Leaf area index time sequence processing method and system

Abstract

The invention relates to a leaf area index time sequence processing method, which comprises the following steps of: inputting a time sequence remote sensing image and suburb forest and economic forest classification data of the same region; calculating a normalized differential vegetation index of the research area and carrying out time sequence sorting; calculating vegetation coverage data of the research area and carrying out time sequence sorting; synthesizing to obtain monthly FVC time sequence data of the suburb forest region and the economic forest region; calculating a suburb forest leaf area index LAI in the research area and performing time sequence sorting; calculating the economic forest leaf area index LAI of the research area and performing time sequence sorting; performing seasonal decomposition on the suburb forest time sequence data after adaptive filtering; performing seasonal decomposition on the economic forest time series data after adaptive filtering; merging the suburb forest time sequence data and the economic forest time sequence data; and outputting to obtain final optimized data. The invention further relates to a leaf area index time sequence processing system. According to the invention, a purification LAI time sequence product which clearly represents long-term trend, mutation and gradual change signals of respective vegetation canopy structures can be output.

Description

Technical Field

[0001] This invention relates to a method and system for time-series processing of leaf area index. Background Technology

[0002] LAI (Leaf Area Index) is defined as the multiple of the total area of ​​plant leaves per unit land area to the total land area. It is a core indicator for quantifying vegetation canopy structure, characterizing vegetation growth, and simulating the exchange of matter and energy on the land surface. In ecological environment monitoring and land use management, LAI derived from remote sensing time-series data plays an irreplaceable role in assessing ecosystem productivity and understanding carbon cycle processes.

[0003] Shenzhen, as a highly urbanized and ecologically sensitive area, has a complex green space system, typically represented by evergreen broad-leaved suburban forests dominated by ecological functions and intensive orchards (such as litchi forests) dominated by economic functions. These two types of vegetation differ fundamentally in canopy structure and spatiotemporal dynamics: suburban forests have continuous, closed, multi-layered canopies, and their temporal variations in the tree area index (LAI) are mainly driven by natural phenological rhythms; while orchards exhibit discrete canopies shaped by artificial pruning, and their LAI dynamics are strongly coupled with the fruit trees' own phenology and periodic human interventions (such as pruning and fruit thinning). Therefore, accurately reversing and analyzing the inherent patterns of their LAI time-series curves is crucial for accurately assessing the structure and function of the urban ecosystem.

[0004] Currently, research on time-series analysis of remotely sensed vegetation area (LAI) suffers from two common limitations. At the parameter inversion level, mainstream approaches rely on single, general inversion models, such as those based on neural networks, lookup tables, or empirical regression. These models are typically built from training data of mixed vegetation types, failing to fully consider the unique influence of specific vegetation types' canopy structure, leaf tilt angle distribution, and other characteristics on radiative transfer processes, leading to systematic biases in inversion results specific to particular vegetation types. At the time-series preprocessing level, the most approximate approach is to directly apply uniform smoothing and decomposition processing (e.g., using SG filtering or STL decomposition) to the LAI time series obtained from the general model inversion, followed by inputting it into algorithms like DBEST for breakpoint detection. While this approach achieves basic processing, its preprocessing stage also ignores the heterogeneity of the time-series signal caused by both inversion model bias and differences in vegetation dynamics. A typical drawback is that the general inversion model may fail to accurately capture the sudden drop and recovery process of LAI caused by pruning in orchards, while the uniform filtering algorithm may misinterpret this real management signal as noise and smooth it out, or confuse the unique seasonal patterns of different vegetation types.

[0005] The core flaw of existing technical solutions lies in their "dual universality" paradigm: that is, they use a unified model in both the parameter inversion and time-series preprocessing stages, ignoring the essential differences in biophysical characteristics and dynamic mechanisms among different vegetation functional types. Specifically: 1) The inversion model is disconnected from the physical mechanism: General LAI inversion models (such as machine learning or empirical regression based on global training sets) fail to be optimized for specific vegetation canopy structures (such as closed canopy of suburban forests and discrete canopy of orchards), resulting in systematic bias in the inversion results, which makes the physical meaning of subsequent time series analysis weak.

[0006] 2) Mismatch between preprocessing methods and time-series dynamics: In the preprocessing stage, the use of fixed-parameter filtering and fixed-period decomposition methods cannot simultaneously solve the complex noise and mode confusion problems caused by the amplification of inversion bias and inherent dynamic differences (such as high-frequency artificial disturbances), making it difficult to accurately extract their true trends and abrupt change signals. Summary of the Invention

[0007] In view of this, it is necessary to provide a time-series processing method and system for leaf area index, which not only realizes the paradigm shift from "single model fitting" to "precise classification processing", but also provides an unprecedented full-chain technical solution for dynamic monitoring of structural parameters and analysis of driving mechanisms of heterogeneous vegetation systems in urban areas by consolidating the foundation of inversion.

