Forest stand attribute analysis method and system based on unmanned aerial vehicle remote sensing data feedback

Abstract

The invention discloses a forest stand attribute analysis method and system based on unmanned aerial vehicle remote sensing data feedback, and relates to the technical field of data processing, and the method comprises the steps: obtaining to-be-analyzed forest region data, querying the historical route data of an unmanned aerial vehicle based on position data, and determining remote sensing data collection equipment information and corresponding multi-source environment information; determining a forest stand attribute influence first sub-value of the to-be-analyzed forest; constructing an attribute analysis space coordinate system, determining to-be-analyzed remote sensing data corresponding to the coordinate points, and obtaining a to-be-analyzed forest stand attribute influence second sub-value; analyzing the remote sensing data to obtain a forest stand attribute index matrix; and based on the matrix and the to-be-analyzed forest stand attribute influence second sub-value, obtaining a forest stand attribute standard matrix and drawing up a forest management strategy. The system comprises an acquisition module and an unmanned aerial vehicle data and environment information association module. According to the method, systematic integration of multi-source data is realized, the forest stand attribute analysis precision is improved, and a scientific basis is provided for forest management.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method and system for analyzing forest stand attributes based on feedback from UAV remote sensing data. Background Technology

[0002] As a core component of the Earth's ecosystem, forests require precise monitoring of their stand attributes (such as tree species composition, diameter at breast height, tree height, biomass, and canopy closure) for ecological protection, carbon sequestration, forestry management, and climate change mitigation. With the development of UAV remote sensing technology and data processing algorithms, forest stand attribute analysis methods and systems based on UAV remote sensing data are gradually becoming key technologies for overcoming traditional bottlenecks.

[0003] Existing methods and systems for analyzing forest stand attributes based on UAV remote sensing data feedback do not adequately consider environmental factors during UAV flight, such as wind speed, temperature, and terrain. This results in significant environmental interference affecting the quality of remote sensing data acquisition, leading to data redundancy and accuracy deviations. Furthermore, the lack of systematic integration and collaborative analysis of multi-source data results in low accuracy of stand attribute inversion, hindering the development of targeted forest management strategies. Therefore, it is necessary to provide a method and system for analyzing forest stand attributes based on UAV remote sensing data feedback to address the aforementioned problems. Summary of the Invention

[0004] To address the aforementioned technical problems, this paper provides a method and system for analyzing forest stand attributes based on UAV remote sensing data feedback. This solution solves the problems of existing methods and systems for analyzing forest stand attributes based on UAV remote sensing data feedback, which fail to adequately consider environmental factors during UAV flight, such as wind speed, temperature, and terrain. This results in significant environmental interference affecting the quality of remote sensing data acquisition, leading to data redundancy and accuracy deviations. Furthermore, the lack of systematic integration and collaborative analysis of multi-source data results in low accuracy of stand attribute inversion, making it difficult to formulate targeted forest management strategies.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for analyzing forest stand attributes based on UAV remote sensing data feedback includes: Acquire data on the forest area to be analyzed, including the location data of the forest to be analyzed; Based on the forest location data to be analyzed, query the historical flight path data of the UAV, and determine the information of the remote sensing data acquisition equipment carried by the corresponding UAV, as well as the multi-source environmental information corresponding to the historical flight path data of the UAV. Based on information from remote sensing data acquisition equipment, historical flight path data of UAVs, and corresponding multi-source environmental information, the first sub-value of the forest stand attribute influence to be analyzed is determined. Construct an attribute analysis spatial coordinate system and determine the remote sensing data to be analyzed corresponding to each coordinate point in the attribute analysis spatial coordinate system. Simultaneously, obtain the second sub-value of the forest stand attribute influence presented by the forest area to be analyzed from the remote sensing data to be analyzed. The remote sensing data to be analyzed is analyzed for each coordinate point in the attribute analysis spatial coordinate system to obtain the forest stand attribute index matrix. Based on the forest stand attribute index matrix and the second sub-value of the influence of the forest stand attributes to be analyzed, the standard matrix of forest stand attributes is obtained, and forest management strategies are formulated.

[0006] In an optional embodiment, the step of querying historical flight path data of the UAV based on the forest location data to be analyzed, and determining the information of the remote sensing data acquisition equipment carried by the corresponding UAV, as well as the multi-source environmental information corresponding to the historical flight path data of the UAV, specifically includes: Based on the location data of the forest to be analyzed, determine the location range information of the forest to be analyzed, including latitude and longitude coordinates and administrative boundaries; Input the latitude and longitude coordinates and administrative boundaries into the spatial analysis tool, and use the spatial analysis function to filter and obtain the historical flight data of UAVs in the corresponding forest area to be analyzed, including basic flight information and flight association files. Based on basic route information, obtain the UAV flight date, UAV number, and UAV flight mission ID, query equipment configuration files, and determine the details of the remote sensing equipment carried on each flight; Based on the detailed information of the remote sensing equipment carried on each flight, the type and parameters of the remote sensing equipment are determined, thereby obtaining the information of the remote sensing data acquisition equipment carried by the corresponding UAV. The historical flight time range and drone number of the drone are obtained from the drone's historical flight route data. The flight log is extracted in combination with the flight date to determine the environmental parameters corresponding to the flight. Based on the flight strip coverage of historical UAV flight data, combined with the location range information of the forest to be analyzed, the geographical location information corresponding to the flight strip coverage is extracted, and the corresponding terrain parameters are determined. By integrating the environmental parameters corresponding to the flight path with the terrain parameters corresponding to the flight path coverage area, multi-source environmental information corresponding to the historical flight path data of the UAV is obtained.

[0007] In an optional embodiment, determining the first sub-value of the forest stand attribute influence to be analyzed based on remote sensing data acquisition equipment information, UAV historical flight path data, and corresponding multi-source environmental information specifically includes: Remote sensing equipment type information and remote sensing equipment parameters are extracted from remote sensing data acquisition equipment information. Simultaneously, the flight path coverage is extracted from UAV historical flight path data, and the environmental parameters corresponding to the UAV historical flight path data are extracted from multi-source environmental information. Based on the environmental adaptability requirements of UAV flight, the historical flight path data of UAVs are classified according to environmental parameters, resulting in two types of flight path data with different environmental adaptability. Based on the environmental parameters corresponding to the classified route data, the risk of route deviation and altitude deviation are assessed, and route data that meets the data quality requirements are further screened. Calculate the actual overlap and planned overlap of different categories of route data after filtering. Based on the matching relationship between the actual overlap and the planned overlap, use an appropriate boundary processing method to determine the actual area value of the flat area. For areas with complex terrain, the actual area value of the complex terrain is determined by combining its terrain parameters and appropriate boundary processing methods through an area calculation method suitable for complex terrain. The summation of the actual area values ​​of flat areas and complex terrains yields the first sub-value of the influence of forest stand attributes on the analysis.

[0008] In an optional embodiment, the step of constructing an attribute analysis spatial coordinate system and determining the remote sensing data to be analyzed corresponding to each coordinate point in the attribute analysis spatial coordinate system, and simultaneously obtaining the second sub-value of the influence of the forest stand attributes of the forest area to be analyzed from the remote sensing data to be analyzed, specifically includes: Based on the first sub-value of the influence of forest stand attributes on the forest to be analyzed and the data of the forest region to be analyzed, the first region of the forest to be analyzed corresponding to flat areas and the second region of the forest to be analyzed corresponding to complex terrain areas are distinguished. Obtain the actual geographical location information of the first and second regions of the forest to be analyzed, and determine the corresponding latitude and longitude ranges; Based on the terrain features and data adaptation requirements of the forest area to be analyzed, a unified spatial coordinate system for attribute analysis is constructed. Through the projection transformation function of the spatial analysis tool, the historical flight path data of UAVs in the first and second areas and the corresponding remote sensing data to be analyzed are uniformly transformed to this coordinate system. In the attribute analysis spatial coordinate system, the forest area to be analyzed is divided into regular grid cells, and a unique spatial index is assigned to each grid cell and the coordinate range is recorded. By using spatial overlay analysis, the remote sensing data to be analyzed corresponding to the coordinate range of each grid cell is extracted, and the correlation mapping relationship between grid cells and remote sensing data is established. For each grid cell of remote sensing data to be analyzed, image processing techniques are used to identify the image boundaries of the forest area to be analyzed, distinguish between forest and non-forest areas, and obtain the two-dimensional outline of the forest area on the image. By combining the resolution of remote sensing images, the feature quantification value of the forest area in the image is statistically calculated, the second sub-value of the forest stand attribute influence to be analyzed is obtained simultaneously, and it is associated with the overall range of the attribute analysis spatial coordinate system.

