46 results about "Tree canopy" patented technology

The invention discloses an ecological restoration area carbon sink increment real-time prediction method and system based on artificial intelligence, and the method comprises the following steps: multi-source data collection: obtaining a monthly vegetation coverage image through a satellite remote sensing platform, collecting the sensor data of soil temperature and humidity, air CO2 concentration and the like through a laid ground sensor network, and carrying out the real-time prediction of the carbon sink increment of an ecological restoration area; adopting an unmanned aerial vehicle laser radar to obtain vegetation canopy point cloud data according to a preset period, and collecting biomass actual measurement data of a restoration area in a historical preset age limit; preprocessing data, performing radiometric calibration, atmospheric correction and cutting splicing on a monthly vegetation coverage image acquired by a satellite remote sensing platform, and extracting a vegetation coverage and vegetation index time sequence; denoising, ground point separation and single tree canopy segmentation are carried out on vegetation canopy point cloud data acquired by the unmanned aerial vehicle laser radar, and the height and crown breadth of single plant vegetation are calculated. According to the invention, carbon sink increment prediction can be realized more accurately.

The invention discloses a forestry remote sensing image recognition method and system based on image recognition, and relates to the technical field of image processing, and the method comprises the steps: collecting and preprocessing a LiDAR point cloud and a multispectral optical image, and obtaining a three-dimensional forest point cloud; forest feature segmentation is carried out through spectral difference analysis, forest feature segmentation boundary points are redistributed based on spectral dissimilarity, a core spectral signature is calculated, and a single-tree point cloud entity set is output; and carrying out topological analysis on the single tree point cloud entity set, and outputting a vector result. According to the method, the three-dimensional forest point cloud is directly operated, each tree is identified as a single-tree point cloud entity containing complete space structure and spectrum information, the terrain influence is fundamentally eliminated, the three-dimensional structure parameters of the single tree can be accurately obtained, a two-stage strategy of geometric rough division and spectrum refinement is creatively provided, and the method is suitable for the three-dimensional tree point cloud entity of the single tree. The deep fusion perfectly combines the macroscopic accuracy of geometric segmentation and the microscopic identification of spectrum segmentation, and greatly improves the segmentation precision of the adhered tree crowns.

The invention discloses an urban low-altitude landing risk assessment method based on three-dimensional space grids, and belongs to the technical field of low-altitude flight risk assessment. The objective of the invention is to solve the problem of accurate assessment of the low-altitude falling risk in a complex city scene. The method comprises the steps of performing three-dimensional rasterization processing on a low-altitude airspace of an evaluation airspace to obtain voxels of the evaluation airspace, and obtaining a voxel set of the evaluation airspace; building shelter formed by population density grids or mobile signaling inversion population, sensitive facility vectors, tree canopy shielding and three-dimensional building models is quantified, values are assigned to voxels of an evaluation airspace, and a voxelization environment parameter database covering the whole domain is obtained; constructing an aircraft failure rate model; carrying out an aircraft falling trajectory uncertainty simulation test to obtain landing coverage probability density distribution and tail end speed of the aircraft; and carrying out voxel risk synthesis by considering the probability of collision with people, and carrying out risk grading on the obtained voxel risk, thereby completing urban low-altitude landing risk assessment based on the three-dimensional space grid.

The invention discloses an early-stage intelligent monitoring system for citrus reiculata Blanco canopy diseases and insect pests based on multispectral imaging, and belongs to the technical field of agricultural information. The system comprises a multispectral imaging module, a three-dimensional point cloud acquisition module, a data fusion module, a time sequence data analysis module, an intelligent identification module and a monitoring result output module. The method comprises the following steps: synchronously acquiring a multispectral image and laser radar point cloud data of a citrus reiculata tree canopy, and generating a point cloud model through spatial registration fusion; continuously recording model data of a plurality of time points, and extracting a time sequence feature vector; identifying disease and pest types and severity by using a deep learning model; and outputting a result to the user terminal. According to the invention, the problem that large-range and high-precision early monitoring of diseases and insect pests of citrus reiculata canopies is difficult to realize in the prior art is solved, early discovery, precise positioning and trend early warning of the diseases and insect pests are realized through air-space-ground integrated data fusion and intelligent analysis, and the intelligent level of orchard management and the disease and insect pest control efficiency are effectively improved.

