Remote sensing monitoring method, device and equipment for grassland green returning period in high-altitude complex terrain area

Abstract

The invention provides a remote sensing monitoring method, a remote sensing monitoring device and remote sensing monitoring equipment for a grassland reviving period in a high-altitude complex terrain area, and relates to the technical field of reviving period estimation. An initial long-time-sequence vegetation index set corresponding to the high-altitude complex terrain area is constructed based on the preprocessed satellite remote sensing image; performing local fitting and global fitting on the initial long-time-sequence vegetation index set to reconstruct the initial long-time-sequence vegetation index set, and obtaining a target long-time-sequence vegetation index set corresponding to the high-altitude complex terrain region; and performing geographical partition on the high-altitude complex terrain area, and determining a green establishment period estimation result corresponding to the high-altitude complex terrain area based on the target long-time sequence vegetation index set, the background field vegetation index threshold and the geographical partition corresponding to the high-altitude complex terrain area. The method effectively solves the problem that a traditional threshold value method is insufficient in adaptability in high-altitude and complex-terrain areas.

Description

Technical Field

[0001] This invention relates to the field of grassland greening period estimation technology, and in particular to a remote sensing monitoring method, device and equipment for grassland greening period in high-altitude complex terrain areas. Background Technology

[0002] The grassland greening-up period is a crucial biological rhythm indicator in grassland ecosystems, signifying the transition of plants from dormancy to rapid growth. It directly impacts grassland productivity, carbon cycling, and ecosystem services. Accurately understanding the spatiotemporal variations of the grassland greening-up period is essential for addressing climate change, guiding grassland management, and optimizing ecological protection. However, due to the diversity of grassland types, complex terrain, and the uncertainty of climate change, traditional ground-based observation methods struggle to achieve large-scale, high-frequency monitoring of the grassland greening-up period. Satellite remote sensing technology, with its advantages of large-scale coverage, multi-temporal observation, and multi-source data fusion, provides an important tool for retrieving the grassland greening-up period.

[0003] In recent years, grassland greening-up period retrieval technology based on satellite remote sensing has developed rapidly, mainly relying on various methods such as optical remote sensing, microwave remote sensing, and multi-source data fusion. Typical research methods include using optical remote sensing features such as vegetation indices (e.g., Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI)) to extract key phenological points of vegetation growth through time series analysis. Currently, there are various methods for grassland greening-up period retrieval from long-term remote sensing data, including thresholding, absolute value, slope, and curve fitting methods. The slope method identifies the greening-up and yellowing times of vegetation growth by taking the first derivative of the vegetation index curve. It is suitable for phenological extraction in areas with stable growth states and minimal influence from the external environment, but its accuracy decreases when external conditions change and affect phenology. The curve fitting method fits the vegetation growth process based on mathematical models and is suitable for areas with minimal changes and good vegetation conditions. However, its accuracy is low in low-vegetation-cover areas and it is not suitable for grasslands with high altitudes and complex terrain. The absolute value method takes a specific absolute value as the start or end of growth, but due to differences in vegetation growth between years, it has a large error margin compared to the actual situation. The threshold method obtains the start and end times of vegetation growth by setting a fixed or dynamic threshold; however, the fixed threshold method is significantly affected by different regions and subjective human factors. Meanwhile, with the introduction of data-driven technologies such as machine learning and deep learning, the accuracy and adaptability of grassland greening-up period inversion have been further improved. Remote sensing inversion technology combined with ecological models has also enhanced the explanatory power of the grassland greening-up period formation mechanism. However, current technologies still face challenges such as data resolution limitations, insufficient model regional applicability, and challenges in multi-source data fusion. While various methods exist for estimating grassland greening-up periods, most are primarily applicable to vast areas with well-developed vegetation and relatively flat terrain. However, in high-altitude, topographically complex areas (such as Tibet), due to their unique natural geographical conditions, complex high-altitude topography, and the fragility and sensitivity of their ecosystems, these areas have become sensitive zones to global climate change. Grassland phenological information and its response to climate change in these areas are not only of significant ecological importance but also a hot topic in global change research. However, a remote sensing method suitable for monitoring grassland greening-up periods in high-altitude, topographically complex regions is currently lacking. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a remote sensing monitoring method, device and equipment for grassland greening period in high-altitude complex terrain areas, which effectively solves the problem of insufficient adaptability of traditional threshold methods in high-altitude and complex terrain areas.

[0005] In a first aspect, embodiments of the present invention provide a remote sensing monitoring method for grassland greening-up period in high-altitude complex terrain areas, including: Satellite remote sensing images corresponding to high-altitude complex terrain areas are preprocessed, and an initial long-term vegetation index set corresponding to high-altitude complex terrain areas is constructed based on the preprocessed satellite remote sensing images. Local and global fitting were performed on the initial long-term vegetation index set to reconstruct the initial long-term vegetation index set and obtain the target long-term vegetation index set corresponding to the high-altitude complex terrain area. Geographically partition the high-altitude complex terrain region, and based on the target long-term vegetation index set, background field vegetation index threshold, and geographic partition, determine the greening period estimation results for the high-altitude complex terrain region.

[0006] In one implementation, the initial long-term vegetation index set includes a subset of initial long-term vegetation indices corresponding to each pixel in a high-altitude complex terrain area; local and global fitting are performed on the initial long-term vegetation index set to reconstruct the initial long-term vegetation index set, resulting in a target long-term vegetation index set corresponding to the high-altitude complex terrain area, including: Perform the following operation on the initial long-term vegetation index subset corresponding to each pixel: A sliding window is established for the initial long-term vegetation index subset corresponding to the pixel. The vegetation index within the sliding window is locally fitted. The sliding window is moved and the vegetation index within the moved sliding window is locally fitted until each vegetation index within the initial long-term vegetation index subset corresponding to the pixel is locally fitted, thus obtaining the local long-term vegetation index subset corresponding to the pixel. In addition, the global trend information represented by the initial long-term vegetation index subset corresponding to the pixel is determined, and the initial long-term vegetation index subset is globally fitted by combining the asymmetric Gaussian function to obtain the global long-term vegetation index subset corresponding to the pixel. The local long-term vegetation index subset and the global long-term vegetation index subset corresponding to the pixel are weighted and reconstructed to obtain the target long-term vegetation index subset corresponding to the pixel.

