Unmanned aerial vehicle monitoring system for forest carbon sink estimation

Abstract

The invention discloses an unmanned aerial vehicle monitoring system for forest carbon sink estimation. The unmanned aerial vehicle monitoring system comprises an unmanned aerial vehicle flight platform, a multi-sensor module, a data transmission module, an edge calculation module, a cloud server and a carbon reserve estimation module. The multi-sensor module is detachably connected to the unmanned aerial vehicle flight platform and used for synchronously collecting multi-dimensional original data of a forest ecosystem, and the multi-sensor module at least comprises a laser radar sensor, a hyperspectral sensor, a thermal infrared sensor and a visible light camera; the edge calculation module is integrated on the unmanned aerial vehicle flight platform. Through fusion of the multi-source sensor, multi-dimensional data such as a three-dimensional structure, biochemical characteristics, thermal characteristics and texture characteristics of a forest can be acquired, and reliable basic data is provided for carbon reserve estimation; local real-time processing of data can be realized through the edge calculation module, so that the processing delay is greatly reduced; and through the carbon reserve estimation module, the estimation precision in different scenes can be ensured according to a forest type optimization model.

Description

Technical Field

[0001] This invention belongs to the field of environmental monitoring technology, specifically relating to a drone monitoring system for estimating forest carbon sinks. Background Technology

[0002] The core of forest carbon storage estimation lies in accurately acquiring forest biomass and then deriving carbon storage through the conversion coefficient between biomass and carbon storage. Traditional forest carbon storage estimation methods are mainly divided into two categories: field plot surveys and satellite remote sensing monitoring. Field plot surveys, which manually measure key parameters such as tree diameter at breast height (DBH), tree height, and canopy closure, and then calculate carbon storage using a biomass model, offer high accuracy but suffer from inherent drawbacks such as high labor intensity, long processing time, limited coverage, and poor accessibility in remote and complex terrain, making them unsuitable for large-scale, dynamic carbon storage monitoring. Satellite remote sensing monitoring utilizes multispectral and hyperspectral data acquired by satellites to invert forest biomass through vegetation indices. It boasts advantages such as wide coverage and short monitoring cycles, but is limited by spatial resolution, making it difficult to accurately identify individual trees, and is susceptible to cloud and atmospheric interference, resulting in lower estimation accuracy in complex terrain areas.

[0003] In recent years, UAV remote sensing technology has been increasingly applied to forest carbon storage monitoring. Existing technologies mostly employ UAVs equipped with a single sensor to acquire forest parameters, combining them with traditional biomass models for carbon storage estimation. For example, CN117893931A discloses a carbon storage monitoring method and system based on UAV-LS, which acquires aboveground vegetation volume using lidar point cloud data and calculates carbon storage by combining basic wood density and carbon content coefficient. However, this system relies solely on lidar data, failing to consider the impact of vegetation spectral information and environmental parameters on carbon storage estimation. Furthermore, data processing depends on the ground-based system, resulting in high processing latency and the inability to output results in real time. Additionally, it does not consider the differences in vegetation growth characteristics among different forest types and geographical regions, making it difficult to guarantee estimation accuracy in diverse forest scenarios.

[0004] Therefore, in response to the aforementioned technical problems, it is necessary to provide a drone monitoring system for forest carbon sequestration estimation.

[0005] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a drone monitoring system for forest carbon sequestration estimation, which can solve the problems of existing systems such as single data collection, high processing latency, and poor adaptability to diversity.

[0007] To achieve the above objectives, a specific embodiment of the present invention provides the following technical solution: A drone monitoring system for forest carbon sequestration estimation includes a drone flight platform, a multi-sensor module, a data transmission module, an edge computing module, a cloud server, and a carbon storage estimation module. The multi-sensor module is detachably connected to the drone flight platform and is used to synchronously collect multi-dimensional raw data of the forest ecosystem. The multi-sensor module includes at least a lidar sensor, a hyperspectral sensor, a thermal infrared sensor, and a visible light camera. The edge computing module is integrated into the drone flight platform and is used for real-time preprocessing and fusion of the multi-dimensional raw data collected by the multi-sensor module. The cloud server stores a basic forest database and a model parameter library, and provides data support and computing power for the carbon storage estimation module. The carbon storage estimation module is used to achieve accurate identification of forest types and targeted carbon storage estimation.

[0008] In one or more embodiments of the present invention, the unmanned aerial vehicle (UAV) flight platform is equipped with a flight control module that can plan the optimal flight path based on forest terrain data and vegetation height data.

