Nearshore Error Suppression Method for Synthetic Aperture Radiometer

Abstract

A method for suppressing nearshore errors of a synthetic aperture radiometer. Aiming at the problems existing in the nearshore detection of the current synthetic aperture radiometer, when it is judged that the observation area is a land-sea boundary area, the modeled brightness temperature is used as prior information, and the method of taking the difference is adopted. The calibrated visibility function is subtracted from the modeled visibility function, and the difference inversion brightness temperature is obtained by using the difference visibility function. Finally, the modeled brightness temperature is added back to the imaging area to achieve the purpose of reducing nearshore errors. The method of the present invention can effectively reduce the nearshore errors caused by the large brightness temperature contrast at the land-sea boundary, thereby improving the detection accuracy of the synthetic aperture radiometer. The method is simple and reliable, can effectively improve the imaging quality of the synthetic aperture radiometer, improve the effective utilization rate of satellite remote sensing data, and has broad market application prospects.

Description

Technical Field

[0001] The present invention belongs to the technical field of space microwave remote sensing, and relates to a method for suppressing near-shore errors of a synthetic aperture radiometer. Background Art

[0002] Microwave radiation measurement technology was first used in the field of radio astronomy and has now found more and more applications in earth remote sensing, medicine, and target detection. To improve the spatial resolution of microwave radiometers, synthetic aperture technology has been introduced into microwave radiometers. A synthetic aperture radiometer is based on two-element interference technology, uses the complex correlation of antennas in pairs to obtain the visibility function, and realizes the observation of brightness temperature through the conversion model between the visibility function and the brightness temperature.

[0003] When a synthetic aperture radiometer detects marine elements in the near-shore area, due to the large brightness temperature gradient between the land area and the ocean area, the retrieved brightness temperature image has Gibbs oscillations. The oscillations spread throughout the brightness temperature image, resulting in too large errors, low utilization rate of near-shore ocean data, and ultimately limiting the inversion accuracy of marine elements. Therefore, solving the near-shore errors is crucial for the high-precision detection of synthetic aperture radiometers. Summary of the Invention

[0004] The technical problem solved by the present invention is: aiming at the problem of large near-shore detection errors of current synthetic aperture radiometers, a method for suppressing near-shore errors of a synthetic aperture radiometer is proposed. This method can suppress near-shore errors by reducing the contrast of brightness temperature between the ocean and land scenes, effectively improve the image quality, and improve the inversion accuracy of synthetic aperture radiometers.

[0005] The technical solution of the present invention is: a method for suppressing near-shore errors of a synthetic aperture radiometer, comprising the following steps:

[0006] (1) Use a synthetic aperture radiometer to observe the target scene, and obtain the original visibility function by using the complex correlation of antennas in pairs;

[0007] (2) Obtain the antenna temperature of the synthetic aperture radiometer;

[0008] (3) Judge whether the target scene belongs to a pure ocean scene, a pure land scene, or an ocean-land mixed scene according to the antenna temperature of the synthetic aperture radiometer; for a pure ocean scene or a pure land scene, directly use the original visibility function to invert the brightness temperature through the conversion model between the visibility function and the brightness temperature and end; for an ocean-land mixed scene, go to step (4);

[0009] (4) Calibrate the original visibility function to obtain the calibrated visibility function;

[0010] (5) Use the model brightness temperature as prior information to obtain the corresponding model visibility function;

[0011] (6) Subtract the calibrated visibility function from the model visibility function to obtain a difference visibility function, and use the difference visibility function to obtain the difference-inverted brightness temperature;

[0012] (7) Superimpose the model brightness temperature on the difference-inverted brightness temperature to obtain the inverted brightness temperature of the target scene of the synthetic aperture radiometer.

[0013] Preferably, the obtaining of the antenna temperature of the synthetic aperture radiometer is specifically as follows: A switch for switching different input signals is arranged at the front end of the receiver of the synthetic aperture radiometer. First, switch the switch to the antenna port, and through the observation of the target scene, obtain the original visibility function matrix V of the target scene 0 , the diagonal value of which is the autocorrelation output V A 0 , and the non-diagonal value is the cross-correlation coefficient M k 0 j ; then switch the switch to the calibration network composed of a calibration source and a power divider network in series, perform two-point calibration by switching between high and low temperatures through the calibration network, and calculate the gain G and the receiver noise temperature T of the synthetic aperture radiometer R ; subsequently, obtain the system noise temperature Finally, obtain the antenna temperature T A ,

[0014]

[0015] where η AN is the radiation efficiency of the antenna, L CA is the insertion loss of the connection cable between the antenna and the switch, T CA is the physical temperature of the connection cable between the antenna and the switch when observing the target scene, and T AN is the physical temperature of the antenna when observing the target scene.

