Near-shore ocean front identification method based on Lagrange particle statistics

Abstract

The invention provides a near-shore ocean front identification method based on Lagrange particle statistics, and belongs to the technical field of ocean environment numerical simulation, and the method comprises the steps: constructing a target sea area hydrodynamic model to simulate a water flow state; performing Lagrange particle tracking according to the constructed target sea area hydrodynamic model to obtain distribution information of particles in time and space; carrying out relative dispersion analysis on the obtained distribution information of the particles in time and space, counting and analyzing the change characteristics of the particles along with time, and obtaining the spatial distribution characteristics of the relative dispersion; and identifying and calibrating the position of the ocean front by using the spatial distribution characteristics of the relative dispersion. According to the method, the Lagrange particle tracking technology is combined with relative dispersion analysis, the method is used for recognizing frontal surface structures such as the offshore capes, the estuary fronts and the land frame fronts, and accurate prediction and recognition of the ocean frontal surfaces are achieved.

Description

Technical Field

[0001] This invention relates to the field of marine environmental numerical simulation technology, and in particular to a method for identifying nearshore marine fronts based on Lagrange particle statistics. Background Technology

[0002] A marine front is a narrow transition zone between two or more bodies of water with distinct characteristics. Its characteristics can be described by horizontal gradients of factors such as temperature, salinity, density, color, and chlorophyll content. Common marine fronts include estuarine fronts, shallow-sea fronts, upwelling fronts, shelf slope break fronts, and western boundary current fronts. Coastal fronts, cape fronts, and basement fronts, especially those occurring in nearshore areas, have important applications in nearshore fisheries and marine environmental protection, and are widely valued by professionals in the marine field.

[0003] The main technologies for identifying ocean fronts include the following: First, the spatial distribution of environmental elements is obtained through remote sensing or field survey data, thereby analyzing the concentration gradient of these elements to determine the location and intensity of fronts. However, satellite remote sensing and ocean surveys have limitations in spatiotemporal resolution, making it difficult to capture small-scale or fine-grained frontal phenomena.

[0004] Second, numerical simulation techniques can be used to better explore frontal dynamics. For example, Liu et al. (Variation in the Current Shear Front and its Potential Effect on Sediment Transport Over the Inner Shelf of the East China Sea in Winter[J]. Journal of Geophysical Research: Oceans, 2018, 123) combined observation and numerical simulation methods to study the characteristics, variation mechanism, and potential impact of ocean current shear fronts on the East China Sea inner shelf on water and sediment transport and the muddy zone of the inner shelf.

[0005] For ocean front detection or prediction, existing related technologies include CN117217073A, which discloses a method for extracting the frontal line of the Kuroshio Current extension below the sea surface, as well as a method for calculating, predicting, and analyzing the frontal trend, solving the problem of accurately extracting the frontal line of the Kuroshio Current extension; CN113484860A, which discloses a method and system for detecting ocean fronts in liquids based on Doppler center anomalies in SAR images, making full use of the velocity characteristics of ocean fronts, solving the problems of missed and false detections of ocean fronts based on amplitude images, and improving the accuracy of ocean front detection in SAR images; and CN114187553A, which discloses an ocean front detection method that integrates scSE and Mask R-CNN networks.

[0006] However, current technologies for detecting or predicting oceanic fronts are mainly based on the perception of the spatiotemporal distribution of environmental elements and the determination of oceanic fronts by the gradient of these elements. However, these technologies have limitations in identifying small-scale oceanic fronts, making it difficult to accurately obtain the distribution of small-scale environmental elements, thus hindering the detection of frontal phenomena.

[0007] The characteristics of oceanic fronts dictate that matter tends to accumulate near the front, forming a barrier that hinders the transport of matter across the front. Therefore, statistical analysis of matter transport paths can reveal the phenomenon of oceanic fronts.

[0008] Lagrange particle tracking, as a research method for ocean material transport, can quickly and effectively identify ocean phenomena by simulating and tracking the trajectories and behaviors of particles near ocean fronts. In recent years, statistical analysis based on Lagrange particle trajectories has made positive progress in revealing ocean current dynamics, material transport mechanisms, and mixing processes. These methods can be categorized into single-particle statistical analysis, two-particle statistical analysis, and multi-particle statistical analysis based on the number of particles in the statistical unit.

