Intelligent forestry monitoring system

Abstract

The invention discloses an intelligent forestry monitoring system, and relates to the technical field of forestry optical monitoring, and the system comprises a multispectral data collection module which is configured to collect optical data by using the optical means of a hyperspectral camera, an infrared thermal imager and LiDAR equipment, a parameter inversion module, calculate a vegetation physiology-structure comprehensive health index, and calculate the vegetation physiology-structure comprehensive health index. The matching degree of a real-time spectrum and a health baseline is calculated in combination with a healthy vegetation spectrum fingerprint database to obtain a spectrum anomaly recognition index, and the risk assessment module is configured to couple the spectrum anomaly recognition index, a pathogenic bacterium development accumulated temperature ratio and environment humidity, calculate a pest and disease outbreak risk index through a time sequence pest and disease outbreak risk index algorithm, and evaluate the disease and disease outbreak risk index. The decision intervention module is configured to divide risk levels according to the pest and disease damage outbreak risk indexes, generate a spatial distribution diagram and output intervention measures, full-chain automation is formed from optical data collection to risk early warning-decision intervention, and then dynamic balance of high-precision early warning-low-cost monitoring is achieved.

Description

Technical Field

[0001] This invention relates to the field of forestry optical monitoring technology, specifically a smart forestry monitoring system. Background Technology

[0002] As a core component of the ecosystem, the health of forestry directly affects biodiversity conservation, carbon sequestration capacity, and sustainable economic development.

[0003] Traditional forestry monitoring relies on manual surveys, which suffers from low efficiency, strong subjectivity, and poor timeliness. While existing monitoring methods have gradually incorporated technologies such as UAV hyperspectral and satellite remote sensing with the development of optical remote sensing, they still face the following shortcomings: First, most existing vegetation indices rely solely on visible-near-infrared reflectance, without considering differences in vegetation structure, making it difficult to achieve synergistic physiological-structural assessment. Second, existing spectral matching algorithms do not consider environmental factors, resulting in "pseudo-anomalies," making it difficult to achieve dynamic adaptation of spectral features to the microenvironment. Finally, existing pest and disease risk models often rely on single factors, such as humidity or temperature, or use simple weighting, without considering the nonlinear coupling effect and time-series cumulative effect between factors, making it difficult to capture the multi-factor threshold synergistic characteristics of pest and disease outbreaks. Summary of the Invention

[0004] The purpose of this invention is to provide a smart forestry monitoring system that solves the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution, comprising: The multispectral data acquisition module is configured to use optical means such as hyperspectral cameras, infrared thermal imagers and LiDAR equipment to acquire optical data including multi-band reflectivity data of vegetation canopy, thermal infrared radiation data of soil surface and three-dimensional structure data of canopy. The parameter inversion module is configured to preprocess the acquired optical data and calculate vegetation health and anomalous parameters through the following steps: Based on multi-band reflectance data and canopy three-dimensional structure data, and using the enhanced vegetation index algorithm, the comprehensive health index of vegetation physiology and structure is calculated. The leaf area index is obtained by inverting the canopy three-dimensional structure data through the gap rate model of LiDAR point cloud. By combining the spectral fingerprint database of healthy vegetation, the matching degree between the real-time spectrum and the healthy baseline is calculated through spectral angle cosine similarity, and a normalization correction term for soil surface thermal infrared radiation data is introduced to obtain the spectral anomaly identification index. The risk assessment module is configured to couple the spectral anomaly identification index, the pathogen development accumulated temperature ratio, and the environmental humidity, and calculate the pest outbreak risk index through the time-series pest risk index algorithm. Among them, the accumulated temperature ratio for pathogen development is calculated by accumulating the daily average temperature data in the soil surface thermal infrared radiation data, and the environmental humidity includes humidity data obtained through an optical humidity sensor. The decision intervention module is configured to classify risk levels based on the pest and disease outbreak risk index, generate spatial distribution maps, and output intervention measures. The risk levels include high risk, low risk, and medium risk. When the risk level is high, the collection cycle of the healthy vegetation spectral fingerprint database needs to be updated.

[0006] Optionally, the multi-band reflectance data includes 700-900nm near-infrared reflectance, 630-690nm red light reflectance, 450-520nm blue light reflectance, n sets of real-time spectral vectors, and fingerprint database reference spectral vectors; The soil surface thermal infrared radiation data includes measured soil temperature, historical average soil temperature for the same period, threshold accumulated temperature, and m-group daily average temperature and the temperature at which pests and diseases develop. The canopy three-dimensional structure data includes a 532nm green laser point cloud; The humidity data includes the normalized average relative humidity.

