A time series SAR data filtering method suitable for broken areas of farmland

Abstract

The application discloses a time sequence SAR data filtering method suitable for broken areas of farmland, and the method comprises the following steps: acquiring time sequence SAR data to be processed, wherein the time sequence SAR data to be processed comprises a plurality of initial SAR images, the order of the plurality of SAR images is consistent with the imaging time sequence, and the plurality of initial SAR images are subjected to first time smoothing processing to obtain a plurality of preliminary filtering images; acquiring similarity data of each pixel in the preliminary filtering images, wherein the similarity data of the pixel comprises the similarity score of other pixels in the N*N neighborhood centered on the pixel, N is an odd positive number, the pixel data in the preliminary filtering images are updated according to the similarity data to obtain a spatial filtering image; and the spatial filtering image is subjected to second time smoothing processing to obtain a final filtering result. The application can effectively suppress noise and effectively preserve the edge and texture characteristics of the farmland area, thereby improving the accuracy of crop monitoring.

Description

Technical Field

[0001] This invention relates to the field of remote sensing data processing technology, and in particular to a time-series SAR data filtering method suitable for broken farmland areas. Background Technology

[0002] Accurate and timely crop monitoring not only ensures grain yield forecasting but also provides information support for the development of modern sustainable agriculture. Currently, remote sensing technology, with its advantage of rapidly and accurately acquiring information over large areas, is widely used in various fields of crop monitoring, such as crop classification, phenological extraction, and yield estimation. With the continuous development of microwave remote sensing technology, synthetic aperture radar (SAR) has been widely applied in agricultural monitoring as a new data source. Compared to optical data, SAR has the advantages of being available all day and all weather, and is unaffected by weather conditions, making it particularly suitable for cloudy and rainy areas. However, when monitoring crops in fragmented farmland areas based on time-series SAR data, the unique imaging principle of SAR results in a certain degree of speckle noise in the SAR images. Furthermore, the unique plant structure of crops increases the randomness of SAR echoes, further exacerbating the noise in the SAR images and affecting the accuracy of crop observation. SAR filtering is a crucial step in eliminating the effects of noise. Current research mainly employs spatial-based filtering methods, such as Lee filtering and Frost filtering. However, for SAR images characterized by fragmented farmland patches and diverse crop types, spatial-based filtering methods struggle to preserve the edge and texture features of farmland areas, thus reducing the accuracy of crop monitoring results.

[0003] Therefore, existing technologies still need to be improved and enhanced. Summary of the Invention

[0004] To address the aforementioned shortcomings of existing technologies, this invention provides a time-series SAR data filtering method suitable for broken farmland areas, aiming to solve the problem of inaccurate crop monitoring results obtained using SAR filtering in existing technologies.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] A first aspect of the present invention provides a time-series SAR data filtering method suitable for farmland fractured areas, the method comprising:

[0007] Acquire time-series SAR data to be processed, which includes multiple initial SAR images. The order of the multiple initial SAR images is consistent with the imaging time sorting. Perform first-time smoothing processing on the multiple initial SAR images to obtain multiple preliminary filtered images.

[0008] The similarity data of each pixel in the preliminary filtered image is obtained. The similarity data of a pixel includes the similarity scores of other pixels in the N*N neighborhood centered on the pixel and the pixel itself, where N is a positive odd number. The pixel data in the preliminary filtered image is updated according to the similarity data to obtain the spatial filtered image.

[0009] The spatially filtered image is then subjected to a second temporal smoothing process to obtain the final filtering result.

[0010] The time-series SAR data filtering method applicable to farmland fractured areas, wherein, before performing a first time smoothing process on the multiple initial SAR images to obtain multiple preliminary filtered images, includes:

[0011] The initial SAR image is preprocessed;

[0012] The preprocessing includes at least one of thermal noise removal, radiometric calibration, and terrain correction.