[0007] In one implementation, local fitting of vegetation indices within a sliding window includes: The parameter values ​​of the convolution operation are fitted using multiple vegetation indices located within a sliding window; Based on the fitted parameter values, a convolution operation is performed on the vegetation index located at the center point of the sliding window. By using the new vegetation index obtained from convolution operations to replace the vegetation index located at the center point of the sliding window, local fitting of the vegetation index located within the sliding window can be achieved.

[0008] In one implementation, the global trend information represented by the initial long-term vegetation index subset corresponding to the pixel is determined, and the initial long-term vegetation index subset is globally fitted using an asymmetric Gaussian function to obtain the global long-term vegetation index subset corresponding to the pixel, including: Determine the amplitude and standard deviation of the vegetation index corresponding to the initial long-term vegetation index subset of the pixel, so as to characterize the global trend information through the amplitude and standard deviation of the vegetation index; By using an asymmetric Gaussian function, based on the time information, amplitude, standard deviation of vegetation index, and center point of the asymmetric Gaussian function corresponding to the initial long-term vegetation index subset, a global fitting is performed on the initial long-term vegetation index subset to obtain the global long-term vegetation index subset corresponding to the pixel.

[0009] In one implementation, based on the target long-term vegetation index set, background field vegetation index thresholds, and geographical zones corresponding to high-altitude complex terrain areas, the estimation results of the greening period corresponding to high-altitude complex terrain areas are determined, including: Obtain the historical long-term vegetation index set corresponding to high-altitude complex terrain areas, and synthesize the historical long-term vegetation index set by rolling mean to obtain the background field vegetation index threshold. Based on the greening period threshold corresponding to each geographic sub-region within the high-altitude complex terrain area, as well as the target long-term vegetation index set and background field vegetation index threshold corresponding to the high-altitude complex terrain area, the greening period of the high-altitude complex terrain area is estimated pixel by pixel, and the greening period estimation result corresponding to the high-altitude complex terrain area is obtained; among which, the greening period estimation result includes the date on which each pixel in the high-altitude complex terrain area enters the greening period.

[0010] In one implementation, the target long-term vegetation index set includes a subset of target long-term vegetation indices corresponding to each pixel within the high-altitude complex terrain region; based on the greening-off period threshold corresponding to each geographic region within the high-altitude complex terrain region, and the target long-term vegetation index set and background field vegetation index threshold corresponding to the high-altitude complex terrain region, the greening-off period of the high-altitude complex terrain region is estimated pixel-by-pixel, yielding the greening-off period estimation result corresponding to the high-altitude complex terrain region, including: Perform the following operation on the subset of target long-term vegetation indices corresponding to each pixel: Determine the vegetation index increase relative to the background field vegetation index threshold for each vegetation index within the target long-term vegetation index subset corresponding to the pixel. Based on the increase in each vegetation index and the greening-up period threshold corresponding to the geographic region where the pixel is located, the date on which the pixel enters the greening-up period is determined.

[0011] In one implementation, the date on which a pixel enters the greening period is determined based on the increase in each vegetation index and the greening period threshold corresponding to the geographic region where the pixel is located, including: Based on the date corresponding to the increase in each vegetation index, it is determined whether the increase in each vegetation index reaches the threshold of the greening period corresponding to the geographic region where the pixel is located. The date corresponding to the increase in vegetation index when the judgment result is first true is determined as the date when that pixel enters the greening period.

[0012] Secondly, embodiments of the present invention also provide a remote sensing monitoring device for the grassland greening period in high-altitude complex terrain areas, comprising: The vegetation index determination module is used to preprocess satellite remote sensing images corresponding to high-altitude complex terrain areas and construct an initial long-term vegetation index set corresponding to high-altitude complex terrain areas based on the preprocessed satellite remote sensing images. The vegetation index reconstruction module is used to perform local and global fitting on the initial long-term vegetation index set to reconstruct the initial long-term vegetation index set and obtain the target long-term vegetation index set corresponding to the high-altitude complex terrain area. The greening period estimation module is used to perform geographical partitioning of high-altitude complex terrain areas, and based on the target long-term vegetation index set, background field vegetation index threshold and geographical partitioning corresponding to the high-altitude complex terrain areas, it determines the greening period estimation results corresponding to the high-altitude complex terrain areas.

[0013] Thirdly, embodiments of the present invention also provide an electronic device, including a processor and a memory, wherein the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement any of the methods provided in the first aspect.

[0014] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement any of the methods provided in the first aspect.

[0015] This invention provides a remote sensing monitoring method, apparatus, and equipment for grassland greening period in high-altitude complex terrain areas. First, satellite remote sensing images corresponding to the high-altitude complex terrain area are preprocessed, and an initial long-term vegetation index set corresponding to the high-altitude complex terrain area is constructed based on the preprocessed satellite remote sensing images. Then, the initial long-term vegetation index set is locally fitted and globally fitted to reconstruct the initial long-term vegetation index set, obtaining a target long-term vegetation index set corresponding to the high-altitude complex terrain area. Finally, the high-altitude complex terrain area is geographically partitioned, and based on the target long-term vegetation index set corresponding to the high-altitude complex terrain area, the background field vegetation index threshold, and the geographical partition, the estimated greening period corresponding to the high-altitude complex terrain area is determined. The above method constructs long-term vegetation indices for high-altitude complex terrain areas and reconstructs the long-term vegetation index set through local and global fitting. Using the reconstructed target long-term vegetation index set, combined with the geographical zoning of high-altitude complex terrain areas and background field vegetation index thresholds, the influence of different altitudes and terrains on vegetation phenology is comprehensively considered. This effectively solves the problem of insufficient adaptability of traditional threshold methods in high-altitude and complex terrain areas. At the same time, the calculation process of this method is simple and efficient, so it also has strong potential for business applications.

