39 results about "Forest structure" patented technology

The invention relates to the technical field of remote sensing, and discloses a forest land health state analysis method and system based on multi-source remote sensing image analysis. The system comprises a multi-source remote sensing acquisition module, a feature extraction and fusion module, a health assessment module, a traceability analysis module, a strategy matching module, a visual reconstruction module, an early warning decision module and an execution feedback module, and constructs a multi-modal forest land observation data set by fusing multi-source remote sensing data of multispectrum, hyperspectrum, radar and thermal infrared. The comprehensive extraction of multi-dimensional information such as vegetation coverage, canopy biochemical characteristics, under-forest structures and surface thermal environments is realized, the limitation of single data source analysis is overcome, and the comprehensiveness and accuracy of forest land health condition evaluation are improved; through dynamic comparison of multi-stage remote sensing images and health risk level mapping, early identification and early warning of forest growth abnormity, degeneration trend and pest and disease risk are realized.

The invention discloses a forest canopy height remote sensing estimation method based on multi-forest feature fusion and a bidirectional stacking model, and the method comprises the steps: obtaining multi-source remote sensing data of a target region, and employing a multi-stage feature selection method MBF-Control to screen an optimal feature subset for spatial distribution data of different forest types, the multi-source remote sensing data comprises GEDI data, Sentinel-1 / 2 data, Landsat-8 data and DEM (Digital Elevation Model) data; a bidirectional stacking model BS-MFTF of multi-forest type feature fusion is constructed; and inputting the optimal feature subset into the multi-forest feature fused bidirectional stacking model BS-MFTF, and outputting a canopy height estimated value of the target area. Structural diversity and spatial patterns are captured, and canopy height estimation and forest structure mapping are improved.

The invention discloses an automatic forest checking method integrating point cloud precise segmentation and parameter inversion, and relates to the field of forest resource checking, and the method comprises the steps: constructing an automatic 3D forest checking frame with a double-branch point cloud segmentation network as a core, carrying out the preprocessing of forest point cloud data, extracting multi-scale features, and carrying out the segmentation of a point cloud segmentation network; performing mask scoring processing on the multi-scale features, and finally outputting a mask score, a semantic segmentation result and an instance segmentation result; dividing the point cloud into three types of semantic tags of ground, branches and leaves according to the semantic segmentation result, dividing the instance segmentation result into independent individual tree instances, performing convex hull or voxelization processing on the accurately segmented point cloud, inverting forest structure parameters, and completing automatic checking of forest resources. The automatic 3D forest checking framework is excellent in performance, efficiently and accurately draws the boundary of a single tree in a point cloud, realizes high-precision identification and separation of branches and leaves on a sample plot and a tree level, and promotes digitization, automation and intelligentization of forestry resource management.

The invention discloses a forest tree DBH (diameter at breast height) inversion method based on a multispectral remote sensing image, which relates to the technical field of remote sensing information inversion, and comprises the following steps of: performing radiometric calibration and geometric correction processing on original multispectral observation data, generating a standardized multispectral data set, and performing single tree identification and canopy segmentation processing to obtain a tree DBH (diameter at breast height); generating single-plant tree crown image data; combining the illumination direction features with the red-edge wave band and the near-infrared wave band, performing multi-scale feature analysis and fusion processing, and outputting canopy feature data; performing nonlinear processing and illumination correction on the canopy feature data to generate breast diameter prediction data; and carrying out space mapping and visual mapping processing on the diameter at breast height prediction data to generate a diameter at breast height distribution diagram result, carrying out precision evaluation and quality inspection processing, and outputting a forest tree diameter at breast height inversion result. The stability, the precision and the environmental adaptability of the breast diameter inversion are improved, and high-reliability technical support is provided for forest structure monitoring and resource evaluation.

The invention provides a forest structure change detection method, a related device and a computer storage medium, and the method comprises the steps: firstly generating target point cloud data based on original laser radar point cloud and multi-source fusion positioning data of a target forest region; determining the current canopy entropy of the target point cloud data based on the spatial distribution of the target point cloud data; and finally, based on the current canopy entropy and the reference canopy entropy, generating a change detection result of the forest structure. And the purpose of rapidly and accurately detecting the forest structure change is achieved.

