Vegetation coverage monitoring method and system fusing lidar and optical remote sensing

Abstract

This invention discloses a method and system for monitoring vegetation cover by integrating LiDAR and optical remote sensing, belonging to the field of vegetation ecological monitoring technology. The method includes: deploying LiDAR sensors and multispectral imaging devices in the vegetation area to be monitored, building a spatiotemporally synchronized acquisition module to simultaneously acquire three-dimensional point cloud and spectral reflectance data of vegetation; extracting vertical structure parameters from the point cloud data, and calculating physiological activity parameters by combining the spectral data; constructing a layered coverage model based on the two types of parameters to generate effective photosynthetic and structural coverage, assessing the vegetation growth status by combining the spatial distribution of the two, and outputting the coverage monitoring results. This invention solves the technical problems of traditional vegetation cover monitoring methods that cannot simultaneously detect the vertical structure and physiological activity of vegetation, and that the acquired data has a single dimension and the results are one-sided. It achieves integrated monitoring of vegetation spatial structure and physiological activity characteristics, and comprehensively and accurately reflects the actual coverage and growth of vegetation.

Description

Technical Field

[0001] This invention relates to the field of vegetation ecological monitoring technology, and in particular to a method and system for monitoring vegetation cover that integrates LiDAR and optical remote sensing. Background Technology

[0002] Vegetation cover is a core indicator for assessing ecological conditions and evaluating vegetation growth status. Its accurate monitoring is crucial for ecological protection and refined vegetation management. Current technologies for vegetation cover detection primarily employ single sensors or remote sensing equipment, and the application of intelligent sensing systems has not yet achieved spatiotemporal fusion of multi-source sensor data. While these methods have played a role in monitoring single structural or physiological characteristics, their application to refined vegetation monitoring has revealed numerous limitations as ecological monitoring requirements increase, failing to meet the needs for accurate assessment of vegetation growth status and effective ecological management. Summary of the Invention

[0003] This application provides a vegetation cover monitoring method and system that integrates LiDAR and optical remote sensing, which solves the technical problem that traditional vegetation cover monitoring methods cannot take into account the coordinated detection of vegetation vertical structure and physiological activity, and the data obtained is of a single dimension and the results are one-sided.

[0004] The first aspect of this application provides a method for monitoring vegetation cover by integrating LiDAR and optical remote sensing. The method includes: deploying a LiDAR sensor and a multispectral imaging device in a vegetation area to be measured, and establishing a spatiotemporally synchronized acquisition module, wherein the LiDAR sensor and the multispectral imaging device maintain a fixed spatial relationship through a rigid support; based on the spatiotemporally synchronized acquisition module, acquiring three-dimensional point cloud data of the vegetation through the LiDAR sensor, and acquiring spectral reflectance data of the vegetation through the multispectral imaging device; extracting vertical structure parameters of the vegetation based on the three-dimensional point cloud data, and calculating physiological activity parameters of the vegetation in conjunction with the spectral reflectance data; constructing a layered coverage model based on the vertical structure parameters and physiological activity parameters to generate effective photosynthetic coverage and structural coverage of the vegetation; assessing the vegetation growth status based on the spatial distribution of the effective photosynthetic coverage and structural coverage, and outputting the coverage monitoring results.

[0005] A second aspect of this application provides a vegetation cover monitoring system integrating LiDAR and optical remote sensing. The system includes: a spatiotemporal synchronous acquisition module construction module, used to deploy LiDAR sensors and multispectral imaging devices in the vegetation area to be measured, and establish a spatiotemporal synchronous acquisition module, wherein the LiDAR sensors and multispectral imaging devices maintain a fixed spatial relationship through a rigid support; a monitoring data acquisition module, based on the spatiotemporal synchronous acquisition module, acquiring three-dimensional point cloud data of vegetation through the LiDAR sensors and acquiring spectral reflectance data of vegetation through the multispectral imaging devices; a physiological activity parameter acquisition module, used to extract vertical structural parameters of vegetation based on the three-dimensional point cloud data, and calculate physiological activity parameters of vegetation in combination with the spectral reflectance data; a layered cover acquisition module, based on the vertical structural parameters and physiological activity parameters, constructing a layered cover model to generate effective photosynthetic cover and structural cover of vegetation; and a cover monitoring result acquisition module, used to assess the vegetation growth status based on the spatial distribution of the effective photosynthetic cover and structural cover, and output the cover monitoring results.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application utilizes LiDAR sensors and multispectral imaging devices deployed in the vegetation area to be tested, along with a spatiotemporally synchronized acquisition module, to collect 3D point cloud and spectral reflectance data of the vegetation. Through height-level stratification and band calculations, parameters related to the vertical structure and physiological activity of the vegetation are extracted. A stratified coverage model is constructed to generate effective photosynthetic coverage and structural coverage. Combining the spatial distribution characteristics of these two types of coverage, vegetation growth status is assessed and anomaly warnings are triggered. This accurately achieves stratified and refined monitoring of vegetation coverage, making the monitoring results more comprehensive and precise. It provides a reliable basis for accurate assessment of vegetation growth status and effective ecological management, achieving integrated monitoring of vegetation spatial structure and physiological activity characteristics, and comprehensively and accurately reflecting the actual coverage and growth of the vegetation. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a flowchart illustrating the vegetation cover monitoring method that integrates LiDAR and optical remote sensing provided in the embodiments of this application.

[0009] Figure 2This is a schematic diagram of the vegetation cover monitoring system that integrates LiDAR and optical remote sensing provided in the embodiments of this application.

[0010] Figure labeling: Spatiotemporal synchronization acquisition module construction module 1, monitoring data acquisition module 2, physiological activity parameter acquisition module 3, stratified coverage acquisition module 4, coverage monitoring result acquisition module 5. Detailed Implementation

[0011] This application provides a vegetation cover monitoring method and system that integrates LiDAR and optical remote sensing, which solves the technical problem that traditional vegetation cover monitoring methods cannot take into account the coordinated detection of vegetation vertical structure and physiological activity, and the data obtained is of a single dimension and the results are one-sided.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0014] Example 1, as Figure 1 As shown, a vegetation cover monitoring method integrating LiDAR and optical remote sensing is described, wherein the method includes: LiDAR sensors and multispectral imaging devices are deployed in the vegetation area to be tested, and a spatiotemporal synchronous acquisition module is established. The LiDAR sensors and multispectral imaging devices maintain a fixed spatial relationship through rigid supports.

[0015] In this embodiment, the LiDAR sensor, or lidar sensor, is an active remote sensing device that accurately acquires the three-dimensional spatial coordinates of a target object and generates three-dimensional point cloud data by emitting laser pulses and recording the flight time and reflection characteristics of the echo signals. The multispectral imaging device is a passive optical remote sensing imaging device capable of simultaneously acquiring ground object reflectance images in multiple discrete spectral bands, such as visible light and near-infrared, and quantitatively acquiring spectral reflectance data of the target object in different bands.

[0016] Specifically, the relative spatial position parameters of the LiDAR sensor and the multispectral imaging device are pre-calibrated, including horizontal offset, vertical offset, axial rotation angle, and field-of-view overlap range. An integrated rigid bracket is fabricated using a low-deformation metal material. Two parallel mounting positions are set on the bracket surface, with the spacing and angle of the two positions matching the pre-calibrated relative spatial position parameters. The LiDAR sensor and multispectral imaging device are then securely mounted in their respective mounting positions. After installation, it is verified that their field-of-view center axes are aligned. The field-of-view overlap range is measured as the ratio of the overlapping area to the total effective field-of-view area of ​​a single device on the effective detection plane of the vegetation canopy. The preset threshold is set to be no less than 90%. For refined canopy monitoring scenarios with an effective detection distance of no more than 50 meters, the preset threshold is set to no less than 95%. For large-scale vegetation monitoring scenarios with an effective detection distance greater than 50 meters but no more than 200 meters, the preset threshold is set to no less than 85%. After verifying that the field of view overlap range meets the preset threshold requirements, all fasteners are locked to ensure that there is no relative displacement between the two during the acquisition process, thus establishing a fixed spatial relationship.

[0017] Next, the monitoring boundary of the vegetation area to be monitored is delineated, and the planar coordinate information of the monitoring range is obtained. Based on the field-of-view parameters and maximum effective detection distance of the LiDAR sensor and multispectral imaging device, the effective monitoring coverage area of ​​a single set of acquisition equipment is calculated. Based on the effective monitoring coverage area, acquisition points are planned within the vegetation area to be monitored, with a 10% to 20% overlap in the effective monitoring coverage of adjacent acquisition points to eliminate blind spots. A horizontal reference plane is set at each acquisition point, and the installed rigid bracket is horizontally placed on the horizontal reference plane. The height of the bracket is adjusted so that the center of the sensor's field of view is aligned with the main canopy area of ​​the vegetation to be monitored, completing the equipment deployment at all acquisition points.

