Methods and Systems for Extracting Dynamic Parameters of Ocean Eddies from SAR and Water Color Remote Sensing Images

Abstract

This invention discloses a method and system for extracting ocean vortex dynamic parameters from SAR and sea color remote sensing images. The method includes: acquiring time-series SAR images and time-series sea color remote sensing images; registering the time-series SAR and sea color remote sensing images; acquiring the sea surface vortex flow field from the time-series SAR image and the sea surface vortex flow field from the time-series sea color image, respectively; weightedly fusing the vortex flow field estimation results from the SAR image and the sea color image to obtain a fused flow field; smoothing the fused flow field to satisfy the principles of continuity and consistency, and extracting ocean vortex dynamic parameters from it. The method and system for extracting ocean vortex dynamic parameters from SAR and sea color remote sensing images provided by this invention effectively utilize the complementary information of SAR and sea color remote sensing images to obtain the sea surface vortex flow field, and extract parameters such as maximum tangential velocity, vorticity, divergence, vortex kinetic energy, and deformation rate from it.

Description

Technical Field

[0001] This invention belongs to the fields of remote sensing image processing and marine remote sensing applications, specifically relating to a method and system for extracting marine eddy dynamic parameters from SAR and water color remote sensing images. Background Technology

[0002] Ocean eddies are ubiquitous in the global oceans, carrying 80% of the total kinetic energy of the ocean. They play a crucial role in the energy exchange and nutrient transport processes at different scales within the ocean, significantly influencing the physical and chemical properties of deep-sea currents and water masses.

[0003] Ocean eddies, due to their rotation, can attract phytoplankton, leading to a significant increase in chlorophyll concentration at the sea surface during periods of abundant phytoplankton growth. Ocean color sensors can invert information about the concentrations of various substances causing changes in ocean color by detecting changes in the signals they receive. Ocean eddies can also cause changes in the distribution of chlorophyll concentration in the ocean; by tracking changes in sea surface chlorophyll concentration acquired by ocean color sensors, the flow field and dynamic parameters of ocean eddies can be extracted. A major challenge in extracting the flow field of ocean eddies from ocean color remote sensing images is the susceptibility of these images to cloud cover, which can result in data loss.

[0004] SAR sensors are unaffected by weather conditions, enabling them to acquire remote sensing images around the clock. Ocean eddies can appear in Synthetic Aperture Radar (SAR) images through an "oil film mechanism." This is because biofilms produced by phytoplankton or oil spills floating on the sea surface suppress the generation of capillary waves, reducing the intensity of SAR backscattering. Under the influence of eddies, the oil film appears as a black "spiral" structure in SAR images. Ocean eddies can drive the redistribution of biofilms on the sea surface; therefore, by tracking changes in the spiral oil film structure in SAR images, the ocean eddy current field can be extracted. However, under conditions such as algal blooms, low wind speeds, and upwelling, large areas of low scattering appear in SAR images, affecting the integrity of the ocean eddy current field extraction and consequently impacting the extraction of ocean eddy dynamic parameters.

[0005] To obtain information on ocean vortex flow fields and dynamic parameters more effectively, existing methods need to be improved and perfected. Summary of the Invention

[0006] Based on the above technical problems, the main objective of this invention is to provide a method and system for extracting ocean vortex dynamic parameters from SAR and water color remote sensing images, thereby obtaining complete vortex flow field information.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for extracting ocean eddy dynamic parameters from SAR and water color remote sensing images, comprising the following steps:

[0009] Step 1: Acquire time-series SAR images and time-series water color remote sensing images;

[0010] Step 2: Register the time-series SAR image and the time-series water color remote sensing image;

[0011] Step 3: Obtain the sea surface vortex current field from time-series SAR images and the sea surface vortex current field from time-series water color remote sensing images, respectively;

[0012] Step 4: Perform weighted fusion of the sea surface vortex current field in the time series SAR image and the sea surface vortex current field in the time series water color remote sensing image to obtain the fused current field;

[0013] Step 5: Smooth the fused flow field and extract the ocean vortex dynamic parameters from it.

[0014] Furthermore, in step 1, the number of time-series SAR images acquired is at least two, and the number of time-series water color remote sensing images acquired is at least two.