[0016] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or understood by practicing the present invention. The purposes and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a remote sensing monitoring method for grassland greening period in high-altitude complex terrain areas, provided by an embodiment of the present invention. Figure 2 This is an overall flowchart of a remote sensing monitoring method for grassland greening period in high-altitude complex terrain areas provided by an embodiment of the present invention; Figure 3 This is a schematic diagram of an accuracy verification result provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of a remote sensing monitoring device for the grassland greening period in a high-altitude complex terrain area provided in an embodiment of the present invention; Figure 5 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Currently, there is a lack of remote sensing monitoring methods for grassland greening-up period that can adapt to high-altitude and complex terrain areas. Based on this, the present invention provides a remote sensing monitoring method, device and equipment for grassland greening-up period in high-altitude and complex terrain areas, which effectively solves the problem of insufficient adaptability of traditional threshold methods in high-altitude and complex terrain areas.

[0022] To facilitate understanding of this embodiment, a remote sensing monitoring method for grassland greening in high-altitude complex terrain areas, as disclosed in this embodiment of the invention, will first be described in detail. (See [link to relevant documentation]). Figure 1 The diagram shows a flowchart of a remote sensing monitoring method for grassland greening period in high-altitude complex terrain areas. The method mainly includes the following steps S102 to S106: S102, preprocess the satellite remote sensing images corresponding to high-altitude complex terrain areas, and construct the initial long-term vegetation index set corresponding to high-altitude complex terrain areas based on the preprocessed satellite remote sensing images.

[0023] High-altitude complex terrain areas refer to regions with high altitudes and diverse topography, such as Tibet. Satellite remote sensing imagery can utilize FY-3D MERSI-2 target imagery at 1km resolution, Level 1 (L1) products, including raw DN (Digital Number) values ​​and geographic data. Preprocessing may include data filtering, quality checks, radiometric correction, and geometric correction. The initial long-term vegetation index set (referred to as the initial long-term NDVI (Normalized Difference Vegetation Index) dataset) includes a subset of the initial long-term vegetation indexes (referred to as the initial long-term NDVI data subset) corresponding to each pixel within the high-altitude complex terrain area. This subset includes the normalized vegetation index corresponding to that pixel on multiple dates.

[0024] In one example, after preprocessing satellite remote sensing images corresponding to high-altitude complex terrain areas, such as data screening, quality inspection, radiometric correction, and geometric correction, pixels affected by cloud interference in the satellite remote sensing images are removed. Normalized vegetation index calculation and synthesis are then performed on the retained pixels to obtain a high-quality initial long-term vegetation index set.

[0025] S104. Local and global fitting are performed on the initial long-term vegetation index set to reconstruct the initial long-term vegetation index set and obtain the target long-term vegetation index set corresponding to the high-altitude complex terrain area.

[0026] Local fitting involves fitting and replacing the normalized vegetation index (NDVI) corresponding to each pixel on any given date, based on multiple neighboring NDVI indices. Global fitting involves fitting and replacing the entire initial long-term vegetation index subset, based on the global trend (such as amplitude and vegetation index standard deviation) represented by the subset. The target long-term vegetation index set (NDVI dataset) is the long-term vegetation index set obtained after reconstructing the initial long-term vegetation index set. This set includes the target long-term vegetation index subset (NDVI dataset) corresponding to each pixel in high-altitude complex terrain areas, which includes the reconstructed normalized vegetation indices for each pixel on multiple dates.

[0027] In one example, for each pixel, after performing local and global fitting on the initial long-term vegetation index subset, two new long-term vegetation index subsets can be obtained. By performing weighted reconstruction on the two new long-term vegetation index subsets, the target long-term vegetation index subset corresponding to that pixel can be obtained. The target long-term vegetation index subsets corresponding to all pixels can constitute the target long-term vegetation index set corresponding to the high-altitude complex terrain area.

[0028] S106. Geographically partition the high-altitude complex terrain area, and determine the greening period estimation result corresponding to the high-altitude complex terrain area based on the target long-term vegetation index set, background field vegetation index threshold and geographic partition.

[0029] The background vegetation index threshold can be understood as a reference value for vegetation indices determined based on a historical long-term vegetation index set. The greening-back period estimation results include the date on which each pixel in the high-altitude complex terrain area enters the greening-back period.

[0030] In one example, DEM (Digital Elevation Model) data can be used to geographically partition high-altitude complex terrain areas, with different thresholds for the greening-out period corresponding to different geographic partitions. For each pixel within the high-altitude complex terrain area, the target long-term vegetation index subset corresponding to that pixel is compared with the background field vegetation index threshold to determine the increase in the vegetation index of that pixel. By determining when the increase in the vegetation index meets the greening-out period threshold corresponding to the geographic partition in which the pixel is located, the date on which the pixel enters the greening-out period can be determined.

[0031] The remote sensing monitoring method for grassland greening period in high-altitude complex terrain areas provided in this invention constructs long-term vegetation indices for high-altitude complex terrain areas, and reconstructs the long-term vegetation index set through local fitting and global fitting. Using the reconstructed target long-term vegetation index set, combined with the geographical zoning of high-altitude complex terrain areas and background field vegetation index thresholds, the method comprehensively considers the impact of different altitudes and terrains on vegetation phenology, thereby effectively solving the problem of insufficient adaptability of traditional threshold methods in high-altitude and complex terrain areas. At the same time, the calculation process of this method is simple and efficient, and therefore it also has strong potential for operational applications.

