Radio frequency interference positioning method and device combining brightness temperature image and subspace

Abstract

The invention discloses a radio frequency interference positioning method and device combining a brightness temperature image and a subspace, and the method comprises the steps: obtaining an original scene brightness temperature image for observing an earth scene, constructing a simulation background brightness temperature image of the earth scene, and obtaining a brightness temperature image after the first SNR is improved; carrying out image processing on the brightness temperature image of the original scene, generating a radio frequency interference (RFI) mask corresponding to the brightness temperature image of the original scene, and obtaining a brightness temperature image after the SNR is improved for the second time; and constructing a covariance matrix of the brightness temperature image after the second SNR improvement, carrying out subspace decomposition on the covariance matrix, and carrying out geolocation on an RFI source observed in an observation image view field. According to the method provided by the invention, the positioning precision of the RFI source under the condition of low signal-to-noise ratio is improved.

Description

Technical Field

[0001] The present invention relates to the field of satellite microwave ocean remote sensing technology, and in particular to a method and device for locating radio frequency interference combining brightness temperature images and subspaces. Background Art

[0002] Due to the existence of illegal radio frequency interference (RFI) sources around the world, SAIR's on-orbit observation data has been seriously affected. Shutting down illegal RFI sources or suppressing RFI sources mostly requires the precise location of the RFI to be known.

[0003] In the prior art, the positioning accuracy and spatial resolution of the RFI source positioning method based on brightness temperature are limited by the angular resolution of the instrument, and there are problems when processing multiple very close sources or sources in non-uniform backgrounds. The subspace-based MUSIC (Multiple Signal Classification) algorithm proposed by H.park et al. applies DOA (Direction of arrival, spatial signal direction of arrival) estimation to SAIR, which can achieve super-resolution positioning and distinguish very close RFI sources. However, the covariance matrix used in the subspace decomposition of the MUSIC algorithm is affected by background noise and channel noise. In the case of low SNR (Signal-to-Noise Ratio) (background is non-uniform, RFI intensity is weak), this effect cannot be ignored, which seriously affects the positioning performance of the algorithm. Therefore, it is necessary to propose a new RFI source geolocation method to achieve higher accuracy in low signal-to-noise ratio conditions. Summary of the invention

[0004] In view of this, the present invention provides a method and device for locating radio frequency interference by combining brightness temperature images and subspaces, which can solve the technical problems in the prior art that affect the positioning performance.

[0005] In order to solve the above technical problems, the present invention is implemented as follows.

[0006] A method for locating radio frequency interference combining brightness temperature image and subspace, the method comprising the following steps:

[0007] Step S1: performing original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene, using a forward model to determine the ocean constant brightness temperature and the land constant brightness temperature in the earth scene, constructing a simulated background brightness temperature image of the earth scene, and performing a difference between the original scene brightness temperature image and the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR improvement;

[0008] Step S2: performing image processing on the original scene brightness temperature image to generate a radio frequency interference (RFI) mask corresponding to the original scene brightness temperature image, performing a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR improvement to obtain the brightness temperature image after the second SNR improvement;

[0009] Step S3: Perform Fourier transform on the brightness temperature image after the second SNR improvement to obtain visibility samples; construct a covariance matrix of the brightness temperature image after the second SNR improvement, perform subspace decomposition on the covariance matrix, and geolocate the RFI source observed in the field of view of the brightness temperature image after the second SNR improvement.

[0010] Preferably, the step S1 comprises:

[0011] Step S11: performing original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene; obtaining the longitude and latitude coordinates corresponding to each pixel point in the earth's field of view of the original scene brightness temperature image, and obtaining the sea and land mask corresponding to the pixel point;

[0012] Step S12: Based on the sea and land masks corresponding to each pixel point, the pixel points in the earth's field of view are classified into two categories: pixel points located at the ocean position and pixel points located at the land position;

[0013] Step S13: determining the ocean constant brightness temperature and the land constant brightness temperature of the observed earth scene based on the forward model;

[0014] Step S14: setting the brightness temperature of pixels located at the ocean position in the original scene brightness temperature image to the ocean constant brightness temperature, and setting the brightness temperature of pixels located at the land position in the original scene brightness temperature image to the land constant brightness temperature, to obtain a simulated background brightness temperature image;

[0015] Step S15: subtract the original scene brightness temperature image from the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR enhancement.