[0008] This invention provides a time-series processing method for leaf area index (LAI), comprising the following steps: S1, inputting time-series remote sensing images of the same region, and inputting classification data of suburban forests and economic forests; S2, calculating the Normalized Difference Vegetation Index (NDVI) of the study area based on the input time-series remote sensing images and performing time-series sorting; S3, calculating the vegetation cover data (FVC) of the study area using the time-series NDVI data and performing time-series sorting; S4, calculating the leaf area index (LAI) of the suburban forests in the study area using the time-series NDVI data and FVC data and performing time-series sorting; S5, processing the time-series NDVI data and FVC data... The process involves: S6, calculating the Leaf Area Index (LAI) of economic forests in the study area and ranking it temporally; S7, performing adaptive filtering on the ranked suburban forest time-series data; S8, performing seasonal decomposition based on a fixed period on the filtered suburban forest time-series data; S9, performing phenological-guided adaptive seasonal decomposition on the filtered economic forest time-series data; S10, merging the suburban forest time-series data obtained from step S7 and the economic forest time-series data obtained from step S9; and S11, outputting the merged time-series data to obtain the final optimized data.

[0009] Preferably, step S2 includes: The Normalized Differential Vegetation Index (NDVI) of the study area was calculated using the input time-series Sentinel-2 data and then sorted by time series. The calculation formula is as follows:

[0010] in, Represents the normalized difference vegetation index; This represents the brightness value of the red band in Sentinel-2 data; This represents the brightness value of the near-infrared spectral band in the Sentinel-2 data.

[0011] Preferably, step S3 includes: The vegetation cover (FVC) data of the study area was calculated using time-series NDVI data and then sorted temporally. The calculation formula is as follows:

[0012] in, Represents a pixel with absolutely no vegetation cover. value; Represents pixels completely covered by vegetation value.

[0013] Preferably, step S4 includes: The Range Area Index (LAI) of the study area was calculated and time-series ranked using temporal NDVI and FVC data. The calculation formula is as follows: .

[0014] Preferably, step S5 includes: The LAI (Location Area Index) of economic forests in the study area was calculated and ranked temporally using time-series NDVI and FVC data. The calculation formula is as follows: .

[0015] Preferably, step S6 includes: For suburban forest time series data Perform adaptive Savitzky-Golay filtering:

[0016] in, The normalization coefficient is the sum of all convolution coefficients within the window. , For the first The convolution coefficients of a point are determined by the order of the polynomial fitting; these coefficients can be generated by specifying the window length and the order. It is half the width of a window.

[0017] Preferably, step S7 includes: For suburban forest time series data A seasonal decomposition based on a fixed period is performed, and the decomposition algorithm is as follows:

[0018] in, It is a trend item; It is a seasonal item; It is the residual term.

[0019] Preferably, step S8 includes: For time series data of economic forests Perform Whittaker filtering; the filtering algorithm is as follows:

[0020] in, To find the fitted value, minimize the objective function. get; For smoothing parameters; It is a d-order difference operator.

[0021] Preferably, step S9 includes: Step S91: Detect the dominant cycle using continuous small transformations. If there is a significant non-12-month cycle, adjust the decomposition basis according to the detection cycle. Step S92: Use BFAST-like methods for iterative decomposition. The decomposition algorithm is as follows:

[0022] in, It is a trend item; It is a seasonal item; It is the residual term; the BFAST iterative decomposition initializes the seasonal components with a period T. Iterate until convergence; Step S93, combine the agricultural calendar to analyze the residual items. The significant drop in the value was identified.

[0023] Preferably, step S10 includes: Merge the two preprocessed groups Timing: .