[0013] The time-series SAR data filtering method applicable to farmland fractured areas, wherein the first time smoothing process of the multiple initial SAR images to obtain multiple preliminary filtered images includes:

[0014] SG filtering is performed on each target's initial SAR image as the center phase to obtain intermediate filtering results. The filtering window is K. The target's initial SAR image is the first of the multiple initial SAR images excluding the first... Zhang Image and After Images other than Zhang's image, ;

[0015] The multiple preliminary filtered images are obtained based on the intermediate filtering results.

[0016] The time-series SAR data filtering method applicable to broken farmland areas, wherein obtaining the multiple preliminary filtered images based on the intermediate filtering results includes:

[0017] The preliminary filtered image is obtained according to the first preset formula;

[0018] The first preset formula is: ;

[0019] in, Indicates the first The preliminary filtered image corresponding to the initial SAR image. Indicates the first The intermediate filtering result corresponding to the i-th initial SAR image obtained by performing SG filtering on the i-th initial SAR image as the center phase.

[0020] The time-series SAR data filtering method applicable to broken farmland areas, wherein obtaining the similarity data of each pixel in the preliminary filtered image includes:

[0021] The similarity data is obtained according to the second preset formula;

[0022] The second preset formula is: ;

[0023] in, For pixels and pixels similarity score, Represents a cell SAR time series vector, Represents a cell SAR time series vector, Represents a cell The SAR time series vector is the transpose of the SAR time series vector, where the SAR time series vector of a pixel is a vector composed of pixel data at that pixel location in each of the preliminary filtered images. This represents the L2 norm.

[0024] The time-series SAR data filtering method applicable to farmland fractured areas, wherein updating the image data in the preliminary filtered image based on the similarity data to obtain a spatially filtered image includes:

[0025] The image data in the preliminary filtered image is updated according to the third preset formula to obtain the spatial filtered image;

[0026] The third preset formula is: ;

[0027] in, This indicates that the position in the j-th spatially filtered image is... Metadata, Indicates the first The position in the preliminary filtered image corresponding to the initial SAR image is The pixel data of the i-th target pixel in the neighborhood of the given pixel, where the target pixel is the pixel whose similarity score is greater than a preset value. This indicates that the i-th target pixel is located at position 1. The similarity score of the pixel, M is the number of pixels. The position in the preliminary filtered image corresponding to the initial SAR image is The number of target pixels in the neighborhood of the target pixel.

[0028] The time-series SAR data filtering method applicable to farmland fractured areas, wherein the second time smoothing process is the same as the first time smoothing process.

[0029] A second aspect of the present invention provides a time-series SAR data filtering device suitable for farmland fractured areas, comprising:

[0030] The first time-smoothing module is used to acquire time-series SAR data to be processed. The time-series SAR data includes multiple initial SAR images. The order of the multiple initial SAR images is consistent with the imaging time sorting. The first time-smoothing process is performed on the multiple initial SAR images to obtain multiple preliminary filtered images.

[0031] The spatial filtering module is used to acquire the similarity data of each pixel in the preliminary filtered image. The similarity data of the pixel includes the similarity scores of other pixels in the N*N neighborhood centered on the pixel and the pixel itself, where N is a positive odd number. The pixel data in the preliminary filtered image is updated according to the similarity data to obtain the spatial filtered image.

[0032] The second temporal smoothing module is used to perform a second temporal smoothing process on the spatially filtered image to obtain the final filtering result.

[0033] A third aspect of the present invention provides a terminal, the terminal including a processor and a computer-readable storage medium communicatively connected to the processor, the computer-readable storage medium being adapted to store a plurality of instructions, the processor being adapted to invoke the instructions in the computer-readable storage medium to perform the steps of implementing the time-series SAR data filtering method for farmland fractured areas as described in any of the preceding claims.

[0034] In a fourth aspect, the present invention provides a computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the steps of the time-series SAR data filtering method for farmland fractured areas as described in any of the preceding claims.