[0032] For ease of understanding, this embodiment of the invention uses observation data from the Fengyun-3 satellite (i.e., FY-3D satellite observation data) as an example to provide a specific implementation method for remote sensing monitoring of grassland greening period in high-altitude complex terrain areas. See [link to relevant documentation]. Figure 2The diagram shows the overall flowchart of a remote sensing monitoring method for grassland greening period in high-altitude complex terrain areas. The general idea of ​​this method is as follows: collect and download one year of FY-3D satellite observation data and preprocess it; construct a long-term NDVI dataset using FY-3D satellite observation data; reconstruct the long-term NDVI dataset; and use the reconstructed long-term NDVI dataset and surface elevation (DEM) data to invert the grassland greening period based on the dynamic threshold method.

[0033] This invention provides a specific implementation of step S102, which includes two steps: data download and data preprocessing, and construction of a long-time-series NDVI dataset. (1.1) Data download and data preprocessing: The FY-3D satellite observation data at 1km resolution, Level 1 (L1) products, including raw DN value data and geographic data, were obtained from the National Meteorological Satellite Center (http: / / satellite.nsmc.org.cn / ). The data format and file name codes are shown in Table 1.

[0034] Table 1. List of FY-3D MERSI-2 input data and related information

[0035] Satellite data preprocessing is a crucial step in ensuring the accuracy and reliability of grassland greening-up period inversion. This invention utilizes FY-3D satellite observation data and, through a series of preprocessing steps, generates high-quality, spatiotemporally consistent long-term satellite data, providing a data foundation for subsequent analysis. The specific preprocessing steps are as follows: (a1) Data screening and quality check; (a2) Radiometric correction and geometric correction.

[0036] (a1) Data screening and quality check: First, screen multi-temporal data covering high-altitude complex terrain areas from the downloaded FY-3D satellite observation data, remove low-quality images affected by adverse factors such as clouds and snow, and retain high-quality data with excellent observation conditions to ensure spatiotemporal continuity.

[0037] (a2) Radiometric and Geometric Correction: Following the "Guide to Data Use of the Medium Resolution Spectral Imager for the Fengyun-3 (Batch 02) Meteorological Satellite Ground Application System Engineering D Satellite" provided on the official website of the National Meteorological Satellite Center, radiometric calibration was performed on the raw DN value data to obtain the zenith apparent reflectance and brightness temperature data for the channel. Simultaneously, atmospheric correction was performed on the data using the 6S radiative transfer model to obtain the true surface reflectance of the satellite data. Finally, the GDAL tool was used to perform geometric correction on the satellite data to correct spatial positioning errors in the image, ensuring accurate alignment between the image and geographic coordinates.

[0038] (1.2) Construction of long-term NDVI dataset: The construction of long-term NDVI dataset is the basis for grassland greening period inversion. It is mainly achieved through the acquisition, processing and integration of time-phase satellite images. The specific process is as follows: (b1) Cloud and snow masking; (b2) NDVI data calculation; (b3) NDVI data synthesis.

[0039] (b1) Cloud and Snow Masking: Based on the following cloud monitoring method, cloud and snow monitoring is performed using preprocessed FY-3D satellite observation data to generate a cloud and snow mask, marking and removing pixels affected by cloud interference. Pixels affected by cloud interference should meet the following conditions: ; ; ; ; In the formula, This is the reflectivity value of the third channel of the FY-3D satellite. This refers to the brightness temperature values ​​of the 24 channels of the FY-3D satellite. This refers to the brightness temperature value of channel 25 of the FY-3D satellite. This is the brightness temperature value of channel 22 of the FY-3D satellite.

[0040] (b2) NDVI data calculation: Using the visible light (red light) and near-infrared bands of the FY-3D satellite, the NDVI value is calculated according to the following formula: ; in, This represents the reflectivity value of the third channel of the FY-3D satellite. This represents the reflectance value of the 4th channel in FY-3D.

[0041] (b3) NDVI Data Synthesis: To improve the quality of the long-term NDVI dataset and reduce the impact of clouds, shadows, and atmospheric interference, a 16-day maximum value synthesis method was further adopted to obtain a long-term NDVI dataset for high-altitude complex terrain areas. Specifically, the 16-day maximum value synthesis method uses NDVI data from each 16-day period as a time window, and calculates the maximum NDVI value of each pixel within each time window as the final NDVI value for that pixel. The calculation formula is shown below, thereby constructing a continuous and stable initial long-term NDVI dataset: NDVImax(i,j,t)=max{NDVI(i,j,t1),NDVI(i,j,t2),…,NDVI(i,j,tn)}; Where (i,j) is the pixel position, and t1,t2,...,tn is the observation time within the time window.

[0042] This invention provides a specific implementation of step S104 described above. Although the initial long-term NDVI dataset processed as described above has eliminated most of the noise, the synthesized initial long-term NDVI dataset still has some outliers. These irregular outliers affect the accuracy of extracting the grassland greening period. Therefore, in order to ensure that the initial long-term NDVI dataset can accurately reflect the seasonal change trend of vegetation, further time series reconstruction of the data is required. Considering that traditional single filtering methods do not adequately consider the prior features of long-term series and have limitations in reconstructing the initial long-term NDVI dataset, this invention uses a combination of Savizky-Golay (S-G) filtering and asymmetric Gaussian (AG) function fitting to reconstruct the initial long-term NDVI dataset, which can greatly reduce the reconstruction error.

[0043] In the specific implementation, the following operation is performed on the initial long-term vegetation index subset corresponding to each pixel: (2.1) Establish a sliding window for the initial long-term vegetation index subset corresponding to the pixel, perform local fitting on the vegetation index within the sliding window, move the sliding window and continue to perform local fitting on the vegetation index within the moved sliding window until each vegetation index within the initial long-term vegetation index subset corresponding to the pixel is locally fitted, and obtain the local long-term vegetation index subset corresponding to the pixel.

[0044] In one implementation, the initial long-term NDVI dataset is locally fitted based on the Savitzky-Golay filtering method, which is a data smoothing method that performs convolution operations on the data within the sliding window based on the least squares principle.

