevation and perform vertical photon filtering. The calculation formula is as follows: in, Indicates the first The predicted elevation of each original point on the trend curve; Indicates the first Latitude coordinates of the original point; Indicates the first A Boolean flag indicating whether to retain a point; Indicates the first The actual elevation of each point; This represents the vertical tolerance threshold.

[0016] Furthermore, step S4 includes the following sub-steps: Step S4-1: Perform refraction correction on the finely denoised underwater photons; Step S4-2: Fuse the refraction-corrected underwater photon data with the finely denoised sea surface photon data to generate a water depth distribution map of the current area.

[0017] Step S4-1-1: Calculate the water depth without refraction correction. The water depth formula is: in, Indicates the water depth before refraction correction; Indicates the original underwater photon altitude; Step S4-1-2: Perform refraction correction on underwater photons (i.e., seabed photons). The refraction correction formula is as follows: in, Indicates the underwater photon height after refraction correction; Indicates the refractive correction factor; It represents the effective refractive index of seawater relative to air.

[0018] Step S4-2-1: Calculate the water depth after refraction correction. The water depth formula is: in, This indicates the water depth after refraction correction; Step S4-2-2: Fuse the refracted underwater photons with the finely denoised sea surface photons to generate a water depth distribution map.

[0019] This invention first processes single-photon point cloud data into a two-dimensional array, classifies it according to photon distribution characteristics and attributes, and selects photons within the target area. It then sets relevant parameters for the DBSCAN clustering algorithm for coarse denoising. Next, it performs fine denoising on surface photons based on three times the standard deviation of Gaussian fitting. Then, it extracts the density peak values ​​of underwater photons and uses cubic spline interpolation for fine denoising. Finally, it performs refraction correction on the underwater photons to eliminate errors caused by seawater refraction. The finely denoised surface photons and underwater photons are then matched and fused to generate a water depth distribution map of the shallow sea area.

[0020] The beneficial effects of this invention are: This invention provides a method for shallow sea depth retrieval based on spaceborne lidar. This method not only uses the DBSCAN clustering algorithm for coarse denoising, but also employs Gaussian fitting for fine denoising of surface photons and cubic spline interpolation for fine denoising of underwater photons. Then, the underwater photons undergo refraction correction, and the refraction-corrected underwater photons are matched and fused with the finely denoised surface photons to obtain a water depth distribution map of the current area. This invention provides a method for shallow sea depth retrieval based on spaceborne lidar. This method employs the DBSCAN clustering algorithm to coarsely denoise the raw photon data acquired by the spaceborne lidar, initially removing significant noise. Next, fine denoising is performed on the coarsely denoised photon data—Gaussian fitting is used for surface photons, and cubic spline interpolation is used for underwater photons to improve the accuracy of both types of photon data. Subsequently, refraction correction is applied to the finely denoised underwater photons to correct positional deviations caused by light refraction. Finally, the refraction-corrected underwater photons are matched and fused with the finely denoised surface photons to generate a water depth distribution map of the current shallow sea area, achieving water depth inversion. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the water depth inversion process according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the coarse noise reduction principle of an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the principle of sea surface photon noise reduction in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the underwater photonic noise reduction principle of an embodiment of the present invention; Figure 5This is a schematic diagram of the refractive correction principle according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the target region extraction results according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the coarse denoising result according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the sea surface photon denoising results according to an embodiment of the present invention; Figure 9 This is a schematic diagram of the underwater photon noise reduction result according to an embodiment of the present invention; Figure 10 This is a schematic diagram of the refraction correction result of the finely denoised underwater photons according to an embodiment of the present invention; Figure 11 This is a schematic diagram of the water depth inversion results according to an embodiment of the present invention. Detailed Implementation

[0023] 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 embodiments of the present invention, and not all embodiments. 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.

[0024] Example

[0025] This embodiment provides a method for shallow sea depth inversion based on spaceborne lidar, such as... Figure 1 As shown, it includes the following steps: Step S1: Data preprocessing: Visualize and classify the data, and select photons in the target area.

[0026] Step S1 includes the following sub-steps: Step S1-1: Read the photon point cloud data of the ATL03 product and project it into a two-dimensional space for visualization. Step S1-2: Based on the spatial distribution characteristics and signal properties of photons, all photons are divided into three categories: sea photons (green), sea surface photons (blue), and underwater photons (red); Steps S1-3: Based on the detection range of shallow sea targets, extract the effective photon region as the range for subsequent noise reduction processing.

[0027] Specifically: Since the ATL03 product has a large data range, step S1 can be used to narrow down the data range and accurately select photons in the target area.

[0028] Step S2, coarse denoising: Use the DBSCAN clustering algorithm to coarsely denoise the target region selected in step S1.