[0018] Finally, the deployed LiDAR sensor and multispectral imaging device are configured with the same main control processing unit. This unit establishes bidirectional data communication links with both the LiDAR sensor and the multispectral imaging device via a high-speed data transmission interface. A unified high-precision temperature-controlled crystal oscillator clock source is configured for the main control processing unit. The same reference clock signal output from this clock source is transmitted to both the LiDAR sensor and the multispectral imaging device via a synchronization clock cable, ensuring that their internal acquisition clocks are synchronized with a synchronization error controlled within 1 microsecond. A synchronous acquisition trigger program is written into the main control processing unit. This program simultaneously sends acquisition trigger commands to both the LiDAR sensor and the multispectral imaging device at a preset acquisition frequency, causing them to initiate 3D point cloud data acquisition and spectral reflectance data acquisition simultaneously, thus completing the establishment of the spatiotemporal synchronous acquisition module.

[0019] Based on the spatiotemporal synchronous acquisition module, the three-dimensional point cloud data of vegetation is acquired through the LiDAR sensor, and the spectral reflectance data of vegetation is acquired through the multispectral imaging device.

[0020] In this embodiment of the application, spectral reflectance refers to the ratio of the radiant flux reflected by the ground object to the incident radiant flux when incident light of a specific wavelength shines on the ground object surface. It is a core optical parameter that can be collected by a multispectral imaging device and can quantitatively reflect the physiological characteristics of vegetation such as chlorophyll and water.

[0021] Optionally, the main control processing unit of the spatiotemporal synchronization acquisition module is first activated, loading the acquisition parameter file pre-configured by those skilled in the art. The acquisition parameter file includes the laser emission frequency, scanning angle range, and point cloud sampling density of the LiDAR sensor, as well as the spectral band range, exposure time, and imaging resolution of the multispectral imaging device. Using a high-precision temperature-controlled crystal oscillator clock source built into the main control processing unit, clock synchronization verification between the LiDAR sensor and the multispectral imaging device is completed, confirming that the internal acquisition clock deviation between the two does not exceed a preset threshold of 1 microsecond. If the clock deviation exceeds the threshold, the clock calibration operation is re-executed until the clock synchronization accuracy of the two meets the acquisition requirements, completing the equipment-ready configuration before acquisition.

[0022] Next, the main control processing unit runs the synchronous acquisition trigger program, simultaneously sending synchronous acquisition trigger commands to both the LiDAR sensor and the multispectral imaging device according to the preset acquisition time interval. Upon receiving the trigger command, the LiDAR sensor immediately activates its laser scanning unit, emitting laser pulses towards the vegetation area to be measured, and receiving the laser echo signals reflected back from the vegetation canopy and the ground. It records the emission time, echo reception time, scanning angle, and echo intensity information of each laser pulse, completing the single scan data acquisition. Upon receiving the same trigger command, the multispectral imaging device synchronously starts the shutter exposure, acquiring multi-band spectral images within the corresponding field of view of the vegetation area to be measured, recording the image grayscale value and exposure time information for each band, completing the single imaging data acquisition.

[0023] After a preset number of data acquisitions are completed at a single acquisition point, the main control processing unit stops sending synchronous acquisition trigger commands and numbers and stores all raw data acquired at that point. The numbering information includes the planar coordinates of the acquisition point, the acquisition time, and the unique identifier of the device. Following the pre-planned sequence of acquisition points, the rigid support equipped with the LiDAR sensor and multispectral imaging device is moved to the next acquisition point. After adjusting the support to a horizontal position and confirming that the field of view coverage meets the requirements, the clock synchronization verification, synchronous trigger acquisition, and data numbering and storage operations are repeated until data acquisition at all planned acquisition points in the vegetation area to be measured is completed.

[0024] Finally, the raw LiDAR data acquired from all collection points were preprocessed. The preprocessing steps were: point cloud denoising, point cloud registration, classification of ground and non-ground point clouds, and coordinate system I. After removing outliers and redundant data caused by environmental noise, complete 3D point cloud data of the vegetation area to be measured was generated. The raw multispectral images acquired from all collection points were also preprocessed. The preprocessing steps were: radiometric calibration, atmospheric correction, image stitching, and geometric correction. The grayscale values ​​of the images were converted to true surface reflectance values. After completing coordinate system I and global image stitching, complete spectral reflectance data of the vegetation area to be measured was generated.

[0025] The vertical structure parameters of the vegetation are extracted based on the three-dimensional point cloud data, and the physiological activity parameters of the vegetation are calculated by combining the spectral reflectance data.

[0026] In one embodiment of this application, the acquired three-dimensional point cloud data of vegetation is subjected to height stratification processing to divide it into three levels: upper, middle and lower canopy layers, and the vertical structure parameters corresponding to each level are extracted. At the same time, the collected vegetation spectral reflectance data is subjected to band calculation to extract chlorophyll index and water content index as physiological activity parameters characterizing the vegetation growth status. This step will be described in detail later.

[0027] Based on the vertical structure parameters and physiological activity parameters, a layered coverage model is constructed to generate the effective photosynthetic coverage and structural coverage of vegetation.

[0028] Specifically, based on the point cloud density and leaf area index extracted from the vertical structure parameters in the aforementioned steps, and combined with the chlorophyll index from the physiological activity parameters obtained simultaneously, a calculation model for photosynthetic effective coverage is established. At the same time, a calculation model for structural coverage is established based on the canopy height distribution and point cloud spatial distribution characteristics in the vertical structure parameters. Finally, the effective photosynthetic coverage and structural coverage of the vegetation are generated through the two corresponding calculation models. This step will be explained in detail in the following sections.

[0029] Based on the spatial distribution of the effective photosynthetic coverage and structural coverage, the vegetation growth status is assessed, and the coverage monitoring results are output.

[0030] Specifically, a coverage-growth status mapping table is pre-constructed and vegetation health levels corresponding to different coverage ranges are defined. Based on this mapping table and the spatial distribution characteristics of the generated effective photosynthetic coverage and structural coverage, vegetation growth status is assessed, and a vegetation health status heat map is generated simultaneously. When the heat map detects that the coverage of a local area within the vegetation area to be tested is lower than a preset threshold, an abnormal growth warning is automatically triggered. This step will be explained in detail later.

[0031] Furthermore, the method provided in this application embodiment includes: The three-dimensional point cloud data is subjected to high-level layering to divide the canopy into upper, middle and lower layers, and the vertical structure parameters of each layer are extracted; the spectral reflectance data is subjected to band calculation to extract the chlorophyll index and water content index as physiological activity parameters.

[0032] Specifically, firstly, the three-dimensional point cloud data of vegetation is divided into three intervals according to height: upper, middle and lower canopy layers. Then, the point cloud density and leaf area index in each interval layer are calculated. The calculation results are used as the vertical structure parameters of the corresponding canopy layer. This step will be explained in detail later.

[0033] Next, retrieve the spectral reflectance data after radiometric calibration, atmospheric correction, and geometric correction. Determine the core bands and parameters required for the data: the red band has a center wavelength of 670 nm and a bandwidth of 10 nm; the red-edge band has a center wavelength of 715 nm and a bandwidth of 10 nm; the near-infrared band has a center wavelength of 850 nm and a bandwidth of 20 nm; the first shortwave infrared band has a center wavelength of 1250 nm and a bandwidth of 10 nm; and the second shortwave infrared band has a center wavelength of 1650 nm and a bandwidth of 10 nm. When the actual center wavelength of the multispectral imaging device deviates from the above reference values, the corresponding band with a deviation of no more than 15 nm from the reference value is preferentially selected for calculation. If there is no corresponding second shortwave infrared band, the first shortwave infrared band with a center wavelength of 1250 nm is selected as a substitute band for the moisture content index calculation.

[0034] Then, the coordinate system of the spectral reflectance data is transformed to a spatial coordinate reference system consistent with that of the 3D point cloud data, ensuring a one-to-one correspondence between the planar pixels of the two types of data and the horizontal projection positions of the spatial point cloud. Normalized vegetation index (NZI) calculation is performed on the spectral reflectance data. The difference between the near-infrared band reflectance and the red band reflectance is divided by their sum to obtain the NZI value for each pixel. Pixels with NZI values ​​greater than 0.2 are retained as initial vegetation pixels. Spatially matching the initial vegetation pixels with the horizontal projection areas of non-ground point clouds in the 3D point cloud data, pixels without corresponding non-ground point clouds are removed, and the retained pixels are determined as valid vegetation pixels. Invalid pixels within the effective vegetation pixel range affected by cloud shadows and light saturation are removed, completing the basic data preparation before band calculations.

