Tree biomass lossless estimation method and system suitable for branch level

Abstract

The invention discloses a tree biomass lossless estimation method and system suitable for a branch level, and relates to the technical field of forestry ecological remote sensing. The method comprises the following steps: collecting point cloud data of a target tree and preprocessing the point cloud data to obtain branch point cloud data and leaf point cloud data; reconstructing the branch point cloud data by using an L1-Tree algorithm to obtain a branch network structure of the target tree, and calculating the branch biomass and the total branch biomass of each grade according to the branch grade by combining the density of the target tree; calculating the leaf point cloud data by using a voxelization algorithm to form a voxel space, and calculating the total biomass of the leaf according to the number of voxels of the leaf point cloud contained in the voxel space and the leaf density; adding the total biomass of the branches and the total biomass of the leaves to obtain the biomass of a single tree; and calculating the proportion of the biomass of each grade in the biomass of the single tree according to the branch biomass of each grade and the biomass of the single tree. The method is used for improving the estimation precision of the tree biomass.

Description

Technical Field

[0001] The present invention relates to the technical field of forest ecological remote sensing, and particularly to a method and system for non-destructively estimating tree biomass applicable to the branch level. Background Art

[0002] Biomass is a key indicator for measuring the carbon storage, productivity, and health status of forest ecosystems. Accurately estimating tree biomass is of great significance for forest resource management, climate change research, and the assessment of carbon markets. The biomass distribution of trees in different organs reflects the growth strategies and health status of the trees. Quantifying the biomass of individual trees at the branch level helps to deeply understand the situation and strategies of tree resource allocation, serving the accurate assessment of tree health and forest tree breeding.

[0003] Currently, the estimation methods of tree biomass mainly rely on two technical paths: the direct destructive sampling method and the estimation method based on allometric equations. The direct destructive sampling method is to cut or collect a certain number of tree samples and measure biomass indicators such as dry weight, branch weight, and leaf weight. The advantage of this method is high accuracy, but its disadvantages are also very obvious: the sampling process will directly damage the trees and the sample size is limited, making it difficult to cover various tree species. More importantly, the destructive sampling method usually requires a large amount of time and manpower, with high costs, and is not suitable for large-scale forest surveys. In addition, for endangered tree species and key protected tree species, it is impossible to obtain tree biomass by using the direct destructive sampling method.

[0004] The allometric equation estimation method is developed on the basis of the direct destructive sampling method. In addition to the disadvantages of the direct destruction method, it is also affected by various factors such as environment, climate, tree species, and age, resulting in large differences in the applicability and accuracy of the model among different regions and tree species. In addition, the tree species used in the allometric equation method are also very limited, and there is a lack of allometric equations for some tree species, and it is not suitable for trees affected by human activities (such as pruning).

[0005] LiDAR can accurately obtain the three-dimensional information of ground objects by emitting laser beams and measuring the time and intensity of the reflected laser beams, and is particularly suitable for large-scale and non-destructive forest resource surveys. LiDAR technology also faces some challenges in practical applications. First, the data volume obtained by LiDAR is huge, and complex algorithms and a large amount of computing resources are required for processing; second, the existing methods for estimating biomass based on LiDAR mostly use the allometric equation method or the traditional volume estimation method, lacking detailed modeling of the internal structure of trees and unable to obtain more refined biomass components. This makes the utilization efficiency and estimation accuracy of LiDAR data need to be further improved.

[0006] Therefore, how to effectively utilize lidar data for high-precision and non-destructive estimation of tree biomass, especially being able to finely distinguish the biomass of different levels of tree branches (such as the main trunk, first-level branches, second-level branches, etc. and leaves), remains an important challenge in current technical research. Summary of the Invention

[0007] The object of the present invention is to provide a method and system for non-destructive estimation of tree biomass applicable to the branch level, which is used to accurately estimate the biomass of different levels of branches and leaves of individual trees, thereby improving the accuracy and reliability of tree biomass estimation, and providing technical support for accurate assessment of tree health conditions, tree carbon storage accounting, and forest tree breeding, etc.