[0009] In an optional embodiment, the step of analyzing the remote sensing data to be analyzed corresponding to each coordinate point in the attribute analysis spatial coordinate system to obtain the forest stand attribute index matrix specifically includes: Based on the dominant tree species composition, stand origin type and preset monitoring targets of the forest to be analyzed, a list of stand attribute indicators to be extracted is defined. Among them, the basic structural indicators include average diameter at breast height (DBH), average tree height, stand density and canopy coverage; the growth indicators include annual DBH growth, single tree volume and volume per unit area; and the health indicators include vegetation coverage, pest and disease infection rate and dead standing tree ratio. Sequentially acquire the remote sensing data to be analyzed corresponding to each regular grid cell in the attribute analysis spatial coordinate system. The remote sensing data to be analyzed includes multispectral remote sensing data, lidar data, and high-resolution optical images. Red band reflectance, near-infrared band reflectance, and vegetation index are extracted from multispectral remote sensing data; canopy height correlation features and point cloud density are extracted from lidar data; and texture features are extracted from high-resolution optical images to determine characteristic parameters related to forest stand attributes. Based on the feature parameters and index list, the dataset is divided into training dataset, validation dataset and test dataset. The training dataset consists of the feature parameters of the regular grid cells and the corresponding ground measured forest stand attribute index values. Based on the type of indicators in the training dataset, a suitable machine learning model is selected: a regression model is used for continuous forest stand attribute indicators, and a classification model is used for categorical forest stand attribute indicators. The training dataset is preprocessed, the selected machine learning model is trained using the preprocessed training dataset, and the model parameters are optimized using the validation dataset. The optimal combination of model parameters is determined with the goal of minimizing the prediction error. The test dataset is input into the optimized machine learning model to evaluate the model's generalization ability. The coefficient of determination is calculated for continuous indicators, and the classification performance score is calculated for categorical indicators. Set a model performance threshold, return unqualified machine learning models to the feature parameter extraction step or adjust the model type to obtain a standard machine learning model that meets the accuracy requirements; Input the feature parameters of all regular grid cells in the attribute analysis spatial coordinate system into the standard machine learning model to obtain the predicted value of forest stand attribute index for each grid cell; Based on the spatial arrangement of regular grid cells in the attribute analysis spatial coordinate system, the predicted forest stand attribute index values ​​of each coordinate point are sequentially used as rows / columns of a matrix to form a two-dimensional forest stand attribute index matrix.

[0010] In an optional embodiment, the step of obtaining a standard matrix of forest stand attributes based on the forest stand attribute index matrix and the second sub-value of the influence of the forest stand attributes to be analyzed, and formulating forest management strategies, specifically includes: Obtain information on tree types and forestry planning objectives for the forest to be analyzed, and determine the standard indicators and threshold ranges for forest stand attributes; Based on the second sub-value of the forest stand attribute influence and the area of ​​the grid cell in the attribute analysis spatial coordinate system, the area proportion of each regular grid cell in the overall forest area is obtained, and the area proportion coefficient is obtained. The forest stand attribute index value of each grid cell in the forest stand attribute index matrix is ​​multiplied by the area proportion coefficient to obtain the weighted index value. Then, the index is summarized across the entire area based on the index type to form a comprehensive attribute index set that reflects the overall characteristics of the forest. Using standard threshold ranges and comprehensive attribute index sets, a standard matrix of forest stand attributes is generated through matrix structuring. The total value of the forest stand attribute index matrix is ​​compared with the threshold of the standard matrix item by item to identify the indicators that deviate from the standard and the degree of deviation, and to formulate forest management strategies.

[0011] Furthermore, a forest stand attribute analysis system based on UAV remote sensing data feedback is proposed to implement any of the analysis methods mentioned above, including: The acquisition module is used to receive and store data of the forest area to be analyzed, including forest location data, and to perform format verification on the data. The UAV data and environmental information association module is used to query historical flight path data of UAVs through spatial analysis tools, associate remote sensing equipment information and multi-source environmental information, and provide a data foundation for area calculation and attribute analysis. The forest stand attribute impact module is used to combine remote sensing equipment information, UAV flight route data and environmental information, classify flight route data according to environmental parameters and evaluate data quality, calculate the actual area value of the forest area to be analyzed, and obtain the first sub-value of the forest stand attribute impact to be analyzed. The attribute analysis spatial coordinate system construction module is used to construct a unified spatial coordinate system based on the topography and data characteristics of the forest area, associate the remote sensing data to be analyzed and divide it into regular grid units, calculate the second sub-value of the influence of the forest stand attribute to be analyzed, and provide a spatial reference for the construction of the attribute index matrix. The forest stand attribute analysis and management strategy formulation module is used to extract feature parameters from remote sensing data, obtain a forest stand attribute index matrix through machine learning model analysis, construct a standard matrix by combining tree type and planning objectives, and formulate forest management strategies.

[0012] In an optional embodiment, the UAV data and environmental information association module includes: The flight route data query unit is used to filter historical flight route data of UAVs in the corresponding area based on the location data of the forest to be analyzed, and extract basic flight route information and flight route related files by using spatial analysis tools. The remote sensing equipment information extraction unit is used to query the equipment configuration file based on the UAV number and flight mission ID in the basic information of the flight route to determine the type and parameters of the remote sensing equipment carried on each flight. The multi-source environmental information integration unit is used to extract environmental parameters from the UAV flight log, obtain corresponding terrain parameters based on the geographical location information of the flight path coverage, and integrate the environmental parameters and terrain parameters to form multi-source environmental information corresponding to the UAV historical flight path data.

[0013] In an optional embodiment, the forest stand attribute influence module includes: The flight path data classification unit is used to classify flight path data into different adaptability categories according to environmental parameters based on the environmental adaptability requirements of UAV flight, and to assess the risk of flight path deviation and altitude deviation based on environmental parameters, and further filter high-quality data. The overlap analysis unit is used to calculate the actual overlap and planned overlap of different types of flight route data, determine whether the actual overlap meets the planning requirements, and provide a boundary processing basis for area calculation. The area value calculation unit is used to calculate the actual area of ​​the flat area based on the overlap matching result and the appropriate boundary processing method. It is also used to calculate the actual area of ​​the complex terrain by combining the terrain parameters of the complex terrain and the area calculation method applicable to the complex terrain, and the boundary processing method. Finally, it is used to summarize the actual areas of the flat area and the complex terrain to obtain the first sub-value of the influence of the forest stand attributes to be analyzed.

[0014] In an optional embodiment, the stand attribute analysis and management strategy formulation module includes: The attribute index extraction unit is used to define the list of forest stand attribute indicators, extract feature parameters from the remote sensing data to be analyzed in each grid unit, calculate the attribute index value of each grid through a machine learning model that meets the accuracy requirements, and integrate them into a forest stand attribute index matrix in spatial order. The standard matrix construction unit is used to determine the standard indicators and threshold ranges of forest stand attributes based on forest tree types and forestry planning objectives, calculate the area proportion coefficient of each grid unit, weight and summarize the indicator matrix to form a comprehensive attribute indicator set, and generate the standard matrix of forest stand attributes through matrix structuring. The management strategy formulation unit is used to compare the total value of the indicator matrix with the threshold of the standard matrix, identify the deviation indicators and the degree of deviation, and formulate targeted management strategies in combination with the actual forest area, topography and environmental information, and clarify the scope and priority of implementation.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This solution proposes a forest stand attribute analysis method based on UAV remote sensing data feedback. By acquiring data of the forest area to be analyzed, querying historical UAV flight path data, and associating it with remote sensing equipment information and multi-source environmental information, it achieves systematic integration of basic forest monitoring data, ensuring the spatial matching and timeliness of the data. By combining remote sensing equipment information, UAV flight path data, and multi-source environmental information, it classifies flight path data according to environmental parameter thresholds and calculates the actual area of ​​flat and complex terrains in a targeted manner, achieving accurate calculation of the actual area of ​​the forest area. It fully considers the impact of environmental factors on data quality and improves the reliability of area calculation. This proposal presents a method for analyzing forest stand attributes based on UAV remote sensing data feedback. By constructing a spatial coordinate system for attribute analysis, multi-source remote sensing data is uniformly transformed into this coordinate system and divided into regular grid units for data association. This achieves refined spatial discretization and data standardization of forest areas, eliminating analysis errors caused by topographic differences and inconsistent data formats. It lays a high-precision spatial benchmark for extracting forest stand attribute indicators. Subsequently, by extracting feature parameters from the grid unit remote sensing data, forest stand attribute indicators are inverted using machine learning models and integrated into an indicator matrix. This achieves quantitative and spatial expression of forest stand attributes, covering multi-dimensional indicators such as structure, growth, and health, providing refined data support for forest quality assessment. Attached Figure Description