[0009] In one or more embodiments of the present invention, the multi-sensor module further includes a synchronization control unit. The synchronization control unit adopts a GPS-BeiDou dual-mode timing and hardware trigger signal linkage mechanism to control the lidar sensor, hyperspectral sensor, thermal infrared sensor and visible light camera to synchronously collect data at the same timestamp.

[0010] In one or more embodiments of the present invention, the synchronization control unit integrates an attitude compensation module, which can receive UAV attitude perception data in real time and perform spatial registration of the data collected by the lidar sensor, hyperspectral sensor, thermal infrared sensor and visible light camera through a coordinate transformation algorithm.

[0011] In one or more embodiments of the present invention, the data transmission module supports 5G and WiFi communication and is equipped with USB and HDMI standardized interfaces for local data export and device debugging.

[0012] In one or more embodiments of the present invention, the preprocessing process of the edge computing module is as follows: an adaptive median filtering algorithm is used to denoise the lidar point cloud data, and the filtering window is dynamically adjusted according to the point cloud density; a radiometric correction combined with a FLAASH atmospheric correction algorithm is used to preprocess the hyperspectral data to eliminate sensor errors and atmospheric interference; a CLAHE histogram equalization algorithm is used to enhance the visible light image and improve details in the shadow area; and a blackbody calibration combined with an atmospheric transmittance correction algorithm is used to correct the accuracy of the thermal infrared data.

[0013] In one or more embodiments of the present invention, the forest basic database includes a forest type sample library, a geographic environment database, and a measured data verification library. The forest type sample library includes multi-sensor labeled feature samples of at least 20 forest types, and the geographic environment database includes climate data, soil carbon pool data, and topographic data for different regions. The model parameter library stores carbon storage estimation models and dynamic parameters corresponding to different forest types and different growth stages. The dynamic parameters include at least the biomass expansion factor, organ carbon content coefficient, root-to-stem ratio, and light energy utilization efficiency.

[0014] In one or more embodiments of the present invention, the carbon storage estimation module includes a forest type identification unit and a multi-model selection unit. The forest type identification unit adopts an improved U-Net network model, takes the fused feature map output by the edge computing module as input, and is trained in combination with a forest type sample library to achieve pixel-level segmentation and identification of forest types.

[0015] In one or more embodiments of the present invention, the multi-model selection unit includes a model fusion module, which uses a weighted fusion algorithm to fuse the carbon storage estimation results corresponding to different types when there are multiple mixed forest types in the monitoring area.

[0016] In one or more embodiments of the present invention, a visualization module is further included, which is used to display the distribution of forest carbon storage, the distribution of forest types, and the quality of monitoring data in the form of three-dimensional maps, heat maps, and statistical tables.

[0017] Compared with existing technologies, the UAV monitoring system for forest carbon sequestration estimation of the present invention can collect multi-dimensional data such as the three-dimensional structure, biochemical characteristics, thermal characteristics and texture characteristics of forests through the fusion of multi-source sensors, providing reliable basic data for carbon storage estimation; the edge computing module can realize local real-time data processing, which greatly reduces processing latency; and the carbon storage estimation module can optimize the model according to forest type to ensure estimation accuracy in different scenarios. Attached Figure Description

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

[0019] Figure 1 This is a structural block diagram of a drone monitoring system for forest carbon sequestration estimation according to an embodiment of the present invention; Figure 2This is a structural block diagram of an unmanned aerial vehicle (UAV) flight platform according to one embodiment of the present invention; Figure 3 This is a block diagram of the forest basic database structure in one embodiment of the present invention; Figure 4 This is a block diagram of the model parameter library structure in one embodiment of the present invention; Figure 5 This is a structural block diagram of a carbon storage estimation module in one embodiment of the present invention. Detailed Implementation

[0020] To enable those skilled in the art to better understand the technical solutions in this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this disclosure.

[0021] like Figures 1 to 5 As shown, an embodiment of the present invention provides a drone monitoring system for forest carbon sink estimation, comprising a drone flight platform, a multi-sensor module, a data transmission module, an edge computing module, a cloud server, and a carbon storage estimation module.

