Forest vegetation coverage growth monitoring system and method based on big data

Abstract

The invention relates to the technical field of environmental monitoring, in particular to a forest vegetation coverage growth monitoring system and method based on big data, and the method comprises the steps: extracting vegetation patch chlorophyll dispersion, texture direction and near-infrared uniformity to generate a microstructure feature set, comparing homogeneity intensity, a reference value and an adjacent condition marker to generate a connected state identifier, and carrying out the analysis of the connected state identifier; and according to the soil moisture capacity and the illumination radiation quantity, abnormal items of chlorophyll reflectivity deviation degree and near-infrared band concentration degree are removed. According to the method, by analyzing the chlorophyll ratio of the sub-pixels and the neighborhood texture, the near-infrared uniformity is quantified, and the heterogeneity precision is improved. And the multi-dimensional homogeneity strength is established by standardized parameters. And establishing a communication criterion by fusing edges, terrains and homogeneity. And performing dynamic screen spectrum abnormity early warning. And integrating the water retention, illumination and correction data of the soil to establish a dynamic model, and improving the prediction temporal-spatial resolution.

Description

Technical Field

[0001] The present invention relates to the field of environmental monitoring technology, and in particular to a forest vegetation cover growth monitoring system and method based on big data. Background Art

[0002] The field of environmental monitoring technology encompasses a technical system for data collection, dynamic analysis, and early warning management of natural environmental elements. Its core focus is on ecosystem status assessment and trend prediction through multi-dimensional data fusion and intelligent analysis. Traditionally, this field has relied on remote sensing image interpretation, ground sensor network deployment, and manual sampling surveys. However, due to the spatial heterogeneity of vegetation growth parameters, data update lags, and insufficient processing capabilities for multi-source, heterogeneous data, it has struggled to meet the demands of refined monitoring. Current technologies focus on building cross-platform data collaboration mechanisms, combining time-series data analysis with spatial modeling to optimize monitoring accuracy and promote the evolution of environmental monitoring from static description to dynamic decision support.

[0003] Among them, a big data-based forest vegetation cover growth monitoring system integrates satellite remote sensing spectral data, meteorological station observation data, and topographic and geomorphological data to establish a quantitative correlation model between vegetation coverage and growth indicators. Based on a standardized preprocessing process for multi-source data, it extracts features of surface reflectance, leaf area index, and soil moisture parameters, and uses time series data aggregation methods to construct a dynamic baseline for vegetation growth. A distributed computing framework enables parallel processing of massive data and real-time detection of abnormal fluctuations, ultimately forming a spatial distribution map of coverage and a growth trend forecast report.

[0004] Existing technologies rely on remote sensing images and ground sensor data, which are insufficient in characterizing the internal microstructural heterogeneity of vegetation patches, resulting in limited accuracy in the assessment of the spatial distribution of growth parameters. Multi-source data fusion lacks a standardized conversion mechanism, and remote sensing and ground data are not compatible enough, which affects the horizontal comparability of coverage models. Time series data analysis focuses on static aggregation, does not establish a dynamic anomaly detection benchmark, and has a delayed response to environmental mutations. Spatial modeling relies on static geographic data and does not integrate patch edge contact and terrain slope differences, resulting in highly subjective connectivity judgments. The lack of correlation between the dynamic changes in canopy thermal radiation and soil water holding capacity makes it impossible to distinguish between natural growth and environmental stress factors, and the prediction results are easily interfered with. The data preprocessing process does not adequately extract near-infrared reflectance uniformity parameters, which limits the microscopic assessment capabilities of vegetation health and reduces the application value of monitoring results in ecological restoration decision-making. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the existing technology and propose a forest vegetation cover growth monitoring system and method based on big data.

[0006] In order to achieve the above objectives, the present invention adopts the following technical solution: a forest vegetation cover growth monitoring system based on big data, the system comprising:

[0007] The patch microstructure analysis module extracts the chlorophyll distribution dispersion of vegetation patches, calculates the texture direction range within the neighborhood based on the dispersion ratio, extracts the near-infrared uniformity value, and generates a patch microstructure feature set;

[0008] The homogeneity fusion module calls the chlorophyll discrete value, texture direction distribution range value, and near-infrared uniformity value in the patch microstructure feature set, performs normalization processing, and fuses them to generate a homogeneity intensity value;

[0009] The connectivity identification module combines the edge lengths and terrain differences between patches, compares the homogeneity strength values of the adjacent patches with the preset benchmark values and the geographical proximity conditions, marks the connectivity status, and generates a patch connectivity status identifier;

[0010] The spectral dynamic screening module extracts the change rate sequence of each type of spectral index, calculates the distribution deviation and concentration, and screens the sequences whose deviation exceeds the preset deviation range or whose concentration does not exceed the preset concentration threshold, thereby generating a set of dynamic abnormal spectral indicators;

[0011] The coverage dynamic integration module calls the patch connectivity status identifier and the dynamic abnormal spectral index set to extract soil water holding capacity and light radiation, eliminate the abnormal items of chlorophyll reflectance offset and near-infrared band concentration, and generate vegetation coverage dynamic monitoring results.

[0012] As a further solution of the present invention, the patch microstructure feature set includes the sub-pixel chlorophyll dispersion ratio, the main texture direction angular distribution range, and the near-infrared uniformity mean. The homogeneity intensity value is specifically the standardized discrete interval value, the angle consistency quantization value, and the uniformity relative grade value. The patch connectivity status identifier covers the edge contact strength threshold, slope difference tolerance, and homogeneity difference judgment threshold. The dynamic anomaly spectral indicator set includes the chlorophyll offset anomaly sequence, the canopy thermal radiation concentration anomaly sequence, and the near-infrared band dynamic anomaly sequence. The corrected coverage dynamic intensity value specifically refers to the fusion of canopy thermal radiation intensity, near-infrared band stability, and water holding capacity correlation correction factor.