[0016] Preferably, the judging of whether the target scene belongs to a pure ocean scene, a pure land scene or an ocean-land mixed scene according to the antenna temperature of the synthetic aperture radiometer is specifically as follows:

[0017]

[0018] where T l is the antenna temperature when observing a pure ocean scene, T h is the antenna temperature when observing a pure land scene. When the antenna temperature of the target scene is within the threshold interval [T l , T h , it is judged that the target scene is an ocean-land mixed scene.

[0019] Preferably, the calibration of the original visibility function to obtain the calibrated visibility function is specifically as follows: Let the calibrated visibility function be Vcal , then where g(0) kj is the value of the fringe elimination function at zero, used to correct the amplitude-phase error of the channel, is the non-diagonal value of the V cal matrix, is the diagonal value of the V cal matrix, equal to the antenna temperature, and the two together constitute the calibrated visibility function V cal .

[0020] Preferably, using the model brightness temperature as prior information to obtain its corresponding model visibility function, specifically:

[0021]

[0022] where V model is the model visibility function, S is the number of antenna elements, T model is the model brightness temperature, P is the number of brightness temperature discretizations, G is the system impulse response matrix, and the number of columns P of G is greater than the number of rows The elements in the G matrix are expressed as:

[0023]

[0024] where F k (ξ,η) and F i (ξ,η) are the antenna patterns of elements k and i, Ω k and Ω i are the corresponding antenna solid angles. The spatial frequency u = (x i -x k ) / λ0, v = (y i -y k ) / λ0, (x k , y k ) and (x i , y i ) are the coordinates of antenna elements k and i respectively, λ0 is the wavelength corresponding to the system center frequency, and the direction cosine , θ and are the azimuth angles.

[0025] Preferably, using the difference visibility function to obtain the difference-inverted brightness temperature, specifically: T Δ = G -1 V Δ , V Δ = V cal -V model , V cal is the calibrated visibility function.

[0026] The advantages of the present invention compared with the prior art are as follows: after determining that the observation area is a land-sea boundary area, the present invention uses the model brightness temperature as prior information and adopts a difference method, which can effectively reduce the near-shore error caused by the large brightness temperature contrast at the land-sea boundary, thereby improving the detection accuracy of the synthetic aperture radiometer. It is an error suppression method applicable to the near-shore detection of synthetic aperture radiometers. The method of the present invention is simple and reliable, can effectively improve the imaging quality of synthetic aperture radiometers, improve the effective utilization rate of satellite remote sensing data, and has broad market application prospects. Description of the Drawings

[0027] Figure 1 It is a flow chart of the traditional processing method of the synthetic aperture radiometer;

[0028] Figure 2 It is a flow block diagram of the method of the present invention;

[0029] Figure 3 It is a principle block diagram of the calibration method of the synthetic aperture radiometer of the present invention;

[0030] Figure 4 It is a schematic diagram of the antenna array arrangement in the embodiment of the present invention;

[0031] Figure 5 They are the results obtained by using the traditional processing method, where Figure (a) is the retrieved brightness temperature and Figure (b) is the brightness temperature error;

[0032] Figure 6 They are the results obtained by using the method of the present invention, where Figure (a) is the retrieved brightness temperature and Figure (b) is the brightness temperature error. Detailed Embodiment

[0033] The present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0034] Figure 1 It shows the traditional processing flow chart of the synthetic aperture radiometer. It mainly observes the target scene, obtains the original visibility function through the complex correlation of antennas in pairs, performs calibration processing to obtain the calibrated visibility function, and then obtains the retrieved brightness temperature of the target scene through the conversion model between the visibility function and the brightness temperature. It can be seen that it does not consider the large brightness temperature gradient between the land area and the ocean area.