[0009] Therefore, how to effectively identify nearshore ocean fronts based on Lagrange particle statistics is a technical problem that urgently needs to be solved. Summary of the Invention

[0010] To address the above problems and improve the accuracy and speed of identifying ocean fronts, this invention provides a method for identifying nearshore ocean fronts based on Lagrange particle statistics.

[0011] This invention provides a method for identifying nearshore ocean fronts based on Lagrange particle statistics, comprising: Construct a hydrodynamic model of the target sea area to simulate the water flow state; Lagrange particle tracking is performed based on the constructed hydrodynamic model of the target sea area to obtain the distribution information of particles in time and space; A relative dispersion analysis is performed on the obtained distribution information of the particles in time and space. The characteristics of particle change over time are statistically analyzed and the spatial distribution characteristics of relative dispersion are obtained. The location of the ocean front is identified and calibrated by utilizing the spatial distribution characteristics of the relative dispersion.

[0012] As a further improvement of the present invention, the construction of a target sea area hydrodynamic model to simulate the water flow state includes constructing a target sea area hydrodynamic model based on FVCOM.

[0013] As a further improvement of the present invention, the target sea area hydrodynamic model adopts Sigma coordinates.

[0014] As a further improvement of the present invention, the governing equations in the Sigma coordinates are as follows: Continuity equation:

[0015] Momentum equation:

[0016]

[0017] Temperature equation and salinity equation:

[0018]

[0019]

[0020]

[0021]

[0022] in, , , These are the coordinate directions. , , This represents the velocity components in the corresponding three directions. It's temperature. Represents salinity. For density, Represents atmospheric pressure. Coriolis parameters, It is the acceleration due to gravity. The vertical eddy viscosity coefficient is... The vertical eddy diffusion coefficient of thermosalinity; and Represents horizontal momentum. and These represent the diffusion terms for temperature and salinity, respectively; Represents turbulent kinetic energy. Represents the turbulence length; Boundary conditions: In the nearshore area, water level boundary conditions are given on the open boundary.

[0023] As a further improvement of the present invention, the Lagrange particle tracking based on the constructed target sea area hydrodynamic model includes using the following formula as the equation of motion of the particles in the flowing water:

[0024] Where x represents the position vector of the particle at time t; u(x, t) is the velocity vector.

[0025] As a further improvement of the present invention, the Lagrange particle tracking based on the constructed target sea area hydrodynamic model includes updating the particle position using a fourth-order Runge-Kutta method:

[0026]

[0027]

[0028]

[0029]

[0030] Among them, the particles in time The position is Particles in The position is .

[0031] As a further improvement of the present invention, the relative discreteness analysis of the obtained distribution information of the particles in time and space includes constructing a relative discreteness field.

[0032] As a further improvement of the present invention, the formula for calculating the relative discreteness field is as follows:

[0033] in, The initial moment of particle release. Represents a time interval. for Moment Particle Location Is time , , , The location of the particle.

[0034] As a further improvement of the present invention, the forward integration of the constructed relative discrete field generates a forward relative discrete field, and the backward integration of the constructed relative discrete field generates a backward relative discrete field.

[0035] As a further improvement of the present invention, the step of identifying and calibrating the position of the ocean front by utilizing the spatial distribution characteristics of the relative dispersion includes using the Canny image edge detection algorithm to identify the position of the ocean front.

[0036] As a further improvement of the present invention, the method of identifying the location of the ocean front using the Canny image edge detection algorithm includes: Preprocessing to remove noise; Calculate the gradient to obtain the intensity and direction of the pixel gradient; Perform nonmaximum suppression to refine the edges; Perform threshold detection to distinguish between strong and weak edges; Lag edge tracking connects edge segments.

[0037] This invention provides a method for identifying nearshore ocean fronts based on Lagrange particle statistics. It combines Lagrange particle tracking technology with relative dispersion analysis, and through the perspective of material transport effects and with the help of numerical simulation technology, it can deeply explore the flow characteristics behind the flow, achieve accurate prediction and identification of ocean fronts, broaden the front identification technology, and provide a new perspective for understanding complex ocean dynamics. It can be used to identify frontal structures such as nearshore cape fronts, estuary fronts, and shelf fronts, and macroscopically and quickly predict the drift paths, diffusion and accumulation range of marine oil spills, marine debris and particulate matter. Attached Figure Description

[0038] Figure 1 This is a schematic diagram illustrating the steps of the nearshore ocean front identification method based on Lagrange particle statistics according to an embodiment of the present invention.