[0035] For the preprocessed effective vegetation pixel spectral reflectance data, the red-edge band reflectance and near-infrared band reflectance corresponding to each pixel are extracted. A red-edge normalized chlorophyll index model is used to perform band calculations. The calculation process involves subtracting the red-edge band spectral reflectance from the near-infrared band spectral reflectance to obtain the first calculation result, and adding the red-edge band spectral reflectance to the near-infrared band spectral reflectance to obtain the second calculation result. Dividing the first calculation result by the second calculation result yields the chlorophyll index value for the corresponding pixel. Band calculations are performed pixel-by-pixel across the entire region according to the above calculation rules, generating continuous chlorophyll index raster data across the entire region. Based on the three-dimensional point cloud data, three canopy layers (upper, middle, and lower) are defined. Layered mask files representing the horizontal projection range of the point cloud in each layer are generated. These layered mask files are spatially overlaid with the chlorophyll index raster data, and the chlorophyll index values ​​within each layered mask range are extracted to obtain the chlorophyll index parameters corresponding to the upper, middle, and lower canopy layers, respectively. Using ground-measured leaf chlorophyll content data, the extracted chlorophyll index was corrected by linear regression to determine the linear fitting equation between the chlorophyll index and the measured chlorophyll content. When the coefficient of determination is greater than 0.7, the chlorophyll index parameter is considered valid, thus completing the extraction of chlorophyll index parameters characterizing vegetation physiological activity.

[0036] For the preprocessed effective vegetation pixel spectral reflectance data, the near-infrared band reflectance and the second shortwave infrared band reflectance corresponding to each pixel are extracted. A normalized water index model is used to perform band calculations. The calculation process involves subtracting the spectral reflectance of the second shortwave infrared band from the near-infrared band spectral reflectance to obtain the first calculation result, and adding the spectral reflectance of the second shortwave infrared band to obtain the second calculation result. Dividing the first calculation result by the second calculation result yields the water content index value for the corresponding pixel. Band calculations are performed pixel-by-pixel across the entire region according to the above calculation rules to generate continuous water content index raster data for the entire region. Based on the layered mask files corresponding to the upper, middle, and lower canopy layers divided from the 3D point cloud data, each layered mask file is spatially overlaid with the water content index raster data. The water content index values ​​within each layered mask range are extracted to obtain the water content index parameters corresponding to the upper, middle, and lower canopy layers, respectively. Using ground-measured equivalent water thickness data of leaves, the extracted water content index was linearly regressed to determine the linear fitting equation between the water content index and the measured leaf water content. When the coefficient of determination is greater than 0.65, the water content index parameter is considered valid, thus completing the extraction of the water content index parameter characterizing vegetation physiological activity.

[0037] Furthermore, the method provided in this application embodiment includes: The three-dimensional point cloud data is divided into three intervals according to height: upper, middle and lower. The point cloud density and leaf area index in each interval are calculated as vertical structure parameters.

[0038] Optionally, from the preprocessed 3D point cloud data, firstly, extract the classified ground point cloud and non-ground point cloud. Perform noise point removal and vegetation feature point cloud filtering on the non-ground point cloud. After removing outlier noise points and non-vegetation feature point clouds, the remaining vegetation canopy point clouds are determined as valid canopy point clouds. Calculate the maximum elevation value of all valid canopy point clouds and determine this value as the canopy top elevation. Calculate the average elevation value of all ground point clouds and determine this value as the ground reference elevation. Calculate the difference between the canopy top elevation and the ground reference elevation, and determine this difference as the total canopy height.

[0039] Next, using the ground reference elevation as the vertical starting point, the total canopy height is divided into three continuous, non-overlapping intervals along the vertical upward direction. The lower canopy interval extends from the ground reference elevation to 30% of the total canopy height; the middle canopy interval extends from 30% to 70% of the total canopy height; and the upper canopy interval extends from 70% to the top of the canopy. Those skilled in the art can adjust the stratification ratio according to the canopy type of the vegetation being measured. For example, for tree canopies, the lower limit of the upper interval can be adjusted to 60% of the total canopy height, and for shrub canopies, the upper limit of the lower interval can be adjusted to 40% of the total canopy height. Based on the elevation value of each effective canopy point cloud, all effective canopy point clouds are assigned to their corresponding intervals, completing the division of the upper, middle, and lower canopy layers.

[0040] For each defined canopy layer, all valid canopy point clouds within that layer are extracted, and the total number of valid canopy point clouds within that layer is calculated. The area of ​​the horizontal projection region corresponding to the pre-defined monitoring boundary of the vegetation area to be monitored is obtained, and this area is used as the unified horizontal projection calculation benchmark for each layer. The total number of valid canopy point clouds within that layer is divided by the corresponding horizontal projection region area to obtain the point cloud density of that layer. The point cloud density calculations for the upper, middle, and lower canopy layers are performed sequentially, and the calculation results are used as the vertical structure parameters for the corresponding layers.

[0041] For each defined interval layer, all valid canopy point clouds belonging to that interval layer are extracted. Full echo data acquired by a LiDAR sensor is retrieved, and the total number of laser pulses perpendicularly incident on the upper boundary of that interval layer is counted. The count includes all laser emission pulses corresponding to the first, middle, and last echoes. The number of laser pulses that completely penetrate the lower boundary of that interval layer without being reflected by the valid canopy point clouds within that interval layer is counted, and this number is determined as the effective laser penetration quantity for that interval layer. The effective laser penetration quantity of that interval layer is divided by the total number of laser pulses incident on the upper boundary of that interval layer to obtain the laser transmittance of that interval layer.

[0042] Subsequently, based on the type of vegetation to be measured and the monitoring platform type of the LiDAR sensor, corresponding canopy extinction coefficients were preset. The extinction coefficient for broad-leaved vegetation collected by the ground-based airborne platform was set to 0.5, for coniferous vegetation to 0.8, and for grassland vegetation to 0.65. The extinction coefficient for the corresponding vegetation type collected by the UAV-based airborne platform was adjusted upwards by 0.05 to 0.1 based on the above baseline values. For mixed vegetation types, the corresponding extinction coefficients were weighted and averaged according to the canopy proportion of each vegetation type to obtain the appropriate canopy extinction coefficient. Based on the Beer-Lambert law, the natural logarithm of the laser transmittance of this interval layer was taken as a negative value, and then divided by the preset canopy extinction coefficient to obtain the leaf area index of this interval layer. The leaf area index calculations were completed sequentially for the upper, middle, and lower canopy layers, and the calculation results were used as the vertical structure parameters of the corresponding layers.

[0043] Furthermore, the method provided in this application embodiment includes: Based on the point cloud density and leaf area index in the vertical structure parameters, and combined with the chlorophyll index in the physiological activity parameters, a photosynthetic effective coverage calculation model is established; based on the canopy height distribution and point cloud spatial distribution characteristics in the vertical structure parameters, a structural coverage calculation model is established; effective photosynthetic coverage is generated through the photosynthetic effective coverage calculation model, and structural coverage is generated through the structural coverage calculation model.

[0044] Specifically, a photosynthetic efficiency estimation model is first established based on the regression equation between chlorophyll index and photosynthetic efficiency. Then, point cloud density and leaf area index in the vertical structure parameters are introduced into the photosynthetic efficiency estimation model as weighting factors. Finally, a photosynthetic effective coverage calculation model that includes point cloud density, leaf area index and chlorophyll index is established through regression analysis. This step will be explained in detail in the following content.

[0045] Next, based on the canopy height distribution analysis, the three-dimensional spatial distribution pattern of the canopy point cloud is analyzed and a height weighting function is established. Then, the spatial occupancy rate of each height layer is calculated according to the spatial distribution characteristics of the point cloud. Finally, the vertical layer weighting is performed by combining the height weighting function and the spatial occupancy rate of each height layer to generate a structural coverage calculation model. This step will also be explained in detail in the following content.

[0046] Finally, the normalized point cloud density, leaf area index, and chlorophyll index data corresponding to each interval of the upper, middle, and lower layers of the canopy in the vegetation area to be tested are retrieved. These data are then input into the photosynthetic effective coverage calculation model that has been constructed and validated. The model is then calculated pixel by pixel to generate raster data of effective photosynthetic coverage for each layer of the canopy and the entire area. Simultaneously, the canopy height distribution and point cloud spatial distribution feature data corresponding to each interval of the canopy in the vegetation area to be tested are retrieved. These data are then input into the structural coverage calculation model that has been constructed and validated. Vertical layer weighted calculation is then performed to generate raster data of structural coverage for each layer of the canopy and the entire area. This process enables the quantitative generation and spatial coordinate matching of the two types of coverage.