The invention relates to a space laser radar forest structure data correction method and system, and belongs to the technical field of space laser remote sensing and forest resource monitoring. Aiming at four kinds of errors of geometric positioning offset, slow distortion, terrain broadening effect and canopy structure influence of GEDI data, UAV control points are arranged by adopting a hierarchical arrangement strategy, a homonymy point relation is established through waveform registration, and weights are calculated; performing overall geometric correction based on the regular grid control points; carrying out local distortion correction based on the encryption control point; establishing a physical compensation model to compensate the terrain broadening effect; based on the canopy structure parameters and GEDI signal quality indexes, establishing a statistical regression model to correct a canopy height residual error; and finally, obtaining a high-precision canopy height convergence result through spatial smoothing processing, and outputting corrected data. The method achieves step-by-step precision improvement from the whole domain to the local domain and from the geometry to the height, and is low in cost, high in precision and wide in application range.

The application discloses a forest structure and intelligent coupling analysis and modeling method for absorbing photosynthetically active radiation, and belongs to the technical field of forest ecology remote sensing, structure analysis and computer simulation. The method focuses on fusing airborne and handheld laser radar information, accurately inverting forest structure indexes, covering single-tree level attributes, neighborhood spatial structure indexes, gap structure, stand density SD and slope SL; three-dimensional radiation transmission simulation is carried out; through the construction of a single-tree three-dimensional model library and a digital sample plot, high-precision three-dimensional forest reconstruction and hierarchical APAR simulation are realized, and photosynthetic spatial heterogeneity parameters are extracted; Spearman correlation analysis is carried out based on the structure and APAR parameters, and PCA dimension reduction and KMeans clustering are combined to identify the dominant structure mode and spatial type; based on the clustering result, a typed structure-APAR coupling integrated learning model is constructed, and the regulation mechanism of single-tree attributes and neighborhood structure on APAR is revealed.

The invention discloses an intelligent coupling analysis and modeling method for a forest structure and absorbed photosynthetically active radiation, and belongs to the technical field of forest ecological remote sensing, structural analysis and computer simulation. Focuses on fusion of airborne and handheld laser radar information, forest structure indexes are accurately inverted, and individual tree level attributes, neighborhood space structure indexes, forest gap structures, forest stand density SD and gradient SL are covered; three-dimensional radiation transmission simulation is carried out, high-precision three-dimensional forest reconstruction and layered APAR simulation are realized by constructing a single-tree three-dimensional model library and a digital sample plot, and photosynthetic spatial heterogeneity parameters are extracted; spearman correlation analysis is carried out based on the structure and APAR parameters, and a dominant structure mode and a space type are identified in combination with PCA dimension reduction and KMeans clustering; and constructing a typing structure-APAR coupling ensemble learning model based on a clustering result, and revealing a regulation and control mechanism of individual tree attributes and a neighborhood structure on APAR.

The application provides a polarization SAR image simulation method for forests under complex terrain conditions, comprising the following steps: (1) setting parameters of a polarization SAR imaging process, including radar imaging parameters, forest parameters and terrain parameters; (2) generating a corresponding terrain model according to the terrain parameters; (3) constructing a forest model according to the forest parameters; (4) calculating a polarization SAR image according to the forest parameter model, and finally simulating the polarization SAR image of the forest under the complex terrain conditions. The method can improve the simulation of the polarization SAR image of the forest under the complex terrain conditions, can be used for understanding the remote sensing mechanism, testing new applications and designing the inversion algorithm of the forest structure, and can further explore the influence of the terrain on the polarization SAR image imaging process.

