CN113156439B - SAR wind field and sea wave joint inversion method and system based on data driving - Google Patents
SAR wind field and sea wave joint inversion method and system based on data driving Download PDFInfo
- Publication number
- CN113156439B CN113156439B CN202110500084.5A CN202110500084A CN113156439B CN 113156439 B CN113156439 B CN 113156439B CN 202110500084 A CN202110500084 A CN 202110500084A CN 113156439 B CN113156439 B CN 113156439B
- Authority
- CN
- China
- Prior art keywords
- sar
- wave
- data
- data set
- average
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 20
- 230000006870 function Effects 0.000 claims description 22
- 238000001228 spectrum Methods 0.000 claims description 21
- 238000004590 computer program Methods 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 12
- 238000010276 construction Methods 0.000 claims description 11
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 4
- 238000000638 solvent extraction Methods 0.000 claims 2
- 238000001514 detection method Methods 0.000 description 11
- 238000004836 empirical method Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 230000004913 activation Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000004106 carminic acid Substances 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
- G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
- G01S13/9021—SAR image post-processing techniques
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/95—Radar or analogous systems specially adapted for specific applications for meteorological use
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/95—Radar or analogous systems specially adapted for specific applications for meteorological use
- G01S13/958—Theoretical aspects
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention belongs to the field of wind field and wave inversion, and provides a data-driven SAR wind field and wave joint inversion method and system. The method comprises the following steps: acquiring SAR data to obtain a plurality of small images in the SAR data; extracting a backscattering coefficient, a normalized variance, an incidence angle and a frequency domain characteristic quantity of the small image, and constructing an input data set; acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the small image in a space-time mode, and constructing an output data set based on the matched wind speed, the effective wave height and the average wave period; inputting the input data set and the output data set into a convolutional neural network model for training to obtain a trained convolutional neural network model; and inputting the SAR data of the sea area to be detected into the trained convolutional neural network model to obtain the wind speed, the effective wave height and the average wave period of the sea area to be detected.
Description
Technical Field
The invention belongs to the technical field of wind field and wave inversion, and particularly relates to a data-driven SAR wind field and wave joint inversion method and system.
Background
The statements in this section merely provide background information related to the present disclosure and may not constitute prior art.
Sea surface wind fields and sea waves are important ocean dynamic process phenomena, play an important role in material and energy exchange between upper ocean processes and an ocean-air interface, and are the ocean phenomena most directly related to human activities. The method has the advantages that observation information of sea surface wind fields and sea waves is comprehensively and systematically acquired, the law of the observation information is mastered, and the method has important significance for ocean science research, disaster prevention and reduction, national defense construction and the like.
The existing detection modes of sea surface wind field and sea wave information comprise field detection, numerical prediction mode and remote sensing detection. The on-site detection comprises a shore-based observation station, a ship and a buoy, the buoy is a recognized high-precision ocean parameter detection method, but the detection range of the buoy is limited and is distributed in a single point mode, and large-range observation cannot be achieved due to high maintenance cost. The numerical prediction mode can obtain parameters such as wind field, sea wave and the like of a target sea area through numerical calculation, the precision of the result is influenced by initial input conditions, and the spatial resolution is not high. With the development of satellite remote sensing technology and synthetic aperture radar technology, the satellite-borne synthetic aperture radar can detect ground objects all day long and all the time, has high spatial resolution (up to several meters to tens of meters), and is a main means for realizing large-range detection of sea surface wind fields and sea waves at present and even in the future. The original information carried in the SAR data is related to the parameters of the wind field and the sea wave, but the parameters of the wind field and the sea wave cannot be directly obtained from the original information, and the inversion means of the wind field and the sea wave is needed, so that the establishment of the SAR wind field and sea wave joint inversion method with high timeliness has important significance.
At present, researchers in various countries develop a great deal of research work in the field of SAR wind field and wave inversion, and establish various methods for extracting wind field and wave parameters from SAR data. The inversion method of the wind field parameters mainly comprises a Geophysical Model Function (GMF), and the inversion method of the sea wave parameters mainly comprises a sea wave spectrum inversion method and a sea wave parameter inversion empirical method.
The geophysical model function is a common wind speed inversion method of the same-polarization SAR, an empirical model is established by counting radar backscattering coefficients, buoy and mode wind field data, and the relation between the radar incidence angle, sea surface wind speed, relative wind direction and backscattering coefficients is described. The geophysical model function is firstly applied to the scatterometer, the scatterometer can observe the same sea surface unit from different azimuth angles, and wind speed and wind direction results can be obtained by establishing a cost function related to the geophysical model function. The azimuth direction of the synthetic aperture radar is unchanged, and the wind direction is required to be used as initialization information, so that the incident angle and the relative wind direction in the geophysical model function can be determined, and the wind speed information of the sea surface can be acquired.