[0045] For specific implementation details, please refer to (c1) to (c3) below: (c1) The parameter values ​​of the convolution operation are fitted using multiple vegetation indices located within a sliding window. In one example, multiple vegetation indices can be randomly selected from the vegetation data located within the sliding window to fit the parameter values ​​of the convolution operation. The expression for the convolution operation is as follows: ; In the formula, The new vegetation index is the fitted value, and k is the normalization factor (usually related to the size of the sliding window). These are the weighting coefficients (i.e., parameter values). The original vegetation index is N, which is the size of the sliding window. N is an odd number, usually N=9.

[0046] (c2) Based on the fitted parameter values, perform a convolution operation on the vegetation indices located at the center point of the sliding window. In one example, the above expression is used to fuse each vegetation index within the sliding window to obtain a new vegetation index at the center point.

[0047] (c3) Use the new vegetation index obtained by convolution operation to replace the vegetation index located at the center point of the sliding window, so as to achieve local fitting of the vegetation index located within the sliding window.

[0048] Furthermore, the vegetation index at the center point of the current sliding window is the vegetation index corresponding to the pixel on the current date. Move the sliding window so that the vegetation index corresponding to the next date adjacent to the current date is located at the center point of the moved sliding window. Continue to perform local fitting on the vegetation index located in the moved sliding window according to the aforementioned (c1) to (c3). Repeat this process until each vegetation index in the initial long-term vegetation index subset corresponding to the pixel has been locally fitted. Then stop moving the sliding window and obtain the local long-term vegetation index subset corresponding to the pixel.

[0049] (2.2) Determine the global trend information represented by the initial long-term vegetation index subset corresponding to the pixel, and combine the asymmetric Gaussian function to perform global fitting on the initial long-term vegetation index subset to obtain the global long-term vegetation index subset corresponding to the pixel.

[0050] In one implementation, although the Savitzky-Golay filtering method can effectively remove local noise, a global trend may still exist in the data, especially when the data exhibits a skewed or asymmetrical distribution. In this case, an asymmetric Gaussian function can be used to fit the global trend of the data.

[0051] In practical implementation, the amplitude and standard deviation of the initial long-term vegetation index subset corresponding to the pixel are determined to represent global trend information. Then, using an asymmetric Gaussian function, based on the time information, amplitude, standard deviation of the vegetation index, and the center point of the asymmetric Gaussian function, a global fit is performed on the initial long-term vegetation index subset to obtain the global long-term vegetation index subset corresponding to the pixel. The expression for the asymmetric Gaussian function is shown below: ; In the formula, It is the new vegetation index after fitting, which is related to the independent variable. Related function values, It is the independent variable, here A represents time information (such as timestamps for annual, monthly, or daily data), and A is the amplitude. It is the center point of the Gaussian function. α is the standard deviation of the vegetation index, and α is a parameter that controls the asymmetry of the Gaussian function.

[0052] (2.3) The local long-term vegetation index subset and the global long-term vegetation index subset corresponding to the pixel are weighted and reconstructed to obtain the target long-term vegetation index subset corresponding to the pixel.

[0053] The local long-term vegetation index subset and the global long-term vegetation index subset are weighted and merged to reconstruct the data, obtaining the target long-term vegetation index subset. The formula for weighted reconstruction is shown below: ; In the formula, This is the final NDVI reconstruction result (i.e., the vegetation index within the subset of the target long-term vegetation index). It is the NDVI value after SG filtering (that is, the vegetation index within the subset of local long-term vegetation indices). It is the NDVI value after asymmetric Gaussian fitting (i.e., the vegetation index within the subset of global long-term vegetation indices). It is the weighting coefficient.

[0054] This invention provides a specific implementation of the aforementioned step S106.

[0055] The estimation of the greening period is first based on DEM data, which divides the high-altitude complex terrain area into geographical zones. Then, based on the dynamic threshold method, the greening period of each zone is estimated using the target long-term NDVI dataset. In order to improve the computational efficiency, this embodiment of the invention modifies the original dynamic threshold method and proposes a background field dynamic threshold greening period estimation method. This method mainly analyzes the vegetation index increment of the current NDVI (i.e., the vegetation index in the target long-term vegetation index set) and the background field NDVI (i.e., the background field vegetation index threshold) of different regions at the current time. If the vegetation index increment of the current NDVI and the background field NDVI is greater than the preset greening period threshold, it is determined that the pixel has entered the greening period at the current time.

[0056] For specific implementation details, please refer to (3.1) to (3.3) below: (3.1) Geographical division of high-altitude complex terrain areas. In one example, based on the wave distribution characteristics of high-altitude complex terrain areas (such as Tibet), the high-altitude complex terrain area can be divided into 5 geographical divisions according to the following division criteria. First geographical region: 0 <DEM<=3200; Second geographical region: 3200 <DEM<=4000; Third geographical region: 4000 <DEM<=4700; 4th geographical region: 4700 <DEM<=5300; Fifth geographic region: DEM>5300.

[0057] (3.2) Obtain the historical long-term vegetation index set corresponding to high-altitude complex terrain areas, and synthesize the historical long-term vegetation index set by rolling mean to obtain the background field vegetation index threshold. Taking Tibet as an example, according to the vegetation growth pattern in Tibet, the long-term average of 108 days (late April) to 190 days (early July) from 2012 to 2021 is selected for rolling mean synthesis to generate the background field vegetation index threshold during the vegetation greening period in Tibet. The formula is shown below: average( , , ..., ).

[0058] (3.3) Based on the greening period threshold corresponding to each geographical sub-region in the high-altitude complex terrain area, as well as the target long-term vegetation index set and background field vegetation index threshold corresponding to the high-altitude complex terrain area, the greening period of the high-altitude complex terrain area is estimated pixel by pixel, and the greening period estimation result corresponding to the high-altitude complex terrain area is obtained.

[0059] In one example, the following operations (d1) to (d2) can be performed on the subset of the target long-term vegetation indices corresponding to each pixel: (d1) Determine the vegetation index increase relative to the background field vegetation index threshold for each vegetation index within the target long-term vegetation index subset corresponding to the pixel. In one example, for the vegetation index of any date within the target long-term vegetation index subset corresponding to the pixel, the difference between the vegetation index of that date and the background field vegetation index threshold can be calculated, and the ratio of that difference to the background field vegetation index threshold can be calculated to obtain the vegetation index increase relative to the background field vegetation index threshold for that date.