[0029] Step S2 includes the following sub-steps: Step S2-1: Load the photon region captured in step S1; Step S2-2: Set the neighborhood radius and minimum number of points parameters for the DBSCAN density clustering algorithm; Step S2-3: Perform coarse noise reduction on the captured photon region according to the set parameters.

[0030] Specifically: Even after selecting the target area, the photons still contain a significant amount of noise. Therefore, the DBSCAN algorithm is used for coarse denoising, serving as the data basis for subsequent steps. The denoising principle is as follows: Figure 2 As shown.

[0031] Step S3, Fine Denoising: Fine denoising is performed on sea surface photons based on three times the standard deviation of Gaussian fitting, and then fine denoising is performed on underwater photons using cubic spline interpolation.

[0032] Step S3 includes the following sub-steps: Step S3-1: Perform fine noise reduction on sea surface photons using a Gaussian fitting algorithm; Step S3-2: Remove the sea surface photons and separate the remaining sea surface photons from the underwater photons; Step S3-3: Perform density analysis on the separated underwater photons and extract the density peak points; Step S3-4: Construct the density trend curve of underwater photons using cubic spline interpolation, and remove noise that deviates from the normal distribution based on the trend curve to complete the fine denoising of underwater photons.

[0033] Specifically: Because surface photons are more stable than underwater photons, different algorithms can be used for denoising to achieve better results. The principles of surface photon denoising and underwater photon denoising are as follows: Figure 3 , Figure 4 As shown.

[0034] Step S4, water depth inversion: In order to eliminate the error caused by seawater refraction, the denoised underwater photons are refraction corrected and then fused with the sea surface photons to produce a water depth distribution map.

[0035] Step S4 includes the following sub-steps: Step S4-1: Perform refraction correction on the finely denoised underwater photons; Step S4-2: Fuse the refraction-corrected underwater photon data with the finely denoised sea surface photon data to generate a water depth distribution map of the current area.

[0036] Specifically: Refraction correction of the denoised underwater photons can more accurately reflect their true location. Fusing them with surface photons yields a water depth distribution map of the target area. The principle of refraction correction is as follows: Figure 5 As shown.

[0037] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0038] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. A method for shallow sea area water depth inversion based on spaceborne lidar, characterized in that, include: Step S1, Data Preprocessing: Read the photon point cloud data of the spaceborne lidar and project it into a two-dimensional space for visualization; based on the spatial distribution characteristics and signal properties of photons, divide all photons into three categories: marine photons, sea surface photons, and underwater photons; according to the detection range of shallow sea targets, extract the effective photon area as the core object for subsequent noise reduction processing. Step S2, coarse denoising: The DBSCAN density clustering algorithm is used to process the denoising region determined in step S1. By setting the neighborhood radius and minimum number of points, discrete noise photons are initially removed to obtain the photon point cloud dataset after coarse denoising. Step S3, Fine Denoising: For sea surface photons in the point cloud data, a Gaussian fitting algorithm is used to further filter effective signals to achieve fine denoising of sea surface photons; after fine denoising, sea photons are removed from the data, leaving only sea surface photons and underwater photons. Based on the differences in spatial location and signal intensity, underwater photons are separated from sea surface photons. Density analysis is performed on the separated underwater photons to extract their density peak points. A density trend curve of underwater photons is constructed using cubic spline interpolation. Based on the trend curve, residual noise that deviates from the normal distribution is removed to complete the overall fine denoising process. Step S4, Water Depth Inversion: Laser signal refraction correction is performed on the finely denoised underwater photon data to eliminate errors caused by seawater refraction effects; the corrected underwater photon data is spatially matched and information fused with the finely denoised sea surface photon data to generate a water depth distribution map of the shallow sea area.

2. The water depth inversion method applied to shallow sea areas according to claim 1, characterized in that, Step S1 includes the following sub-steps: Step S1-1: Read the photon point cloud data of the ATL03 product and project it into a two-dimensional space for visualization. Step S1-2: Based on the spatial distribution characteristics and signal properties of photons, all photons are divided into three categories: sea photons (green), sea surface photons (blue), and underwater photons (red); Steps S1-3: Based on the detection range of shallow sea targets, extract the effective photon region as the range for subsequent noise reduction processing.

3. The water depth inversion method applied to shallow sea areas according to claim 1, characterized in that, Step S2 includes the following sub-steps: Step S2-1: Load the photon region captured in step S1; Step S2-2: Set the neighborhood radius and minimum number of points parameters for the DBSCAN density clustering algorithm; Step S2-3: Perform coarse noise reduction on the captured photon region according to the set parameters.

