Method, device and electronic equipment for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images

Abstract

The present invention provides a method, device and electronic device for extracting ocean ice vortex dynamic parameters from dual-polarization SAR images, and relates to the field of marine engineering technology. The method comprises: obtaining continuous time series of HH polarization and HV polarization SAR image ocean ice vortex data; SAR image preprocessing; feature tracking and abnormal feature point filtering; pattern matching and feature fusion; ice vortex dynamic parameter extraction. The present invention realizes automatic extraction of ocean ice vortex dynamic parameters from dual-polarization SAR images without human intervention, and high-resolution ice vortex dynamic parameters are crucial for a better understanding of ice vortex dynamics.

Description

Technical Field

[0001] The invention relates to the technical field of marine engineering, and in particular to a method, a device and electronic equipment for extracting ocean ice vortex dynamic parameters from dual-polarization SAR images. Background Art

[0002] Ocean eddies are a common marine phenomenon that are active near the coast. They play an important role in the exchange of materials and energy in the ocean and are an important object of marine science research. In high-latitude polar regions, due to the presence of sea ice, the traditional method of using altimeter data to observe eddies is complicated, which makes this data type almost unsuitable for ice eddy detection in high-latitude polar regions. Since the sea ice on the surface of the ice eddies presents the characteristics of ocean current structure, the surface characteristics of eddies in the marginal ice zone can be effectively tracked in high-resolution SAR images. The MIZ dynamics in SAR images are caused by the different radar backscattering intensities between open water and low-concentration sea ice, as well as the accumulation of sea ice at the eddy boundary along the surface current convergence zone. In areas with low sea ice concentration, sea ice tends to accumulate in major cyclonic eddies, and the sea ice velocity is close to the surface ocean current velocity, allowing people to deduce the upper ocean eddy dynamics parameters from SAR image sea ice observations.

[0003] Therefore, the analysis of ocean ice eddies based on SAR images is currently a research hotspot. Summary of the invention

[0004] In view of this, the object of the present invention is to provide a method, device and electronic equipment for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images, so as to realize the automatic extraction of ocean ice eddy dynamic parameters in high-latitude polar regions.

[0005] A method for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images comprises the following steps:

[0006] S1: Obtain continuous time series of dual-polarization SAR image ocean ice eddy data, and decode them to obtain HH and HV dual-polarization data;

[0007] S2: SAR image preprocessing, including orbit correction, HH polarization incident angle correction, HV polarization thermal noise removal, radiation calibration, adaptive Wiener filtering, decibel processing, normalization processing, and image downsampling processing for SAR ocean ice eddy images;

[0008] S3: Feature tracking and abnormal feature point filtering. First, the ORB (Oriented Fast and Rotated Brief) algorithm is used to detect feature points on the normalized HH polarization and HV polarization images respectively to obtain sparse feature points. Secondly, abnormal feature points are filtered out.

[0009] S4: Pattern matching and feature fusion: Based on S3, further refined pattern matching and dual-polarization feature fusion are performed to obtain a more accurate ice vortex motion vector;

[0010] S5: Extraction of ice vortex dynamic parameters.

[0011] Furthermore, the step S1 is specifically as follows:

[0012] The continuous time series of dual-polarization SAR image ocean ice eddy data: obtain the Sentienl-1A / 1B satellite EW mode L1GRDM continuous time series of dual-polarization SAR image ocean ice eddy data, the time interval of the continuous time series SAR images is less than 1 hour, and the spatial coverage overlap rate is greater than 0.3; decode the continuous time series dual-polarization SAR images that meet the requirements to obtain HH and HV dual-polarization data.