[0035] Compared with existing technologies, this invention provides a time-series SAR data filtering method suitable for broken farmland areas. The method provided by this invention comprehensively considers time and space filtering for SAR data, and measures the spatial similarity of the time-series data after preliminary time filtering. The filtered neighbors in the spatial filtering have temporal consistency, avoiding the imbalance in the time dimension in the spatial filtering. The final filtering result effectively suppresses noise while effectively preserving the edge and texture features of the farmland area, which can improve the accuracy of crop monitoring. Attached Figure Description

[0036] Figure 1 A flowchart illustrating an embodiment of the time-series SAR data filtering method for farmland fractured areas provided by the present invention;

[0037] Figure 2 A schematic diagram illustrating the filtering effect of an embodiment of the time-series SAR data filtering method for farmland fractured areas provided by the present invention. Figure 1 ;

[0038] Figure 3 A schematic diagram illustrating the filtering effect of an embodiment of the time-series SAR data filtering method for farmland fractured areas provided by the present invention. Figure 2 ;

[0039] Figure 4 A schematic diagram of the structural principle of an embodiment of the time-series SAR data filtering device for farmland fractured areas provided by the present invention;

[0040] Figure 5 A schematic diagram illustrating the principle of an embodiment of the terminal provided by the present invention. Detailed Implementation

[0041] To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0042] The time-series SAR data filtering method for broken farmland areas provided by this invention can be applied to terminals with computing capabilities, including but not limited to various computers, servers, mobile devices, etc.

[0043] Example 1

[0044] like Figure 1 As shown, one embodiment of the time-series SAR data filtering method applicable to farmland fractured areas includes the following steps:

[0045] S100. Obtain the time series SAR data to be processed, which includes multiple initial SAR images. The order of the multiple initial SAR images is consistent with the imaging time sorting. Perform a first time smoothing process on the multiple initial SAR images to obtain multiple preliminary filtered images.

[0046] The time-series SAR data to be processed includes multiple initial SAR images. The multiple initial SAR images are sorted according to the time of the remote sensing data used to generate the images to obtain the time-series SAR data to be processed.

[0047] After generating the initial SAR image based on remote sensing data, each image can be preprocessed separately, and subsequent filtering can be performed based on the preprocessed initial SAR image. That is, before performing a first-time smoothing process on the multiple initial SAR images to obtain multiple preliminary filtered images, the process includes:

[0048] The initial SAR image is preprocessed;

[0049] The preprocessing includes at least one of thermal noise removal, radiometric calibration, and terrain correction.

[0050] In the method provided in this embodiment, a first time filtering is performed to achieve time smoothing, with the aim of obtaining the overall long-term variation trend of each pixel. The first time smoothing process on the multiple initial SAR images to obtain multiple preliminary filtered images includes:

[0051] SG filtering is performed on each target's initial SAR image as the center phase to obtain intermediate filtering results. The filtering window is K. The target's initial SAR image is the first of the multiple initial SAR images excluding the first... Zhang Image and After Images other than Zhang's image, ;

[0052] The multiple preliminary filtered images are obtained based on the intermediate filtering results.

[0053] In this embodiment, Savitzky-Golay (SG) filtering is used to implement the first time smoothing process. In SG filtering, a center phase within the filtering window needs to be specified. Unlike traditional SG filtering, this method does not only output the filtering result of the center phase, but also outputs the filtering results of five consecutive time phases. Finally, the output results of all time phases are averaged to obtain the preliminary filtering result. The filtering window K is a positive odd number; for example, K can be 5. The polynomial order in SG filtering can be set to 2.