[0013] As a further solution of the present invention, the patch microstructure analysis module includes a chlorophyll difference integration submodule, a texture distribution span submodule, and a reflection uniformity integration submodule;

[0014] The chlorophyll difference integration submodule detects the chlorophyll distribution data of sub-pixels within vegetation patches, analyzes the difference between the chlorophyll value of each sub-pixel and the overall mean of the patch, integrates the correlation between the difference value and the mean, and generates a dispersion-mean ratio;

[0015] The texture distribution span submodule extracts the main texture direction angle data of the pixel unit in the neighborhood, determines the distribution boundary of the main texture direction angle, evaluates the concentrated distribution span range of the angle value, and generates the main texture angle span;

[0016] The reflection uniformity comprehensive submodule calls the discreteness mean ratio and the main texture angle span, combines the distribution state of the near-infrared reflectivity in each sub-pixel unit, integrates the discreteness influencing factor and the texture span weight, evaluates the spatial consistency of the reflectivity within the patch, and generates a patch microstructure feature set.

[0017] As a further solution of the present invention, the homogeneous fusion module includes a parameter normalization submodule, an index conversion submodule, and a homogeneous intensity synthesis submodule;

[0018] The parameter normalization submodule calls the chlorophyll discrete values in the patch microstructure feature set, adjusts the distribution range of the discrete values through linear mapping according to the upper and lower limits of the preset standard interval, and generates a standard discrete coefficient;

[0019] An index conversion submodule extracts the texture direction distribution range value from the patch microstructure feature set, calculates the ratio coefficient between the direction distribution range and the theoretical maximum angle, and generates an angle consistency index by combining the angle concentration weight factor;

[0020] The homogeneity intensity synthesis submodule calls the standard discrete coefficient and angle consistency index, superimposes the near-infrared uniformity value in the plaque microstructure feature set, converts the uniformity value into a relative level according to the level segmentation rule, and integrates the weighted calculation results of each parameter to generate a homogeneity intensity value.

[0021] As a further solution of the present invention, the connectivity determination module includes a homogeneity difference calculation submodule, a connectivity benchmark comparison submodule, and a state identification generation submodule;

[0022] The homogeneity difference calculation submodule obtains the homogeneity strength values of the adjacent patches, calculates the absolute value of the difference between the two, combines the edge contact length between the patches and the terrain slope difference value, and superimposes the correction coefficient of the edge contact length to the slope difference to generate the homogeneity difference degree;

[0023] A connectivity benchmark comparison submodule calls the homogeneity difference degree, compares a preset connectivity benchmark value with a geographic proximity condition, divides the degree of deviation between the homogeneity difference degree and the benchmark value into discrete levels, and generates a connectivity judgment benchmark based on the spatial constraint range of the geographic proximity condition;

[0024] The status identification generation submodule selects the patch pairs that meet the benchmark value and the geographical proximity condition based on the connectivity judgment benchmark, matches the preset connectivity status classification rules according to the deviation degree level, and generates the patch connectivity status identification.

[0025] As a further solution of the present invention, the spectral dynamic screening module includes a sequence offset calculation submodule, an abnormal marker screening submodule, and an indicator set generation submodule;

[0026] The sequence offset calculation submodule extracts the continuous monitoring ratio sequence of chlorophyll reflectance change, canopy thermal radiation change, and near-infrared band change, calculates the distribution offset and concentration of each sequence, integrates the numerical relationship between the offset and concentration, and generates spectral distribution parameters;

[0027] The abnormal marking screening submodule calls the spectral distribution parameters, compares the distribution deviation with the preset deviation range, and the concentration with the preset concentration threshold, screens out sequences with deviations exceeding the range or concentrations not reaching the threshold, marks the corresponding sequence types, and generates abnormal spectral sequences;

[0028] The indicator set generation submodule integrates the chlorophyll reflectance, canopy thermal radiation, and near-infrared band abnormal data marked in the abnormal spectral sequence, stores them by sequence type, summarizes the associated parameters of all abnormal sequences, and generates a dynamic abnormal spectral indicator set.

[0029] As a further solution of the present invention, the coverage dynamic integration module includes an abnormal item elimination submodule, a dynamic strength fusion submodule, and a coverage conflict determination submodule;

[0030] The abnormal item removal submodule calls the chlorophyll reflectance deviation abnormal item and the near-infrared band concentration abnormal item in the dynamic abnormal spectral indicator set, filters the soil water holding capacity and light radiation data of the non-connected area, removes the monitoring units containing the above abnormal items, retains the remaining canopy thermal radiation and near-infrared band change data, and generates the filtered spectral data;

[0031] The dynamic intensity fusion submodule extracts the canopy thermal radiation change and near-infrared band change sequences based on the screened spectral data, calculates the mean of the fluctuation amplitudes of the two sequences, and superimposes the correction coefficient of soil water holding capacity to light radiation to generate the coverage dynamic intensity value;

[0032] The coverage conflict determination submodule calls the coverage dynamic intensity value and the soil water holding capacity data of the non-connected area in the patch connectivity status identification, compares the distribution range of the coverage dynamic intensity value with the matching trend of the soil water holding capacity, marks the conflicting areas and classifies and integrates them to generate dynamic monitoring results of vegetation coverage.

[0033] A method for monitoring forest vegetation cover growth based on big data is provided. The method is performed based on the above-mentioned forest vegetation cover growth monitoring system based on big data and comprises the following steps:

[0034] S1: Extract the chlorophyll distribution dispersion of vegetation patches, calculate the texture direction range in the neighborhood based on the dispersion ratio, extract the near-infrared uniformity value, and generate a patch microstructure feature set;

[0035] S2: calling the chlorophyll discrete value, texture direction distribution range value, and near-infrared uniformity value in the patch microstructure feature set, performing normalization processing, and fusing them to generate a homogeneity intensity value;

[0036] S3: combining the edge lengths and terrain differences between patches, based on the homogeneity strength values of the adjacent patches, comparing the preset benchmark values with the geographical proximity conditions, marking the connectivity status, and generating a patch connectivity status identifier;

[0037] S4: extracting the change rate sequence of each type of spectral index, calculating the distribution deviation and concentration, screening the sequences whose deviation exceeds the preset deviation range or whose concentration does not exceed the preset concentration threshold, and generating a dynamic abnormal spectral index set;

[0038] S5: Call the patch connectivity status identifier and dynamic abnormal spectral indicator set, extract soil water holding capacity and light radiation, remove abnormal items of chlorophyll reflectance offset and near-infrared band concentration, and generate dynamic monitoring results of vegetation coverage.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are:

