A method and system for evaluating urban forest community carbon storage based on multi-source data

Abstract

The present invention is applicable to the technical field of forest assessment methods, and provides an urban forest community carbon storage assessment method and system based on multi-source data to obtain research area information; after the equipment is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud is obtained, a single tree data unit is generated according to the relationship model, and the single tree data unit is bound to a unique number; the contour features of the image area corresponding to the single tree data unit are extracted, and whether there are contradictions in the contour features, the abnormal contours are secondary peeled off to obtain the final single tree data unit, and species classification is performed based on the final single tree data unit; a structural weighting factor is constructed in combination with the area of ​​the sample plot, the corresponding level is determined according to the total carbon storage value of the entire sample plot, the theoretical carbon storage upper limit is calculated, and the relative deviation rate of the current carbon storage is calculated according to the theoretical carbon storage upper limit and the total carbon storage value of the sample plot. The method has the advantages of refined structural expression, flexible model matching, and the ability to judge and improve the assessment results.

Description

Technical Field

[0001] The present invention belongs to the technical field of forest assessment methods, and in particular relates to a method and system for assessing carbon reserves of urban forest communities based on multi-source data. Background Art

[0002] As urban ecosystems receive increasing attention, urban forests, as crucial green infrastructure, play a core role in carbon sequestration, microclimate regulation, and ecological buffering. To scientifically assess their carbon storage capacity, precise measurement of forest community carbon storage has become a core technical focus for urban forestry planning and ecological management.

[0003] In existing technologies, forest carbon storage assessment mainly relies on LiDAR point clouds to obtain vegetation structure data, and combines them with a unified biomass estimation model to convert carbon storage for trees in the sample plot. However, this type of method has multiple limitations: on the one hand, a single data source is prone to misjudgment or omission in areas with dense forest obstruction and interlaced structures, resulting in incomplete extraction of single tree structural parameters; on the other hand, fixed universal models ignore the differential effects of species, forest types, developmental stages, etc. on carbon storage composition, and cannot achieve a precise expression of "model selection based on trees". In addition, current assessment methods generally lack participatory analysis of spatial structure, cannot reflect the distribution contribution of different structural areas to the overall carbon storage, and are difficult to provide effective support for strategies to enhance carbon sequestration potential.

[0004] At the same time, most existing solutions remain at the level of "quantity calculation" and lack the ability to categorize and provide feedback on assessment results, making it impossible to form an "evaluation-diagnosis-intervention" closed-loop mechanism for practical governance and optimization. Therefore, it is urgent to develop a new urban forest carbon storage assessment method that integrates multi-source data collection, image-assisted recognition, dynamic model adaptation, structural zoning integration, and grade determination feedback to improve the accuracy, adaptability, and management value of the assessment, and provide a decision-making basis for improving the carbon sequestration capacity of urban ecosystems. Summary of the Invention

[0005] The purpose of the embodiment of the present invention is to provide a method for carbon storage assessment of urban forest communities based on multi-source data, aiming to solve the problem raised in the third part of the background technology.

[0006] The embodiment of the present invention is implemented as follows: a method for assessing carbon storage of urban forest communities based on multi-source data, the method comprising:

[0007] Acquire research area information, wherein the research area information includes the research area, research time, and research requirements, wherein the research time includes data collection time and plot setting duration;

[0008] Install the equipment according to the installation requirements. After the equipment is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud is obtained. According to the relationship model, a single tree data unit is generated and a unique number is bound to the single tree data unit.

[0009] Extract the contour features of the image area corresponding to the single tree data unit, identify whether there are contradictions in the contour features, perform secondary stripping on abnormal contours, obtain the final single tree data unit, and perform species classification based on the final single tree data unit;

[0010] The structural weighting factor is constructed based on the area of ​​the sample plot, the corresponding level is determined according to the total carbon storage value of the entire sample plot, the theoretical upper limit of carbon storage is calculated, and the relative deviation rate of current carbon storage is calculated based on the theoretical upper limit of carbon storage and the total carbon storage value of the sample plot.

[0011] Preferably, the steps of installing the device according to the installation requirements, obtaining the geometric relationship between the orientation in the panoramic image and the center direction of the point cloud after the device is installed, generating a single tree data unit according to the relationship model, and binding the single tree data unit to a unique number specifically include:

[0012] Obtain device information, obtain installation requirements, and install the device according to the installation requirements, including installation height and calibration plate;

[0013] After the device is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud obtained by laser radar scanning is obtained, and the image recognition confidence and acquisition integrity thresholds are obtained. The image recognition confidence and acquisition integrity of the panoramic image are evaluated according to the thresholds;

[0014] A projection envelope relationship model between the image and the point cloud is constructed based on the qualified geometric relationship, a single tree data unit is generated based on the relationship model, and a unique number is bound to the single tree data unit. The single tree data unit includes the diameter at breast height, tree height and crown width.

[0015] Preferably, the steps of extracting contour features of the image area corresponding to the single tree data unit, identifying whether there are contradictions in the contour features, performing secondary stripping on abnormal contours to obtain the final single tree data unit, and performing species classification based on the final single tree data unit specifically include:

[0016] Extract the contour features of the image area corresponding to the single tree data unit. The contour features include crown shape, texture symmetry, and edge continuity. Identify whether there is a contradiction between the contour features. The contradiction is the inconsistency between the image display and the point cloud clustering. If a contradiction is found, it is determined that the current clustering is abnormal.

[0017] Perform secondary stripping on the abnormal contours. The secondary stripping is guided by the main trunk direction or crown edge in the image, constructing an auxiliary cutting plane in the point cloud, and re-dividing the cluster blocks.

[0018] After the division is completed, the diameter at breast height, tree height, and crown width of each sub-block are recalculated to obtain the final single tree data unit. Species classification is performed based on the final single tree data unit, and specific tree species are identified through species classification.

[0019] Preferably, the steps of constructing a structural weighting factor based on the area of ​​the sample plot, determining the corresponding grade based on the total carbon storage value of the entire sample plot, calculating the theoretical upper limit of carbon storage, and calculating the relative deviation rate of current carbon storage based on the theoretical upper limit of carbon storage and the total carbon storage value of the sample plot specifically include:

[0020] Obtain the sample plot area and divide the sample plot into different levels of structural regions based on the final single tree data unit. The structural regions include structural homogeneous regions, structural transition regions, and structural heterogeneous regions. Calculate the average single tree carbon storage and total number of trees within the structural regions.

[0021] Combine the area of ​​the sample plot to construct a structural weighting factor, integrate the total carbon storage value of the entire sample plot based on the structural weighting factor, obtain the sample plot carbon storage grade evaluation rules, and determine the corresponding grade based on the total carbon storage value of the entire sample plot;

[0022] Calculate the theoretical carbon storage upper limit, calculate the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample site, obtain the deviation rate threshold, and if the relative deviation rate is greater than the deviation rate threshold, obtain a list of improvement suggestions and send the list to the terminal.

[0023] Preferably, the device information includes a lidar and a panoramic camera.

[0024] Another object of an embodiment of the present invention is to provide an urban forest community carbon storage assessment system based on multi-source data, the system comprising:

[0025] A research preparation module, which obtains research area information, including the research area, research time, and research requirements. The research time includes data collection time and plot setting duration;

[0026] The single tree data module installs the equipment according to the installation requirements. After the equipment is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud is obtained. The single tree data unit is generated according to the relationship model and the single tree data unit is bound to a unique number.

[0027] The secondary stripping module extracts the contour features of the image area corresponding to the single tree data unit, identifies whether there are any contradictions in the contour features, performs secondary stripping on abnormal contours, obtains the final single tree data unit, and performs species classification based on the final single tree data unit;

[0028] The grade determination module constructs a structural weighting factor based on the area of ​​the sample plot, determines the corresponding grade according to the total carbon storage value of the entire sample plot, calculates the theoretical upper limit of carbon storage, and calculates the relative deviation rate of current carbon storage based on the theoretical upper limit of carbon storage and the total carbon storage value of the sample plot.