[0016] Figure 1 This is a flowchart of a forest stand attribute analysis method based on UAV remote sensing data feedback proposed in this invention; Figure 2 This is a flowchart illustrating the acquisition of multi-source environmental information in this invention. Figure 3 This is a flowchart illustrating the process of obtaining the actual overlap of the first and second routes in this invention. Figure 4 This is a flowchart illustrating the process of obtaining the first sub-value of the influence of forest stand attributes on the forest to be analyzed in this invention. Figure 5 This is a system framework diagram of a forest stand attribute analysis system based on UAV remote sensing data feedback proposed in this invention. Detailed Implementation

[0017] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0018] Reference Figure 1 - Figure 5 As shown, a method for analyzing forest stand attributes based on UAV remote sensing data feedback includes: Acquire data on the forest area to be analyzed, including the location data of the forest to be analyzed; Based on the forest location data to be analyzed, query the historical flight path data of the UAV, and determine the information of the remote sensing data acquisition equipment carried by the corresponding UAV, as well as the multi-source environmental information corresponding to the historical flight path data of the UAV. Based on information from remote sensing data acquisition equipment, historical flight path data of UAVs, and corresponding multi-source environmental information, the first sub-value of the forest stand attribute influence to be analyzed is determined. Construct an attribute analysis spatial coordinate system and determine the remote sensing data to be analyzed corresponding to each coordinate point in the attribute analysis spatial coordinate system. Simultaneously, obtain the second sub-value of the forest stand attribute influence presented by the forest area to be analyzed from the remote sensing data to be analyzed. The remote sensing data to be analyzed is analyzed for each coordinate point in the attribute analysis spatial coordinate system to obtain the forest stand attribute index matrix. Based on the forest stand attribute index matrix and the second sub-value of the influence of the forest stand attributes to be analyzed, the standard matrix of forest stand attributes is obtained, and forest management strategies are formulated.

[0019] Furthermore, based on the forest location data to be analyzed, historical flight path data of the UAVs are queried, and the information of the remote sensing data acquisition equipment carried by the corresponding UAVs is determined, as well as the multi-source environmental information corresponding to the historical flight path data of the UAVs, specifically including: Based on the location data of the forest to be analyzed, determine the location range information of the forest to be analyzed, including latitude and longitude coordinates and administrative boundaries; Input the latitude and longitude coordinates and administrative region boundaries (Shapefile vector boundaries) into the geographic information system, and use the spatial analysis function of GIS to filter and display the historical flight path data of UAVs in the corresponding forest area to be analyzed on the map, including basic flight path information (flight date, UAV number, take-off and landing point coordinates, flight path coverage) and flight path associated files (storage path, data volume, resolution, etc. of original image / point cloud data). Based on basic route information, obtain the UAV flight date, UAV number, and UAV flight mission ID, query equipment configuration files, and determine the details of the remote sensing equipment carried on each flight; Based on the detailed information of the remote sensing equipment carried on each flight, the type and parameters of the remote sensing equipment are determined, thereby obtaining the information of the remote sensing data acquisition equipment carried by the corresponding UAV. Obtain the historical flight time range and drone number from the drone's historical flight path data; Based on the drone's flight date, flight logs are extracted from the drone's historical flight route data to determine the environmental parameters corresponding to the drone's flight date, including wind speed, wind pressure, flight altitude, and temperature. Based on historical flight path data of UAVs, the flight path coverage is extracted from the basic flight path information, and the latitude and longitude coordinates corresponding to the flight path coverage are extracted from the location range information of the forest to be analyzed. Based on the latitude and longitude coordinates corresponding to the coverage area of ​​the flight strip, determine the corresponding terrain parameters; By integrating the environmental parameters corresponding to the drone's flight date and the terrain parameters corresponding to the flight path coverage, multi-source environmental information corresponding to the drone's historical flight path data is obtained.

[0020] Specifically, in this solution, after determining the latitude and longitude coordinates and administrative boundaries based on the location data of the forest to be analyzed, the Shapefile vector boundary is imported into the GIS system. Spatial overlay analysis is used to filter historical UAV flight paths covering the area, extracting basic flight path information (flight date, UAV number, etc.) and associated files (image / point cloud storage paths, etc.). Then, based on the UAV number and mission ID in the flight path information, the equipment file is queried to determine the remote sensing equipment type and parameters. Simultaneously, the flight time range is extracted from the flight path data, and environmental parameters such as wind speed and wind pressure are obtained from the flight logs in conjunction with the flight date. Topographic parameters are extracted from the latitude and longitude coordinates corresponding to the flight path coverage area. Finally, environmental and topographic parameters are integrated to form multi-source environmental information. This process accurately links forest areas and historical UAV data through GIS spatial analysis, achieving a systematic integration of remote sensing equipment information, flight path data, and multi-source environmental information. This ensures both the spatial matching and timeliness of the data, and provides comprehensive and high-quality basic data for subsequent forest area calculation and stand attribute analysis. It effectively solves the problems of low analytical accuracy caused by data dispersion and insufficient consideration of environmental factors in traditional methods, improving the scientificity and reliability of forest stand attribute analysis.

[0021] Furthermore, based on information from remote sensing data acquisition equipment, historical flight path data of UAVs, and corresponding multi-source environmental information, the first sub-value affecting the forest stand attributes to be analyzed is determined, specifically including: Remote sensing equipment type information and remote sensing equipment parameters are extracted from remote sensing data acquisition equipment information. Simultaneously, the flight path coverage is extracted from UAV historical flight path data, and the environmental parameters corresponding to the UAV historical flight path data are extracted from multi-source environmental information. Set a threshold for safe flight parameters for drones. Use historical flight path data of drones with environmental parameters lower than the threshold as the first historical flight path data, and use historical flight path data of drones with environmental parameters higher than the threshold as the second historical flight path data. By using the environmental parameters corresponding to the first data of the drone's historical flight path, the risk value of flight path deviation is assessed, and the preset flight path data of the drone corresponding to the first data of the drone's historical flight path is determined simultaneously. Using the environmental parameters corresponding to the first data of the UAV's historical flight path, the historical flight altitude of the UAV is obtained. Simultaneously, the preset planned altitude of the UAV is obtained from the preset flight path data of the UAV, and the deviation value between the historical flight altitude of the UAV and the preset planned altitude of the UAV is obtained and recorded as the altitude deviation value. Set a flight path deviation risk threshold and an altitude deviation threshold. Record the first data of the UAV's historical flight path with a flight path deviation risk value exceeding the flight path deviation risk threshold as the third data of the UAV's historical flight path, and record the first data of the UAV's historical flight path with an altitude deviation value exceeding the altitude deviation threshold as the fourth data of the UAV's historical flight path. Obtain the coverage area of ​​the flight path corresponding to the second and third historical flight path data of the UAV, and then determine the actual overlap of the first flight path corresponding to the second and third historical flight path data of the UAV. Obtain the coverage area of ​​the flight path corresponding to the second and fourth historical flight path data of the UAV, and then determine the actual overlap of the second flight path corresponding to the second and fourth historical flight path data of the UAV. Obtain the overlap of the first route planning corresponding to the second and third historical route data of the UAV from the historical route data of the UAV; obtain the overlap of the second route planning corresponding to the second and fourth historical route data of the UAV from the historical route data of the UAV. If the actual overlap of the first route is greater than or equal to the planned overlap of the first route, the actual area value of the flat area is determined by the historical route data of the UAV and the coverage area of ​​each historical route. If the actual overlap of the first route is less than the planned overlap of the first route, then the adjacent historical route data of the UAV corresponding to the second and third historical route data of the UAV are obtained from the historical route data of the UAV, and the corresponding boundaries are filled in by the adjacent historical route data of the UAV to determine the actual area value of the flat area. If the actual overlap of the second route is greater than or equal to the planned overlap of the second route, then obtain the route coverage area corresponding to the second data of the UAV historical route and the fourth data of the UAV historical route, and then obtain all terrain parameters within the coverage area of ​​the route. Simultaneously, use the surface area formula to determine the actual area value of the complex terrain. If the actual overlap of the second route is less than the planned overlap of the second route, then the corresponding ground survey data corresponding to the second and fourth historical data of the UAV route are obtained to correct the corresponding boundary. Then, all terrain parameters within the route coverage area of ​​the historical UAV route data after the boundary correction are obtained, and the surface area formula is used simultaneously to determine the actual area value of the complex terrain. The summation of the actual area values ​​of flat areas and complex terrains yields the first sub-value of the influence of forest stand attributes on the analysis.