[0013] Furthermore, in step S2, the SAR image preprocessing includes:

[0014] (1) Adaptive Wiener filtering: A 5×5 adaptive Wiener filter is used for smoothing. The formula is as follows:

[0015] g(x,y)=f(x,y)*h(x,y)+n(x,y)

[0016]

[0017] Among them, f(x,y) is the input original SAR image, h(x,y) is the fuzzy degradation matrix, n(x,y) is the noise signal added later, g(x,y) is the sampled signal after fuzzy processing, and w(x,y) is the filtering operating system. is the estimated output signal;

[0018] (2) Decibel processing: Calculation of standard radar backscatter cross section σ 0 , the formula is as follows:

[0019]

[0020] Among them, σ 0 represents the normalized radar backscatter cross section, DN represents the brightness value of the pixel, which reflects the reflection characteristics of the target object, and A represents the effective area of ​​the radar antenna receiving the signal;

[0021] (3) Normalization processing: σ 0 The value is normalized to the range of 0 to 255 and converted from a 32-bit floating point number to an 8-bit integer data. The conversion formula is as follows:

[0022]

[0023] in, and are the σ of the HH and HV polarization images respectively. 0 The maximum and minimum values ​​of , Z represents the set of integers, that is The value must be an integer;

[0024] (4) Image downsampling: The SAR image is downsampled twice in the horizontal x and vertical y directions. The pixel resolution after downsampling is 80m×80m.

[0025] Furthermore, in step S3, the feature tracking and abnormal feature point filtering are specifically as follows:

[0026] First, the ORB algorithm is used to track the features of the preprocessed HH polarization and HV polarization images respectively to obtain sparse feature points:

[0027] The first step is to find the key points in the image: determine the value of the threshold parameter h for the selected feature points. For any pixel point p on the image, FAST compares the 16 pixels in the circle with point p as the center. If the gray value of the circle is less than l p -h(l p That is, the gray value of point p) or the gray value is greater than l p If there are more than 8 pixels with +h, pixel p is selected as the key point.

[0028] The second step is to use Harris corner point scoring to screen feature points. The Harris response function is defined as:

[0029] R=det(M)-k·trace 2 (M);

[0030]

[0031] Where M is the autocorrelation matrix:

[0032]

[0033] I x and I y are the gradients of the image pixels in the horizontal x and vertical y directions, respectively. k is an empirical parameter, generally ranging from 0.04 to 0.06. det(M) is the determinant of the matrix M, and trace(M) is the trace of the matrix M.

[0034] The third step is to use the Hamming distance for feature matching. For each key point in the first image, the ratio of the minimum and second minimum Hamming distances between the key points in the second image is calculated. If the ratio is lower than the threshold of 0.7, the key point with the minimum Hamming distance is considered a match.

[0035] Secondly, abnormal feature points are filtered out: (1) If the maximum drift speed F is exceeded 1 m / s is considered as an error vector and is removed. Maximum ice drift F 1 Set to 0.5m / s; (2) Filter out inconsistent drift vectors. 1 and 1 By x 2 and 2 Apply a quadratic polynomial function and use the least squares method to approximate and compare the approximate coordinates with the actual coordinates over the simulated point F 2 km drift vectors will be filtered out, with a maximum drift distance of F 2 Set to 25km.

[0036]

[0037] in, and are the coordinates of the point extrapolated from image 2, A i,j and B i,j is the polynomial coefficient obtained by applying the least squares method to all valid feature point matches. i and j are the power exponents in the polynomial terms, indicating the x of the corresponding terms. 1 and 1 n represents the highest degree of the polynomial, and N represents the limit on the number of valid terms in the polynomial.

[0038] Finally, the characteristic points of HH polarization and HV polarization images and the preliminary displacement estimation of ice vortex are obtained.

[0039] Furthermore, in step S4, pattern matching and feature fusion are specifically as follows:

[0040] On the basis of S3, further refined pattern matching is performed and dual-polarization feature fusion is performed to obtain a more accurate ice vortex motion vector.

[0041] First, the maximum cross-correlation algorithm (MCC) is used to determine the best matching area. The MCC calculation formula is as follows:

[0042]

[0043] Among them, f(x, y) and g(x, y) are the reference image, the image to be matched and the calculation template respectively, D x,y is the variance of f(x,y), and D is the variance of g(x,y).

[0044] Secondly, calculate the speed v of sea ice drift speed 、Eastward component of sea ice drift velocity U ice, the vertical northward component of sea ice drift velocity V ice , obtained by the following formula:

[0045] d x =x 2 -x 1 ;

[0046] d y =y 2 -y 1 ;

[0047]

[0048] Among them, d x and d y are the horizontal and vertical displacement components of sea ice (in pixels), with a resolution of 80 m; (x 1 ,y 1 ) and (x 2 ,y 2 ) are respectively the time t 1 and t 2 The position of the target in the two images; t is the time interval for monitoring sea ice drift (t 2 ,t 1 ).