[0029] Preferably, the single tree data module includes:

[0030] The equipment installation unit obtains equipment information and installation requirements, and installs the equipment according to the installation requirements, including the installation height and calibration plate;

[0031] An information acquisition unit, after the device is installed, obtains the geometric relationship between the position in the panoramic image and the center direction of the point cloud obtained by laser radar scanning, obtains the image recognition confidence and acquisition integrity thresholds, and evaluates the image recognition confidence and acquisition integrity of the panoramic image based on the thresholds;

[0032] A single tree binding unit constructs a projection envelope relationship model between the image and the point cloud based on the qualified geometric relationship, generates a single tree data unit based on the relationship model, and binds the single tree data unit to a unique number. The single tree data unit includes the diameter at breast height, tree height and crown width.

[0033] Preferably, the secondary stripping module includes:

[0034] The contour feature unit extracts the contour features of the image area corresponding to the single tree data unit. The contour features include crown shape, texture symmetry, and edge continuity. It identifies whether there is a contradiction in the contour features. The contradiction is the inconsistency between the image display and the point cloud clustering. If a contradiction is found, it is determined that the current clustering is abnormal.

[0035] A secondary stripping unit performs secondary stripping on abnormal contours. The secondary stripping is guided by the main trunk direction or crown edge in the image, constructs an auxiliary cutting plane in the point cloud, and re-divides the cluster blocks;

[0036] After the species classification unit is divided, the diameter at breast height, tree height, and crown width of each sub-block are recalculated to obtain the final single tree data unit. Species classification is performed based on the final single tree data unit, and specific tree species are identified through species classification.

[0037] Preferably, the level determination module includes:

[0038] Structural region division unit: obtain the sample plot area, and divide the sample plot into different levels of structural regions according to the final single tree data unit. The structural regions include structural homogeneous regions, structural transition regions, and structural heterogeneous regions. Calculate the average single tree carbon storage and total number of trees within the structural regions.

[0039] The grade determination unit constructs a structural weighting factor based on the area of ​​the sample plot, integrates the total carbon storage value of the entire sample plot based on the structural weighting factor, obtains the sample plot carbon storage grade evaluation rules, and determines the corresponding grade based on the total carbon storage value of the entire sample plot;

[0040] The deviation rate unit calculates the theoretical carbon storage upper limit, calculates the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample site, obtains the deviation rate threshold, and if the relative deviation rate is greater than the deviation rate threshold, obtains the improvement suggestion list and sends the list to the terminal.

[0041] Preferably, the device information includes a lidar and a panoramic camera.

[0042] An embodiment of the present invention provides a method for assessing carbon reserves in urban forest communities based on multi-source data. During the data acquisition phase, the method utilizes a coaxial mounting structure of a rotating handheld laser radar and a 360-degree panoramic camera, combined with a QR code calibration plate and differential GPS, to achieve real-time spatial synchronization of point cloud and image data. It also constructs a projection envelope relationship between the image and point cloud, providing a precise spatial reference for subsequent individual tree identification and species classification. During the individual tree segmentation phase, an image-assisted stripping mechanism is integrated, and based on the geometric consistency between image contour features and point cloud clustering results, a secondary segmentation is performed on abnormal clustering components. This method is particularly suitable for dense forest environments with complex structures such as shadow dislocation and interlaced trunks, significantly improving the accuracy and completeness of individual tree identification.

[0043] In the carbon storage calculation stage, based on the species labels, diameter at breast height data and forest type classification information of individual trees, the optimal biomass equation is dynamically matched, and species or genus level models are called first. In the sample site integration stage, a structure-dominated sample site zoning strategy is proposed. The sample sites are divided into homogeneous areas, transition areas and heterogeneous areas according to the diameter at breast height diversity, intermingling degree and angular scale. A structural weighting factor is constructed, and the carbon storage is weighted and integrated according to the area ratio of each region to ensure that the carbon storage assessment can reflect the actual contribution of structural heterogeneity to ecological functions. The relative deviation rate of the current carbon storage is calculated, and management thresholds are introduced for grading. When the deviation is significant, an improvement suggestion list is automatically generated and pushed to the terminal, realizing a closed-loop management path from identification to intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A flowchart of a method for assessing carbon storage in urban forest communities based on multi-source data provided by an embodiment of the present invention;

[0045] Figure 2 A flowchart of the steps of generating a single data unit according to a relational model and binding the single data unit to a unique number provided in an embodiment of the present invention;

[0046] Figure 3A flowchart of the steps of extracting contour features of an image region corresponding to a single tree data unit, performing secondary stripping of abnormal contours, and performing species classification based on the final single tree data unit, provided by an embodiment of the present invention;

[0047] Figure 4 A flowchart of the steps of determining the corresponding grade based on the total carbon storage value of all sample plots and calculating the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample plots provided in an embodiment of the present invention;

[0048] Figure 5 This is an architecture diagram of an urban forest community carbon storage assessment system based on multi-source data provided by an embodiment of the present invention;

[0049] Figure 6 This is an architectural diagram of a single data module provided by an embodiment of the present invention;

[0050] Figure 7 An architectural diagram of a secondary stripping module provided in an embodiment of the present invention;

[0051] Figure 8 This is an architectural diagram of the level determination module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0053] It is understood that the terms "first," "second," etc., used herein may be used to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script without departing from the scope of this application.

[0054] like Figure 1 As shown, an embodiment of the present invention provides a method for assessing carbon storage of urban forest communities based on multi-source data, the method comprising:

[0055] S100, obtaining research area information, wherein the research area information includes the research area, research time and research requirements, wherein the research time includes data collection time and plot setting duration.

[0056] In this step, information about the study area is obtained to ensure spatiotemporal consistency and ecological adaptability for subsequent data collection, plot layout, and model selection. This information includes not only the geographic scope and spatial dimensions, but also the data collection window and plot setup period, as well as the temporal dimensions, research focus, and functional dimensions, providing a complete set of background parameters for carbon storage modeling.

[0057] First, the administrative boundaries and ecological zoning of the research area need to be clear. For example, it can be set as "an urban forest area dominated by park-type green spaces within a main urban area." Geographically, it is necessary to obtain basic GIS data of the region, including vegetation cover maps, DEM elevation maps, road and building vector boundaries, etc., for subsequent sample plot planning and point cloud registration. At the same time, the research time needs to be refined into two categories: data collection time and sample plot setting duration. The former refers to the specific date of field sampling or sensor deployment, such as the completion of all sample point cloud and image data collection from July to September 2025; the latter refers to the time period during which the sample plot remains unchanged under the same structural standard, ensuring that subsequent stages can carry out time series comparison, re-measurement and update, or long-term carbon storage evolution modeling. For example, the sample plot setting maintenance time is 3 years, during which the sample plot number and location will not be changed.

[0058] Furthermore, research requirements must clearly define the data focus of the areas being collected, such as whether to emphasize structurally heterogeneous forestlands, or whether to focus on artificially regenerated forests with greater carbon storage growth potential or on disturbed edge areas. For example, if the research requirement is to "prioritize the carbon storage enhancement potential of fragmented forestlands on the urban fringe," the study area selection will tend to favor ecological transition zones such as border parks and residential green corridors, and the coverage of structurally heterogeneous areas will be increased in the sample design.

[0059] S200, installing the equipment according to the installation requirements, obtaining the geometric relationship between the orientation in the panoramic image and the center direction of the point cloud after the equipment is installed, generating a single tree data unit according to the relationship model, and binding the single tree data unit with a unique number.

[0060] In this step, the equipment was installed according to the installation requirements. An integrated support platform was constructed at the center of the plot, and the handheld LiDAR and 360° panoramic camera were coaxially mounted to ensure a unified directional reference for the collected data in physical space. High-contrast QR code calibration plates were placed at the four corners of the plot, combined with high-precision coordinate acquisition to establish a clear mapping between image pixels and point cloud coordinates. This allowed the construction of an image-point cloud projection envelope model, providing stable geometric support for subsequent image-assisted analysis.

[0061] Based on this registration model, the system constructs point cloud structural parameters and image recognition results into "single tree data units" and embeds unique numbers to enable data traceability across stages and modules. Each single tree data unit contains spatial location, geometric structure, species characteristics, and image outline information, enabling the pairing of structural data and semantic information during the collection process. This mechanism not only significantly improves data integrity and consistency, but also lays a standardized and traceable data foundation for subsequent single tree segmentation, biomass matching, and spatial zoning integration, with strong system integration and engineering feasibility.

[0062] S300, extracting the contour features of the image area corresponding to the single tree data unit, identifying whether there are contradictions in the contour features, performing secondary stripping on abnormal contours, obtaining the final single tree data unit, and performing species classification based on the final single tree data unit.