[0015] Furthermore, the time interval of the time-series remote sensing images is shorter than the decorrelation time of the image features.

[0016] Furthermore, the sea surface vortex current field of SAR images is estimated from time-series SAR images, and the sea surface vortex current field of water color images is estimated from time-series water color remote sensing images.

[0017] Furthermore, taking time-series SAR images as an example, obtaining the sea surface vortex current field of SAR images includes:

[0018] The first SAR image in the time series SAR image is divided into many small template windows. Any template window T is selected, and there is partial overlap between different template windows T.

[0019] The template window T is matched within the search region of the second SAR image selected in the time series SAR images, and the cross-correlation coefficient is calculated using the following formula:

[0020]

[0021] Where S′ represents the search sub-window corresponding to the template window within the search area, (x,y) are the center pixel coordinates of the template window, and p and q represent the displacements of the center of the search sub-window relative to the center of the search window in the distance and azimuth directions, respectively. This represents the average pixel value within the search sub-window. This represents the average pixel value within the template window;

[0022] Obtain the maximum value r(x,y) in the correlation coefficient matrix corresponding to the template window centered at pixel (x,y), and estimate the vortex horizontal velocity vector based on the positional offset of the maximum value relative to the center point:

[0023]

[0024] θ=tan -1 ((q max Δy) / (p max Δx))

[0025] Where Δt is the time interval of the time-series SAR images, Δx and Δy are the pixel intervals of the SAR images in the range and azimuth directions, respectively, c is the estimated sea surface vortex current field vector, d is the estimated sea surface displacement vector, and p max and q max θ represents the number of pixels that the template window T moves to the right and down within the search window when the maximum cross-correlation coefficient is obtained, respectively, and θ is the direction of the estimated sea surface vortex flow field vector.

[0026] Select the next template window and repeat the above operation until all template windows in the first SAR image have been traversed, and the sea surface vortex flow field of the time series SAR image can be obtained.

[0027] Furthermore, the sea surface vortex current fields of the acquired time-series SAR images and time-series water color remote sensing images are filtered to obtain the effective vortex current field vector. The effective correlation coefficient w at any center pixel (x,y) of the template window after filtering is then calculated. n (x,y) can be represented as:

[0028]

[0029] Where, r n (x,y) represents the maximum cross-correlation coefficient corresponding to the nth flow field vector, m is the correlation coefficient threshold, and n is the number of flow fields at the center pixel of the template window, n=1,2.

[0030] Furthermore, the sea surface eddy current field of the time-series SAR image is fused with the sea surface eddy current field of the time-series water color remote sensing image to obtain the fused flow field. The latitudinal and meridional components of the fused flow field at any central pixel (x,y) of the template window are as follows:

[0031]

[0032] Among them, u n (x,y) and v n(x,y) represent the nth latitudinal and longitudinal components at the center pixel (x,y) of the template window, respectively, where n is the number of flow fields at the center pixel of the template window.

[0033] Furthermore, the merged flow field is smoothed to obtain the complete sea surface vortex flow field.

[0034]

[0035] in, It is an M-dimensional column vector.

[0036] Furthermore, in step 5, extracting the dynamic parameters of the ocean eddy includes: using the vorticity of the eddy to reflect the strength of the eddy rotation; the calculation method for vorticity is as follows: The vorticity field of the vortex is obtained from the fused flow field as follows:

[0037]

[0038] in, For Hamiltonian operators, ζ Ci (i = 1, 2, ..., N) is an M-dimensional column vector.

[0039] Furthermore, in step 5, extracting the dynamic parameters of the ocean eddy includes: using the divergence of the eddy to reflect the convergence and divergence in the eddy velocity field; the divergence calculation method is as follows: The divergence field of the vortex is obtained from the fused flow field as follows:

[0040]

[0041] Where, ξ Ci (i = 1, 2, ..., N) is an M-dimensional column vector.

[0042] Further, in step 5, extracting the ocean vortex dynamic parameters includes: obtaining the vortex kinetic energy KE of the vortex velocity field at any template window center pixel (x,y) using the above-mentioned fused flow field.

