Method and device for reconstructing time series of ecological remote sensing parameters from environment satellite no. 2

Abstract

This invention discloses a method and apparatus for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite, belonging to the field of remote sensing monitoring technology. The method includes: acquiring environmental remote sensing parameter data from the Environment-2 satellite and MODIS ecological remote sensing parameter data covering the same study area at the same time series, and preprocessing them; downscaling the MODIS ecological remote sensing parameter data and then interleaving it with the Environment-2 satellite ecological remote sensing parameter data according to time to obtain interleaved sequence data; performing sliding filtering on the interleaved sequence data according to a set time filtering window along the time dimension to obtain smoothed sequence data; and extracting the Environment-2 satellite ecological remote sensing parameter data from the smoothed sequence data to obtain reconstructed sequence data. This invention makes the reconstructed time series of Environment-2 satellite ecological remote sensing parameter products smoother and more stable, more realistically reflecting the changing patterns of ground features, and improving the usability of the 16-meter ecological remote sensing parameter data from the Environment-2 satellite.

Description

Technical Field

[0001] This invention relates to the field of remote sensing monitoring technology, specifically to a method and apparatus for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite. Background Technology

[0002] The HJ-2 A and B satellites are each equipped with four 16-meter resolution multispectral cameras, with a single satellite imaging swath of 800 kilometers, offering advantages such as wide coverage and short revisit cycles. Among them, the long-term 16-meter resolution ecological remote sensing parameter products can provide more effective spatial information support for rapid dynamic monitoring of the ecological environment, large-scale ecological status assessment, and resource management.

[0003] Although the processing technology for ecological parameter products from single-scene data of the Environment-2 satellite is relatively mature, there are obvious stitching gaps between multi-track images due to multiple factors such as weather conditions, sensor stability, shooting angle, atmospheric correction algorithms, and radiometric consistency processing technology for long-term series. In areas where cloud and rain weather has a large number of invalid values, there are also many invalid values.

[0004] Currently, a system for automated production of continuous products (such as monthly NDVI) for many years has not yet been established nationwide or even globally. The diversity, standardization, and continuous update services of the products still lag behind mature data sources abroad (such as MODIS, Landsat, Sentinel-2, etc.), which directly restricts the application and development of the Environment-2 satellite in large-area ecological environment monitoring and spatiotemporal pattern analysis. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method and apparatus for reconstructing the time series of ecological remote sensing parameters from the Environment-2 satellite. This method and apparatus result in a smoother and more stable time series of ecological remote sensing parameter products from the Environment-2 satellite, which more accurately reflects the patterns of ground feature changes and improves the usability of the 16-meter ecological remote sensing parameter data from the Environment-2 satellite.

[0006] The technical solution provided by this invention is as follows:

[0007] A method for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite, the method comprising:

[0008] S1: Acquire environmental remote sensing parameter data from the Environment-2 satellite and MODIS ecological remote sensing parameter data covering the same study area at the same time series;

[0009] S2: Preprocess the ecological remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data;

[0010] S3: Downscale the MODIS ecological remote sensing parameter data to the same spatial resolution as the environmental satellite ecological remote sensing parameter data;

[0011] S4: The ecological remote sensing parameter data from the Environment-2 satellite and the downscaled MODIS ecological remote sensing parameter data are sorted in an interleaved order according to time to obtain interleaved sequence data;

[0012] S5: Perform sliding filtering on the interleaved sequence data according to the time dimension based on the set time filtering window to obtain smooth sequence data;

[0013] S6: Extract the ecological remote sensing parameter data of the Environment-2 satellite from the smoothed sequence data to obtain the reconstructed sequence data of the ecological remote sensing parameters of the Environment-2 satellite.

[0014] Furthermore, S5 includes:

[0015] S51: Set the window length of the time filtering window and polynomial order ;

[0016] in, , , ;

[0017] S52: Calculate the filter coefficient vector based on the least squares principle, according to the set window length and polynomial order. ;

[0018]

[0019]

[0020]

[0021]

[0022] S53: Slide the time filtering window across the interleaved sequence data along the time dimension. Within each time filtering window, convolve the filtering coefficient vector with the interleaved sequence data to obtain smoothed sequence data. ;

[0023]

[0024] in, For the smoothed sequence data of the first One data point, , The total number of sequences in the smoothed sequence data. This is the position index of the relative center point within the time filtering window. ; The filter coefficient vector The Each element value For the interleaved sequence data of the first Data.