[0040] In the present invention, the near-infrared reflectance uniformity is quantified by analyzing the ratio of sub-pixel chlorophyll dispersion to the mean value, combined with the distribution range of neighborhood texture directions, to improve the accuracy of characterizing internal heterogeneity of vegetation. The discrete values are standardized, the texture directions and uniformity parameters are converted, and multi-dimensional homogeneity intensity is constructed to enhance data fusion compatibility. The patch edge contact length and terrain slope differences are fused, and the absolute value of the homogeneity difference is combined to establish a spatial connectivity criterion to optimize the objectivity of ecosystem interaction analysis. Dynamic screening of spectral anomaly indicators, establishment of a real-time monitoring mechanism, and strengthening of early warning capabilities for environmental mutations. The soil water holding capacity, light radiation, and canopy thermal radiation and near-infrared changes after removing anomalies are integrated to construct a dynamic coverage intensity correction model to improve the spatiotemporal resolution and anti-interference ability of growth trend prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a system flow chart of the present invention;

[0042] Figure 2 This is a flowchart of obtaining the plaque microstructure analysis module of the present invention;

[0043] Figure 3 This is a flowchart of obtaining the homogeneous fusion module of the present invention;

[0044] Figure 4This is a flowchart of obtaining the connectivity determination module of the present invention;

[0045] Figure 5 This is a flowchart of the acquisition of the spectrum dynamic screening module of the present invention;

[0046] Figure 6 This is a flowchart of obtaining the dynamic integration module covered by the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0048] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0049] In the embodiments of the present invention, the terms "image" and "picture" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same. The terms "of," "corresponding," and "corresponding" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same.

[0050] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0051] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0052] See also Figure 1 The present invention provides a technical solution: a forest vegetation cover growth monitoring system based on big data, the system comprising:

[0053] The patch microstructure analysis module extracts the chlorophyll distribution dispersion of sub-pixel units within vegetation patches, calculates the ratio of the dispersion to the mean, and calculates the distribution range of the main texture direction angle within the neighborhood to obtain the mean value of the near-infrared reflectance uniformity parameter and generate a patch microstructure feature set;

[0054] The homogeneity fusion module calls the chlorophyll discrete value, texture direction distribution range value, and near-infrared uniformity value in the patch microstructure feature set, maps the discrete value to the standard interval, converts the texture direction range value into the angle consistency index, and converts the uniformity value into a relative level to generate a homogeneity intensity value;

[0055] The connectivity identification module combines the edge contact length between patches and the difference in terrain slope, calculates the absolute value of the homogeneity difference based on the homogeneity strength value of adjacent patches, compares the preset benchmark value with the geographical proximity condition, marks the connectivity status, and generates a patch connectivity status identifier;

[0056] The spectral dynamic screening module extracts the continuous monitoring ratio sequence of chlorophyll reflectance changes, canopy thermal radiation changes, and near-infrared band changes, calculates the distribution deviation and concentration of each sequence, marks the sequence with deviation exceeding the preset deviation range or concentration not exceeding the preset concentration threshold as an abnormal indicator, and generates a dynamic abnormal spectral indicator set;

[0057] The coverage dynamic integration module extracts the soil water holding capacity and light radiation of non-connected areas under the patch connectivity status mark, combines the dynamic abnormal spectral indicator set, eliminates the chlorophyll reflectance offset anomaly and the near-infrared band concentration anomaly, integrates the remaining canopy thermal radiation and near-infrared band changes, obtains the coverage dynamic intensity value, determines whether it conflicts with the soil water holding capacity, and generates vegetation coverage dynamic monitoring results.

[0058] The patch microstructure feature set includes the sub-pixel chlorophyll dispersion ratio, the main texture direction angular distribution range, and the near-infrared uniformity mean. The homogeneity intensity value is specifically the standardized discrete interval value, the angle consistency quantification value, and the uniformity relative grade value. The patch connectivity status identification covers the edge contact strength threshold, slope difference tolerance, and homogeneity difference judgment threshold. The dynamic anomaly spectral indicator set includes the chlorophyll offset anomaly sequence, the canopy thermal radiation concentration anomaly sequence, and the near-infrared band dynamic anomaly sequence. The corrected coverage dynamic intensity value specifically refers to the fusion of canopy thermal radiation intensity, near-infrared band stability, and water holding capacity correlation correction factor.

[0059] See also Figure 2 ,The patch microstructure analysis module includes the chlorophyll difference integration submodule, the texture distribution span submodule, and the reflection uniformity integration submodule;

[0060] The chlorophyll difference integration submodule detects the chlorophyll distribution data of sub-pixels within vegetation patches, analyzes the difference between the chlorophyll value of each sub-pixel and the overall mean of the patch, integrates the correlation between the difference value and the mean, and generates a dispersion-mean ratio;

[0061] The chlorophyll content data of each sub-pixel in the target vegetation patch was obtained through satellite imagery. Each sub-pixel covered 1 square meter. The chlorophyll value was calculated using the red and near-infrared band reflectance ratio method. For example, the reflectance ratios of the sub-pixels in a patch were 1.8, 2.1, and 1.5, respectively, corresponding to chlorophyll values of 40, 45, and 35 μg / cm 2 The calculated mean value of the plaque was 40 μg / cm 2 , perform the following operations on each sub-pixel: calculate the absolute difference between its chlorophyll value and the mean value, for example, the value of sub-pixel A is 45μg / cm 2 When the absolute difference is 5, the absolute differences of all sub-pixels are averaged to obtain the average absolute difference (e.g. 3.2 μg / cm 2 ), the mean absolute difference is divided by the mean to obtain the dispersion mean ratio (such as 3.2 / 40=0.08). If the ratio is greater than 0.1, it is determined to be a high-difference patch. The threshold interval is set according to historical data. The dispersion mean ratio of 0-0.05 is low difference, 0.05-0.1 is medium difference, and greater than 0.1 is high difference. The final output is the dispersion mean ratio.