The application discloses a kind of visual Token compression method and device of multi-modal large language model based on space-time forest, comprising: each frame video frame is input into multi-modal large language model, after the visual encoder in multi-modal large language model, the visual Token embedding sequence corresponding to each frame video frame is obtained, from the visual Token embedding sequence corresponding to each frame video frame, several key visual Token embeddings are selected and node feature is constructed;Node feature is input into space-time forest algorithm, each key visual Token embedding is regarded as node, the directed edge between two nodes is constructed based on the joint constraint of semantic similarity, spatial distance and time sequence, node is constructed into space-time tree by directed edge and is organized into space-time forest structure, and space-time forest structure is processed under preset constraint, obtain final node set, the key visual Token embedding corresponding to the node in final node set is regarded as compressed visual Token embedding sequence.The application can reduce the amount of calculation of model.

This invention discloses a method for registering forest point clouds from UAV-LiDAR during leafy and leafless periods, relating to photogrammetry, remote sensing mapping, and forest resource monitoring. The method extracts ground points from two phases of UAV-LiDAR point clouds, constructs multi-directional radial topographic profiles, and extracts local profile segments using a sliding window. It calculates a comprehensive similarity based on trend consistency, robustness to local shape deformation, and curvature similarity, and selects stable matching segments by combining radial distance and topographic undulation amplitude consistency constraints. Three-dimensional corresponding point pairs are generated from the matching segments, and the initial rigid body transformation is solved using SVD. The registration is then optimized using ICP to obtain a multi-temporal fused point cloud. This invention can achieve reliable registration under conditions of phenological differences, missing vegetation structure, and initial biases, providing support for the reconstruction of three-dimensional forest structures and refined resource monitoring.

The application discloses a method for setting forest monitoring standard plots based on airborne laser radar data, comprising the following steps: S1: airborne laser radar surveying and mapping is carried out on the survey monitoring target forest stand; S2: surveying and mapping laser radar point cloud achievement data are acquired; S3: point cloud data are preprocessed; S4: a grid effective layout range is identified; S5: grid layout and effective grid screening are carried out; S6: forest structure parameter mean value extraction and high density index calculation are carried out; S7: effective grid layering is carried out; S8: HDI distribution is calculated layer by layer, and index concentrated distribution range grid is extracted; S9: typical grid screening is carried out; and S10: standard compliance verification and final determination are carried out. The method provided by the application can accurately extract forest stand factor information, and can provide standard plot settable plots according to standard plot setting principles, and can extract terrain factors of the plots, so that the typicality of the plots can be ensured, and the uncertainty brought by complex terrain to forest survey work can be avoided.

The present invention belongs to the technical field of river water body treatment facilities, and in particular relates to an underwater forest ecological restoration structure. The present invention makes the ecological restoration effect of the underwater forest more prominent, flexible, and comprehensive by arranging a lifting suspension box unit, a lifting frame unit, and a porous strip unit on the water-retaining wooden pile unit. The present invention has the following advantages: First, the water-retaining wooden pile unit is high on both sides and low in the middle, so that the cleaning boat can pass normally, ensuring that the overall sewage treatment effect of the river is more prominent and comprehensive; second, the underwater forest structure includes submerged plants, emergent plants, and microorganisms, so that many water purification effects such as reducing the nitrogen and phosphorus content in the water, increasing dissolved oxygen, absorbing heavy metals, and settling suspended pollutants can be achieved.

The invention relates to the technical field of electrical digital data processing, and discloses a form filling method based on a tree forest structure, which is characterized by comprising the following steps of: acquiring multi-source form data, converting the multi-source form data into semantic vectors by using a pre-training language model, and constructing a two-dimensional incidence matrix by combining co-occurrence frequency analysis and correlation strength calculation; dynamically constructing a tree forest containing conflict detection; according to the form type filled by the user, in combination with a caching strategy of a session mode or a historical mode, matching and processing conflicts according to meta-information vector priorities; the forest node relation is adjusted through a visual platform, and the association accuracy is verified in real time; and monitoring the modification rate to trigger incremental learning, and iterating the forest structure according to a service isolation strategy. The problems of time consumption, field conflict, repeated labor and poor user experience of form filling in the prior art are solved, and the purposes of intelligent multiplexing, efficient filling, conflict avoidance and experience improvement are achieved.