[0025] Furthermore, the interleaved sequence data for:

[0026]

[0027] in, For the first Ecological remote sensing parameter data from the Environment-2 satellite. For the first MODIS ecological remote sensing parameter data after downscaling transformation , This refers to the total number of periods of the environmental remote sensing parameter data from the Environment-2 satellite / MODIS environmental remote sensing parameter data.

[0028] Furthermore, S2 includes:

[0029] S21: The scale factor of the ecological remote sensing parameter data of the Environment-2 satellite and the ecological remote sensing parameter data of MODIS is uniformly set to 10000.

[0030] S22: Set the pixels with pixel values ​​in the range of [-3000, 10000] of the environmental remote sensing parameter data of the Environment-2 satellite and the MODIS ecological remote sensing parameter data as valid values, and set other pixel values ​​as invalid values.

[0031] Furthermore, S3 includes:

[0032] Using land use type data with a spatial resolution of 16 meters as the trend surface factor, the MODIS ecological remote sensing parameter data with a spatial resolution of 250 meters was downscaled to a spatial resolution of 16 meters using the TsHARP downscaling method.

[0033] Furthermore, the method also includes:

[0034] S7: Perform post-processing and data inspection on the reconstructed sequence data;

[0035] Furthermore, S7 includes:

[0036] S71: The first part of the reconstructed sequence data One and last Boundary fitting is performed on each data point;

[0037] S72: Reassign values ​​to data points in the reconstructed sequence data that are outside the data range;

[0038] S72: For data points in the reconstructed sequence data that are null, use the average value of data points within a predetermined range before and after the null value.

[0039] A device for reconstructing time series ecological remote sensing parameters from the Environment-2 satellite, the device comprising:

[0040] The data acquisition module is used to acquire environmental remote sensing parameter data from the Environment-2 satellite and MODIS ecological remote sensing parameter data covering the same study area and the same time series.

[0041] The preprocessing module is used to preprocess the ecological remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data.

[0042] The downscaling conversion module is used to downscale the MODIS ecological remote sensing parameter data to the same spatial resolution as the environmental satellite ecological remote sensing parameter data of the Environment-2 satellite.

[0043] The staggered sorting module is used to stagger the ecological remote sensing parameter data of the Environment-2 satellite and the downscaled MODIS ecological remote sensing parameter data according to time to obtain staggered sequence data.

[0044] The sliding filter module is used to perform sliding filtering on the interleaved sequence data according to the time dimension according to the set time filtering window to obtain smooth sequence data;

[0045] The data extraction module is used to extract the ecological remote sensing parameter data of the Environment-2 satellite from the smoothed sequence data to obtain the reconstructed sequence data of the ecological remote sensing parameters of the Environment-2 satellite.

[0046] Furthermore, the sliding filter module includes:

[0047] The setting unit is used to set the window length of the time filtering window. and polynomial order ;

[0048] in, , , ;

[0049] The filter coefficient vector determination unit is used to calculate the filter coefficient vector based on the least squares principle, according to the set window length and polynomial order. ;

[0050]

[0051]

[0052]

[0053]

[0054] A sliding filter unit is used to slide the time filtering window across the interleaved sequence data along the time dimension. Within each time filtering window, the filter coefficient vector is convolved with the interleaved sequence data to obtain smoothed sequence data. ;

[0055]

[0056] in, For the smoothed sequence data of the first One data point, , The total number of sequences in the smoothed sequence data. This is the position index of the relative center point within the time filtering window. ; The filter coefficient vector The Each element value For the interleaved sequence data of the first Data.

[0057] Furthermore, the interleaved sequence data for:

[0058]

[0059] in, For the first Ecological remote sensing parameter data from the Environment-2 satellite. For the first MODIS ecological remote sensing parameter data after downscaling transformation , This refers to the total number of periods of the environmental remote sensing parameter data from the Environment-2 satellite / MODIS environmental remote sensing parameter data.

[0060] Furthermore, the preprocessing module includes:

[0061] The scaling factor unification unit is used to unify the scaling factor of the ecological remote sensing parameter data of the Environment-2 satellite and the ecological remote sensing parameter data of MODIS to 10000 respectively.

[0062] The pixel range setting unit is used to set the pixel values ​​of the environmental remote sensing parameter data of the Environment-2 satellite and the MODIS ecological remote sensing parameter data to be in the range of [-3000, 10000] as valid values, and set other pixel values ​​to invalid values.