[0043] KE=(u 2 (x,y)+v 2 (x,y) / 2

[0044] Where u(x,y) and v(x,y) represent the latitudinal and longitudinal components of the fused flow field at the center pixel (x,y) of the template window, respectively.

[0045] Further, in step 5, extracting the dynamic parameters of the ocean eddy includes: the characteristics of the eddy are characterized by the deformation rate and the Okubo-Weiss parameter, wherein the shear deformation rate at the center pixel (x,y) of any template window is calculated as follows:

[0046]

[0047] in, This represents the partial derivative of the meridional component of the fused flow field with respect to x. This represents the partial derivative of the latitudinal component of the fused flow field with respect to y.

[0048] The formula for calculating the stretch deformation rate at the center pixel (x, y) of any template window is:

[0049]

[0050] The formula for calculating the total deformation rate is:

[0051]

[0052] The formula for calculating Okubo-Weiss parameters is:

[0053]

[0054] This invention also provides a system for extracting ocean eddy dynamic parameters from SAR and water color remote sensing images, comprising:

[0055] SAR image vortex flow field extraction module, used for extracting sea surface vortex flow fields from time-series SAR images;

[0056] The water color image vortex flow field extraction module is used for extracting sea surface vortex flow fields from time-series water color remote sensing images;

[0057] The flow field fusion module is used for fusing data from various sea surface vortex flow fields.

[0058] The ocean vortex dynamic parameter extraction module is used for extracting ocean vortex dynamic parameters.

[0059] The present invention has the following beneficial effects:

[0060] 1. This invention starts from the fusion of multi-source remote sensing images, solves the problem of data loss caused by cloud cover in water color remote sensing images, and obtains a complete vortex flow field.

[0061] 2. Starting from the fusion of multi-source remote sensing images, this invention solves the problem of feature loss in low-scattering regions of SAR images and obtains a complete vortex flow field.

[0062] 3. This invention improves the accuracy of extracting ocean eddy dynamic parameters by starting from the fusion of multi-source remote sensing images. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the process for extracting ocean vortex dynamic parameters from SAR and water color remote sensing images according to an embodiment of the present invention;

[0064] Figure 2 This is a detailed flowchart illustrating a method for extracting ocean vortex dynamic parameters from SAR and water color remote sensing images, as proposed in an embodiment of the present invention.

[0065] Figure 3(a) shows the region of interest extracted from an ENVISAT SAR image after preprocessing, according to an embodiment of the present invention.

[0066] Figure 3(b) shows the region of interest extracted from an ERS SAR image after preprocessing, according to an embodiment of the present invention.

[0067] Figure 3(c) shows the result of estimating and filtering the sea surface vortex current field from a time-series SAR image according to an embodiment of the present invention;

[0068] Figure 4(a) shows the region of interest extracted from a chlorophyll a (Chl-a) concentration image obtained by a Medium Resolution Imaging Spectrometer Instrument (MERIS) according to an embodiment of the present invention.

[0069] Figure 4(b) shows the region of interest extracted from a Chl-a image obtained by a Moderate-resolution Imaging Spectroradiometer (MODIS) according to an embodiment of the present invention.

[0070] Figure 4(c) shows the result of estimating and filtering the sea surface eddy current field from the time series Chl-a image according to an embodiment of the present invention;

[0071] Figure 5(a) is a result of smoothing the fused flow field according to an embodiment of the present invention, with an ENVISAT SAR image as the background;

[0072] Figure 5(b) is a result of smoothing the fused flow field according to an embodiment of the present invention, with a MERIS Chl-a image as the background.

[0073] Figure 6(a) is a diagram showing the results of extracting the maximum tangential velocity from an ocean vortex flow field according to an embodiment of the present invention;

[0074] Figure 6(b) is a diagram showing the vorticity field extraction result of an ocean vortex flow field according to an embodiment of the present invention;

[0075] Figure 6(c) is a diagram showing the results of extracting the divergence field from an ocean vortex flow field according to an embodiment of the present invention;

[0076] Figure 6(d) is a diagram showing the results of extracting eddy kinetic energy from an ocean vortex flow field according to an embodiment of the present invention.