[0039] Figure 2 This is a schematic diagram illustrating the implementation process of the nearshore ocean front identification method based on Lagrange particle statistics according to an embodiment of the present invention.

[0040] Figure 3 This is a particle distribution diagram under the Biklee jet flow field obtained by Lagrange particle tracking in an embodiment of the present invention.

[0041] Figure 4 This is a spatial distribution diagram of the relative dispersion in the Bikley jet verification test of an embodiment of the present invention.

[0042] Figure 5 This is a schematic diagram of the relative dispersion edge extracted by image edge detection in the Bickle jet verification experiment of the present invention.

[0043] Figure 6 This is a topographic map of the shoreline and water depth of the Yellow River Delta region in this embodiment of the invention.

[0044] Figure 7 This is a schematic diagram of the hydrodynamic field simulation results of the Yellow River Delta region in an embodiment of the present invention.

[0045] Figure 8 This is a backward relative dispersion spatial distribution map of the Yellow River Delta region in an embodiment of the present invention.

[0046] Figure 9 This is a schematic diagram of the relative dispersion of the image edge extraction based on the spatial distribution of relative dispersion in the Yellow River Delta region, as described in an embodiment of the present invention. Detailed Implementation

[0047] The following describes specific embodiments and appendices. Figure 1-9 The invention is described in detail so that those skilled in the art can more fully understand its purpose, features and effects.

[0048] Unless otherwise specified, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. In the event of any discrepancy between the definitions of terms in this invention and their commonly understood meaning by one of ordinary skill in the art to which this invention pertains, the definitions set forth herein shall prevail.

[0049] Currently, the detection and prediction technologies for nearshore ocean fronts mainly rely on the observation and simulation of environmental variables (such as temperature, salinity, chlorophyll concentration, and suspended sediment content). The core of these methods lies in using gradient changes in environmental factors to define the location of ocean fronts. However, the source-sink dynamics of environmental factors and their complex interactions significantly affect the effectiveness and accuracy of front detection. Especially in nearshore areas, small-scale ocean fronts triggered by topographic features such as headlands, due to their small size and dynamic characteristics, are often difficult to clearly show in traditional spatial distribution maps of environmental factors, posing a problem in accurately capturing such fine structures.

[0050] This invention provides a method for identifying nearshore ocean fronts based on Lagrange particle statistics, which improves upon existing technologies and can greatly enhance the accuracy and speed of identifying nearshore ocean fronts, especially small-scale ocean fronts.

[0051] Example 1 As a specific embodiment of the present invention, this embodiment provides a method for identifying nearshore ocean fronts based on Lagrange particle statistics, referring to... Figure 1 , Figure 2 The specific steps are as follows: S100. Construct a hydrodynamic model of the target sea area to simulate the water flow state; S200. Based on the constructed hydrodynamic model of the target sea area, Lagrange particle tracking is performed to obtain the distribution information of particles in time and space. S300. Perform relative dispersion analysis on the obtained particle distribution information in time and space, statistically analyze the particle change characteristics over time, and obtain the spatial distribution characteristics of relative dispersion. S400: Using the spatial distribution characteristics of relative dispersion, the location of ocean fronts is identified and calibrated.

[0052] This invention presents a method for identifying nearshore ocean fronts based on Lagrange particle statistics. It creatively combines Lagrange particle tracking technology with relative dispersion analysis, achieving intuitive visualization of the ocean front formation process through quantitative statistics of particle behavior characteristics. This enables more accurate and rapid identification of ocean fronts. This method not only overcomes the limitations of traditional methods but also more objectively and accurately depicts the dynamic evolution of fronts.

[0053] This invention is applicable to marine environmental monitoring and disaster early warning, especially marine oil spill prediction, and is particularly suitable for the fine detection and analysis of small-scale frontal structures in nearshore areas, including but not limited to typical marine phenomena such as coastal cape fronts and estuary fronts.