[0063] Furthermore, the downscaling conversion module is used for:

[0064] Using land use type data with a spatial resolution of 16 meters as the trend surface factor, the MODIS ecological remote sensing parameter data with a spatial resolution of 250 meters was downscaled to a spatial resolution of 16 meters using the TsHARP downscaling method.

[0065] Furthermore, the device also includes:

[0066] The post-processing module is used to perform post-processing and data inspection on the reconstructed sequence data;

[0067] Furthermore, the post-processing module includes:

[0068] Boundary fitting unit, used for the first part of the reconstructed sequence data One and last Boundary fitting is performed on each data point;

[0069] A data range checking unit is used to reassign values ​​to data points in the reconstructed sequence data that are outside the data range.

[0070] The null value checking unit is used to replace null values ​​in the reconstructed sequence data with the average value of data points within a predetermined range before and after the null values.

[0071] The present invention has the following beneficial effects:

[0072] Compared to the limitations of traditional single-scene products of 16-meter ecological remote sensing parameters from the Environment-2 satellite, this invention innovatively combines the 16-meter ecological remote sensing parameter products from the Environment-2 satellite with MODIS ecological parameter products through spatiotemporal correlation and multi-satellite data interleaving and recombination. By employing a time-filtering smoothing algorithm, data smoothing is performed within a time sliding window, thereby removing noise while preserving data details and peak values ​​to the greatest extent possible. This can reduce the adverse effects of weather conditions and data quality factors on time series.

[0073] This invention ingeniously integrates the temporal characteristics of two data sources. It retains the spatial information of the high spatial resolution data of the environmental remote sensing parameter products from the Environment-2 satellite, while using MODIS products to remove obvious seams between adjacent scenes and fill in missing pixel values ​​caused by weather factors such as clouds and rain. This makes the reconstructed environmental remote sensing parameter products more smooth and stable in time series, more realistically reflecting the changing patterns of ground features. It improves the usability of the 16-meter ecological remote sensing parameter data from the Environment-2 satellite, and lays a solid technical foundation for the automated and standardized production of a long-term, seamless national-scale 16-meter ecological parameter product system. This significantly enhances the application value of the Environment-2 satellite data in the field of ecological and environmental monitoring. Attached Figure Description

[0074] Figure 1 This is a flowchart of the time series reconstruction method for ecological remote sensing parameters from the Environment-2 satellite of the present invention;

[0075] Figure 2 A schematic diagram of the high-quality 16-meter NDVI product of the Environment-2 satellite after reconstruction in Hongyuan County, Sichuan Province from January to December 2022;

[0076] Figure 3 This is a schematic diagram of the time series reconstruction device for ecological remote sensing parameters of the Environment-2 satellite according to the present invention. Detailed Implementation

[0077] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0078] This invention provides a method for reconstructing time series ecological remote sensing parameters from the Environment-2 satellite. The aim is to solve the technical problem that existing 16-meter ecological remote sensing parameter products from the Environment-2 satellite have obvious stitching gaps and severe pixel loss due to factors such as weather and imaging conditions, making them unsuitable for direct use in long-term, large-area ecological environment monitoring and analysis.

[0079] like Figure 1 As shown, the method includes:

[0080] S1: Acquire ecological remote sensing parameter data from the HJ-2 satellite and MODIS ecological remote sensing parameter data covering the same study area and the same time series.

[0081] Taking the Normalized Difference Vegetation Index (NDVI) as an example, 120 monthly NDVI products with HJ-2 16-meter spatial resolution from 2010 to 2020 and 120 monthly NDVI products with MODIS 250-meter spatial resolution from 2010 to 2020 were obtained. Simultaneously, land use type data at 16-meter resolution for the study area were collected as a benchmark for subsequent downscaling.

[0082] S2: Preprocess the ecological remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data.

[0083] Preprocessing includes standardizing the scaling factor and setting invalid values ​​to obtain NDVI data in the same unit of measurement for the study area.

[0084] A specific implementation process includes:

[0085] S21: The scale factor of the ecological remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data is set to 10000.

[0086] S22: Set the pixel values ​​in the range of [-3000, 10000] of the environmental remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data as valid values, and set other pixel values ​​as the invalid value "32767".

[0087] S3: Downscale the MODIS ecological remote sensing parameter data to achieve the same spatial resolution as the environmental remote sensing parameter data from the Environment-2 satellite.