[0022] Specifically, the system extracts equipment type (e.g., visible light camera, LiDAR) and parameters (e.g., focal length, point cloud density) from remote sensing data acquisition equipment information. Simultaneously, it extracts the spatial range covered by the flight path from historical UAV flight path data and environmental parameters such as wind speed and wind pressure from multi-source environmental information. It sets safe flight parameter thresholds for UAVs (e.g., wind speed ≤ 5), and accordingly divides the flight path data into first data that meets environmental parameter standards and second data that exceeds them. Based on the environmental parameters of the first data, it assesses the risk value of flight path deviation using the wind pressure calculation formula (wind pressure = 0.613 × wind speed²), while simultaneously acquiring the corresponding preset flight path data and calculating the altitude deviation between historical flight altitude and preset planned altitude. Finally, it sets flight path deviation risk thresholds (e.g., deviation > 5m) and altitude deviation thresholds (e.g., deviation...). If the deviation exceeds 10%, the first data is further divided into the third data (offset exceeding the threshold) and the fourth data (height deviation exceeding the threshold). The actual overlap of the first route between the second and third data and the second route between the second and fourth data are calculated by GIS spatial overlay, and the corresponding planned overlap is extracted. If the first actual overlap is greater than or equal to the planned value, the flat area is calculated directly using the route range; otherwise, the boundary is completed by adjacent routes before calculation. If the second actual overlap is greater than or equal to the planned value, the complex terrain area is calculated using the surface area formula (actual area = projected area / cos(slope angle)) in combination with terrain parameters (slope, aspect); otherwise, the boundary is corrected by combining ground survey data before calculation using the formula. Finally, the areas of the two types of areas are summed to obtain the first sub-value of the influence of the forest stand attributes to be analyzed. This step, by refining the correlation analysis between remote sensing equipment parameters and environmental parameters, combined with quantitative threshold classification (such as safe flight threshold and drift risk threshold) and precise overlap comparison, adopts differentiated area calculation methods for different terrains (direct calculation for flat areas and surface formulas for complex terrains), and ensures data integrity through boundary completion or correction. This effectively improves the accuracy and reliability of the impact of forest stand attributes, providing accurate basic data support for subsequent forest stand attribute analysis and management strategy formulation.

[0023] It is understandable that "wind pressure = 0.613 × wind speed²" is a simplified calculation formula for drone flight scenarios, where the wind speed is in m / s and the result is wind pressure (in Pa). The derivation of this formula is based on standard air density (1.225 kg / m³) and a simplification factor (0.5 × air density ≈ 0.613). Historical flight path data for drones is a structured collection of data recording past flight missions of drones. It mainly includes basic flight parameters: flight date, time, drone number, flight mission ID, take-off and landing coordinates, flight altitude (actual altitude and planned altitude), flight speed, and planned flight path; spatial coverage information: flight path coverage (latitude and longitude boundaries), heading and lateral overlap (planned value and actual value), spatial resolution of image / point cloud data, etc.; associated data index: storage path, data volume, and file format of raw remote sensing data (such as RGB imagery, multispectral data, LiDAR point clouds), as well as the corresponding metadata (such as sensor operating status at the time of shooting); flight status records: attitude data (roll angle, pitch angle, yaw angle), GPS positioning accuracy, battery level, and other equipment operating information during flight.

[0024] Furthermore, the calculation of the actual overlap of the first route (second and third data) and the actual overlap of the second route (second and fourth data) are both based on spatial overlay analysis of the route coverage area. The specific steps are as follows: S1. Obtain the coverage area of ​​the flight strip corresponding to the second, third (or fourth) data from the historical flight route data of the UAV, and represent it as a polygonal vector boundary (such as a rectangle or irregular polygon) in latitude and longitude coordinates. S2. Calculate the actual geographic area (unit: square meters or hectares) of the overlapping area (intersection) of the two flight path boundaries using GIS spatial analysis tools (such as ArcGIS's Intersect function). S3. Calculate the independent coverage area (i.e., the spatial range area of ​​a single route) of the flight strip corresponding to the second data and the third data (or the fourth data). S4. Actual overlap of the first route = (intersection area of ​​the second and third data ÷ smaller single route area of ​​the two routes) × 100%; S5. Actual overlap of the second route = (intersection area of ​​the second data and the fourth data ÷ area of ​​the smaller single route of the two routes) × 100%.

[0025] Furthermore, an attribute analysis spatial coordinate system is constructed, and the remote sensing data to be analyzed corresponding to each coordinate point in the attribute analysis spatial coordinate system is determined. Simultaneously, the second sub-value of the forest stand attribute influence presented by the forest area to be analyzed is obtained from the remote sensing data to be analyzed, specifically including: Based on the first sub-value of the influence of forest stand attributes on the forest to be analyzed and the data of the forest region to be analyzed, the first forest region to be analyzed corresponding to the actual area value of flat areas is obtained, and the second forest region to be analyzed corresponding to the actual area value of complex terrain is obtained simultaneously. Obtain all actual geographical location information of the first region of the forest to be analyzed, corresponding to the actual area value of the flat area, and record it as the latitude and longitude range of the first region of the forest to be analyzed. Obtain all actual geographical location information of the second forest region to be analyzed, corresponding to the actual area value of the complex terrain, and record it as the latitude and longitude range of the second forest region to be analyzed. The latitude and longitude range of the first forest region to be analyzed is projected onto the basic coordinate system of the GIS software through the Gauss-Kruger 3-degree zone, and the central meridian control points are densified to obtain the spatial coordinate system for attribute analysis. An elevation datum is set, and the GPS geodetic height corresponding to the latitude and longitude range of the second forest region to be analyzed is corrected for deviation. The remote sensing data corresponding to the historical flight path data of UAVs in the latitude and longitude range of the first and second forest areas to be analyzed were uniformly converted into the attribute analysis spatial coordinate system using the projection conversion function in the GIS software. In the attribute analysis spatial coordinate system, the forest area to be analyzed is divided into regular grid cells, each grid cell is assigned a unique spatial index, and the coordinate range of each regular grid cell is recorded. By using spatial overlay analysis, the remote sensing data corresponding to the coordinate range of each regular grid cell is extracted and used as the remote sensing data to be analyzed for the regular grid cell. A grid cell-remote sensing data association mapping table is then established. For each regular grid cell of remote sensing data to be analyzed, an image segmentation algorithm is used to identify the image boundary of the forest area to be analyzed, distinguish between forest and non-forest areas, and obtain the two-dimensional contour of the forest area on the image. Based on the two-dimensional outline of the forest area in the image, combined with the resolution of the remote sensing image, the total number of pixels occupied by the forest area in the image is counted, the second sub-value of the forest stand attribute influence to be analyzed is obtained simultaneously, and associated with the overall range of the attribute analysis spatial coordinate system.