The invention discloses a natural secondary forest tree species classification method based on multi-modal data fusion. The method comprises the following steps: S1, building a tree species space-spectrum feature extraction module; s2, building a tree canopy global context sensing fusion module; s3, building a tree perception detail optimization module; s4, building a tree species classification network based on multi-modal data fusion; s5, after the tree species classification network with multi-modal data fusion is built, corresponding hyperspectral and laser radar data sets are used for training and testing; according to the method, the problems of fuzzy crown boundary, difficulty in distinguishing tree species with similar structures and low classification precision caused by difficulty in considering long-distance global dependence modeling and fine-grained local detail reservation when an existing tree species classification technology is used for processing a complex forest stand environment are solved.

The embodiment of the invention relates to the technical field of wind field simulation, and discloses a tree canopy wind field assessment method based on wind environment simulation, comprising: acquiring a three-dimensional data model of a to-be-assessed target environment; hydrodynamic simulation is executed to simulate wind field distribution when wind flows through the three-dimensional data model, and the tree model of the at least one tree is configured to be a porous medium model in the wind field simulation; obtaining a result based on wind field analog simulation; and determining a risk assessment result for the corresponding tree according to the wind load response index. According to the wind field evaluation method in the embodiment of the invention, at least one wind load response index, such as wind speed, pressure, wind load and the like, related to the corresponding tree is calculated according to a wind field simulation result. The quantified indexes can visually reflect the wind action condition of the tree in the wind field, the mechanical response of the tree in different wind environments can be deeply known, and a specific numerical basis is provided for evaluating the wind resistance of the tree.

The invention belongs to the technical field of fruit tree cultivation, and particularly discloses a durian micro-spraying humidity control cultivation method which comprises the following steps: selecting a durian tree planting field, and digging planting holes; a distributed micro-spray head array is arranged around each durian tree, and the micro-spray heads are vortex atomization type micro-spray heads; installing temperature regulation and control equipment in the planting land; installing a soil moisture content sensor in the soil, and arranging a canopy temperature and humidity sensor on the canopy of the durian tree; the bottoms of the planting holes are filled with decomposed organic fertilizer, and durian seedlings are planted; regulation and control are carried out according to requirements of durian on temperature and humidity in different growth stages; the durian tree is patrolled regularly, and when the signs of leaf tip burning and dry heat burning are found, cultivation intervention is conducted in combination with temperature and humidity regulation and control. According to the micro-spraying humidity-control cultivation method for the durians, through an accurate micro-spraying system and a humidity control strategy, a suitable humidity environment is created for growth of the durians, the growth speed of the durians is increased, and the occurrence probability of scorched tips of leaves, burning caused by dry heat and runt seedlings is reduced.

The invention discloses a FasterNet and BiFPN fused litchi tree canopy segmentation method, and relates to the technical field of deep learning and intelligent agriculture. A lightweight FasterNet module is used as a backbone network and is deeply integrated with a bidirectional feature pyramid network BiFPN, an efficient multi-scale feature extraction and fusion architecture is constructed, and the method is used for segmenting the canopy of a litchi tree. The FasterNet module adopts a combination strategy of partial channel deep convolution and point-by-point convolution, preferentially performs deep convolution operation on partial channels of an input feature map, captures spatial textures and edge details of litchi crown leaves, realizes cross-channel information interaction and dimension transformation through light-weight point-by-point convolution, and obtains the spatial texture and edge details of the litchi crown leaves. According to the design, the calculation load brought by dense convolution operation is remarkably reduced, so that the model can process high-resolution input of 5280 * 3956 pixels at a high speed, the BiFPN network aggregates multi-stage features which are extracted by FasterNet and come from different depths through a bidirectional cross-scale connection path, and collaborative optimization of calculation efficiency and multi-scale feature representation capability is realized.