[0047] Furthermore, the method provided in this application embodiment includes: Based on the regression equation between the chlorophyll index and photosynthetic efficiency, a photosynthetic efficiency estimation model is established; the point cloud density and leaf area index are introduced into the photosynthetic efficiency estimation model as weighting factors; regression analysis is performed on the photosynthetic efficiency estimation model to establish a photosynthetic effective coverage calculation model that includes point cloud density, leaf area index and chlorophyll index.

[0048] In this embodiment, photosynthetic efficiency is the ability of plant leaves to convert absorbed photosynthetically active radiation into organic dry matter through photosynthesis. It is usually quantified by the net photosynthetic rate per unit time and per unit leaf area and is a core indicator that directly characterizes the physiological activity and growth potential of plants.

[0049] Specifically, firstly, the pre-processed chlorophyll index raster data corresponding to the upper, middle, and lower layers of the canopy are retrieved, and simultaneously, measured data of leaf photosynthetic efficiency at the corresponding canopy level within the vegetation area to be tested are acquired. Photosynthetic efficiency is quantified using the net photosynthetic rate of leaves. The measured data are collected using a portable photosynthesis measurement system. Following pre-defined canopy stratification rules, no fewer than thirty sample points are selected within the vegetation area to be tested. Net photosynthetic rate data of functional leaves at each sample point corresponding to the canopy level are collected, and the collected net photosynthetic rate data are used as the measured baseline value for photosynthetic efficiency.

[0050] The measured baseline value of photosynthetically active coverage is defined as the proportion of the projected area of ​​photosynthetically active leaves on the horizontal projection surface within the corresponding layer of the canopy to the total horizontal projection area of ​​the quadrat. The measured data are obtained by using a canopy analyzer combined with the stratified quadrat survey method. At each sample point, the photosynthetically active radiation interception data of the upper, middle and lower layers of the canopy are measured by the canopy analyzer. Combined with the destructive sampling survey results of leaves in the corresponding layer, the measured value of photosynthetically active coverage of the corresponding canopy layer at each sample point is calculated.

[0051] Next, the chlorophyll index and corresponding measured photosynthetic efficiency values ​​for each sample point were paired. Outliers were removed using the Grubbs' test, with a significance level set at 0.05, completing the preprocessing of the sample data. A univariate linear regression analysis was then performed, with chlorophyll index as the independent variable and measured photosynthetic efficiency as the dependent variable, to construct a regression equation between chlorophyll index and photosynthetic efficiency. The general formula for the regression equation is: Where P is the estimated photosynthetic efficiency, CI is the chlorophyll index, a is the regression slope coefficient, and b is the regression intercept coefficient. The coefficients of the regression equation are solved using the least squares method. At the same time, the coefficient of determination of the regression equation is calculated. When the coefficient of determination is greater than or equal to 0.7, the regression equation is considered valid. Based on the valid regression equation, a photosynthetic efficiency estimation model is established to realize the quantitative estimation of photosynthetic efficiency of each layer of the canopy based on the chlorophyll index.

[0052] Then, point cloud density and leaf area index (LAI) data for each layer of the canopy were retrieved. The two types of data were normalized using a min-max normalization method, mapping the values ​​of point cloud density and LAI to the range of 0 to 1, thus eliminating the influence of dimensional differences on model calculations. The core input of the photosynthetic efficiency estimation model is clearly defined as the chlorophyll index. The normalized point cloud density and LAI are used as multiplication weighting factors and introduced into the calculation logic of the photosynthetic efficiency estimation model. Point cloud density represents the spatial density of leaves in the corresponding layer of the canopy, and leaf area index represents the total photosynthetic tissue in the corresponding layer of the canopy. The general formula of the photosynthetic efficiency correction estimation model after introducing the weighting factors is: ,in P is the corrected photosynthetic efficiency estimate, D is the normalized point cloud density, and LAI is the normalized leaf area index. This calculation logic is used to implement the hierarchical weighted correction of the canopy photosynthetic efficiency, resulting in the photosynthetic efficiency correction estimation model after introducing structural weights.

[0053] Next, based on the photosynthetic efficiency correction estimation model with weighted factors, the corrected photosynthetic efficiency estimates for each sample point within the vegetation area to be tested were obtained, and the measured photosynthetically effective coverage values ​​for the corresponding canopy level of each sample point were simultaneously matched. Using normalized point cloud density, normalized leaf area index, and chlorophyll index as independent variables, and the measured photosynthetically effective coverage value of the corresponding sample point as the dependent variable, a multiple linear regression analysis method was used to construct a multiple regression equation. The general formula of the regression equation is: Where FPEC is the calculated value of photosynthetically effective coverage. , , , respectively, are the regression coefficients for point cloud density, leaf area index, and chlorophyll index, and d is the regression constant term.

[0054] Furthermore, the regression coefficients were calculated using partial least squares (PLS) to eliminate the effects of multicollinearity among independent variables. A significance test was performed on the regression coefficients, with a significance level set at 0.05. Insignificant independent variables were removed, and significant independent variables were retained to obtain the final photosynthetically effective coverage calculation model. For different vegetation types—broadleaf, coniferous, and shrub-grass—separate stratified photosynthetically effective coverage calculation models were constructed. For mixed vegetation types, a weighted average was calculated based on the canopy proportion of each vegetation type and the corresponding regression coefficients to obtain suitable model parameters. The model's calculation accuracy was verified using at least twenty reserved validation sample points. A model was considered valid and usable for quantitative calculation of photosynthetically effective coverage of each canopy layer in the tested vegetation area when the coefficient of determination between the calculated and measured values ​​was greater than or equal to 0.75 and the root mean square error was less than or equal to 0.1.

[0055] By clarifying the quantitative definition and measurement standards of photosynthetic effective coverage, the complete expression of the photosynthetic efficiency estimation model, the logic of introducing structural weight factors, and the general formula and construction process of the final photosynthetic effective coverage calculation model, the establishment process of the photosynthetic effective coverage calculation model has achieved a deep integration of the vertical layering structural characteristics and physiological activity characteristics of the vegetation canopy, effectively improving the accuracy of the layering calculation of vegetation photosynthetic effective coverage.

[0056] Furthermore, the method provided in this application embodiment includes: Based on the canopy height distribution, the three-dimensional spatial distribution pattern of the canopy point cloud is analyzed, and a height weighting function is established. The spatial occupancy rate of each height layer is calculated according to the spatial distribution characteristics of the point cloud. Based on the height weighting function and the spatial occupancy rate of each height layer, vertical layer weighting is performed to generate the structure coverage calculation model.

[0057] Specifically, firstly, the preprocessed, ground-point and non-ground-point classification, and noise-removed 3D canopy point cloud data is retrieved. Simultaneously, the effective canopy point cloud data for the three canopy layers (upper, middle, and lower) obtained in the previous steps, along with the extracted point cloud density and leaf area index data for the corresponding layers, are retrieved. The average elevation value of all ground point clouds is extracted as the ground reference elevation. The difference between the absolute elevation of each effective canopy point cloud and the ground reference elevation is calculated to obtain the relative canopy height of the corresponding point cloud. Taking the total canopy height as the overall range and using a fixed height step of 10 cm, the height range of each of the three canopy layers (upper, middle, and lower) is further subdivided at equal intervals to generate continuous subdivided height statistical intervals. The number of effective canopy point clouds within each subdivided height statistical interval is counted, generating a distribution sequence of canopy point cloud quantity as a function of relative height. This sequence is the canopy height distribution data.

[0058] Then, the proportion of point cloud quantity in each subdivided height statistical interval to the total number of effective canopy point clouds is calculated. The subdivided height statistical interval with the largest proportion, which is greater than or equal to 15%, is determined as the peak height interval. Based on the hierarchical classification of the peak height interval, the three-dimensional spatial distribution pattern of the canopy point cloud is determined. Peak height intervals belonging to the upper layer of the canopy are classified as upper-layer clustered, peak height intervals belonging to the middle layer of the canopy are classified as middle-layer clustered, and so on. The coefficient of variation of the proportion of point cloud quantity in all subdivided height statistical intervals is less than or equal to 20%, and the distribution is classified as uniform.