[0024] This invention provides a time-series processing system for leaf area index (LAI). The system includes an input module, a sorting module, a filtering module, a decomposition module, a merging module, and an output module. The input module is used to input time-series remote sensing images of the same region and to input classification data for suburban forests and economic forests. The sorting module is used to calculate the Normalized Differential Vegetation Index (NDVI) of the study area based on the input time-series remote sensing images and to sort it time-series. The sorting module is also used to calculate the vegetation cover data (FVC) of the study area using the time-series NDVI data and to sort it time-series. Furthermore, the sorting module is used to calculate the leaf area index (LAI) of the suburban forests in the study area using the time-series NDVI data and FVC data and to sort it time-series. The study uses time-series NDVI and FVC data to calculate the Leaf Area Index (LAI) of economic forests in the study area and sort it temporally. A filtering module performs adaptive filtering on the sorted suburban forest time-series data. A decomposition module performs seasonal decomposition based on a fixed period on the filtered suburban forest time-series data. The filtering module also filters the sorted economic forest time-series data. The decomposition module further performs phenological-guided adaptive seasonal decomposition on the filtered economic forest time-series data. A merging module merges the suburban forest and economic forest time-series data obtained from the decomposition module. An output module outputs the merged time-series data to obtain the final optimized data.

[0025] The core breakthrough of this application lies in the following: First, by introducing nonlinear empirical models adapted to the canopy structure characteristics of suburban forests and economic forests respectively, accurate initial LAI values ​​are inverted, ensuring the physical accuracy of the basic data from the source. Then, based on this, differentiated adaptive filtering and phenological-guided decomposition are implemented for the temporal characteristics of the two types to extract their pure temporal "skeleton" information. This invention, by coupling differentiated inversion with differentiated preprocessing, completely solves the problems of model mismatch and information loss when processing heterogeneous vegetation types in a single process, providing a reliable LAI time-series product foundation for high-precision change detection and attribution analysis. This invention has the following beneficial effects: (1) This invention represents a fundamental leap from "appearance processing" to "mechanism fusion," enhancing the physical consistency and accuracy of the product. Existing technologies only perform post-processing smoothing and decomposition at the time-series level, failing to address the systematic errors at the source of the inversion. This invention pioneered a differentiated inversion model that matches the physical mechanism of vegetation canopy structure, designing dedicated nonlinear models for the saturation effect of closed canopies in suburban forests and the dynamic range of discrete canopies in orchards. This fundamentally corrects the inversion bias caused by the singular structural assumptions of general models, enabling the initial LAI time series to more realistically reflect the biophysical properties of vegetation, laying a reliable physical foundation for subsequent analysis, and solving the core pain point of "input is distortion."

[0026] (2) A differentiated pipeline integrating "inversion-preprocessing" was constructed, achieving minimization of information loss and maximization of signal fidelity. The existing "unified inversion + unified preprocessing" scheme will cause error propagation and information confusion. This invention integrates the concept of differentiation throughout the entire chain: first, an initial sequence with clear physical meaning is generated through a dedicated model, and then a preprocessing process is "tailor-made" for it (such as configuring conformal filtering for the smoothed suburban forest sequence after inversion, and configuring strong smoothing and adaptive decomposition for the orchard sequence with violent fluctuations after inversion). This differentiated design with front-end and back-end collaboration ensures that each processing link is highly adapted to the inherent characteristics of the vegetation type, and achieves maximum preservation of effective information and systematic suppression of noise in the process from the original remote sensing signal to the pure time series product.

[0027] (3) The time-series products significantly enhance their ability to indicate driving factors such as management activities and natural changes, expanding their application depth. Because the LAI time-series products produced by this invention possess both high physical accuracy and a high temporal signal-to-noise ratio, their value in interpreting the driving factors of vegetation dynamics far exceeds that of existing technologies. For example, abrupt drops in the LAI sequence of economic forests, accurately inverted and discriminatively preserved by the model, can be directly correlated with pruning or harvesting events; subtle trend changes in the LAI sequence of suburban forests can also more reliably indicate natural disturbances or environmental stresses. This enables subsequent applications such as breakpoint detection and trend analysis to yield more accurate and explanatory conclusions, providing stronger data support for decisions such as precision agriculture and ecological assessment.

[0028] In summary, this invention achieves a technological paradigm upgrade by deeply integrating the concept of differentiation into the entire process from inversion to preprocessing. Its advantages are reflected in the comprehensive improvement of product accuracy, information fidelity, and final application value. Attached Figure Description

[0029] Figure 1 This is a flowchart of the leaf area index time-series processing method of the present invention; Figure 2 This is a hardware architecture diagram of the leaf area index time-series processing system of the present invention. Detailed Implementation

[0030] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0031] See Figure 1 The diagram shown is a flowchart of a preferred embodiment of the leaf area index time-series processing method of the present invention.

[0032] Step S1: Input time-series remote sensing images of the same area, and input classification data for suburban forests and economic forests. Specifically: In this embodiment, time-series Sentinel-2 remote sensing images of the same region are input, covering at least 5 complete years in time and the spatial range covers the entire study area; classification data of suburban forests and economic forests are input, with the spatial range covering the entire study area.

