A Method for Improving Brightness Temperature Near Shore of One-Dimensional Synthetic Aperture Microwave Radiometer Based on Node Sampling

Abstract

The present invention belongs to the technical field of one-dimensional synthetic aperture nearshore data processing, and relates to a method for improving the nearshore brightness temperature of a one-dimensional synthetic aperture microwave radiometer based on node sampling, which includes the following steps: parameter initialization, oversampling processing of the frequency-domain signal containing Gibbs contamination; finding nodes in the oversampled brightness temperature image through the positive and negative values of the second derivative, and screening ordinary nodes and mutation nodes; performing iterative correction on ordinary nodes, including calculating the high-resolution Laplacian operator, updating the brightness temperature value of ordinary nodes, and circular iteration; performing brightness temperature correction on mutation nodes, calculating the trend factor according to the brightness temperature difference value of the nearest adjacent nodes, and updating the brightness temperature value of mutation nodes; outputting the corrected nearshore data brightness temperature value. The present invention uses the method of node sampling to eliminate Gibbs oscillations and reduce the land-sea buffer zone. On the basis of the completion of iteration for all ordinary nodes, the mutation nodes are corrected by calculating the trend factor to achieve the purpose of reducing the land-sea buffer zone.

Description

Technical Field

[0001] The present invention relates to the technical field of one-dimensional synthetic aperture nearshore data processing, and particularly to the field of eliminating nearshore influence of one-dimensional synthetic aperture microwave radiometers. Background Art

[0002] The application of synthetic aperture technology in microwave radiometers has successfully solved the contradiction between spatial resolution and antenna size, greatly improving the spatial resolution of observations, and at the same time significantly reducing the volume and weight of the radiometer. However, these advantages come at the cost of increased system design and signal processing complexity. Compared with traditional real-aperture microwave radiometers, synthetic aperture microwave radiometers use complex antenna arrays and interferometric imaging principles, thus introducing various errors. These errors not only come from the instrument itself but are also affected by external factors, posing certain challenges to high-precision brightness temperature measurement.

[0003] The advantage of synthetic aperture radiometer technology is to improve the spatial resolution of low-frequency radiometers without increasing the physical size of the antenna through aperture synthesis technology. However, due to its high system complexity, the difficulty of data processing also increases accordingly, especially showing discreteness and frequency-domain truncation problems in satellite payload interferometry. When there is a brightness temperature jump in the field of view, large-scale Gibbs phenomena are likely to occur, which is unacceptable for salinity measurement with extremely high requirements for brightness temperature accuracy.

[0004] Synthetic aperture microwave radiometers will face the problem that land and ocean areas are aliased into the field of view of synthetic aperture measurement in the nearshore area. Due to the characteristics of aperture synthesis, frequency-domain information loss will occur during interferometric measurement. On the one hand, it causes the field of view aliasing phenomenon of the instrument, narrowing the available mapping width; on the other hand, if there is a brightness temperature jump in the field of view, it will lead to Gibbs contamination in the retrieved brightness temperature image. Therefore, more intractable land-sea mutual contamination problems will be faced in the nearshore area than real-aperture radiometers. And salinity inversion requires high-precision brightness temperature input, and the land contamination problem in nearshore measurement brightness temperature must be corrected.

[0005] The information disclosed in this background art section is only intended to enhance the overall understanding of the present invention and should not be regarded as an admission or any form of implication that this information constitutes prior art already known to those of ordinary skill in the art. Summary of the Invention

[0006] In order to overcome the Gibbs contamination problem caused by the bright temperature jump phenomenon in the field of view during interferometric measurement and the problem of land-sea boundary ambiguity in the prior art, the present invention proposes a method for improving the bright temperature near the shore of a one-dimensional synthetic aperture microwave radiometer based on node sampling. After initializing the inversion parameters, the method performs oversampling on the frequency-domain signal, and then screens the nodes of the oversampled bright temperature image to identify the mutation nodes and non-mutation nodes. For non-mutation nodes, iterative correction operations are performed. On this basis, for mutation nodes, adjustment is made according to the trend factor calculated based on the bright temperature difference between the mutation nodes and their adjacent nodes, and finally the correction of the bright temperature at the mutation nodes is achieved. After the entire process is completed, the corrected bright temperature value of the near-shore data is output.