[0016] Preferably, the ocean constant brightness temperature and the land constant brightness temperature are calculated as follows:

[0017]

[0018]

[0019]

[0020] in, is the constant brightness temperature over land, is the antenna temperature; Calculate the visibility of 1K brightness temperature for the Earth's surface, visibility of cold sky, ocean, sun and / or moon, solar reflection, and antenna back lobe; are the ocean constant brightness temperatures for vertical and horizontal polarizations, is the constant brightness temperature of the polarized ocean, is the constant brightness temperature of the H-polarized ocean, Γ v (SST, SSS, θ) is the V polarization Fresnel reflection coefficient, Γ h (SST, SSS, θ) is the H-polarization Fresnel reflection coefficient, SST is the sea surface temperature, SSS is the sea surface salinity, WS is the wind speed, the default values ​​of SST, SSS, and WS are set to the average values ​​of 15°C, 35psu, and 4m / s, respectively, θ is the zenith angle of the target in the sensor direction, and r(·) is the Fresnel reflection coefficient model, which is calculated by:

[0021]

[0022]

[0023] where ε s is the complex dielectric constant of the ocean.

[0024] Preferably, the step S2 comprises:

[0025] Step S21: binarizing the original scene brightness temperature image according to the brightness temperature threshold;

[0026] Step S22: dilating the binarized brightness temperature image;

[0027] Step S23: The expanded binary brightness temperature image is inverted by HFFT to obtain a brightness temperature image of the supplementary harmonic component, where:

[0028] V HCS =F(T dilated ) Formula 12

[0029] T HCS =F -1 (V HCS ) Formula 13

[0030] V HCS represents the brightness temperature image of the complementary harmonic component, F(T dilated ) is the response function, T dilated is the brightness temperature image after expansion, T HCS To supplement the brightness temperature image of the harmonic component, F -1 (V HCS ) is HFFT transformation, V HCS To supplement the visibility of harmonic components;

[0031] Step S24: transform the threshold range of the brightness temperature image of the supplementary harmonic component to [0.3, 3] to generate an RFI mask

[0032]

[0033] Where T mask is the RFI mask, T HCS To supplement the brightness temperature image of the harmonic component;

[0034] Step S25: performing a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR enhancement to obtain a brightness temperature image after the second SNR enhancement.

[0035] Preferably, the step S3 comprises:

[0036] Step S31: Perform Fourier transform on the brightness temperature image after the second SNR enhancement to obtain visibility samples.

[0037]

[0038] in, represents visibility samples, F(·) represents Fourier transform, This is the brightness temperature image after the second SNR improvement;

[0039] Step S32: construct the covariance matrix of the brightness temperature image after the second SNR enhancement.

[0040]

[0041] in, represents the SNR covariance matrix, is the element of the covariance matrix, (u ij , v ij ) is the baseline formed by antenna i and antenna j, which is equal to the wavelength-normalized spacing between the two antennas i and j on the xy plane. is the visibility of the corresponding antenna i and antenna j after SNR improvement, H is the Hermilte transpose, i, j are the antenna numbers, and N is the number of antennas; (u ji , v ji ) is the baseline formed by antenna j and antenna i, which is equal to the wavelength-normalized spacing between the two antennas j and i on the xy plane. is the Hermilte transpose of the visibility of the corresponding antenna j and antenna i after SNR improvement.

[0042] The covariance matrix is ​​subjected to subspace decomposition to obtain a noise subspace, including:

[0043] Perform eigenvalue decomposition on the covariance matrix, that is

[0044]

[0045] Among them, λ k is the kth eigenvalue, u k is λ k The corresponding eigenvector, H is the Hermilte transpose;

[0046] Extract the eigenvectors of NM minimum eigenvalues ​​to form the noise subspace

[0047]

[0048] Where M is the number of RFI sources, U s is the N×M signal subspace, U n is the noise subspace of N×(N-M+1), ∧ s and ∧ n are diagonal matrices with eigenvalues ​​of signal subspace and noise subspace respectively;

[0049] Step S33: Calculate the multi-signal classification pseudo-spectrum of the signal for the target area observed in the field of view of the brightness temperature image after the second SNR enhancement, that is,

[0050]

[0051] Where P M (ξ, η) represents the multi-signal classification pseudo-spectrum of the target area, a(ξ, η) represents the signal guidance vector, H is the Hermilte transpose, U n is the noise subspace;

[0052] Step S34: Perform peak detection and interpolation on the multi-signal classification pseudo-spectrum to obtain the geographic positioning coordinates of the RFI source, and convert the geographic positioning coordinates into longitude and latitude forms.

[0053] The present invention provides a device for locating radio frequency interference by combining brightness temperature images and subspaces, the device comprising:

[0054] The first enhancement module is configured to perform original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene, determine the ocean constant brightness temperature and the land constant brightness temperature in the earth scene using a forward model, construct a simulated background brightness temperature image of the earth scene, and perform a difference between the original scene brightness temperature image and the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR enhancement;

[0055] A second enhancement module is configured to perform image processing on the original scene brightness temperature image, generate a radio frequency interference (RFI) mask corresponding to the original scene brightness temperature image, and perform a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR enhancement to obtain a brightness temperature image after the second SNR enhancement;

[0056] Positioning module: configured to perform Fourier transform on the brightness temperature image after the second SNR improvement to obtain visibility samples; construct a covariance matrix of the brightness temperature image after the second SNR improvement, perform subspace decomposition on the covariance matrix, and geolocate the RFI source observed in the field of view of the brightness temperature image after the second SNR improvement.