[0033] Step S2: Based on the input temporal remote sensing images, calculate the Normalized Differential Vegetation Index (NDVI) for the study area and perform temporal sorting. Specifically: In this embodiment, the Normalized Differential Vegetation Index (NDVI) of the study area is calculated and sorted temporally using the input time-series Sentinel-2 data. The calculation formula is as follows:

[0034] in, Represents the normalized difference vegetation index; This represents the brightness value of the red band in Sentinel-2 data; This represents the brightness values ​​in the near-infrared spectral band of Sentinel-2 data, which are then synthesized monthly. Time series data .

[0035] Step S3: Calculate the vegetation cover (FVC) data for the study area using time-series NDVI data and perform time-series sorting. Specifically: In this embodiment, vegetation cover data (FVC) of the study area is calculated using time-series NDVI data and then sorted temporally. The calculation formula is as follows:

[0036] in, Represents a pixel with absolutely no vegetation cover. value; Represents pixels completely covered by vegetation Values; then, using the input data on the classification of suburban forests and economic forests, monthly FVC time-series data for suburban forest areas and economic forest areas are synthesized on a monthly basis. .

[0037] Step S4: Calculate the Leaf Area Index (LAI) of the suburban forests in the study area and rank it temporally using time-series NDVI and FVC data. Specifically: The Range Area Index (LAI) of the study area was calculated and time-series ordered using temporal NDVI and FVC data. The calculation formula is as follows:

[0038] Step S5: Calculate the Leaf Area Index (LAI) of economic forests in the study area using time-series NDVI and FVC data, and rank them temporally. Specifically: In this embodiment, the LAI of economic forests in the study area is calculated and sorted temporally using time-series NDVI and FVC data. The calculation formula is as follows: .

[0039] Step S6 involves adaptive filtering of the sorted suburban forest time-series data. Specifically: In this embodiment, the time series data of suburban forests are used. Adaptive Savitzky-Golay filtering (hereinafter referred to as SG filtering) is performed, and the filtering algorithm is as follows:

[0040] in, The normalization coefficient is the sum of all convolution coefficients within the window (i.e., ...). In actual calculations, no explicit settings are required because the convolution kernel is automatically normalized. For the first The convolution coefficients of the points are determined by the order of the polynomial fitting (usually 2 to 3). The coefficients can be generated by specifying the window length and the order. With a window width of half, this method allows for dynamic adjustment of the window size based on the sequence noise level. The window length can be adjusted from 5 to 9. Users can evaluate the results through trial and error. Users can calculate the ratio of the root mean square error (RMSE) of the filtered sequence to the standard deviation of the original sequence. If the ratio is greater than 0.8, the window can be appropriately increased to 7 or even 9.

[0041] Step S7 involves performing a seasonal decomposition based on a fixed period on the filtered suburban forest time-series data. Specifically: In this embodiment, the time series data of suburban forests are used. A seasonal decomposition based on a fixed period is performed, and the decomposition algorithm is as follows:

[0042] in, It is a trend item; It is a seasonal item; These are residual terms, and the calculation method for each term includes: Step S71, Trend Item The calculation requires fitting a trend line using LOESS. LOESS needs to have a smoothing parameter `span` (the local window span, typically 0.2-0.3 for medium- to long-term trends). For a sequence length N, the trend value... The weighted polynomial fit value (weights are a three-kernel function) is the spanN data points near point t. The fitting algorithm is as follows:

[0043]

[0044] Where h is the width of the local window (h = spanN). This is the kernel function.

[0045] Step S72, Seasonal Items The calculation method is as follows: (1) Detrending:

[0046] (2) Calculate the seasonal component for each month. (k=1,...,12), calculate the average of the detrended values ​​for this month across all years:

[0047] in, It is to satisfy data points The number of; It means that for all satisfying index Summation; These are the detrended data values; Data points The corresponding months; based on the relatively stable annual cycle of 12 months in subtropical evergreen broad-leaved forests, k=1,...,12 is calculated, and the typical seasonal components of each month are extracted through a moving average window; (3) Adjust the seasonal components so that their annual sum is 0:

[0048] (4) Expand to the full sequence (That is, the component value of t corresponding to the month).