[0008] In the first aspect, a method for non-destructive estimation of tree biomass applicable to the branch level provided by the present invention adopts the following technical solutions: A method for non-destructive estimation of tree biomass applicable to the branch level includes: Collect the point cloud data of the target tree, and preprocess the point cloud data to obtain branch point cloud data and leaf point cloud data; Provide the L1-Tree algorithm, use the L1-Tree algorithm to reconstruct the branch point cloud data to obtain the branch network structure of the target tree, and calculate the branch biomass and the total branch biomass of each level according to the branch network structure and the density of the target tree; Provide a voxelization algorithm, use the voxelization algorithm to calculate the leaf point cloud data to obtain a voxel space containing the leaf point cloud, and calculate the total leaf biomass according to the number of voxels containing the leaf point cloud and the leaf density in the voxel space; Add the total branch biomass and the total leaf biomass to obtain the biomass of the individual tree; according to the branch biomass of each level and the biomass of the individual tree, calculate the proportion of the branch biomass of each level in the biomass of the individual tree.

[0009] A further technical solution lies in that the preprocessing of the point cloud data to obtain branch point cloud data and leaf point cloud data specifically includes: Provide a statistical denoising algorithm and a branch-leaf separation algorithm; Use the statistical denoising algorithm and / or manual cropping method to remove the noise points in the point cloud data; Use the branch-leaf separation algorithm and / or manual editing method to divide the point cloud data into branch point cloud and leaf point cloud.

[0010] A further technical solution lies in that the calculation of the branch biomass and the total branch biomass of each level according to the branch network structure and the density of the target tree specifically includes: Extract the branch skeleton lines and corresponding branch point clouds of each level from the branch network structure according to the branch level; take the length of each branch skeleton line as the branch length, take the fitted thickness of each branch point cloud as the branch thickness, and multiply the branch length and branch thickness of each branch to obtain the volume of each branch; According to the type of the target tree, obtain the branch density of the target tree from the existing database; Multiply the branch density of the target tree by the volume of each branch to obtain the branch biomass of each level; Accumulate the branch biomass of each level to obtain the total branch biomass.

[0011] A further technical solution lies in that the voxelization algorithm is used to calculate the leaf point cloud data to obtain a voxel space containing the leaf point cloud, and the total leaf biomass is calculated according to the number of voxels containing the leaf point cloud and the leaf density in the voxel space, specifically including: Calculate the average point spacing of the leaf point cloud; and use 3 times the length of the average point spacing of the leaf point cloud as the voxel side length to voxelize the point cloud to obtain a voxel space containing the leaf point cloud; Calculate the number of voxels containing the leaf point cloud in the voxel space, and multiply the number of voxels by the volume of a single voxel to obtain the leaf volume; Obtain the leaf density in the voxel space by the actual measurement method, and multiply the leaf volume by the leaf density to obtain the total leaf biomass.

[0012] A further technical solution lies in that, according to the branch biomass of each level and the biomass of a single tree, calculate the proportion of the branch biomass of each level in the biomass of a single tree, including: Divide the branch biomass of each level and the leaf biomass by the biomass of a single tree respectively to obtain the proportion of the branches and leaves of each level in the biomass of a single tree.

[0013] Compared with the prior art, in the method provided by the present invention, by adopting the L1-tree branch network reconstruction algorithm and the leaf voxelization technology, the biomass of branches and leaves at different levels of trees can be accurately estimated, thereby improving the accuracy and reliability of tree biomass estimation, and providing technical support for accurate assessment of tree health conditions, tree carbon storage accounting, forest tree breeding, etc.

[0014] In a second aspect, a tree biomass non-destructive estimation system applicable to the branch level provided by the present invention adopts the following technical solution: A tree biomass non-destructive estimation system applicable to the branch level, characterized in that it includes a processor and a memory connected to the processor; a computer program capable of being loaded and executed by the processor as any one of the methods in the first aspect is stored on the memory.

[0015] Compared with the prior art, the beneficial effects of the system provided by the present invention are the same as those of the method for non-destructively estimating the biomass of trees applicable to the branch level described in the above technical solution, and will not be elaborated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings: Figure 1 is a flowchart of a method provided by an embodiment of the present invention; Figure 2 is a schematic diagram of the ground-based lidar point cloud data of trees and the tree branch point cloud and leaf point cloud data obtained after preprocessing provided by an embodiment of the present invention; Figure 3 is a schematic diagram of the tree branch network generated from the branch network data provided by an embodiment of the present invention; Figure 4 is a schematic diagram of the voxelization of the leaf point cloud provided by an embodiment of the present invention; Figure 5 is a distribution diagram of the proportion of the biomass of different levels of branches and leaves provided by an embodiment of the present invention; Figure 6 is a comparison chart of the accuracy of estimating the biomass by the method of the present invention and the allometric equation method. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The present invention will be further described in detail below with reference to the accompanying drawings.