[0026] Specifically, based on the first sub-value of the forest stand attribute influence to be analyzed and the regional basic data, the area corresponding to the actual area of ​​flat areas is divided into the first region of the forest to be analyzed, and the area corresponding to the actual area of ​​complex terrain is divided into the second region of the forest to be analyzed. The actual geographical location information of the first and second regions is extracted to obtain the corresponding latitude and longitude ranges (e.g., the first region is 116°-117°E, 39°-40°N). The latitude and longitude range of the first region is transformed to the GIS basic coordinate system through Gauss-Kruger 3-degree zone projection (the central meridian is determined according to the longitude, e.g., 117°E corresponds to the 3-degree zone central meridian 117°). At the same time, the central meridian control points are densified (one control point is added every 5km) to improve accuracy, forming an attribute analysis spatial coordinate system. The elevation datum is set as the 1985 National Elevation Datum, and the GPS geodetic height of the second region is calculated. Line deviation correction (correction amount = measured elevation - geodetic height); the remote sensing data (images, point clouds) of the two regions are uniformly converted to this coordinate system through the GIS projection conversion function; the coordinate system is divided into regular grid units of 10m×10m, a unique spatial index is assigned (such as row and column numbers R1C1, R1C2) and the coordinate range is recorded; the remote sensing data corresponding to each grid is extracted through spatial overlay analysis, and a "grid-data" mapping table is established; the remote sensing data of each grid is used to identify the forest boundary using an image segmentation algorithm based on NDVI threshold (NDVI>0.3 is judged as forest) to obtain a two-dimensional outline; combined with the resolution of the remote sensing image (such as 1 pixel = 0.5m×0.5m, the area of ​​a single pixel is 0.25㎡), the total number of pixels in the forest area is counted, the second sub-value of the influence of the forest stand attribute to be analyzed (total number of pixels × 0.25㎡) is calculated and associated with the overall range of the coordinate system.

[0027] Understandably, this scheme constructs a unified spatial coordinate system using Gauss-Kruger projection, and combines elevation correction and data transformation to ensure spatial consistency of multi-source remote sensing data. Regular grid division enables refined spatial discretization of forest areas, facilitating point-by-point analysis of subsequent attribute indicators. Image segmentation and area calculation methods accurately distinguish between forest and non-forest areas, ensuring the accuracy of image area occupancy. The overall process provides a standardized, high-precision spatial benchmark for constructing the forest stand attribute indicator matrix, effectively eliminating analytical errors caused by topographic differences and inconsistent data formats, and improving the reliability and spatial correlation of subsequent forest stand attribute analysis.

[0028] Furthermore, the remote sensing data corresponding to each coordinate point in the attribute analysis spatial coordinate system is analyzed to obtain the forest stand attribute index matrix, specifically including: Based on the dominant tree species composition, stand origin type and preset monitoring targets of the forest to be analyzed, a list of stand attribute indicators to be extracted is defined. Among them, the basic structural indicators include average diameter at breast height (DBH), average tree height, stand density and canopy coverage; the growth indicators include annual DBH growth, single tree volume and volume per unit area; and the health indicators include vegetation coverage, pest and disease infection rate and dead standing tree ratio. The remote sensing data to be analyzed is obtained sequentially for each regular grid cell in the attribute analysis spatial coordinate system. The remote sensing data to be analyzed includes multispectral remote sensing data, lidar data, and high-resolution optical images. Red band reflectance, near-infrared band reflectance, and vegetation index are extracted from multispectral remote sensing data; maximum canopy height, average canopy height, canopy volume, and point cloud density are extracted from lidar data; and texture features are extracted from high-resolution optical images to determine characteristic parameters related to forest stand attributes. Based on the feature parameters and index list, the training dataset, validation dataset and test dataset are divided. The training dataset consists of the feature parameters of the regular grid cells and the corresponding ground measured forest stand attribute index values. The validation dataset is used for model parameter tuning and the test dataset is used to evaluate model accuracy. Based on the training dataset, machine learning models are selected. For continuous forest stand attribute indicators, a random forest regression model is used, and for categorical forest stand attribute indicators, a random forest classification model is used. The training dataset is preprocessed, including standardizing the feature parameters (making the mean of the feature values ​​0 and the standard deviation 1), supplementing missing values ​​with K-nearest neighbor interpolation, and identifying and removing outliers using the Z-score method. The selected machine learning model is trained using the training dataset, and the trained machine learning model is optimized, including adjusting the model hyperparameters through grid search or Bayesian optimization. For the random forest model, the number of decision trees, the maximum tree depth, and the minimum number of split samples are adjusted. For the GBDT model, the learning rate, the number of iterations, and the number of leaf nodes are adjusted. The optimal parameter combination is determined with the goal of minimizing the prediction error of the dataset (RMSE is used for continuous indicators, and classification accuracy is used for categorical indicators). Input the test dataset into the optimized machine learning model to evaluate the model's generalization ability. Calculate the coefficient of determination R² for continuous metrics and the F1 score for categorical metrics. Set a threshold for the coefficient of determination and a threshold for the F1 score. Return machine learning models that do not meet the threshold to the feature parameter extraction step or change the model type to obtain a standard machine learning model. R² ≥ 0.7 is considered a valid model, and F1 ≥ 0.7 is considered a valid model. Input the feature parameters of all regular grid cells in the attribute analysis spatial coordinate system into the standard machine learning model to obtain the predicted value of forest stand attribute index for each grid cell; According to the spatial arrangement order of the regular grid cells in the attribute analysis spatial coordinate system, the predicted values ​​of forest stand attribute indicators of each coordinate point in the attribute analysis spatial coordinate system are used as the rows / columns of the matrix in turn, and integrated to form a two-dimensional forest stand attribute indicator matrix. Among them, continuous indicators are retained to two decimal places, and categorical indicators are represented by category labels (such as "conifers" and "healthy").

[0029] Specifically, based on the type of forest to be analyzed (e.g., coniferous forest, broadleaf forest) and monitoring objectives (e.g., ecological protection, carbon sequestration), basic structural indicators (tree height, diameter at breast height, crown width, stand density), growth indicators (biomass, volume), and health indicators (NDVI, pest and disease severity) are defined to form an indicator list. For the remote sensing data to be analyzed in each regular grid cell of the attribute analysis spatial coordinate system, canopy texture features (contrast and entropy values ​​calculated through gray-level co-occurrence matrix) and color features (RGB band mean) are extracted from optical images, and elevation features (maximum elevation, average elevation, point cloud density) are extracted from LiDAR point clouds. The vegetation index (NDVI = (near-infrared band - red band) / (near-infrared band + red band)) is calculated from multispectral data. The extracted feature parameters are input into a trained machine learning model (e.g., random forest, CNN), and the model outputs the indicator value for each grid cell, such as tree height. The difference between the LiDAR maximum elevation and the ground DEM is calculated. Biomass is estimated by combining diameter at breast height (DBH) and tree height with the allometric growth equation W=a×DBH^b×H^c. Here, W represents the biomass of a single tree or stand (usually in kilograms or tons), which is the dependent variable of the equation, i.e., the target value obtained through the equation. DBH represents the diameter at breast height (usually in centimeters), which is the diameter of the trunk at 1.3 meters above the ground and is a key structural parameter reflecting the growth status of trees. H represents the tree height (usually in meters), which is the vertical height of the tree from the ground to the crown. a, b, and c are parameters (constants) of the equation. The parameter values ​​are different for different tree species or stand types and need to be obtained by fitting measured data. They are used to quantify the nonlinear relationship between DBH, tree height, and biomass. According to the spatial arrangement order (row and column number) of the grid cells, the index value vector of each cell is used as the row of the matrix to form a forest stand attribute index matrix.

[0030] Understandably, this approach selects forest stand attribute indicators in a targeted manner, extracts feature parameters from multi-source remote sensing data, and uses machine learning models to achieve accurate inversion of indicator values, ultimately forming a structured indicator matrix. This ensures the comprehensiveness of indicator extraction (covering structure, growth, and health dimensions) and realizes the spatial expression of indicators through the spatial correlation of grid cells. It provides refined and quantitative basic data for subsequent comparison with standard matrices and the formulation of management strategies, effectively improving the systematicness and accuracy of forest stand attribute analysis.