[0059] Next, for the upper, middle, and lower canopy layers, effective canopy point clouds were extracted from each layer. The 3D point clouds of each layer were then vertically projected onto a horizontal plane to generate a horizontal projection grid that perfectly matches the spatial reference of the spectral reflectance data acquired by the multispectral imaging device. The grid resolution was set to be the same as the pixel resolution of the multispectral imaging device to ensure spatial matching between the two types of data. For each cell within the horizontal projection grid, the number of effective canopy point clouds within that cell was counted, and the proportion of grid cells containing effective canopy point clouds to the total number of grid cells was calculated to obtain the horizontal coverage ratio of the point clouds. Simultaneously, the ratio of the standard deviation to the mean of the number of point clouds in all grid cells was calculated to obtain the horizontal distribution variation coefficient of the point clouds. Combining the vertical canopy height distribution data, the horizontal point cloud coverage ratio and distribution variation coefficient, and the extracted point cloud density and leaf area index for the corresponding layers, a complete spatial distribution characteristic of the canopy point clouds was obtained, providing basic data for subsequent calculations of spatial occupancy.

[0060] Next, based on the determined 3D spatial distribution pattern of the canopy point cloud, the appropriate height weighting function type is determined. For example, a linearly increasing height weighting function is used for the upper-layer clustered tree canopy; a normal distribution height weighting function is used for the middle-layer clustered shrub canopy; and an equal-weighted height weighting function is used for the uniformly distributed grassland canopy. In the linearly increasing height weighting function, the height weight value at a corresponding relative height is equal to the relative canopy height of the canopy point cloud divided by the total canopy height, and the weight value increases linearly from 0 to 1 as the relative height increases. In the normal distribution height weighting function, the height weight value at a corresponding relative height is calculated with the natural constant as the base and the square of the difference between the negative relative canopy height and the center height of the peak height interval divided by the square of twice the standard deviation of the canopy height distribution sequence as the exponent. The weight value reaches its maximum at the peak height and gradually decreases upwards and downwards. In the equal-weighted height weighting function, the weight value corresponding to all heights within the total canopy height range remains at 1. For the three intervals of the canopy (upper, middle, and lower), the weight values ​​corresponding to all subdivided height statistical intervals within each interval are extracted. The weighting coefficient is the ratio of the number of valid canopy point clouds in each subdivided height statistical interval to the total number of point clouds in the corresponding interval. The weight values ​​of all subdivided height statistical intervals within the interval are weighted and averaged to obtain the layer weight coefficient of the corresponding interval. When there are no valid canopy point clouds in a certain interval, the layer weight coefficient of that interval is assigned to 0.

[0061] For the upper, middle, and lower canopy layers, horizontal projection grid data for each layer is extracted. Each grid cell is binarized, with a value of 1 for cells containing valid canopy point clouds and 0 for cells without valid canopy point clouds. The total number of grid cells with a value of 1 in each layer is counted, and this number is divided by the total number of grid cells in the corresponding layer's horizontal projection plane to obtain the horizontal occupancy rate of that layer. The total number of valid canopy point clouds in each layer is counted, and this number is divided by the total number of valid canopy point clouds in the entire canopy to obtain the vertical point cloud percentage of that layer. The point cloud density and leaf area index of the corresponding layer, which have undergone min-max normalization, are retrieved, and their arithmetic mean is used as the structure correction coefficient. The horizontal occupancy rate, vertical point cloud percentage, and structure correction coefficient of each layer are multiplied consecutively to obtain the spatial occupancy rate of that height layer. When there are no valid canopy point clouds in a certain layer, the spatial occupancy rate of that layer is set to 0. The spatial occupancy rate calculations for the upper, middle, and lower canopy layers are completed sequentially.

[0062] Next, the hierarchical weight coefficients corresponding to the upper, middle, and lower canopy layers, as well as the spatial occupancy rates of the corresponding layers, are retrieved. A formula for calculating structural coverage is constructed using a hierarchical weighted summation method. The calculated structural coverage value is equal to the sum of the products of the hierarchical weight coefficients of each canopy layer and their corresponding spatial occupancy rates, with the summation covering the upper, middle, and lower canopy layers. For mixed vegetation types, the spatial distribution pattern and suitable height weight function of each vegetation type are determined based on the proportion of point cloud quantity within the canopy, and the corresponding hierarchical weight coefficients are calculated. Then, the hierarchical weight coefficients of the mixed vegetation are obtained by weighting according to the proportion of vegetation types, and substituted into the formula to complete the construction of the structural coverage calculation model for mixed vegetation types. The accuracy of the constructed model was verified using ground-measured canopy structure coverage data. The ground-measured data was obtained by combining quadrat survey with a canopy analyzer. No fewer than thirty sample quadrats were selected in the vegetation area to be measured, and the total coverage ratio of the canopy on the horizontal projection plane in each quadrat was measured as the measured baseline value. When the coefficient of determination between the model calculated value and the measured baseline value is greater than or equal to 0.8, and the root mean square error is less than or equal to 0.08, the model is deemed valid and can be used for quantitative calculation of the structure coverage of the vegetation area to be measured.

[0063] By clarifying the acquisition process of canopy height distribution and point cloud spatial distribution characteristics, the quantitative judgment criteria of canopy spatial distribution patterns, the specific expression of the height weight function and scene adaptation rules, the quantitative calculation method of space occupancy rate and the complete logic of vertical layer weighting model construction, the accurate quantitative representation of the three-dimensional spatial structure of vegetation canopy is realized, which effectively improves the accuracy and scene adaptability of vegetation structure coverage calculation.

[0064] Furthermore, the method provided in this application embodiment includes: A vegetation cover-growth status mapping table is constructed to define the vegetation health level corresponding to different coverage ranges. Based on the vegetation cover-growth status mapping table, the vegetation growth status is evaluated by combining the spatial distribution of effective photosynthetic coverage and structural coverage, and a vegetation health status heat map is generated. When the vegetation health status heat map detects that the coverage of a local area is lower than a preset threshold, an abnormal growth warning is triggered.

[0065] In one embodiment, firstly, effective photosynthetic coverage and structural coverage data for each interval of the upper, middle, and lower canopy layers, which have already undergone spatial coordinate matching, are retrieved, along with continuous raster data of effective photosynthetic coverage and structural coverage across the entire area. The numerical distribution intervals of the two types of coverage are statistically analyzed, confirming that the effective numerical range for both types of coverage is 0 to 1. Combining the forestry industry standard "Technical Regulations for Forest Resource Planning, Design, and Survey" with general specifications in the field of vegetation remote sensing monitoring, and the growth characteristics of the vegetation type to be measured, the vegetation health level is divided into five consecutive levels: excellent health, good health, moderate health, poor health, and extremely poor health.

[0066] For the three main vegetation types—trees, shrubs, and grasslands—numerical ranges for effective photosynthetic cover and structural cover were determined for each health level. Specifically, excellent health corresponds to both cover values ​​being greater than or equal to 0.8; good health corresponds to both cover values ​​being greater than or equal to 0.6 and less than 0.8; moderate health corresponds to both cover values ​​being greater than or equal to 0.4 and less than 0.6; poor health corresponds to both cover values ​​being greater than or equal to 0.2 and less than 0.4; and very poor health corresponds to both cover values ​​being less than 0.2. Cover values ​​exactly at the threshold of the range are assigned to the lower health level range.

[0067] For mixed vegetation types, the weighting coefficients for the coverage intervals of each vegetation type are first determined based on the proportion of point cloud data of different vegetation types within the canopy. A weighted average is then used to calculate the coverage value intervals corresponding to each health level of the mixed vegetation. Differential weighting coefficients are set for the coverage data of different canopy layers: the weighting coefficient for effective photosynthetic coverage in the upper canopy is set to 0.5, in the middle canopy to 0.3, and in the lower canopy to 0.2. Structural coverage is determined using the arithmetic mean of the three canopy layers. When effective photosynthetic coverage and structural coverage belong to different level intervals, the lower level is used as the initial health level for that area. If one type of coverage is in the poor health range, regardless of the other type of coverage, the area is directly classified as being in the poor health range. This process completes the construction of a coverage-growth status mapping table for different vegetation types and canopy layers, establishing a one-to-one correspondence between coverage values ​​and vegetation health levels.

[0068] Next, the vegetation photosynthetic intensity is assessed based on the spatial distribution characteristics of effective photosynthetic coverage, and the vegetation space occupancy is assessed simultaneously based on the spatial distribution characteristics of structural coverage. Then, the two assessment results of vegetation photosynthetic intensity and space occupancy are combined with the constructed coverage-growth status mapping table to complete the vegetation growth status assessment, and finally generate a vegetation health status heat map. This step will be explained in detail in the following content.