[0057] The present invention provides a computer-readable storage medium, wherein a plurality of instructions are stored in the storage medium; the plurality of instructions are used for a processor to load and execute the method as described above.

[0058] An electronic device provided by the present invention is characterized in that the electronic device comprises:

[0059] A processor, which is used to execute multiple instructions;

[0060] A memory for storing a plurality of instructions;

[0061] The plurality of instructions are used to be stored by the memory and loaded and executed by the processor to implement the method as described above.

[0062] Beneficial technical effects brought by the present invention:

[0063] (1) The present invention suppresses background noise in the image domain, enhances RFI signals, improves the signal-to-noise ratio, and uses the MUSIC algorithm for positioning in the spatial domain, thereby improving the positioning accuracy of the RFI source under low signal-to-noise ratio conditions.

[0064] (2) The present invention performs two signal-to-noise ratio enhancements in the image domain, improving the signal-to-noise ratio by removing the earth background and enhancing the RFI source target. The high signal-to-noise ratio covariance matrix obtained thereby reduces the influence of the earth background and improves the RFI source positioning accuracy.

[0065] (3) The present invention reduces the influence of the earth background by including power, and the MUSIC pseudo-spectrum of the present invention obtains a sharper peak and lower side lobes, which significantly improves the contrast. Therefore, compared with the prior art (traditional MUSIC method), the present invention can obtain higher positioning accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A flow chart of a method for locating radio frequency interference by combining brightness temperature images and subspaces provided in an embodiment of the present invention;

[0067] Figure 2 A BT snapshot image reconstructed by standard HFFT (hexagram fast Fourier transform) provided in the first embodiment of the present invention;

[0068] Figure 3 The simulated original brightness temperature scene image with RFI simulation added provided in the first embodiment of the present invention;

[0069] Figure 4 A background brightness temperature image estimated using the sea-land forward model provided in the first embodiment of the present invention;

[0070] Figure 5 The brightness temperature image after the first SNR improvement provided in the first embodiment of the present invention;

[0071] Figure 6 The image obtained by binarizing the original brightness temperature scene image provided in the first embodiment of the present invention;

[0072] Figure 7 The expanded image provided by the first embodiment of the present invention;

[0073] Figure 8 The image after the harmonic components are supplemented provided in the first embodiment of the present invention;

[0074] Fig. 9 The brightness temperature image after the second SNR improvement provided in the first embodiment of the present invention;

[0075] Fig.10 A pseudo spectrum of a traditional MUSIC method obtained from a visibility sample of a snapshot provided in the first embodiment of the present invention;

[0076] Fig.11 A 3D view of a pseudo spectrum of a conventional MUSIC method obtained from visibility samples of a snapshot provided in the first embodiment of the present invention;

[0077] Fig.12 The MUSIC pseudo spectrum of the method of the present invention obtained from the visibility sample after SNR enhancement of the snapshot provided in the first embodiment of the present invention;

[0078] Fig.13 A 3D view of the MUSIC pseudo-spectrum of the method of the present invention obtained from the SNR-enhanced visibility samples of the snapshot provided in the first embodiment of the present invention.

[0079] Fig.14 A BT snapshot image reconstructed by standard HFFT provided in the second embodiment of the present invention;

[0080] Fig.15 The brightness temperature scene image after the SNR is improved twice provided in the second embodiment of the present invention;

[0081] Fig.16 The pseudo spectrum of the traditional MUSIC method obtained from the visibility samples of the snapshot provided in the second embodiment of the present invention;

[0082] Fig.17 A 3D view of a pseudo spectrum of a conventional MUSIC method obtained from visibility samples of a snapshot provided in the second embodiment of the present invention.

[0083] Fig.18 The MUSIC pseudo spectrum of the method of the present invention obtained from the visibility sample after SNR enhancement of the snapshot provided in the second embodiment of the present invention;

[0084] Fig.19 A 3D view of the MUSIC pseudo-spectrum of the method of the present invention obtained from the SNR-enhanced visibility samples of the snapshot provided in the second embodiment of the present invention.

[0085] Fig. 20 The BT snapshot image reconstructed by standard HFFT provided in the third embodiment of the present invention;

[0086] Fig.21 The brightness temperature scene image after the SNR is improved twice provided in the third embodiment of the present invention;

[0087] Fig. 22 The pseudo spectrum of the traditional MUSIC method obtained from the visibility samples of the snapshot provided in the third embodiment of the present invention;

[0088] Fig.23 A 3D view of a pseudo spectrum of a conventional MUSIC method obtained from visibility samples of a snapshot provided in the third embodiment of the present invention.