[0007] Thus, the present invention uses node sampling and iterative calculation of the high-resolution Laplacian operator of the nodes to suppress Gibbs contamination. On this basis, the purpose of reducing the land-sea fuzzy area is achieved by calculating the trend factor.

[0008] The technical solution of the present invention is as follows:

[0009] A method for improving the bright temperature near the shore of a one-dimensional synthetic aperture microwave radiometer based on node sampling, comprising the following steps:

[0010] Step 1, initialize the near-shore data parameters of the one-dimensional synthetic aperture microwave radiometer, and perform oversampling on the frequency-domain signal containing Gibbs contamination, set the oversampling factor, and obtain a one-dimensional oversampled bright temperature image;

[0011] Step 2, find the nodes in the oversampled bright temperature image through the positive and negative values of the second derivative, set the threshold S, and screen the ordinary nodes and mutation nodes; when the differences on both sides of the node are greater than the threshold S at the same time, this node is regarded as a mutation node, and other nodes are regarded as ordinary nodes;

[0012] Step 3, perform iterative correction on the ordinary nodes, including calculating the high-resolution Laplacian operator, updating the bright temperature value of the ordinary nodes, and performing loop iteration; the calculation expression of the high-resolution Laplacian operator is: (1)

[0013] Wherein, represents the high-resolution Laplacian operator, T represents the bright temperature value, represents the node position, λ represents the calculation range of the high-resolution Laplacian operator, T(u,v) represents the bright temperature value at the node position, and respectively represent the bright temperature values at the corresponding positions;

[0014] Step 4: Perform brightness temperature correction on the mutation nodes, calculate the trend factor ρ based on the brightness temperature difference values of the nearest neighboring nodes, and then adjust and update the brightness temperature values of the mutation nodes; the expression of the trend factor ρ is as follows: (2); or, (3);

[0015] where, represents the relative distance between the node and the mutation node, and k represents the range of trend factor calculation; and respectively represent the brightness temperature values at the corresponding positions;

[0016] Step 5: Output the brightness temperature values of the corrected nearshore data.

[0017] Further, the parameter initialization in Step 1 includes the definition of radiometer system parameters, scene definition, channel and error parameter definition, and inversion parameter definition.

[0018] Further, in Step 1, before performing the oversampling process, determine the number of zero-padding in the parameter definition. The visibility function of the nearshore data of the one-dimensional synthetic aperture microwave radiometer is a one-dimensional matrix with side length N, and after the oversampling process, it becomes a new one-dimensional matrix with side length N L and there is an integer relationship between N and N L : N L = N×β, where β is the oversampling factor.

[0019] Further, in Step 2, since there is a zero value between the positive and negative values of the second derivative, find all the nodes, and set the threshold as S according to the brightness temperature change characteristics in the nearshore area. The range of S is 40 - 60. Screen the nodes according to the threshold S. If the differences on both sides of a node exceed S at the same time, then regard this node as a mutation node, and obtain a one-dimensional matrix of the positions of the mutation nodes with size 1×n, where n is the number of mutation nodes.

[0020] Further, the number of loop iterations in Step 3 is determined according to the correction effect; an interruption condition is introduced during the iterative loop process, and the number of iterations is limited to m, where m is any integer between 20 and 30, and the specific value of m is adjusted according to the correction result.

[0021] Further, in Step 4, if the number of mutation nodes is even, then correct the two mutation nodes to the positions calculated according to the trend factors obtained from the nearest non-mutation nodes. At this time, for the mutation node with the relatively left position, the trend factor ρ is calculated using formula (2), and for the mutation node with the relatively right position, the trend factor ρ is calculated using formula (3).

[0022] Further, in the fourth step, if the number of mutation nodes is odd, first calculate the differences between its brightness temperature values and those of the ordinary nodes on both sides respectively. If the difference between the mutation node and the ordinary node on the left side is smaller, then select formula (2) to calculate the trend factor; if the difference between the mutation node and the ordinary node on the right side is smaller, then select formula (3) to calculate the trend factor.

[0023] Further, in the fourth step, if the number of ordinary nodes on the corresponding side is less than k (k ranges from 1 to 3) when calculating the trend factor, then skip the calculation of the trend factor here and retain the original node brightness temperature value.