4. The water depth inversion method for shallow sea areas according to claim 1, characterized in that, Step S3 includes the following sub-steps: Step S3-1: Perform fine noise reduction on sea surface photons using a Gaussian fitting algorithm; Step S3-2: Remove the sea surface photons and separate the remaining sea surface photons from the underwater photons; Step S3-3: Perform density analysis on the separated underwater photons and extract the density peak points; Step S3-4: Construct the density trend curve of underwater photons using cubic spline interpolation, and remove noise that deviates from the normal distribution based on the trend curve to complete the fine denoising of underwater photons.

5. The water depth inversion method applied to shallow sea areas according to claim 1, characterized in that, Step S4 includes the following sub-steps: Step S4-1: Perform refraction correction on the finely denoised underwater photons; Step S4-2: Fuse the refraction-corrected underwater photon data with the finely denoised sea surface photon data to generate a water depth distribution map of the current area.

6. The water depth inversion method for shallow sea areas according to claim 4, characterized in that, Step S3-1 includes the following sub-steps: Step S3-1-1: Extract points from the data that are close to the initial sea level, construct a height histogram, and obtain the sea level height by fitting a one-dimensional Gaussian function to the histogram. The formula for the one-dimensional Gaussian function is: in, Indicates amplitude; This represents the mean, i.e., the fitted sea level height; Standard deviation indicates the degree of dispersion. Step S3-1-2: Based on the three-standard-deviation principle of the Gaussian distribution curve, the sea surface height range is limited to three standard deviations. The formula for the sea surface height range is: in, Indicates the lower boundary of the range of sea surface height; Indicates the upper boundary of the range of sea surface height.

7. The water depth inversion method for shallow sea areas according to claim 4, characterized in that, Step S3-2 includes the following sub-steps: Step S3-2-1: Remove sea photons based on the sea surface height range determined in step S3-1; Step S3-2-2: Separate the denoised sea surface photons from the underwater photons.

8. The water depth inversion method for shallow sea areas according to claim 4, characterized in that, Step S3-3 includes the following sub-steps: Step S3-3-1: Set the underwater photons as the original data point set, divide the window by latitude, and calculate using the following formula: in, Represents the original set of data points; Indicates latitude; Indicates elevation; Indicates the first The start or end boundary of a window; Represents the smallest of all original points. coordinate; Indicates the window width, 0.0005°; Indicates the first The center of each window coordinate; Indicates the first The left boundary of the window; Indicates the first The right boundary of the window; Step S3-3-2: Extract the elevation direction density peak value within each window, construct a histogram of density peak values, and obtain the elevation corresponding to the density peak value. The calculation formula is as follows: in, Indicates the first The set of all elevations within a window; Indicates the first The range of a window; Indicates the first The probability density values ​​of each bin; Indicates falling into the first Points for each bin; This indicates the width of the elevation bin within the current window; Indicates the first Peak density elevation of each window; Indicates the left boundary of the bin with the highest density; Indicates the right boundary of the bin with the highest density; The bin number representing the bin with the highest density; The bin index represents the bin with the highest density value.

9. The water depth inversion method for shallow sea areas according to claim 4, characterized in that, Step S3-4 includes the following sub-steps: Step S3-4-1: Sort the peak points in ascending order. The calculation formula is as follows: in, An array representing the center coordinates of all active windows; Represents an index sequence sorted in ascending order; Step S3-4-2: Construct the curve using the cubic spline interpolation function. The calculation formula is as follows: in, The function representing the constructed trend curve (interpolation function); Indicates the sorted order of the first... The center of each window (horizontal coordinate); This represents the corresponding peak density elevation (vertical axis). This indicates that the curve is second-order continuously differentiable, ensuring a smooth curve. Step S3-4-3: Predict the trend elevation and perform vertical photon filtering. The calculation formula is as follows: in, Indicates the first The predicted elevation of each original point on the trend curve; Indicates the first Latitude coordinates of the original point; Indicates the first A Boolean flag indicating whether to retain a point; Indicates the first The actual elevation of each point; This represents the vertical tolerance threshold.

10. The water depth inversion method applied to shallow sea areas according to claim 5, characterized in that, Step S4-1 includes the following sub-steps: Step S4-1-1: Calculate the water depth without refraction correction. The water depth formula is: in, Indicates the water depth before refraction correction; Indicates the original underwater photon altitude; Step S4-1-2: Perform refraction correction on underwater photons (i.e., seabed photons). The refraction correction formula is as follows: in, Indicates the underwater photon height after refraction correction; Indicates the refractive correction factor; It represents the effective refractive index of seawater relative to air.

11. The water depth inversion method for shallow sea areas according to claim 5, characterized in that, Step S4-2 includes the following sub-steps: Step S4-2-1: Calculate the water depth after refraction correction. The water depth formula is: in, This indicates the water depth after refraction correction; Step S4-2-2: Fuse the refracted underwater photons with the finely denoised sea surface photons to generate a water depth distribution map.

Images