[0007] Optionally, the parameter inversion module includes: The inversion submodule is configured to use a LiDAR device to emit a 532nm green laser to acquire three-dimensional point cloud data of the canopy and simultaneously record the laser echo intensity and coordinate information. A cloth-simulation filtering algorithm was used to remove ground points while retaining the vegetation canopy point cloud. The gap ratio is calculated based on the proportion of laser penetration points under the canopy to the total point cloud. The leaf area index is calculated based on the gap ratio and the average leaf tilt angle factor according to the tree species type; The vegetation health calculation submodule is configured to take the near-infrared reflectance, the red light reflectance, the blue light reflectance, and the leaf area index as input, and calculate the comprehensive physiological-structural health index of vegetation through the following steps: The vegetation photosynthetic activity and chlorophyll content are calculated based on the difference between the near-infrared reflectance and the red reflectance. The weighted red light reflectance is added to the near-infrared reflectance to calculate and obtain the vegetation structure correction value that increases the red light weight to suppress saturation. After weighting the blue light reflectance, a correction benchmark value for reducing the blue light weight to correct the scattering environmental interference is calculated. The anti-interference benchmark value is obtained by subtracting the vegetation structure correction value from the environmental interference correction benchmark value. The basic physiological index is obtained by dividing the vegetation photosynthetic activity by the chlorophyll content by the anti-interference benchmark value. After weighting the basic physiological indices, a correction term (1+0.1×Y) of the leaf area index is introduced to compensate for the saturation phenomenon of the reflectance of high leaf area vegetation, and the comprehensive health index of vegetation physiology and structure is calculated. Where Y is the leaf area index.

[0008] Optionally, the parameter inversion module further includes: The spectral anomaly identification submodule is configured to compare the real-time spectral vector with the fingerprint database reference spectral vector and calculate the spectral angle matching degree using the spectral angle cosine function. After calculating the deviation between the measured soil temperature and the historical average for the same period, normalization is performed to obtain the soil temperature deviation index. The spectral similarity coupling index is calculated based on the product of the vegetation physiological-structural comprehensive health index and the spectral angle matching degree. The soil temperature deviation index is weighted and calculated to obtain the temperature correction index; The spectral anomaly identification index is calculated by adding the spectral similarity coupling index and the temperature correction index.

[0009] Optionally, the risk assessment module further includes an accumulated temperature calculation submodule and an early warning submodule. The accumulated temperature calculation submodule is configured to collect the threshold accumulated temperature and m groups of daily average temperature and the starting temperature of pest and disease development by an infrared thermal imager, accumulate the effective accumulated temperature within a 30-day sliding window, and calculate the ratio with the threshold accumulated temperature to obtain the accumulated temperature ratio. The early warning submodule is configured to calculate the pest and disease outbreak risk index using the spectral anomaly identification index, the accumulated temperature ratio, and the average relative humidity through the following steps: Using the aforementioned spectral anomaly identification index as the basic weight, it reflects the current degree of stress on the vegetation; The accumulated temperature ratio is multiplied by the spectral anomaly identification index to calculate the accumulated temperature coupling index, highlighting the risk during the active phase of pathogens; The average relative humidity is weighted and the humidity amplification index is calculated. The pest and disease outbreak risk index is obtained by calculating the product of the accumulated temperature coupling index and the humidity amplification index.

[0010] Optionally, the risk level classification and intervention measures are as follows: Low risk: Maintain the current base cycle; Medium risk: Increase ground-based verification sampling while maintaining the current basic cycle; High risk: Automatically generate a plan for felling diseased trees and spraying pesticides, and push it to the forestry management terminal. The plan includes accurate operation coordinates and dosage calculations, and calculates the update cycle by combining the pest and disease outbreak risk index and the basic cycle. The base period is the system's default spectral acquisition period, with an initial value of 30 days.

[0011] Optionally, the multispectral data acquisition module includes: The synchronization control unit is configured to achieve spatiotemporal registration of hyperspectral cameras, infrared thermal imagers, and LiDAR devices when authorized by GPS. The data preprocessing unit is configured to perform radiometric calibration, atmospheric correction, and point cloud denoising on the raw optical data.