The method for performing sea wave parameter inversion based on the sea wave spectrum is that firstly, the SAR data is used for inverting the sea wave spectrum, and then the sea wave parameters are calculated by utilizing the sea wave spectrum. The empirical method for sea wave parameter inversion is to establish the relationship between SAR data and sea wave parameters through regression prediction to obtain an empirical model, and after the SAR data is input into the empirical model, the empirical model can give accurate sea wave parameters. The empirical method does not need to carry out complex wave spectrum inversion and does not need to rely on an external data source.
When extracting wind speed in the co-polarized SAR data using the geophysical model function, an initial wind direction needs to be input, which typically needs to be obtained from an external data source. This results in the accuracy of SAR wind speed inversion being affected by more factors and makes SAR unable to independently achieve detection of sea surface wind fields. Most SAR ocean wave spectrum inversion methods rely on initial guess spectra, and the initial guess spectra need to be provided by other data sources besides SAR data. This situation leads to the following problems: firstly, acquiring a primary guess spectrum from an external data source introduces more uncertain factors influencing an inversion result; secondly, extra time is needed in the process of obtaining the initial guess spectrum, so that the timeliness of inversion is reduced; finally, SAR does not allow for the detection of ocean waves as an independent source of observation.
In summary, most conventional methods for inverting the wind field and the sea wave of the SAR at present rely on wind direction and initial guess spectrum information (the wind direction is initial information required to be input when inverting the wind speed, and the initial guess spectrum is initial spectrum information required to be input when inverting the sea wave spectrum), and most of the information of the wind direction and the initial guess spectrum needs to be obtained from an external data source, which results in that the SAR cannot independently realize the detection of the wind field and the sea wave. Extra time is needed for acquiring data from an external data source, and the accuracy of the external data is influenced by more factors, so that the timeliness of SAR wind field and sea wave inversion is reduced, and the inversion accuracy of the wind field and sea wave parameters is easily influenced by more factors.
Disclosure of Invention
In order to solve the technical problems in the background art, the invention provides a data-driven SAR WIND field and sea wave joint inversion method and a data-driven SAR WIND field and sea wave joint inversion system, wherein 23 characteristic quantities related to WIND field and sea wave parameters are extracted from SAR data, and an empirical method CWAVE _ WIND of WIND field and sea wave joint inversion is established by using a convolutional neural network method. The method is irrelevant to wind direction and initial guess spectrum, so that SAR wind field and sea wave inversion do not depend on an external data source, and information such as wind speed, effective wave height, average wave period and the like can be synchronously output.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a data-driven SAR wind field and sea wave joint inversion method.
A data-driven SAR wind field and sea wave joint inversion method comprises the following steps:
acquiring SAR data, and acquiring a plurality of small images in the SAR data;
extracting a backscattering coefficient, a normalized variance, an incidence angle and a frequency domain characteristic quantity of the small image, and constructing an input data set;
acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the small image in a space-time mode, and constructing an output data set based on the matched wind speed, the effective wave height and the average wave period;
inputting the input data set and the output data set into a convolutional neural network model for training to obtain a trained convolutional neural network model;
and inputting the SAR data of the sea area to be tested into the trained convolutional neural network model to obtain the wind speed, the effective wave height and the average wave period of the sea area to be tested.
Further, the preprocessing process comprises: controlling the quality of the SAR data; for the WV1 data, images with normalized variances between 1-2 were considered high quality images; for the WV2 data, images with normalized variances between 1-1.4 were considered high quality images.
Further, the constructing of the input data set comprises: and obtaining a high-quality small image from the plurality of small images, calculating the average value of the backscattering coefficient, the average value of the normalized variance, the central incident angle and the average value of the frequency domain characteristic quantity of all pixels in the high-quality small image, and taking the average value of the backscattering coefficient, the average value of the normalized variance, the central incident angle and the average value of the frequency domain characteristic quantity as the backscattering coefficient, the normalized variance, the incident angle and the frequency domain characteristic quantity of the small image to construct an input data set.
Further, the constructing of the output data set comprises: acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the high-quality small image in a space-time mode, calculating the average value of the wind speed, the average value of the effective wave height and the average value of the average wave period of all pixels in the small image, taking the wind average value of the wind speed, the average value of the effective wave height and the average value of the average wave period as the wind speed, the effective wave height and the average wave period of the small image, and constructing an output data set.