[0088] Specifically, land use type data with a spatial resolution of 16 meters can be used as trend surface factors. The TsHARP downscaling method can be used to convert MODIS ecological remote sensing parameter data with a spatial resolution of 250 meters to a spatial resolution of 16 meters. This yields MODIS NDVI data at the same scale as the HJ-2 16-meter NDVI data.

[0089] S4: The ecological remote sensing parameter data from the Environment-2 satellite and the downscaled MODIS ecological remote sensing parameter data are interleaved according to time to obtain interleaved sequence data.

[0090] In the interleaved sorting process, the data can be renamed and reorganized according to time series rules to obtain data sorted in a staggered order according to the satellite sensor sequence.

[0091] One specific implementation method is as follows:

[0092] 1. Rename the MODIS NDVI products for the first month of 2020 after downscaling conversion, such as naming the product "NDVI_2020_01_MODIS.tiff", and so on.

[0093] 2. Rename the HJ-2 NDVI products according to the same rules, such as renaming the HJ-2 products for the first month of 2020 as "NDVI_2020_01_HJ.tiff", and so on.

[0094] 3. Place the renamed MODIS NDVI and HJ-2 NDVI data in the same directory, and sort them in the order of "NDVI_2020_01_HJ.tiff", "NDVI_2020_01_MODIS.tiff", etc., to obtain interleaved sequence data sorted in a staggered order according to the different satellite sensor sequences. .

[0095]

[0096] in, For the first Ecological remote sensing parameter data from the Environment-2 satellite. For the first MODIS ecological remote sensing parameter data after downscaling transformation , This refers to the total number of periods of the environmental remote sensing parameter data from the Environment-2 satellite / MODIS environmental remote sensing parameter data. The total number of data sequences is .

[0097] S5: Perform sliding filtering on the interleaved sequence data according to the time dimension based on the set time filtering window to obtain smooth sequence data.

[0098] This step involves setting a time-based filtering window and performing a sliding filter along the time dimension. The value of the center point of the fitted time-based filtering window is then calculated as the filtering result for that point. By sliding the time-based filtering window and repeating the above process, smoothed sequence data is obtained.

[0099] As an example, this step includes:

[0100] S51: Set the window length of the time filtering window and polynomial order ;

[0101] Interleaved sequence data by If we represent it, then the interleaved sequence data can be written as Set the sliding window length to... Window half width is ( It is an odd number. ); set the order of the polynomial fitting as (satisfy ).

[0102] S52: Calculate the filter coefficient vector based on the least squares principle, according to the set window length and polynomial order. ;

[0103]

[0104]

[0105]

[0106]

[0107] S53: Apply time filtering window to interleaved sequence data The process slides along the time dimension, and within each time filtering window, the filter coefficient vector is convolved with the interleaved sequence data to obtain the smoothed sequence data. ;

[0108]

[0109] in, For the smoothed sequence data Each data point is also the center point of its corresponding time filter window. The filtering results , This represents the total number of sequences in the smoothed sequence data.

[0110] The relative center point within the time filtering window Location index, ,Right now , .

[0111] These are the filter convolution kernel coefficients (smoothing coefficients), determined by the window length. and the order of the fitted polynomial The unique determination is obtained through the least squares method, which is the filter coefficient vector. The Each element value.

[0112] For the interleaved sequence data Data.

[0113] By moving the window and repeating the above process, smoothing filtering of NDVI data across the entire study area can be achieved, resulting in smoothed sequence data. .

[0114] S6: Extract the ecological remote sensing parameter data of the Environment-2 satellite from the smoothed sequence data to obtain the reconstructed sequence data of the ecological remote sensing parameters of the Environment-2 satellite.

[0115] by For example, the extracted reconstructed sequence data .

[0116] S7: Perform post-processing and data inspection on the reconstructed sequence data.

[0117] Specific implementation methods include:

[0118] S71: The first step in reconstructing the sequence data One and last Each data point is fitted to the boundary using forward / backward difference or mirror symmetry methods to improve the accuracy of the boundary points.

[0119] S72: Reassign values ​​to data points in the reconstructed sequence data that are outside the data range.

[0120] The NDVI data range should be [-3000, 10000]. Pixel values ​​less than -3000 should be assigned the value -3000, and pixel values ​​greater than 10000 should be assigned the value 10000 (except for invalid values).