The application uses images to construct a forest sample plot measurement model. The three-dimensional reconstruction of images obtained by using a smart phone has the advantages of low measurement cost and simple actual operation suitable for field operation. The improved DBSCAN algorithm improves the accuracy of clustering, so that the constructed three-dimensional measurement model can be used for obtaining forest structure parameters, and the measurement accuracy of the structure parameters is much higher than the national forest second-class survey standard.

The invention discloses an ICESat-2 adaptive photon denoising method and an ICESat-2 adaptive photon denoising system based on vegetation layering. According to the method, a vegetation layering strategy is introduced, the photon data is divided into different gradient grades according to the vegetation coverage, and the problem that a traditional global denoising method is insufficient in adaptability in a complex vegetation environment is effectively solved. And the optimal denoising parameter of each vegetation level is adaptively determined in combination with the Bayesian optimization algorithm, so that the photon denoising precision and efficiency of the DBSCAN clustering in the vegetation region with uneven density are remarkably improved. According to the method, the advantages of multi-source remote sensing data are fully utilized, efficient screening and separation of signal photons and noise under different vegetation density conditions are achieved, and a higher-quality data basis is provided for subsequent forest structure parameter inversion and ground feature classification.

The application belongs to the technical field of ecological monitoring, and discloses a forest ULS-TLS point cloud automatic registration method based on ground context description; the application uses ground points as registration primitives, constructs ground context, and can directly register TLS single scanning data to ULS point cloud without pre-registration among different scanning point clouds of TLS; the application has high reliability: the extraction process of ground points is relatively standardized, and stable and reliable results can be usually obtained, and is not affected by the distribution, density and complexity of shapes of trees in the forest; the application improves the calculation efficiency: compared with the extraction of tree attributes, the use of ground points as registration primitives reduces the calculation cost and avoids high calculation overhead caused by the extraction of tree attributes; the application is dedicated to providing a forest ULS-TLS point cloud automatic registration method based on ground context description, so as to quickly, accurately and automatically realize the registration of forest ULS and TLS point clouds, comprehensively acquire forest structure information, and promote the application of laser radar technology in forest investigation.

The present invention proposes a non-continuous entity recognition method based on sequence to forest. In order to better characterize the dependencies between internal blocks of entities, a forest structure is used to model entity sets in texts. A neural network-based "encoder-decoder" generation framework is adopted to realize the sequence to forest generation process. On the encoder side, a Transformer-based encoder is first used to capture the global dependency features of each word in the text, and then a convolutional neural network is used to further capture the local dependency features of the words. On the decoder side, an attention module based on a "note" mechanism is designed to capture the semantic relevance between the output and input at each decoding moment. The beneficial effects of the present invention are: it can ensure the disorder between entities and the validity within the entity, it can effectively capture the relevance between entity blocks, it can effectively improve the model's recognition ability for non-continuous entities, and it can be adapted to the scenario of continuous entity recognition.

The invention discloses a carbon sink measurement method and system for a coastal protection forest, and the method comprises the following steps: obtaining field data of a target coastal protection forest, and carrying out the preprocessing of the field data, and obtaining a three-dimensional point cloud model of the protection forest; based on the three-dimensional point cloud model of the protection forest, adaptive forest structure parameter extraction is carried out, and a vegetation type and a corresponding tree height are obtained; the adaptive forest tree structure parameter extraction comprises parameter adaptive single tree segmentation and segmentation-fitting mutual feedback optimization; and performing carbon sink index calculation based on the vegetation type and the corresponding tree height to obtain the carbon density of the protection forest. According to the invention, by introducing a parameter adaptive initialization and segmentation-fitting mutual feedback optimization mechanism driven by a forest type feature vector, internal self-checking and iterative optimization are realized, and an automatic high-precision carbon sink measurement method adaptively coping with changes of arbors and shrubs in a coastal area is realized.