Further, before training the model, the method further comprises: the input data set and the output data set are processed, and the processing comprises dividing the data set and normalizing the data.
The invention provides a data-driven SAR wind field and sea wave joint inversion system.
A SAR wind field and sea wave joint inversion system based on data driving comprises:
an acquisition module configured to: acquiring SAR data, and acquiring a plurality of small images in the SAR data;
an input data set construction module configured to: extracting a backscattering coefficient, a normalized variance, an incidence angle and a frequency domain characteristic quantity of the small image, and constructing an input data set;
an output data set construction module configured to: acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the small image in a space-time mode, and constructing an output data set based on the matched wind speed, the effective wave height and the average wave period;
a model training module configured to: inputting the input data set and the output data set into a convolutional neural network model for training to obtain a trained convolutional neural network model;
an inversion module configured to: and inputting the SAR data of the sea area to be tested into the trained convolutional neural network model to obtain the wind speed, the effective wave height and the average wave period of the sea area to be tested.
A third aspect of the invention provides a computer-readable storage medium.
A computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps in the data-driven SAR wind field and wave joint inversion based method according to the first aspect.
A fourth aspect of the invention provides a computer apparatus.
A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the data-driven SAR wind field and wave joint inversion method according to the first aspect.
Compared with the prior art, the invention has the beneficial effects that:
1. the SAR wind field and sea wave joint inversion method established by the invention is suitable for Sentinel-1 satellite SAR wave mode data, and has nothing to do with wind direction and initial guess spectrum, so that the wind field and sea wave inversion of the SAR does not depend on an external data source, the information of the wind field and sea wave can be independently detected, the parameters of the wind field and the sea wave can be synchronously output, and the timeliness of the SAR wind field and sea wave parameter inversion is improved.
2. The method can simultaneously calculate the wind speed, the effective wave height and the average wave period of the sea surface, and improves the timeliness of the SAR wind field wave inversion.
3. The method has high calculation efficiency and low cost, is suitable for various sea conditions, and has good universality.
Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
FIG. 1 is a flow chart of a data-driven SAR wind field and sea wave joint inversion method;
FIG. 2 is a graph comparing inverted wind speed to NDBC wind speed in an embodiment of the present invention;
FIG. 3 is a graph of the comparison of the inversion and NDBC effective wave heights in an embodiment of the present invention;
FIG. 4 is a graph comparing the inversion average wave period and the NDBC average wave period in an embodiment of the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
According to the method, 23 characteristic quantities related to a WIND field and sea waves are extracted from Sentinil-1 satellite SAR wave mode data, a relation between the characteristic quantities and the WIND field and sea wave parameters is established by using a convolutional neural network, and a SAR WIND field and sea wave joint inversion method CWAVE _ WIND is obtained, so that SAR WIND field and sea wave inversion does not depend on an external data source, and the WIND field and sea wave parameters can be synchronously output.
Example one
As shown in fig. 1, the present embodiment provides a data-driven SAR wind field and wave joint inversion method, which is applied to a server for illustration, and it is understood that the method may also be applied to a terminal, and may also be applied to a system including a terminal and a server, and is implemented by interaction between the terminal and the server. The server may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a network server, cloud communication, middleware service, a domain name service, a security service CDN, a big data and artificial intelligence platform, and the like. The terminal may be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and the like. The terminal and the server may be directly or indirectly connected through wired or wireless communication, and the application is not limited herein. In this embodiment, the method includes the steps of:
s101: acquiring SAR data, and acquiring a plurality of small images in the SAR data;
firstly, parameters related to wind speed, effective wave height and average wave period in SAR data are selected and used as input characteristic quantities of the model. The geophysical model function is a classical homopolarization SAR wind speed inversion empirical method, and the method has a specific function form and establishes a relation between the sea surface wind speed and SAR backscattering coefficients, incidence angles and relative wind directions. In the actual wind speed inversion process, a data source other than the SAR data is often used to provide wind direction information, and the wind speed inversion cannot be completed by using only the SAR data. The research of the invention aims to extract sea surface wind speed by taking SAR data as a unique data source, so that SAR backscattering coefficients and incidence angles are selected as input characteristic quantities related to the wind speed. The CWAVE method is a classic SAR ocean wave inversion empirical method, and uses some SAR information related to ocean wave parameters. The backscattering coefficient and the normalized variance (CVAR) are 2 SAR spatial domain characteristic parameters related to wave parameters and can reflect the energy of the SAR image. The normalized variance is calculated as follows:
wherein CVAR is a normalized variance; a is the intensity of the SAR image; std is the standard deviation.