[0022] The system comprises a multi-sensor module, detachably connected to the UAV flight platform, for simultaneously collecting multi-dimensional raw data on the forest ecosystem. This module includes at least a lidar sensor, a hyperspectral sensor, a thermal infrared sensor, and a visible light camera. An edge computing module, integrated into the UAV flight platform, performs real-time preprocessing and fusion of the multi-dimensional raw data collected by the sensor module. A cloud server stores the forest database and model parameter library, providing data support and computing power for the carbon storage estimation module. The carbon storage estimation module enables accurate identification of forest types and targeted carbon storage estimation.

[0023] This application utilizes the fusion of multiple sensor sources to collect multi-dimensional data on forest structure, biochemical characteristics, thermal features, and texture features, providing reliable foundational data for carbon inventory estimation. The edge computing module enables local real-time data processing, significantly reducing processing latency. The carbon inventory estimation module optimizes the model based on forest type, ensuring estimation accuracy across different scenarios.

[0024] In addition, the drone monitoring system also includes a visualization module and an early warning module. The visualization module is used to display the distribution of forest carbon storage, forest type distribution and monitoring data quality in the form of three-dimensional maps, heat maps and statistical tables.

[0025] For example, a three-dimensional map can be used to display the three-dimensional structure of a forest and the spatial distribution of its carbon storage. A heat map can be used to show the differences in carbon storage density across different regions, where the formula for calculating carbon storage density is: In the formula, CD is the carbon storage density, with units of tC / hm. 2 C 总 The total carbon storage in the monitored area is expressed in tC; A 总 The total area of ​​the monitored region is expressed in hectares (hm²). 2 .

[0026] Specifically, the early warning module can set an abnormal change threshold for carbon storage based on the comparative analysis of carbon storage estimation results and historical data (the abnormal change threshold can be customized according to different forest types, such as ±10% for coniferous forests and ±15% for broadleaf forests).

[0027] In this embodiment, the formula for calculating the abnormal change rate of carbon reserves is: In the formula, ΔC represents the abnormal change rate of carbon reserves, in %; C 本 This is the estimated carbon reserves for this study, in tC. This represents the historical average carbon reserves for the same period, expressed in tC.

[0028] Furthermore, The calculation formula is: In the formula, Let n be the estimated carbon storage for the same period in year i, and n be the number of years of historical data for the same period. The average estimated carbon storage for the same period in history is calculated using carbon storage data from the same monitoring period (such as the same quarter or the same growth stage) within the past 3-5 years. If the historical data is less than 3 years, existing data from the same period can be used instead.

[0029] This application utilizes a visualization module to make forest carbon storage information more intuitive and easier to understand, enabling managers to quickly grasp the status and spatial distribution characteristics of forest carbon storage. Through the early warning module, when changes in carbon storage in the monitored area exceed a threshold, the module will issue an alert, marking the location of the abnormal area, forest type, and magnitude of the change. Simultaneously, it will push alert information to relevant management personnel, providing timely decision support for forest resource management and carbon storage project supervision.

[0030] In this embodiment, the UAV flight platform adopts a multi-rotor heavy-duty UAV design, which has the capabilities of vertical take-off and landing, fixed-point hovering and low-speed cruising, and can adapt to the low-altitude monitoring needs of complex forest terrain such as mountains and hills.

[0031] like Figure 2As shown, the UAV flight platform is equipped with a flight control module. Based on GIS map data and forest terrain and vegetation height data acquired by lidar sensors, the flight control module uses an improved A* algorithm to plan the optimal flight path, ensuring full coverage of the monitoring area. The overlap rate of the flight path can be customized according to the monitoring accuracy requirements. Simultaneously, the UAV flight platform integrates a lidar obstacle avoidance unit. By collecting real-time distance information of surrounding obstacles, it uses a dynamic obstacle avoidance algorithm to achieve autonomous obstacle avoidance, ensuring flight safety.

[0032] In addition, the drone flight platform is powered by a lithium battery pack with a capacity of ≥20000mAh, a flight time of ≥4 hours, and an operating radius of ≥20km, which can meet the endurance requirements for large-scale forest monitoring. The drone flight platform has a battery management system with real-time power monitoring, overcharge and over-discharge protection, and power warning functions.

[0033] In addition, the UAV flight platform is equipped with a positioning module that integrates GPS and Beidou dual-mode positioning with a positioning accuracy of ≤1m. It can accurately record the UAV's flight position and attitude data, providing a positioning benchmark for the spatial registration of sensor data.

[0034] Furthermore, the UAV flight platform also integrates an attitude perception module. This module, through the cooperation of gyroscopes, accelerometers, and magnetometers, can collect attitude data such as pitch angle, roll angle, and heading angle of the UAV in real time, providing data support for attitude compensation of sensor data.