The application discloses a forest tree diameter at breast height inversion method based on multispectral remote sensing images, relates to the technical field of remote sensing information inversion, and comprises the following steps: performing radiation calibration and geometric correction processing on original multispectral observation data, generating a standardized multispectral data set, and performing single-tree recognition and crown layer segmentation processing to generate single-tree crown image data; combining the illumination direction feature with the red edge band and the near-infrared band through multiscale feature analysis and fusion processing to output crown layer feature data; performing nonlinear processing and illumination correction on the crown layer feature data to generate diameter at breast height prediction data; performing spatial mapping and visual mapping processing on the diameter at breast height prediction data to generate diameter at breast height distribution map results, and performing precision evaluation and quality inspection processing to output forest tree diameter at breast height inversion results. The stability, precision and environmental adaptability of diameter at breast height inversion are improved, and high-reliability technical support is provided for forest structure monitoring and resource evaluation.

The invention relates to a discontinuous microscopic image automatic splicing method based on multi-seed region growth and global layout, which comprises the following steps: acquiring a microscopic image sequence set to be spliced and preprocessing the microscopic image sequence set, processing the preprocessed image based on a multi-seed region growth method to form an associated forest structure, and carrying out local and global optimization on the associated forest structure, and generating and outputting a result. According to the scheme of the invention, an image set is intelligently divided into a plurality of connected sub-graphs (forest structures), local and global optimization is carried out on the connected sub-graphs, and global coordinate optimization of sub-graph nodes is carried out on adjacent nodes crossing the connected sub-graphs based on prior expected relative displacement. And carrying out splicing and coordinate optimization on the internal and global macroscopic layouts of each sub-graph, inhibiting error accumulation, and realizing high-precision splicing.

The invention relates to the technical field of forest remote sensing and forest structures, in particular to a forest window extraction and analysis method based on GF-2 images and object-oriented machine learning, which comprises the following steps: S1, determining a research area and collecting data; s2, data preprocessing; s3, constructing an object-oriented multi-level structure segmentation classification model; s4, forest window extraction based on an object-oriented multi-layer structure segmentation classification model; s5, exploring forest window and vegetation index response; and S6, estimating long-time-sequence forest window evolution. Through GF-2 image high-precision forest window extraction, feature statistics, spatial pattern analysis and forest window and vegetation index response modeling, the problems of low forest window identification efficiency, incomplete analysis, difficult dynamic monitoring and insufficient vegetation relationship quantification are solved, and support is provided for dynamic monitoring and scientific management of a forest ecosystem.

The invention discloses a randomly configured afforestation method and system for forest stand vertical fault restoration. The afforestation method comprises the following steps: constructing a random complementary planting structure unit; the forest windows are graded according to the area scale, and different numbers of random complementary planting structure units are supplemented in a random configuration mode; selecting nursery stocks to be subjected to complementary planting; standardizing seedling planting and management; reasonably selecting complementary planting tree species of the random complementary planting structure units; the planting and afforestation time is stipulated. According to the method, dumbbell-shaped and torch-shaped configuration units are constructed, the forest windows with different areas and scales are graded, and different numbers of random complementary planting structure units are supplemented by adopting a random configuration mode, so that the stability and ecological diversity of the forest are remarkably improved. By simulating a random structure of a natural forest and constructing a random structural body, a long-acting and stable microenvironment is provided for the forest, the overall health and stability of the forest stand are maintained, the natural succession of the forest structure is accelerated, and rapid formation of a multi-tree mixed forest is promoted.

The invention provides a forest structure parameter inversion model obtaining method, a forest structure parameter inversion model using method and related equipment, and is applied to the technical field of artificial intelligence. According to the method, a training set containing space lag variables is constructed based on actual measurement data of current environment factors and forest structure parameters of target area samples; a neighborhood sample is determined through spatial neighborhood matching, nonlinear mapping is performed by using a multi-layer perceptron (MLP) network structure based on a spatial Durbin model, a gradient descent optimization model is adopted, a spatial autocorrelation effect is effectively considered, spatial dependence among pixels is eliminated, and the precision of forest structure parameter inversion is effectively improved.