[0063] In this step, the outline features of the image region corresponding to the individual tree data units are extracted. Each clustered block in the point cloud data is back-projected onto the corresponding 360° panoramic image to extract the crown width outline of the area occupied by the tree in the image space. This process uses a geometric algorithm based on spatial projection transformations and incorporates the known reference coordinates provided by the calibration plate to ensure that the outline extraction results have a stable geometric correspondence.

[0064] The CNN contour recognition model is then used to extract features from each image contour region, obtaining its edge connectivity, shape regularity, and boundary sharpness. This feature is then compared for consistency with the spatial morphology of the crown width of the point cloud cluster. If multiple opposing crown widths, overlapping occlusions, or structural breaks are found in the image contour, but are classified as the same single tree in the point cloud, this is considered an abnormal cluster. This triggers a secondary peeling operation, repartitioning the point cloud based on the principal direction projection rule until the image and point cloud crown width features are highly consistent.

[0065] After finalizing the single-tree data unit structure, species classification is performed. This process combines the image feature vector with the trained model and is supplemented by label library matching.

[0066] S400: construct a structural weighting factor based on the area of ​​the sample plot, determine the corresponding level based on the total carbon storage value of the entire sample plot, calculate the theoretical upper limit of carbon storage, and calculate the relative deviation rate of current carbon storage based on the theoretical upper limit of carbon storage and the total carbon storage value of the sample plot.

[0067] In this step, the structural weighting factor is constructed based on the area of ​​the sample plot, and the sample plot is divided into several structural areas, such as structural homogeneous area, structural transition area and structural heterogeneous area, and the area proportion A of each area is calculated respectively. i , and combined with the structural complexity index of the area, the structural weight w can be set by comprehensively weighting the breast diameter diversity Hd, the mixed degree M and the angular scale W i, calculate the weighting factor that can reflect the structural complexity, and use it to correct the total carbon storage value, so that the assessment results have stronger structural adaptability and ecological responsiveness.

[0068] After obtaining the weighted total carbon storage value, a grading system is created based on the mean and standard deviation of the regional statistical sample, forming a horizontally comparable and vertically traceable carbon storage grading assessment system. Simultaneously, a theoretical carbon storage upper limit model is constructed, estimating the maximum carbon sequestration potential of the sample site under its current structure based on the optimal carbon storage density per unit area and a structural weighting factor. By comparing actual carbon storage with the theoretical upper limit and calculating the relative deviation rate, the gap between the site structure and carbon storage can be identified.

[0069] like Figure 2 As shown, as a preferred embodiment of the present invention, the steps of installing the device according to the installation requirements, obtaining the geometric relationship between the position in the panoramic image and the center direction of the point cloud after the device is installed, generating a single tree data unit according to the relationship model, and binding the single tree data unit to a unique number specifically include:

[0070] S201, obtaining device information, obtaining installation requirements, and installing the device according to the installation requirements, wherein the installation requirements include installation height and calibration plate.

[0071] In this step, obtain device information and key parameters for the selected acquisition equipment, including the functional specifications, interface standards, and mounting methods for the LiDAR model (such as the LiGrip V100), 360° panoramic camera model (such as the Insta360 Pro), and differential GPS device (such as the SR1 Pro). This information should include the LiDAR's scanning frequency, field of view, and point cloud density, as well as the panoramic camera's image resolution and synchronization mechanism. This ensures that subsequent installation procedures meet technical requirements. Furthermore, the device's power supply method and data synchronization interface standards must be obtained to provide the hardware foundation for integrated installation.

[0072] Before installation, obtain clear installation requirements and set the equipment installation height to a uniform 1.3 meters. The bracket structure should have height locking and anti-drift features, and a level should be used for installation correction, ensuring the equipment's scanning axis is perpendicular to the ground and the pitch angle is 0°. High-contrast QR code calibration plates should be placed at the four corners of the sample site. These plates should be made of high-strength anti-reflective material and placed at a height of approximately 1.2–1.5 meters.

[0073] Taking a 0.1-hectare urban plot as an example, a platform-type bracket was set up in the center of the plot, and the LiGrip V100 and Insta360 Pro were coaxially mounted. Calibration plates numbered A1 to A4 were placed at the four corners of the plot, ensuring a GPS data acquisition error of no more than ±5cm. After installation, each device entered a test run to confirm that the image and point cloud output were synchronized and fully covered before the formal data acquisition process began.

[0074] S202: After the device is installed, the geometric relationship between the orientation in the panoramic image and the center direction of the point cloud is obtained. The point cloud is obtained by laser radar scanning, and the image recognition confidence and acquisition integrity thresholds are obtained. The image recognition confidence and acquisition integrity of the panoramic image are evaluated according to the thresholds.

[0075] In this step, the geometric relationship between the orientation of the panoramic image and the center direction of the point cloud is determined after the equipment is installed. After the coaxial installation of the LiDAR and 360° panoramic camera is completed, a spatial calibration process is used to establish the geometric correspondence between the orientation of each pixel in the panoramic image and the center direction of the LiDAR point cloud. This correspondence is based on the spatial mapping of the calibration plate coordinates to the image pixels, and a projection transformation matrix is ​​constructed between the image coordinate system and the point cloud coordinate system.

[0076] A dual evaluation mechanism of image recognition confidence and image acquisition completeness is introduced. Image recognition confidence refers to the classification probability output given by the species recognition results of each single tree image area based on the deep learning model, expressed as P class The acquisition completeness refers to whether the pixel coverage of the crown structure in the image space is complete, whether there is occlusion, overexposure, underexposure, etc. that affect the extraction of structural information. It is usually evaluated comprehensively through edge connectivity, image entropy, and brightness histogram balance, and the threshold T is set. c and T i They are used to determine whether the recognition is reliable and whether the image is complete.

[0077] For example, if the species recognition model gives a confidence score of 0.92 for a single tree image, and its outline shows clear edges and a complete pixel distribution, both the confidence and completeness of the recognition are considered to be met. Conversely, if the confidence score is lower than 0.7, or if the image edges are broken due to lighting or the number of edge pixels falls below a preset threshold, the image is considered unreliable input and can be marked as "needing additional sampling." This will result in downgrading or additional sampling in subsequent carbon storage calculations and species matching.

[0078] S203: construct a projection envelope relationship model between the image and the point cloud based on the qualified geometric relationship, generate a single tree data unit based on the relationship model, and bind the single tree data unit to a unique number. The single tree data unit includes the diameter at breast height, tree height, and crown width.

[0079] In this step, a projective envelope relationship model between the image and point cloud is constructed based on the evaluated geometric relationships. This model describes the corresponding envelope region between the crown outline of an individual tree in image space and its 3D structure in point cloud space. This construction process relies on a geometric calibration matrix acquired coaxially. This process projects the 3D spatial information from the point cloud into image space and simultaneously reverse-projects the crown outline extracted from the image back into point cloud space, forming a spatially consistent closed region of the individual tree outline.

[0080] Based on this, the segmented point cloud clusters are analyzed for projection overlap with the image crown outline. Image-point cloud matches with an overlap greater than a set threshold (e.g., 85%) are identified as individual tree instances, and a single tree data unit is generated. This data unit includes three core structural parameters: diameter at breast height (D) extracted from the horizontal profile of the point cloud, tree height (H) derived from the vertical range of the point cloud, and crown width (W) calculated from the image outline and the horizontal projection of the point cloud, ensuring spatial integrity and recognition accuracy.

[0081] To ensure traceability and consistency in subsequent calculations, each individual tree data unit is automatically assigned a unique number. This numbering scheme uses a three-level structure: region number + plot number + tree number. For example, "XZ001-03-T12" represents the 12th tree in plot number 3 in Xicheng District. All numbers and corresponding structural parameters are automatically written to a structured database, enabling direct indexing and retrieval in subsequent steps, such as biomass model invocation, species classification, and spatial structure partitioning.

[0082] like Figure 3 As shown, as a preferred embodiment of the present invention, the steps of extracting contour features of the image area corresponding to the single tree data unit, identifying whether there are contradictions in the contour features, performing secondary stripping on abnormal contours, obtaining the final single tree data unit, and performing species classification based on the final single tree data unit specifically include:

[0083] S301, extracting the contour features of the image area corresponding to the single tree data unit, the contour features including crown shape, texture symmetry and edge continuity, and identifying whether there is a contradiction in the contour features, the contradiction being inconsistency between the image display and the point cloud clustering. If a contradiction is identified, it is determined that the current clustering is abnormal.