[0121] S72: For data points that are null in the reconstructed sequence data, use the average value of data points within a specified range before and after the null value.

[0122] Specifically, for the few remaining null values, a method of averaging the values ​​of the preceding and following months (mean filter) is used for rapid processing. The final result is the HJ-2 16-meter resolution NDVI product after time-series reconstruction and denoising.

[0123] Compared to the limitations of traditional single-scene products of 16-meter ecological remote sensing parameters from the Environment-2 satellite, this invention innovatively combines the 16-meter ecological remote sensing parameter products from the Environment-2 satellite with MODIS ecological parameter products through spatiotemporal correlation and multi-satellite data interleaving and recombination. By employing a time-filtering smoothing algorithm, data smoothing is performed within a time sliding window, thereby removing noise while preserving data details and peak values ​​to the greatest extent possible. This can reduce the adverse effects of weather conditions and data quality factors on time series.

[0124] This invention ingeniously integrates the temporal characteristics of two data sources. It retains the spatial information of the high spatial resolution data of the environmental remote sensing parameter products from the Environment-2 satellite, while using MODIS products to remove obvious seams between adjacent scenes and fill in missing pixel values ​​caused by weather factors such as clouds and rain. This makes the reconstructed environmental remote sensing parameter products more smooth and stable in time series, more realistically reflecting the changing patterns of ground features. It improves the usability of the 16-meter ecological remote sensing parameter data from the Environment-2 satellite, and lays a solid technical foundation for the automated and standardized production of a long-term, seamless national-scale 16-meter ecological parameter product system. This significantly enhances the application value of the Environment-2 satellite data in the field of ecological and environmental monitoring.

[0125] To illustrate the solution and effects of the present invention in detail, a specific embodiment is described below:

[0126] 1. Data preparation.

[0127] HJ-2 16-meter NDVI data and MODIS 250-meter NDVI data were acquired for Hongyuan County, Sichuan Province from January to December 2022. Land use type data at 16-meter resolution were also collected.

[0128] 2. HJ-2 and MODIS data preprocessing.

[0129] The acquired HJ-2 16-meter NDVI data and MODIS 250-meter NDVI data were preprocessed, including unifying the scaling factor and setting invalid values, to obtain NDVI data of the same measurement unit for the study area.

[0130] (1) The scale factor of HJ-2 16-meter NDVI data and MODIS 250-meter NDVI data is unified to 10000.

[0131] (2) Set the pixels with a value range of [-3000, 10000] as valid values ​​and set other pixel values ​​as invalid values ​​"32767".

[0132] 3. Downscaling conversion of low spatial resolution MODIS NDVI data.

[0133] Land use type data with a resolution of 16 meters was selected as the trend surface factor. The TsHARP classic downscaling method was used to downscale the low spatial resolution MODIS 250-meter NDVI data to a resolution of 16 meters, resulting in MODIS NDVI data at the same scale as the HJ-2 16-meter NDVI data.

[0134] 4. Interleaved and recombined data from two stars.

[0135] The downscaled MODIS NDVI and the original HJ-2 16-meter NDVI data were renamed and reorganized according to time series rules to obtain data sorted in a staggered order according to the satellite sensor sequence.

[0136] 5. Filter timing reconstruction.

[0137] For the sorted multi-source remote sensing vegetation index time series data in the file directory, a window width of 5 and a polynomial order of 3 are selected. The value of the fitted polynomial at the center point of the window is calculated and used as the filtering result for that point. By moving the window and repeating the above process, the NDVI data reconstructed from the time series through iterative calculation is obtained.

[0138] 6. HJ-2 data extraction and post-processing.

[0139] Extract the filtered 16-meter NDVI data corresponding to HJ-2 from the processed file directory, and perform data quality checks and post-processing. The final result is a high-quality HJ-2 16-meter resolution NDVI product after time-series reconstruction and denoising processing, such as... Figure 2 As shown.

[0140] This invention also provides a device for reconstructing time series ecological remote sensing parameters from the Environment-2 satellite, such as... Figure 3 As shown, the device includes:

[0141] Data acquisition module 1 is used to acquire environmental remote sensing parameter data from the Environment 2 satellite and MODIS ecological remote sensing parameter data covering the same study area and the same time series.

[0142] Preprocessing module 2 is used to preprocess the ecological remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data.

[0143] Downscaling module 3 is used to downscale MODIS ecological remote sensing parameter data to the same spatial resolution as the environmental satellite ecological remote sensing parameter data.