[0031] Furthermore, based on the forest stand attribute index matrix and the second sub-value of the influence of the forest stand attributes to be analyzed, a standard matrix of forest stand attributes is obtained, and forest management strategies are formulated, specifically including: Obtain information on tree types and forestry planning objectives for the forest to be analyzed, and determine the standard indicators and threshold ranges for forest stand attributes; Based on the second sub-value of the forest stand attribute influence and the area of ​​the grid cell in the attribute analysis spatial coordinate system, the area proportion of each regular grid cell in the overall forest area is obtained, and the area proportion coefficient is obtained. The forest stand attribute index value of each grid cell in the forest stand attribute index matrix is ​​multiplied by the area proportion coefficient to obtain the weighted index value. Then, the index is summarized across the entire area based on the index type to form a comprehensive attribute index set that reflects the overall characteristics of the forest. Using standard threshold ranges and comprehensive attribute index sets, a standard matrix of forest stand attributes is generated through matrix structuring. The total value of the forest stand attribute index matrix is ​​compared with the threshold of the standard matrix item by item to identify the indicators that deviate from the standard and the degree of deviation, and to formulate forest management strategies.

[0032] Specifically, this scheme obtains the tree types (such as Masson pine and Chinese fir) and forestry planning objectives (such as timber forest cultivation and ecological protection zone construction) of the forest to be analyzed through field surveys or forestry archives. Based on this, it determines the standard indicators of forest stand attributes (such as diameter at breast height and volume of timber forests, and NDVI and canopy closure of ecological forests) and the standard threshold range (such as the NDVI threshold of 0.6-0.8 for healthy Masson pine stands). According to the second sub-value of the influence of forest stand attributes on the forest to be analyzed (such as 1000 hectares) and the area of ​​the grid unit in the attribute analysis spatial coordinate system (such as 10m×10m=0.01 hectares), the area proportion coefficient of each grid is calculated (0.01÷1000=0.00001). The index value of each grid in the forest stand attribute index matrix (such as NDVI=0.5 for a certain grid) is compared with the corresponding Multiply the area proportion coefficients to obtain the weighted index value (0.5 × 0.00001 = 0.000005). Then, summarize the weighted values ​​of all grids according to the index type (such as NDVI, diameter at breast height) to form a comprehensive attribute index set (such as the weighted average NDVI of the whole area = 0.55). Perform matrix structuring with the standard threshold range as columns and the comprehensive attribute index as columns (such as rows representing "NDVI", columns representing "excellent (0.6-0.8), medium (0.4-0.6), poor (<0.4)") to generate a standard matrix of forest stand attributes. Compare the total value of the index matrix of the whole area (such as 0.55) with the threshold of the standard matrix to identify the deviation index (NDVI is lower than the "excellent" level) and the degree of deviation (0.05), and formulate targeted management strategies (such as replanting in areas with NDVI < 0.6).

[0033] Understandably, this step ensures the relevance and practicality of the standard matrix by combining forest type and planning objectives to determine standard indicators and thresholds; it achieves precise integration of grid-scale data into full-area characteristics by using area proportion coefficients to weight and summarize indicator values; and it clarifies indicator deviations through matrix comparison, providing a quantitative basis for management strategies. The overall process closely integrates forest stand attribute analysis with actual forestry needs, enhancing the scientific nature and operability of management strategies and effectively solving the problems of vague standards and insufficient targeted measures in traditional management.

[0034] Furthermore, the total values ​​of the forest stand attribute index matrix are compared item by item with the threshold values ​​of the standard matrix to identify indicators that deviate from the standard and the degree of deviation, and to formulate forest management strategies, specifically including: The total values ​​of the forest stand attribute index matrix (such as overall average tree height, weighted NDVI, average biomass, etc.) are compared one by one with the threshold ranges of the forest stand attribute standard matrix (such as NDVI of 0.6-0.8 for healthy stands and average tree height ≥15m for high-quality stands) to determine whether each index is within the standard threshold. For example, if the total weighted NDVI is 0.5, which is lower than the lower limit of the standard threshold of 0.6, it is marked as a "deviation index", and the degree of deviation is calculated (0.6-0.5=0.1); if the average biomass of a certain area is 120 tons / hectare, which is within the standard threshold of 100-150 tons / hectare, it is marked as "compliant with the standard". The deviation indicators are classified according to the degree of deviation (e.g., slight deviation: deviation value < 10% of the standard threshold; moderate deviation: 10% to 30%; severe deviation: > 30%), and combined with the attribute analysis spatial coordinate system, the specific grid unit corresponding to the deviation indicator is located (e.g., the area with NDVI < 0.6 is concentrated in the grid between coordinates (X1, Y1) and (X2, Y2)), and the spatial distribution characteristics of the deviation area are clarified (e.g., concentrated in steep slope areas or forest edge areas).

[0035] The targeted management strategies include: for minor deviations from the indicators (such as canopy closure being slightly below the standard): develop monitoring enhancement strategies, increase the frequency of drone inspections (such as once per quarter), and focus on tracking the changing trends of the indicators; For moderate deviations from the standard (e.g., tree height is 20% below the standard): Develop growth promotion measures based on terrain parameters (e.g., slope < 25°), such as precision fertilization (calculating fertilization amount based on grid unit area) and pruning competing branches; For severely deviated indicators (such as NDVI < 0.4 and concentrated in arid areas): develop ecological restoration plans, including replanting suitable tree species (such as drought-resistant conifers), constructing irrigation facilities (planning the irrigation range based on the second sub-value of the forest stand attributes to be analyzed), and clarifying the implementation priority (such as restoring degraded forest edge areas first). For indicators that meet the standards (such as biomass being in the excellent range): formulate maintenance strategies, such as setting logging ban periods and controlling human interference, to ensure the stability of forest stand attributes.

[0036] Afterwards, it is necessary to implement the strategy and evaluate its effectiveness. The proposed management strategy will be broken down into specific tasks according to spatial grid units (e.g., 50 seedlings need to be replanted in a certain grid unit). An implementation plan will be formulated by combining the first sub-value of the forest stand attributes to be analyzed and the terrain parameters (e.g., manual tending instead of mechanical operations will be used in areas with a slope > 25°). Subsequently, the forest stand attribute index matrix will be recalculated through regular monitoring using UAV remote sensing data (e.g., once every six months) to evaluate the effectiveness of the strategy implementation and dynamically adjust management measures.

[0037] Furthermore, a forest stand attribute analysis system based on UAV remote sensing data feedback is proposed to implement any of the analysis methods mentioned above, including: The acquisition module is used to receive and store data of the forest area to be analyzed, including forest location data, and to perform data format verification. The module for linking UAV data with environmental information is used to query historical flight data of UAVs, link remote sensing equipment information and multi-source environmental information, and provide a data foundation for area calculation and attribute analysis. The Forest Stand Attribute Impact Module is used to calculate the actual area of ​​the forest area to be analyzed by combining information from remote sensing equipment, UAV flight path data and environmental information. The attribute analysis spatial coordinate system construction module is used to construct a unified spatial coordinate system, associate the remote sensing data to be analyzed, calculate the second sub-value of the attribute influence of the forest stand to be analyzed, and provide a spatial benchmark for the construction of the attribute index matrix. The forest stand attribute analysis and management strategy formulation module is used to analyze remote sensing data to obtain a forest stand attribute index matrix, construct a standard matrix, and formulate forest management strategies.

[0038] Furthermore, the module for linking drone data with environmental information includes: The flight route data query unit is used to filter historical UAV flight route data for the corresponding area based on the forest location data to be analyzed, and extract basic flight route information and flight route related files through GIS spatial analysis functions. The remote sensing equipment information extraction unit is used to query the equipment configuration file based on the UAV number and flight mission ID in the basic information of the flight route, and determine the type and parameters of the remote sensing equipment carried on each flight. The multi-source environmental information integration unit is used to extract environmental parameters from the UAV flight logs, obtain the corresponding terrain parameters based on the latitude and longitude coordinates of the flight path coverage, and integrate the environmental parameters and terrain parameters to form multi-source environmental information corresponding to the UAV's historical flight path data.