[0054] Example 2 As another specific embodiment of the present invention, this embodiment provides a method for identifying nearshore ocean fronts based on Lagrange particle statistics, referring to... Figure 1 , Figure 2 Based on Example 1, it includes: In S100, a hydrodynamic model of the target sea area is constructed based on FVCOM (Finite-Volume Coastal Ocean Model) to match the complex and winding coastline of the nearshore region. Due to the unique design of FVCOM's unstructured triangular mesh, its use enables fast and efficient computation while maintaining the conservation of physical quantities. The model's flexibility and adaptability are particularly suitable for waters with varied shoreline morphologies, such as estuaries and bays, ensuring high accuracy and reliability of simulation results under complex terrain conditions.

[0055] Furthermore, the hydrodynamic model for the target sea area adopts Sigma coordinates. The governing equations in Sigma coordinates are: Continuity equation:

[0056] Momentum equation:

[0057]

[0058] Temperature equation and salinity equation:

[0059]

[0060]

[0061]

[0062]

[0063] in, , , These are the coordinate directions. , , This represents the velocity components in the corresponding three directions. It's temperature. Represents salinity. For density, Represents atmospheric pressure. Coriolis parameters, It is the acceleration due to gravity. The vertical eddy viscosity coefficient is... The vertical eddy diffusion coefficient of thermosalinity; and Represents horizontal momentum. and These represent the diffusion terms for temperature and salinity, respectively; Represents turbulent kinetic energy. Represents the turbulence length; Boundary conditions: In the nearshore area, water level boundary conditions are given on the open boundary.

[0064] By using Sigma coordinates in the FVCOM model, the coordinate system defines the vertical coordinates as a scale to the water depth, allowing the model to naturally adapt to various complex topographic variations. The coordinate system can more accurately describe the undulations of the seabed topography, thereby improving the model's simulation accuracy for nearshore and shallow water areas. Furthermore, by dynamically adjusting the vertical stratification, the thickness of each layer varies with water depth, thus reducing errors caused by topographic changes.

[0065] In other embodiments, hydrodynamic models, such as MIKE21, can also be used to simulate the flow state of water bodies.

[0066] In S200, the motion of particles in flowing water adopts and follows the following equations of motion:

[0067] Where x represents the particle in time t Position vector; u(x, t ) is the velocity vector.

[0068] Furthermore, a fourth-order Runge-Kutta method is used to approximate the particle position update. It is assumed that the particle's position is updated in time... The position is Then in the next time step Location It can be obtained using the fourth-order Runge-Kutta method, with the specific steps as follows:

[0069]

[0070]

[0071]

[0072]

[0073] In S300, a relative discreteness field is constructed, and based on the constructed relative discreteness field, the characteristics of particle changes over time are statistically analyzed to obtain the spatial distribution characteristics of the relative discreteness.

[0074] Relative Dispersion (RD) is a statistical method for quantifying the time-varying characteristics of a particle swarm, particularly suitable for evaluating the changes in particle distribution from initial to final distribution. Specifically, based on four-particle statistics, the relative dispersion field of the entire computational domain is constructed by analyzing the particle distribution changes of the four neighboring grid points around each discrete grid point. The calculation formula is as follows:

[0075] in, The initial moment of particle release. Represents a time interval. for Moment Particle Location Is time , , , The location of the particle.

[0076] The forward integration of the constructed relative discrete field generates the forward relative discrete field (Forward RDField), and the backward integration of the constructed relative discrete field generates the backward relative discrete field (Backward RDField).

[0077] The forward integral starts from the initial release time of the particle and tracks its dynamic changes up to the current time point, reflecting the degree of "stretching" or "contraction" that the particle undergoes over time. The backward integral traces back from the current time point to the initial release time of the particle and is mainly used to identify the regions where the particle swarm contraction is most significant.

[0078] This embodiment employs a dual analytical framework of forward and backward integration. Forward integration focuses on the formation and development of fronts, while backward integration aims to reveal the precise location of fronts. This strategy not only enhances the comprehensiveness of front prediction but also provides rich data support for subsequent in-depth analysis.