This utility model discloses an agricultural remote sensing measurement device, including a control box, a solar charging device, and a tree measurement device. This agricultural remote sensing measurement device uses measuring branches to measure data such as leaf density, light radiation within the tree canopy, temperature, humidity, and stem flow. The controller stores the corresponding data in a memory, and periodically uploads the data in the memory to a remote control center via a 5G communication module. This facilitates researchers' study of tree transpiration, and the remote control center can view the current data at any time via the 5G communication module for convenient sampling. The use of tree branches facilitates the installation of the measuring branches and control box, ensuring installation stability for long-term monitoring. The solar charging device charges the battery, reducing the measurement device's dependence on external power and is energy-saving and environmentally friendly.

This invention discloses a method for assessing urban low-altitude fall risks based on a three-dimensional spatial grid, belonging to the field of low-altitude flight risk assessment technology. It aims to address the problem of accurately assessing low-altitude fall risks in complex urban scenarios. The invention includes: 1) 3) Rasterizing the low-altitude airspace of the assessment area into voxels, obtaining a voxel set; 2) quantifying population density grids or mobile signaling-derived population, sensitive facility vectors, tree canopy shading, and building shelters formed by three-dimensional building models, and assigning these values ​​to the voxels of the assessment airspace, resulting in a voxel-based environmental parameter database covering the entire area; 3) constructing an aircraft failure rate model; 4) conducting an uncertainties simulation experiment of the aircraft's fall trajectory to obtain the probability density distribution of the aircraft's landing coverage and terminal velocity; 5) considering the probability of collision with people, synthesizing voxel risks, and classifying the voxel risks to complete an urban low-altitude fall risk assessment based on a three-dimensional spatial grid.

The invention relates to the technical field of atmospheric mode and numerical simulation, in particular to a tree canopy drag wind speed attenuation parameterization method based on a turbulence physical model, which comprises the following steps: calculating a three-dimensional equivalent leaf area index by acquiring street geometric parameters and tree structure data, and decomposing the three-dimensional equivalent leaf area index into leaf area density changing along with height; further converting into equivalent windward area density, substituting the equivalent windward area density into a momentum equation, defining a canopy drag source item and calculating a canopy turbulence production item; and finally, coupling a canopy turbulence production item with a turbulence equation, and carrying out numerical solution by combining the speed and shear stress conditions of the ground, the crown edge and the roof to obtain a street height stratification wind speed profile. According to the three-dimensional parameterization method, the wind speed attenuation simulation capability of the tree streets is remarkably improved, the wind speed profile is more continuous and reasonable and highly coincides with the CFD simulation result, and the method is suitable for urban wind environment assessment and air pollution simulation.

1. Name of the product in this design: Assembled Toy (Tree Crown). 2. Purpose of this design: Assembled toys for entertainment and play. 3. The key design features of this product are the combination of shape and pattern. 4. The image or photograph that best illustrates the design's key points: a 3D model.

The present application relates to the technical field of atmospheric model and numerical simulation, and particularly relates to a tree crown drag wind speed attenuation parameterization method based on a turbulent flow physical model, which obtains street geometric parameters and tree structure data, calculates a three-dimensional equivalent leaf area index, and decomposes the three-dimensional equivalent leaf area index into a leaf area density varying with height; after being further converted into an equivalent windward area density, the equivalent windward area density is substituted into a momentum equation, a crown layer drag source term is defined and a crown layer turbulent production term is calculated; finally, the crown layer turbulent production term is coupled with a turbulent flow equation, and velocity and shear stress conditions of the ground, the tree crown edge and the roof are combined to obtain a street height layered wind speed profile by numerical solution. The present application significantly improves the wind speed attenuation simulation capability of a street with trees by a three-dimensional parameterization method, makes the wind speed profile more continuous and reasonable, and is highly consistent with a CFD simulation result, and is suitable for urban wind environment evaluation and air pollution simulation.