[0039] Furthermore, the forest stand attribute impact module includes: The flight path data classification unit is used to set threshold values ​​for safe flight parameters of the UAV, and divides the flight path data into first data and second data. The first data is used to evaluate the risk value of flight path deviation and altitude deviation value, and is further classified into third data and fourth data. The overlap analysis unit is used to calculate the actual overlap of the first route between the second data and the third data, and the actual overlap of the second route between the second data and the fourth data. It is used to extract the corresponding planning overlap and determine whether the actual overlap meets the planning requirements. The area calculation unit is used to calculate the actual area of ​​flat areas directly based on the coverage of the first route if the actual overlap is greater than or equal to the planned overlap. If not, the boundary is filled in by adjacent route data before calculation. If the actual overlap of the second route is greater than or equal to the planned overlap, the actual area of ​​complex terrain is calculated by combining terrain parameters and the surface area formula. If not, the boundary is corrected by combining ground survey data before calculation. This unit is used to summarize the actual areas of flat areas and complex terrain to obtain the first sub-value of the influence of forest stand attributes to be analyzed.

[0040] Furthermore, the stand attribute analysis and management strategy formulation module includes: The attribute index extraction unit is used to define the list of forest stand attribute indicators, extract feature parameters from the remote sensing data to be analyzed in each grid unit, and calculate the attribute index value of each grid unit through a machine learning model, and integrate them into a forest stand attribute index matrix in spatial order. The standard matrix construction unit is used to determine the standard indicators and threshold ranges of forest stand attributes based on forest tree types and forestry planning objectives. It is also used to calculate the area proportion coefficient of each grid unit, weight and summarize the indicator matrix to form a comprehensive attribute indicator set, which is used to generate the standard matrix of forest stand attributes through matrix structuring. The management strategy formulation unit compares the total value of the indicator matrix with the threshold of the standard matrix to identify deviation indicators and their degree. Combining the actual forest area, topography, and environmental information, it formulates targeted management strategies and clarifies the scope and priority of implementation.

[0041] The advantages of this invention are as follows: By systematically integrating historical flight path data from UAVs, remote sensing equipment information, and multi-source environmental information, and combining environmental parameter thresholds to classify and process flight path data, the actual area of ​​different terrains is calculated in a targeted manner, thus achieving accurate calculation of forest area. By constructing a unified attribute analysis spatial coordinate system, multi-source remote sensing data is standardized and associated with regular grid cells. A matrix of forest stand attribute indicators is formed using machine learning models. A standard matrix is ​​constructed by combining tree type and planning objectives, and management strategies are formulated. This comprehensively covers the entire process from data acquisition, processing, and analysis to strategy formulation. It fully considers the impact of environmental factors on data quality and achieves quantitative and spatial expression of forest stand attributes, effectively improving the accuracy of analysis and the pertinence of management strategies, providing efficient and scientific technical support for the refined management of forest resources.

[0042] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for analyzing forest stand attributes based on UAV remote sensing data feedback, characterized in that, include: Acquire data on the forest area to be analyzed, including the location data of the forest to be analyzed; Based on the forest location data to be analyzed, query the historical flight path data of the UAV, and determine the information of the remote sensing data acquisition equipment carried by the corresponding UAV, as well as the multi-source environmental information corresponding to the historical flight path data of the UAV. Based on information from remote sensing data acquisition equipment, historical flight path data of UAVs, and corresponding multi-source environmental information, the first sub-value of the forest stand attribute influence to be analyzed is determined. Construct an attribute analysis spatial coordinate system and determine the remote sensing data to be analyzed corresponding to each coordinate point in the attribute analysis spatial coordinate system. Simultaneously, obtain the second sub-value of the forest stand attribute influence presented by the forest area to be analyzed from the remote sensing data to be analyzed. The remote sensing data to be analyzed is analyzed for each coordinate point in the attribute analysis spatial coordinate system to obtain the forest stand attribute index matrix. Based on the forest stand attribute index matrix and the second sub-value of the influence of the forest stand attributes to be analyzed, the standard matrix of forest stand attributes is obtained, and forest management strategies are formulated.

2. The method for analyzing forest stand attributes based on UAV remote sensing data feedback according to claim 1, characterized in that, The process of querying historical flight path data of drones based on the forest location data to be analyzed, and determining the information of the remote sensing data acquisition equipment carried by the corresponding drones, as well as the multi-source environmental information corresponding to the historical flight path data of the drones, specifically includes: Based on the location data of the forest to be analyzed, determine the location range information of the forest to be analyzed, including latitude and longitude coordinates and administrative boundaries; Input the latitude and longitude coordinates and administrative boundaries into the spatial analysis tool, and use the spatial analysis function to filter and obtain the historical flight data of UAVs in the corresponding forest area to be analyzed, including basic flight information and flight association files. Based on basic route information, obtain the UAV flight date, UAV number, and UAV flight mission ID, query equipment configuration files, and determine the details of the remote sensing equipment carried on each flight; Based on the detailed information of the remote sensing equipment carried on each flight, the type and parameters of the remote sensing equipment are determined, thereby obtaining the information of the remote sensing data acquisition equipment carried by the corresponding UAV. The historical flight time range and drone number of the drone are obtained from the drone's historical flight route data. The flight log is extracted in combination with the flight date to determine the environmental parameters corresponding to the flight. Based on the flight strip coverage of historical UAV flight data, combined with the location range information of the forest to be analyzed, the geographical location information corresponding to the flight strip coverage is extracted, and the corresponding terrain parameters are determined. By integrating the environmental parameters corresponding to the flight path with the terrain parameters corresponding to the flight path coverage area, multi-source environmental information corresponding to the historical flight path data of the UAV is obtained.

3. The method for analyzing forest stand attributes based on UAV remote sensing data feedback according to claim 1, characterized in that, The process of determining the first sub-value of the forest stand attribute influence to be analyzed based on remote sensing data acquisition equipment information, UAV historical flight path data, and corresponding multi-source environmental information specifically includes: Remote sensing equipment type information and remote sensing equipment parameters are extracted from remote sensing data acquisition equipment information. Simultaneously, the flight path coverage is extracted from UAV historical flight path data, and the environmental parameters corresponding to the UAV historical flight path data are extracted from multi-source environmental information. Based on the environmental adaptability requirements of UAV flight, the historical flight path data of UAVs are classified according to environmental parameters, resulting in two types of flight path data with different environmental adaptability. Based on the environmental parameters corresponding to the classified route data, the risk of route deviation and altitude deviation are assessed, and route data that meets the data quality requirements are further screened. Calculate the actual overlap and planned overlap of different categories of route data after filtering. Based on the matching relationship between the actual overlap and the planned overlap, use an appropriate boundary processing method to determine the actual area value of the flat area. For areas with complex terrain, the actual area value of the complex terrain is determined by combining its terrain parameters and appropriate boundary processing methods through an area calculation method suitable for complex terrain. The summation of the actual area values ​​of flat areas and complex terrains yields the first sub-value of the influence of forest stand attributes on the analysis.

4. The method for analyzing forest stand attributes based on UAV remote sensing data feedback according to claim 1, characterized in that, The process of constructing an attribute analysis spatial coordinate system and determining the remote sensing data to be analyzed corresponding to each coordinate point in the attribute analysis spatial coordinate system, and simultaneously obtaining the second sub-value of the forest stand attribute influence of the forest area to be analyzed from the remote sensing data to be analyzed, specifically includes: Based on the first sub-value of the influence of forest stand attributes on the forest to be analyzed and the data of the forest region to be analyzed, the first region of the forest to be analyzed corresponding to flat areas and the second region of the forest to be analyzed corresponding to complex terrain areas are distinguished. Obtain the actual geographical location information of the first and second regions of the forest to be analyzed, and determine the corresponding latitude and longitude ranges; Based on the terrain features and data adaptation requirements of the forest area to be analyzed, a unified spatial coordinate system for attribute analysis is constructed. Through the projection transformation function of the spatial analysis tool, the historical flight path data of UAVs in the first and second areas and the corresponding remote sensing data to be analyzed are uniformly transformed to this coordinate system. In the attribute analysis spatial coordinate system, the forest area to be analyzed is divided into regular grid cells, and a unique spatial index is assigned to each grid cell and the coordinate range is recorded. By using spatial overlay analysis, the remote sensing data to be analyzed corresponding to the coordinate range of each grid cell is extracted, and the correlation mapping relationship between grid cells and remote sensing data is established. For each grid cell of remote sensing data to be analyzed, image processing techniques are used to identify the image boundaries of the forest area to be analyzed, distinguish between forest and non-forest areas, and obtain the two-dimensional outline of the forest area on the image. By combining the resolution of remote sensing images, the feature quantification value of the forest area in the image is statistically calculated, the second sub-value of the forest stand attribute influence to be analyzed is obtained simultaneously, and it is associated with the overall range of the attribute analysis spatial coordinate system.