[0049] Step S73, Residual Term The calculation method is as follows:

[0050] Among them, residual term Including noise and outliers, the residual terms need to be analyzed. application The criteria exclude significant outliers. The criterion assumes that the residuals follow a normal distribution, calculates the mean μ and standard deviation σ of the residual sequence, and eliminates those that satisfy the following criteria. Points (considered outliers) will be used to identify outliers. Replace with The corrected seasonal and trend components are then reconstructed into a purified sequence. .

[0051] Step S8 involves filtering the sorted time-series data of economic forests. Specifically: In this embodiment, time-series data of economic forests are used. Perform Whittaker filtering; the filtering algorithm is as follows:

[0052] in, To find the fitted value, minimize the objective function. get; For smoothing parameters, it is recommended to set it to [value]. To enhance smoothness; For a d-order difference operator, d is usually set to 2; by adjusting... It effectively suppresses high-frequency man-made disturbance noise.

[0053] Step S9 involves performing phenological-guided adaptive seasonal decomposition on the filtered economic forest time-series data. Specifically: In this embodiment, time-series data of economic forests are used. Perform phenology-guided adaptive seasonal decomposition, the decomposition steps of which include: Step S91: The dominant cycle is detected using the Continuous Wavelet Transform (CWT). CWT analysis involves selecting Morlet wavelets as the basis function and performing a continuous wavelet transform on the detrended sequence. The wavelet power spectrum of the cycle corresponding to scale s (cycle = scale × sampling interval) is calculated. By detecting the dominant cycle, if a significant non-12-month cycle exists (e.g., double-cropping due to management), the decomposition basis is adjusted according to the detected cycle. The principle for determining a significant cycle is: using a red noise background spectrum test, if the power of a certain cycle exceeds the 95% confidence level, it is considered a significant cycle. If the largest significant cycle is not 12 months (e.g., 6 months), then this cycle T is recorded as the seasonal decomposition basis.

[0054] Step S92: Use a BFAST (Breaks For Additive Season and Trend) class method for iterative decomposition. The decomposition algorithm is as follows:

[0055] in, It is a trend item; It is a seasonal item; It is the residual term; the BFAST iterative decomposition initializes the seasonal components with a period T (detection value or default 12). The iteration continues until convergence. The iterative decomposition steps are as follows: (1) Fitting trend: for Using LOESS (span=0.3) yields

[0056] (2) Detrending:

[0057] (3) Seasonal updates: for Aggregate according to period T (same as the suburban forest method) to obtain

[0058] (4) Calculate the changes: Stop when (ε=0.001) (5) Calculation results:

[0059]

[0060]

[0061] Step S93: Combine the agricultural calendar (such as known pruning and harvest times) to analyze the residual terms. Identify significant drops in the data: If If so, it indicates synchronization with agricultural events and is considered a "true signal" and retained; if This is considered noise removal. The final output is a purified sequence. .

[0062] Step S10: Merge the time-series data of suburban forests obtained in step S7 and the time-series data of economic forests obtained in step S9. Specifically: In this embodiment, the two preprocessed groups are merged. Timing:

[0063] Step S11: Output the merged time-series data to obtain the final optimized data. Specifically: In this embodiment, the final optimized data product is output. This data product can be directly used in subsequent research such as breakpoint detection and driving factor analysis of the DBEST algorithm.

[0064] See Figure 2 The diagram shown is a hardware architecture diagram of the leaf area index time-series processing system 10 of the present invention. The system includes: an input module 101, a sorting module 102, a filtering module 103, a decomposition module 104, a merging module 105, and an output module 106.

[0065] The input module 101 is used to input time-series remote sensing images of the same area, and to input classification data of suburban forests and economic forests. Specifically: In this embodiment, the input module 101 inputs time-series Sentinel-2 remote sensing images of the same region, covering at least 5 complete years in time and the entire research area in spatial scope; it also inputs classification data of suburban forests and economic forests, with the spatial scope covering the entire research area.

[0066] The sorting module 102 is used to calculate the Normalized Differential Vegetation Index (NDVI) of the study area and perform temporal sorting based on the input temporal remote sensing images. Specifically: In this embodiment, the sorting module 102 calculates the Normalized Difference Vegetation Index (NDVI) of the study area using the input time-series Sentinel-2 data and performs time-series sorting. The calculation formula is as follows:

[0067] in, Represents the normalized difference vegetation index; This represents the brightness value of the red band in Sentinel-2 data; This represents the brightness values ​​in the near-infrared spectral band of Sentinel-2 data, which are then synthesized monthly. Time series data .