[0144] The staggered sorting module 4 is used to stagger the ecological remote sensing parameter data of the Environment-2 satellite and the downscaled MODIS ecological remote sensing parameter data according to time to obtain staggered sequence data.

[0145] The sliding filter module 5 is used to perform sliding filtering on the interleaved sequence data according to the time dimension according to the set time filtering window to obtain smooth sequence data.

[0146] Data extraction module 6 is used to extract the ecological remote sensing parameter data of the Environment-2 satellite from the smoothed sequence data to obtain the reconstructed sequence data of the ecological remote sensing parameters of the Environment-2 satellite.

[0147] As an example, the sliding filter module includes:

[0148] The setting unit is used to set the window length of the time filtering window. and polynomial order .

[0149] in, , , .

[0150] The filter coefficient vector determination unit is used to calculate the filter coefficient vector based on the least squares principle, according to the set window length and polynomial order. .

[0151]

[0152]

[0153]

[0154]

[0155] The sliding filter unit slides the time filtering window across the interleaved sequence data along the time dimension. Within each time filtering window, the filter coefficient vector is convolved with the interleaved sequence data to obtain smoothed sequence data. .

[0156]

[0157] in, For the smoothed sequence data One data point, , This represents the total number of sequences in the smoothed sequence data. This is the position index of the relative center point within the time filtering window. ; The filter coefficient vector The Each element value For the interleaved sequence data Data.

[0158] Specifically, the aforementioned interleaved sequence data for:

[0159]

[0160] in, For the first Ecological remote sensing parameter data from the Environment-2 satellite. For the first MODIS ecological remote sensing parameter data after downscaling transformation , This refers to the total number of periods of the environmental remote sensing parameter data from the Environment-2 satellite / MODIS environmental remote sensing parameter data.

[0161] In this invention, the preprocessing module includes:

[0162] The scaling factor unification unit is used to unify the scaling factor of the ecological remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data to 10000.

[0163] The pixel range setting unit is used to set the pixel values ​​in the range of [-3000, 10000] of the environmental remote sensing parameter data of the Environment-2 satellite and the MODIS ecological remote sensing parameter data as valid values, and set other pixel values ​​as invalid values.

[0164] The downscaling conversion module is used for:

[0165] Using land use type data with a spatial resolution of 16 meters as the trend surface factor, the MODIS ecological remote sensing parameter data with a spatial resolution of 250 meters was downscaled to a spatial resolution of 16 meters using the TsHARP downscaling method.

[0166] As an improvement to an embodiment of the present invention, the apparatus further includes:

[0167] The post-processing module is used to perform post-processing and data inspection on the reconstructed sequence data.

[0168] Specifically, the post-processing module includes:

[0169] Boundary fitting unit, used for the first step of reconstructing sequence data One and last Boundary fitting is performed on each data point.

[0170] The data range checking unit is used to reassign values ​​to data points in the reconstructed sequence data that are outside the data range.

[0171] The null value checking unit is used to replace null values ​​in the reconstructed sequence data with the average value of data points within a specified range before and after the null value.

[0172] The apparatus provided in this embodiment of the invention operates on the same principle and produces the same technical effects as the aforementioned method embodiments. For the sake of brevity, any parts not mentioned in the apparatus embodiments can be referred to the corresponding content in the aforementioned method embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the apparatus and units described above can all be referred to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

[0173] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit them. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the scope of the technology disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, 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, and should all be covered within the scope of protection of the present invention.

Claims

1. A method for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite, characterized in that, The method includes: S1: Acquire environmental remote sensing parameter data from the Environment-2 satellite and MODIS ecological remote sensing parameter data covering the same study area at the same time series; S2: Preprocess the ecological remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data; S3: Downscale the MODIS ecological remote sensing parameter data to the same spatial resolution as the environmental satellite ecological remote sensing parameter data; S4: The ecological remote sensing parameter data from the Environment-2 satellite and the downscaled MODIS ecological remote sensing parameter data are sorted in an interleaved order according to time to obtain interleaved sequence data; S5: Perform sliding filtering on the interleaved sequence data according to the time dimension based on the set time filtering window to obtain smooth sequence data; S6: Extract the ecological remote sensing parameter data of the Environment-2 satellite from the smoothed sequence data to obtain the reconstructed sequence data of the ecological remote sensing parameters of the Environment-2 satellite.