5. The method for analyzing forest stand attributes based on UAV remote sensing data feedback according to claim 1, characterized in that, The analysis of the remote sensing data to be analyzed corresponding to each coordinate point in the attribute analysis spatial coordinate system to obtain the forest stand attribute index matrix specifically includes: Based on the dominant tree species composition, stand origin type and preset monitoring targets of the forest to be analyzed, a list of stand attribute indicators to be extracted is defined. Among them, the basic structural indicators include average diameter at breast height (DBH), average tree height, stand density and canopy coverage; the growth indicators include annual DBH growth, single tree volume and volume per unit area; and the health indicators include vegetation coverage, pest and disease infection rate and dead standing tree ratio. Sequentially acquire the remote sensing data to be analyzed corresponding to each regular grid cell in the attribute analysis spatial coordinate system. The remote sensing data to be analyzed includes multispectral remote sensing data, lidar data, and high-resolution optical images. Red band reflectance, near-infrared band reflectance, and vegetation index are extracted from multispectral remote sensing data; canopy height correlation features and point cloud density are extracted from lidar data; and texture features are extracted from high-resolution optical images to determine characteristic parameters related to forest stand attributes. Based on the feature parameters and index list, the dataset is divided into training dataset, validation dataset and test dataset. The training dataset consists of the feature parameters of the regular grid cells and the corresponding ground measured forest stand attribute index values. Based on the type of indicators in the training dataset, a suitable machine learning model is selected: a regression model is used for continuous forest stand attribute indicators, and a classification model is used for categorical forest stand attribute indicators. The training dataset is preprocessed, the selected machine learning model is trained using the preprocessed training dataset, and the model parameters are optimized using the validation dataset. The optimal combination of model parameters is determined with the goal of minimizing the prediction error. The test dataset is input into the optimized machine learning model to evaluate the model's generalization ability. The coefficient of determination is calculated for continuous indicators, and the classification performance score is calculated for categorical indicators. Set a model performance threshold, return unqualified machine learning models to the feature parameter extraction step or adjust the model type to obtain a standard machine learning model that meets the accuracy requirements; Input the feature parameters of all regular grid cells in the attribute analysis spatial coordinate system into the standard machine learning model to obtain the predicted value of forest stand attribute index for each grid cell; Based on the spatial arrangement of regular grid cells in the attribute analysis spatial coordinate system, the predicted forest stand attribute index values ​​of each coordinate point are sequentially used as rows / columns of a matrix to form a two-dimensional forest stand attribute index matrix.

6. The method for analyzing forest stand attributes based on UAV remote sensing data feedback according to claim 1, characterized in that, The process involves obtaining a standard matrix of forest stand attributes based on the forest stand attribute index matrix and the second sub-values ​​influencing the attributes of the forest stands to be analyzed, and formulating forest management strategies, specifically including: Obtain information on tree types and forestry planning objectives for the forest to be analyzed, and determine the standard indicators and threshold ranges for forest stand attributes; Based on the second sub-value of the forest stand attribute influence and the area of ​​the grid cell in the attribute analysis spatial coordinate system, the area proportion of each regular grid cell in the overall forest area is obtained, and the area proportion coefficient is obtained. The forest stand attribute index value of each grid cell in the forest stand attribute index matrix is ​​multiplied by the area proportion coefficient to obtain the weighted index value. Then, the index is summarized across the entire area based on the index type to form a comprehensive attribute index set that reflects the overall characteristics of the forest. Using standard threshold ranges and comprehensive attribute index sets, a standard matrix of forest stand attributes is generated through matrix structuring. The total value of the forest stand attribute index matrix is ​​compared with the threshold of the standard matrix item by item to identify the indicators that deviate from the standard and the degree of deviation, and to formulate forest management strategies.

7. A forest stand attribute analysis system based on UAV remote sensing data feedback, used to implement the analysis method as described in any one of claims 1-6, characterized in that, include: The acquisition module is used to receive and store data of the forest area to be analyzed, including forest location data, and to perform format verification on the data. The UAV data and environmental information association module is used to query historical flight path data of UAVs through spatial analysis tools, associate remote sensing equipment information and multi-source environmental information, and provide a data foundation for area calculation and attribute analysis. The forest stand attribute impact module is used to combine remote sensing equipment information, UAV flight route data and environmental information, classify flight route data according to environmental parameters and evaluate data quality, calculate the actual area value of the forest area to be analyzed, and obtain the first sub-value of the forest stand attribute impact to be analyzed. The attribute analysis spatial coordinate system construction module is used to construct a unified spatial coordinate system based on the topography and data characteristics of the forest area, associate the remote sensing data to be analyzed and divide it into regular grid units, calculate the second sub-value of the influence of the forest stand attribute to be analyzed, and provide a spatial reference for the construction of the attribute index matrix. The forest stand attribute analysis and management strategy formulation module is used to extract feature parameters from remote sensing data, obtain a forest stand attribute index matrix through machine learning model analysis, construct a standard matrix by combining tree type and planning objectives, and formulate forest management strategies.

8. A forest stand attribute analysis system based on UAV remote sensing data feedback according to claim 7, characterized in that, The module for associating UAV data with environmental information includes: The flight route data query unit is used to filter historical flight route data of UAVs in the corresponding area based on the location data of the forest to be analyzed, and extract basic flight route information and flight route related files by using spatial analysis tools. The remote sensing equipment information extraction unit is used to query the equipment configuration file based on the UAV number and flight mission ID in the basic information of the flight route to determine the type and parameters of the remote sensing equipment carried on each flight. The multi-source environmental information integration unit is used to extract environmental parameters from the UAV flight log, obtain corresponding terrain parameters based on the geographical location information of the flight path coverage, and integrate the environmental parameters and terrain parameters to form multi-source environmental information corresponding to the UAV historical flight path data.

9. A forest stand attribute analysis system based on UAV remote sensing data feedback according to claim 7, characterized in that, The forest stand attribute influence module includes: The flight path data classification unit is used to classify flight path data into different adaptability categories according to environmental parameters based on the environmental adaptability requirements of UAV flight, and to assess the risk of flight path deviation and altitude deviation based on environmental parameters, and further filter high-quality data. The overlap analysis unit is used to calculate the actual overlap and planned overlap of different types of flight route data, determine whether the actual overlap meets the planning requirements, and provide a boundary processing basis for area calculation. The area value calculation unit is used to calculate the actual area of ​​the flat area based on the overlap matching result and the appropriate boundary processing method. It is also used to calculate the actual area of ​​the complex terrain by combining the terrain parameters of the complex terrain and the area calculation method applicable to the complex terrain, and the boundary processing method. Finally, it is used to summarize the actual areas of the flat area and the complex terrain to obtain the first sub-value of the influence of the forest stand attributes to be analyzed.

10. A forest stand attribute analysis system based on UAV remote sensing data feedback according to claim 7, characterized in that, The forest stand attribute analysis and management strategy formulation module includes: The attribute index extraction unit is used to define the list of forest stand attribute indicators, extract feature parameters from the remote sensing data to be analyzed in each grid unit, calculate the attribute index value of each grid through a machine learning model that meets the accuracy requirements, and integrate them into a forest stand attribute index matrix in spatial order. The standard matrix construction unit is used to determine the standard indicators and threshold ranges of forest stand attributes based on forest tree types and forestry planning objectives, calculate the area proportion coefficient of each grid unit, weight and summarize the indicator matrix to form a comprehensive attribute indicator set, and generate the standard matrix of forest stand attributes through matrix structuring. The management strategy formulation unit is used to compare the total value of the indicator matrix with the threshold of the standard matrix, identify the deviation indicators and the degree of deviation, and formulate targeted management strategies in combination with the actual forest area, topography and environmental information, and clarify the scope and priority of implementation.

Images