This utility model relates to the field of environmental monitoring equipment technology, and discloses an air quality monitoring device for tree maintenance, including a housing, an air monitoring unit, a fan, and an air duct. The housing is used to connect to the tree and has a cavity, an air inlet, and an air outlet. The air inlet and air outlet are both connected to the cavity. The air monitoring unit and the fan are both connected to the cavity. The fan is used to draw air from outside the housing into the cavity. The air duct is connected to the housing and is connected to the air inlet. The air duct extends to or beyond the tree canopy. When the fan operates, it causes the air from or beyond the canopy to form an airflow that enters the cavity through the air duct and the air inlet and exits through the air outlet. During this process, the air monitoring unit detects this airflow to obtain the composition information of the air in or beyond the canopy. In other words, the air quality monitoring device for tree maintenance can monitor the air around the tree canopy, providing data support for the scientific maintenance of trees.

This invention relates to the field of automated litchi harvesting technology, specifically a high-low staggered harvesting device and method for harvesting litchi from the entire tree canopy. The device includes a walking mechanism, a high-low staggered double robotic arm, a flexible clamping mechanism, a curved guide rail mechanism, a shearing device, a control device, and an identification device. The identification device and shearing device are mounted on the robotic arms. The low-position 6-DOF robotic arm is mounted on the slide of the curved guide rail mechanism. The high-position 3-DOF robotic arm consists of three linear motion modules along the X, Y, and Z axes. The flexible clamping mechanism is placed on the X-axis linear motion mechanism, which is placed on the Y-axis linear motion mechanism, and the Y-axis linear motion mechanism is placed on the Z-axis linear motion mechanism. The Z-axis linear motion mechanism, control device, and curved guide rail mechanism are mounted on the walking mechanism. This invention can cover the entire litchi canopy for fruit harvesting, is suitable for harvesting in high-branch scenarios, and significantly improves the efficiency of litchi harvesting.

This application relates to a method and system for detecting the diameter at breast height (DBH) of individual trees in forests based on the organic fusion of multi-source sensors, belonging to the field of forestry remote sensing technology. It performs terrain filtering, orthorectification, and independent tree instance segmentation on airborne lidar point cloud data and multispectral satellite imagery, respectively. Then, through local semantic topology graph construction and cross-modal graph matching, object-level identity binding is achieved to obtain a set of homologous binding pairs for individual trees. Subsequently, a Gaussian weight field for the tree canopy is constructed using the 3D point cloud spatial density distribution of the bound trees. A pixel-by-pixel probabilistic soft-assignment weighting is performed on the multispectral image object blocks to generate a soft-fusion feature vector for individual trees. Finally, this feature vector is input into a regression model constrained by the forest allometric growth law for DBH detection. This application effectively suppresses edge noise pollution caused by spatiotemporal misalignment of multi-source data through object-level flexible anchoring and probabilistic feature fusion, improving the accuracy and robustness of DBH detection in complex forest areas.

The invention relates to the field of intelligent agriculture, and particularly discloses a macadamia nut disease detection and recognition method based on image recognition, which comprises the following steps: S100: image data acquisition: acquiring high-definition digital images of leaves and fruits at the middle and upper parts of a macadamia nut tree canopy from multiple angles under a natural light condition, the method comprises the following steps: collecting an object covering healthy leaves and disease samples which are confirmed as anthracnose, leaf blight and stem canker on site by agricultural experts, and accurately marking each collected image to generate a data set containing image file names and corresponding disease category labels, wherein the disease samples comprise different disease severity degrees, leaf growth stages and background environments; s200, performing image input processing, and taking the data set obtained in S100 as input data; s300, image processing is carried out; s400, obtaining a final recognition result; according to the technical scheme, disease features can be learned in a self-adaptive mode, complex environment interference is resisted, and recognition precision and efficiency are both considered.