[0018] This specific embodiment is only an interpretation of the present invention and does not limit the present invention. Those skilled in the art can make modifications to this embodiment without creative contributions according to needs after reading this specification, but as long as they are within the scope of the claims of the present invention, they are protected by the patent law.

[0019] It should be noted that in the present invention, words such as "exemplary" or "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly speaking, using words such as "exemplary" or "for example" is intended to present relevant concepts in a specific manner.

[0020] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "And / or" describes the relationship between related objects and indicates that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the related objects before and after.

[0021] The present invention provides a method for non-destructively estimating the biomass of trees at the branch level. This method requires the use of ground-based lidar point clouds, and then through steps such as preprocessing, foliage separation, branch network reconstruction, branch volume estimation, leaf point cloud voxelization, and leaf volume estimation, the estimation of the biomass of different levels of branches, leaf biomass, and total biomass of trees is realized.

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0023] An embodiment of the present invention provides a method for non-destructively estimating the biomass of trees at the branch level. The main process of the method is described as follows: Please refer to Figure 1 : Step S1: Collect the point cloud data of the target tree, and preprocess the point cloud data to obtain branch point cloud data and leaf point cloud data.

[0024] Among them, the point cloud data is a three-dimensional space data set collected and processed by lidar. The point cloud data of the target tree can be obtained through currently commonly used ground-based lidar. During the process of obtaining the data, a three-station scanning method is adopted around the target tree. For taller trees, an elevation scanning step needs to be considered to ensure data integrity. After obtaining the data, the point cloud of the target tree needs to be manually cropped to remove the point clouds of other non-target trees.

[0025] Among them, the preprocessing of the point cloud data includes two steps: denoising and foliage separation. By preprocessing the point cloud data, branch point cloud data and leaf point cloud data are obtained, which specifically includes the following steps: Step S11: Provide a statistical denoising algorithm and a foliage separation algorithm; Step S12: Use the statistical denoising algorithm and / or manual cropping method to remove the noise points in the point cloud data; Step S13: Use the branch and leaf separation algorithm and / or manual editing method to divide the point cloud data into branch point clouds and leaf point clouds.

[0026] Among them, noise points are caused by branch shaking, multipath effects, etc. during the acquisition of ground-based lidar data. Statistical denoising methods can be used to process them. The statistical denoising algorithm is a technique for image denoising through statistical methods, mainly including mean filters and median filters. The statistical denoising algorithm calculates the spatial distance distribution between a single point in the overall data and other points in its neighborhood. If the spatial distance between a certain point and other points in its neighborhood is greater than the average value + 3 times the standard deviation, it can be considered a noise point and removed.

[0027] The branch and leaf separation algorithm is a technique for separating branches and leaves in tree point cloud data, mainly used for accurately calculating above-ground biomass and leaf area index, as well as for three-dimensional tree modeling.

[0028] In the embodiment of the present invention, the branch and leaf separation algorithm can adopt the LeWos model method. First, use the LeWos model to perform preliminary branch and leaf separation on the point cloud, separating branch and leaf point clouds. Then, perform fine separation on the mixed point cloud through the path tracing detection algorithm, select the optimal path length to execute the path tracing detection algorithm to achieve a higher separation accuracy.

[0029] The branch and leaf separation algorithm can also be carried out by using a threshold-based branch and leaf separation algorithm and manual editing. By calculating the line features of the point cloud, set a threshold to separate the leaf point cloud from the branch point cloud. When the separation accuracy is not high, further separate the difficult-to-classify leaf point clouds through manual editing, and manually supplement the broken part of the branch point cloud. Finally, generate branch point clouds and leaf point clouds, as Figure 2 shown.

[0030] Step S2: Provide the L1-Tree algorithm, use the L1-Tree algorithm to reconstruct the branch point cloud data to obtain the branch network structure of the target tree, and calculate the branch biomass and total branch biomass of each level according to the branch network structure and the density of the target tree.

[0031] Among them, the L1-Tree algorithm uses the classic method L1-Median in the field of computer graphics to extract the skeleton of the current branch point cloud data, and on this basis, uses the tree growth strategy to optimize the skeleton, accurately estimate the fine structure parameters of the branches, and realize the reconstruction of the branch point cloud data, as Figure 3 shown. Among them, the branch level in botany refers to the classification method of tree branches, usually divided according to the order of branching in the crown. The primary branches are the branches growing on the central trunk, the secondary branches grow on the primary branches, the tertiary branches grow on the secondary branches, and so on. That is, the primary branches are usually thicker than the secondary and tertiary branches.