[0060] (d2) Determine the date when the pixel enters the greening period based on the increase of each vegetation index and the greening period threshold corresponding to the geographic region where the pixel is located. In one example, the date corresponding to the increase of each vegetation index can be used to determine whether the increase of each vegetation index reaches the greening period threshold corresponding to the geographic region where the pixel is located; the date corresponding to the increase of the vegetation index when the determination result is yes for the first time is determined as the date when the pixel enters the greening period.

[0061] In a specific embodiment, according to the divided geographical regions, the green-up period is calculated according to different geographical regions, where: for the region with 0 < DEM <= 3200, this region is basically forest visually and does not participate in the calculation, or the green-up period is set to the minimum value, and this part will be masked by the forest mask file later without affecting the calculation; for the region with 3200 < DEM <= 4000, analyze the values of the current long-term NDVI at each moment and compare them with the background field NDVI. If the increase reaches 20%, it is considered that the green-up period starts on that day; for the region with 4000 < DEM <= 4700, if the increase reaches 15%, it is considered that the green-up period starts on that day; for the region with 4700 < DEM <= 5300, if the increase reaches 10%, it is considered that the green-up period starts on that day; for the region with > 5300, if the increase reaches 5%, it is considered that the green-up period starts on that day.

[0062] Furthermore, the green-up period results can be corrected for different regions respectively to obtain the final estimated result of the green-up period.

[0063] The embodiment of the present invention addresses the technical problems in the remote sensing monitoring of the grassland green-up period in high-altitude complex terrain regions, and proposes a method for retrieving the grassland green-up period based on FY-3D satellite observation data. This method makes full use of the advantages of the high temporal resolution of the FY-3D satellite, constructs a long-term normalized difference vegetation index (NDVI) dataset, and combines digital elevation model (DEM) data, comprehensively considering the influence of terrain factors on vegetation phenology. By using the dynamic threshold method, the long-term NDVI dataset is deeply analyzed to dynamically determine the key thresholds reflecting the grassland green-up characteristics, thus effectively solving the problem of insufficient adaptability of traditional threshold methods in complex terrain and different vegetation types. At the same time, the calculation process of this method is simple and efficient, so it also has strong potential for business applications.

[0064] For example, through the above method, the predicted results of the green-up period in Tibet in 2021 are calculated. From the results, it can be seen that there is an obvious trend of earlier in the east and later in the west in the grassland green-up period in Tibet. From the southeast to the northwest, the start time of the green-up period is gradually postponed. The average green-up period of the alpine grassland in Tibet is concentrated in the middle and late May. The green-up period in the east is mainly from April to May, while in the west it is mainly in June. This is mainly affected by the altitude. In regions with higher latitudes or altitudes, the green-up period of the pasture is later.

[0065] At the same time, the observed data of the forage phenology in Tibet provided by 3 pasture stations are used as the true values, and 30 sample points in 10 years from 2012 to 2021 are selected. The correlation coefficient R2, absolute error (Bias), and root mean square error RMSE are used to verify the accuracy of the green-up period retrieved by the satellite. See Figure 3The diagram shows a schematic of the accuracy verification results: As can be seen from the figure, the results of the greening period retrieved by the satellite are quite consistent with the data from the ground stations, with a correlation coefficient greater than 0.7. Therefore, the two have a good correlation. The greening period extracted based on this method can accurately retrieve the spatiotemporal variation characteristics of the phenological period of the alpine grassland in the Tibetan region.

[0066] Based on the foregoing embodiments, this invention provides a remote sensing monitoring device for grassland greening-up period in high-altitude complex terrain areas. (See also...) Figure 4 The diagram shows a remote sensing monitoring device for grassland greening in high-altitude, complex terrain areas. The device mainly includes the following parts: The vegetation index determination module 402 is used to preprocess satellite remote sensing images corresponding to high-altitude complex terrain areas and construct an initial long-term vegetation index set corresponding to high-altitude complex terrain areas based on the preprocessed satellite remote sensing images. The vegetation index reconstruction module 404 is used to perform local and global fitting on the initial long-term vegetation index set to reconstruct the initial long-term vegetation index set and obtain the target long-term vegetation index set corresponding to the high-altitude complex terrain area. The greening period estimation module 406 is used to perform geographical partitioning of high-altitude complex terrain areas, and determine the greening period estimation results corresponding to high-altitude complex terrain areas based on the target long-term vegetation index set, background field vegetation index threshold and geographical partitioning.

[0067] The remote sensing monitoring device for grassland greening period in high-altitude complex terrain areas provided in this invention constructs long-term vegetation indices for high-altitude complex terrain areas, and reconstructs the long-term vegetation index set through local fitting and global fitting. Using the reconstructed target long-term vegetation index set, combined with the geographical zoning of high-altitude complex terrain areas and background field vegetation index thresholds, the device comprehensively considers the influence of different altitudes and terrains on vegetation phenology, thereby effectively solving the problem of insufficient adaptability of traditional threshold methods in high-altitude and complex terrain areas. At the same time, the calculation process of this device is simple and efficient, and therefore it also has strong potential for operational applications.

[0068] In one implementation, the initial long-term vegetation index set includes a subset of the initial long-term vegetation indices corresponding to each pixel in a high-altitude complex terrain area; the vegetation index reconstruction module 404 includes: The local fitting unit is used to: establish a sliding window for the initial long-term vegetation index subset corresponding to the pixel, perform local fitting on the vegetation index located in the sliding window, move the sliding window and continue to perform local fitting on the vegetation index located in the moved sliding window until each vegetation index in the initial long-term vegetation index subset corresponding to the pixel is locally fitted, thereby obtaining the local long-term vegetation index subset corresponding to the pixel. The global fitting unit is used to: determine the global trend information represented by the initial long-term vegetation index subset corresponding to the pixel, and combine the asymmetric Gaussian function to perform global fitting on the initial long-term vegetation index subset to obtain the global long-term vegetation index subset corresponding to the pixel. The weighted reconstruction unit is used to: perform weighted reconstruction of the local long-term vegetation index subset and the global long-term vegetation index subset corresponding to the pixel to obtain the target long-term vegetation index subset corresponding to the pixel.