[0049] Finally, the ice vortex drift vectors obtained by HH polarization and HV polarization are fused to obtain a more accurate dual-polarization ice vortex motion vector.

[0050] Furthermore, in step S5, the ice vortex dynamic parameter extraction is specifically as follows:

[0051] According to the horizontal eastward component of sea ice drift velocity U ice , the vertical component of sea ice drift velocity V ice The specific calculation formulas for the dynamic parameters of ice vortex such as relative vorticity, normalized vorticity, instantaneous kinetic energy, and divergence are as follows:

[0052] (1) Relative vorticity ζ ice :

[0053]

[0054] Among them, ice is the relative vorticity, U ice and V ice represent the horizontal eastward (x direction) component and the vertical component (y direction) of the sea ice drift speed, respectively. Indicates U ice The rate of change in the x direction, Indicates V ice The rate of change in the y direction.

[0055] (2) Normalized vorticity

[0056]

[0057] Where f is the Coriolis force parameter, ζ ice is the relative vorticity.

[0058] (3) Ice vortex instantaneous kinetic energy KE ice :

[0059]

[0060] Among them, KE ice is the instantaneous kinetic energy of the ice vortex, U ice and V ice They represent the horizontal eastward (x direction) and vertical (y direction) components of the sea ice drift speed, respectively.

[0061] (4) Ice vortex divergence field

[0062]

[0063] in, represents the velocity field, U ice and V ice represent the horizontal eastward (x direction) component and the vertical component (y direction) of the sea ice drift speed, respectively. Indicates U ice The rate of change in the x direction, Indicates V ice The rate of change in the y direction.

[0064] A device for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images, the device comprising:

[0065] The continuous time series dual-polarization SAR image acquisition module is used to obtain HH and HV dual-polarization SAR ocean ice eddy images in the high-latitude polar sea ice edge area;

[0066] SAR image preprocessing module, used to realize SAR ocean ice eddy image orbit correction, HH polarization incident angle correction, HV polarization thermal noise removal, radiation calibration, adaptive Wiener filtering, decibel processing, image normalization processing, and obtain dual-polarization SAR image after data preprocessing;

[0067] Feature tracking and abnormal feature point filtering module, used to detect ice vortex feature points and filter abnormal feature points in dual-polarization SAR images;

[0068] Pattern matching and feature fusion module, used for refined pattern matching and dual-polarization feature fusion to obtain more accurate ice vortex motion vector;

[0069] The ice vortex dynamic parameter extraction module is used to automatically extract ice vortex dynamic parameters, including relative vorticity, normalized vorticity, divergence and kinetic energy.

[0070] An electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0071] A computer-readable storage medium stores a computer program, wherein the computer program executes the steps of the above method when executed by a processor.

[0072] Advantages and beneficial effects of the present invention:

[0073] According to the response of sea ice drift to the passage of mesoscale eddies, the present invention proposes a new method for extracting ice eddy dynamic parameters, which includes acquiring continuous time series of HH polarization and HV polarization SAR image ocean ice eddy data; SAR image preprocessing; feature tracking and abnormal feature point filtering; pattern matching and feature fusion; and ice eddy dynamic parameter extraction, and finally obtaining the ice eddy dynamic parameter results.

[0074] The present invention can effectively realize the automatic extraction of ocean vortex dynamic parameters in high-latitude polar marginal ice zones. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 A flow chart of extracting ocean ice eddy dynamic parameters from dual-polarization SAR images provided by an embodiment of the present invention.

[0076] Figure 2 A schematic diagram of a device for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images provided in an embodiment of the present invention.

[0077] Figure 3 A schematic diagram of an electronic device for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images provided by an embodiment of the present invention.

[0078] Figure 4 This is a dual-polarization feature point fusion effect diagram provided by an embodiment of the present invention. The red dots represent the feature points extracted by HH polarization, and the yellow dots represent the feature points extracted by HV. The features are superimposed on the HH polarization SAR image.