[0069] Next, a pre-set coverage warning threshold is established. This threshold corresponds to the critical value between the poor health level and the extremely poor health level in the coverage-growth status mapping table. The basic warning threshold is set to 0.2. For special scenarios such as deciduous vegetation and grasslands in arid and semi-arid regions, the warning threshold can be lowered by 0.05 to 0.1. For trees in their full-leaf stage, the warning threshold can be raised by 0.05 to 0.1. A sliding window matching the pixel resolution of the multispectral imaging device is used to scan the entire area of ​​the generated vegetation health status heatmap row by row and column by column. The sliding window size is set to 3 pixels × 3 pixels. The weighted average of the effective photosynthetic coverage and the average of the structural coverage of all pixels within each sliding window are calculated.

[0070] When the sliding window scans to the boundary of the area to be measured, if there are fewer than 3×3 pixels at the boundary, the single-pixel coverage value is directly compared with the warning threshold and included in the warning detection range. When the weighted average of effective photosynthetic coverage or the average of structural coverage within a single sliding window is lower than the warning threshold, the location of the window is marked. When three consecutive adjacent sliding windows in the horizontal, vertical, or diagonal directions are marked, the spatial area corresponding to the consecutive windows is determined to be a vegetation growth abnormality area, automatically triggering a growth abnormality warning, and simultaneously recording and outputting complete information such as the spatial coordinates of the abnormal area, the average coverage value, and the corresponding vegetation health level.

[0071] Furthermore, the method provided in this application embodiment includes: Based on the spatial distribution characteristics of the effective photosynthetic coverage, the vegetation photosynthetic intensity is assessed, and based on the spatial distribution characteristics of the structural coverage, the vegetation space occupancy status is assessed; combining the photosynthetic intensity and the vegetation space occupancy status, a vegetation health status heatmap is generated.

[0072] Optionally, a quantitative relationship model between effective photosynthetic coverage and photosynthetic intensity is first established. Then, based on this quantitative relationship model and the spatial distribution characteristics of effective photosynthetic coverage, regions with high photosynthetic efficiency and regions with low photosynthetic efficiency are identified, and a spatial distribution map of photosynthetic intensity is generated. At the same time, the spatial heterogeneity characteristics of structural coverage are analyzed, regions with sparse vegetation cover and regions with dense vegetation cover are identified, and an assessment map of vegetation spatial distribution is generated. This step will be explained in detail in the following sections.

[0073] Next, the spatial distribution map of photosynthetic intensity and the assessment map of vegetation spatial distribution, which have been unified with the spatial coordinate reference, are retrieved. Simultaneously, the constructed cover-growth status mapping table, along with the corresponding effective photosynthetic cover raster data, structural cover raster data, and canopy layer cover data for the entire area, are retrieved. Spatial registration processing is performed on all raster data. The nearest neighbor interpolation method is used to uniformly adjust the pixel resolution, spatial coordinate system, and projection reference of all raster data to be completely consistent with the original data acquired by the multispectral imaging device, ensuring that the pixel positions of all data correspond one-to-one without spatial offset. The photosynthetic intensity values ​​in the spatial distribution map of photosynthetic intensity and the structural cover values ​​corresponding to the vegetation spatial distribution assessment map are processed using the min-max normalization method, uniformly mapping both types of values ​​to the interval between 0 and 1, eliminating the impact of dimensional differences on subsequent comprehensive assessment.

[0074] Then, the weighting rules for the vegetation physiological activity dimension and the vegetation structure dimension were determined. The base weighting coefficient for the vegetation physiological activity dimension, corresponding to photosynthetic intensity, was set to 0.6, and the base weighting coefficient for the vegetation spatial occupancy dimension, corresponding to structural coverage, was set to 0.4. Differential adjustments were made to the weighting coefficients for different vegetation types. For example, the weighting coefficient for the physiological activity dimension of arbor vegetation was adjusted to 0.65, and the weighting coefficient for the structure dimension was adjusted to 0.35; the base weighting coefficient for shrub vegetation remained unchanged; and the weighting coefficient for the physiological activity dimension of grassland vegetation was adjusted to 0.55, and the weighting coefficient for the structure dimension was adjusted to 0.45.

[0075] For mixed vegetation types, the proportion of point cloud data for trees, shrubs, and grasslands within the mixed area is first statistically analyzed using canopy 3D point cloud data. Using the proportion of point cloud data for each type of vegetation as a weight, a weighted average is calculated for the dimensional weight coefficients of the corresponding vegetation type to obtain the appropriate physiological activity and structural dimension weight coefficients for the mixed vegetation area. Based on these weight coefficients, a weighted summation operation is performed pixel by pixel. The normalized photosynthetic intensity value is multiplied by the corresponding physiological activity dimension weight coefficient, and the normalized structural coverage value is added multiplied by the corresponding structural dimension weight coefficient to obtain the comprehensive vegetation health index for each pixel location, generating continuous raster data of the comprehensive vegetation health index for the entire region.

[0076] Based on the established coverage-growth status mapping table, the basic numerical ranges of the comprehensive health index corresponding to the five vegetation health levels were defined. For different phenological stages of vegetation, the numerical ranges corresponding to the health levels were adjusted accordingly. For trees and shrubs, the lower limit of the numerical range corresponding to each health level during the leaf-spreading stage was uniformly lowered by 0.1, and the lower limit during the leaf-falling stage was uniformly lowered by 0.15. For grasslands, the basic numerical range remained unchanged during the peak growing season, and the lower limit during the withering stage was uniformly lowered by 0.2. The comprehensive vegetation health index of each pixel was matched one-to-one with the corresponding vegetation type and phenological stage numerical range in the coverage-growth status mapping table to determine the vegetation health level corresponding to each pixel. When the comprehensive vegetation health index was exactly at the interval threshold, it was assigned to the lower health level interval. Standardized color mapping values ​​were set for each of the five health levels. Based on the health level obtained for each pixel, the corresponding color value was assigned pixel-by-pixel, generating continuous vegetation health level assigned raster data across the entire area.

[0077] Finally, based on the acquired vegetation health level assignment raster data, a linear gradient color mapping rule was used to convert discrete health level assignments into continuous color gradient effects. Combined with geographic coordinate information and projection benchmarks consistent with the original data, a visualized vegetation health status heatmap was generated. The spatial coverage and pixel resolution of this heatmap perfectly matched the original data acquired by the multispectral imaging device. The measured vegetation health level data obtained through ground quadrat surveys were used to verify the accuracy of the generated vegetation health status heatmap. At least twenty evenly distributed verification sample points were selected within the vegetation area to be measured. The vegetation health level at the corresponding sample point location in the heatmap was compared with the measured health level on the ground. When the level matching accuracy was greater than or equal to 85%, the heatmap was deemed valid, completing the final generation and output of the vegetation health status heatmap.

[0078] Furthermore, the method provided in this application embodiment includes: A quantitative relationship model between effective photosynthetic coverage and photosynthetic intensity is established. Based on the quantitative relationship model, regions with high photosynthetic efficiency and regions with low photosynthetic efficiency are identified according to spatial distribution characteristics, and a spatial distribution map of photosynthetic intensity is generated. The spatial heterogeneity characteristics of the structural coverage are analyzed to identify sparse and dense vegetation coverage regions, and an assessment map of vegetation spatial distribution is generated.

[0079] In one embodiment, effective photosynthetic coverage data for each interval of the upper, middle, and lower canopy layers, after spatial coordinate matching, are first retrieved. Simultaneously, measured photosynthetic intensity data for the corresponding locations in the vegetation area to be tested are acquired. Photosynthetic intensity is quantified using the net photosynthetic rate of leaves. A portable photosynthetic measurement system is used, selecting no fewer than thirty sample points within the vegetation area to be tested. Net photosynthetic rate data of functional leaves at each sample point corresponding to the canopy layer are collected as the measured baseline value for photosynthetic intensity. During the measurement process, three to five fully expanded, disease-free, and mechanically undamaged functional leaves are selected for each sample point corresponding to the canopy layer. The measurement time window is selected from 9:00 AM to 11:30 AM on a sunny day, and the ambient photosynthetically active radiation is kept stable at 1000 ppm during the measurement process. The ambient temperature is controlled between 25 and 30 degrees Celsius, and the relative humidity is controlled between 40 and 60 percent to ensure the stability and consistency of the measured data.