[0054] The step of obtaining the multiple preliminary filtered images based on the intermediate filtering results includes:

[0055] The preliminary filtered image is obtained according to the first preset formula;

[0056] The first preset formula is: ;

[0057] in, Indicates the first The preliminary filtered image corresponding to the initial SAR image. Indicates the first The intermediate filtering result corresponding to the i-th initial SAR image obtained by performing SG filtering on the i-th initial SAR image as the center phase. It is easy to see that, for the case of a filtering window of 5, only filtering of 5 phases such as j-2, j-1, j, j+1, and j+2 will produce the filtering result of the j-th phase.

[0058] After performing the first time smoothing process on the SAR data, the method provided in this embodiment then performs spatial filtering, specifically including the following steps:

[0059] S200. Obtain the similarity data of each pixel in the preliminary filtered image. The similarity data of a pixel includes the similarity scores of other pixels in the N*N neighborhood centered on the pixel and the pixel itself, where N is a positive odd number. Update the pixel data in the preliminary filtered image according to the similarity data to obtain the spatial filtered image.

[0060] The step of obtaining the similarity data of each pixel in the preliminary filtered image includes:

[0061] The similarity data is obtained according to the second preset formula;

[0062] The second preset formula is: ;

[0063] Among them, pixels For pixels A pixel within the N*N neighborhood of the center. For pixels and pixels similarity score, Represents a cell SAR time series vector, Represents a cell SAR time series vector, Represents a cell The SAR time series vector is the transpose of the SAR time series vector, where the SAR time series vector of a pixel is a vector composed of pixel data at that pixel location in each of the preliminary filtered images. This represents the L2 norm.

[0064] For the preliminary filtered image, spatial averaging is performed based on the time series similarity within an N*N neighborhood. To further improve filtering accuracy, only pixels with high similarity participate in the averaging process in this step; that is, only pixel data with a similarity score greater than a preset threshold participates in the averaging. The preset threshold can be 0.5. Updating the pixel data in the preliminary filtered image based on the similarity data to obtain the spatially filtered image includes:

[0065] The image data in the preliminary filtered image is updated according to the third preset formula to obtain the spatial filtered image;

[0066] The third preset formula is: ;

[0067] in, This indicates that the position in the j-th spatially filtered image is... Metadata, Indicates the first The position in the preliminary filtered image corresponding to the initial SAR image is The pixel data of the i-th target pixel in the neighborhood of the given pixel, where the target pixel is the pixel whose similarity score is greater than a preset value. This indicates that the i-th target pixel is located at position 1. The similarity score of the pixel, M is the number of pixels. The position in the preliminary filtered image corresponding to the initial SAR image is The number of target pixels in the neighborhood of the target pixel.

[0068] In practical applications, N can be 7; however, those skilled in the art will understand that N can also take other values. In step S200, the neighborhood similarity of the preliminary time-filtered result is measured by time series similarity, and pixels with similarity scores greater than a preset threshold are used as the neighborhood for spatial filtering of the entire time series. In existing SAR filtering methods, some adaptive Lee filtering is used to determine the neighborhood range to maintain edge features. However, the neighborhood range for each time phase is obtained separately. For the fragmented agricultural areas of South China, the echo intensity of different adjacent crop plots varies at different time phases, which can easily lead to an imbalance in the filtering neighborhood between different time phases, affecting the final filtering effect. The method provided in this embodiment measures spatial similarity based on the time series curve after preliminary time filtering. Compared with the existing time-by-time Lee filtering or Frost filtering to find the neighborhood space, the spatial filtering neighborhood of the method provided in this embodiment has temporal series consistency, avoiding imbalance in the time dimension.

[0069] Please refer to it again. Figure 1 The method provided in this embodiment further includes the following steps:

[0070] S300. Perform a second time smoothing process on the spatially filtered image to obtain the final filtering result.

[0071] The second time smoothing process can be the same as the first time smoothing process. That is, for the spatially filtered image, the same improved SG filtering method as in step S100 is used to filter and obtain the final filtering result.

[0072] In the method provided in this embodiment, the second time filtering can smooth the spatially averaged result again in time, so that the SAR data can more accurately reflect the growth curve of crops.