[0089] Fig.24 The MUSIC pseudo spectrum of the method of the present invention obtained from the visibility sample after SNR enhancement of the snapshot provided in the third embodiment of the present invention;

[0090] Fig.25 A 3D view of the MUSIC pseudo-spectrum of the method of the present invention obtained from the visibility samples of the snapshot after SNR enhancement provided in the third embodiment of the present invention.

[0091] Fig.26 A bar chart showing the detection rate and false detection rate of the traditional MUSIC method and the method of the present invention obtained by Monte Carlo simulation according to the fourth embodiment of the present invention;

[0092] Fig. 27 The original brightness temperature scene image of two very close RFI sources located at the sea-land boundary provided in the fourth embodiment of the present invention;

[0093] Fig.28The original brightness temperature scene image of two very close RFI sources located on land provided in the fourth embodiment of the present invention;

[0094] Fig.29 The positioning result of two very close RFI sources located at the sea-land boundary scene provided in the fourth embodiment of the present invention;

[0095] Fig.30 This is the positioning result of the scenario where two very close RFI sources are located on land, provided in the fourth embodiment of the present invention. DETAILED DESCRIPTION

[0096] The present invention is described in detail below in conjunction with the accompanying drawings and embodiments.

[0097] like Figure 1 As shown, the present invention proposes a method for locating radio frequency interference by combining brightness temperature image and subspace, and the method comprises the following steps:

[0098] Step S1: performing original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene, using a forward model to determine the ocean constant brightness temperature and the land constant brightness temperature in the earth scene, constructing a simulated background brightness temperature image of the earth scene, and performing a difference between the original scene brightness temperature image and the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR improvement;

[0099] Step S2: performing image processing on the original scene brightness temperature image to generate a radio frequency interference (RFI) mask corresponding to the original scene brightness temperature image, performing a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR improvement to obtain the brightness temperature image after the second SNR improvement;

[0100] Step S3: Perform Fourier transform on the brightness temperature image after the second SNR improvement to obtain visibility samples; construct a covariance matrix of the brightness temperature image after the second SNR improvement, perform subspace decomposition on the covariance matrix, and geolocate the RFI source observed in the field of view of the brightness temperature image after the second SNR improvement.

[0101] Furthermore, the step S1 comprises:

[0102] Step S11: performing original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene; obtaining the longitude and latitude coordinates corresponding to each pixel point in the earth's field of view of the original scene brightness temperature image, and obtaining the sea and land mask corresponding to the pixel point;

[0103] Step S12: Based on the sea and land masks corresponding to each pixel point, the pixel points in the earth's field of view are classified into two categories: pixel points located at the ocean position and pixel points located at the land position;

[0104] Step S13: determining the ocean constant brightness temperature and the land constant brightness temperature of the observed earth scene based on the forward model;

[0105] Step S14: setting the brightness temperature of pixels located at the ocean position in the original scene brightness temperature image to the ocean constant brightness temperature, and setting the brightness temperature of pixels located at the land position in the original scene brightness temperature image to the land constant brightness temperature, to obtain a simulated background brightness temperature image;

[0106] Step S15: subtract the original scene brightness temperature image from the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR enhancement.

[0107] Furthermore, in step S11, the latitude and longitude coordinates corresponding to each pixel in the original scene brightness temperature image are obtained, and the sea and land mask corresponding to the pixel is obtained from the ISEA grid point closest to the latitude and longitude coordinates corresponding to the pixel. ISEA is Icosahedral Snyder Equal Area, which is a global map projection method.

[0108] The original scene brightness temperature image is:

[0109] T raw =F -1 (V raw ) Formula 1

[0110] where F -1 (·) represents HFFT inversion, T raw represents the original scene brightness temperature image, V raw Represents the raw measured visibility of microwave radiation signals emitted from the Earth's surface.

[0111] In step S13, the ocean constant brightness temperature and the land constant brightness temperature are calculated as follows:

[0112]

[0113]

[0114]

[0115] in, is the constant brightness temperature over land, is the antenna temperature; Calculate the visibility of 1K brightness temperature for the Earth's surface, visibility of cold sky, ocean, sun and / or moon, solar reflection, and antenna back lobe; are the ocean constant brightness temperatures for vertical and horizontal polarizations, is the constant brightness temperature of the polarized ocean, is the constant brightness temperature of the H-polarized ocean, Γ v (SST, SSS, θ) is the V polarization Fresnel reflection coefficient, Γ h (SST, SSS, θ) is the H-polarization Fresnel reflection coefficient, SST represents the sea surface temperature, SSS represents the sea surface salinity, WS represents the wind speed, the default values ​​of SST, SSS, and WS are set to the average values ​​of 15°C, 35psu, and 4m / s, respectively, θ is the zenith angle of the target in the sensor direction, and r(·) is the Fresnel reflection coefficient model, which is calculated by:

[0116]

[0117]

[0118] where ε s is the complex dielectric constant of the ocean.