[0077] Figure 7(a) is a diagram showing the results of extracting shear deformation rate from an ocean vortex flow field according to an embodiment of the present invention;

[0078] Figure 7(b) is a diagram showing the results of extracting the tensile deformation rate of an ocean vortex flow field according to an embodiment of the present invention;

[0079] Figure 7(c) shows the results of extracting the total deformation rate of an ocean vortex flow field according to an embodiment of the present invention;

[0080] Figure 7(d) is a diagram showing the results of extracting Okubo-Weiss parameters from an ocean vortex flow field according to an embodiment of the present invention. Detailed Implementation

[0081] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0082] When extracting ocean eddy dynamic parameters using remote sensing data, optical remote sensing images have certain limitations in detection and information acquisition, resulting in data loss in cloud-covered areas. SAR images are unaffected by weather, but in conditions such as algal blooms, low-wind-speed areas, and upwelling, SAR images exhibit large areas of low scattering, affecting the extraction of ocean eddy dynamic parameters.

[0083] Based on the above reasons, this invention discloses a method and system for extracting ocean eddy dynamic parameters from SAR and water color remote sensing images.

[0084] On the one hand, such as Figure 1 As shown, this invention provides a method for extracting ocean eddy dynamic parameters from SAR and water color remote sensing images, which includes the following steps:

[0085] Step 1: Acquire time-series SAR images and time-series water color remote sensing images;

[0086] Step 2: Register the time-series SAR image and the time-series water color remote sensing image;

[0087] Step 3: Estimate the sea surface eddy current field from time-series SAR images and time-series water color remote sensing images, respectively;

[0088] Step 4: Weightedly fuse the sea surface eddy current field of the time series SAR image with the eddy current field estimation results of the time series water color remote sensing image to obtain the fused flow field;

[0089] Step 5: Smooth the fused flow field and extract the ocean vortex dynamic parameters from it.

[0090] In some embodiments, chlorophyll a concentration (Chl-a) remote sensing image products from water color data are selected as time-series water color remote sensing images, i.e., time-series Chl-a images, which can be downloaded from publicly available data websites.

[0091] The time-series SAR images should consist of at least two scenes, and the time-series Chl-a images should also consist of at least two scenes. The time interval between time-series remote sensing images must be less than the decorrelation time of the image features; otherwise, mismatches are likely to occur when estimating sea surface eddy current fields. Generally, the time interval between time-series SAR images should not exceed one hour, and the time interval between time-series Chl-a images should be within 12 hours for better results.

[0092] After preprocessing, time-series SAR images and time-series Chl-a images are used to estimate the sea surface vortex current field in selected study areas. Taking the time-series SAR image as an example, the first SAR image in the time-series SAR image is divided into many small template windows. Any template window T is selected, and there may be partial overlap between template windows T.

[0093] The template window T performs matching within the search region of the second SAR image selected in the time-series SAR images, and calculates the cross-correlation coefficient for each match. The calculation formula is as follows:

[0094]

[0095] Where S′ represents the search sub-window corresponding to the template window within the search area, (x,y) are the center pixel coordinates of the template window, and p and q represent the displacements of the center of the search sub-window relative to the center of the search window in the distance and azimuth directions, respectively. This represents the average pixel value within the search sub-window. This represents the average pixel value within the template window.

[0096] After the template window has traversed the entire search area, the maximum value r(x,y) in the correlation coefficient matrix corresponding to the template window is obtained. The horizontal velocity vector of the vortex is estimated based on the positional offset of the maximum value relative to the center point.

[0097]

[0098] θ=tan -1 ((q max Δy) / (p max Δx)) (3)

[0099] Where Δt is the time interval of the time-series SAR images, Δx and Δy are the pixel intervals of the SAR images in the range and azimuth directions, respectively, c is the estimated sea surface vortex current field vector, d is the estimated sea surface displacement vector, and p max and q max θ represents the number of pixels that the template window T moves to the right and down within the search window when the maximum cross-correlation coefficient is obtained, respectively, and θ is the direction of the estimated sea surface vortex flow field vector.

[0100] Select the next template window and repeat the above operation until all template windows in the first SAR image have been traversed, and the sea surface vortex flow field of the time series SAR image can be obtained.