The backscattering coefficient and the normalized variance of the spatial domain cannot fully contain the information of the sea waves, and the sea wave related information can be further supplemented by searching the characteristic quantity of the SAR image from the frequency domain. The characteristic parameters of the frequency domain are decomposed and extracted by a group of orthogonal functions, and 20 parameters extracted by orthogonal decomposition are closely related to long waves and short waves of the SAR wave mode image. The calculation formula of the 20 characteristic parameters is as follows:
wherein,
is a normalized SAR image spectrum and is,
is an orthogonal function, k x And k y Is the azimuth and range wavenumbers in the SAR image spectrum.
After determining the input and output features of the empirical method, preparing a data set for training and testing the empirical method, comprising: input data set, output data set. The first step in preparing the input data set is to download the SAR data, the coverage area of the SAR data is the indian ocean sea area, and the acquisition time of the data is from 1 month 2020 to 10 months 2020. When a download area is selected, the latitude of the SAR data is limited within 60 degrees, and the influence of sea ice on the detection quality of the SAR data is avoided.
S102: and extracting the backscattering coefficient, the normalized variance, the incident angle and the frequency domain characteristic quantity of the small image to construct an input data set.
After obtaining the Sentinel-1A satellite SAR wave mode data, preprocessing the SAR data, and extracting a backscattering coefficient, a normalized variance, an incidence angle and 20 frequency domain characteristic quantities from each small image. In the preprocessing process, the quality of the SAR image is controlled, and for WV1 data, the image with the normalized variance between 1 and 2 is regarded as a high-quality image; for the WV2 data, images with normalized variances between 1-1.4 were considered high quality images. And taking the backscattering coefficient average value, the normalized variance average value, the central incidence angle and the frequency domain characteristic quantity average value of all pixels in the high-quality small image as the backscattering coefficient, the normalized variance, the incidence angle and the frequency domain characteristic quantity of the small image, wherein each small image provides 1 group of input data, and the characteristic quantities extracted from all high-quality SAR wave mode small images form an input data set.
S103: acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the small image in a space-time mode, and constructing an output data set based on the matched wind speed, the effective wave height and the average wave period.
After the input data set is prepared, the output data set is prepared. The output variables are wind speed, effective wave height and average wave period, and the corresponding data are the effective wave height and average wave period data which are provided by the ECMWF and are combined by wind speed of 10 meters, wind waves and surge waves. The ECMWF data had a temporal resolution of 1 hour, a spatial resolution of 0.125 ° × 0.125 °, a temporal range of 2020.01.01-2020.10.31, a spatial range of 70 ° S-30 ° N, 30 ° E-120 ° E. After the ECMWF data are downloaded, the wind speed, the effective wave height and the average wave period of the ECMWF are matched with each pixel of the high-quality small image in a space-time mode, and then the average value of the wind speed, the effective wave height and the average wave period of all pixels in the small image is calculated and used as the wind speed, the effective wave height and the average wave period of the small image. The space-time matching result of each small image provides 1 group of output data, and an output data set is formed by the wind speed, the effective wave height and the average wave period of the space-time matching of all high-quality SAR wave mode small images.
S104: inputting the input data set and the output data set into a convolutional neural network model for training to obtain a trained convolutional neural network model;
as one or more embodiments, after preparing the input data set and the output data set for the training method, the data set needs to be preprocessed, and the preprocessing of the data set performed by the invention mainly comprises the following steps: and dividing and normalizing the data set. The data set division refers to dividing a prepared data set into a training data set and a test data set, wherein the common division ratio is 7: 3. data normalization is the conversion of input and output data sets of different value ranges into the same value range. Since the coverage of data values of different samples is different, directly taking them as input will affect the convergence and speed of the training process.
After data preprocessing is completed, an empirical method is established, and the CWAVE _ WIND method is established by using a convolutional neural network. The convolutional neural network algorithm was implemented using a neural network toolbox in MATLAB. When the tool box is used for realizing the convolutional neural network algorithm, a network layer of the convolutional neural network is constructed. The convolutional neural network constructed by the invention is a 7-layer regression network, and comprises the following steps: input layer, convolution layer, activation function layer, full connection layer 1, full connection layer 2, output layer, regression layer. Wherein the input layer is used for receiving input data from the outside, and is set to be [ 2311 ]; the convolution layer has the functions of local perception and weight sharing, and is composed of 16 filters with the size of 3 multiplied by 3, and the edge filling mode in the convolution process is the same filling; the activation function layer is used for introducing nonlinearity into a neural network, and The activation function selects a modified Linear Unit (ReLU); the fully connected layer is used for integrating the features and preparing for regression, and the number of neurons in the layer is set to be 400; the output layer is used for outputting results to the outside, the number of the neurons of the layer is set to be 3, and the number is the number of output variables; the role of the regression layer is to calculate the loss values. In training the network, the learning rate of the neural network is set to 0.001.