[0032] Calculate the branching biomass and the total branch biomass of each level according to the branch network structure and the density of the target tree, which specifically includes the following steps: Step S21: Extract the branching skeleton lines and the corresponding branching point clouds of each level from the branch network structure according to the branch level; take the length of each branching skeleton line as the branch length, take the fitted thickness of each branching point cloud as the branch thickness, and multiply the branch length and branch thickness of each branch to obtain the volume of each branch.

[0033] Among them, there are two ways to calculate the branch volume: First, calculate the total branch length according to the path of the skeleton line from the starting point to the ending point, project all the point clouds onto the direction of the branch, and use the Hough transform method to calculate the average branch thickness, and multiply the two to get V 枝 ; Second, calculate the thickness at a certain interval in the direction of the branch, the interval is the length, multiply the length and the thickness to obtain the volume of a section of the branch, and sum the volumes of each section to obtain the volume V of the corresponding branch 枝 。

[0034] Step S22: According to the type of the target tree, obtain the branch density of the target tree from the existing database; Step S23: Multiply the branch density of the target tree and the volume of each branch to obtain the branching biomass of each level; Step S24: Sum the branching biomass of each level to obtain the total branch biomass.

[0035] Among them, the existing database includes literature, databases or the branch density ρ obtained by the actual measurement method 枝 。 The biomass AGB of a branch of any level (i, j) 枝i,j can be calculated by the following formula: AGB 枝i,j =V 枝i,j *ρ 枝。

[0036] Among them, i represents the level of a certain branch, and j is a certain branch in the branches of this level. Summarize the biomass of branches of different levels to obtain the branch biomass AGB of this tree 枝 。

[0037] Step S3: Provide a voxelization algorithm, and use the voxelization algorithm to calculate the leaf point cloud data to obtain a voxel space containing the leaf point cloud. Calculate the total leaf biomass according to the number of voxels containing the leaf point cloud and the leaf density in the voxel space.

[0038] Among them, voxelization is the process of converting a three-dimensional geometric model (such as a polygon mesh, surface, or point cloud) into a uniform cubic grid, and each cube is called a voxel, as Figure 4 shown. The size of the voxel can be set according to specific requirements. Smaller voxels can better capture the details of the object, but will increase the computational amount; larger voxels will simplify the calculation but may lose important geometric information.

[0039] In the embodiment of the present invention, the voxelization algorithm is used to calculate the leaf point cloud data to obtain a voxel space containing the leaf point cloud. Specifically, after the voxelization process of the leaf point cloud, calculate the average point spacing of the leaf point cloud, and set the voxel length according to three times the average point spacing of the leaf point cloud to obtain a voxel space containing the leaf point cloud.

[0040] In this step S3, calculating the total leaf biomass according to the number of voxels containing the leaf point cloud and the leaf density in the voxel space specifically includes: Step S31: Calculate the number of voxels containing the leaf point cloud in the voxel space, and multiply the number of voxels by the volume of a single voxel to obtain the leaf volume V 叶 ; Step S32: Obtain the leaf density in the voxel space through the actual measurement method, and multiply the leaf volume by the leaf density to obtain the total leaf biomass.

[0041] Among them, obtaining the leaf density in the voxel space through the drainage method specifically means obtaining the measured density ρ of the leaves of this tree species by weighing a small number of leaves and dividing by the number of voxels they occupy 叶 , and then obtaining the total leaf biomass AGB through the following formula 叶 : AGB 叶 =V 叶 *ρ 叶。

[0042] Step S4: Add the total branch biomass and the total leaf biomass to obtain the biomass of a single tree; calculate the proportion of the branch biomass of each level in the biomass of a single tree according to the branch biomass of each level and the biomass of a single tree.

[0043] Among them, the biomass of a single tree is the above-ground biomass of the target tree, excluding the root biomass of the target tree.

[0044] Calculate the proportion of the branch biomass at each level in the biomass of a single tree. Specifically, divide the branch biomass and leaf biomass at each level by the biomass of the single tree respectively to obtain the proportion of the branches and leaves at each level in the biomass of the single tree, as Figure 5 shown.

[0045] Compared with the traditional biomass estimation method, the method provided by the present invention avoids the damage to trees caused by obtaining biomass through the actual measurement method, and is especially applicable to the biomass monitoring of endangered tree species lacking allometric equations. At the same time, it also overcomes the problems such as large errors in the use of traditional allometric equations in different regions and tree species; the proportion information of the biomass of branches and leaves at different levels is of great significance for understanding the situation and strategy of tree resource allocation, for the accurate assessment of the tree health condition, and for guiding forest tree breeding.