The invention discloses a stay wire type precise pesticide spraying unmanned aerial vehicle, and belongs to the technical field of unmanned aerial vehicles, the stay wire type precise pesticide spraying unmanned aerial vehicle comprises an unmanned aerial vehicle body, the unmanned aerial vehicle body is provided with a pesticide box and an airborne intelligent camera, the lower end of the unmanned aerial vehicle body is provided with at least one profiling pesticide spraying arm, and the profiling pesticide spraying arms can conduct profiling according to tree crowns; sprayers and position sensors are distributed on the profiling pesticide spraying arm, an outlet of the pesticide box is communicated with the sprayers through pipelines, a transmission box, a sensor control system and a pesticide spraying arm control system are arranged at the position, corresponding to the profiling pesticide spraying arm, of the upper end of the unmanned aerial vehicle body, the transmission box is connected with the profiling pesticide spraying arm, and the position sensors are electrically connected with the sensor control system. The sensor control system is electrically connected with the pesticide spraying arm control system, and the pesticide spraying arm control system is electrically connected with the transmission case. Through profiling of the profiling pesticide spraying arm, the position of the spray head on the profiling pesticide spraying arm can be controlled to be matched with the shape of a fruit tree crown, the comprehensiveness and uniformity of the spraying coverage range are ensured, and precise pesticide spraying is achieved.

This disclosure relates to a firebreak clearing device and a control method for the firebreak clearing device. The firebreak clearing device includes: an off-road chassis assembly (1) having wheels (11) and a non-load-bearing frame (12); a boom assembly (2) connected to the frame (12) and capable of telescoping and rotating relative to the frame (12); a preparation assembly (3) connected to the frame (12) and configured to prepare a flame retardant; a spraying assembly (4) disposed at the end of the boom assembly (2) and connected to the preparation assembly (3) via piping, and configured to spray the flame retardant; and a processor (5) configured to cause the spraying assembly (4) to spray the flame retardant onto the tree canopy during the movement of the off-road chassis assembly (1) to form a firebreak. The firebreak clearing device sprays while driving, meeting the requirements of rapid response in fire prevention and effectively preventing the spread of fire.

This invention relates to the field of ecological and environmental protection technology, and in particular to a method for predicting forest carbon sink change trends based on multi-source heterogeneous data analysis. The method includes: acquiring multi-source remote sensing images and completing geometric correction and registration to generate a spatially aligned raster set; extracting canopy boundaries to form a forest mask; extracting three-dimensional structural parameters of the tree canopy using LiDAR point cloud data; identifying tree species and ages by combining color histograms, texture features, and multispectral features; calculating static carbon storage and generating a carbon density map based on canopy area and carbon density coefficient; identifying transpiration anomalies and calculating leaf water potential by combining thermal infrared images to predict carbon absorption rates; and superimposing the carbon increment onto the carbon density map to form a time series, identifying carbon sink hotspots, transition zones, and carbon-depleted zones through difference and clustering. This invention, by integrating multi-source remote sensing data and efficient algorithms, achieves real-time monitoring, dynamic prediction, and refined management of forest carbon sinks, providing accurate data support for carbon markets, ecological compensation, and carbon neutrality decision-making.