[0073] To verify the effectiveness of the method provided in this embodiment, the inventors conducted experimental verification. Taking sugarcane fields in South China as an example, they used Sentinel-1A time series data, with a time period of 12 days per scene, for verification. Figure 2 and Figure 3 The figures show the filtering effects of VH polarization and VV polarization data, respectively. "smoothed" indicates the filtering effect of the method provided in this embodiment, and "Original" indicates the filtering effect of existing multi-channel SAR filtering methods. Experimental results show that the method provided by this invention has a better filtering effect. While eliminating the influence of noise, the filtering results effectively preserve the edge and texture features of farmland areas, and provide a more accurate description of crop growth curves.

[0074] In summary, this embodiment provides a time-series SAR data filtering method suitable for broken farmland areas. For SAR data, time and spatial filtering are comprehensively considered, and spatial similarity is measured in the time-series data after preliminary time filtering. The filtered neighbors in the spatial filtering have temporal consistency, avoiding the imbalance in the time dimension in the spatial filtering. The final filtering result effectively suppresses noise while effectively preserving the edge and texture features of the farmland area, which can improve the accuracy of crop monitoring.

[0075] It should be understood that although the steps in the flowcharts shown in the accompanying drawings are displayed sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0076] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0077] Example 2

[0078] Based on the above embodiments, the present invention also provides a time-series SAR data filtering device suitable for farmland fractured areas, such as... Figure 4 As shown, the time-series SAR data filtering device suitable for farmland fractured areas includes:

[0079] The first time-smoothing module is used to acquire time-series SAR data to be processed. The time-series SAR data includes multiple initial SAR images. The order of the multiple initial SAR images is consistent with the imaging time sorting. The multiple initial SAR images are subjected to first time-smoothing processing to obtain multiple preliminary filtered images, as specifically described in Embodiment 1.

[0080] The spatial filtering module is used to acquire the similarity data of each pixel in the preliminary filtered image. The similarity data of the pixel includes the similarity scores of other pixels in the N*N neighborhood centered on the pixel and the pixel itself, where N is a positive odd number. The pixel data in the preliminary filtered image is updated according to the similarity data to obtain the spatial filtered image, as specifically described in Embodiment 1.

[0081] The second temporal smoothing module is used to perform a second temporal smoothing process on the spatially filtered image to obtain the final filtering result, as described in Embodiment 1.

[0082] Example 3

[0083] Based on the above embodiments, the present invention also provides a terminal, such as... Figure 5 As shown, the terminal includes a processor 10 and a memory 20. Figure 5 Only some of the terminal components are shown; however, it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0084] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory. In other embodiments, the memory 20 may be an external storage device of the terminal, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc. Further, the memory 20 may include both internal and external storage devices. The memory 20 is used to store application software and various types of data installed on the terminal. The memory 20 can also be used to temporarily store data that has been output or will be output. In one embodiment, the memory 20 stores a time-series SAR data filtering program 30 suitable for broken farmland areas. This time-series SAR data filtering program 30 can be executed by the processor 10 to implement the time-series SAR data filtering method for broken farmland areas described in this application.

[0085] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor, or other chip, used to run program code stored in the memory 20 or process data, such as executing the time-series SAR data filtering method applicable to farmland fractured areas.

[0086] In one embodiment, when processor 10 executes time-series SAR data filtering program 30 in memory 20 suitable for farmland fractured areas, the following steps are performed:

[0087] Acquire time-series SAR data to be processed, which includes multiple initial SAR images. The order of the multiple initial SAR images is consistent with the imaging time sorting. Perform first-time smoothing processing on the multiple initial SAR images to obtain multiple preliminary filtered images.

[0088] The similarity data of each pixel in the preliminary filtered image is obtained. The similarity data of a pixel includes the similarity scores of other pixels in the N*N neighborhood centered on the pixel and the pixel itself, where N is a positive odd number. The pixel data in the preliminary filtered image is updated according to the similarity data to obtain the spatial filtered image.