[0012] Optionally, the real-time spectral vector is specifically the spectral reflectance of the current monitoring point obtained by preprocessing after being acquired by a hyperspectral camera; The fingerprint database reference spectral vector is specifically the standard spectral reflectance of healthy vegetation obtained through laboratory detached leaf spectral measurements and field calibration.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: I. This invention introduces a leaf area index correction term after weighting basic physiological indices to compensate for the saturation phenomenon of reflectance in high leaf area vegetation. This coupled design can simultaneously quantify the physiological state (chlorophyll, cell structure) and structural characteristics (canopy coverage) of vegetation. The gap rate model based on LiDAR point cloud can reduce the error of optical inversion.

[0014] Second, the spectral anomaly identification index of the present invention first calculates the cosine similarity based on hyperspectral data and the spectral fingerprint database of healthy vegetation, and then measures the soil temperature by infrared thermal imager. The high temperature interference is eliminated by normalization. In this way, by adding the spectral similarity coupling index and the temperature correction index as two factors, the synergistic judgment of spectral fingerprint matching and microenvironmental stress can be achieved.

[0015] Third, the pest and disease outbreak risk index of this invention is first based on the daily average temperature data of an infrared thermal imager, accumulating 30 days of effective accumulated temperature and comparing it with the pathogen development threshold. Then, based on the measured average relative humidity of an optical humidity sensor, the spore diffusion rate is quantified by weighting the average relative humidity. In this way, through the cascade coupling of the accumulated temperature coupling index and the humidity amplification index, the three key factors of spectral anomaly, pathogen development stage, and transmission conditions are integrated, which in particular improves the accuracy of early warning of pine wilt disease outbreaks and solves the problem of "lagging assessment of single factors". Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the detailed embodiments of the invention to explain the invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a system framework diagram of the intelligent forestry monitoring system; Figure 2 This is a schematic diagram of the parameter inversion module in this invention; Figure 3 This is a schematic diagram of the risk assessment module of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1, please refer to Figures 1 to 3 This embodiment provides a smart forestry monitoring system, comprising a multispectral data acquisition module, a parameter inversion module, a risk assessment module, and a decision intervention module, wherein: The multispectral data acquisition module is configured to use optical means such as hyperspectral cameras, infrared thermal imagers and LiDAR equipment to acquire optical data including multi-band reflectivity data of vegetation canopy, thermal infrared radiation data of soil surface and three-dimensional structure data of canopy. The multispectral data acquisition module includes: The synchronization control unit is configured to achieve spatiotemporal registration of hyperspectral cameras, infrared thermal imagers, and LiDAR devices when authorized by GPS. The data preprocessing unit is configured to perform radiometric calibration, atmospheric correction, and point cloud denoising on the raw optical data.

[0019] The multi-band reflectance data includes 700-900nm near-infrared reflectance, 630-690nm red light reflectance, 450-520nm blue light reflectance, n sets of real-time spectral vectors, and fingerprint database reference spectral vectors. Soil surface thermal infrared radiation data include measured soil temperature, historical average soil temperature for the same period, threshold accumulated temperature, and daily average temperature and pest and disease development initiation temperature of group m. The canopy three-dimensional structure data includes 532nm green laser point clouds.

[0020] In this embodiment, the system's multispectral data acquisition module collects multi-band reflectance data of the vegetation canopy using a hyperspectral camera (450-900nm band), simultaneously acquiring thermal infrared radiation (8-14μm) from the soil surface using an infrared thermal imager and 532nm green laser light emitted by a LiDAR device, thus achieving three-dimensional point cloud data acquisition. A GPS synchronization control unit is provided to achieve spatiotemporal registration of multiple devices, and combined with radiometric calibration, atmospheric correction, and point cloud denoising preprocessing, data accuracy is ensured.

[0021] Hyperspectral data: Acquire near-infrared (700-900nm), red (630-690nm), and blue (450-520nm) reflectance and n sets of real-time spectral vectors to capture the physiological and biochemical characteristics of vegetation; Thermal infrared data: Collects measured soil temperature, historical average for the same period, and daily average temperature of m groups to reflect changes in microenvironment heat; LiDAR data: Generate 532nm green laser point cloud and invert canopy three-dimensional structural parameters (such as leaf area index).