[0024] Advantages of the present invention:

[0025] (1) The present invention provides a method for improving the brightness temperature near the shore of a one-dimensional synthetic aperture microwave radiometer based on node sampling. The method of node sampling is used to eliminate Gibbs oscillations and reduce the land-sea buffer zone. In each iteration process, by calculating the high-resolution Laplace operator, the purpose of eliminating Gibbs oscillations is achieved, thereby suppressing Gibbs contamination; on the basis of the completion of the iteration of all non-mutation nodes, the mutation nodes are corrected by calculating the trend factor to achieve the purpose of reducing the land-sea buffer zone.

[0026] (2) The present invention uses the method of node sampling to suppress Gibbs contamination and obtains an oversampled brightness temperature image by zero-padding; the suppression process is an iterative process, and the number of iterations plays a role in the quality of the suppression effect and the speed of the program; λ represents the calculation range of the high-resolution Laplace operator; when correcting the brightness temperature of the mutation nodes, the calculation range k of the trend factor is used as a calculation parameter for the brightness temperature of the mutation nodes;

[0027] When suppressing the Gibbs effect, the present invention abandons the windowing operation in the traditional method and operates on the one-dimensional matrix of the entire oversampled brightness temperature image, making full use of the difference information of the brightness temperature values of adjacent nodes, effectively avoiding the phenomenon that the brightness temperature value of a single node is too high or too low, and can retain the boundary information to a certain extent when correcting the brightness temperature. Description of the drawings

[0028] Figure 1 It is a schematic diagram of the method flow provided by the present invention.

[0029] Figure 2 It is a comparison image of the correction results in Embodiment 1 of the present invention, where the solid line represents the set scene brightness temperature, the dashed line represents the original inversion brightness temperature, and the dotted line represents the corrected brightness temperature; the abscissa ξ represents the direction cosine.

[0030] Figure 3 It is a comparison chart of the differences before and after correction in Embodiment 1 of the present invention, where the solid line represents the brightness temperature difference between before correction and the set scene, and the dashed line represents the brightness temperature difference between after correction and the set scene; the abscissa ξ represents the direction cosine. Detailed implementation manners

[0031] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0032] To further understand the present invention, the present invention will be further described in conjunction with the accompanying drawings and embodiments.

[0033] As Figure 1 shown, the present invention provides a method for improving the nearshore brightness temperature of a one-dimensional synthetic aperture microwave radiometer based on node sampling, including the following steps:

[0034] Step 1: Initialize the nearshore data parameters of the one-dimensional synthetic aperture microwave radiometer, that is, initialize the system inversion parameters, including: defining the radiometer system parameters, scene definition, channel and error parameter definition, and inversion parameter definition. And perform oversampling processing on the frequency domain signal containing Gibbs contamination.

[0035] Accurate parameter definition can obtain good inversion results, so that the brightness temperature image can be more accurately reconstructed in subsequent iterations and corrections.

[0036] The spatial oversampling of the brightness temperature image is achieved by embedding the Fourier coefficients of the basic one-dimensional matrix into a larger one-dimensional matrix and filling new spatial frequency domain zeros. In this process, the original Fourier coefficients are retained, ensuring that the original information is completely retained during the oversampling process. During the oversampling process, no assignment operation is performed on the Fourier coefficients. Define the side length of the one-dimensional matrix of the visibility function of the original nearshore data of the one-dimensional synthetic aperture microwave radiometer as N, and the original Fourier coefficient matrix is embedded in a larger one-dimensional matrix with a side length of N L and there is an integer relationship β between N and N L : N L = N×β. Name β as the oversampling factor. All new Fourier coefficients are set to zero. Apply the inverse Fourier transform to the larger Fourier domain, and finally obtain the oversampled brightness temperature image, which has the same spatial coverage as the original image, but the density is β times.

[0037] Step 2: A node is a position where the second derivative of the data curve is zero and the direction changes. Since there is a zero value between the positive and negative values of the second derivative, all nodes can be found, and a threshold S is set according to the brightness temperature change characteristics in the nearshore area to screen the nodes in the oversampled image to distinguish ordinary nodes from mutation nodes.

[0038] The range of the threshold S is 40 - 60, and this range is determined according to the difference between the sea surface brightness temperature and the land brightness temperature. In the simulation system, since the scene brightness temperature can be freely set, the size of the threshold should be changed according to the setting of the scene brightness temperature, and the setting of the threshold is relatively flexible. Screen the nodes. If the differences on both adjacent sides of a node exceed S at the same time, this node is regarded as a mutation node.