[0079] Figure 5 An embodiment of the present invention provides an ice vortex motion vector field based on a Sentinel-1 dual-polarization SAR image and superimposed HH polarization SAR image.

[0080] Figure 6 It is an ice vortex velocity field based on Sentinel-1 dual-polarization SAR image provided by an embodiment of the present invention.

[0081] Figure 7 It is an ice vortex normalized vorticity map based on Sentinel-1 dual-polarization SAR image provided by an embodiment of the present invention.

[0082] Figure 8 It is an ice vortex kinetic energy map based on Sentinel-1 dual-polarization SAR image provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0083] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0084] Embodiment 1:

[0085] This embodiment provides a method for extracting ocean ice eddy parameters from dual-polarization SAR images. Figure 1 As shown, the method for extracting ocean ice vortex dynamic parameters includes the following steps:

[0086] Step S101, continuous time series dual-polarization SAR image ocean ice vortex data acquisition

[0087] The continuous time series of dual-polarization SAR image ocean ice eddy data: obtain the continuous time series of dual-polarization SAR image ocean ice eddy data of Sentienl-1A / 1B satellite EW mode L1GRDM, the time interval of the continuous time series SAR images is less than 1 hour, and the spatial coverage overlap rate is greater than 0.3. Decode the continuous time series dual-polarization SAR images that meet the requirements to obtain HH and HV dual-polarization data.

[0088] Step S102, SAR ocean ice vortex image preprocessing: orbit correction, HH polarization incident angle correction, HV polarization thermal noise removal, radiation calibration, adaptive Wiener filtering, decibel processing, normalization processing, image downsampling processing are performed on the SAR ocean ice vortex image. The specific steps of the adaptive Wiener filtering, decibel processing, normalization processing, and image downsampling processing are as follows:

[0089] (1) Adaptive Wiener filtering: A 5×5 adaptive Wiener filter is used for smoothing. The formula is as follows:

[0090] g(x,y)=f(x,y)*h(x,y)+n(x,y)

[0091]

[0092] Among them, f(x,y) is the input original SAR image, h(x,y) is the fuzzy degradation matrix, n(x,y) is the noise signal added later, g(x,y) is the sampled signal after fuzzy processing, and w(x,y) is the filtering operating system. is the estimated output signal.

[0093] (2) Decibel processing: Calculation of standard radar backscatter cross section σ 0 , the formula is as follows:

[0094]

[0095] Among them, σ 0 It represents the normalized radar backscatter cross section, DN represents the brightness value of the pixel, which reflects the reflection characteristics of the target object. A represents the effective area of ​​the radar antenna receiving the signal.

[0096] (3) Normalization processing: σ 0 The value is normalized to the range of 0 to 255 and converted from a 32-bit floating point number to an 8-bit integer data. The conversion formula is as follows:

[0097]

[0098] in, and are the σ of the HH and HV polarization images respectively. 0 The maximum and minimum values ​​of . Z represents the set of integers, that is The value must be an integer.

[0099] (4) Image downsampling: The SAR image is downsampled twice in the horizontal x and vertical y directions. The pixel resolution after downsampling is 80m×80m.

[0100] Step S103, the specific steps of feature tracking and abnormal feature point filtering are as follows:

[0101] First, the ORB algorithm is used to track the features of the preprocessed HH polarization and HV polarization images respectively to obtain sparse feature points:

[0102] The first step is to find the key points in the image: determine the value of the threshold parameter h for the selected feature points. For any pixel point p on the image, FAST compares the 16 pixels in the circle with point p as the center. If the gray value of the circle is less than l p -h(l p That is, the gray value of point p) or the gray value is greater than l p If there are more than 8 pixels with +h, pixel p is selected as the key point.

[0103] The second step is to use Harris corner point scoring to screen feature points. The Harris response function is defined as:

[0104] R=det(M)-k·trace 2 (M);

[0105]

[0106] Where M is the autocorrelation matrix:

[0107]

[0108] I x and I y are the gradients of the image pixels in the horizontal x and vertical y directions, respectively. k is an empirical parameter, generally ranging from 0.04 to 0.06. det(M) is the determinant of the matrix M, and trace(M) is the trace of the matrix M.