[0119] ε s The calculation formula is

[0120]

[0121] In this embodiment, It is the visibility measured considering the land area within the unit circle as a uniform source with BT=1K.

[0122] The step S15: subtracting the original scene brightness temperature image from the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR enhancement, wherein:

[0123]

[0124]

[0125] is the brightness temperature image after the first SNR improvement, T raw is the original scene brightness temperature image.

[0126] Furthermore, the step S2 comprises:

[0127] Step S21: binarizing the original scene brightness temperature image according to the brightness temperature threshold.

[0128]

[0129] Where T this the brightness temperature image after binarization;

[0130] Step S22: dilate the binarized brightness temperature image.

[0131] T dilated =max{T th (ξ-s, η-t)+B(s,t)+a(bd / r)} Formula 11

[0132] Among them, T dilated is the binary brightness temperature image after expansion, B(s, t) is a disk-shaped structural element with radius r = (Δξ, Δη), d is the distance from (ξ, η) to the center of the structural element B(s, t), a and b are coefficients related to non-planar expansion, a = 1, b = 1.5, T th (ξ-s, η-t) is the brightness temperature image after binarization, ξ and η are the horizontal and vertical coordinates of the pixel points in the field of view of the inversion image, s and t are the horizontal and vertical coordinates of the pixel points in the structuring element B, and Δξ and Δη are the unit pixel spacings of the field of view of the inversion image.

[0133] Step S23: The expanded binary brightness temperature image is inverted by HFFT to obtain a brightness temperature image of the supplementary harmonic component, where:

[0134] V HCS =F(T dilated ) Formula 12

[0135] Y HCS =F -1 (V HCS ) Formula 13

[0136] V HCS represents the brightness temperature image of the complementary harmonic component, F(T dilated ) is the response function, T dilated is the brightness temperature image after expansion, T HCS To supplement the brightness temperature image of the harmonic component, F -1 (V HCS ) is HFFT transformation, V HCS To supplement the visibility of harmonic components.

[0137] In this embodiment, the expanded binary brightness temperature image is input as a source display into the SAIR system for inversion to obtain a brightness temperature image of the supplementary harmonic component.

[0138] Step S24: transform the threshold range of the brightness temperature image of the supplementary harmonic component to [0.3, 3] to generate an RFI mask

[0139]

[0140] Where Tmask is the RFI mask, T HCS It is a brightness temperature image that supplements the harmonic components.

[0141] Step S25: performing a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR enhancement to obtain a brightness temperature image after the second SNR enhancement.

[0142]

[0143] in represents the brightness temperature image after the second SNR improvement, Tmask represents the RFI mask, ° represents the Hadamard product operation, This is the brightness temperature image after the first SNR improvement.

[0144] The step S3 comprises:

[0145] Step S31: Perform Fourier transform on the brightness temperature image after the second SNR enhancement to obtain visibility samples.

[0146]

[0147] in, represents visibility samples, F(·) represents Fourier transform, This is the brightness temperature image after the second SNR improvement.

[0148] Step S32: construct the covariance matrix of the brightness temperature image after the second SNR enhancement.

[0149]

[0150] in, represents the SNR covariance matrix, is the element of the covariance matrix, (uij, vij) is the baseline formed by antenna i and antenna j, which is equal to the wavelength-normalized spacing between the two antennas i and j on the xy plane, is the visibility of the corresponding antenna i and antenna j after SNR improvement, H is the Hermilte transpose, i, j are the antenna numbers, and N is the number of antennas; (u ji , v ji ) is the baseline formed by antenna j and antenna i, which is equal to the wavelength-normalized spacing between the two antennas j and i on the xy plane. is the Hermilte transpose of the visibility of the corresponding antenna j and antenna i after SNR improvement.

[0151] The covariance matrix is ​​subjected to subspace decomposition to obtain a noise subspace, including:

[0152] Perform eigenvalue decomposition on the covariance matrix, that is

[0153]

[0154] Among them, λ k is the kth eigenvalue, u k is λ k The corresponding eigenvector, H is the Hermilte transpose.

[0155] Extract the eigenvectors of NM minimum eigenvalues ​​to form the noise subspace

[0156]

[0157] Where N is the number of antennas, M is the number of RFI sources, and U is the s is the N×M signal subspace, U n is the noise subspace of N×(N-M+1), ∧ s and ∧ n are diagonal matrices with the eigenvalues ​​of the signal subspace and the noise subspace, respectively.

[0158] Step S33: Calculate the multi-signal classification (MUSIC) pseudo-spectrum of the signal for the target area observed in the field of view of the brightness temperature image after the second SNR improvement, that is,

[0159]

[0160] Where P M (ξ, η) represents the MUSIC pseudo-spectrum of the target area, a(ξ, η) represents the signal-directed vector, H is the Hermilte transpose, and U n is the noise subspace.