[0022] Thus, the multispectral data acquisition module realizes multi-dimensional collaborative acquisition of optical data, eliminating spatiotemporal registration errors (positioning accuracy ≤0.5m); it breaks through the limitations of a single sensor by fusing multi-source data, providing high signal-to-noise ratio input for subsequent parameter inversion; and the standardized preprocessing process ensures cross-platform comparability of data, providing basic data support for forestry monitoring in a three-in-one manner of "spectral-structure-thermal environment".

[0023] Please see Figures 1 to 2 The parameter inversion module is configured to preprocess the acquired optical data and calculate vegetation health and anomalous parameters through the following steps: Based on multi-band reflectance data and canopy three-dimensional structure data, and using the enhanced vegetation index algorithm, the comprehensive health index of vegetation physiology and structure is calculated. The leaf area index is obtained by inverting the canopy three-dimensional structure data through the gap rate model of LiDAR point cloud. By combining the spectral fingerprint database of healthy vegetation, the matching degree between the real-time spectrum and the healthy baseline is calculated through spectral angle cosine similarity, and a normalization correction term for soil surface thermal infrared radiation data is introduced to obtain the spectral anomaly identification index. Humidity data includes normalized average relative humidity.

[0024] The parameter inversion module includes: The inversion submodule is configured to use a LiDAR device to emit a 532nm green laser to acquire three-dimensional point cloud data of the canopy and simultaneously record the laser echo intensity and coordinate information. A cloth-simulation filtering algorithm was used to remove ground points while retaining the vegetation canopy point cloud. The gap ratio is calculated based on the proportion of laser penetration points under the canopy to the total point cloud. The leaf area index is calculated based on the gap ratio and the average leaf tilt angle factor according to tree species. The vegetation health calculation submodule is configured to take near-infrared reflectance, red light reflectance, blue light reflectance, and leaf area index as input, and calculate the comprehensive physiological-structural health index of vegetation through the following steps: The difference between near-infrared reflectance and red reflectance is used to calculate the photosynthetic activity and chlorophyll content of vegetation. The red light reflectance is weighted and then added to the near-infrared reflectance to calculate the vegetation structure correction value that increases the red light weight to suppress saturation. After weighting the blue light reflectance, the environmental interference correction benchmark value for reducing the blue light weight to correct the scattering is calculated. The anti-interference benchmark value is obtained by subtracting the vegetation structure correction value from the environmental interference correction benchmark value. The basic physiological index is obtained by dividing the vegetation photosynthetic activity by the chlorophyll content by the anti-interference benchmark value. After weighting the basic physiological indices, a correction term of leaf area index (1+0.1×Y) is introduced to compensate for the saturation phenomenon of reflectivity in vegetation with high leaf area, and the comprehensive health index of vegetation physiology and structure is calculated.

[0025] The parameter inversion module also includes: The spectral anomaly identification submodule is configured to compare the real-time spectral vector with the reference spectral vector in the fingerprint database, and calculate the spectral angle matching degree using the spectral angle cosine function. After calculating the deviation between the measured soil temperature and the historical average for the same period, the soil temperature deviation index was obtained by normalization. The spectral similarity coupling index is calculated by multiplying the vegetation physiological-structural comprehensive health index with the spectral angle matching degree. The soil temperature deviation index is weighted and calculated to obtain the temperature correction index. The spectral anomaly identification index is calculated by adding the spectral similarity coupling index and the temperature correction index. Specifically, the real-time spectral vector is the spectral reflectance of the current monitoring point obtained by preprocessing after being acquired by a hyperspectral camera. The fingerprint database reference spectral vector is specifically the standard spectral reflectance of healthy vegetation obtained through laboratory detached leaf spectral measurements and field calibration.

[0026] In this embodiment, the initial calculation of vegetation health and abnormal parameters has the following shortcomings: The risk assessment module is configured to couple the spectral anomaly identification index, the pathogen development accumulated temperature ratio, and the environmental humidity, and calculate the pest outbreak risk index through the time-series pest risk index algorithm. Among them, the accumulated temperature ratio for pathogen development is calculated by accumulating the daily average temperature data in the soil surface thermal infrared radiation data, and the environmental humidity includes humidity data obtained through an optical humidity sensor.