[0035] Since the error caused by Gibbs oscillation is related to the contrast of the scene brightness temperature, in order to suppress this error, the method of the present invention adopts a difference method to reduce the brightness temperature mutation caused by the land-sea boundary in the image, thereby minimizing the influence of the error and improving the detection accuracy of the synthetic aperture radiometer. Figure 2The flowchart of the method of the present invention is given, and the specific method steps are as follows:

[0036] 1) Visibility function calibration

[0037] Figure 3 It is the principle block diagram for calibrating the synthetic aperture radiometer.

[0038] The switch before the receiver is used to select between different input signals. When the switch is switched to port A (antenna port), the system is in the observation mode. By observing the target scene, the original visibility function matrix V of the target scene is obtained 0 , the diagonal value of which is the autocorrelation output V A 0 , and the non-diagonal value is the cross-correlation coefficient M k 0 j .

[0039] When the switch is switched to port C (calibration port), the system is in the calibration mode. By switching the high and low temperatures through the calibration network for two-point calibration, the system gain G and the receiver noise temperature T can be calculated R .

[0040] Using the high and low temperatures output by the calibration network for two-point calibration to determine the system gain and the receiver noise temperature, specifically:

[0041]

[0042]

[0043] Among them, G C is the system gain at the calibration end, is the receiver noise temperature at the calibration end, (T hot ) C is the high temperature output by the calibration network, (T cold ) C is the low temperature output by the calibration network, V hot is the autocorrelation output of the correlator when observing the high-temperature calibration source, V cold is the autocorrelation output of the correlator when observing the low-temperature calibration source.

[0044] Through port conversion, the system gain and the system noise temperature at the calibration end are converted to the antenna port.

[0045]

[0046]

[0047] Among them, T sw is the physical temperature of the switch, S LA is the S parameter of the switch at the antenna port, S LCThe switch S-parameters for the calibration port. G is the system gain at the antenna end, and T R is the receiver noise temperature at the antenna end. For detailed information, please refer to the reference: Francesc Torres, Ignasi Corbella, Adriano Camps, and et al., “Denormalization of Visibilities for In-Orbit Calibration of Interferometric Radiometers”, IEEE Trans. Geosci. Remote Sens.

[0048] The system noise temperature can be obtained from the system gain G and the output of the visibility function as follows:

[0049]

[0050] where T SYS is the system noise temperature, is the autocorrelation output when observing the target scene.

[0051] Considering the insertion loss of each device in the system link, the antenna temperature T A is solved and its expression is:

[0052]

[0053] where η AN is the radiation efficiency of the antenna, L CA is the insertion loss of the connecting cable between the antenna and the switch, T CA is the physical temperature of the connecting cable between the antenna and the switch in the system's observation mode, and T AN is the physical temperature of the antenna when observing the scene.

[0054] The original visibility function is subjected to absolute calibration to obtain the calibrated visibility function V cal .

[0055]

[0056]

[0057] where g(0) kj is the value of the fringe-nulling function (the fringe-nulling function can be solved by a public method) at zero, which is used to correct the channel amplitude-phase error. is the off-diagonal value of the calibrated visibility function matrix, is the diagonal value of the calibrated visibility function matrix, which is equal to the antenna temperature. The two together constitute the calibrated visibility function V cal .

[0058] 2) Sea-land boundary discrimination

[0059] Use the antenna temperature T obtained in the previous step A to discriminate the scene. Since the antenna temperature T A is the weighted value of the scene brightness temperature and the antenna pattern within the observation range of the radiometer, therefore, for different observation scenes, the antenna temperature output by the synthetic aperture radiometer is different.

[0060] Due to the large difference in brightness temperature between land and sea, the output result of the antenna temperature T A can be divided into the following three cases:

[0061]

[0062] where T l is the antenna temperature output when observing a pure ocean area, and T h is the antenna temperature output when observing a pure land area.

[0063] Set the discrimination threshold interval [T l , T h for the sea-land boundary. When the antenna temperature of the target scene is within the threshold interval [T l , T h , it can be judged that the target scene is neither a pure ocean nor a pure land area, that is, the sea-land boundary area.