[0101] The acquired ocean vortex flow field contains some erroneous flow field vectors, therefore filtering is required to obtain valid vortex flow field vectors. The effective correlation coefficient w at any center pixel (x, y) of the template window after filtering is... n (x,y) can be represented as:

[0102]

[0103] Where, r n (x,y) represents the maximum cross-correlation coefficient corresponding to the nth flow field vector, m is the correlation coefficient threshold, and n is the number of flow fields at the center pixel of the template window. In this invention, n = 1, 2.

[0104] The sea surface eddy current field of the filtered time-series SAR image and the sea surface eddy current field of the time-series Chl-a image are weighted and fused using the correlation coefficient as the weight to obtain the fused flow field. The latitudinal and meridional components of the fused flow field at the center pixel (x,y) of any template window are as follows:

[0105]

[0106] Among them, u n (x,y) and v n (x,y) represent the nth latitudinal component and the longitudinal component at the center pixel (x,y) of the template window, respectively, where n is the number of flow fields at the center pixel of the template window;

[0107] To meet the principles of flow field continuity and local consistency, the merged flow field needs to be smoothed to obtain the complete sea surface flow field.

[0108]

[0109] in, It is an M-dimensional column vector.

[0110] Based on the aforementioned fused flow field, dynamic parameters of the ocean vortex are extracted. The vorticity of a vortex reflects the strength of its rotation; the vorticity is maximum at the vortex center and zero at the edges. The vorticity is calculated as follows: The vorticity field of the vortex is obtained from the fused flow field as follows:

[0111]

[0112] in, For Hamiltonian operators, ζ Ci (i = 1, 2, ..., N) is an M-dimensional column vector.

[0113] The vortex divergence field can reflect the convergence and divergence in the vortex velocity field. The divergence is calculated as follows: The divergence field of the vortex is obtained from the fused flow field as follows:

[0114]

[0115] Where, ξ Ci (i = 1, 2, ..., N) is an M-dimensional column vector.

[0116] The vortex kinetic energy KE of the vortex velocity field at the center pixel (x,y) of any template window is obtained from the fused flow field:

[0117] KE=(u 2 (x,y)+v 2 (x,y)) / 2 (9)

[0118] Where u(x,y) and v(x,y) represent the latitudinal and longitudinal components of the fused flow field at the center pixel (x,y) of the template window, respectively.

[0119] The characteristics of the vortex can also be characterized by the shear rate and the Okubo-Weiss parameter. The shear rate at the center pixel (x,y) of any template window is calculated as follows:

[0120]

[0121] in, This represents the partial derivative of the meridional component of the fused flow field with respect to x. This represents the partial derivative of the latitudinal component of the fused flow field with respect to y.

[0122] The formula for calculating the stretch deformation rate at the center pixel (x, y) of any template window is:

[0123]

[0124] The formula for calculating the total deformation rate is:

[0125]

[0126] The formula for calculating Okubo-Weiss parameters is:

[0127]

[0128] On the other hand, the present invention provides a system for extracting ocean eddy dynamic parameters from SAR and water color remote sensing images, the system comprising:

[0129] A SAR image vortex flow field extraction module is used for extracting sea surface vortex flow fields from time-series SAR images of the system.

[0130] A water color image vortex flow field extraction module is used for extracting the sea surface vortex flow field from the time-series water color remote sensing images of the system.

[0131] A flow field fusion module is used to fuse various flow field data of the system.

[0132] The ocean vortex dynamic parameter extraction module is used for extracting ocean vortex dynamic parameters of the system.

[0133] The following detailed description of the method and system for extracting ocean eddy dynamic parameters from SAR and water color remote sensing images proposed in this invention is provided through specific embodiments:

[0134] like Figure 2 As shown in Figures 3(a) and 3(b), firstly, ENVISAT and ERS SAR images are selected as the first and second SAR images, respectively. The time-series SAR images are registered and the study area is cropped, as shown in Figures 3(a) and 3(b). Large areas of low scattering are present in the images. The water color remote sensing images selected in this embodiment are from the MERIS water color sensor and the MODIS medium resolution imaging spectrometer Chl-a images, which are used as the first and second water color images, respectively. After preprocessing, the time-series Chl-a images are used to select the same region of interest as the time-series SAR images, as shown in Figures 4(a) and 4(b). Two areas of data loss due to cloud contamination are present in the images.