After the CWAVE _ WIND method is established, the method is used for respectively carrying out WIND field and sea wave parameter joint inversion on SAR data in a training data set, a testing data set and a buoy data set, comparing the inverted WIND speed, the effective wave height and the average wave period with the reference WIND speed, the effective wave height and the average wave period, and analyzing the availability of the CWAVE _ WIND method. The comparison of the parameters of the buoy data set is shown in fig. 2-4.
As can be seen from the results of comparing the wind field sea wave parameters as shown in FIGS. 2-4, the wind speed range of the buoy data set is 3-15m/s, and the root mean square error of the inversion wind speed and the NDBC wind speed is 1.38 m/s; the range of the effective wave height is 0-5m, and the root mean square error of the inversion effective wave height and the NDBC effective wave height is 0.47 m; the average wave period ranges from 4 s to 9s, and the root mean square error of the inversion average wave period and the NDBC average wave period is 0.48 s. The correlation coefficient of the inversion result of the wind field sea wave parameters compared with the NDBC result is not less than 0.8, which shows that the correlation coefficient is strong. The above results illustrate that: the empirical method trained through the convolutional neural network algorithm is suitable for the Sentinel-1 satellite SAR wave mode data.
S105: and inputting the SAR data of the sea area to be tested into the trained convolutional neural network model to obtain the wind speed, the effective wave height and the average wave period of the sea area to be tested.
Example two
The embodiment provides a data-driven SAR wind field and sea wave joint inversion system.
A SAR wind field and sea wave joint inversion system based on data driving comprises:
an acquisition module configured to: acquiring SAR data, and acquiring a plurality of small images in the SAR data;
an input data set construction module configured to: extracting a backscattering coefficient, a normalized variance, an incidence angle and a frequency domain characteristic quantity of the small image, and constructing an input data set;
an output data set construction module configured to: acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the small image in a time-space mode, and constructing an output data set based on the matched wind speed, the effective wave height and the average wave period;
a model training module configured to: inputting the input data set and the output data set into a convolutional neural network model for training to obtain a trained convolutional neural network model;
an inversion module configured to: and inputting the SAR data of the sea area to be tested into the trained convolutional neural network model to obtain the wind speed, the effective wave height and the average wave period of the sea area to be tested.
It should be noted here that the acquisition module, the input data set construction module, the output data set construction module, the model training module and the inversion module correspond to steps S101 to S105 in the first embodiment, and the modules are the same as the corresponding steps in the implementation example and application scenarios, but are not limited to the disclosure in the first embodiment. It should be noted that the modules described above as part of a system may be implemented in a computer system such as a set of computer-executable instructions.
EXAMPLE III
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps in the data-driven SAR wind field and wave joint inversion-based method according to the first embodiment.
Example four
The embodiment provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the program, the processor implements the steps in the data-driven SAR wind field and wave joint inversion method according to the first embodiment.
As will be appreciated by one skilled in the art, embodiments of the present invention may provide a method, system, or computer program product. Accordingly, the present invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (9)
1. A data-driven SAR wind field and sea wave joint inversion method is characterized by comprising the following steps:
acquiring SAR data, and acquiring a plurality of small images in the SAR data;
extracting a backscattering coefficient, a normalized variance, an incidence angle and a frequency domain characteristic quantity of the small image, and constructing an input data set;
acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the small image in a space-time mode, and constructing an output data set based on the matched wind speed, the effective wave height and the average wave period;
inputting the input data set and the output data set into a convolutional neural network model for training to obtain a trained convolutional neural network model;
inputting SAR data of a sea area to be detected into the trained convolutional neural network model to obtain the wind speed, the effective wave height and the average wave period of the sea area to be detected;
and the acquisition of the frequency domain characteristic quantity comprises the step of decomposing and extracting the SAR image spectrum through a group of orthogonal functions to obtain 20 frequency domain characteristic quantities related to the long wave and the short wave of the SAR wave mode image.
2. The data-driven SAR wind field and wave joint inversion method based on the claim 1 is characterized in that SAR data are preprocessed, and the preprocessing process comprises the following steps: controlling the quality of the SAR data; for the WV1 data, images with normalized variances between 1-2 were considered high quality images; for the WV2 data, images with normalized variances between 1-1.4 were considered high quality images.