[0161] Step S34: Perform peak detection and interpolation on the MUSIC pseudo-spectrum to obtain the geographic location coordinates of the RFI source, and convert the geographic location coordinates into longitude and latitude forms.

[0162] The following four embodiments, combined Figure 2-Figure 30 The method of locating radio frequency interference combining brightness temperature image and subspace according to the present invention is described.

[0163] Example 1: Testing based on simulation data

[0164] The simulation was carried out with reference to the relevant parameters of the SMOS (Soil Moisture and Ocean Salinity) satellite radiometer system. Figure 2 is a reconstructed BT image of a certain earth scene. Figure 1 Add two RFIs with intensities of 300K and 350K to get the simulated original brightness temperature image Figure 3 . Figure 4 is the scene brightness temperature estimated using the forward model. Figure 5 This is the brightness temperature image after the first SNR improvement. Figure 6 , 7 , 8 are the images after binarization, expansion, and harmonic component supplementation, respectively. Fig. 9 This is the brightness temperature image after the second SNR boost. After two SNR boosts, you can see that the RFI source is more prominent in the background. Fig.10 , Fig.12 It is the MUSIC pseudo-spectral image and its 3D image obtained by the traditional MUSIC algorithm. Fig.11 , Fig.13 The pseudo-spectral image of MUSIC and its 3D image obtained by the method of the present invention have relatively large side lobes around the RFI position in the pseudo-spectral image of the traditional MUSIC algorithm, while the corresponding peak of RFI in the pseudo-spectral image of the method of the present invention is steeper and has smaller side lobes.

[0165] Example 2: Test based on SMOS L1A data (measured at 11:47:13 on August 18, 2013)

[0166] Fig.14 is the BT snapshot image reconstructed by standard HFFT, Fig.15 It is the brightness temperature image after the SNR is improved twice. Fig.16 , Fig.18 They are the original brightness temperature image and the MUSIC pseudo-spectrum calculated corresponding to the brightness temperature after SNR enhancement. Fig.17 , Fig.19 The 3D views are the original brightness temperature image and the MUSIC pseudo-spectrum calculated for the brightness temperature after SNR enhancement. It can be clearly seen that due to the influence of excessive background noise in the original image, the MUSIC spectrum cannot reflect the spatial distribution of RFI. The brightness temperature image after the second SNR enhancement is significantly improved, and the corresponding peak of RFI in the MUSIC pseudo-spectrum is very obvious, which can reflect the location of RFI.

[0167] Example 3: Test based on SMOS L1A data (measured at 11:10:23 on August 18, 2013)

[0168] Fig. 20 is the BT snapshot image reconstructed by standard HFFT, Fig.21 It is the brightness temperature image after the SNR is improved twice. Fig. 22 , Fig.24 They are the original brightness temperature image and the MUSIC pseudo-spectrum calculated corresponding to the brightness temperature after SNR enhancement. Fig.23 , Fig.25They are the 3D views of the original brightness temperature image and the MUSIC pseudo-spectrum calculated corresponding to the brightness temperature after SNR enhancement. For this snapshot, the traditional MUSIC algorithm cannot locate the position of RFI, but the method of the present invention can achieve RFI positioning, and the MUSIC pseudo-spectrum can reflect the RFI position, indicating that the method of the present invention is better than the traditional MUSIC algorithm in RFI positioning under low SNR conditions.

[0169] Example 4: Testing based on simulation data

[0170] The simulation was carried out with reference to the relevant parameters of the SMOS satellite radiometer system. The Monte Carlo simulation method was used to evaluate the detection rate and false alarm rate of the proposed algorithm and the traditional MUSIC algorithm. Figure 2 RFI with intensities ranging from 300K to 500K was randomly introduced into the experiment. The instrument response of a 1K point source located at (ξ,η) used in the experiment has a BT of about 0.64K in the BT map, resulting in an increase of about 192K to 320K in the BT at the added location. Both algorithms were used to locate the RFI, and their detection and false detection rates were evaluated. This process was repeated 1000 times. Fig.26 The left side of the middle bar graph shows the detection rate, and the right side shows the false positive rate. The proposed method is represented by the dashed column, while the traditional MUSIC algorithm is represented by the solid column. The bar graph shows that under the same conditions, the detection rate of the proposed method reaches 84.5%, which is significantly higher than the 43.1% of the traditional MUSIC algorithm. At the same time, the false positive rate can be kept at a relatively low level of 18.6%. This shows that the proposed method achieves a higher detection rate while maintaining a lower false positive rate than the traditional MUSIC algorithm. Fig. 27 and Fig.28 These are two scenarios where the RFIs are located close to each other, one at the sea-land boundary and the other on land. Fig.29 and Fig.30 The positioning results of the two scenes using the traditional brightness temperature algorithm, the traditional MUSIC algorithm, and the method of the present invention are respectively plotted, wherein the red cross represents the true position of the RFI, the square represents the positioning result of the method of the present invention, the circle represents the positioning result of the traditional MUSIC algorithm, and the triangle represents the positioning result of the traditional brightness temperature method. Overall, the positioning results of the method of the present invention in the two scenes are closer to the true position of the RFI, indicating that the positioning accuracy of the method of the present invention is better than that of the traditional brightness temperature algorithm and the traditional MUSIC algorithm in the low SNR scene.