[0062] The texture distribution span submodule extracts the main texture direction angle data of the pixel unit in the neighborhood, determines the distribution boundary of the main texture direction angle, evaluates the concentrated distribution span range of the angle value, and generates the main texture angle span;

[0063] Extract the main texture direction angle of the gray-level co-occurrence matrix for the 3×3 pixel window in the neighborhood. For example, the grayscale value of the pixels in a certain window is [120, 130, 125; 115, 128, 135; 110, 122, 140]. Calculate the contrast in the horizontal, vertical, and diagonal directions. Select the direction corresponding to the maximum contrast as the main texture direction. If the horizontal contrast is 50, the vertical is 60, and the diagonal is 55, then the main texture direction is the vertical direction (angle 90 degrees). Count the main texture angles of all windows in the neighborhood. For example, a certain spot The main texture angle in the block is [85,92,88,95,83]. The minimum angle is determined to be 83 degrees and the maximum angle is 95 degrees. The span range is calculated to be 12 degrees. Statistically, 80% of the angle values are concentrated in the range of 85-93 degrees. The span range accounts for (93-85) / 12=0.67. Multiply the span range by the proportion coefficient to obtain the main texture angle span (12×0.67=8 degrees). If the span is less than 5 degrees, it is judged as a high consistency texture, 5-10 degrees as medium consistency, and greater than 10 degrees as low consistency. Finally, the main texture angle span is output.

[0064] The reflection uniformity synthesis submodule uses the dispersion mean ratio and the main texture angle span, combines the distribution of near-infrared reflectivity in each sub-pixel unit, integrates the dispersion influencing factor and the texture span weight, evaluates the spatial consistency of the reflectivity within the patch, and generates a set of patch microstructure features;

[0065] Call the discreteness mean ratio (such as 0.08) and the main texture angle span (such as 8 degrees), set the discreteness influence factor to 0.6 and the texture span weight to 0.4, calculate the standard deviation of the reflectance of each sub-pixel in the near-infrared band (such as the standard deviation is 0.05), multiply the reflectance standard deviation by the discreteness mean ratio (0.05×0.08=0.004), and then add it to the value after multiplying the main texture angle span by the weight (8×0.4=3.2). The final comprehensive index is 0.004+3.2=3.204. If the comprehensive index is less than 2.5, it is judged to be high spatial consistency, 2.5-4.0 is medium consistency, and greater than 4.0 is low consistency. For example, the comprehensive index of a patch is 3.2, which is classified as medium consistency. The output includes a patch microstructure feature set containing the discreteness mean ratio, main texture angle span, and reflectance standard deviation.

[0066] See also Figure 3 ,The homogeneous fusion module includes the parameter normalization submodule, the ,index conversion submodule, and the homogeneous intensity synthesis submodule;

[0067] The parameter normalization submodule calls the chlorophyll discrete values in the patch microstructure feature set, adjusts the distribution range of the discrete values through linear mapping according to the upper and lower limits of the preset standard interval, and generates a standard discrete coefficient;

[0068] Receive the chlorophyll discrete value (for example, 0.08) in the patch microstructure feature set, preset the standard interval to [0,1], call the maximum value 0.15 and the minimum value 0.03 of the patch discrete value in the historical statistics, and adjust the discrete value to the standard interval through the linear mapping formula. For example, the mapping result corresponding to the current discrete value 0.08 is (0.08-0.03) / (0.15-0.03)=0.416. If the discrete value of a patch is 0.12, the mapping result is (0.12-0.03) / (0.15-0.03)=0.75. When the standard discrete coefficient exceeds 1, it is forced to be truncated to 1, and when it is less than 0, it is forced to be 0. The preset standard interval is determined based on the 95% quantile of the historical data of similar vegetation patches, and the standard discrete coefficient is finally output.

[0069] The index conversion submodule extracts the texture direction distribution range value from the patch microstructure feature set, calculates the ratio coefficient between the direction distribution range and the theoretical maximum angle, and combines it with the angle concentration weight factor to generate the angle consistency index;

[0070] The texture direction distribution range value (for example, 12 degrees) in the patch microstructure feature set is extracted, and the theoretical maximum angle is set to 180 degrees (because the texture direction angle is periodically distributed between 0 and 180 degrees). The calculated proportional coefficient is 12 / 180 = 0.067, and the angle concentration weight factor is set according to the span range ratio (such as the span range ratio of 0.67 in the previous example), weight factor = 1-span range ratio = 0.33. When the span range ratio is ≤0.5, the weight factor is 0.5, when it is 0.5-0.7, it is 0.3, and when it is >0.7, it is 0.1. For example, the span range ratio of 0.67 corresponds to a weight factor of 0.3. The angle consistency index = proportional coefficient × weight factor = 0.067 × 0.3 = 0.0201. If the span range of a patch is 8 degrees and the ratio is 0.6, it is calculated as (8 / 180) × 0.3 = 0.0133, and the angle consistency index is finally output.

[0071] The homogeneity intensity synthesis submodule uses the standard dispersion coefficient and angle consistency index to superimpose the near-infrared uniformity values in the patch microstructure feature set, converts the uniformity values into relative levels according to the grade segmentation rules, and generates a homogeneity intensity value by combining the weighted calculation results of the three parameters;

[0072] Call the standard dispersion coefficient 0.416 and the angle consistency index 0.0201, extract the near-infrared uniformity value (for example, 0.05), the uniformity level segmentation rule is: 0-0.02 is level 1 (assigned as 0.2), 0.02-0.04 is level 2 (0.5), 0.04-0.06 is level 3 (0.8), the current uniformity 0.05 corresponds to level 3 (0.8), set the dispersion coefficient weight 0.5, the angle index weight 0.3, and the uniformity weight 0.2, and calculate the uniformity. The homogeneity strength value = 0.416×0.5+0.0201×0.3+0.8×0.2=0.208+0.00603+0.16=0.374. If the parameters of a certain patch are the standard dispersion coefficient of 0.75, the angle consistency of 0.0133, and the uniformity level of 2 (0.5), then the calculation result = 0.75×0.5+0.0133×0.3+0.5×0.2=0.375+0.004+0.1=0.479, and the final output is the homogeneity strength value.