[0109] The third step is to use the Hamming distance for feature matching. For each key point in the first image, the ratio of the minimum and second minimum Hamming distances between the key points in the second image is calculated. If the ratio is lower than the threshold of 0.7, the key point with the minimum Hamming distance is considered a match.

[0110] Secondly, abnormal feature points are filtered out: (1) If the maximum drift speed F is exceeded 1 m / s is considered as an error vector and is removed. Maximum ice drift F 1 Set to 0.5m / s; (2) Filter out inconsistent drift vectors. 1 and 1 By x 2 and 2 Apply a quadratic polynomial function and use the least squares method to approximate and compare the approximate coordinates with the actual coordinates over the simulated point F 2 km drift vectors will be filtered out, with a maximum drift distance of F 2 Set to 25km.

[0111]

[0112] in, and are the coordinates of the point extrapolated from image 2, A i,j and B i,j is the polynomial coefficient obtained by applying the least squares method to all valid feature point matches. i and j are the power exponents in the polynomial terms, indicating the x of the corresponding terms. 1 and 1 n represents the highest degree of the polynomial, and N represents the limit on the number of valid terms in the polynomial.

[0113] Finally, the characteristic points of HH polarization and HV polarization images and the preliminary displacement estimation of ice vortex are obtained.

[0114] Step S104, the pattern matching and feature fusion, based on S3, further refines the pattern matching and performs dual-polarization feature fusion to obtain a more accurate ice vortex motion vector. The specific steps are as follows:

[0115] First, the maximum cross-correlation algorithm (MCC) is used to determine the best matching area. The MCC calculation formula is as follows:

[0116]

[0117] Among them, f(x, y) and g(x, y) are the reference image, the image to be matched and the calculation template respectively, D x,y is the variance of f(x,y), and D is the variance of g(x,y).

[0118] Secondly, calculate the speed v of sea ice drift speed 、Eastward component of sea ice drift velocity U ice , the vertical northward component of sea ice drift velocity V ice , can be obtained by the following formula:

[0119] d x =x 2 -x 1 ;

[0120] d y =y 2 -y 1 ;

[0121]

[0122] Among them, d x and d y are the horizontal and vertical displacement components of sea ice (in pixels), with a resolution of 80 m; (x 1 ,y 1 ) and (x 2 ,y 2 ) are respectively the time t 1 and t 2 The position of the target in the two images; t is the time interval for monitoring sea ice drift (t 2 ,t 1 ).

[0123] Finally, the ice vortex drift vectors obtained by HH polarization and HV polarization are fused to obtain a more accurate dual-polarization ice vortex motion vector.

[0124] Step S105, extracting ice vortex dynamic parameters, based on the horizontal eastward component of sea ice drift velocity U ice , the vertical component of sea ice drift velocity V ice The specific calculation formulas for the dynamic parameters of ice vortex such as relative vorticity, normalized vorticity, instantaneous kinetic energy, and divergence are as follows:

[0125] (1) Relative vorticity ζ ice :

[0126]

[0127] Among them, ice is the relative vorticity, U ice and V ice represent the horizontal eastward (x direction) component and the vertical component (y direction) of the sea ice drift speed, respectively. Indicates U ice The rate of change in the x direction, Indicates V ice The rate of change in the y direction.

[0128] (2) Normalized vorticity

[0129]

[0130] Where f is the Coriolis force parameter, ζ ice is the relative vorticity.

[0131] (3) Ice vortex instantaneous kinetic energy KE ice :

[0132]

[0133] Among them, KE ice is the instantaneous kinetic energy of the ice vortex, U ice and V ice They represent the horizontal eastward (x direction) and vertical (y direction) components of the sea ice drift speed, respectively.

[0134] (4) Ice vortex divergence field

[0135]

[0136] in, represents the velocity field, U ice and V ice represent the horizontal eastward (x direction) component and the vertical component (y direction) of the sea ice drift speed, respectively. Indicates U ice The rate of change in the x direction, Indicates V iceThe rate of change in the y direction.