[0069] In one implementation, the local fitting unit is specifically used for: The parameter values ​​of the convolution operation are fitted using multiple vegetation indices located within a sliding window; Based on the fitted parameter values, a convolution operation is performed on the vegetation index located at the center point of the sliding window. By using the new vegetation index obtained from convolution operations to replace the vegetation index located at the center point of the sliding window, local fitting of the vegetation index located within the sliding window can be achieved.

[0070] In one implementation, the global fitting unit is specifically used for: Determine the amplitude and standard deviation of the vegetation index corresponding to the initial long-term vegetation index subset of the pixel, so as to characterize the global trend information through the amplitude and standard deviation of the vegetation index; By using an asymmetric Gaussian function, based on the time information, amplitude, standard deviation of vegetation index, and center point of the asymmetric Gaussian function corresponding to the initial long-term vegetation index subset, a global fitting is performed on the initial long-term vegetation index subset to obtain the global long-term vegetation index subset corresponding to the pixel.

[0071] In one implementation, the greening period estimation module 406 includes: Background field synthesis unit is used to: obtain the historical long-term vegetation index set corresponding to high-altitude complex terrain area, perform rolling mean synthesis on the historical long-term vegetation index set, and obtain the background field vegetation index threshold. The greening period estimation unit is used to: estimate the greening period of each pixel in the high-altitude complex terrain area based on the greening period threshold corresponding to each geographic sub-region within the high-altitude complex terrain area, as well as the target long-term vegetation index set and background field vegetation index threshold corresponding to the high-altitude complex terrain area, and obtain the greening period estimation result corresponding to the high-altitude complex terrain area; wherein, the greening period estimation result includes the date on which each pixel in the high-altitude complex terrain area enters the greening period.

[0072] In one implementation, the target long-term vegetation index set includes a subset of the target long-term vegetation indices corresponding to each pixel in a high-altitude complex terrain area; the greening period estimation unit includes: The increment determination subunit is used to: determine the increment of each vegetation index in the target long-term vegetation index subset corresponding to the pixel, relative to the background field vegetation index threshold. The greening period estimation sub-unit is used to determine the date when a pixel enters the greening period based on the increase in each vegetation index and the greening period threshold corresponding to the geographic region where the pixel is located.

[0073] In one implementation, the greening period estimation subunit is specifically used for: Based on the date corresponding to the increase in each vegetation index, it is determined whether the increase in each vegetation index reaches the threshold of the greening period corresponding to the geographic region where the pixel is located. The date corresponding to the increase in vegetation index when the judgment result is first true is determined as the date when that pixel enters the greening period.

[0074] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0075] This invention provides an electronic device, specifically, the electronic device includes a processor and a storage device; the storage device stores a computer program, and the computer program, when run by the processor, executes the method described in any of the above embodiments.

[0076] Figure 5 The present invention provides a schematic diagram of the structure of an electronic device 100, which includes a processor 50, a memory 51, a bus 52 and a communication interface 53. The processor 50, the communication interface 53 and the memory 51 are connected through the bus 52. The processor 50 is used to execute executable modules, such as computer programs, stored in the memory 51.

[0077] The memory 51 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 53 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0078] Bus 52 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 5 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0079] The memory 51 is used to store programs. After receiving an execution instruction, the processor 50 executes the programs. The method executed by the device for defining the flow process disclosed in any of the foregoing embodiments of the present invention can be applied to the processor 50 or implemented by the processor 50.

[0080] Processor 50 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 50 or by instructions in software form. Processor 50 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 51. The processor 50 reads the information in memory 51 and, in conjunction with its hardware, completes the steps of the above method.

[0081] The computer program product of the readable storage medium provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, please refer to the foregoing method embodiments, which will not be repeated here.

[0082] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0083] 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 it. 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 technical scope 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. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A remote sensing monitoring method for grassland greening period in high altitude complex terrain areas, characterized in that: include: Preprocessing the satellite remote sensing images corresponding to the high-altitude complex terrain area, and constructing an initial long-time series vegetation index set corresponding to the high-altitude complex terrain area based on the preprocessed satellite remote sensing images; Performing local fitting and global fitting on the initial long-term vegetation index set respectively to reconstruct the initial long-term vegetation index set to obtain a target long-term vegetation index set corresponding to the high-altitude complex terrain area; Geographically partitioning the high-altitude complex terrain area, and determining the greening period estimation result corresponding to the high-altitude complex terrain area based on the target long-term vegetation index set corresponding to the high-altitude complex terrain area, the background field vegetation index threshold and the geographical partitioning; The initial long-term vegetation index set includes an initial long-term vegetation index subset corresponding to each pixel in the high-altitude complex terrain area; the initial long-term vegetation index set is locally fitted and globally fitted respectively to reconstruct the initial long-term vegetation index set to obtain a target long-term vegetation index set corresponding to the high-altitude complex terrain area, including: The following operations are performed on the initial long-term vegetation index subset corresponding to each pixel: A sliding window is established for the initial long-term vegetation index subset corresponding to the pixel, and the vegetation index located in the sliding window is locally fitted. The sliding window is moved and the vegetation index located in the moved sliding window is continuously locally fitted until each vegetation index in the initial long-term vegetation index subset corresponding to the pixel is locally fitted to obtain the local long-term vegetation index subset corresponding to the pixel; And, determining the global trend information represented by the initial long-term vegetation index subset corresponding to the pixel, and performing global fitting on the initial long-term vegetation index subset in combination with an asymmetric Gaussian function to obtain the global long-term vegetation index subset corresponding to the pixel; The local long-term vegetation index subset and the global long-term vegetation index subset corresponding to the pixel are weighted reconstructed to obtain the target long-term vegetation index subset corresponding to the pixel.