[0073] See also Figure 4 ,The connectivity discrimination module includes a homogeneity difference calculation submodule, a connectivity benchmark comparison submodule, and a state identifier generation submodule;

[0074] The homogeneity difference calculation submodule obtains the homogeneity strength values of adjacent patches, calculates the absolute value of the difference between the two, combines the edge contact length between patches with the terrain slope difference value, and superimposes the correction coefficient of the edge contact length to the slope difference to generate the homogeneity difference degree;

[0075] Receive the homogeneity strength values of adjacent patches A and B (for example, A is 0.4 and B is 0.6), calculate the absolute difference as |0.4-0.6|=0.2, extract the edge contact length between patches (such as 50 meters) and the terrain slope difference value (such as patch A has a slope of 5 degrees and patch B has a slope of 15 degrees, the difference is 10 degrees), set the correction coefficient of edge contact length to slope difference as 0.1 for every 10 meters increase in contact length, and the current correction coefficient for 50 meters is 0.5, and the slope difference is converted to The value is multiplied by the correction coefficient (10×0.5=5), and the absolute homogeneity difference of 0.2 is added to the corrected slope difference of 5 (0.2+5=5.2). If the slope difference exceeds 20 degrees, the maximum correction value is limited to 2. For example, when the slope difference is 25 degrees and the contact length is 30 meters, the correction coefficient is 0.3, and the corrected slope difference is 25×0.3=7.5. However, when the slope difference exceeds 20 degrees, the slope difference correction value is forced to 2, then the total difference is 0.2+2=2.2, and the homogeneity difference is finally output.

[0076] The connectivity benchmark comparison submodule calls the homogeneity difference degree to compare the preset connectivity benchmark value with the geographical proximity condition, divides the degree of deviation between the homogeneity difference degree and the benchmark value into discrete levels, and generates a connectivity judgment benchmark based on the spatial constraint range of the geographical proximity condition.

[0077] The homogeneity difference (e.g., 5.2) is called, and the connectivity benchmark value is preset to 3.0 (based on the fact that 80% of the patch pairs in the historical data have a difference of less than 3.0). The deviation degree from the benchmark value is calculated as 5.2-3.0=2.2. The deviation degree is divided into three levels: ≤1.0 is level 1 (low deviation), 1.0-2.0 is level 2 (medium deviation), and >2.0 is level 3 (high deviation). The geographical proximity condition is set to a patch spacing of less than 100 meters. If the current patch spacing is 80 meters and the difference deviation level is 3, the connectivity judgment standard is "high deviation-proximity". If the patch spacing is 150 meters, the geographical proximity condition is ignored and the judgment standard is "high deviation-non-proximity". For example, if a patch pair has a difference of 2.2 (level 3) and a spacing of 90 meters, the judgment standard is "high deviation-proximity". Another patch pair has a difference of 1.8 (level 2) and a spacing of 70 meters, the judgment standard is "medium deviation-proximity". The final output is the connectivity judgment standard.

[0078] The status identification generation submodule, based on the connectivity judgment benchmark, screens the patch pairs that meet the benchmark value and meet the geographical proximity conditions, matches the preset connectivity status classification rules according to the deviation degree level, and generates the patch connectivity status identification;

[0079] The screening criteria for patch pairs are "medium deviation-proximity" or "low deviation-proximity". For example, a patch pair with a difference of 1.8 (level 2) and a spacing of 70 meters meets the conditions. The preset connectivity status classification rules are: level 1 and proximity are marked as "high connectivity", level 2 and proximity are marked as "medium connectivity", level 3 or non-proximity are marked as "low connectivity". The current example patch pair is marked as "medium connectivity". If a patch pair has a difference of 0.8 (level 1) and a spacing of 90 meters, it is marked as "high connectivity". Another patch pair has a difference of 3.5 (level 3) and a spacing of 50 meters, and is marked as "low connectivity". The patch connectivity status identifier is output.

[0080] See also Figure 5 ,The spectral dynamic screening module includes a sequence offset calculation submodule, an abnormal marker screening submodule, and an indicator set generation submodule;

[0081] The sequence offset calculation submodule extracts the continuous monitoring ratio sequence of chlorophyll reflectance change, canopy thermal radiation change, and near-infrared band change, calculates the distribution offset and concentration of each sequence, integrates the numerical relationship between the offset and concentration, and generates spectral distribution parameters;

[0082] Calculate the distribution deviation and concentration of each sequence using the formula:

[0083]

[0084] Among them, μ c Represents the arithmetic mean of the continuous monitoring ratio sequence of chlorophyll reflectance changes in the current monitoring period, μ h represents the arithmetic mean of the chlorophyll reflectance change series in the same historical period and region, σ c represents the standard deviation of the chlorophyll reflectance change sequence in the current monitoring period, n represents the number of data points in the continuous monitoring ratio sequence, x i represents the i-th monitoring ratio data in the sequence, k represents the weight coefficient dynamically adjusted based on the sequence length (take k = 0.5, dynamically set according to the sequence length n: when n ≥ 10, k = 0.8, when n < 10, k = 0.5), D s represents the distribution deviation of the chlorophyll reflectance change series, C s represents the concentration of the chlorophyll reflectance variation series;

[0085] Taking the chlorophyll reflectance change sequence as an example, according to the satellite monitoring data of a certain area, the chlorophyll reflectance change ratio sequence for 5 consecutive days is 0.85, 0.82, 0.88, 0.90, and 0.75, and the historical mean value for the same period is μ h =0.80 (calculated based on data from the same region and time period from 2020 to 2022), the current series mean μ cCalculated as (0.85+0.82+0.88+0.90+0.75) / 5=0.84, standard deviation σ c Calculated as Sequence length n = 5, adjustment coefficient k = 0.5 (because n = 5 < 10, according to the preset rule when n < 10 k = 0.5), calculate D s Item 1 Item 2 Final D s =1.492+0.022=1.514, concentration C s Calculated as Integration D s with C s When the spectral distribution parameter is D s 0.6+C s ·0.4=1.514·0.6+17.717·0.4=0.908+7.087=7.995. The results show that the distribution deviation of the current chlorophyll reflectance change sequence compared with the historical mean is 1.514, the concentration is 17.717, and the weighted spectral distribution parameter 7.995 provides a quantitative basis for the subsequent abnormal mark screening submodule. The deviation D s The value of reflects the comprehensive difference between the current sequence and historical data (including mean shift and fluctuation range), and the concentration C s The value reflects the stability and mean reliability of the current data. After the two are integrated through preset weights (0.6 and 0.4), the spectral distribution parameters directly determine whether the sequence exceeds the preset offset range or concentration threshold. For example, when the spectral distribution parameter exceeds the preset threshold of 8.0, it is judged to be abnormal. The current value of 7.995 does not exceed the limit. This result is used as the quantitative output of the "spectral distribution parameter" and is called by subsequent modules to determine the abnormality of the sequence.