2. The method for reconstructing time series ecological remote sensing parameters from the Environment-2 satellite according to claim 1, characterized in that, S5 includes: S51: Set the window length of the time filtering window and polynomial order ; in, , , ; S52: Calculate the filter coefficient vector based on the least squares principle, according to the set window length and polynomial order. ; S53: Slide the time filtering window across the interleaved sequence data along the time dimension. Within each time filtering window, convolve the filtering coefficient vector with the interleaved sequence data to obtain smoothed sequence data. ; in, For the smoothed sequence data of the first One data point, , The total number of sequences in the smoothed sequence data. This is the position index of the relative center point within the time filtering window. ; The filter coefficient vector The Each element value For the interleaved sequence data of the first Data.

3. The method for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite according to claim 2, characterized in that, The interleaved sequence data for: in, For the first Ecological remote sensing parameter data from the Environment-2 satellite. For the first MODIS ecological remote sensing parameter data after downscaling transformation , This refers to the total number of periods of the environmental remote sensing parameter data from the Environment-2 satellite / MODIS environmental remote sensing parameter data.

4. The method for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite according to claim 3, characterized in that, S2 includes: S21: The scale factor of the ecological remote sensing parameter data of the Environment-2 satellite and the ecological remote sensing parameter data of MODIS is uniformly set to 10000. S22: Set the pixels with pixel values ​​in the range of [-3000, 10000] of the environmental remote sensing parameter data of the Environment-2 satellite and the MODIS ecological remote sensing parameter data as valid values, and set other pixel values ​​as invalid values.

5. The method for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite according to claim 3, characterized in that, S3 includes: Using land use type data with a spatial resolution of 16 meters as the trend surface factor, the MODIS ecological remote sensing parameter data with a spatial resolution of 250 meters was downscaled to a spatial resolution of 16 meters using the TsHARP downscaling method.

6. The method for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite according to any one of claims 3-5, characterized in that, The method further includes: S7: Perform post-processing and data inspection on the reconstructed sequence data.

7. The method for reconstructing time series of ecological remote sensing parameters from the Environment-2 satellite according to claim 6, characterized in that, S7 includes: S71: The first part of the reconstructed sequence data One and last Boundary fitting is performed on each data point; S72: Reassign values ​​to data points in the reconstructed sequence data that are outside the data range; S72: For data points in the reconstructed sequence data that are null, use the average value of data points within a predetermined range before and after the null value.

8. A time series reconstruction device for ecological remote sensing parameters from the Environment-2 satellite, characterized in that, The device includes: The data acquisition module is used to acquire environmental remote sensing parameter data from the Environment-2 satellite and MODIS ecological remote sensing parameter data covering the same study area and the same time series. The preprocessing module is used to preprocess the ecological remote sensing parameter data from the Environment-2 satellite and the MODIS ecological remote sensing parameter data. The downscaling conversion module is used to downscale the MODIS ecological remote sensing parameter data to the same spatial resolution as the environmental satellite ecological remote sensing parameter data of the Environment-2 satellite. The staggered sorting module is used to stagger the ecological remote sensing parameter data of the Environment-2 satellite and the downscaled MODIS ecological remote sensing parameter data according to time to obtain staggered sequence data. The sliding filter module is used to perform sliding filtering on the interleaved sequence data according to the time dimension according to the set time filtering window to obtain smooth sequence data; The data extraction module is used to extract the ecological remote sensing parameter data of the Environment-2 satellite from the smoothed sequence data to obtain the reconstructed sequence data of the ecological remote sensing parameters of the Environment-2 satellite.

9. The device for reconstructing time series ecological remote sensing parameters from the Environment-2 satellite according to claim 8, characterized in that, The sliding filter module includes: The setting unit is used to set the window length of the time filtering window. and polynomial order ; in, , , ; The filter coefficient vector determination unit is used to calculate the filter coefficient vector based on the least squares principle, according to the set window length and polynomial order. ; A sliding filter unit is used to slide the time filtering window across the interleaved sequence data along the time dimension. Within each time filtering window, the filter coefficient vector is convolved with the interleaved sequence data to obtain smoothed sequence data. ; in, For the smoothed sequence data of the first One data point, , The total number of sequences in the smoothed sequence data. This is the position index of the relative center point within the time filtering window. ; The filter coefficient vector The Each element value For the interleaved sequence data of the first Data.

Images