[0046] Taking the direct destruction sampling method as the actual measured biomass, compare the method of the present invention with the estimation method based on the allometric equation. The estimation accuracies of the two are as Figure 6 shown. Obviously, the present invention has a higher estimation accuracy.

[0047] The embodiment of the present application also provides a non-destructive estimation system for the biomass of trees applicable to the branch level, including a processor and a memory connected to the processor; a computer program capable of being loaded and executed by the processor as any one of the foregoing methods is stored on the memory.

[0048] Although the present invention has been described in conjunction with specific features and their embodiments, it is obvious that various modifications and combinations can be made without departing from the spirit and scope of the present invention. Accordingly, this specification and the drawings are merely exemplary illustrations of the present invention defined by the appended claims, and are considered to have covered any and all modifications, variations, combinations, or equivalents within the scope of the present invention. Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A non-destructive estimation method for tree biomass at the branch level, characterized in that: include: Collecting point cloud data of the target tree, and preprocessing the point cloud data to obtain branch point cloud data and leaf point cloud data; Providing an L1-Tree algorithm, using the L1-Tree algorithm to reconstruct the branch point cloud data to obtain the branch network structure of the target tree, and calculating the branch biomass and the total branch biomass of each level according to the branch network structure and the density of the target tree according to the branch level; Providing a voxelization algorithm, using the voxelization algorithm to calculate the leaf point cloud data to obtain a voxel space containing the leaf point cloud, and calculating the total leaf biomass according to the number of voxels containing the leaf point cloud in the voxel space and the leaf density; The total biomass of the branches and the total biomass of the leaves are added together to obtain the biomass of the individual tree; based on the biomass of the branches at each level and the biomass of the individual tree, the proportion of the branch biomass at each level to the biomass of the individual tree is calculated.

2. The non-destructive estimation method of tree biomass at the branch and trunk level according to claim 1, characterized in that: The preprocessing of the point cloud data to obtain branch point cloud data and leaf point cloud data specifically includes: Provide statistical denoising algorithm and branch-leaf separation algorithm; Using the statistical denoising algorithm and / or manual cropping to remove noise points in the point cloud data; The point cloud data is divided into branch point cloud and leaf point cloud using the branch-leaf separation algorithm and / or manual editing.

3. The non-destructive estimation method of tree biomass at the branch and trunk level according to claim 1, characterized in that: The step of calculating the biomass of branches at each level and the total biomass of branches according to the branch network structure and the density of the target tree specifically includes: Extracting branch skeleton lines and corresponding branch point clouds of each level from the branch network structure according to the branch level classification; taking the length of each branch skeleton line as the branch length, taking the fitting thickness of each branch point cloud as the branch thickness, and multiplying the branch length and branch thickness of each branch to obtain the volume of each branch; According to the type of the target tree, obtaining the branch density of the target tree from an existing database; Multiplying the trunk density of the target tree and the volume of each branch to obtain the branch biomass of each level; The branch biomass of each level is accumulated to obtain the total branch biomass.

4. The non-destructive estimation method of tree biomass applicable to branch and trunk level according to claim 1, characterized in that: The method of calculating the leaf point cloud data by using a voxelization algorithm to obtain a voxel space containing the leaf point cloud, and calculating the total leaf biomass according to the number of voxels containing the leaf point cloud in the voxel space and the leaf density, specifically includes: Calculating the average point spacing of the leaf point cloud; voxelizing the point cloud using three times the length of the average point spacing of the leaf point cloud as the voxel side length to obtain a voxel space containing the leaf point cloud; Calculating the number of voxels in the voxel space that contain the leaf point cloud, and multiplying the number of voxels by the volume of a single voxel to obtain the leaf volume; The leaf density in the voxel space is obtained by actual measurement, and the leaf volume is multiplied by the leaf density to obtain the total leaf biomass.

5. The non-destructive estimation method of tree biomass applicable to branch and trunk level according to claim 1, characterized in that: According to the branch biomass of each level and the biomass of individual trees, the proportion of branch biomass of each level to the biomass of individual trees was calculated, including: The branch biomass and leaf biomass of each level were divided by the biomass of the individual tree to obtain the proportion of branches and leaves of each level in the biomass of the individual tree.

6. A non-destructive estimation system for tree biomass at the branch level, characterized in that: The invention comprises a processor and a memory connected to the processor; the memory stores a computer program which can be loaded by the processor and executes any one of the methods as claimed in claims 1 to 5.

Images