[0027] In this embodiment: First, based on multispectral reflectance data, the vegetation physiological-structural comprehensive health index DJ is calculated using the enhanced vegetation index algorithm, and the leaf area index Y is retrieved using the LiDAR point cloud gap rate model; combined with the healthy vegetation spectral fingerprint database, the matching degree between the real-time spectrum and the healthy baseline is calculated using the spectral angle cosine similarity cosθ, and a soil temperature normalization correction term ΔT is introduced. norm Generate the spectral anomaly identification index DZP; The formula for calculating the vegetation physiological-structural comprehensive health index is as follows: ; In the formula: DJ is the vegetation physiological-structural comprehensive health index, A is the near-infrared reflectance, B is the red light reflectance, C is the blue light reflectance, and Y is the leaf area index; The formula for calculating the spectral anomaly detection index is as follows: ; ; ; In the formula: DZP is the spectral anomaly detection index, cosθ is the spectral angular matching degree, and θ is the x i With y i The included angle, where n is the number of spectral vector comparison groups, n≥10; x i For the real-time spectral vector, y i The reference spectral vector for the fingerprint database; △T norm T is the soil temperature deviation index. s The measured soil temperature is T0, which is the historical average for the same period.

[0028] Physiological-structural coupling assessment: To quantify the synergistic effect of chlorophyll content, photosynthetic activity, and canopy structure (leaf area index); Spectral anomaly detection: It achieves an early identification accuracy rate of over 95% for vegetation stress (such as pests and diseases, drought); Environmental interference correction: The influence of atmospheric scattering is eliminated by weighting the blue light reflectance (7.5×C), and the red light weighting is amplified (6×B) to suppress signal saturation in areas with high vegetation cover.

[0029] In summary, the parameter inversion module breaks through the limitations of existing vegetation indices that rely solely on spectral reflectance, and integrates three-dimensional structural parameters to enhance anti-interference capabilities; spectral anomaly identification enables dynamic correlation between "physiological stress and environmental factors," providing early warning of vegetation anomalies 7-10 days earlier than the traditional threshold method; and it provides quantitative and traceable intermediate parameters for subsequent risk assessment, supporting the scientific validity and repeatability of monitoring results.

[0030] Additionally, it is worth noting that 2.5 is used to normalize the vegetation index results to the range of -1 to 2 (healthy vegetation is usually distributed in the range of 0.5 to 1.5), to avoid numerical fluctuations caused by differences in the original reflectance. The Y range is typically 0-10, and the contribution range of 0.1×Y is 0-1, ensuring that the correction of Y to DJ does not exceed 100% and avoiding over-amplification of the influence of Y. and Both 5 and 0.05 are corrections for soil temperature to ensure uniformity of unit quantity.

[0031] Please see Figure 1 and Figure 3 The risk assessment module also includes a cumulative temperature calculation submodule and an early warning submodule. The accumulated temperature calculation submodule is configured to collect the threshold accumulated temperature, m groups of daily average temperature and the starting temperature of pest and disease development by an infrared thermal imager, accumulate the effective accumulated temperature within a 30-day sliding window, and calculate the ratio with the threshold accumulated temperature to obtain the accumulated temperature ratio. The early warning submodule is configured to use spectral anomaly identification index, accumulated temperature ratio, and average relative humidity to calculate the pest and disease outbreak risk index through the following steps: The spectral anomaly identification index is used as the basic weight to reflect the current stress level of vegetation; The accumulated temperature ratio is multiplied by the spectral anomaly identification index to calculate the accumulated temperature coupling index, highlighting the risk during the active phase of pathogens; The average relative humidity is weighted to calculate the humidity amplification index; The risk index of pest and disease outbreaks is obtained by calculating the product of the accumulated temperature coupling index and the humidity amplification index.

[0032] In this embodiment, the algorithm unit first calculates the pest and disease outbreak risk index using the following formula: ; ; In the formula: X is the pest and disease outbreak risk index, JW is the effective accumulated temperature, JW0 is the threshold accumulated temperature, and RH avg The average relative humidity; m is the number of temperature comparison groups, m=30, Ti Let be the average daily temperature of the i-th group, and FS be the temperature at which pests and diseases begin to develop. Pathogen development stage assessment: accumulated temperature ratio It reflects the progress of pathogen infection, and the threshold accumulated temperature JW0 can be adapted to typical forestry diseases and pests such as pine wilt disease. Synergistic effect of environmental factors: High humidity (RH) avg When the risk index is >80%, the risk index increases by 30%, accurately capturing the temperature and humidity coupling conditions for disease transmission.