3. The data-driven SAR wind field and wave joint inversion method based on claim 2, characterized in that the construction of the input data set comprises: and obtaining a high-quality small image from the plurality of small images, calculating the average value of the backscattering coefficient, the average value of the normalized variance, the central incident angle and the average value of the frequency domain characteristic quantity of all pixels in the high-quality small image, and taking the average value of the backscattering coefficient, the average value of the normalized variance, the central incident angle and the average value of the frequency domain characteristic quantity as the backscattering coefficient, the normalized variance, the incident angle and the frequency domain characteristic quantity of the small image to construct an input data set.
4. The data-driven SAR wind field and wave joint inversion method based on claim 3, characterized in that the construction of the output data set comprises: acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the high-quality small image in a time-space mode, calculating the wind mean value, the average value of the effective wave height and the average value of the average wave period of the wind speed of all pixels in the small image, taking the average value of the wind speed, the average value of the effective wave height and the average value of the average wave period as the wind speed, the effective wave height and the average wave period of the small image, and constructing an output data set.
5. The data-driven SAR wind field and wave joint inversion method based on claim 1, characterized in that before training the model, the method further comprises: the input data set and the output data set are processed, and the processing comprises dividing the data set and normalizing the data.
6. The data-driven SAR wind field and wave joint inversion based on claim 5, characterized in that the data set partitioning comprises partitioning the prepared data set into a training data set and a testing data set.
7. A SAR wind field and sea wave joint inversion system based on data driving is characterized by comprising:
an acquisition module configured to: acquiring SAR data, and acquiring a plurality of small images in the SAR data;
an input data set construction module configured to: extracting a backscattering coefficient, a normalized variance, an incidence angle and a frequency domain characteristic quantity of the small image, and constructing an input data set;
an output data set construction module configured to: acquiring ECMWF data, matching the wind speed, the effective wave height and the average wave period for each pixel of the small image in a space-time mode, and constructing an output data set based on the matched wind speed, the effective wave height and the average wave period;
a model training module configured to: inputting the input data set and the output data set into a convolutional neural network model for training to obtain a trained convolutional neural network model;
an inversion module configured to: inputting SAR data of the sea area to be tested into the trained convolutional neural network model to obtain the wind speed, the effective wave height and the average wave period of the sea area to be tested;
and the acquisition of the frequency domain characteristic quantity comprises the step of decomposing and extracting the SAR image spectrum through a group of orthogonal functions to obtain 20 frequency domain characteristic quantities related to the long wave and the short wave of the SAR wave mode image.
8. A computer readable storage medium having stored thereon a computer program, wherein the program when executed by a processor implements the steps in the data-driven SAR wind field and wave joint inversion method according to any of claims 1-6.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps in the data-driven SAR wind field and wave joint inversion based method according to any of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110500084.5A CN113156439B (en) | 2021-05-08 | 2021-05-08 | SAR wind field and sea wave joint inversion method and system based on data driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110500084.5A CN113156439B (en) | 2021-05-08 | 2021-05-08 | SAR wind field and sea wave joint inversion method and system based on data driving |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113156439A CN113156439A (en) | 2021-07-23 |
| CN113156439B true CN113156439B (en) | 2022-08-30 |
Family
ID=76873815