[0137] Embodiment 2:

[0138] This embodiment provides a device for extracting ocean ice eddy parameters from dual-polarization SAR images, and its structural diagram is shown in FIG. Figure 2 As shown, wherein the device includes:

[0139] The continuous time series dual-polarization SAR image acquisition module 710 is used to acquire HH and HV dual-polarization SAR ocean ice eddy images of the sea ice edge area in the high-latitude polar region;

[0140] SAR image preprocessing module 720, used to implement SAR ocean ice eddy image orbit correction, HH polarization incident angle correction, HV polarization thermal noise removal, radiation calibration, adaptive Wiener filtering, decibel processing, image normalization processing, and obtain dual-polarization SAR image after data preprocessing;

[0141] The feature tracking and abnormal feature point filtering module 730 is used to detect ice vortex feature points and filter abnormal feature points in dual-polarization SAR images;

[0142] The pattern matching and feature fusion module 740 is used for fine pattern matching and dual-polarization feature fusion to obtain a more accurate ice vortex motion vector;

[0143] The ice vortex dynamic parameter extraction module 750 is used to automatically extract ice vortex dynamic parameters, including relative vorticity, normalized vorticity, divergence and kinetic energy.

[0144] The device for extracting dynamic parameters of ocean ice eddies from dual-polarized SAR images provided in the embodiment of the present invention has the same technical features as the method for extracting dynamic parameters of ocean ice eddies from dual-polarized SAR images provided in the above embodiment, so it can also solve the same technical problems and achieve the same technical effects. For the sake of brief description, for matters not mentioned in the embodiment part, reference can be made to the corresponding contents in the above method embodiment.

[0145] Embodiment 3:

[0146] This embodiment also provides an electronic device, including a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the method described in the above embodiment 1, and the processor is configured to execute the program stored in the memory.

[0147] See also Figure 3 This embodiment also provides an electronic device 100, including: a processor 50, a memory 51, a bus 52 and a communication interface 53, wherein the processor 50, the communication interface 53 and the memory 51 are connected via the bus 52; the processor 50 is used to execute an executable module stored in the memory 51, such as a computer program.

[0148] The memory 51 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 53 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0149] The bus 52 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0150] Among them, the memory 51 is used to store programs, and the processor 50 executes the program after receiving the execution instruction. The method executed by the device for flow process definition disclosed in any embodiment of the above-mentioned embodiment of the present invention can be applied to the processor 50 or implemented by the processor 50.

[0151] The processor 50 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 50. The above processor 50 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in the embodiments of the present invention can be directly embodied as a hardware decoding processor to be executed, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 51, and the processor 50 reads the information in the memory 51 and completes the steps of the above method in combination with its hardware.

[0152] Embodiment 4:

[0153] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the above embodiment 1 are executed.

[0154] Embodiment 5:

[0155] This embodiment conducts verification experiments based on the methods and devices provided in the above embodiments.

[0156] Two consecutive Sentinel-1A and Sentinel-1A1B images acquired over the Fram Strait on September 17, 2017 were selected, with a time interval of 48 minutes between the two images. Dual-polarization SAR images were used to track feature points and remove outliers to obtain valid feature points (such as Figure 4 Through refined pattern matching and dual-polarization feature fusion, the ice vortex motion velocity field is obtained (as shown in Figure 5 and Figure 6Further calculation of ice vortex dynamic parameters, such as ice vortex normalized vorticity diagram (such as Figure 7 As shown in), ice vortex kinetic energy (as shown in Figure 8 As shown in the above figure, it can be seen that the method and device provided by the present invention can effectively realize the automatic extraction of ocean eddy dynamic parameters in high-latitude polar marginal ice zones, and provide high-resolution SAR image ice eddy motion parameter estimation, which is crucial for the study of mesoscale and sub-mesoscale dynamic processes in polar oceans.