[0033] In summary, the multi-factor nonlinear coupling model of the risk assessment module overcomes the limitations of assessing a single environmental factor and improves the spatiotemporal resolution of risk prediction; the accumulated temperature-humidity synergistic mechanism reveals the ecological mechanism of pest and disease outbreaks, providing a scientific basis for the formulation of prevention and control measures; and the dynamic risk index provides priority decision support for forestry management departments and optimizes resource allocation efficiency.

[0034] Example 2, please refer to Figures 1 to 3 The decision intervention module is configured to classify risk levels based on the pest and disease outbreak risk index, generate spatial distribution maps, and output intervention measures. The risk levels include high risk, low risk, and medium risk. When the risk level is high, the collection cycle of the healthy vegetation spectral fingerprint database needs to be updated and output. The risk level classification and intervention measures are as follows: Low risk: Maintain the current base cycle; Medium risk: Increase ground-based verification sampling while maintaining the current basic cycle; High risk: Automatically generate plans for felling diseased trees and spraying pesticides, and push them to the forestry management terminal. The plans include precise operation coordinates and dosage calculations, and calculate and obtain the update cycle by combining the pest and disease outbreak risk index and the basic cycle. The base period is the system's default spectral acquisition period, with an initial value of 30 days.

[0035] In this embodiment, the formula for calculating the update cycle is as follows: ; In the formula: Day new For the update cycle, Day now Based on the basic cycle.

[0036] Based on the risk index X for pest and disease outbreaks, risk levels are classified, spatial distribution maps are generated, and differentiated intervention measures are output. This enables closed-loop management of "risk warning - precise intervention - dynamic feedback," reducing prevention and control costs by 60% compared to traditional manual inspections. Differentiated measures avoid over-intervention (such as routine monitoring in low-risk areas) and missed prevention (such as precision operations in high-risk areas), balancing ecological protection and economic costs. Spatial decision support enhances the intelligence level of forestry management and provides a replicable technical paradigm for disaster prevention and control in large-scale forest areas.

[0037] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart forestry monitoring system, characterized in that, Include: The multispectral data acquisition module is configured to use optical means such as hyperspectral cameras, infrared thermal imagers and LiDAR equipment to acquire optical data including multi-band reflectivity data of vegetation canopy, thermal infrared radiation data of soil surface and three-dimensional structure data of canopy. The parameter inversion module is configured to preprocess the acquired optical data and calculate vegetation health and anomalous parameters through the following steps: Based on multi-band reflectance data and canopy three-dimensional structure data, and using the enhanced vegetation index algorithm, the comprehensive health index of vegetation physiology and structure is calculated. The leaf area index is obtained by inverting the canopy three-dimensional structure data through the gap rate model of LiDAR point cloud. By combining the spectral fingerprint database of healthy vegetation, the matching degree between the real-time spectrum and the healthy baseline is calculated through spectral angle cosine similarity, and a normalization correction term for soil surface thermal infrared radiation data is introduced to obtain the spectral anomaly identification index. The risk assessment module is configured to couple the spectral anomaly identification index, the pathogen development accumulated temperature ratio, and the environmental humidity, and calculate the pest outbreak risk index through the time-series pest risk index algorithm. Among them, the accumulated temperature ratio for pathogen development is calculated by accumulating the daily average temperature data in the soil surface thermal infrared radiation data, and the environmental humidity includes humidity data obtained through an optical humidity sensor. The decision intervention module is configured to classify risk levels based on the pest and disease outbreak risk index, generate spatial distribution maps, and output intervention measures. The risk levels include high risk, low risk, and medium risk. When the risk level is high, the collection cycle of the healthy vegetation spectral fingerprint database needs to be updated.

2. The intelligent forestry monitoring system according to claim 1, characterized in that, The multi-band reflectance data includes 700-900nm near-infrared reflectance, 630-690nm red light reflectance, 450-520nm blue light reflectance, n sets of real-time spectral vectors, and fingerprint database reference spectral vectors. The soil surface thermal infrared radiation data includes measured soil temperature, historical average soil temperature for the same period, threshold accumulated temperature, and m-group daily average temperature and the starting temperature for pest and disease development. The canopy three-dimensional structure data includes a 532nm green laser point cloud; The humidity data includes the normalized average relative humidity.