[0089] The spatially filtered image is then subjected to a second temporal smoothing process to obtain the final filtering result.

[0090] Before performing a first-time smoothing process on the multiple initial SAR images to obtain multiple preliminary filtered images, the process includes:

[0091] The initial SAR image is preprocessed;

[0092] The preprocessing includes at least one of thermal noise removal, radiometric calibration, and terrain correction.

[0093] The step of performing a first-time smoothing process on the multiple initial SAR images to obtain multiple preliminary filtered images includes:

[0094] SG filtering is performed on each target's initial SAR image as the center phase to obtain intermediate filtering results. The filtering window is K. The target's initial SAR image is the first of the multiple initial SAR images excluding the first... Zhang Image and After Images other than Zhang's image, ;

[0095] The multiple preliminary filtered images are obtained based on the intermediate filtering results.

[0096] The step of obtaining the multiple preliminary filtered images based on the intermediate filtering results includes:

[0097] The preliminary filtered image is obtained according to the first preset formula;

[0098] The first preset formula is: ;

[0099] in, Indicates the first The preliminary filtered image corresponding to the initial SAR image. Indicates the first The intermediate filtering result corresponding to the i-th initial SAR image obtained by performing SG filtering on the i-th initial SAR image as the center phase.

[0100] The step of obtaining the similarity data of each pixel in the preliminary filtered image includes:

[0101] The similarity data is obtained according to the second preset formula;

[0102] The second preset formula is: ;

[0103] in, For pixels and pixels similarity score, Represents a cell SAR time series vector, Represents a cell SAR time series vector, Represents a cell The SAR time series vector is the transpose of the SAR time series vector, where the SAR time series vector of a pixel is a vector composed of pixel data at that pixel location in each of the preliminary filtered images. This represents the L2 norm.

[0104] The step of updating the pixel data in the preliminary filtered image based on the similarity data to obtain the spatially filtered image includes:

[0105] The image data in the preliminary filtered image is updated according to the third preset formula to obtain the spatial filtered image;

[0106] The third preset formula is: ;

[0107] in, This indicates that the position in the j-th spatially filtered image is... Metadata, Indicates the first The position in the preliminary filtered image corresponding to the initial SAR image is The pixel data of the i-th target pixel in the neighborhood of the given pixel, where the target pixel is the pixel whose similarity score is greater than a preset value. This indicates that the i-th target pixel is located at position 1. The similarity score of the pixel, M is the number of pixels. The position in the preliminary filtered image corresponding to the initial SAR image is The number of target pixels in the neighborhood of the target pixel.

[0108] The second time smoothing process is the same as the first time smoothing process.

[0109] Example 4

[0110] The present invention also provides a computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the steps of the time-series SAR data filtering method applicable to farmland fractured areas as described above.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A time series SAR data filtering method suitable for broken areas of farmland, characterized in that, The method comprises: acquiring time series SAR data to be processed, the time series SAR data comprising a plurality of initial SAR images, the order of the plurality of initial SAR images being consistent with the imaging time sequence, performing first time smoothing processing on the plurality of initial SAR images to obtain a plurality of preliminary filtered images; acquiring similarity data of each pixel in the preliminary filtered image, the similarity data of the pixel comprising a similarity score of other pixels in an N*N neighborhood centered on the pixel, N being an odd positive number, updating the pixel data in the preliminary filtered image according to the similarity data to obtain a spatial filtered image; performing second time smoothing processing on the spatial filtered image to obtain a final filtering result; the updating of the pixel data in the preliminary filtered image according to the similarity data to obtain the spatial filtered image comprises: updating the pixel data in the preliminary filtered image according to a third preset formula to obtain the spatial filtered image; The third preset formula is: ; in, This indicates that the position in the j-th spatially filtered image is... Metadata, Indicates the first The position in the preliminary filtered image corresponding to the initial SAR image is The pixel data of the i-th target pixel in the neighborhood of the given pixel, where the target pixel is the pixel whose similarity score is greater than a preset value. This indicates that the i-th target pixel is located at position 1. The similarity score of the pixel, M is the number of pixels. The position in the preliminary filtered image corresponding to the initial SAR image is The number of target pixels in the neighborhood of the target pixel.