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110500084.5A Active CN113156439B (en) | 2021-05-08 | 2021-05-08 | SAR wind field and sea wave joint inversion method and system based on data driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113156439B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114265087A (en) * | 2021-12-24 | 2022-04-01 | 青岛镭测创芯科技有限公司 | Sea wave parameter inversion method and related assembly |
| CN114580509B (en) * | 2022-02-21 | 2022-10-28 | 国家海洋环境预报中心 | Sea wave macroscopic characteristic quantity prediction system based on convolution length memory network |
| CN114895309B (en) * | 2022-05-12 | 2023-08-15 | 中科卫星应用德清研究院 | Full-polarization SAR sea surface wind field inversion method and system considering SAR amplitude-phase imbalance |
| CN114814779B (en) * | 2022-06-29 | 2022-11-11 | 国家海洋技术中心 | Buoy surge wave height observation data error evaluation method, system, equipment and medium |
| CN115983141B (en) * | 2023-03-21 | 2023-07-14 | 国家海洋局北海预报中心((国家海洋局青岛海洋预报台)(国家海洋局青岛海洋环境监测中心站)) | Deep learning-based method, medium and system for inverting wave height of ocean waves |
| CN117687044B (en) * | 2023-11-02 | 2024-08-30 | 昆明理工大学 | Deep learning method for satellite-borne GNSS-R global sea surface effective wave height estimation |
| CN117969881B (en) * | 2024-03-29 | 2024-06-25 | 中国石油大学(华东) | Method, device and medium for inverting typhoon sea state and sea surface wind speed based on remote sensing data |
| CN118033590B (en) * | 2024-04-12 | 2024-06-28 | 南京信息工程大学 | Short-term precipitation prediction method based on improved VIT neural network |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012170091A1 (en) * | 2011-06-08 | 2012-12-13 | Chevron U.S.A. Inc. | System and method for seismic data inversion |
| CN104331588A (en) * | 2014-03-06 | 2015-02-04 | 浙江海洋学院 | A joint inversion device of sea surface wind field and sea waves |
| CN109886217A (en) * | 2019-02-26 | 2019-06-14 | 上海海洋大学 | A method of it is high that wave being detected from Nearshore Wave video based on convolutional neural networks |
| CN110991087A (en) * | 2019-12-24 | 2020-04-10 | 中国石油大学(华东) | Wind field inversion method and system based on multi-incidence-angle networking SAR satellite data |
| WO2020164270A1 (en) * | 2019-02-15 | 2020-08-20 | 平安科技（深圳）有限公司 | Deep-learning-based pedestrian detection method, system and apparatus, and storage medium |
| CN111583214A (en) * | 2020-04-30 | 2020-08-25 | 江苏科技大学 | Method for inverting sea surface wind speed by using marine radar image based on RBF neural network |
| CN112014842A (en) * | 2020-09-16 | 2020-12-01 | 国家海洋技术中心 | Effective wave height estimation method and system for high-resolution three-signal full-polarization SAR data |
| CN112180369A (en) * | 2020-09-23 | 2021-01-05 | 中国人民解放军国防科技大学 | Depth learning-based sea surface wind speed inversion method for one-dimensional synthetic aperture radiometer |
| EP3805794A1 (en) * | 2019-10-10 | 2021-04-14 | Orbital Insight, Inc. | Analysis of sar data using neural networks |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9423307B2 (en) * | 2013-02-20 | 2016-08-23 | Mesa Photonics, LLC | Method and apparatus for determining wave characteristics using interaction with a known wave |
| US10705208B2 (en) * | 2018-06-11 | 2020-07-07 | Augmented Radar Imaging, Inc. | Vehicle location determination using synthetic aperture radar |
| CN110543815B (en) * | 2019-07-22 | 2024-03-08 | 平安科技（深圳）有限公司 | Training method of face recognition model, face recognition method, device, equipment and storage medium |
-
2021
- 2021-05-08 CN CN202110500084.5A patent/CN113156439B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012170091A1 (en) * | 2011-06-08 | 2012-12-13 | Chevron U.S.A. Inc. | System and method for seismic data inversion |
| CN104331588A (en) * | 2014-03-06 | 2015-02-04 | 浙江海洋学院 | A joint inversion device of sea surface wind field and sea waves |
| WO2020164270A1 (en) * | 2019-02-15 | 2020-08-20 | 平安科技（深圳）有限公司 | Deep-learning-based pedestrian detection method, system and apparatus, and storage medium |
| CN109886217A (en) * | 2019-02-26 | 2019-06-14 | 上海海洋大学 | A method of it is high that wave being detected from Nearshore Wave video based on convolutional neural networks |
| EP3805794A1 (en) * | 2019-10-10 | 2021-04-14 | Orbital Insight, Inc. | Analysis of sar data using neural networks |
| CN110991087A (en) * | 2019-12-24 | 2020-04-10 | 中国石油大学(华东) | Wind field inversion method and system based on multi-incidence-angle networking SAR satellite data |
| CN111583214A (en) * | 2020-04-30 | 2020-08-25 | 江苏科技大学 | Method for inverting sea surface wind speed by using marine radar image based on RBF neural network |
| CN112014842A (en) * | 2020-09-16 | 2020-12-01 | 国家海洋技术中心 | Effective wave height estimation method and system for high-resolution three-signal full-polarization SAR data |