[0039] Since there will be cases where mutation nodes are continuous in complex high and low brightness temperature data (offshore data), after screening out the continuous mutation nodes, if the number of mutation nodes is odd, the middle mutation node is regarded as the final mutation node; if the number of continuous mutation nodes is even, the two middle mutation nodes are retained.

[0040] Step 3: Iteratively correct the ordinary nodes. Except for the mutation nodes and the nodes one before and after them, for other nodes, calculate the high-resolution Laplacian operator on the oversampled image. The calculation method is the difference between the average of the brightness temperature values of the first adjacent points of each node on the image and this point. If there are no adjacent non-mutation nodes at the node position, the original value is retained. The expression is as follows:

[0041] Among them, represents the high-resolution Laplacian operator, T represents the brightness temperature value, λ represents the calculation range of the high-resolution Laplacian operator, represents the node position. Among them, u represents the position number of this node in the node sequence. Since the one-dimensional synthetic aperture imaging is a one-dimensional line, so is set to 0 here. The high-resolution Laplacian operator calculated each time of iteration needs to be added to the brightness temperature of the ordinary node to obtain the updated brightness temperature. The brightness temperature is updated once for each iteration. Finally, the iteratively corrected ordinary nodes and the uncorrected mutation nodes are represented as T1.

[0042] The number of loop iterations should be determined according to the correction effect. When the number of iterations is relatively early, it can be visually detected that large fluctuations are being eliminated, and then it gradually stagnates. At this time, from the i-th iteration to the i + 1-th iteration, there may be a fluctuation phenomenon: the value of the same point is updated to a new value, and then after another iteration, it returns to the previous value. The value of this point changes between these two values and repeats in this cycle. Therefore, an interruption condition should be introduced here to avoid invalid iterations, and the number of iterations is limited to m, where m is any integer between 20 and 30 and can be adjusted according to the correction effect.

[0043] For simplicity, introduce a very simple interruption condition: the algorithm stops iterating when it reaches 25 iterations. Because from the test results, no obvious image improvement was observed after 25 iterations.

[0044] Step 4: When calibrating the bright temperature of the mutation node, the bright temperatures of the ordinary nodes and the mutation node involved in the calculation are both obtained from T1. If the number of mutation nodes is even, the two mutation nodes are respectively calibrated to the positions calculated according to the trend factors obtained from the nearest ordinary nodes; at this time, the calculation formula for the trend factor ρ corresponding to the mutation node with a relatively left position is ; for the calculation of the trend factor ρ corresponding to the mutation node with a relatively right position, the formula ;

[0045] If the number of mutation nodes is odd, first calculate the differences between its bright temperature values and those of the ordinary nodes on both sides, and then the trend factor is calculated based on the side with the smaller difference.

[0046] If the difference between the mutation node and the left ordinary node is smaller, the trend factor is calculated based on the left node of its relative position. At this time, the calculation formula for the trend factor ρ is: ;

[0047] If the difference between the mutation node and the right ordinary node is smaller, the trend factor is calculated based on the right node of its relative position. At this time, the calculation formula for the trend factor ρ is: ; where ρ is the bright temperature value of the trend factor, represents the relative distance between the node and the mutation node, and k represents the calculation range of the trend factor; and respectively represent the bright temperature values at the corresponding positions.

[0048] If the number of ordinary nodes on the corresponding side is less than k during the calculation of the trend factor, skip the calculation of the trend factor here and retain the original node bright temperature value. The value range of k is between 1 and 3.

[0049] Generally, k is taken as 2. The reason is that the values of the ordinary nodes have been iteratively calibrated by calculating the high-resolution Laplace operator in the above steps, and the differences between them and the adjacent ordinary nodes are not very large. Therefore, taking k as 2 is sufficient to meet the requirements for calculating the trend factor ρ.

[0050] Step 5: Output the calibrated bright temperature value of the nearshore data and end. Example 1

[0051] This example illustrates the method for improving the nearshore bright temperature of a one-dimensional synthetic aperture microwave radiometer based on node sampling through a one-dimensional synthetic aperture microwave radiometer simulation system.

[0052] The observation scenario of a one-dimensional synthetic aperture microwave radiometer is a one-dimensional line. During the simulation process, various parameter inputs will affect the calibration effect of the nodes and the final brightness temperature image result. If too few parameters are defined, it will have a negative impact on the screening and calibration of the nodes; if too many parameters are defined, it may cause too much redundancy in the calibration process, reduce the operation efficiency, and increase the algorithm calibration time.