[0064] 3) Calculation of the model visibility function

[0065] The Earth is composed of land and ocean areas. From the long-term observation data of other satellites, the brightness temperature distribution of the entire Earth can be obtained as the brightness temperature of the Earth model. The brightness temperature of the Earth model can be updated periodically according to the situation to be as close as possible to the true brightness temperature distribution of the current observation scene. According to the longitude and latitude position information of the observation scene of the synthetic aperture radiometer, find the model brightness temperature distribution T M of the corresponding area of the observation scene in the brightness temperature of the Earth model.

[0066] From the conversion relationship between the brightness temperature and the visibility function, the model visibility function corresponding to the model brightness temperature can be obtained:

[0067]

[0068] where V model is the model visibility function, S is the number of antenna elements. T model is the model brightness temperature, and P is the number of brightness temperature discretizations. G is the system impulse response matrix. To ensure the data processing accuracy, it is generally required that the number of columns P of the matrix G is greater than the number of rows The G matrix is obtained from the actual antenna pattern of the system and is generally used for the brightness temperature inversion of synthetic aperture radiometers. Its expression is as follows:

[0069]

[0070] where F k (ξ,η) and F i (ξ,η) are the antenna patterns of elements k and i, and Ω k and Ω i are the corresponding antenna solid angles. The spatial frequency u = (x i -x k ) / λ0, v = (y i -y k ) / λ0, (x k , y k ) and (x i , y i ) are the coordinates of antenna elements k and i respectively, and λ0 is the wavelength corresponding to the center frequency of the system. The direction cosines θ and are the azimuth angles.

[0071] 4) Obtaining the brightness temperature by difference inversion

[0072] Subtract the calibrated visibility function V cal obtained in step 1) from the model visibility function V model obtained in step 3) to get the difference visibility function:

[0073] V Δ = V cal - V model (6)

[0074] where V Δ is the difference visibility function.

[0075] Since the amplitude of the difference visibility function is smaller than that of the actual visibility function, the difference visibility function can be used to obtain the brightness temperature with the minimum error. Through the brightness temperature inversion of the G matrix, the difference inversion brightness temperature T Δ can be obtained:

[0076] T Δ = G -1 V Δ (7)

[0077] Since the G matrix is an ill-conditioned matrix, in order to obtain a stable approximate solution of the above formula, the inversion of the G matrix can be solved by various regularization methods such as truncated singular value and Tikhonov regularization.

[0078] 5) Obtaining the brightness temperature of the target scene

[0079] Finally, on the retrieved brightness temperature T obtained in step 4), the modeled brightness temperature T Δ is superimposed to obtain the retrieved brightness temperature T model of the target scene of the synthetic aperture radiometer: Inv :

[0080] T Inv = T Δ + T model (8)

[0081] Embodiment

[0082] To further illustrate a method for suppressing near - shore errors of a synthetic aperture radiometer provided by the present invention, it is described in detail below in combination with specific examples:

[0083] In this embodiment, a Y - type array with 18 elements in a single arm is used for simulation, and the array layout is as shown in Figure 4 . A rectangular scene with a brightness temperature of 250K is set in the field of view as the land area; the brightness temperature of the rest of the scene is set to 100K as the ocean area. That is, the scene setting simulates the situation where there is a piece of land in the ocean.

[0084] Using the traditional processing method, directly perform brightness temperature inversion after calibration from the visibility function of the scene, and the obtained result is as shown in Figure 5 . It can be seen from the Figure 5 result that due to the large brightness temperature difference between the land (the red rectangular part in the image) and the ocean (the rest of the blue part in the image), there are large Gibbs oscillation errors in the retrieved brightness temperature, and the error is the largest near the shore, and the oscillation spreads to the entire image.

[0085] Using the near - shore error suppression method proposed by the present invention, set the modeled brightness temperature consistent with the brightness temperature distribution of the scene, subtract the modeled visibility function from the calibrated scene visibility function, and perform brightness temperature inversion using the obtained difference visibility function. The result is as shown in Figure 6 . It can be seen that the error is effectively suppressed, the land boundary of the retrieved brightness temperature is clear, and the obtained image quality is high.

[0086] Therefore, the near - shore error suppression method of the synthetic aperture radiometer proposed by the present invention can effectively suppress the near - shore error and improve the quality of the retrieved image on the premise of ensuring the accuracy of the modeled brightness temperature.

[0087] The content not described in detail in the specification of the present invention belongs to the well - known technology of those skilled in the art.