[0068] The sorting module 102 is also used to calculate the vegetation cover data (FVC) of the study area using time-series NDVI data and perform time-series sorting. Specifically: In this embodiment, the sorting module 102 calculates the vegetation cover data (FVC) of the study area using time-series NDVI data and performs time-series sorting. The calculation formula is as follows:

[0069] in, Represents a pixel with absolutely no vegetation cover. value; Represents pixels completely covered by vegetation Values; then, using the input data on the classification of suburban forests and economic forests, monthly FVC time-series data for suburban forest areas and economic forest areas are synthesized on a monthly basis. .

[0070] The sorting module 102 is also used to calculate the Leaf Area Index (LAI) of the suburban forests in the study area and perform time-series sorting using time-series NDVI and FVC data. Specifically: In this embodiment, the sorting module 102 calculates the LAI (Local Area Area) of the study area using temporal NDVI and FVC data and performs temporal sorting. The calculation formula is as follows:

[0071] The sorting module 102 is also used to calculate the Leaf Area Index (LAI) of economic forests in the study area and perform time-series sorting using time-series NDVI and FVC data. Specifically: In this embodiment, the sorting module 102 calculates the LAI of economic forests in the study area and performs time-series sorting using temporal NDVI and FVC data. The calculation formula is as follows: .

[0072] The filtering module 103 is used to perform adaptive filtering on the sorted time-series data of suburban forests. Specifically: In this embodiment, the filtering module 103 is used for time-series data of suburban forests. Adaptive Savitzky-Golay filtering (hereinafter referred to as SG filtering) is performed, and the filtering algorithm is as follows:

[0073] in, The normalization coefficient is the sum of all convolution coefficients within the window (i.e., ...). In actual calculations, no explicit settings are required because the convolution kernel is automatically normalized. For the first The convolution coefficients of the points are determined by the order of the polynomial fitting (usually 2 to 3). The coefficients can be generated by specifying the window length and the order. With a window width of half, this method allows for dynamic adjustment of the window size based on the sequence noise level. The window length can be adjusted from 5 to 9. Users can evaluate the results through trial and error. Users can calculate the ratio of the root mean square error (RMSE) of the filtered sequence to the standard deviation of the original sequence. If the ratio is greater than 0.8, the window can be appropriately increased to 7 or even 9.

[0074] The decomposition module 104 is used to perform seasonal decomposition based on a fixed period on the filtered suburban forest time-series data. Specifically: In this embodiment, the decomposition module 104 targets the time-series data of suburban forests. A seasonal decomposition based on a fixed period is performed, and the decomposition algorithm is as follows:

[0075] in, It is a trend item; It is a seasonal item; These are residual terms, and the calculation of each term specifically includes: The decomposition module 104 uses the LOESS fitting trend line to calculate the trend term. For LOESS, a smoothing parameter `span` needs to be set (the local window span, typically 0.2 to 0.3 for medium- to long-term trends). For the sequence length N, the trend value... The weighted polynomial fit value (weights are a three-kernel function) is the spanN data points near point t. The fitting algorithm is as follows:

[0076]

[0077] Where h is the width of the local window (h = spanN). This is the kernel function.

[0078] The decomposition module 104 calculates seasonal items. Specifically, it includes: (1) Detrending:

[0079] (2) Calculate the seasonal component for each month. (k=1,...,12), calculate the average of the detrended values ​​for this month across all years:

[0080] in, It is to satisfy data points The number of; It means that for all satisfying index Summation; These are the detrended data values; Data points The corresponding months; based on the relatively stable annual cycle of 12 months in subtropical evergreen broad-leaved forests, k=1,...,12 is calculated, and the typical seasonal components of each month are extracted through a moving average window; (3) Adjust the seasonal components so that their annual sum is 0:

[0081] (4) Expand to the full sequence (That is, the component value of t corresponding to the month).

[0082] The decomposition module 104 calculates the residual term. Specifically, it includes:

[0083] Among them, residual term Including noise and outliers, the residual terms need to be analyzed. application The criteria exclude significant outliers. The criterion assumes that the residuals follow a normal distribution, calculates the mean μ and standard deviation σ of the residual sequence, and eliminates those that satisfy the following criteria. Points (considered outliers) will be used to identify outliers. Replace with The corrected seasonal and trend components are then reconstructed into a purified sequence. .

[0084] The filtering module 103 is also used to filter the sorted time-series data of economic forests. Specifically: In this embodiment, the filtering module 103 targets the time-series data of economic forests. Perform Whittaker filtering; the filtering algorithm is as follows:

[0085] in, To find the fitted value, minimize the objective function. get; For smoothing parameters, it is recommended to set it to [value]. To enhance smoothness; For a d-order difference operator, d is usually set to 2; by adjusting... It effectively suppresses high-frequency man-made disturbance noise.