[0171] The present invention also provides a device for guiding a microwave radiometer to detect salinity based on real-time spectrum monitoring results, the device comprising:

[0172] The first enhancement module is configured to perform original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene, determine the ocean constant brightness temperature and the land constant brightness temperature in the earth scene using a forward model, construct a simulated background brightness temperature image of the earth scene, and perform a difference between the original scene brightness temperature image and the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR enhancement;

[0173] A second enhancement module is configured to perform image processing on the original scene brightness temperature image, generate a radio frequency interference (RFI) mask corresponding to the original scene brightness temperature image, and perform a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR enhancement to obtain a brightness temperature image after the second SNR enhancement;

[0174] Positioning module: configured to perform Fourier transform on the brightness temperature image after the second SNR improvement to obtain visibility samples; construct a covariance matrix of the brightness temperature image after the second SNR improvement, perform subspace decomposition on the covariance matrix, and geolocate the RFI source observed in the field of view of the brightness temperature image after the second SNR improvement.

[0175] The above specific embodiments only describe the design principle of the present invention. The shapes and names of the components in the description may be different and are not limited. Therefore, those skilled in the art in the field of the present invention may modify or replace the technical solutions recorded in the above embodiments; and these modifications and replacements do not deviate from the creative purpose and technical solutions of the present invention and should all fall within the protection scope of the present invention.

Claims

1. A method for locating radio frequency interference by combining brightness temperature image and subspace, characterized in that: The method comprises the following steps: Step S1: performing original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene, using a forward model to determine the ocean constant brightness temperature and the land constant brightness temperature in the earth scene, constructing a simulated background brightness temperature image of the earth scene, and performing a difference between the original scene brightness temperature image and the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR improvement; Step S2: performing image processing on the original scene brightness temperature image to generate a radio frequency interference (RFI) mask corresponding to the original scene brightness temperature image, performing a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR improvement to obtain the brightness temperature image after the second SNR improvement; Step S3: Perform Fourier transform on the brightness temperature image after the second SNR improvement to obtain visibility samples; construct a covariance matrix of the brightness temperature image after the second SNR improvement, perform subspace decomposition on the covariance matrix, and geolocate the RFI source observed in the field of view of the brightness temperature image after the second SNR improvement.

2. The method according to claim 1, characterized in that The step S1 comprises: Step S11: performing original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene; obtaining the longitude and latitude coordinates corresponding to each pixel point in the earth's field of view of the original scene brightness temperature image, and obtaining the sea and land mask corresponding to the pixel point; Step S12: Based on the sea and land masks corresponding to each pixel point, the pixel points in the earth's field of view are classified into two categories: pixel points located at the ocean position and pixel points located at the land position; Step S13: determining the ocean constant brightness temperature and the land constant brightness temperature of the observed earth scene based on the forward model; Step S14: setting the brightness temperature of pixels located at the ocean position in the original scene brightness temperature image to the ocean constant brightness temperature, and setting the brightness temperature of pixels located at the land position in the original scene brightness temperature image to the land constant brightness temperature, to obtain a simulated background brightness temperature image; Step S15: subtract the original scene brightness temperature image from the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR enhancement.

3. The method according to any one of claims 1 to 2, characterized in that: The calculation method of the ocean constant brightness temperature and the land constant brightness temperature is: in, is the constant brightness temperature over land, is the antenna temperature; Calculate the visibility of 1K brightness temperature for the Earth's surface, visibility of cold sky, ocean, sun and / or moon, solar reflection, and antenna back lobe; are the ocean constant brightness temperatures for vertical and horizontal polarizations, is the constant brightness temperature of the polarized ocean, is the constant brightness temperature of the H-polarized ocean, Γ v (SST,SSS,θ) is the V polarization Fresnel reflection coefficient, Γ h (SST, SSS, θ) is the H-polarization Fresnel reflection coefficient, SST is the sea surface temperature, SSS is the sea surface salinity, WS is the wind speed, and the default values ​​of SST, SSS, and WS are set to the average values ​​of 15°C, 35 psu, and 4 m / s, respectively. θ is the zenith angle of the target in the sensor direction, and Γ(·) is the Fresnel reflection coefficient model, which is calculated by: where ε s is the complex dielectric constant of the ocean.