Claims

1. A method for suppressing near - shore errors of a synthetic aperture radiometer, characterized in that It includes the following steps: (1) Use a synthetic aperture radiometer to observe the target scene, and obtain the original visibility function by pairwise complex correlation of antennas; (2) Obtain the antenna temperature of the synthetic aperture radiometer; (3) Judge whether the target scene belongs to a pure ocean scene, a pure land scene or an ocean-land mixed scene according to the antenna temperature of the synthetic aperture radiometer; for a pure ocean scene or a pure land scene, directly use the conversion model between the visibility function and the brightness temperature, and use the original visibility function to invert the brightness temperature and end; for an ocean-land mixed scene, go to step (4); (4) Calibrate the original visibility function to obtain the calibrated visibility function; (5) Use the model brightness temperature as prior information to obtain the corresponding model visibility function; (6) Subtract the model visibility function from the calibrated visibility function to obtain the difference visibility function, and use the difference visibility function to obtain the difference-inverted brightness temperature; (7) Superimpose the model brightness temperature on the difference-inverted brightness temperature to obtain the inverted brightness temperature of the target scene of the synthetic aperture radiometer; The using the model brightness temperature as prior information to obtain the corresponding model visibility function is specifically: Among them, V model is the model visibility function, S is the number of antenna elements, and T model is the model brightness temperature, P is the number of brightness temperature discretizations, G is the system impulse response matrix, and the number of columns P of G is greater than the number of rows The elements in the G matrix are expressed as: Among them, F k (ξ, η) and F i (ξ, η) are the antenna patterns of element k, i, Ω k and Ω i are the corresponding antenna solid angles; the spatial frequency u = (x i - x k ) / λ0, v = (y i - y k ) / λ0, (x k , y k ) and (x i , y i ) are the coordinates of antenna elements k, i respectively, λ0 is the wavelength corresponding to the system center frequency, and the direction cosines θ and are the azimuth angles.

2. The method for suppressing near - shore errors of a synthetic aperture radiometer according to claim 1, characterized in that: The method for obtaining the antenna temperature of a synthetic aperture radiometer is specifically as follows: A switch for switching different input signals is provided at the front end of the receiver of the synthetic aperture radiometer. First, the switch is switched to the antenna port, and through the observation of the target scene, the original visibility function matrix V of the target scene is obtained 0 , and the diagonal values thereof are autocorrelation outputs The non-diagonal values are cross-correlation coefficients Then, the switch is switched to a calibration network composed of a calibration source and a power divider network connected in series. Two-point calibration is performed by switching between high and low temperatures through the calibration network, and the gain G and the receiver noise temperature T of the synthetic aperture radiometer are calculated R ; Subsequently, the system noise temperature is obtained Finally, the antenna temperature T is obtained A , where η AN is the radiation efficiency of the antenna, L CA is the insertion loss of the connecting cable between the antenna and the switch, T CA is the physical temperature of the connecting cable between the antenna and the switch when observing the target scene, T AN is the physical temperature of the antenna when observing the target scene.

3. The method for suppressing near - shore errors of a synthetic aperture radiometer according to claim 2, characterized in that: The judging whether the target scene belongs to a pure ocean scene, a pure land scene or an ocean-land mixed scene according to the antenna temperature of the synthetic aperture radiometer is specifically: where T l is the antenna temperature when observing a pure ocean scene, and T h is the antenna temperature when observing a pure land scene. When the antenna temperature of the target scene is within the threshold range [T l , T h , it is determined that the target scene is an ocean-land mixed scene.

4. The method for suppressing near - shore errors of a synthetic aperture radiometer according to claim 2, characterized in that: Calibrate the original visibility function to obtain the calibrated visibility function, specifically: Let the calibrated visibility function be V cal , then where g(0) kj is the value of the fringe-nulling function at zero, used to correct the channel amplitude-phase error, is the non-diagonal value of the V cal matrix, is the diagonal value of the V cal matrix, equal to the antenna temperature, and the two together constitute the calibrated visibility function V cal .

5. The method for suppressing near - shore errors of a synthetic aperture radiometer according to claim 1, characterized in that: The method for obtaining the differential inversion brightness temperature using the differential visibility function is specifically as follows: T Δ = G -1 V Δ V Δ = V cal - V model V cal is the calibration visibility function.

Images