[0157] The above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, it should be understood by those skilled in the art that any person familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or perform equivalent replacements on some of the technical features thereof; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images, characterized in that: The following steps are involved: S1: Obtain continuous time series of dual-polarization SAR image ocean ice eddy data, and decode them to obtain HH and HV dual-polarization data; S2: SAR image preprocessing, including orbit correction, HH polarization incident angle correction, HV polarization thermal noise removal, radiation calibration, adaptive Wiener filtering, decibel processing, normalization processing, and image downsampling processing for SAR ocean ice eddy images; S3: Feature tracking and abnormal feature point filtering. First, the ORB algorithm is used to detect feature points on the normalized HH polarization and HV polarization images to obtain sparse feature points. Secondly, abnormal feature points are filtered out. S4: Pattern matching and feature fusion: Based on S3, further refined pattern matching and dual-polarization feature fusion are performed to obtain a more accurate ice vortex motion vector; the refined pattern matching and dual-polarization feature fusion are specifically as follows: First, the maximum cross-correlation algorithm MCC is used to determine the best matching area. The MCC calculation formula is as follows: Among them, f(x, y) and g(x, y) are the reference image, the image to be matched and the calculation template respectively, D x，y is the variance of f(x, y), and D is the variance of g(x, y); Secondly, calculate the speed v of sea ice drift speed 、Eastward component of sea ice drift velocity U ice , the vertical northward component of sea ice drift velocity V ice , obtained by the following formula: d x =x2-x1; d y =y2-y1; Among them, d x and d y are the horizontal and vertical displacement components of sea ice, respectively, with a resolution of 80 m; (x1, y1) and (x2, y2) are the positions of the target at time t1 and t2 in the two images, respectively; t is the time interval for monitoring sea ice drift (t2, t1); Finally, the ice vortex drift vectors obtained by HH polarization and HV polarization are fused to obtain a more accurate dual-polarization ice vortex motion vector. S5: Extraction of ice vortex dynamic parameters.

2. The method for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images as claimed in claim 1, characterized in that: The step S1 specifically comprises: the continuous time series of dual-polarization SAR image ocean ice eddy data: obtaining the Sentienl-1A / 1B satellite EW mode L1 GRDM continuous time series of dual-polarization SAR image ocean ice eddy data, the time interval of the continuous time series SAR images is less than 1 hour, and the spatial coverage overlap rate is greater than 0.3; decoding the continuous time series of dual-polarization SAR images that meet the requirements to obtain HH and HV dual-polarization data.

3. The method for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images as claimed in claim 1, characterized in that: In step S2, the SAR image preprocessing includes: (1) Adaptive Wiener filtering: A 5×5 adaptive Wiener filter is used for smoothing. The formula is as follows: g(x,y)=f(x,y)*h(x,y)+n(x,y) Among them, f(x, y) is the input original SAR image, h(x, y) is the fuzzy degradation matrix, n(x, y) is the noise signal added later, g(x, y) is the sampled signal after fuzzy processing, and w(x, y) is the filtering operating system. is the estimated output signal; (2) Decibel processing: Calculation of standard radar backscatter cross section σ 0 , the formula is as follows: Among them, σ 0 represents the normalized radar backscatter cross section, DN represents the brightness value of the pixel, and A represents the effective area of ​​the radar antenna receiving the signal; (3) Normalization processing: σ 0 The value is normalized to the range of 0 to 255 and converted from a 32-bit floating point number to an 8-bit integer data. The conversion formula is as follows: in, and are the σ of the HH and HV polarization images respectively. 0 The maximum and minimum values ​​of Z are integer sets, that is, The value must be an integer; (4) Image downsampling: The SAR image is downsampled twice in the horizontal x and vertical y directions. The pixel resolution after downsampling is 80m×80m.

4. The method for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images as claimed in claim 1, characterized in that: In step S3, the feature tracking and abnormal feature point filtering include: Firstly, the ORB algorithm is used to track the features of the preprocessed HH polarization and HV polarization images respectively to obtain sparse feature points; Secondly, abnormal feature points are filtered out; Finally, the characteristic points of HH polarization and HV polarization images and the preliminary displacement estimation of ice vortex are obtained.