[0035] In this embodiment, the lidar sensor is a pulse lidar with a laser wavelength of 1550nm, a ranging range of 0.5-200m, a ranging accuracy of ±2cm, and a point cloud density of ≥100 points / ㎡. It is used to acquire three-dimensional point cloud data of the forest and accurately extract three-dimensional structural parameters such as tree height, diameter at breast height, crown width, and stand density.

[0036] Among them, the tree diameter at breast height (DBH) is inverted based on crown width and tree height parameters using the allometric growth equation, and the formula is: In the formula, DBH is the diameter at breast height of the tree in cm; H is the height of the tree in m; C is the crown width of the tree in m; a, b, and c are the parameters of the allometric growth equation, and the values ​​of the parameters are different for different forest types (e.g., for Chinese fir forest, a=0.235, b=0.621, c=0.315).

[0037] The hyperspectral sensor has a spectral range of 400-1000 nm, a spectral resolution of ≤3 nm, ≥256 bands, and a signal-to-noise ratio of ≥500:1, and is used to acquire hyperspectral data of forests. By analyzing the spectral characteristics, biochemical parameters such as chlorophyll content, nitrogen content, and water content of vegetation are retrieved to reflect the growth status and health of the forest. The retrieval formula is as follows: ; In the formula, Chl represents the chlorophyll content of vegetation, in mg / g; NDVI RE The red-edged normalized vegetation index, MC represents vegetation moisture content, expressed as %; WI represents the moisture index. k1, k2, k3, and k4 are the parameters of the inversion model, which are determined by fitting measured data.

[0038] The thermal infrared sensor has a detection band of 8-14μm, a temperature measurement range of -20℃ to 60℃, and a temperature resolution of ≤0.05℃. It is used to acquire thermal infrared images of forests, monitor the surface temperature distribution of forests, reflect the water stress status of vegetation, and provide a basis for judging the health status and carbon storage capacity of forests.

[0039] Visible light cameras with a resolution of ≥20 million pixels and a lens focal length of 24mm are equipped with autofocus and exposure functions. They are used to acquire high-definition visible light images of forests, record information such as forest texture features and vegetation coverage, and provide intuitive data support for forest type identification.

[0040] The synchronization control unit employs a GPS-BeiDou dual-mode time synchronization and hardware trigger signal linkage mechanism. It acquires a standard time signal via GPS to generate a synchronization trigger signal, controlling the lidar sensor, hyperspectral sensor, thermal infrared sensor, and visible light camera to collect data at the same timestamp. Simultaneously, this unit receives attitude data from the UAV's attitude perception module and performs spatial registration of the data collected by each sensor using a coordinate transformation algorithm.

[0041] Furthermore, the data transmission module supports 5G and WiFi communication and is equipped with standardized USB and HDMI interfaces for local data export and device debugging.

[0042] In this embodiment, the preprocessing process of the edge computing module is as follows: An adaptive median filtering algorithm is used to denoise the lidar point cloud data, with the filtering window dynamically adjusted according to the point cloud density. Radiometric correction combined with the FLAASH atmospheric correction algorithm is used to preprocess the hyperspectral data to eliminate sensor errors and atmospheric interference. The CLAHE histogram equalization algorithm is used to enhance the visible light image, improving details in shadow areas. Blackbody calibration combined with atmospheric transmittance correction is used to correct the accuracy of the thermal infrared data.

[0043] Then, key features were extracted from the preprocessed sensor data. Three-dimensional structural features such as tree height, diameter at breast height (DBH), crown width, and stand density were extracted from the lidar point cloud data. Vegetation indices and vegetation biochemical parameters were extracted from the hyperspectral data. Thermal features such as forest surface temperature and temperature standard deviation were extracted from the thermal infrared data. Texture features and vegetation cover features were extracted from the visible light images.

[0044] The data fusion process of the edge computing module adopts a multimodal data fusion algorithm based on the attention mechanism, which dynamically allocates the feature weights of each sensor through the convolutional attention module.

[0045] like Figure 3 As shown, the forest basic database includes a forest type sample library, a geographic environment database, and a measured data validation library. The forest type sample library contains multi-sensor labeled feature samples of at least 20 forest types, including three major categories: coniferous forests, broad-leaved forests, and mixed forests, as well as their sub-species. Each sample includes corresponding lidar point cloud features, hyperspectral features, thermal infrared features, visible light texture features, and a forest type label.