[0084] In this step, we extract the contour features of the image region corresponding to the individual tree data units. These contour features primarily include three aspects: crown morphology, texture symmetry, and edge continuity. These features are extracted using a deep convolutional neural network coupled with an edge detection operator and represented as vectors.

[0085] During the recognition process, the contour features extracted from the image are spatially aligned with the point cloud clustering results. If a cluster block appears intact in the point cloud, but its corresponding image region exhibits significant fragmentation, occlusion, or boundary expansion, or if a distinct single tree outline is present in the image but the point cloud fails to detect the corresponding structure, this indicates a discrepancy between the image display and the point cloud clustering. Such discrepancies indicate possible tree merging in the point cloud clustering, such as misidentifying two trees as one, or tree omission, such as occlusion leading to unrecognized trees.

[0086] For example, in a dense forest belt, a point cloud cluster may be identified as a single large tree in the LiDAR data. However, the image area shows two distinct crown boundaries with low texture continuity in between. This discrepancy suggests that two adjacent trees may have been mistakenly merged into one. In this case, the system automatically marks the cluster as "abnormal" and initiates a secondary segmentation process using image-assisted stripping to ensure the independence and accuracy of individual tree identification.

[0087] S302 , performing secondary stripping on the abnormal contours. The secondary stripping is guided by the main trunk direction or the crown edge in the image, constructing an auxiliary cutting plane in the point cloud, and re-dividing the cluster blocks.

[0088] In this step, the abnormal contour is stripped twice. First, the structural features of the abnormal cluster area in the image are analyzed, and its main trunk direction, that is, the longitudinal direction from the crown top to the base and the crown edge contour direction are extracted first.

[0089] Next, the extracted image principal direction information is back-projected into the point cloud space, where an auxiliary cutting plane perpendicular to the main trunk direction is constructed. This cutting plane is typically based on the boundary between the two crown sub-contours in the image or the minimum closed area of ​​the edge angle, determining its spatial position and normal direction. The auxiliary cutting operation re-divides the original cluster by intercepting the set of points in the point cloud cluster whose distance from the cutting plane is greater than a threshold, thereby separating the two mistakenly merged trees into independent units.

[0090] For example, if an unusual area in an image is identified as having two symmetrically arranged crowns with an angle of approximately 60° between the crown edge contours, back-projection can be used to locate two local density centers in the point cloud. A cutting plane is established with the perpendicular bisector connecting these two centers as its normal vector. By intercepting the point cloud around this section line, image-driven re-segmentation of the cluster block is completed, reorganizing the two actually independent trees into new single-tree data units and rebinding their numbers and structural parameters.

[0091] S303: After the division is completed, the diameter at breast height, tree height, and crown width of each sub-block are recalculated to obtain the final single tree data unit, and species classification is performed based on the final single tree data unit to identify the specific tree species through species classification.

[0092] In this step, after the division is completed, the structural parameters of each re-divided sub-block are independently calculated to generate the final valid single tree data unit. A horizontal section is extracted at a height of 1.3 meters above the ground. The circular or elliptical cross-section is fitted using the point cloud slice density profile to calculate the diameter. The Z-axis range is calculated from the highest point on the ground to the highest point of the crown. The sub-block is projected onto the horizontal plane, the maximum envelope boundary of the point cloud on the XY plane is extracted, and the average diameter of the major and minor axes is calculated.

[0093] After parameter calculation, the sub-blocks are combined with the contour, texture, and color features of the corresponding regions in the image to further identify species. This recognition is based on a pre-trained convolutional neural network model. It uses the image's texture directionality, color distribution, and leaf morphology to generate feature vectors, which are then compared for similarity with a database of standard species image vectors. The output category is the species name, such as Cinnamomum camphora, Ginkgo biloba, and Hackberry, along with a confidence score to assess the reliability of the recognition result.

[0094] For example, for two sub-plots obtained from a single stripping operation, the calculated DBH values ​​were 24.3cm and 17.5cm, respectively, with tree heights of 12.7m and 9.6m, and crown widths of 5.1m and 3.8m, respectively. Image recognition confirmed that the first sub-plot matched a camphor tree with a confidence score of 0.94, while the second sub-plot matched a Chinese elm tree with a confidence score of 0.89. Both exceeded the threshold and were confirmed as valid individual trees. Ultimately, these two independent data units were stored in the database and subsequently used by the carbon storage model.

[0095] like Figure 4 As shown, as a preferred embodiment of the present invention, the steps of constructing a structural weighting factor based on the area of ​​the sample plot, determining the corresponding level based on the total carbon storage value of the entire sample plot, calculating the theoretical carbon storage upper limit, and calculating the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample plot specifically include:

[0096] S401, obtaining a sample plot area, dividing the sample plot into different levels of structural areas according to the final single tree data unit, wherein the structural areas include structural homogeneous areas, structural transition areas, and structural heterogeneous areas, and calculating the average single tree carbon storage and the total number of trees in the structural areas.

[0097] In this step, the plot area is obtained and spatially divided based on the distribution of individual tree structural characteristics. First, the spatial boundaries of the current plot are obtained to determine its area, shape, and geographic coordinate distribution. Then, based on the structural parameters (DBH D, tree height H, crown width W) and spatial distribution coordinates of the individual tree data units, the entire plot is scanned for structural characteristics using a local statistical method (e.g., a 3×3m or 5×5m sliding window).

[0098] The plots were divided into three structural areas: structural homogeneous area: characterized by low H0, low M, and low W, with highly uniform tree species and sizes, commonly found in plantations or neat green belts; structural transition area: characterized by medium H0 and M, commonly found in partially regenerated or small-scale mixed areas; structural heterogeneous area: characterized by high H0, high M, and high W, with complex structure and mixed tree species, typical of natural secondary forests or long-term natural growth areas;

[0099] For each structural area, the total carbon storage and total number of trees in all individual tree data units are counted, and the average carbon storage of individual trees is calculated. Carbon storage calculation is based on the individual carbon storage value generated by the dynamic biomass matching mechanism (such as B j =a(D 2 H) b ×CF, where CF is the carbon factor).

[0100] The above H0 is the degree of diversity; M is the degree of intermixing, which is the degree of interlaced distribution of different tree species in a certain area; B j is the carbon storage of the jth tree; a and b are the parameters of the dynamic biomass model, corresponding to the regression parameters of a species, genus or universal carbon storage model, which need to be called from the regional model library;

[0101] For example, in a certain urban green space sample plot, the structurally heterogeneous area contained 35 trees with a total carbon storage of 6.3 tons, with an average carbon storage of 0.18 tons per tree. The structurally homogeneous area contained 46 trees with a total carbon storage of 5.2 tons, with an average of 0.11 tons. The structural transition area was somewhere in between. These data provided the basis for subsequent regional weighting and overall performance evaluation.

[0102] S402: construct a structural weighting factor based on the area of ​​the sample plot, integrate the total carbon storage value of the entire sample plot based on the structural weighting factor, obtain the sample plot carbon storage grade evaluation rule, and determine the corresponding grade based on the total carbon storage value of the entire sample plot.

[0103] In this step, the structural weighting factor is constructed based on the area of ​​the sample plot, and the actual area A of each type of structural area in the sample plot (such as structural homogeneous area, structural transition area, and structural heterogeneous area) is calculated. i and the total carbon storage in the region C i Based on the proportion of area to the total area of ​​the plot w i =A i / A total , construct the structural weighting factor w i , and use this to perform weighted integration of the carbon storage values ​​of each region:

[0104] ;

[0105] Among them, C totalThe weighted total carbon storage value of the plot reflects the different contributions of different structural regions to carbon storage capacity. After integrating the carbon storage value, it is necessary to introduce a carbon storage grade evaluation rule. This rule is based on historical plot statistical data of similar areas (such as urban green spaces, ecological isolation zones, secondary forests, etc.) and sets a theoretical carbon storage upper limit C. max , median C mid With bottom line C min . Change the current plot's C total Comparing with these benchmark values, carbon storage levels are classified, for example:

[0106] Level 1 (Excellent): C total ≥0.8×C max ;

[0107] Level 2 (good): 0.6×C max ≤C total <0.8×C max ;

[0108] Level 3 (medium): 0.4×C max ≤C total <0.6×C max ;

[0109] Level 4 (poor): C total <0.4×C max ;

[0110] For example, a 1-hectare urban plot consists of 40% structurally heterogeneous areas, 35% structurally transitional areas, and 25% structurally homogeneous areas. The total carbon storage in these three areas is 8.2 tonnes, 5.4 tonnes, and 3.1 tonnes, respectively. A weighted calculation yields a weighted total carbon storage of 6.18 tonnes. If the historical upper limit of carbon storage in this area is 8.5 tonnes, then 6.18 tonnes is approximately 73% of that, placing it at a "good" level.