[0086] The abnormal marking and screening submodule calls the spectral distribution parameters, compares the distribution deviation with the preset deviation range, and the concentration with the preset concentration threshold, screens out the sequences with deviations exceeding the range or concentrations not reaching the threshold, marks the corresponding sequence types, and generates abnormal spectral sequences;

[0087] Call the spectral distribution parameters (such as chlorophyll parameter 6.692, canopy parameter offset 2, concentration 0.5), compare the offset with the preset range of ±0.2, chlorophyll offset 0.04 is within the range, canopy offset 2 is out of range, concentration threshold 10, chlorophyll concentration 16.67 meets the standard, canopy concentration 0.5 does not meet the standard, screen the canopy thermal radiation sequence as abnormal, mark its type as "canopy thermal radiation abnormality", near-infrared band offset 0.02 is within the range, concentration 50 meets the standard, no mark, for example, another chlorophyll sequence offset 0.25 is out of range, concentration 8 is below the threshold, then it is marked as "chlorophyll reflectance abnormality", and finally output the abnormal spectral sequence marking result.

[0088] The indicator set generation submodule integrates the chlorophyll reflectance, canopy thermal radiation, and near-infrared band abnormal data marked in the abnormal spectral sequence, stores them by sequence type, summarizes the associated parameters of all abnormal sequences, and generates a dynamic abnormal spectral indicator set;

[0089] The canopy thermal radiation anomaly sequence with a mean of 32°C and a standard deviation of 2°C is stored in the "thermal radiation anomaly category." If the chlorophyll sequence is marked, its mean of 0.85 and standard deviation of 0.06 are stored in the "chlorophyll anomaly category." The associated parameters of the anomaly sequence are extracted during aggregation. For example, the canopy anomaly associated parameters include a slope difference of 5 degrees and an edge contact length of 50 meters. The chlorophyll anomaly associated parameter is a dispersion mean ratio of 0.08. The integrated dynamic anomaly spectral indicator set includes {anomaly type: canopy thermal radiation, mean: 32, standard deviation: 2, associated parameters: slope difference 5, contact length 50} and {anomaly type: chlorophyll reflectance, mean: 0.85, standard deviation: 0.06, associated parameters: dispersion 0.08}. For example, unmarked near-infrared anomalies are not included in the set. The final output is a classified and stored indicator set.

[0090] See also Figure 6 ,The coverage dynamic integration module includes the abnormal item elimination submodule, the dynamic strength fusion submodule, and the coverage conflict determination submodule;

[0091] The abnormal item removal submodule calls the chlorophyll reflectance deviation abnormal item and the near-infrared band concentration abnormal item in the dynamic abnormal spectral indicator set, filters the soil water holding capacity and light radiation data of the non-connected area, removes the monitoring units containing the above abnormal items, retains the remaining canopy thermal radiation and near-infrared band change data, and generates the filtered spectral data;

[0092] Obtain the chlorophyll reflectance deviation anomaly items marked in the dynamic abnormal spectral index set (such as the mean value of 0.85 exceeds the historical range of 0.70-0.80) and the near-infrared band concentration anomaly items (such as the concentration of 50 is lower than the threshold of 60), call the area marked as "low connectivity" in the patch connectivity status identification (such as patch numbers P5-P8), extract its soil water holding capacity data (such as P5 water holding capacity 20%, P6 water holding capacity 18%), and screen the water holding capacity below the preset threshold of 25% and the light radiation above 600W / m 2 Monitoring unit (e.g. P5 water holding capacity 20% < 25%, radiation 650W / m 2 >600, meeting the elimination conditions), the units that meet both the abnormal items and the screening conditions (such as P5's chlorophyll deviation abnormality + water holding capacity / radiation exceeding the standard) are removed from the data set, and the canopy thermal radiation series (such as P6 thermal radiation mean 32℃) and near-infrared band change data (such as P6 near-infrared mean 0.05) are retained. For example, P7 with a water holding capacity of 22% but no abnormal item mark is retained, and the final output is the canopy thermal radiation data [32, 30, 28]℃ and near-infrared data [0.05, 0.06, 0.04] after elimination.

[0093] The dynamic intensity fusion submodule extracts the canopy thermal radiation change and near-infrared band change sequences based on the filtered spectral data, calculates the mean of the fluctuation amplitude of the two sequences, and superimposes the correction coefficient of soil water holding capacity to light radiation to generate the dynamic intensity value of the cover;

[0094] The canopy thermal radiation change sequence after screening was extracted (such as three-day data of 32℃, 30℃, and 28℃), and the average fluctuation amplitude was calculated as the average absolute value of the daily temperature difference (|32-30|+|30-28|) / 2=2℃. The fluctuation amplitude average of the near-infrared band change sequence (0.05, 0.06, 0.04) was (|0.05-0.06|+|0.06-0.04|) / 2=0.015. The correction coefficient of soil water holding capacity to light radiation was set as 0.1 for every 5% decrease in water holding capacity (for example, the water holding capacity of P6 was 18% lower than the benchmark 25%, and the corresponding coefficient of the difference was 7%. 0.14), superimpose the correction coefficient to the mean fluctuation amplitude: thermal radiation fluctuation 2℃×(1+0.14)=2.28, near-infrared fluctuation 0.015×(1+0.14)=0.0171, covering dynamic intensity value=2.28+0.0171=2.2971. If the water holding capacity of a unit is 20% (the difference of 5% corresponds to a coefficient of 0.1), thermal radiation fluctuation 3℃×1.1=3.3, near-infrared fluctuation 0.02×1.1=0.022, intensity value=3.3+0.022=3.322, and finally output the intensity value set of each unit [2.2971,3.322].