5. The method for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images as claimed in claim 4, characterized in that: The ORB algorithm is used to perform feature tracking on the preprocessed HH polarization and HV polarization images respectively, including: The first step is to find the key points in the image: determine the value of the threshold parameter h for the selected feature points. For any pixel point p on the image, FAST compares the 16 pixels in the circle with point p as the center. If the grayscale value of the surrounding 16 pixels is less than l p -h,l p That is, the gray value of point p, or the gray value is greater than l p +h has more than 8 pixels in total, then pixel p is selected as the key point; The second step is to use Harris corner point scoring to screen feature points. The Harris response function is defined as: R=det(M)-k trace 2 (M); Where M is the autocorrelation matrix: I x and I y are the gradients of the image pixels in the horizontal x and vertical y directions, respectively. k is an empirical parameter. det(M) is the determinant of the matrix M. trace(M) is the trace of the matrix M. The third step is to use the Hamming distance for feature matching. For each key point in the first image, the ratio of the minimum and second minimum Hamming distances between the key points in the second image is calculated. If the ratio is lower than the threshold of 0.7, the key point with the minimum Hamming distance is considered to be matched. The abnormal feature point filtering includes: (1) if the maximum drift speed F1m / s is exceeded, it is regarded as an error vector and is removed; the maximum ice drift F1 is set to 0.5m / s; (2) inconsistent drift vectors are filtered out; x1 and y1 are approximated by applying a quadratic polynomial function to x2 and y2 and using the least squares method to compare the approximate coordinates with the actual coordinates. Drift vectors exceeding the simulation point F2km will be filtered out, and the maximum drift distance F2 is set to 25km; in, and are the coordinates of the extrapolated point, A i，j and B i，j are the polynomial coefficients obtained by applying the least squares method to all valid feature point matches; i and j are the power exponents in the polynomial terms, representing the powers of x1 and y1 of the corresponding terms; n represents the highest degree of the polynomial, and N represents the limit on the number of valid terms in the polynomial.

6. The method for extracting ocean ice eddy dynamic parameters from dual-polarization SAR images as claimed in claim 1, characterized in that: In step S5, the ice vortex dynamic parameter extraction is specifically as follows: according to the horizontal eastward component U of the sea ice drift velocity ice , vertical component of sea ice drift velocity V ice , the specific calculation formula is as follows: (1) Relative vorticity ζ ice : Among them, ice is the relative vorticity, U ice and V ice represent the horizontal and vertical components of the sea ice drift velocity, respectively. Indicates U ice The rate of change in the x direction, Indicates V ice The rate of change in the y direction; (2) Normalized vorticity Where f is the Corioli force parameter, ζ ice is the relative vorticity; (3) Ice vortex instantaneous kinetic energy KE ice : Among them, KE ice is the instantaneous kinetic energy of the ice vortex, U ice and V ice Represent the horizontal and vertical components of sea ice drift velocity respectively; (4) Ice vortex divergence field in, represents the velocity field, U ice and V ice represent the horizontal and vertical components of the sea ice drift velocity, respectively. Indicates U ice The rate of change in the x direction, Indicates V ice The rate of change in the y direction.

7. An extraction device based on the method for extracting ocean ice eddy parameters from dual-polarization SAR images according to claim 1, characterized in that: The device includes: The continuous time series dual-polarization SAR image acquisition module is used to obtain HH and HV dual-polarization SAR ocean ice eddy images in the high-latitude polar sea ice edge area; SAR image preprocessing module, used to realize SAR ocean ice eddy image orbit correction, HH polarization incident angle correction, HV polarization thermal noise removal, radiation calibration, adaptive Wiener filtering, decibel processing, image normalization processing, and obtain dual-polarization SAR image after data preprocessing; Feature tracking and abnormal feature point filtering module, used to detect ice vortex feature points and filter abnormal feature points in dual-polarization SAR images; Pattern matching and feature fusion module, used for refined pattern matching and dual-polarization feature fusion to obtain more accurate ice vortex motion vector; The ice vortex dynamic parameter extraction module is used to automatically extract ice vortex dynamic parameters, including relative vorticity, normalized vorticity, divergence and kinetic energy.

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory is used to store a program for supporting the processor to execute the method according to any one of claims 1 to 6, and the processor is configured to execute the program stored in the memory.

9. A computer-readable storage medium, characterized in that: A computer program is stored on a computer-readable storage medium, wherein the computer program executes the steps of the method according to any one of claims 1 to 6 when the computer program is executed by a processor.

Images