[0135] Then, the sea surface vortex flow field is estimated from the time series SAR image and the time series Chl-a image according to formulas (1)-(3), and filtered. The sea surface vortex flow field of the filtered time series SAR image is shown in Figure 3(c), and the sea surface vortex flow field of the filtered time series Chl-a image is shown in Figure 4(c).

[0136] Next, to obtain a more refined sea surface current field, the sea surface vortex current field of the time series Chl-a image was interpolated to the same latitude and longitude grid as the sea surface vortex current field of the time series SAR image. Then, the sea surface vortex current field of the time series SAR image and the interpolated sea surface vortex current field of the time series Chl-a image were fused according to formulas (4)-(5). In order to meet the principles of flow field continuity and local consistency, the fused flow field was smoothed. The smoothed results are shown in Figures 5(a) and 5(b), where Figures 5(a) and 5(b) are the flow field fusion results displayed with ENVISAT SAR image and MERIS Chl-a image as backgrounds, respectively. It can be seen from the figures that the sea surface vortex current field of the time series Chl-a image and the sea surface vortex current field of the time series SAR image complement each other, and the vortex structure at the bottom of the fused flow field is relatively complete.

[0137] Finally, the vortex dynamic parameters were extracted using the vortex at the bottom of the image (the boxed areas in Figures 5(a) and 5(b)). The extracted vortex dynamic parameters include the maximum tangential velocity, vorticity field, divergence field, vortex kinetic energy, shear deformation rate, tensile deformation rate, total deformation rate, and Okubo-Weiss parameters. Figure 6(a) shows the vortex velocity field, with the maximum tangential velocity of the vortex marked. Figure 6(b) shows the calculated vorticity field; the absolute value of vorticity is larger at the vortex center, indicating a greater degree of rotation. The positive or negative sign of vorticity indicates the direction of rotation. Figure 6(c) shows the calculated divergence field, reflecting the convergence (negative) and divergence (positive) of the vortex velocity field. Figure 6(d) shows the vortex kinetic energy; the center point of the figure indicates the location of the vortex center. Figure 7(a) shows the shear deformation rate of the vortex flow field, Figure 7(b) shows the tensile deformation rate of the vortex flow field, Figure 7(c) shows the total deformation rate of the vortex flow field, and Figure 7(d) shows the Okubo-Weiss parameter results.

[0138] This completes the extraction of ocean eddy dynamic parameters from SAR and water color remote sensing images.

[0139] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for extracting oceanic eddy kinetic parameters from SAR and ocean color remote sensing images, characterized in that, The method comprises the following steps: Step 1, obtaining time-series SAR images and time-series water color remote sensing images; Step 2, registering the time-series SAR images and the time-series water color remote sensing images; Step 3, respectively obtaining time-series SAR image sea surface vortex flow fields and time-series water color remote sensing image sea surface vortex flow fields; estimating the SAR image sea surface vortex flow field from the time-series SAR images, and estimating the water color image sea surface vortex flow field from the time-series water color remote sensing images; For the time-series SAR images, obtaining the SAR image sea surface vortex flow field comprises: The first SAR image in the time series SAR images is divided into a plurality of small template windows, and any template window is selected ; Template window The matching is performed in the search area of the second SAR image selected in the time series SAR images, and the cross-correlation coefficient is calculated, and the calculation formula is as follows: wherein, represents the corresponding search sub-window of the template window within the search window, is the center pixel coordinate of the template window, and respectively represent the displacement of the search sub-window center relative to the search window center in the distance direction and the azimuth direction, represents the pixel mean value within the search sub-window, represents the pixel mean value within the template window; obtaining a maximum value in a correlation coefficient matrix corresponding to the template window of the center pixel point coordinate obtaining a maximum value in a correlation coefficient matrix corresponding to the template window of the center pixel point coordinate estimating a vortex horizontal velocity vector according to a position offset of the maximum value relative to the center point wherein, is the time interval of the time series SAR images, and are the pixel intervals of the SAR images in the range and azimuth direction, respectively, is the estimated sea surface vorticity field vector, is the estimated sea surface displacement vector, and are the template window sizes at which the maximum cross-correlation coefficient is obtained is the number of pixel points moved rightward and downward within the search window, is the direction of the estimated sea surface vorticity field vector; The above operation is repeated by selecting the next template window until all the template windows in the first SAR image are traversed, and the time-series SAR image sea surface vortex flow field is obtained; Step 4, performing weighted fusion on the time-series SAR image sea surface vortex flow field and the time-series water color remote sensing image sea surface vortex flow field to obtain a fused flow field; Step 5, smoothing the fused flow field and extracting marine eddy dynamic parameters therefrom.