4. The method according to claim 1, characterized in that The step S2 comprises: Step S21: binarizing the original scene brightness temperature image according to the brightness temperature threshold; Step S22: dilating the binarized brightness temperature image; Step S23: The expanded binary brightness temperature image is inverted by HFFT to obtain a brightness temperature image of the supplementary harmonic component, where: V HCS =F(T dilated ) Formula 12 T HCS =F -1 (V HCS ) Formula 13 V HCS represents the brightness temperature image of the complementary harmonic component, F(T dilated ) is the response function, T dilated is the brightness temperature image after expansion, T HCS To supplement the brightness temperature image of the harmonic component, F -1 (V HCS ) is HFFT transformation, V HCS To supplement the visibility of harmonic components; Step S24: transform the threshold range of the brightness temperature image of the supplementary harmonic component to [0.3, 3] to generate an RFI mask Where T mask is the RFI mask, T HCS To supplement the brightness temperature image of the harmonic component; Step S25: performing a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR enhancement to obtain a brightness temperature image after the second SNR enhancement.

5. The method as claimed in claim 4, characterized in that The step S3 comprises: Step S31: Perform Fourier transform on the brightness temperature image after the second SNR enhancement to obtain visibility samples. Among them, V SNRup represents the visibility sample, F(·) represents the Fourier transform, This is the brightness temperature image after the second SNR improvement; Step S32: construct the covariance matrix of the brightness temperature image after the second SNR enhancement. Among them, R SNRup represents the SNR covariance matrix, is the element of the covariance matrix, (u ij ,v ij ) is the baseline formed by antenna i and antenna j, which is equal to the wavelength-normalized spacing between the two antennas i and j on the xy plane. SNRup (u ij , v ij ) is the visibility of the corresponding antenna i and antenna j after SNR improvement, H is the Hermilte transpose, i, j are the antenna numbers, and N is the number of antennas; (u ji , v ji ) is the baseline formed by antenna j and antenna i, which is equal to the wavelength-normalized spacing between the two antennas j and i on the xy plane. SNRup (u ji ,v ji ) H is the Hermilte transpose of the visibility of the corresponding antenna j and antenna i after SNR improvement; The covariance matrix is ​​subjected to subspace decomposition to obtain a noise subspace, including: Perform eigenvalue decomposition on the covariance matrix, that is Among them, λ k is the kth eigenvalue, u k is λ k The corresponding eigenvector, H is the Hermilte transpose; Extract the eigenvectors of NM minimum eigenvalues ​​to form the noise subspace Where M is the number of RFI sources, U s is the N×M signal subspace, U n is the noise subspace of N×(N-M+1), ∧ s and ∧ n are diagonal matrices with eigenvalues ​​of signal subspace and noise subspace, respectively; Step S33: Calculate the multi-signal classification pseudo-spectrum of the signal for the target area observed in the field of view of the brightness temperature image after the second SNR enhancement, that is, Where P M (ξ,η) represents the multi-signal classification pseudo-spectrum of the target area, a(ξ,η) represents the signal guidance vector, H is the Hermilte transpose, and U n is the noise subspace; Step S34: Perform peak detection and interpolation on the multi-signal classification pseudo-spectrum to obtain the geographic positioning coordinates of the RFI source, and convert the geographic positioning coordinates into longitude and latitude forms.

6. A device for locating radio frequency interference by combining brightness temperature image and subspace, characterized in that: The device comprises: The first enhancement module is configured to perform original measurement on the microwave radiation signal emitted from the earth's surface to obtain an original scene brightness temperature image of the observed earth scene, determine the ocean constant brightness temperature and the land constant brightness temperature in the earth scene using a forward model, construct a simulated background brightness temperature image of the earth scene, and perform a difference between the original scene brightness temperature image and the simulated background brightness temperature image to obtain a brightness temperature image after the first SNR enhancement; A second enhancement module is configured to perform image processing on the original scene brightness temperature image, generate a radio frequency interference (RFI) mask corresponding to the original scene brightness temperature image, and perform a Hadamard product operation on the RFI mask and the brightness temperature image after the first SNR enhancement to obtain a brightness temperature image after the second SNR enhancement; Positioning module: configured to perform Fourier transform on the brightness temperature image after the second SNR improvement to obtain visibility samples; construct a covariance matrix of the brightness temperature image after the second SNR improvement, perform subspace decomposition on the covariance matrix, and geolocate the RFI source observed in the field of view of the brightness temperature image after the second SNR improvement.

7. A computer-readable storage medium, wherein a plurality of instructions are stored in the storage medium; the plurality of instructions are used for a processor to load and execute the method as claimed in any one of claims 1 to 5.

8. An electronic device, characterized in that: The electronic device comprises: A processor, which is used to execute multiple instructions; A memory for storing a plurality of instructions; The plurality of instructions are used to be stored in the memory and loaded and executed by the processor according to any one of claims 1 to 5.

Images