[0095] The coverage conflict determination submodule calls the coverage dynamic intensity value and the soil water holding capacity data of the non-connected area in the patch connectivity status mark, compares the distribution range of the coverage dynamic intensity value with the change trend of soil water holding capacity, marks the conflicting areas, classifies and integrates them, and generates the dynamic monitoring results of vegetation coverage;

[0096] For the matching degree between the distribution range of dynamic coverage intensity value and the trend of soil water holding capacity, the formula is used:

[0097]

[0098] Among them, I d Represents the coverage dynamic strength value output by the coverage dynamic strength fusion submodule, I mid Represents the arithmetic midpoint value of the reasonable range of preset coverage dynamic strength, I range Represents the difference between the maximum and minimum values of the preset interval, S w represents the linear regression slope of the continuous monitoring data of soil water holding capacity in non-connected areas, k1 represents the intensity deviation correction coefficient dynamically adjusted according to soil type, k2 represents the weight coefficient set according to the trend direction of water holding capacity change, M c A parameter representing the matching degree between the dynamic intensity of cover and the trend of soil water holding capacity;

[0099] To override the dynamic intensity value I d =2.2971 (calculated by the dynamic intensity fusion submodule), the preset coverage dynamic intensity reasonable range is [1.5,3.5], calculate I mid =(1.5+3.5) / 2=2.5, I range =3.5-1.5=2.0, the three-day monitoring data of soil water holding capacity is 18%, 17%, and 16%, and the slope S is calculated by linear regression. w :

[0100]

[0101] The soil type is sandy soil (determined according to the soil texture classification standard), k1 is set to 0.6 (sandy soil has strong permeability and is sensitive to strength deviation. According to experimental data, the k1 range of sandy soil is 0.5-0.7). The change in water holding capacity is a negative trend, and k2 is set to 0.4 (negative trends have lower weight than positive trends, based on the setting of ecological response research). The matching parameter M is calculated. c :

[0102] Item I mid ·|S w |=2.5·1=2.5;

[0103] Item 2

[0104] Final M c =2.5+0.02435=2.52435. The result shows that the matching parameter of the dynamic strength of coverage and the trend of soil water holding capacity is 2.52435, and the matching parameter M c By comparing with the preset threshold (such as threshold 2.5), it is determined whether there is a conflict. The current M c =2.52435>2.5, marked as a conflict area. The matching parameter value reflects the combined influence of the deviation degree of coverage dynamic intensity within a reasonable range and the trend of water holding capacity change. The higher the value, the greater the conflict risk. The result is directly used in the logical judgment of "marking conflict areas" and becomes the quantitative basis for generating dynamic monitoring results of vegetation cover.

[0105] A forest vegetation cover growth monitoring method based on big data comprises the following steps:

[0106] S1: Extract the chlorophyll distribution dispersion of vegetation patches, calculate the texture direction range in the neighborhood based on the dispersion ratio, extract the near-infrared uniformity value, and generate a patch microstructure feature set;

[0107] S2: Call the chlorophyll discrete value, texture direction distribution range value, and near-infrared uniformity value in the patch microstructure feature set, perform normalization processing, and fuse them to generate a homogeneity intensity value;

[0108] S3: Combine the edge length and terrain differences between patches, based on the homogeneity strength values of adjacent patches, compare the preset benchmark value with the geographical proximity conditions, mark the connectivity status, and generate the patch connectivity status identification;

[0109] S4: extracting the change rate sequence of each type of spectral index, calculating the distribution deviation and concentration, screening the sequences whose deviation exceeds the preset deviation range or whose concentration does not exceed the preset concentration threshold, and generating a dynamic abnormal spectral index set;

[0110] S5: Call the patch connectivity status identification and dynamic abnormal spectral index set to extract soil water holding capacity and light radiation, eliminate the abnormal items of chlorophyll reflectance offset and near-infrared band concentration, and generate dynamic monitoring results of vegetation cover.

[0111] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A forest vegetation cover growth monitoring system based on big data, characterized in that: The system comprises: The patch microstructure analysis module extracts the chlorophyll distribution dispersion of vegetation patches, calculates the texture direction range within the neighborhood based on the dispersion ratio, extracts the near-infrared uniformity value, and generates a patch microstructure feature set; The homogeneity fusion module calls the chlorophyll discrete value, texture direction distribution range value, and near-infrared uniformity value in the patch microstructure feature set, performs normalization processing, and fuses them to generate a homogeneity intensity value; The connectivity identification module combines the edge lengths and terrain differences between patches, compares the homogeneity strength values of the adjacent patches with the preset benchmark values and the geographical proximity conditions, marks the connectivity status, and generates a patch connectivity status identifier; The spectral dynamic screening module extracts the change rate sequence of each type of spectral index, calculates the distribution deviation and concentration, and screens the sequences whose deviation exceeds the preset deviation range or whose concentration does not exceed the preset concentration threshold, thereby generating a set of dynamic abnormal spectral indicators; The coverage dynamic integration module calls the patch connectivity status identifier and the dynamic abnormal spectral index set to extract soil water holding capacity and light radiation, eliminate the abnormal items of chlorophyll reflectance offset and near-infrared band concentration, and generate vegetation coverage dynamic monitoring results.

2. The forest vegetation cover growth monitoring system based on big data according to claim 1 is characterized by: The patch microstructure feature set includes the sub-pixel chlorophyll dispersion ratio, the main texture direction angular distribution range, and the near-infrared uniformity mean. The homogeneity intensity value is specifically the standardized discrete interval value, the angle consistency quantification value, and the uniformity relative grade value. The patch connectivity status identifier covers the edge contact strength threshold, slope difference tolerance, and homogeneity difference judgment threshold. The dynamic anomaly spectral indicator set includes the chlorophyll offset anomaly sequence, the canopy thermal radiation concentration anomaly sequence, and the near-infrared band dynamic anomaly sequence. The corrected coverage dynamic intensity value specifically refers to the fusion of canopy thermal radiation intensity, near-infrared band stability, and water holding capacity correlation correction factor.