[0080] Then, the effective photosynthetic coverage and the corresponding measured photosynthetic intensity for each sample point were paired. Outlier samples were removed using the Grubbs' test, and a significance level of 0.05 was set to complete the preprocessing of the sample data. A univariate linear regression analysis method was used, with effective photosynthetic coverage as the independent variable and the measured photosynthetic intensity as the dependent variable, to construct a quantitative relationship model between effective photosynthetic coverage and photosynthetic intensity. The general formula of the model is: Where PN is the estimated photosynthetic intensity, FPEC is the effective photosynthetic coverage, e is the regression slope coefficient, and f is the regression intercept coefficient. The regression coefficients are solved using the least squares method. For the three main vegetation types—trees, shrubs, and grasslands—quantitative relationship models are constructed for each vegetation type. For mixed vegetation types, the regression coefficients of each vegetation type are weighted and averaged according to the proportion of point cloud data of different vegetation types within the canopy to obtain model parameters suitable for mixed vegetation. The model accuracy is verified using at least twenty reserved validation sample points. The model is considered valid when the coefficient of determination between the calculated and measured values ​​is greater than or equal to 0.7 and the root mean square error is less than or equal to 0.1, thus establishing the quantitative relationship model between effective photosynthetic coverage and photosynthetic intensity.

[0081] Next, effective photosynthetic coverage raster data corresponding to each interval of the upper, middle, and lower layers of the canopy in the vegetation area to be tested are retrieved. Combined with the canopy layer weighting coefficient, continuous effective photosynthetic coverage weighted raster data is calculated for the entire area. The weighted raster data is then input pixel by pixel into a quantized relational model that has passed validity verification to complete the model calculation and obtain the estimated value of photosynthetic intensity corresponding to each pixel location, generating continuous photosynthetic intensity raster data for the entire area. Based on the growth characteristics, phenological period, and photosynthetic intensity numerical distribution characteristics of the vegetation type to be tested, photosynthetic intensity grading thresholds are set. The upper and lower limits of the basic grading thresholds are set to 1.2 times and 0.8 times the average photosynthetic intensity of the entire area, respectively. For example, for arbor vegetation, the upper and lower limits of the grading threshold during the peak leafing period are adjusted to 1.3 times and 0.7 times the average value of the entire region, respectively, and the upper and lower limits of the grading threshold during the leaf-falling period are adjusted to 1.1 times and 0.9 times the average value of the entire region, respectively; for shrub vegetation, the basic grading threshold remains unchanged; for grassland vegetation, the upper and lower limits of the grading threshold during the peak growing season are adjusted to 1.25 times and 0.75 times the average value of the entire region, respectively, and the upper and lower limits of the grading threshold during the yellowing period are adjusted to 1.1 times and 0.9 times the average value of the entire region, respectively.

[0082] Then, regions with photosynthetic intensity values ​​greater than or equal to the upper limit of the grading threshold are identified as high photosynthetic efficiency regions, while regions with photosynthetic intensity values ​​less than or equal to the lower limit of the grading threshold are identified as low photosynthetic efficiency regions. Different values ​​are assigned to regions with different photosynthetic efficiency levels, and combined with a spatial coordinate reference consistent with the multispectral imaging device, a continuous spatial distribution map of photosynthetic intensity over the entire region is generated, completing the spatial quantification and regional identification of vegetation photosynthetic intensity.

[0083] Subsequently, continuous structural cover raster data of the entire vegetation area under test were retrieved. The spatial heterogeneity characteristics of structural cover were analyzed using the moving window method. The basic size of the moving window was set to 3 pixels × 3 pixels, consistent with the pixel resolution of the multispectral imaging device. For high-resolution multispectral data with a spatial resolution higher than 1 meter, the moving window size was adjusted to 5 pixels × 5 pixels; for low-resolution multispectral data with a spatial resolution lower than 5 meters, the basic window size remained unchanged; for large-scale monitoring scenarios at the county level, the moving window size could be adjusted to 7 pixels × 7 pixels.

[0084] A pre-defined moving window is used to scan the entire area of ​​the structural cover raster data row by row and column by column. The coefficient of variation (COP) of the structural cover values ​​within each moving window is calculated to characterize the spatial heterogeneity of structural cover in the corresponding area. The larger the COP, the greater the difference in the spatial distribution of vegetation cover in the corresponding area. When the moving window reaches the boundary of the area to be measured, if there are not enough pixels of the pre-defined window size at the boundary, the local heterogeneity is calculated directly using the structural cover values ​​of a single pixel and included in the coverage degree identification range. Based on the global numerical distribution characteristics of structural cover and the vegetation type to be measured, vegetation cover degree classification thresholds are set. The upper and lower limits of the basic classification thresholds are set to 1.2 times and 0.8 times the average value of the global structural cover, respectively. For arbor vegetation, the upper and lower limits of the classification thresholds are adjusted to 1.25 times and 0.75 times the average value of the global structural cover, respectively.

[0085] Finally, areas with structural cover values ​​greater than or equal to the upper limit of the grading threshold are identified as densely vegetated areas, while areas with structural cover values ​​less than or equal to the lower limit of the grading threshold are identified as sparsely vegetated areas. Different values ​​are assigned to areas with different cover levels, and combined with a spatial coordinate reference consistent with the multispectral imaging device, a continuous vegetation spatial distribution assessment map is generated across the entire region, thereby completing the quantitative assessment and regional identification of vegetation spatial occupancy.

[0086] In summary, the vegetation cover monitoring method integrating LiDAR and optical remote sensing provided in this application has the following technical effects: This application deploys LiDAR sensors and multispectral imaging devices in the vegetation area to be tested, simultaneously acquiring three-dimensional point cloud and spectral reflectance data, extracting vegetation vertical structure and physiological activity parameters, constructing a layered coverage model to generate effective photosynthetic coverage and structural coverage, and combining spatial distribution to complete the assessment of vegetation growth status, making vegetation coverage monitoring results more accurate and reliable. It achieves integrated monitoring of vegetation spatial structure and physiological activity characteristics, comprehensively and accurately reflecting the actual coverage and growth of vegetation.

[0087] Example 2, as Figure 2 As shown, based on the same inventive concept as in Embodiment 1 above, this application provides a vegetation cover monitoring system integrating LiDAR and optical remote sensing, the system comprising: Spatiotemporal synchronous acquisition module construction module 1 is used to deploy LiDAR sensors and multispectral imaging devices in the vegetation area to be measured, and to establish a spatiotemporal synchronous acquisition module, wherein the LiDAR sensors and multispectral imaging devices maintain a fixed spatial relationship through rigid supports.

[0088] The monitoring data acquisition module 2, based on the spatiotemporal synchronous acquisition module, acquires three-dimensional point cloud data of vegetation through the LiDAR sensor and acquires spectral reflectance data of vegetation through the multispectral imaging device.

[0089] The physiological activity parameter acquisition module 3 is used to extract the vertical structure parameters of vegetation based on the three-dimensional point cloud data, and calculate the physiological activity parameters of vegetation in combination with the spectral reflectance data.

[0090] The layered coverage acquisition module 4 constructs a layered coverage model based on the vertical structure parameters and physiological activity parameters, and generates the effective photosynthetic coverage and structural coverage of the vegetation.

[0091] Coverage monitoring result acquisition module 5 is used to assess the vegetation growth status based on the spatial distribution of the effective photosynthetic coverage and structural coverage, and output the coverage monitoring results.

[0092] Furthermore, the physiological activity parameter acquisition module 3 is used to perform the following steps: The three-dimensional point cloud data is subjected to high-level layering to divide the canopy into upper, middle and lower layers, and the vertical structure parameters of each layer are extracted; the spectral reflectance data is subjected to band calculation to extract the chlorophyll index and water content index as physiological activity parameters.

[0093] Furthermore, the physiological activity parameter acquisition module 3 is used to perform the following steps: The three-dimensional point cloud data is divided into three intervals according to height: upper, middle and lower. The point cloud density and leaf area index in each interval are calculated as vertical structure parameters.

[0094] Furthermore, the layer coverage acquisition module 4 is used to perform the following steps: Based on the point cloud density and leaf area index in the vertical structure parameters, and combined with the chlorophyll index in the physiological activity parameters, a photosynthetic effective coverage calculation model is established; based on the canopy height distribution and point cloud spatial distribution characteristics in the vertical structure parameters, a structural coverage calculation model is established; effective photosynthetic coverage is generated through the photosynthetic effective coverage calculation model, and structural coverage is generated through the structural coverage calculation model.

[0095] Furthermore, the layer coverage acquisition module 4 is used to perform the following steps: Based on the regression equation between the chlorophyll index and photosynthetic efficiency, a photosynthetic efficiency estimation model is established; the point cloud density and leaf area index are introduced into the photosynthetic efficiency estimation model as weighting factors; regression analysis is performed on the photosynthetic efficiency estimation model to establish a photosynthetic effective coverage calculation model that includes point cloud density, leaf area index and chlorophyll index.

[0096] Furthermore, the layer coverage acquisition module 4 is used to perform the following steps: Based on the canopy height distribution, the three-dimensional spatial distribution pattern of the canopy point cloud is analyzed, and a height weighting function is established. The spatial occupancy rate of each height layer is calculated according to the spatial distribution characteristics of the point cloud. Based on the height weighting function and the spatial occupancy rate of each height layer, vertical layer weighting is performed to generate the structure coverage calculation model.