[0079] In S400, continuous linear features with significant gradient changes in the relative discreteness field are captured. These continuous linear features with significant gradient changes in the relative discreteness field are referred to as "ridges." Specifically, due to their robustness in noisy environments and effective detection of weak edges, the Canny image edge detection algorithm is used to capture these continuous linear features with significant gradient changes in the relative discreteness field.

[0080] Furthermore, the Canny algorithm is used to capture the continuous linear features of significant gradient changes in the discrete field, including: S401. Preprocessing, noise removal Specifically, a Gaussian filter is used to smooth the image, reducing the impact of noise and ensuring the accuracy of subsequent edge detection.

[0081] S402. Calculate the gradient to obtain the intensity and direction of the pixel gradient. Specifically, the Sobel operator or other similar differential operators are used to calculate the gradient intensity and direction of each pixel in the image. The gradient intensity reflects the sharpness of the edge, while the direction indicates the orientation of the edge.

[0082] S403. Perform non-maximum suppression and refine the edges. Specifically, by comparing the gradient direction of each pixel with the gradient intensity in its neighborhood, non-local maxima are removed, thereby achieving edge refinement and ensuring that each edge is represented by only a single pixel-wide line.

[0083] S404. Perform threshold detection to distinguish between strong and weak edges. Specifically, two thresholds are set: a high threshold and a low threshold, used to identify strong edges and potential weak edges, respectively. Pixels above the high threshold are considered strong edge points, while pixels below the low threshold are discarded.

[0084] S405, Lag Edge Tracking, Connecting Edge Fragments Specifically, by using hysteresis edge tracing, weak edge points that meet certain conditions are linked into complete edge lines to obtain a coherent ridge structure and a front structure in a relatively discrete field.

[0085] Ultimately, the frontal structure captured in the relative discrete field is taken as the nearshore ocean front.

[0086] To accurately capture frontal structures in a relatively discrete field, this embodiment employs the classic Canny algorithm from the field of image processing. Through high-precision edge detection technology, key "ridge" structures in the relatively discrete field can be effectively identified as oceanic fronts, improving not only the accuracy of front identification but also providing a new path for the automated detection of oceanic fronts.

[0087] The present invention provides a method for identifying nearshore ocean fronts based on Lagrange particle statistics. Unlike existing methods that rely on the distribution of environmental factors, this method focuses on tracking the trajectory of Lagrange particles and reveals the existence of ocean fronts by analyzing the statistical characteristics of particle behavior. This method not only avoids errors caused by the uncertainty of environmental variables but also efficiently identifies small-scale ocean front phenomena that are difficult to capture using traditional methods, making it particularly suitable for complex and variable nearshore environments.

[0088] Furthermore, in one embodiment, to verify the feasibility of the relative discreteness method in identifying material transport structures, an experiment was conducted using the classic ideal flow field experiment—the Biklee jet case. The Biklee jet is a kinematically idealized model, a meandering, ribbon-like jet composed of counter-rotating vortices above and below it. The Biklee jet is described by the transient velocity field generated by the stream function, whose equation is:

[0089] in, t For time, This represents a stable background flow field. Represents a periodically changing flow field;

[0090] The specific values ​​of these parameters can be set according to the specific circumstances.

[0091] The particle distribution under the Bickle jet flow field obtained from Lagrange particle tracking is as follows: Figure 3 As shown. The relative discreteness field is obtained through particle tracking, as shown. Figure 4 As shown, edges with relative dispersion are then extracted using an image edge detection algorithm, resulting in the following: Figure 5 The spatial structure of relative dispersion.

[0092] Example 3 As another specific embodiment of the present invention, this embodiment provides a specific example of a method for identifying nearshore ocean fronts based on Lagrange particle statistics, taking the Yellow River Delta region as an example, to identify nearshore cape fronts.

[0093] Obtain basic geographic data.

[0094] Coastline and water depth topography: The coastline direction of the target area is extracted using high-resolution nautical charts or satellite imagery software such as Google Earth. Simultaneously, detailed water depth data is collected using nautical chart data or on-site underwater topographic surveys to construct a coastline and water depth topographic map of the calculation area, such as... Figure 6 As shown.

[0095] Construct a hydrodynamic model for the target sea area.