The invention discloses a global self-adaptive parallel mechanical arm apple picking and follow-up collecting device and method.The device comprises a central controller, a longitudinal multi-speed lifting module, a transverse telescopic module, a parallel mechanical arm and a visual control system, and the longitudinal multi-speed lifting module, the transverse telescopic module and the parallel mechanical arm are electrically connected with the central controller; the visual control system is used for obtaining three-dimensional coordinates and depth information of fruits, and the central controller controls the longitudinal speed-multiplying lifting module to drive the picking operation platform to move to a plurality of height stations in sequence. Meanwhile, after the transverse telescopic module is controlled to transversely move the picking operation platform according to three-dimensional coordinates and depth information obtained by the visual control system at each height station, the central controller controls the parallel mechanical arm to pick the fruits, and the picked fruits are collected in a follow-up mode through the follow-up fruit collecting module. According to the device, full-area covering picking of fruits at different heights and different depth positions of tree crowns is achieved, and the environmental suitability of the device is greatly improved.

This invention discloses a forest plant protection drone, relating to the field of drone technology. It includes a drone, a spraying assembly, at least one delivery device, and at least one pesticide pack. The spraying assembly includes at least one nozzle, each connected to the drone's landing gear, and each nozzle is used to spray pesticides onto the forest. Delivery devices are mounted on the drone's fuselage and can apply force to the pesticide pack to attach it to the fuselage. Each delivery device can also release the force on the pesticide pack, allowing it to fall under gravity. Each pesticide pack can burst open at the center of the tree canopy or the center of the tree trunk. This invention improves the uniformity and comprehensiveness of pesticide application, reduces the overall structural weight of the drone, and helps increase the drone's pesticide payload.

The invention relates to the technical field of insect sampling devices, in particular to an orchard canopy insect three-dimensional sampling device and method based on insect taxis guidance, and the orchard canopy insect three-dimensional sampling device comprises a support adjusting system which comprises a central main rod and a height adjusting mechanism; the intercepting and guiding net body comprises an intercepting net and a guiding net, the intercepting net is provided with a dark-color net surface and is used for forming visual impairment and intercepting insects, the guiding net is provided with a light-color net surface and is upwards inclined and gathered at the top end, and the upper edge of the intercepting net is connected to the inner side of the guiding net; the collecting system is arranged at the top end of the guide net and used for allowing insects to enter and collecting the insects. Through the height-adjustable structural design, the application limitation of the fixed height of a traditional Markov net is broken through, and the capturing capacity of the orchard tree canopy flying insects is remarkably enhanced; the system is used in cooperation with a ground Markov net, a three-dimensional sampling system can be constructed, and the integrity and monitoring precision of orchard arthropod community investigation are effectively improved.

This invention discloses a collaborative inversion method for multiple attributes of individual trees based on transfer learning, relating to the field of forest ecological remote sensing monitoring technology. The method acquires remote sensing images and training samples, generates a canopy height model based on LiDAR point clouds, constructs and trains a single-tree segmentation model based on the U-Net++ architecture, and fine-tunes the encoder weights of the single-tree segmentation model by transferring them to a tree height prediction model and a forest type classification model. Finally, the remote sensing images are input into the three models to obtain the single-tree canopy boundary, predicted tree height, and forest type, and the integrated output is the multi-attribute inversion result for individual trees. This method addresses the problems of complex data dependencies, non-reusable features, low inversion accuracy and efficiency, and easy error accumulation inherent in traditional independent inversion of single-tree attributes.

This invention discloses a tapping harvesting method for various terrain environments and types of dried fruits. A lifting device is mounted on a walking platform; both the tapping device and the amplitude-limiting device are connected to the lifting device; the tapping bar in the tapping device passes through the gap between the two limiting rods in the amplitude-limiting device; the walking platform has multi-degree-of-freedom mechanical legs for moving the entire mechanical device and positioning the tapping device appropriately; the lifting device adjusts the height of the tapping device and the amplitude-limiting device to accommodate trees of different heights; the tapping device taps the tree canopy to cause the fruits to fall and controls the tapping force. This tapping harvesting method can adapt to various harvesting environments (sand, steep slopes, etc.), and can adjust the tapping height and amplitude for trees of different heights or canopy sizes (i.e., multiple types of dried fruits). It achieves a high harvesting rate, reduces damage to the tree, and overcomes the risk of falls associated with manual harvesting.