[0097] Furthermore, the coverage monitoring result acquisition module 5 is used to perform the following steps: A vegetation cover-growth status mapping table is constructed to define the vegetation health level corresponding to different coverage ranges. Based on the vegetation cover-growth status mapping table, the vegetation growth status is evaluated by combining the spatial distribution of effective photosynthetic coverage and structural coverage, and a vegetation health status heat map is generated. When the vegetation health status heat map detects that the coverage of a local area is lower than a preset threshold, an abnormal growth warning is triggered.

[0098] Furthermore, the coverage monitoring result acquisition module 5 is used to perform the following steps: Based on the spatial distribution characteristics of the effective photosynthetic coverage, the vegetation photosynthetic intensity is assessed, and based on the spatial distribution characteristics of the structural coverage, the vegetation space occupancy status is assessed; combining the photosynthetic intensity and the vegetation space occupancy status, a vegetation health status heatmap is generated.

[0099] Furthermore, the coverage monitoring result acquisition module 5 is used to perform the following steps: A quantitative relationship model between effective photosynthetic coverage and photosynthetic intensity is established. Based on the quantitative relationship model, regions with high photosynthetic efficiency and regions with low photosynthetic efficiency are identified according to spatial distribution characteristics, and a spatial distribution map of photosynthetic intensity is generated. The spatial heterogeneity characteristics of the structural coverage are analyzed to identify sparse and dense vegetation coverage regions, and an assessment map of vegetation spatial distribution is generated.

[0100] The vegetation cover monitoring system integrating LiDAR and optical remote sensing provided in this embodiment of the invention can execute the vegetation cover monitoring method integrating LiDAR and optical remote sensing provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0101] Although this application makes various references to certain modules in the system according to the embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy distinction between each other and are not used to limit the scope of protection of this invention.

[0102] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application. In some cases, the actions or steps described in this application can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A vegetation cover monitoring method integrating LiDAR and optical remote sensing, characterized in that, The method includes: LiDAR sensors and multispectral imaging devices are deployed in the vegetation area to be tested, and a spatiotemporal synchronous acquisition module is established. The LiDAR sensors and multispectral imaging devices maintain a fixed spatial relationship through rigid supports. Based on the spatiotemporal synchronous acquisition module, the three-dimensional point cloud data of vegetation is acquired through the LiDAR sensor, and the spectral reflectance data of vegetation is acquired through the multispectral imaging device. The vertical structure parameters of the vegetation are extracted based on the three-dimensional point cloud data, and the physiological activity parameters of the vegetation are calculated by combining the spectral reflectance data. Based on the vertical structure parameters and physiological activity parameters, a layered coverage model is constructed to generate the effective photosynthetic coverage and structural coverage of vegetation. Based on the spatial distribution of the effective photosynthetic coverage and structural coverage, the vegetation growth status is assessed, and the coverage monitoring results are output.

2. The vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in claim 1, characterized in that, Vertical structure parameters of vegetation are extracted from the three-dimensional point cloud data, and physiological activity parameters of vegetation are calculated by combining the spectral reflectance data, including: The three-dimensional point cloud data is subjected to height layering processing to divide it into upper, middle and lower canopy layers, and the vertical structure parameters of each layer are extracted. Band operations were performed on the spectral reflectance data to extract the chlorophyll index and water content index as physiological activity parameters.

3. The vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in claim 2, characterized in that, The 3D point cloud data is subjected to height-level layering, dividing it into upper, middle, and lower canopy layers, and the vertical structure parameters of each layer are extracted, including: The three-dimensional point cloud data is divided into three intervals according to height: upper layer, middle layer and lower layer. Point cloud density and leaf area index within each interval layer are calculated as vertical structure parameters.

4. The vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in claim 3, characterized in that, Based on the aforementioned vertical structure parameters and physiological activity parameters, a stratified coverage model is constructed to generate the effective photosynthetic coverage and structural coverage of vegetation, including: Based on the point cloud density and leaf area index in the vertical structure parameters, and combined with the chlorophyll index in the physiological activity parameters, a calculation model for photosynthetic effective coverage is established. Based on the canopy height distribution and point cloud spatial distribution characteristics in the vertical structure parameters, a structural coverage calculation model is established. Effective photosynthetic coverage is generated using the photosynthetic effective coverage calculation model, and structural coverage is generated using the structural coverage calculation model.

5. The vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in claim 4, characterized in that, Based on the point cloud density and leaf area index in the vertical structure parameters, and combined with the chlorophyll index in the physiological activity parameters, a calculation model for photosynthetically effective coverage is established, including: Based on the regression equation between the chlorophyll index and photosynthetic efficiency, a photosynthetic efficiency estimation model is established. The point cloud density and leaf area index are used as weighting factors in the photosynthetic efficiency estimation model. Regression analysis was performed on the photosynthetic efficiency estimation model to establish a photosynthetic effective coverage calculation model that includes point cloud density, leaf area index, and chlorophyll index.

6. The vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in claim 4, characterized in that, Based on the canopy height distribution and point cloud spatial distribution characteristics in the vertical structure parameters, a structure coverage calculation model is established, including: Based on the canopy height distribution, the three-dimensional spatial distribution pattern of the canopy point cloud is analyzed, and a height weighting function is established. Calculate the spatial occupancy rate of each height layer based on the spatial distribution characteristics of the point cloud; Based on the height weight function and the space occupancy rate of each height layer, vertical layer weighting is performed to generate the structure coverage calculation model.

7. The vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in claim 1, characterized in that, Based on the spatial distribution of effective photosynthetic cover and structural cover, vegetation growth status is assessed, and cover monitoring results are output, including: Construct a vegetation cover-growth status mapping table and define the vegetation health levels corresponding to different coverage ranges; Based on the coverage-growth status mapping table, and combined with the spatial distribution of effective photosynthetic coverage and structural coverage, vegetation growth status is assessed, and a vegetation health status heat map is generated. When the vegetation health status heatmap detects that the coverage of a local area is lower than a preset threshold, an abnormal growth warning is triggered.

8. The vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in claim 7, characterized in that, Based on the aforementioned cover-growth status mapping table, and combined with the spatial distribution of effective photosynthetic cover and structural cover, vegetation growth status is assessed, generating a vegetation health status heatmap, including: Based on the spatial distribution characteristics of the effective photosynthetic coverage, the intensity of vegetation photosynthesis is evaluated, and based on the spatial distribution characteristics of the structural coverage, the spatial occupancy of vegetation is evaluated. By combining the photosynthetic intensity and vegetation space occupancy, a heat map of vegetation health status is generated.

9. The vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in claim 8, characterized in that, Based on the spatial distribution characteristics of the effective photosynthetic coverage, the vegetation photosynthetic intensity is assessed, and based on the spatial distribution characteristics of the structural coverage, the vegetation spatial occupancy status is assessed, including: Establish a quantitative relationship model between effective photosynthetic coverage and photosynthetic intensity; Based on the quantification relationship model, high photosynthetic efficiency regions and low photosynthetic efficiency regions are identified according to spatial distribution characteristics, and a spatial distribution map of photosynthetic intensity is generated. Analyze the spatial heterogeneity characteristics of the structural coverage, identify sparse and dense vegetation areas, and generate an assessment map of vegetation spatial distribution.

10. A vegetation cover monitoring system integrating LiDAR and optical remote sensing, characterized in that, The system for implementing the vegetation cover monitoring method integrating LiDAR and optical remote sensing as described in any one of claims 1-9 includes: A spatiotemporal synchronous acquisition module construction module is used to deploy LiDAR sensors and multispectral imaging devices in the vegetation area to be measured, and to establish a spatiotemporal synchronous acquisition module, wherein the LiDAR sensors and multispectral imaging devices maintain a fixed spatial relationship through rigid supports. The monitoring data acquisition module, based on the spatiotemporal synchronous acquisition module, acquires three-dimensional point cloud data of vegetation through the LiDAR sensor and acquires spectral reflectance data of vegetation through the multispectral imaging device. The physiological activity parameter acquisition module is used to extract the vertical structure parameters of vegetation based on the three-dimensional point cloud data, and calculate the physiological activity parameters of vegetation in combination with the spectral reflectance data. The layered coverage acquisition module constructs a layered coverage model based on the vertical structure parameters and physiological activity parameters, and generates the effective photosynthetic coverage and structural coverage of the vegetation. The coverage monitoring result acquisition module is used to assess the vegetation growth status based on the spatial distribution of the effective photosynthetic coverage and structural coverage, and output the coverage monitoring results.

Images