[0111] S403, calculate the theoretical carbon storage upper limit, calculate the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample site, obtain the deviation rate threshold, and if the relative deviation rate is greater than the deviation rate threshold, obtain an improvement suggestion list and send the list to the terminal.

[0112] In this step, the theoretical carbon storage upper limit is calculated. Based on the structural area division results, the theoretical optimal single tree carbon storage is introduced into each type of area, and then multiplied by the total number of corresponding trees to obtain the theoretical carbon storage upper limit C. max For example, the theoretical carbon storage of a single tree in a structurally heterogeneous area is 0.28 t, and the actual number of trees is 40, so the upper limit of this area is 11.2 t.

[0113] Conduct deviation analysis on the carbon storage performance of the current plot and calculate its relative deviation rate:

[0114] ;

[0115] The deviation rate reflects the distance between the carbon storage of the sample site and its potential. In order to distinguish whether intervention is needed, a deviation rate threshold (such as 25%) is introduced as a management trigger standard. 偏离 >R 阈值 , the system determines that there is room for optimization in this area.

[0116] Based on this, reference plots with similar structures to the current plot but superior carbon storage performance are retrieved from the database. Their species composition, density, interplant ratio, age structure, and other characteristics are analyzed, and a list of improvement recommendations is automatically generated, such as increasing interplant ratios, introducing high-carbon storage dominant species like Hackberry and Pistacia, and optimizing plant spacing. This list is formatted and automatically pushed to forestry work terminals for managers to conduct feasibility assessments and make restoration decisions.

[0117] For example, the theoretical upper limit of a certain plot is 9.4t, but the actual value is 6.3t, with a deviation rate of 32.9%, which is higher than the set threshold of 25%. The system determines it as "medium-low adaptation" and pushes a suggestion: replant 30 camphor trees in the structural transition zone to increase the mixed planting index, with an estimated carbon storage gain of 0.9t.

[0118] like Figure 5 As shown, an urban forest community carbon storage assessment system based on multi-source data provided by an embodiment of the present invention includes:

[0119] The research preparation module 100 is used to obtain research area information, wherein the research area information includes the research area, research time and research requirements. The research time includes data collection time and plot setting duration.

[0120] In this system, the research preparation module 100 obtains information about the study area to ensure the temporal and spatial consistency and ecological adaptability of subsequent data collection, plot layout, and model selection. This information includes not only the geographic scope and spatial dimensions, but also the data collection time window and plot setup period, as well as the temporal dimension, research focus, and functional dimension, providing comprehensive background parameters for carbon storage modeling.

[0121] First, the administrative boundaries and ecological zoning of the research area need to be clear. For example, it can be set as "an urban forest area dominated by park-type green spaces within a main urban area." Geographically, it is necessary to obtain basic GIS data of the region, including vegetation cover maps, DEM elevation maps, road and building vector boundaries, etc., for subsequent sample plot planning and point cloud registration. At the same time, the research time needs to be refined into two categories: data collection time and sample plot setting duration. The former refers to the specific date of field sampling or sensor deployment, such as the completion of all sample point cloud and image data collection from July to September 2025; the latter refers to the time period during which the sample plot remains unchanged under the same structural standard, ensuring that subsequent stages can carry out time series comparison, re-measurement and update, or long-term carbon storage evolution modeling. For example, the sample plot setting maintenance time is 3 years, during which the sample plot number and location will not be changed.

[0122] Furthermore, research requirements must clearly define the data focus of the areas being collected, such as whether to emphasize structurally heterogeneous forestlands, or whether to focus on artificially regenerated forests with greater carbon storage growth potential or on disturbed edge areas. For example, if the research requirement is to "prioritize the carbon storage enhancement potential of fragmented forestlands on the urban fringe," the study area selection will tend to favor ecological transition zones such as border parks and residential green corridors, and the coverage of structurally heterogeneous areas will be increased in the sample design.

[0123] The single tree data module 200 is used to install the equipment according to the installation requirements, obtain the geometric relationship between the orientation in the panoramic image and the center direction of the point cloud after the equipment is installed, generate a single tree data unit according to the relationship model, and bind the single tree data unit to a unique number.

[0124] In this system, the single tree data module 200 is installed according to installation requirements. An integrated support platform is constructed at the center of the plot, and a handheld laser radar and a 360° panoramic camera are coaxially mounted, ensuring a unified directional reference for the collected data in physical space. High-contrast QR code calibration plates are placed at the four corners of the plot, combined with high-precision coordinate acquisition to establish a clear mapping relationship between image pixels and point cloud coordinates. This allows the construction of an image-point cloud projection envelope model, providing stable geometric support for subsequent image-assisted analysis.

[0125] Based on this registration model, the system constructs point cloud structural parameters and image recognition results into "single tree data units" and embeds unique numbers to enable data traceability across stages and modules. Each single tree data unit contains spatial location, geometric structure, species characteristics, and image outline information, enabling the pairing of structural data and semantic information during the collection process. This mechanism not only significantly improves data integrity and consistency, but also lays a standardized and traceable data foundation for subsequent single tree segmentation, biomass matching, and spatial zoning integration, with strong system integration and engineering feasibility.

[0126] The secondary stripping module 300 is used to extract the contour features of the image area corresponding to the single tree data unit, identify whether there are contradictions in the contour features, perform secondary stripping on abnormal contours, obtain the final single tree data unit, and perform species classification based on the final single tree data unit.

[0127] In this system, the secondary peeling module 300 extracts the contour features of the image region corresponding to the individual tree data units. Each clustered block in the point cloud data is back-projected onto the corresponding 360° panoramic image to extract the crown width contour of the area occupied by the clustered block in the image space. This process utilizes a geometric algorithm based on spatial projection transformations, combined with the known reference coordinates provided by a calibration plate, to ensure that the contour extraction results have a stable geometric correspondence.

[0128] The CNN contour recognition model is then used to extract features from each image contour region, obtaining its edge connectivity, shape regularity, and boundary sharpness. This feature is then compared for consistency with the spatial morphology of the crown width of the point cloud cluster. If multiple opposing crown widths, overlapping occlusions, or structural breaks are found in the image contour, but are classified as the same single tree in the point cloud, this is considered an abnormal cluster. This triggers a secondary peeling operation, repartitioning the point cloud based on the principal direction projection rule until the image and point cloud crown width features are highly consistent.

[0129] After finalizing the single-tree data unit structure, species classification is performed. This process combines the image feature vector with the trained model and is supplemented by label library matching.

[0130] The grade determination module 400 is used to construct a structural weighting factor based on the area of ​​the sample plot, determine the corresponding grade based on the total carbon storage value of the entire sample plot, calculate the theoretical carbon storage upper limit, and calculate the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample plot.

[0131] In this system, the grade determination module 400 constructs a structural weighting factor based on the area of ​​the sample plot, divides the sample plot into several structural areas, such as structural homogeneous area, structural transition area and structural heterogeneous area, and calculates the area proportion A of each area. i , and combined with the structural complexity index of the area, the structural weight w can be set by comprehensively weighting the breast diameter diversity Hd, the mixed degree M and the angular scale W i , calculate the weighting factor that can reflect the structural complexity, and use it to correct the total carbon storage value, so that the assessment results have stronger structural adaptability and ecological responsiveness.

[0132] After obtaining the weighted total carbon storage value, a grading system is created based on the mean and standard deviation of the regional statistical sample, forming a horizontally comparable and vertically traceable carbon storage grading assessment system. Simultaneously, a theoretical carbon storage upper limit model is constructed, estimating the maximum carbon sequestration potential of the sample site under its current structure based on the optimal carbon storage density per unit area and a structural weighting factor. By comparing actual carbon storage with the theoretical upper limit and calculating the relative deviation rate, the gap between the site structure and carbon storage can be identified.