3. The forest vegetation cover growth monitoring system based on big data according to claim 1 is characterized by: The patch microstructure analysis module includes a chlorophyll difference integration submodule, a texture distribution span submodule, and a reflection uniformity integration submodule; The chlorophyll difference integration submodule detects the chlorophyll distribution data of sub-pixels within vegetation patches, analyzes the difference between the chlorophyll value of each sub-pixel and the overall mean of the patch, integrates the correlation between the difference value and the mean, and generates a dispersion-mean ratio; The texture distribution span submodule extracts the main texture direction angle data of the pixel unit in the neighborhood, determines the distribution boundary of the main texture direction angle, evaluates the concentrated distribution span range of the angle value, and generates the main texture angle span; The reflection uniformity comprehensive submodule calls the discreteness mean ratio and the main texture angle span, combines the distribution state of the near-infrared reflectivity in each sub-pixel unit, integrates the discreteness influencing factor and the texture span weight, evaluates the spatial consistency of the reflectivity within the patch, and generates a patch microstructure feature set.

4. The forest vegetation cover growth monitoring system based on big data according to claim 1 is characterized by: The homogeneous fusion module includes a parameter normalization submodule, an index conversion submodule, and a homogeneous intensity synthesis submodule; The parameter normalization submodule calls the chlorophyll discrete values in the patch microstructure feature set, adjusts the distribution range of the discrete values through linear mapping according to the upper and lower limits of the preset standard interval, and generates a standard discrete coefficient; An index conversion submodule extracts the texture direction distribution range value from the patch microstructure feature set, calculates the ratio coefficient between the direction distribution range and the theoretical maximum angle, and generates an angle consistency index by combining the angle concentration weight factor; The homogeneity intensity synthesis submodule calls the standard dispersion coefficient and angle consistency index, superimposes the near-infrared uniformity value in the plaque microstructure feature set, converts the uniformity value into a relative grade according to the grade segmentation rule, and generates a homogeneity intensity value by combining the weighted calculation results of the three parameters.

5. The forest vegetation cover growth monitoring system based on big data according to claim 1 is characterized in that: The connectivity determination module includes a homogeneity difference calculation submodule, a connectivity benchmark comparison submodule, and a state identification generation submodule; The homogeneity difference calculation submodule obtains the homogeneity strength values of the adjacent patches, calculates the absolute value of the difference between the two, combines the edge contact length between the patches and the terrain slope difference value, and superimposes the correction coefficient of the edge contact length to the slope difference to generate the homogeneity difference degree; A connectivity benchmark comparison submodule calls the homogeneity difference degree, compares a preset connectivity benchmark value with a geographic proximity condition, divides the degree of deviation between the homogeneity difference degree and the benchmark value into discrete levels, and generates a connectivity judgment benchmark based on the spatial constraint range of the geographic proximity condition; The status identification generation submodule selects the patch pairs that meet the benchmark value and the geographical proximity condition based on the connectivity judgment benchmark, matches the preset connectivity status classification rules according to the deviation degree level, and generates the patch connectivity status identification.

6. The forest vegetation cover growth monitoring system based on big data according to claim 1 is characterized by: The spectrum dynamic screening module includes a sequence offset calculation submodule, an abnormal marker screening submodule, and an indicator set generation submodule; The sequence offset calculation submodule extracts the continuous monitoring ratio sequence of chlorophyll reflectance change, canopy thermal radiation change, and near-infrared band change, calculates the distribution offset and concentration of each sequence, integrates the numerical relationship between the offset and concentration, and generates spectral distribution parameters; The abnormal marking screening submodule calls the spectral distribution parameters, compares the distribution deviation with the preset deviation range, and the concentration with the preset concentration threshold, screens out sequences with deviations exceeding the range or concentrations not reaching the threshold, marks the corresponding sequence types, and generates abnormal spectral sequences; The indicator set generation submodule integrates the chlorophyll reflectance, canopy thermal radiation, and near-infrared band abnormal data marked in the abnormal spectral sequence, stores them by sequence type, summarizes the associated parameters of all abnormal sequences, and generates a dynamic abnormal spectral indicator set.

7. The forest vegetation cover growth monitoring system based on big data according to claim 1 is characterized by: The coverage dynamic integration module includes an abnormal item elimination submodule, a dynamic strength fusion submodule, and a coverage conflict determination submodule; The abnormal item removal submodule calls the chlorophyll reflectance deviation abnormal item and the near-infrared band concentration abnormal item in the dynamic abnormal spectral indicator set, filters the soil water holding capacity and light radiation data of the non-connected area, removes the monitoring units containing the above abnormal items, retains the remaining canopy thermal radiation and near-infrared band change data, and generates the filtered spectral data; The dynamic intensity fusion submodule extracts the canopy thermal radiation change and near-infrared band change sequences based on the screened spectral data, calculates the mean of the fluctuation amplitudes of the two sequences, and superimposes the correction coefficient of soil water holding capacity to light radiation to generate the coverage dynamic intensity value; The coverage conflict determination submodule calls the coverage dynamic intensity value and the soil water holding capacity data of the non-connected area in the patch connectivity status identification, compares the distribution range of the coverage dynamic intensity value with the matching trend of the soil water holding capacity, marks the conflicting areas and classifies and integrates them to generate dynamic monitoring results of vegetation coverage.

8. A forest vegetation cover growth monitoring method based on big data, characterized in that: The method is used in the forest vegetation cover growth monitoring system based on big data according to any one of claims 1 to 7, comprising the following steps: S1: Extract the chlorophyll distribution dispersion of vegetation patches, calculate the texture direction range in the neighborhood based on the dispersion ratio, extract the near-infrared uniformity value, and generate a patch microstructure feature set; S2: calling the chlorophyll discrete value, texture direction distribution range value, and near-infrared uniformity value in the patch microstructure feature set, performing normalization processing, and fusing them to generate a homogeneity intensity value; S3: combining the edge lengths and terrain differences between patches, based on the homogeneity strength values of the adjacent patches, comparing the preset benchmark values with the geographical proximity conditions, marking the connectivity status, and generating a patch connectivity status identifier; S4: extracting the change rate sequence of each type of spectral index, calculating the distribution deviation and concentration, screening the sequences whose deviation exceeds the preset deviation range or whose concentration does not exceed the preset concentration threshold, and generating a dynamic abnormal spectral index set; S5: Call the patch connectivity status identifier and dynamic abnormal spectral indicator set, extract soil water holding capacity and light radiation, remove abnormal items of chlorophyll reflectance offset and near-infrared band concentration, and generate dynamic monitoring results of vegetation coverage.

Images