[0053] Definition of radiometer system parameters: The antenna of the one-dimensional synthetic aperture microwave radiometer is set to an antenna array with 10 feed units [0 2 4 5 6 7 16 19 21 24]. The minimum antenna spacing between each feed unit is 0.618du of the L-band wavelength. During the simulation process, the center frequency of the RF signal is 1.413 GHz, the bandwidth is set to 20 MHz, and the receiver noise is 140K.

[0054] Scene definition: During the simulation process, in order to simulate the land-sea boundary, the simulation scene is set to a shape that is high in the middle and low on both sides. The extended source is a set of points with evenly distributed pixels. The brightness temperature of the two side scenes is set to 0k, and the brightness temperature of the middle scene is set to 200k.

[0055] Definition of channel and error parameters: This part is an ideal simulation of various hardware, and all errors are set to 0.

[0056] Definition of reaction parameters: The inversion method uses ideal Fourier inversion, the number of zero-padding is determined to be 370, and windowless processing is selected.

[0057] After defining all kinds of parameters, the simulated one-dimensional synthetic aperture L1A-level data is obtained through the simulation system and oversampled. The oversampling factor is β, which is set to 37 here. Then, the inverse Fourier transform is performed to obtain the one-dimensional oversampled brightness temperature image T.

[0058] According to Figure 1 , by the method that there is a zero value between the second-order positive and negative derivatives, all nodes in the brightness temperature image T are found. When the difference between the adjacent sides of the node exceeds the threshold S at the same time, where S is set to 50 here, the node is marked as a mutation node. Note that the threshold here cannot be set too large, otherwise the mutation nodes cannot be screened out, and at the same time, it cannot be too small, otherwise the screened mutation nodes cannot play the role of marking the land-sea boundary. After marking all mutation nodes, calculate the high-resolution Laplacian operator of the ordinary nodes , after calculation, add the Laplacian operator to the node brightness temperature T, and the brightness temperature value of the mutation node remains unchanged, and finally form the calibrated brightness temperature T1. The calculation formula is: where T is the oversampled brightness temperature value, (u, v) represents the node position, λ represents the calculation range of the high-resolution Laplacian operator, which is set to 1 here. Since the one-dimensional synthetic aperture imaging is a one-dimensional line, so herev Set to 0, the number of iterations m here is set to 25 times, and T1 represents the brightness temperature value obtained after adding the node and the Laplacian operator.

[0059] Based on T1, continue to correct the mutation nodes. If the number of mutation nodes is even, the two mutation nodes are respectively corrected to the positions calculated according to the trend factor obtained from the nearest normal node; if the number of mutation nodes is odd, first calculate the difference between its brightness temperature value and the brightness temperature values of the nodes on both sides, and then correct it to the side with the smaller difference, and its value will be calculated according to the trend factor. The trend factor The expression is as follows: is the brightness temperature value of the trend factor, (u, v) represents the node position. Since the one-dimensional synthetic aperture imaging is a one-dimensional line, so here is set to 0, j represents the relative distance between the node and the mutation node, k represents the trend factor calculation range, and the trend factor calculation range k is set to 2. Update the brightness temperature value of the mutation node and output the brightness temperature.

[0060] As Figure 2 shown, the Gibbs contamination is significantly suppressed; at the brightness temperature mutation, the land-sea boundary is more obvious and the blurred area is reduced. As Figure 3 shown, at the place where the brightness temperature of the set scene is gentle, the Gibbs error is significantly corrected for the corrected brightness temperature.