[0133] like Figure 6 As shown, as a preferred embodiment of the present invention, the single tree data module 200 includes:

[0134] The equipment installation unit 201 is used to obtain equipment information, obtain installation requirements, and install the equipment according to the installation requirements. The installation requirements include installation height and calibration plate.

[0135] In this module, the device installation unit 201 obtains device information, including the core parameters of the selected acquisition equipment. This includes the functional specifications, interface standards, and mounting methods of the LiDAR model (e.g., LiGrip V100), 360° panoramic camera model (e.g., Insta360 Pro), and differential GPS device (e.g., SR1 Pro). This device information should include the LiDAR's scanning frequency, field of view, and point cloud density, as well as the panoramic camera's image resolution and synchronization mechanism, to ensure that subsequent installation operations meet technical requirements. Furthermore, the device's power supply method and data synchronization interface standards must be obtained to provide the hardware foundation for integrated installation.

[0136] Before installation, obtain clear installation requirements and set the equipment installation height to a uniform 1.3 meters. The bracket structure should have height locking and anti-drift features, and a level should be used for installation correction, ensuring the equipment's scanning axis is perpendicular to the ground and the pitch angle is 0°. High-contrast QR code calibration plates should be placed at the four corners of the sample site. These plates should be made of high-strength anti-reflective material and placed at a height of approximately 1.2–1.5 meters.

[0137] Taking a 0.1-hectare urban plot as an example, a platform-type bracket was set up in the center of the plot, and the LiGrip V100 and Insta360 Pro were coaxially mounted. Calibration plates numbered A1 to A4 were placed at the four corners of the plot, ensuring a GPS data acquisition error of no more than ±5cm. After installation, each device entered a test run to confirm that the image and point cloud output were synchronized and fully covered before the formal data acquisition process began.

[0138] The information acquisition unit 202 is used to obtain the geometric relationship between the position in the panoramic image and the center direction of the point cloud after the device is installed. The point cloud is obtained by lidar scanning, obtain the image recognition confidence and acquisition integrity thresholds, and evaluate the image recognition confidence and acquisition integrity of the panoramic image based on the thresholds.

[0139] In this module, the information acquisition unit 202 determines the geometric relationship between the orientation of the panoramic image and the center direction of the point cloud after the device is installed. After the coaxial installation of the LiDAR and the 360° panoramic camera is completed, a spatial calibration process is performed to establish a geometric correspondence between the orientation of each pixel in the panoramic image and the center direction of the LiDAR point cloud. This correspondence is based on the spatial mapping of the calibration plate coordinates to the image pixels, and a projection transformation matrix is ​​constructed between the image coordinate system and the point cloud coordinate system.

[0140] A dual evaluation mechanism of image recognition confidence and image acquisition completeness is introduced. Image recognition confidence refers to the classification probability output given by the species recognition results of each single tree image area based on the deep learning model, expressed as P class The acquisition completeness refers to whether the pixel coverage of the crown structure in the image space is complete, whether there is occlusion, overexposure, underexposure, etc. that affect the extraction of structural information. It is usually evaluated comprehensively through edge connectivity, image entropy, and brightness histogram balance, and the threshold T is set. c and T i They are used to determine whether the recognition is reliable and whether the image is complete.

[0141] For example, if the species recognition model gives a confidence score of 0.92 for a single tree image, and its outline shows clear edges and a complete pixel distribution, both the confidence and completeness of the recognition are considered to be met. Conversely, if the confidence score is lower than 0.7, or if the image edges are broken due to lighting or the number of edge pixels falls below a preset threshold, the image is considered unreliable input and can be marked as "needing additional sampling." This will result in downgrading or additional sampling in subsequent carbon storage calculations and species matching.

[0142] The single tree binding unit 203 is used to construct a projection envelope relationship model between the image and the point cloud based on the qualified geometric relationship, generate a single tree data unit based on the relationship model, and bind the single tree data unit to a unique number. The single tree data unit includes the diameter at breast height, tree height and crown width.

[0143] In this module, the tree binding unit 203 constructs a projected envelope relationship model between the image and point cloud based on the evaluated geometric relationships. This model describes the corresponding envelope region between a tree's crown profile in image space and its three-dimensional structure in point cloud space. This construction process relies on a coaxially acquired geometric calibration matrix, projecting the three-dimensional spatial information from the point cloud into image space while simultaneously inversely mapping the crown profile extracted from the image back into point cloud space, forming a spatially consistent closed region of the tree's profile.

[0144] Based on this, the segmented point cloud clusters are analyzed for projection overlap with the image crown outline. Image-point cloud matches with an overlap greater than a set threshold (e.g., 85%) are identified as individual tree instances, and a single tree data unit is generated. This data unit includes three core structural parameters: diameter at breast height (D) extracted from the horizontal profile of the point cloud, tree height (H) derived from the vertical range of the point cloud, and crown width (W) calculated from the image outline and the horizontal projection of the point cloud, ensuring spatial integrity and recognition accuracy.

[0145] To ensure traceability and consistency in subsequent calculations, each individual tree data unit is automatically assigned a unique number. This numbering scheme uses a three-level structure: region number + plot number + tree number. For example, "XZ001-03-T12" represents the 12th tree in plot number 3 in Xicheng District. All numbers and corresponding structural parameters are automatically written to a structured database, enabling direct indexing and retrieval in subsequent steps, such as biomass model invocation, species classification, and spatial structure partitioning.

[0146] like Figure 7 As shown, as a preferred embodiment of the present invention, the secondary stripping module 300 includes:

[0147] The contour feature unit 301 is used to extract the contour features of the image area corresponding to the single tree data unit. The contour features include crown shape, texture symmetry and edge continuity, and identify whether there is a contradiction in the contour features. The contradiction is the inconsistency between the image display and the point cloud clustering. If a contradiction is identified, it is determined that the current clustering is abnormal.

[0148] In this module, the contour feature unit 301 extracts the contour features of the image region corresponding to the individual tree data units. Contour features primarily include three aspects: crown morphology, texture symmetry, and edge continuity. These features are extracted using a deep convolutional neural network coupled with an edge detection operator and represented as vectors.

[0149] During the recognition process, the contour features extracted from the image are spatially aligned with the point cloud clustering results. If a cluster block appears intact in the point cloud, but its corresponding image region exhibits significant fragmentation, occlusion, or boundary expansion, or if a distinct single tree outline is present in the image but the point cloud fails to detect the corresponding structure, this indicates a discrepancy between the image display and the point cloud clustering. Such discrepancies indicate possible tree merging in the point cloud clustering, such as misidentifying two trees as one, or tree omission, such as occlusion leading to unrecognized trees.

[0150] For example, in a dense forest belt, a point cloud cluster may be identified as a single large tree in the LiDAR data. However, the image area shows two distinct crown boundaries with low texture continuity in between. This discrepancy suggests that two adjacent trees may have been mistakenly merged into one. In this case, the system automatically marks the cluster as "abnormal" and initiates a secondary segmentation process using image-assisted stripping to ensure the independence and accuracy of individual tree identification.

[0151] The secondary stripping unit 302 is used to perform secondary stripping on the abnormal contour. The secondary stripping is guided by the main trunk direction or the crown edge in the image, constructs an auxiliary cutting plane in the point cloud, and re-divides the cluster blocks.

[0152] In this module, the secondary stripping unit 302 performs secondary stripping on the abnormal contour. First, the structural features of the abnormal cluster area in the image are analyzed, and its main trunk direction, that is, the longitudinal direction from the crown top to the base and the crown edge contour direction are preferentially extracted.

[0153] Next, the extracted image principal direction information is back-projected into the point cloud space, where an auxiliary cutting plane perpendicular to the main trunk direction is constructed. This cutting plane is typically based on the boundary between the two crown sub-contours in the image or the minimum closed area of ​​the edge angle, determining its spatial position and normal direction. The auxiliary cutting operation re-divides the original cluster by intercepting the set of points in the point cloud cluster whose distance from the cutting plane is greater than a threshold, thereby separating the two mistakenly merged trees into independent units.

[0154] For example, if an unusual area in an image is identified as having two symmetrically arranged crowns with an angle of approximately 60° between the crown edge contours, back-projection can be used to locate two local density centers in the point cloud. A cutting plane is established with the perpendicular bisector connecting these two centers as its normal vector. By intercepting the point cloud around this section line, image-driven re-segmentation of the cluster block is completed, reorganizing the two actually independent trees into new single-tree data units and rebinding their numbers and structural parameters.