3. The intelligent forestry monitoring system according to claim 2, characterized in that: The parameter inversion module includes: The inversion submodule is configured to use a LiDAR device to emit a 532nm green laser to acquire three-dimensional point cloud data of the canopy and simultaneously record the laser echo intensity and coordinate information. A cloth-simulation filtering algorithm was used to remove ground points while retaining the vegetation canopy point cloud. The gap ratio is calculated based on the proportion of laser penetration points under the canopy to the total point cloud. The leaf area index is calculated based on the gap ratio and the average leaf tilt angle factor according to the tree species type; The vegetation health calculation submodule is configured to take the near-infrared reflectance, the red light reflectance, the blue light reflectance, and the leaf area index as input, and calculate the comprehensive physiological-structural health index of vegetation through the following steps: Based on the difference between the near-infrared reflectance and the red reflectance, the photosynthetic activity and chlorophyll content of the vegetation are calculated. The weighted red light reflectance is added to the near-infrared reflectance to calculate and obtain the vegetation structure correction value that increases the red light weight to suppress saturation. After weighting the blue light reflectance, the environmental interference correction benchmark value for reducing the blue light weight to correct the scattering is calculated. The anti-interference benchmark value is obtained by subtracting the vegetation structure correction value from the environmental interference correction benchmark value. The basic physiological index is obtained by dividing the vegetation photosynthetic activity by the chlorophyll content by the anti-interference benchmark value. After weighting the basic physiological indices, a correction term (1+0.1×Y) of the leaf area index is introduced to compensate for the saturation phenomenon of the reflectance of high leaf area vegetation, and the comprehensive health index of vegetation physiology and structure is calculated. Where Y is the leaf area index.

4. The intelligent forestry monitoring system according to claim 3, characterized in that: The parameter inversion module also includes: The spectral anomaly identification submodule is configured to compare the real-time spectral vector with the fingerprint database reference spectral vector and calculate the spectral angle matching degree using the spectral angle cosine function. After calculating the deviation between the measured soil temperature and the historical average for the same period, normalization is performed to obtain the soil temperature deviation index. The spectral similarity coupling index is calculated based on the product of the vegetation physiological-structural comprehensive health index and the spectral angle matching degree. The soil temperature deviation index is weighted and calculated to obtain the temperature correction index; The spectral anomaly identification index is calculated by adding the spectral similarity coupling index and the temperature correction index.

5. The intelligent forestry monitoring system according to claim 4, characterized in that: The risk assessment module also includes an accumulated temperature calculation submodule and an early warning submodule. The accumulated temperature calculation submodule is configured to collect the threshold accumulated temperature and m groups of daily average temperature and the starting temperature of pest and disease development by an infrared thermal imager, accumulate the effective accumulated temperature within a 30-day sliding window, and calculate the ratio with the threshold accumulated temperature to obtain the accumulated temperature ratio. The early warning submodule is configured to calculate the pest and disease outbreak risk index using the spectral anomaly identification index, the accumulated temperature ratio, and the average relative humidity through the following steps: The spectral anomaly identification index is used as the basic weight; The accumulated temperature ratio is multiplied by the spectral anomaly identification index to calculate the accumulated temperature coupling index; The average relative humidity is weighted and the humidity amplification index is calculated. The pest and disease outbreak risk index is obtained by calculating the product of the accumulated temperature coupling index and the humidity amplification index.

6. The intelligent forestry monitoring system according to claim 5, characterized in that: The risk level classification and intervention measures are as follows: Low risk: Maintain the current base cycle; Medium risk: Increase ground-based verification sampling while maintaining the current basic cycle; High risk: Automatically generate a plan for felling diseased trees and spraying pesticides, and push it to the forestry management terminal. The plan includes accurate operation coordinates and dosage calculations, and calculates the update cycle by combining the pest and disease outbreak risk index and the basic cycle. The base period is the system's default spectral acquisition period, with an initial value of 30 days.

7. The intelligent forestry monitoring system according to claim 1, characterized in that: The multispectral data acquisition module includes: The synchronization control unit is configured to achieve spatiotemporal registration of hyperspectral cameras, infrared thermal imagers, and LiDAR devices when authorized by GPS. The data preprocessing unit is configured to perform radiometric calibration, atmospheric correction, and point cloud denoising on the raw optical data.

8. The intelligent forestry monitoring system according to claim 4, characterized in that: The real-time spectral vector is specifically the spectral reflectance of the current monitoring point obtained by preprocessing after being acquired by a hyperspectral camera. The fingerprint database reference spectral vector is specifically the standard spectral reflectance of healthy vegetation obtained through laboratory detached leaf spectral measurements and field calibration.

Images