2. The remote sensing monitoring method for grassland greening period in high altitude complex terrain area according to claim 1 is characterized in that: Performing local fitting on the vegetation index within the sliding window includes: Using a plurality of vegetation indices within the sliding window, fitting the parameter values ​​of the convolution operation; Based on the parameter value obtained by fitting, a convolution operation is performed on the vegetation index located at the center point of the sliding window; The new vegetation index obtained by the convolution operation is used to replace the vegetation index at the center point of the sliding window, so as to achieve local fitting of the vegetation index within the sliding window.

3. The remote sensing monitoring method for grassland greening period in high altitude complex terrain area according to claim 1 is characterized in that: Determine the global trend information represented by the initial long-term vegetation index subset corresponding to the pixel, and perform global fitting on the initial long-term vegetation index subset in combination with an asymmetric Gaussian function to obtain the global long-term vegetation index subset corresponding to the pixel, including: Determine the amplitude and the vegetation index standard deviation corresponding to the initial long-term vegetation index subset corresponding to the pixel, so as to characterize the global trend information through the amplitude and the vegetation index standard deviation; Through an asymmetric Gaussian function, based on the time information corresponding to the initial long-term vegetation index subset, the amplitude, the vegetation index standard deviation and the center point of the asymmetric Gaussian function, the initial long-term vegetation index subset is globally fitted to obtain the global long-term vegetation index subset corresponding to the pixel.

4. The remote sensing monitoring method for grassland greening period in high altitude complex terrain area according to claim 1 is characterized in that: Based on the target long-term vegetation index set corresponding to the high-altitude complex terrain area, the background field vegetation index threshold and the geographical partition, determining the greening period estimation result corresponding to the high-altitude complex terrain area, including: Acquire a historical long-term vegetation index set corresponding to the high-altitude complex terrain area, perform rolling mean synthesis on the historical long-term vegetation index set, and obtain a background field vegetation index threshold; According to the greening period threshold corresponding to each geographical division in the high-altitude complex terrain area, as well as the target long-term vegetation index set and the background field vegetation index threshold corresponding to the high-altitude complex terrain area, the greening period of the high-altitude complex terrain area is estimated pixel by pixel to obtain the greening period estimation result corresponding to the high-altitude complex terrain area; wherein the greening period estimation result includes the date when each pixel in the high-altitude complex terrain area enters the greening period.

5. The remote sensing monitoring method for grassland greening period in high altitude complex terrain area according to claim 4 is characterized in that: The target long-term vegetation index set includes a target long-term vegetation index subset corresponding to each pixel in the high-altitude complex terrain area; according to the greening period threshold corresponding to each geographical sub-area in the high-altitude complex terrain area, and the target long-term vegetation index set and the background field vegetation index threshold corresponding to the high-altitude complex terrain area, the greening period of the high-altitude complex terrain area is estimated pixel by pixel, and the greening period estimation result corresponding to the high-altitude complex terrain area is obtained, including: The following operations are performed on the target long-term vegetation index subset corresponding to each pixel: Determine the vegetation index increase of each vegetation index in the target long-term vegetation index subset corresponding to the pixel relative to the background field vegetation index threshold; Based on the increase of each vegetation index and the greening period threshold value corresponding to the geographical division where the pixel is located, the date when the pixel enters the greening period is determined.

6. The remote sensing monitoring method for grassland greening period in high altitude complex terrain area according to claim 5 is characterized in that: Based on the increase of each vegetation index and the greening period threshold value corresponding to the geographical zone where the pixel is located, determining the date when the pixel enters the greening period, including: According to the date corresponding to each vegetation index increase, it is judged in turn whether each vegetation index increase reaches the greening period threshold value corresponding to the geographical division where the pixel is located; The date corresponding to the increase in the vegetation index when the judgment result is yes for the first time is determined as the date when the pixel enters the greening period.

7. A remote sensing monitoring device for grassland greening period in high altitude complex terrain areas, characterized in that: include: A vegetation index determination module is used to preprocess the satellite remote sensing images corresponding to the high-altitude complex terrain area, and construct an initial long-time series vegetation index set corresponding to the high-altitude complex terrain area based on the preprocessed satellite remote sensing images; A vegetation index reconstruction module is used to perform local fitting and global fitting on the initial long-time series vegetation index set respectively, so as to reconstruct the initial long-time series vegetation index set and obtain a target long-time series vegetation index set corresponding to the high-altitude complex terrain area; A greening period estimation module is used to geographically partition the high-altitude complex terrain area, and determine the greening period estimation result corresponding to the high-altitude complex terrain area based on the target long-term vegetation index set corresponding to the high-altitude complex terrain area, the background field vegetation index threshold and the geographical partition; The initial long-term vegetation index set includes an initial long-term vegetation index subset corresponding to each pixel in the high-altitude complex terrain area; the vegetation index reconstruction module is specifically used for: The following operations are performed on the initial long-term vegetation index subset corresponding to each pixel: A sliding window is established for the initial long-term vegetation index subset corresponding to the pixel, and the vegetation index located in the sliding window is locally fitted. The sliding window is moved and the vegetation index located in the moved sliding window is continuously locally fitted until each vegetation index in the initial long-term vegetation index subset corresponding to the pixel is locally fitted to obtain the local long-term vegetation index subset corresponding to the pixel; And, determining the global trend information represented by the initial long-term vegetation index subset corresponding to the pixel, and performing global fitting on the initial long-term vegetation index subset in combination with an asymmetric Gaussian function to obtain the global long-term vegetation index subset corresponding to the pixel; The local long-term vegetation index subset and the global long-term vegetation index subset corresponding to the pixel are weighted reconstructed to obtain the target long-term vegetation index subset corresponding to the pixel.

8. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to implement the method according to any one of claims 1 to 6.

Images