[0155] The species classification unit 303 is used to recalculate the diameter at breast height, tree height, and crown width of each sub-block after the division is completed to obtain the final single tree data unit, perform species classification based on the final single tree data unit, and identify specific tree species through species classification.

[0156] In this module, after the species classification unit 303 is divided, the structural parameters of each re-divided sub-block are independently calculated to generate the final valid single tree data unit. A horizontal section is extracted at a height of 1.3 meters above the ground. The circular or elliptical cross-section is fitted using the point cloud slice density profile to calculate the diameter. The Z-axis range is calculated from the highest point on the ground to the highest point of the crown. The sub-block is projected onto the horizontal plane, the maximum envelope boundary of the point cloud on the XY plane is extracted, and the average diameter of the major and minor axes is calculated.

[0157] After parameter calculation, the sub-blocks are combined with the contour, texture, and color features of the corresponding regions in the image to further identify species. This recognition is based on a pre-trained convolutional neural network model. It uses the image's texture directionality, color distribution, and leaf morphology to generate feature vectors, which are then compared for similarity with a database of standard species image vectors. The output category is the species name, such as Cinnamomum camphora, Ginkgo biloba, and Hackberry, along with a confidence score to assess the reliability of the recognition result.

[0158] For example, for two sub-plots obtained from a single stripping operation, the calculated DBH values ​​were 24.3cm and 17.5cm, respectively, with tree heights of 12.7m and 9.6m, and crown widths of 5.1m and 3.8m, respectively. Image recognition confirmed that the first sub-plot matched a camphor tree with a confidence score of 0.94, while the second sub-plot matched a Chinese elm tree with a confidence score of 0.89. Both exceeded the threshold and were confirmed as valid individual trees. Ultimately, these two independent data units were stored in the database and subsequently used by the carbon storage model.

[0159] like Figure 8 As shown, as a preferred embodiment of the present invention, the level determination module 400 includes:

[0160] The structural region division unit 401 is used to obtain the sample plot area and divide the sample plot into different levels of structural regions based on the final single tree data unit. The structural regions include structural homogeneous regions, structural transition regions, and structural heterogeneous regions, and calculate the average single tree carbon storage and total number of trees in the structural regions.

[0161] In this module, the structural region division unit 401 obtains the plot area and divides the plot into spatial structural regions based on the distribution of individual tree structural characteristics. First, the spatial boundary of the current plot is obtained to determine its area, shape, and geographic coordinate distribution. Then, based on the structural parameters (diameter at breast height D, tree height H, crown width W) and spatial distribution coordinates in the individual tree data unit, the entire plot is scanned for structural characteristics using a local statistical method (e.g., a 3×3m or 5×5m sliding window).

[0162] The plots were divided into three structural areas: structural homogeneous area: characterized by low H0, low M, and low W, with highly uniform tree species and sizes, commonly found in plantations or neat green belts; structural transition area: characterized by medium H0 and M, commonly found in partially regenerated or small-scale mixed areas; structural heterogeneous area: characterized by high H0, high M, and high W, with complex structure and mixed tree species, typical of natural secondary forests or long-term natural growth areas;

[0163] For each structural area, the total carbon storage and total number of trees in all individual tree data units are counted, and the average carbon storage of individual trees is calculated. Carbon storage calculation is based on the individual carbon storage value generated by the dynamic biomass matching mechanism (such as B j =a(D 2 H) b ×CF, where CF is the carbon factor).

[0164] For example, in a certain urban green space sample plot, the structurally heterogeneous area contained 35 trees with a total carbon storage of 6.3 tons, with an average carbon storage of 0.18 tons per tree. The structurally homogeneous area contained 46 trees with a total carbon storage of 5.2 tons, with an average of 0.11 tons. The structural transition area was somewhere in between. These data provided the basis for subsequent regional weighting and overall performance evaluation.

[0165] The grade determination unit 402 is used to construct a structural weighting factor based on the area of ​​the sample plot, integrate the total carbon storage value of the entire sample plot based on the structural weighting factor, obtain the sample plot carbon storage grade evaluation rule, and determine the corresponding grade based on the total carbon storage value of the entire sample plot.

[0166] In this module, the grade determination unit 402 constructs a structural weighting factor based on the area of ​​the sample plot, and calculates the actual area A of each type of structural area (such as structural homogeneous area, structural transition area, and structural heterogeneous area) in the sample plot. i and the total carbon storage in the region C i Based on the proportion of area to the total area of ​​the plot w i =A i / A total , construct the structural weighting factor w i , and use this to perform weighted integration of the carbon storage values ​​of each region:

[0167] ;

[0168] Among them, Ctotal is the weighted total carbon storage value of the sample site, which reflects the different contributions of different structural regions to carbon storage capacity. After integrating the carbon storage values, it is necessary to introduce a carbon storage grade evaluation rule. This rule is based on the historical sample site statistical data of similar areas (such as urban green spaces, ecological isolation zones, secondary forests, etc.) and sets a theoretical carbon storage upper limit C max , median C mid With bottom line C min . Change the current plot's C total Comparing with these benchmark values, carbon storage levels are classified, for example:

[0169] Level 1 (Excellent): C total ≥0.8×C max ;

[0170] Level 2 (good): 0.6×C max ≤C total <0.8×C max ;

[0171] Level 3 (medium): 0.4×C max ≤C total <0.6×C max ;

[0172] Level 4 (poor): C total <0.4×C max ;

[0173] For example, a 1-hectare urban plot consists of 40% structurally heterogeneous areas, 35% structurally transitional areas, and 25% structurally homogeneous areas. The total carbon storage in these three areas is 8.2 tonnes, 5.4 tonnes, and 3.1 tonnes, respectively. A weighted calculation yields a weighted total carbon storage of 6.18 tonnes. If the historical upper limit of carbon storage in this area is 8.5 tonnes, then 6.18 tonnes is approximately 73% of that, placing it at a "good" level.

[0174] The deviation rate unit 403 is used to calculate the theoretical carbon storage upper limit, calculate the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample site, obtain the deviation rate threshold, and if the relative deviation rate is greater than the deviation rate threshold, obtain an improvement suggestion list and send the list to the terminal.

[0175] In this module, the deviation rate unit 403 calculates the theoretical carbon storage upper limit. Based on the structural area division results, the theoretical optimal single tree carbon storage is introduced into each type of area, and then multiplied by the total number of corresponding trees to obtain the theoretical carbon storage upper limit C. max For example, the theoretical carbon storage of a single tree in a structurally heterogeneous area is 0.28 t, and the actual number of trees is 40, so the upper limit of this area is 11.2 t.

[0176] Conduct deviation analysis on the carbon storage performance of the current plot and calculate its relative deviation rate:

[0177] ;

[0178] The deviation rate reflects the distance between the carbon storage of the sample site and its potential. In order to distinguish whether intervention is needed, a deviation rate threshold (such as 25%) is introduced as a management trigger standard. 偏离 >R 阈值 , the system determines that there is room for optimization in this area.

[0179] Based on this, reference plots with similar structures to the current plot but superior carbon storage performance are retrieved from the database. Their species composition, density, interplant ratio, age structure, and other characteristics are analyzed, and a list of improvement recommendations is automatically generated, such as increasing interplant ratios, introducing high-carbon storage dominant species like Hackberry and Pistacia, and optimizing plant spacing. This list is formatted and automatically pushed to forestry work terminals for managers to conduct feasibility assessments and make restoration decisions.

[0180] For example, the theoretical upper limit of a certain plot is 9.4t, but the actual value is 6.3t, with a deviation rate of 32.9%, which is higher than the set threshold of 25%. The system determines it as "medium-low adaptation" and pushes a suggestion: replant 30 camphor trees in the structural transition zone to increase the mixed planting index, with an estimated carbon storage gain of 0.9t.

[0181] In one embodiment, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the following steps are performed:

[0182] Acquire research area information, wherein the research area information includes the research area, research time, and research requirements, wherein the research time includes data collection time and plot setting duration;

[0183] Install the equipment according to the installation requirements. After the equipment is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud is obtained. According to the relationship model, a single tree data unit is generated and a unique number is bound to the single tree data unit.