The application discloses a kind of structural features integrated radar forest height inversion random forest method, belong to the technical field of radar detection;With the basis feature of polarization interferometric synthetic aperture radar PolInSAR multi-polarization observation data, the influence of analyzing physical model expression on PolInSAR forest height inversion is analyzed, mining associated physical model parameter is introduced into training feature dataset, supplementary input feature, consider complex forest structure scene, utilize Fourier-Legendre polynomial to depict the influence of forest vertical structure, extract the polynomial coefficient representing the vertical structure feature of forest and integrate into the training feature set, further supplement input feature;With a small amount of LiDAR forest height information as supervision label, based on random forest learning method to establish regression model, realize forest height inversion.The application proposes a forest height inversion method based on random forest, which integrates physical model features and vertical structure features, to improve the accuracy and stability of forest height inversion.

The invention relates to the technical field of remote sensing image processing, in particular to a forest structure remote sensing parameter-based post-fire recovery process evaluation method, which specifically comprises the following steps of: S1, recognition and severity recognition of burned patches: setting a research area, selecting a burned area and a non-burned area in the research area, taking Landsat time sequence data as a source, and carrying out evaluation on the severity of the burned patches; after preprocessing, synthesizing a Landsat time sequence stack and calculating an NBR time sequence, then using a LandTrendr algorithm to obtain NBR values before and after occurrence of a forest disturbance pixel, and combining an RdNBR index to distinguish the severity degree of the burned plaque into a low grade, a medium grade and a high grade; s2, forest structure parameter remote sensing inversion; and S3, reconstructing the post-firing recovery process. According to the method, the dNBR is introduced to serve as a predictive variable, and the defect of an existing remote sensing inversion model in the aspect of fire disturbance intensity characterization is effectively overcome by enhancing the response capacity of the forest structure parameter remote sensing inversion model to the fire disturbance intensity.

The present application relates to the technical field of three-dimensional reconstruction, and proposes a forest structure reconstruction system based on three-dimensional laser scanning point cloud modeling. The system acquires forest point cloud data through multi-source laser scanning, and completes data preprocessing in combination with auxiliary perception information; on this basis, a hierarchical feature expression mode is constructed, a three-constraint joint mechanism and a multi-source feature balance mechanism are introduced, and the accurate separation of tree trunks, branches and canopies is realized; further, the cross-scale structure expression capability is enhanced through structure perception feature propagation; in the three-dimensional reconstruction stage, based on skeleton topology and parameter-driven modeling, in combination with complex area compensation and adaptive refinement strategy, high-fidelity reconstruction of forest structure is realized; finally, single-tree structure model and forest structure parameters are output. The present application improves the point cloud segmentation precision and reconstruction quality, and is suitable for forest resource investigation, carbon storage evaluation and ecological monitoring scenes.

The invention relates to the technical field of forest resource monitoring and management, and discloses a forest stand structure optimization and complementary planting method fusing ULS and BLS point clouds, which improves the acquisition precision of complex mountain forest structure information by fusing the ULS and BLS point clouds. The continuity and reliability of forest window identification are improved through the processes of threshold screening, edge detection, smoothing and the like; quantification of the forest stand structure quality is realized by constructing a comprehensive evaluation index composed of five types of space structure parameters; and complementary planting optimization is carried out by introducing a genetic algorithm, so that the globality and effectiveness of scheme search are enhanced. The method provides an integrated technical framework for forest window operation and space structure optimization of the mountain natural forest, and has good application potential.

The application discloses a forestry unmanned aerial vehicle point cloud and ground point cloud non-target automatic registration method and device. The method comprises the following steps: S1, extracting two kinds of data source breast height position point cloud slice data, respectively using DBSCAN clustering algorithm to extract single tree coordinate information; S2, determining the corresponding relationship of the two based on the starburst diagram composed of the single tree coordinates in the two kinds of data sources, and solving the initial transformation matrix; S3, using ICP algorithm to perform fine matching on the two kinds of data, and obtaining the final transformation matrix, S4, and then performing space coordinate system conversion on the ground laser scanning point cloud based on the unmanned aerial vehicle laser point cloud. In the case that complete forest structure information cannot be obtained from a single point cloud data source, the initial transformation matrix is obtained by extracting single tree coordinates and using the starburst diagram composed of the geometric relationship between the single trees, the transformation matrix is calculated by combining ICP fine registration, and finally the transformation from the ground point cloud coordinate system to the unmanned aerial vehicle point cloud coordinate system is realized.