[0046] The geographic environment database includes climate data (such as annual average temperature, annual precipitation, sunshine hours, etc.), soil carbon pool data (such as soil type, soil organic matter content, soil pH value, etc.) and topographic data (such as altitude, slope, aspect, etc.) for different regions.

[0047] In addition, the measured data validation library collects measured data on forest carbon storage, including biomass data obtained through plot surveys and carbon storage flux data obtained through carbon flux monitoring stations. The measured data covers different forest types, regions, and growth stages, and is used to validate carbon storage estimation results and optimize model parameters.

[0048] like Figures 4 to 5 As shown, the model parameter library stores carbon storage estimation models and dynamic parameters corresponding to different forest types and growth stages. The carbon storage estimation models include biomass-based models, photosynthetic efficiency-based models, and volume-based models. Dynamic parameters include at least the biomass expansion factor, organ carbon content coefficient, root-to-tiller ratio, and light energy utilization efficiency.

[0049] The carbon storage estimation module includes a forest type identification unit and a multi-model selection unit. The forest type identification unit uses an improved U-Net network model, taking the fused feature map output by the edge computing module as input and training it in conjunction with a forest type sample library to achieve pixel-level segmentation and identification of forest types.

[0050] In addition, the multi-model selection unit, based on the forest type results output by the forest type identification unit and combined with the model parameter library on the cloud server, achieves adaptive selection and parameter matching of carbon storage estimation models.

[0051] For example, for coniferous forests, due to their long growth cycle and stable biomass structure, a biomass-based carbon storage model is chosen. The formula for the biomass-based carbon storage model is:

[0052] In the formula, C b Carbon storage is based on biomass, in tC; B is the total stand biomass, in t; CF is the carbon content coefficient, which varies for different forest types (e.g., CF = 0.5180 for Masson pine forests); W 干 W 枝 W 叶 W 根 , respectively, represent the biomass of the trunk, branches, leaves, and roots of the i-th tree, in t; n is the number of trees in the stand.

[0053] Specifically, Based on the diameter at breast height (DBH) and tree height (H) parameters obtained from the inversion, the allometric growth equations for different tree species are used for calculation, as shown in the following formula:

[0054] In the formula, a1-a4, b1-b4, and c1-c4 are the parameters of the allometric growth equation for different tree species. The parameter values ​​for different forest types are referenced from the model parameter library.

[0055] Specifically, the multi-model selection unit includes a model fusion module. When multiple forest types are mixed in the monitoring area, a weighted fusion algorithm is used to fuse the carbon storage estimation results corresponding to different types. The weights are determined based on the area proportion of each forest type, as shown in the following formula:

[0056] In the formula, C 总 The total carbon storage of the mixed forest area is expressed in tC; C k Estimated carbon storage for forest type k, in tC; W k A is the area weighting coefficient for the k-th forest type; k The area of ​​forest type k in the mixed forest region is expressed in hectares (hm²). 2 m represents the number of forest types within the mixed forest area.

[0057] In practical use, managers send monitoring task instructions to the drone flight platform via a cloud server, including the monitoring area, flight altitude, flight speed, and sensor parameter settings. After receiving the instructions, the drone flight platform completes initialization tasks such as sensor preheating, positioning calibration, and flight path planning.

[0058] The drone flight platform flies autonomously along a planned path. A synchronous control unit controls lidar sensors, hyperspectral sensors, thermal infrared sensors, and visible light cameras to simultaneously collect multi-dimensional raw data of the forest, while also recording the drone's flight position and attitude data. The edge computing module preprocesses the collected multi-sensor raw data, extracting key features from each sensor and generating a fused feature map using an attention-based multimodal data fusion algorithm. The data transmission module transmits the fused feature map, drone positioning data, and attitude data processed by the edge computing module to a cloud server.

[0059] The cloud server's carbon storage estimation module calls the forest type identification unit, using a fused feature map as input and combining it with a forest type sample library to achieve pixel-level segmentation and identification of forest types, outputting a forest type distribution map and area percentage. The multi-model selection unit matches the corresponding carbon storage estimation model and dynamic parameters from the model parameter library based on the forest type identification results. For mixed forest areas, a weighted fusion algorithm is used for multi-model fusion estimation. Error correction is performed using a measured data validation library, and the final carbon storage estimation result is output.