[0184] Extract the contour features of the image area corresponding to the single tree data unit, identify whether there are contradictions in the contour features, perform secondary stripping on abnormal contours, obtain the final single tree data unit, and perform species classification based on the final single tree data unit;

[0185] The structural weighting factor is constructed based on the area of ​​the sample plot, the corresponding level is determined according to the total carbon storage value of the entire sample plot, the theoretical upper limit of carbon storage is calculated, and the relative deviation rate of current carbon storage is calculated based on the theoretical upper limit of carbon storage and the total carbon storage value of the sample plot.

[0186] In one embodiment, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor performs the following steps:

[0187] Acquire research area information, wherein the research area information includes the research area, research time, and research requirements, wherein the research time includes data collection time and plot setting duration;

[0188] Install the equipment according to the installation requirements. After the equipment is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud is obtained. According to the relationship model, a single tree data unit is generated and a unique number is bound to the single tree data unit.

[0189] Extract the contour features of the image area corresponding to the single tree data unit, identify whether there are contradictions in the contour features, perform secondary stripping on abnormal contours, obtain the final single tree data unit, and perform species classification based on the final single tree data unit;

[0190] The structural weighting factor is constructed based on the area of ​​the sample plot, the corresponding level is determined according to the total carbon storage value of the entire sample plot, the theoretical upper limit of carbon storage is calculated, and the relative deviation rate of current carbon storage is calculated based on the theoretical upper limit of carbon storage and the total carbon storage value of the sample plot.

[0191] It should be understood that, although the various steps in the flow chart of each embodiment of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0192] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When executed, the program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0193] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0194] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0195] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for assessing carbon storage in urban forest communities based on multi-source data, characterized in that: The method comprises: Acquire research area information, wherein the research area information includes the research area, research time, and research requirements, wherein the research time includes data collection time and plot setting duration; Install the equipment according to the installation requirements. After the equipment is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud is obtained. According to the relationship model, a single tree data unit is generated and a unique number is bound to the single tree data unit. Extract the contour features of the image area corresponding to the single tree data unit, identify whether there are contradictions in the contour features, perform secondary stripping on abnormal contours, obtain the final single tree data unit, and perform species classification based on the final single tree data unit; Specifically, it includes extracting the contour features of the image area corresponding to the single tree data unit, the contour features include crown shape, texture symmetry and edge continuity, identifying whether there is a contradiction in the contour features, the contradiction is the inconsistency between the image display and the point cloud clustering, if the contradiction is identified, it is determined that the current clustering is abnormal; performing secondary stripping on the abnormal contour, the secondary stripping is guided by the main trunk direction or crown edge in the image, constructing an auxiliary cutting plane in the point cloud, and re-dividing the cluster block; after the division is completed, recalculating the diameter at breast height, tree height, and crown width of each sub-block to obtain the final single tree data unit, performing species classification based on the final single tree data unit, and identifying specific tree species through species classification; The structural weighting factor is constructed based on the area of ​​the sample plot. The corresponding grade is determined based on the total carbon storage value of the entire sample plot. The theoretical upper limit of carbon storage is calculated. The relative deviation rate of current carbon storage is calculated based on the theoretical upper limit of carbon storage and the total carbon storage value of the sample plot. Specifically, the following are included: Obtain the sample plot area and divide the sample plot into different levels of structural regions based on the final single tree data unit. The structural regions include structural homogeneous regions, structural transition regions, and structural heterogeneous regions. Calculate the average single tree carbon storage and total number of trees within the structural regions. Combine the area of ​​the sample plot to construct a structural weighting factor, integrate the total carbon storage value of the entire sample plot based on the structural weighting factor, obtain the sample plot carbon storage grade evaluation rules, and determine the corresponding grade based on the total carbon storage value of the entire sample plot; Calculate the theoretical carbon storage upper limit, calculate the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample site, obtain the deviation rate threshold, and if the relative deviation rate is greater than the deviation rate threshold, obtain a list of improvement suggestions and send the list to the terminal.

2. The urban forest community carbon storage assessment method based on multi-source data according to claim 1 is characterized in that: The steps of installing the device according to the installation requirements, obtaining the geometric relationship between the position in the panoramic image and the center direction of the point cloud after the device is installed, generating a single tree data unit according to the relationship model, and binding the single tree data unit to a unique number specifically include: Obtain device information, obtain installation requirements, and install the device according to the installation requirements, including installation height and calibration plate; After the device is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud obtained by laser radar scanning is obtained, and the image recognition confidence and acquisition integrity thresholds are obtained. The image recognition confidence and acquisition integrity of the panoramic image are evaluated according to the thresholds; A projection envelope relationship model between the image and the point cloud is constructed based on the qualified geometric relationship, a single tree data unit is generated based on the relationship model, and a unique number is bound to the single tree data unit. The single tree data unit includes the diameter at breast height, tree height and crown width.

3. The urban forest community carbon storage assessment method based on multi-source data according to claim 1 is characterized in that: The device information includes lidar and panoramic camera.

4. An urban forest community carbon storage assessment system based on multi-source data, characterized in that: The system comprises: A research preparation module, which obtains research area information, including the research area, research time, and research requirements. The research time includes data collection time and plot setting duration; The single tree data module installs the equipment according to the installation requirements. After the equipment is installed, the geometric relationship between the position in the panoramic image and the center direction of the point cloud is obtained. The single tree data unit is generated according to the relationship model and the single tree data unit is bound to a unique number. The secondary stripping module extracts the contour features of the image area corresponding to the single tree data unit, identifies whether there are any contradictions in the contour features, performs secondary stripping on abnormal contours, obtains the final single tree data unit, and performs species classification based on the final single tree data unit; The secondary stripping module includes: The contour feature unit extracts the contour features of the image area corresponding to the single tree data unit. The contour features include crown shape, texture symmetry, and edge continuity. It identifies whether there is a contradiction in the contour features. The contradiction is the inconsistency between the image display and the point cloud clustering. If a contradiction is found, it is determined that the current clustering is abnormal. A secondary stripping unit performs secondary stripping on abnormal contours. The secondary stripping is guided by the main trunk direction or crown edge in the image, constructs an auxiliary cutting plane in the point cloud, and re-divides the cluster blocks; After the species classification unit is divided, the diameter at breast height, tree height, and crown width of each sub-block are recalculated to obtain the final single tree data unit. Species classification is performed based on the final single tree data unit, and specific tree species are identified through species classification; The grade determination module builds a structural weighting factor based on the area of ​​the sample plot, determines the corresponding grade based on the total carbon storage value of the entire sample plot, calculates the theoretical upper limit of carbon storage, and calculates the relative deviation rate of current carbon storage based on the theoretical upper limit of carbon storage and the total carbon storage value of the sample plot; The level determination module includes: Structural region division unit: obtain the sample plot area, and divide the sample plot into different levels of structural regions according to the final single tree data unit. The structural regions include structural homogeneous regions, structural transition regions, and structural heterogeneous regions. Calculate the average single tree carbon storage and total number of trees within the structural regions. The grade determination unit constructs a structural weighting factor based on the area of ​​the sample plot, integrates the total carbon storage value of the entire sample plot based on the structural weighting factor, obtains the sample plot carbon storage grade evaluation rules, and determines the corresponding grade based on the total carbon storage value of the entire sample plot; The deviation rate unit calculates the theoretical carbon storage upper limit, calculates the relative deviation rate of the current carbon storage based on the theoretical carbon storage upper limit and the total carbon storage value of the sample site, obtains the deviation rate threshold, and if the relative deviation rate is greater than the deviation rate threshold, obtains the improvement suggestion list and sends the list to the terminal.

5. The urban forest community carbon storage assessment system based on multi-source data according to claim 4 is characterized in that: The single tree data module includes: The equipment installation unit obtains equipment information and installation requirements, and installs the equipment according to the installation requirements, including the installation height and calibration plate; An information acquisition unit, after the device is installed, obtains the geometric relationship between the position in the panoramic image and the center direction of the point cloud obtained by laser radar scanning, obtains the image recognition confidence and acquisition integrity thresholds, and evaluates the image recognition confidence and acquisition integrity of the panoramic image based on the thresholds; A single tree binding unit constructs a projection envelope relationship model between the image and the point cloud based on the qualified geometric relationship, generates a single tree data unit based on the relationship model, and binds the single tree data unit to a unique number. The single tree data unit includes the diameter at breast height, tree height and crown width.

6. The urban forest community carbon storage assessment system based on multi-source data according to claim 5 is characterized in that: The device information includes lidar and panoramic camera.

Images