[0086] The decomposition module 104 is also used to perform phenological-guided adaptive seasonal decomposition on the filtered economic forest time-series data. Specifically: In this embodiment, the decomposition module 104 targets the time-series data of economic forests. Perform phenology-guided adaptive seasonal decomposition, including: The decomposition module 104 uses Continuous Wavelet Transform (CWT) to detect the dominant cycle. CWT analysis involves selecting Morlet wavelets as the basis function and performing a continuous wavelet transform on the detrended sequence. The wavelet power spectrum of the cycle corresponding to scale s (cycle = scale × sampling interval) is calculated. By detecting the dominant cycle, if a significant non-12-month cycle exists (such as double-cropping due to management), the decomposition basis is adjusted according to the detected cycle. The principle for determining a significant cycle is: using a red noise background spectrum test, if the power of a certain cycle exceeds the 95% confidence level, it is considered a significant cycle. If the largest significant cycle is not 12 months (such as 6 months), then this cycle T is recorded as the seasonal decomposition basis.

[0087] The decomposition module 104 uses the BFAST (Breaks For Additive Season and Trend) class method for iterative decomposition:

[0088] in, It is a trend item; It is a seasonal item; It is the residual term; the BFAST iterative decomposition initializes the seasonal components with a period T (detection value or default 12). Iterate until convergence. The iterative decomposition includes: (1) Fitting trend: for Using LOESS (span=0.3) yields

[0089] (2) Detrending:

[0090] (3) Seasonal updates: for Aggregate according to period T (same as the suburban forest method) to obtain

[0091] (4) Calculate the changes: Stop when (ε=0.001) (5) Calculation results:

[0092]

[0093]

[0094] The decomposition module 104 combines the agricultural calendar (such as known pruning and harvest times) to analyze the residual terms. Identify significant drops in the data: If If so, it indicates synchronization with agricultural events and is considered a "true signal" and retained; if This is considered noise removal. The final output is a purified sequence. .

[0095] The merging module 105 is used to merge the time-series data of suburban forests and economic forests obtained by the decomposition module 104. Specifically: In this embodiment, the merging module 105 merges the two preprocessed sets. Timing:

[0096] The output module 106 is used to output the merged time-series data to obtain the final optimized data. Specifically: In this embodiment, the output module 106 outputs the final optimized data product. This data product can be directly used in subsequent research such as breakpoint detection and driving factor analysis of the DBEST algorithm.

[0097] This invention realizes a complete chain of differentiated LAI product production, from inversion source to temporal purification. Its core objective lies not only in subsequent temporal preprocessing, but more importantly, in constructing a differentiated inversion foundation that matches vegetation type: specifically, establishing dedicated nonlinear empirical models for suburban forests and economic forests to ensure the physical accuracy and type specificity of the initial LAI values ​​from the source. Based on this, this invention further designs a highly adapted differentiated preprocessing pipeline for these two types of LAI sequences with clear physical meaning but distinct temporal characteristics. For the inverted suburban forest LAI time series, shape-preserving smoothing and fixed-period decomposition are used to precisely preserve its natural growth rhythm; for the economic forest LAI time series, strong noise suppression and adaptive periodic decomposition are used, and agricultural calendars are integrated for signal discrimination to accurately capture management activity signals. Ultimately, by achieving the innovative coupling of "differentiated inversion" and "differentiated preprocessing", pure LAI time-series products that can clearly reflect the "structure-dynamic" relationship of each vegetation system are produced. This provides full-process technical support from the data source to advanced products for high-precision change detection and analysis of driving factors, and completely solves the key problem that a single process cannot meet the needs of heterogeneous vegetation type analysis.

[0098] Although the present invention has been described with reference to the present preferred embodiments, those skilled in the art should understand that the above preferred embodiments are only used to illustrate the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A time-series processing method for leaf area index, characterized in that, The method includes the following steps: S1, Input time-series remote sensing images of the same area, and input classification data of suburban forests and economic forests; S2, Based on the input time-series remote sensing images, calculate the Normalized Differential Vegetation Index (NDVI) of the study area and perform time-series sorting. S3, using time-series NDVI data, calculates the vegetation cover data (FVC) of the study area and sorts it by time series; S4. Using time-series NDVI and FVC data, the leaf area index (LAI) of the suburban forests in the study area was calculated and sorted over time. S5. Using time-series NDVI and FVC data, the leaf area index (LAI) of economic forests in the study area was calculated and sorted over time. S6, adaptive filtering is performed on the sorted suburban forest time series data; S7 performs seasonal decomposition based on a fixed period on the filtered suburban forest time series data; S8, filtering is performed on the sorted time-series data of economic forests; S9 performs phenological-guided adaptive seasonal decomposition on the filtered time-series data of economic forests; S10, merge the suburban forest time series data obtained from step S7 and the economic forest time series data obtained from step S9. S11 outputs the merged time series data to obtain the final optimized data.