[0060] The visualization module displays information such as forest type distribution and carbon storage estimation results in various forms, compares the carbon storage estimation results with historical data, and issues early warnings for areas with abnormal changes through the early warning module.

[0061] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0062] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0063] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0064] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0065] It will be apparent to those skilled in the art that this disclosure is not limited to the details of the exemplary embodiments described above, and that this disclosure can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of this disclosure is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within this disclosure. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0066] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment includes only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A drone monitoring system for forest carbon sequestration estimation, characterized in that, It includes a drone flight platform, a multi-sensor module, a data transmission module, an edge computing module, a cloud server, and a carbon storage estimation module; The multi-sensor module is detachably connected to the UAV flight platform and is used to synchronously collect multi-dimensional raw data of the forest ecosystem. The multi-sensor module includes at least a lidar sensor, a hyperspectral sensor, a thermal infrared sensor, and a visible light camera. The edge computing module is integrated into the UAV flight platform and is used to perform real-time preprocessing and fusion of the multi-dimensional raw data collected by the multi-sensor module. The cloud server is used to store the forest basic database and model parameter library, and to provide data support and computing power for the carbon storage estimation module. The carbon storage estimation module is used to achieve accurate identification of forest types and targeted carbon storage estimation.

2. The UAV monitoring system for forest carbon sequestration estimation according to claim 1, characterized in that, The drone flight platform is equipped with a flight control module, which can plan the optimal flight path based on forest terrain data and vegetation height data.

3. The UAV monitoring system for forest carbon sequestration estimation according to claim 1, characterized in that, The multi-sensor module also includes a synchronization control unit, which adopts a GPS-BeiDou dual-mode timing and hardware trigger signal linkage mechanism to control the lidar sensor, hyperspectral sensor, thermal infrared sensor and visible light camera to synchronously collect data at the same timestamp.

4. The UAV monitoring system for forest carbon sequestration estimation according to claim 3, characterized in that, The synchronization control unit integrates an attitude compensation module, which can receive UAV attitude perception data in real time and perform spatial registration of the data collected by the lidar sensor, hyperspectral sensor, thermal infrared sensor and visible light camera through a coordinate transformation algorithm.

5. The UAV monitoring system for forest carbon sequestration estimation according to claim 1, characterized in that, The data transmission module supports 5G and WiFi communication and is equipped with USB and HDMI standardized interfaces for local data export and device debugging.

6. The UAV monitoring system for forest carbon sequestration estimation according to claim 1, characterized in that, The preprocessing process of the edge computing module is as follows: an adaptive median filtering algorithm is used to denoise the lidar point cloud data, with the filtering window dynamically adjusted according to the point cloud density; a radiometric correction combined with the FLAASH atmospheric correction algorithm is used to preprocess the hyperspectral data to eliminate sensor errors and atmospheric interference; a CLAHE histogram equalization algorithm is used to enhance the visible light image, improving details in shadow areas; and a blackbody calibration combined with an atmospheric transmittance correction algorithm is used to correct the accuracy of the thermal infrared data.

7. The UAV monitoring system for forest carbon sequestration estimation according to claim 1, characterized in that, The forest basic database includes a forest type sample library, a geographic environment database, and a measured data verification library; The forest type sample library includes multi-sensor labeled feature samples of at least 20 forest types, and the geographic environment database includes climate data, soil carbon pool data and topographic data of different regions. The model parameter library stores carbon storage estimation models and dynamic parameters corresponding to different forest types and different growth stages. The dynamic parameters include at least the biomass expansion factor, organ carbon content coefficient, root-to-tuber ratio, and light energy utilization efficiency.

8. A drone monitoring system for forest carbon sequestration estimation according to claim 7, characterized in that, The carbon storage estimation module includes a forest type identification unit and a multi-model selection unit. The forest type identification unit adopts an improved U-Net network model, takes the fused feature map output by the edge computing module as input, and is trained in combination with the forest type sample library to achieve pixel-level segmentation and identification of forest types.

9. A drone monitoring system for forest carbon sequestration estimation according to claim 8, characterized in that, The multi-model selection unit includes a model fusion module. When multiple forest types are mixed in the monitoring area, a weighted fusion algorithm is used to fuse the carbon storage estimation results corresponding to different types.

10. A drone monitoring system for forest carbon sequestration estimation according to claim 1, characterized in that, It also includes a visualization module, which is used to display the distribution of forest carbon storage, forest type distribution and monitoring data quality in the form of three-dimensional maps, heat maps and statistical tables.

Images