[0061] The above description is only the preferred embodiment of the present invention and is not a limitation of the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modification, equivalent replacement, modification, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for improving nearshore brightness temperature using a one-dimensional synthetic aperture microwave radiometer based on node sampling, characterized in that: The steps include: Step 1: Initialize the nearshore data parameters of the one-dimensional synthetic aperture microwave radiometer, oversample the frequency domain signal containing Gibbs contamination, set the oversampling factor, and obtain a one-dimensional oversampled brightness temperature image; Step 2: Find the nodes in the oversampled brightness temperature image through the positive and negative values ​​of the second-order derivative, set the threshold S, and filter the common nodes and mutation nodes; when the difference between the two adjacent sides of the node is greater than the threshold S at the same time, the node is regarded as a mutation node, and the other nodes are regarded as common nodes; Step 3: perform iterative correction on the common nodes, including calculating the high-resolution Laplace operator, updating the brightness temperature value of the common nodes, and iterating in a loop; the high-resolution Laplace operator calculated in each iteration needs to be added to the brightness temperature of the common node to obtain the updated brightness temperature, and the brightness temperature needs to be updated once each iteration; the calculation expression of the high-resolution Laplace operator is: Among them, Δ H T(u,v) represents the high-resolution Laplace operator, T represents the brightness temperature, (u,v) represents the node position, λ represents the calculation range of the high-resolution Laplace operator, T(u,v) represents the brightness temperature value of the node position, T(u-λ,v) and T(u+λ,v) represent the brightness temperature values ​​of the corresponding positions respectively; Step 4: perform brightness temperature correction on the mutation node, calculate the trend factor ρ according to the brightness temperature difference of the nearest adjacent node, and then adjust and update the brightness temperature value of the mutation node; the expression of the trend factor ρ is as follows: Among them, j represents the relative distance between the node and the mutation node, k represents the calculation range of the trend factor, T(uj,v), T(u-(j+1),v), T(u-1,v), T(u+j,v), T(u+(j+1),v) and T(u+1,v) represent the brightness temperature values ​​of the corresponding positions respectively; If the number of mutation nodes is even, the two mutation nodes are corrected to the positions calculated based on the trend factor obtained from the nearest non-mutation node. At this time, the trend factor ρ corresponding to the mutation node on the left is calculated using formula (2), and the trend factor ρ corresponding to the mutation node on the right is calculated using formula (3); If the number of mutation nodes is an odd number, first calculate the difference between the brightness temperature of the mutation node and the ordinary nodes on both sides. If the difference between the mutation node and the ordinary node on the left is smaller than the difference between the mutation node and the ordinary node on the right, then use formula (2) to calculate the trend factor; if the difference between the mutation node and the ordinary node on the right is smaller than the difference between the mutation node and the ordinary node on the left, then use formula (3) to calculate the trend factor. Step 5: Output the corrected brightness temperature value of the nearshore data.

2. The method for improving nearshore brightness temperature of a one-dimensional synthetic aperture microwave radiometer based on node sampling according to claim 1 is characterized in that: The parameter initialization of step one includes radiometer system parameter definition, scene definition, channel and error parameter definition, and inversion parameter definition.

3. The method for improving nearshore brightness temperature of a one-dimensional synthetic aperture microwave radiometer based on node sampling according to claim 1 is characterized in that: In the step 1, before the oversampling process is performed, the number of zero padding is determined in the parameter definition. The visibility function of the nearshore data of the one-dimensional synthetic aperture microwave radiometer is a one-dimensional matrix with a side length of N. After the sampling process, it becomes a one-dimensional matrix with a side length of N. L The new one-dimensional matrix, N and N L There is an integer relationship between them: N L =N×β, β is the oversampling factor.

4. The method for improving nearshore brightness temperature of a one-dimensional synthetic aperture microwave radiometer based on node sampling according to claim 1 is characterized in that: In the step 2, all nodes are found according to the existence of zero value between positive and negative values ​​of the second-order derivative, and a threshold value S is set according to the brightness temperature variation characteristics of the nearshore area, and the range of S is 40-60. The nodes are screened according to the threshold value S. If the difference between the two adjacent sides of the node exceeds S at the same time, the node is regarded as a mutation node, and a 1×n one-dimensional matrix of mutation node positions is obtained, where n is the number of mutation nodes.

5. The method for improving nearshore brightness temperature of a one-dimensional synthetic aperture microwave radiometer based on node sampling according to claim 1 is characterized in that: The number of loop iterations in step three is determined according to the correction effect; an interrupt condition is introduced during the iterative loop process to limit the number of iterations to m, where m is any integer between 20 and 30, and the specific value of m is adjusted according to the correction result.

6. The method for improving nearshore brightness temperature of a one-dimensional synthetic aperture microwave radiometer based on node sampling according to claim 1 is characterized in that: In the step 4, if the number of ordinary nodes on the corresponding side is less than k when the trend factor is calculated, and k is 1-3, the trend factor calculation here is skipped and the original node brightness temperature value is retained.

Images