2. The method according to claim 1, wherein, In the step 1, the number of the obtained time-series SAR images is at least two, and the number of the obtained time-series water color remote sensing images is at least two.

3. The method of claim 1, wherein the method further comprises: The time interval of the time-series remote sensing images is less than the decorrelation time of image features.

4. The method of claim 1, wherein the method further comprises: The step 3 comprises: filtering the acquired time series SAR image sea surface vortex flow field and time series water color remote sensing image sea surface vortex flow field to obtain effective vortex flow field vectors, and the effective correlation coefficient at the center pixel point of any template window after filtering is represented as: wherein, represents the maximum cross-correlation coefficient corresponding to the is a correlation coefficient threshold value, is the number of flow field vectors at the center pixel point of the template window, ; and is the center pixel coordinate of the template window.​ 5. The method according to claim 4, wherein, In the step 4, the latitudinal component and the longitudinal component of the fused flow field at an arbitrary template window center pixel point are respectively: wherein, and respectively represent the first weft and warp components at the center pixel point of the template window, is the number of flow fields at the center pixel point of the template window, is the center pixel coordinate of the template window.

6. The method according to claim 5, wherein, In step 5, the fused flow field is smoothed to obtain a complete sea surface vortex flow field : wherein is an M-dimensional column vector.

7. The method according to claim 6, wherein, In the step 5, the marine eddy dynamic parameters comprise: The vortex degree of the vortex is used to reflect the strength of the vortex rotation, and the calculation method of the vortex degree is: The vortex degree field of the vortex is obtained from the fusion flow field as follows: wherein is a Hamiltonian operator, is an M-dimensional column vector.

8. The method according to claim 7, wherein, In the step 5, the marine eddy dynamic parameters comprise: The divergence of the vortex is used to reflect the convergence and divergence in the vortex velocity field, and the divergence calculation method is: The divergence field of the vortex is obtained from the fusion flow field as follows: wherein is an M-dimensional column vector.

9. The method according to claim 8, wherein, In the step 5, the marine eddy dynamic parameters comprise: The vortex energy of the vortex velocity field at the center pixel point of any template window is obtained by using the above fusion flow field :​ wherein, and respectively represent the zonal component and the meridional component of the fusion flow field at the center pixel point of the template window, is the center pixel coordinate of the template window.

10. The method of claim 1, wherein the method further comprises: In the step 5, the marine eddy dynamic parameters comprise: The characteristics of the vortex are characterized by the deformation rate and the Okubo-Weiss parameter, wherein the shear deformation rate at the center pixel point of any template window is calculated by the following formula: wherein, the shear deformation rate at the center pixel point of any template window is calculated by the following formula: wherein, denotes the pair of meridional components of the fused flow field is taken with respect to the parameter denotes the pair of zonal components of the fused flow field is taken with respect to the parameter The formula for calculating the stretch deformation rate at any template window center pixel point is: wherein, is the center pixel coordinate of the template window; The formula for calculating the total deformation rate is: The formula for calculating the Okubo-Weiss parameter is: 。 11. A system for implementing the method for extracting the dynamic parameters of marine eddies from SAR and ocean color remote sensing images according to any one of claims 1-10, characterized in that, comprise: The SAR image vortex flow field extraction module is used for time-series SAR image sea surface vortex flow field extraction; The water color image vortex flow field extraction module is used for time-series water color remote sensing image sea surface vortex flow field extraction; The flow field fusion module is used for data fusion of each sea surface vortex flow field; The marine eddy dynamic parameter extraction module is used for marine eddy dynamic parameter extraction.

Images