[0096] Select an ocean model: Choose an ocean computing model suitable for complex coastlines, such as FVCOM (Finite Volume Community Ocean Model) or MIKE 21, to simulate the hydrodynamic field of the study area. Ensure the model is fully validated and the simulation results are in high agreement with field observation data. The hydrodynamic field of the target sea area is as follows: Figure 7 As shown.

[0097] To conduct Lagrange particle tracking.

[0098] Virtual particles are deployed within the target sea area and continuously tracked to record their trajectories, providing foundational data for subsequent analysis.

[0099] Relative dispersion analysis is performed, calculating the relative dispersion of particles among nodes within the computational region, and generating a spatial distribution map of the relative dispersion. For example... Figure 8 As shown, the spatial distribution of relative dispersion was presented in high-definition image form using professional drawing software.

[0100] Structural extraction and analysis utilize image processing techniques, such as the Canny edge detection algorithm, to perform detailed analysis of images with relatively discrete spatial distributions, extracting key structural features and identifying the specific location and morphology of coastal headlands. Figure 9 As shown.

[0101] This embodiment is based on the relative dispersion near the Yellow River Delta. Its spatial distribution reflects the blocking and shielding effect of the coastal headlands on material transport. This structure is similar to the distribution structure of pollutants or suspended sediment obtained by remote sensing and can be regarded as the coastal headland front.

[0102] This embodiment can not only accurately detect the coastal cape front structure in the Yellow River Delta region, but also further explore its dynamic evolution process, providing a scientific basis for marine environmental management and ecological restoration in the region.

[0103] The method for identifying nearshore ocean fronts based on Lagrange particle statistics of the present invention can be used to identify frontal structures such as nearshore cape fronts, estuary fronts and shelf fronts, and to macroscopically and quickly predict the drift paths, diffusion and accumulation range of marine oil spills, marine debris and particulate matter.

[0104] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A method for identifying nearshore ocean fronts based on Lagrange particle statistics, characterized in that, The method includes: Construct a hydrodynamic model of the target sea area to simulate the water flow state; Lagrange particle tracking is performed based on the constructed hydrodynamic model of the target sea area to obtain the distribution information of particles in time and space; A relative dispersion analysis is performed on the obtained distribution information of the particles in time and space. The characteristics of particle change over time are statistically analyzed and the spatial distribution characteristics of relative dispersion are obtained. The location of the ocean front is identified and calibrated by utilizing the spatial distribution characteristics of the relative dispersion.

2. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 1, characterized in that, The construction of a target sea area hydrodynamic model to simulate water flow includes constructing a target sea area hydrodynamic model based on FVCOM.

3. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 2, characterized in that, The hydrodynamic model of the target sea area uses Sigma coordinates.

4. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 1, characterized in that, The Lagrange particle tracking based on the constructed target sea area hydrodynamic model includes using the following formula as the equation of motion for the particles in flowing water: Where x represents the position vector of the particle at time t; u(x, t) is the velocity vector.

5. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 4, characterized in that, The Lagrange particle tracking based on the constructed target sea area hydrodynamic model includes updating the particle position using a fourth-order Runge-Kutta method: Among them, the particles in time The position is Particles in The position is .

6. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 1, characterized in that, The relative discreteness analysis of the obtained particle distribution information in time and space includes constructing a relative discreteness field.

7. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 6, characterized in that, The formula for calculating the relative discreteness field is: in, The initial moment of particle release. Represents a time interval. for Moment Particle Location Is time , , , The location of the particle.

8. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 7, characterized in that, The forward integration of the constructed relative discrete field generates the forward relative discrete field, and the backward integration of the constructed relative discrete field generates the backward relative discrete field.

9. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 1, characterized in that, The process of identifying and calibrating the location of the ocean front by utilizing the spatial distribution characteristics of the relative dispersion includes using the Canny image edge detection algorithm to identify the location of the ocean front.

10. The method for identifying nearshore ocean fronts based on Lagrange particle statistics according to claim 9, characterized in that, The method of identifying the location of ocean fronts using the Canny image edge detection algorithm includes: Preprocessing to remove noise; Calculate the gradient to obtain the intensity and direction of the pixel gradient; Perform nonmaximum suppression to refine the edges; Perform threshold detection to distinguish between strong and weak edges; Lag edge tracking connects edge segments.

Images