2. The method as described in claim 1, characterized in that, Step S2 includes: The Normalized Differential Vegetation Index (NDVI) of the study area was calculated using the input time-series Sentinel-2 data and then sorted by time series. The calculation formula is as follows: in, Represents the normalized difference vegetation index; This represents the brightness value of the red band in Sentinel-2 data; This represents the brightness value of the near-infrared spectral band in the Sentinel-2 data.

3. The method as described in claim 2, characterized in that, Step S3 includes: The vegetation cover (FVC) data of the study area was calculated using time-series NDVI data and then sorted temporally. The calculation formula is as follows: in, Represents a pixel with absolutely no vegetation cover. value; Represents pixels completely covered by vegetation value.

4. The method as described in claim 3, characterized in that, Step S4 includes: The Range Area Index (LAI) of the study area was calculated and time-series ranked using temporal NDVI and FVC data. The calculation formula is as follows: 。 5. The method as described in claim 4, characterized in that, Step S5 includes: The LAI (Location Area Index) of economic forests in the study area was calculated and ranked temporally using time-series NDVI and FVC data. The calculation formula is as follows: 。 6. The method as described in claim 5, characterized in that, Step S6 includes: For suburban forest time series data Perform adaptive Savitzky-Golay filtering: in, The normalization coefficient is the sum of all convolution coefficients within the window. , For the first The convolution coefficients of a point are determined by the order of the polynomial fitting; these coefficients can be generated by specifying the window length and the order. It is half the width of a window.

7. The method as described in claim 6, characterized in that, Step S7 includes: For suburban forest time series data A seasonal decomposition based on a fixed period is performed, and the decomposition algorithm is as follows: in, It is a trend item; It is a seasonal item; It is the residual term.

8. The method as described in claim 7, characterized in that, Step S8 includes: For time series data of economic forests Perform Whittaker filtering; the filtering algorithm is as follows: in, To find the fitted value, minimize the objective function. get; For smoothing parameters; It is a d-order difference operator.

9. The method as described in claim 8, characterized in that, Step S9 includes: Step S91: Detect the dominant cycle using continuous small transformations. If there is a significant non-12-month cycle, adjust the decomposition basis according to the detection cycle. Step S92: Use BFAST-like methods for iterative decomposition. The decomposition algorithm is as follows: in, It is a trend item; It is a seasonal item; It is the residual term; the BFAST iterative decomposition initializes the seasonal components with a period T. Iterate until convergence; Step S93, combine the agricultural calendar to analyze the residual items. The significant drop in the value was identified.

10. The method as described in claim 9, characterized in that, Step S10 includes: Merge the two preprocessed groups Timing: 。 11. A time-series processing system for leaf area index, characterized in that, The system includes an input module, a sorting module, a filtering module, a decomposition module, a merging module, and an output module, wherein: The input module is used to input time-series remote sensing images of the same area and input classification data of suburban forests and economic forests. The sorting module is used to calculate the Normalized Differential Vegetation Index (NDVI) of the study area and sort it in time series based on the input time-series remote sensing images. The sorting module is also used to calculate the vegetation cover data (FVC) of the study area using time-series NDVI data and to sort it in time series. The sorting module is also used to calculate the Leaf Area Index (LAI) of the suburban forests in the study area and perform time-series sorting using time-series NDVI and FVC data. The sorting module is also used to calculate the leaf area index (LAI) of economic forests in the study area and perform time-series sorting using time-series NDVI and FVC data. The filtering module is used to perform adaptive filtering on the sorted suburban forest time series data. The decomposition module is used to perform seasonal decomposition based on a fixed period on the filtered suburban forest time series data. The filtering module is also used to filter the sorted time-series data of economic forests; The decomposition module is also used to perform phenological-guided adaptive seasonal decomposition on the filtered economic forest time series data; The merging module is used to merge the time-series data of suburban forests and economic forests obtained by the decomposition module. The output module is used to output the merged time series data to obtain the final optimized data.

Images