2. The time series SAR data filtering method suitable for broken area of farmland according to claim 1, characterized in that, before the performing of the first time smoothing processing on the plurality of initial SAR images to obtain the plurality of preliminary filtered images, the method comprises: performing preprocessing on the initial SAR image; the preprocessing comprises at least one of thermal noise removal, radiation scaling and terrain correction. 3.The time series SAR data filtering method for broken areas in farmland according to claim 1, wherein, the performing of the first time smoothing processing on the plurality of initial SAR images to obtain the plurality of preliminary filtered images comprises: S-G filtering is performed on each target initial SAR image as the center to obtain an intermediate filtering result, and the filtering window is K, the target initial SAR image being an image other than the first and last images in the plurality of initial SAR images, ;​​ acquiring the plurality of preliminary filtered images according to the intermediate filtering result.

4. The method for filtering time series SAR data suitable for broken areas of farmland according to claim 3, characterized in that, the acquiring of the plurality of preliminary filtered images according to the intermediate filtering result comprises: acquiring the preliminary filtered image according to a first preset formula; The first preset formula is: ; in, Indicates the first The preliminary filtered image corresponding to the initial SAR image. Indicates the first The intermediate filtering result corresponding to the i-th initial SAR image obtained by performing SG filtering on the i-th initial SAR image as the center phase.

5. The method for filtering time series SAR data suitable for broken areas in farmland according to claim 1, characterized in that, the acquiring of the similarity data of each pixel in the preliminary filtered image comprises: acquiring the similarity data according to a second preset formula; The second preset formula is: ; wherein a pixel and a pixel a similarity score of a pixel denotes a SAR time series vector of a pixel denotes a SAR time series vector of a pixel denotes a SAR time series vector of a pixel denotes a SAR time series vector of a pixel denotes the transpose of a SAR time series vector of a pixel denotes the transpose of a SAR time series vector of a pixel denotes a two-norm.

6. The method for filtering time series SAR data suitable for broken areas in farmland according to claim 1, characterized in that, the second time smoothing processing process is consistent with the first time smoothing processing process.

7. A time series SAR data filtering device suitable for use in broken areas of farmland, characterized by, The time series SAR data filtering device suitable for farmland fragmented areas is used to implement the time series SAR data filtering method suitable for farmland fragmented areas according to any one of claims 1-6, and comprises: a first time smoothing module, configured to acquire time series SAR data to be processed, the time series SAR data comprising a plurality of initial SAR images, the order of the plurality of initial SAR images being consistent with the imaging time sequence, perform first time smoothing processing on the plurality of initial SAR images to obtain a plurality of preliminary filtered images; a spatial filtering module, configured to acquire similarity data of each pixel in the preliminary filtered image, the similarity data of the pixel comprising a similarity score of other pixels in an N*N neighborhood centered on the pixel, N being an odd positive number, update the pixel data in the preliminary filtered image according to the similarity data to obtain a spatial filtered image; a second time smoothing module, configured to perform second time smoothing processing on the spatial filtered image to obtain a final filtering result.

8. A terminal, characterized by comprising: The terminal comprises a processor, a computer readable storage medium connected with the processor in communication, the computer readable storage medium is suitable for storing a plurality of instructions, the processor is suitable for calling the instructions in the computer readable storage medium to execute the steps of the time series SAR data filtering method suitable for the broken area of farmland in any one of the above claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the time series SAR data filtering method suitable for the broken area of farmland in any one of claims 1-6.

Images