This application discloses a method and system for auditing the cost of clearing transmission line corridors based on UAV point clouds. The method includes: acquiring three-dimensional laser point cloud data of the transmission line corridor before and after the clearing operation, obtained by a lidar on a UAV; performing ground point separation on the three-dimensional laser point cloud data before and after the clearing operation, respectively; constructing a tree canopy height model based on the three-dimensional laser point cloud data and the separated ground points; performing individual tree segmentation based on the tree canopy height model; identifying the trees to be cleared based on the individual tree segmentation results before and after the clearing operation, and generating a list of point cloud analysis results corresponding to the cleared trees; and auditing the submitted materials for the transmission line corridor clearing based on the point cloud analysis result list using a large model to obtain the audit results for the transmission line corridor clearing.

The invention belongs to the technical field of statistical methods, and particularly relates to a method for regulating and controlling a tree canopy structure. According to the method, a nonlinear approximate ircorrelation equation of the tree height, the branch height and the effective crown height of the man-made forest and environmental factors of the average temperature in spring, the moisture loss in autumn and the moisture loss in spring is constructed, and the equation quantifies an interaction mechanism of the environmental factors and the relative sectional area on the dynamic change of the vertical structure of the crown; therefore, the canopies can be accurately managed.

The invention provides an urban tree species classification and mapping method based on LiDAR and hyperspectral fusion. The method comprises the steps of multi-source data acquisition and collaborative preprocessing; vegetation area extraction and non-tree ground object masking; individual tree crown segmentation and contour optimization are carried out; performing multi-dimensional feature extraction and feature engineering; feature selection and classification model construction; multi-category tree species identification and classification result generation; carrying out precision verification and uncertainty analysis; the advantages of LiDAR data in the aspect of representing a canopy three-dimensional structure and the characteristics of hyperspectral data in the aspect of detecting vegetation biochemical parameters are fully utilized, information complementation is achieved through feature level fusion, and the classification precision is remarkably improved.

The invention provides a litchi tree canopy pest and disease damage intelligent identification method, which comprises the steps of obtaining a litchi tree canopy high-resolution image, repairing an information missing region in the image by adopting an image repairing algorithm, and taking the repaired image as an input image for subsequent pest and disease damage identification; for the underexposure problem existing in the input image, performing illumination compensation on the underexposure image by adaptively adjusting the brightness and the contrast ratio of the input image to obtain an image after illumination compensation; aiming at the de-noised image, through an attention mechanism in the litchi tree canopy disease and insect pest identification model, automatically learning distinguishing characteristics of different disease and insect pest symptoms, improving identification precision of similar disease and insect pests, and obtaining a disease and insect pest identification result; the method comprises the following steps: acquiring multi-view images shot at different angles of a litchi tree canopy, and carrying out registration and fusion processing on the multi-view images to obtain a fusion image containing more comprehensive information.

The invention provides an anti-interference processing method and system for a tree below an overhead line, and belongs to the technical field of power transmission lines, and the method comprises the steps: arranging a sensing unit close to a tree canopy on the overhead line, and collecting the related data of a target branch; according to the data, spraying a load growth inhibitor through the unmanned aerial vehicle; a clamp is attached to the surface of the branch, the direction of the clamp is adjusted to enable the branch to be stressed, and the clamp is connected with the fixing pile through a rope; monitoring the branch acting force according to a stress sensor arranged in the clamp; a memory alloy braid layer is arranged on the overhead line, and the braid layer generates temperature change so as to generate elastic shrinkage; after the inhibitor in the step S2 is dried, the inhibitor spraying area and the clamp covering area are coated to form a biocompatible protective film. According to the anti-interference processing method for the tree below the overhead line, the crown shape and the greening ecology of the tree are protected while the growth of the target risky branches is precisely controlled, and environmental pollution and vegetation damage are reduced through precise spraying of the load type inhibitor.