| CN112180369A (en) * | 2020-09-23 | 2021-01-05 | 中国人民解放军国防科技大学 | Depth learning-based sea surface wind speed inversion method for one-dimensional synthetic aperture radiometer |
Non-Patent Citations (11)
| Title |
|---|
| Assessing the Bayesian inversion technique of C-band synthetic aperture radar data for the retrieval of wind fields in marine coastal areas;M Adamo 等;《Journal of Applied Remote Sensing》;20141106;第1-16页 * |
| Classification of the global Sentinel-1 SAR vignettes for ocean surface process studies;Chen Wang 等;《Remote Sensing of Environment》;20191231;第1-11页 * |
| Detection of Ocean Wave Parameters Using Synthetic Aperture Radar (SAR) Data;Sudhir Kumar Chaturvedi 等;《The Journal of Navigation》;20121217;第283-293页 * |
| Research on Wind Speed Inversion Method for X-Band Networked SAR Satellite;Yong Wan 等;《Journal of Marine Science and Engineering》;20200819;第1-12页 * |
| X-波段雷达近海海浪频谱反演的神经网络模型;王静 等;《海洋学报(中文版)》;20130331;第35卷(第2期);第43-51页 * |
| 基于BP神经网络的SAR海表风场反演——以长江入海口及临近海域为例;胡敏 等;《海洋湖沼通报》;20200430;第35-42页 * |
| 基于RADARSAT-2 SAR卫星数据的海浪谱反演方法及其海试实验验证;万勇 等;《实验室研究与探索》;20200731;第39卷(第7期);第16-19、27页 * |
| 基于神经网络模型的GNSS-R海面风速反演算法研究;王斯嘉;《中国优秀博硕士学位论文全文数据库(硕士)基础科学辑》;20210215;第1-77页 * |
| 多传感器卫星海面风场遥感研究;冯倩;《中国优秀博硕士学位论文全文数据库(博士)基础科学辑》;20050315;第1-104页 * |
| 星载SAR海洋场景仿真与反演方法综述;史伟哲;《航天器工程》;20110131;第20卷(第1期);第50-56页 * |
| 机载SAR探测海面风场、海浪和海洋内波方法的分析与研究;刘丽娜 等;《气象水文海洋仪器》;20120630;第6-10页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113156439A (en) | 2021-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113156439B (en) | SAR wind field and sea wave joint inversion method and system based on data driving | |
| CN104613944B (en) | Distributed water depth prediction method based on GWR (geographically weighted regression) and BP (back propagation) neural network | |
| CN111738347B (en) | Sea wave direction spectrum correction method and device, storage medium and electronic equipment | |
| CN111950438B (en) | Depth learning-based effective wave height inversion method for Tiangong No. two imaging altimeter | |
| CN115077656B (en) | Reservoir water reserve retrieval method and device | |
| CN109033543A (en) | A kind of heterogeneous area's vegetation fraction estimation method, device and equipment of earth's surface | |
| Beauchamp et al. | 4DVarNet-SSH: End-to-end learning of variational interpolation schemes for nadir and wide-swath satellite altimetry | |
| CN116343046A (en) | Multi-source remote sensing urban river water quality monitoring method based on self-optimizing machine learning | |
| CN116482317B (en) | Lake water nutrition state real-time monitoring method, system, equipment and medium | |
| CN116719062A (en) | Satellite signal quality evaluation method and device, equipment and storage medium | |
| CN113984198B (en) | Shortwave radiation prediction method and system based on convolutional neural network | |
| Mooneyham et al. | SWRL Net: A spectral, residual deep learning model for improving short-term wave forecasts | |
| CN115097451A (en) | Sea wave multi-parameter inversion method and system based on SAR satellite data | |
| CN117148360A (en) | Lightning approach prediction method and device, electronic equipment and computer storage medium | |
| CN110991087A (en) | Wind field inversion method and system based on multi-incidence-angle networking SAR satellite data | |
| CN117274831B (en) | Offshore turbid water body depth inversion method based on machine learning and hyperspectral satellite remote sensing image | |
| CN114037905A (en) | Estuary area water meter salinity remote sensing inversion method and system based on satellite-ground cooperation | |
| CN117664874A (en) | Estuary area salt tide tracing monitoring method and system based on space-earth combination | |
| CN116299247B (en) | InSAR atmospheric correction method based on sparse convolutional neural network | |
| CN117371316A (en) | Deep learning-based stationary satellite solar short wave radiation inversion method and readable storage medium | |
| CN109063760B (en) | Polarization SAR classification method based on random forest multi-scale convolution model | |
| CN113822895B (en) | ScanSAR image scallop effect inhibition method based on self-attention mechanism and CycleGAN | |
| CN112861355B (en) | Sea surface net radiation estimation method and system | |
| Etala et al. | An investigation of ensemble-based assimilation of satellite altimetry and tide gauge data in storm surge prediction | |
| Li et al. | WSPTGAN for Global Ocean Surface Wind Speed Generation with High Temporal Resolution and Spatial Coverage |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |