Leaf area index calculation method and device, equipment and storage medium

Abstract

The invention relates to a leaf area index calculation method and device, equipment and a storage medium. According to the method, the quantitative relation between the fractal dimension and the uniform canopy leaf area index in the canopy space is constructed based on the point cloud data of the canopy space, and the canopy heterogeneity index is defined to correct the leaf area index assumed by the uniform canopy, so that the influence of leaf aggregation in the three-dimensional space is considered; and the accuracy of leaf area index calculation is improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of data processing, and particularly to a method, apparatus, device, and storage medium for calculating leaf area index. Background Art

[0002] Leaf area index (LAI), defined as half of the total leaf area on a unit horizontal ground, is one of the important vegetation structure parameters in the carbon cycle and water cycle.

[0003] In recent years, the technology of terrestrial lidar (TLS) has developed rapidly. Compared with passive optical instruments, TLS is less sensitive to ambient light and can provide detailed three-dimensional structure information of the vegetation canopy, which is beneficial to the indirect measurement of LAI. Currently, the methods for estimating LAI based on TLS point clouds can generally be divided into two categories: voxel-based methods and projection-based methods. Among them, the estimation results of the voxel method are greatly affected by the voxel size used and it is difficult to fully correct the influence of the aggregation effect. While the projection method can better consider the influence of the aggregation effect, it is essentially a method based on passive optical images and fails to fully utilize the three-dimensional structure information described by the point cloud.

[0004] Therefore, the existing methods for calculating leaf area index are affected by various factors and their accuracy is insufficient. Summary of the Invention

[0005] To solve the above technical problems, the present disclosure provides a method, apparatus, device, and storage medium for calculating leaf area index to improve the accuracy of leaf area index calculation.

[0006] In a first aspect, an embodiment of the present disclosure provides a method for calculating leaf area index, including:

[0007] Constructing a quantitative relationship between the fractal dimension and the homogeneous canopy leaf area index in the canopy space based on the canopy point cloud data in the canopy space;

[0008] Calculating the leaf area index of the canopy space under the assumption of a homogeneous canopy according to the canopy point cloud data and the quantitative relationship;

[0009] Projecting the canopy point cloud data onto a plurality of preset projection planes in the canopy space to obtain a projection image corresponding to each preset projection plane;

[0010] Determining the canopy heterogeneity index of the canopy space according to the ratio of the number of point cloud pixels to the number of pixels in the projection image in the projection image, where the point cloud pixel is a pixel containing the canopy point cloud;

[0011] The leaf area index under the assumption of a uniform canopy is corrected according to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space.

[0012] In some embodiments, constructing the quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space based on the canopy point cloud data of the canopy space includes:

[0013] Dividing the canopy space into a plurality of sample cubes;

[0014] Calculating the maximum movement range when a single leaf intersects with a single sample cube;

[0015] According to the maximum movement range and the volume of the sample cube, calculating the probability that the sample cube intersects with the single leaf;

[0016] According to the leaf distribution characteristics in the canopy space and the probability that the sample cube intersects with the single leaf, constructing the quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space.

[0017] In some embodiments, the canopy space is a hexahedron, and the preset projection plane is three pairwise adjacent planes in the hexahedron; determining the canopy heterogeneity index of the canopy space according to the ratio of the number of point cloud pixels in the projection image to the number of pixels in the projection image includes:

[0018] For each projection image, calculating the ratio of the number of point cloud pixels to the number of pixels in the projection image to obtain the point cloud ratio corresponding to the projection image;

[0019] Calculating the product of the point cloud ratios corresponding to each projection image to obtain the canopy heterogeneity index of the canopy space.

[0020] In some embodiments, correcting the leaf area index under the assumption of a uniform canopy according to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space includes:

[0021] Calculating the ratio of the leaf area index under the assumption of a uniform canopy to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space.

[0022] In some embodiments, calculating the leaf area index of the canopy space under the assumption of a uniform canopy according to the canopy point cloud data and the quantitative relationship includes:

[0023] According to the canopy point cloud data, calculating the fractal dimension of the canopy space;

[0024] Calculate the leaf area index based on the uniform canopy assumption according to the fractal dimension of the canopy space and the quantitative relationship.

[0025] In some embodiments, calculating the fractal dimension of the canopy space based on the canopy point cloud data includes:

[0026] Determine a plurality of preset side length values;

[0027] For each preset side length value, divide the canopy space into a plurality of sample cubes with side lengths equal to the preset side length value, and count the number of point cloud cubes containing the point cloud data among the plurality of sample cubes;

[0028] Perform a linear regression on the plurality of preset side length values and the natural logarithm values of the number of point cloud cubes corresponding to each preset side length value to obtain a linear regression equation;

[0029] Determine the absolute value of the slope of the linear regression equation as the fractal dimension of the canopy space.

[0030] In some embodiments, the method further includes:

[0031] Layer the canopy space according to a preset layer height to obtain a plurality of canopy sub-spaces;

[0032] For each canopy sub-space, calculate the actual leaf area index of the canopy sub-space based on the canopy sub-space point cloud data corresponding to the canopy sub-space;

[0033] Use the actual leaf area index of the canopy sub-space as the index weight of the canopy sub-space;

[0034] Calculate the stratified leaf area index of the canopy sub-space according to the index weight of the canopy sub-space and the actual leaf area index of the canopy space.

[0035] In a second aspect, an embodiment of the present disclosure provides a leaf area index calculation device, including:

[0036] A construction module for constructing a quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space based on the canopy point cloud data of the canopy space;

[0037] A first calculation module for calculating the leaf area index of the canopy space under the assumption of a uniform canopy according to the canopy point cloud data and the quantitative relationship;

[0038] A projection module for projecting the canopy point cloud data onto a plurality of preset projection planes of the canopy space to obtain a projection image corresponding to each preset projection plane;

[0039] A determination module, configured to determine a canopy heterogeneity index of the canopy space according to a ratio of the number of point cloud pixels in the projection image to the number of pixels in the projection image, where the point cloud pixels are pixels containing canopy point clouds;

[0040] A correction module, configured to correct the leaf area index under the assumption of a uniform canopy according to the canopy heterogeneity index to obtain an actual leaf area index of the canopy space.

[0041] In a third aspect, an embodiment of the present disclosure provides an electronic device, including:

[0042] A memory;

[0043] A processor; and

[0044] A computer program;

[0045] wherein, the computer program is stored in the memory and is configured to be executed by the processor to implement the method as described in the first aspect.

[0046] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium, on which a computer program is stored, and the computer program is executed by a processor to implement the method as described in the first aspect.

[0047] In a fifth aspect, an embodiment of the present disclosure further provides a computer program product, which includes a computer program or instruction, and when the computer program or instruction is executed by a processor, the leaf area index calculation method as described above is implemented.

[0048] The leaf area index calculation method, device, equipment, and storage medium provided by the embodiments of the present disclosure construct a quantitative relationship between the fractal dimension and the leaf area index of a uniform canopy in the canopy space based on the point cloud data of the canopy space, and define a canopy heterogeneity index to correct the leaf area index under the assumption of a uniform canopy, so as to consider the influence of leaf aggregation in three-dimensional space and improve the accuracy of leaf area index calculation. Description of the Drawings

[0049] The drawings here are incorporated into the description and form a part of this description, showing embodiments consistent with the present disclosure and used together with the description to explain the principles of the present disclosure.

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0051] Figure 1Flowchart of the leaf area index calculation method provided by the embodiments of the present disclosure;

[0052] Figure 2 Example diagrams of a uniform canopy and an aggregated canopy with the same leaf area index provided by the embodiments of the present disclosure;

[0053] Figure 3 Schematic diagram of the three-dimensional coordinate system space provided by the embodiments of the present disclosure;

[0054] Figure 4 Schematic diagram of a leaf tangent to the surface of a sample cube provided by the embodiments of the present disclosure;

[0055] Figure 5 Schematic diagram of a leaf tangent to the edge of a cube provided by the embodiments of the present disclosure;

[0056] Figure 6 Schematic diagram of the TLS site setting provided by the embodiments of the present disclosure;

[0057] Figure 7 Example diagram of the distribution of fish-eye image sampling points in a sub-plot provided by the embodiments of the present disclosure;

[0058] Figure 8 Schematic diagram of the LAI result estimation of a uniform plot provided by the embodiments of the present disclosure;

[0059] Figure 9 Schematic diagram of the PAI result of the RAMI sub-plot estimated by using six comparison methods provided by the embodiments of the present disclosure;

[0060] Figure 10 Schematic diagram of the PAI result of the RAMI sub-plot estimated based on the complete canopy point cloud;

[0061] Figure 11 Schematic diagram of the comparison result between the average stratified PAI calculated by using the FDTLS method and the accurate average stratified PAI calculated by using LESS;

[0062] Figure 12 Schematic diagram of the structure of the leaf area index calculation device provided by the embodiments of the present disclosure;

[0063] Figure 13 Schematic diagram of the structure of the electronic device provided by the embodiments of the present disclosure. Detailed implementation manners

[0064] In order to be able to more clearly understand the above-mentioned objects, features, and advantages of the present disclosure, the solutions of the present disclosure will be further described below. It should be noted that, without conflict, the embodiments of the present disclosure and the features in the embodiments may be combined with each other.

[0065] Numerous specific details are set forth in the following description to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only a part of the embodiments of the present disclosure, rather than all the embodiments.

[0066] Ground measurements of LAI can be carried out by direct and indirect methods. Since the indirect method has the advantages of being convenient, efficient, non-destructive and suitable for large-scale repeated measurements, the indirect method is more widely used than the direct method when obtaining ground validation data for satellite LAI products. Indirect measurement methods are generally based on the Beer-Lambertian law. Considering the influence of the woody part in the vegetation canopy, the indirect measurement method usually obtains the plant area index (PAI).

[0067] In recent years, the technology of terrestrial laser scanning (TLS) has developed rapidly. Compared with passive optical instruments, TLS is less sensitive to environmental light, can provide detailed three-dimensional structural information of the vegetation canopy, and is beneficial to the indirect measurement of LAI. In addition, the vertical distribution of LAI, i.e., stratified LAI, can be estimated using canopy point cloud data. Currently, the methods for estimating LAI based on TLS point cloud can generally be divided into two categories: voxel-based methods and projection-based methods.

[0068] The voxel-based method first divides the canopy point cloud into small voxels, and then calculates the proportion of voxels without point cloud in each layer as the gap fraction of that layer. Then, according to the gap fraction, the Beer-Lambertian law is used to estimate the stratified LAI of each layer, and finally the total LAI is obtained through the LAI values of each layer. In addition to this method, the voxel-based method can also use ray tracing technology to calculate the gap fraction of each voxel, so that the leaf area volume density in each voxel can be calculated and converted into LAI. However, the voxel-based method is very sensitive to the voxel size, and there is currently no consensus on the optimal voxel size. In addition, the voxel-based method cannot fully consider the influence of the aggregation effect on LAI estimation because it assumes that the leaves in each voxel are randomly distributed.

[0069] The projection-based method projects the point cloud from a single-station scan into a fisheye-like photo. Then, methods such as the limited length averaging method (LX), the gap size distribution method (CC), the combination of the CC method and the LX method (CLX), or the path length distribution method (PATH) are used to estimate LAI. Although the projection-based method can consider the influence of the aggregation effect, it does not fully utilize the three-dimensional information of the point cloud and is essentially not much different from the method based on passive optical instruments.

[0070] Among them, the estimation result of the voxel method is greatly affected by the voxel size used, and it is difficult to fully correct the influence of the aggregation effect. Although the projection method can better consider the influence of the aggregation effect, it is essentially a method based on passive optical images and does not fully utilize the three-dimensional structure information described by the point cloud.

[0071] In view of the above problems, the embodiments of the present disclosure provide a method for calculating the leaf area index, and the method will be introduced below in combination with specific embodiments.

[0072] Figure 1 It is a flowchart of the method for calculating the leaf area index provided by the embodiments of the present disclosure. This method can be applied to any terminal device with data processing functions, such as smart phones, personal digital assistants, tablet computers, wearable devices with displays, desktop computers, laptop computers, all-in-one computers, etc. It can be understood that the method for calculating the leaf area index provided by the embodiments of the present disclosure can also be applied in other scenarios.

[0073] Next, Figure 1 the method for calculating the leaf area index shown will be introduced, and the specific steps included in the method are as follows:

[0074] S101. Construct a quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space based on the canopy point cloud data in the canopy space.

[0075] The fractal dimension (FD) is used to describe the complexity and self-similarity of natural objects. In this step, the fractal dimension is used to describe the complexity and self-similarity of the vegetation canopy.

[0076] Specifically, the quantitative relationship between the fractal dimension and the uniform canopy leaf area index is the functional relationship between the fractal dimension and the uniform canopy leaf area index, and the fractal dimension and the uniform canopy leaf area index are positively correlated.

[0077] S102. Calculate the leaf area index of the canopy space under the assumption of a uniform canopy according to the canopy point cloud data and the quantitative relationship.

[0078] A uniform canopy means that the leaves in the canopy space are randomly distributed, and the probability of each leaf being distributed at various positions in space is relatively uniform. Such a vegetation canopy is called a uniform canopy; the leaf area index under the assumption of a uniform canopy is the leaf area index of the canopy space calculated based on the quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the state of a uniform tree crown.

[0079] S103. Project the canopy point cloud data onto a plurality of preset projection planes in the canopy space to obtain a projection image corresponding to each preset projection plane.

[0080] S104. Determine the canopy heterogeneity index of the canopy space according to the ratio of the number of point cloud pixels in the projection image to the number of pixels in the projection image, where the point cloud pixels are pixels containing canopy point clouds.

[0081] In actual situations, the leaves in the canopy space tend to be aggregated rather than randomly distributed. Compared with a uniform canopy with the same leaf area index, the horizontal projection area of the aggregated canopy is smaller and the FD is higher.

[0082] Figure 2 This is an example diagram of a uniform canopy and an aggregated canopy with the same leaf area index provided by an embodiment of the present disclosure. As Figure 2 shown, the fractal dimension of the uniform canopy is lower than that of the aggregated canopy. The FD of a uniform tree canopy with an LAI of 2 is 1.80, while the FD of an aggregated canopy with an LAI of 2 is 2.49. In the projection image of the uniform canopy, each pixel contains point cloud data, while in the projection image of the aggregated canopy, there are some pixels that do not contain point cloud data.

[0083] Therefore, the leaf area index calculated based on the assumption of a uniform canopy in the above steps will be higher than the actual leaf area index of the canopy space.

[0084] Traditional methods based on the Beer-Lambertian law use a clumping index (CI) to correct the influence of leaf aggregation in the two-dimensional plane. However, since calculating FD based on point clouds is carried out in three-dimensional space, CI is not applicable.

[0085] To consider the influence of leaf aggregation in three-dimensional space, a canopy heterogeneity index (CHI) is defined.

[0086] Specifically, the canopy space is a hexahedron, and the preset projection plane is three pairwise adjacent planes in the hexahedron. The determining of the canopy heterogeneity index of the canopy space according to the ratio of the number of point cloud pixels in the projection image to the number of pixels in the projection image includes: for each of the projection images, calculate the ratio of the number of point cloud pixels to the number of pixels in the projection image to obtain the point cloud ratio corresponding to the projection image; calculate the product of the point cloud ratios corresponding to each projection image to obtain the canopy heterogeneity index of the canopy space.

[0087] Among them, the canopy space is a hexahedron, specifically a cuboid or a cube, and the embodiments of the present disclosure do not limit this.

[0088] As Figure 2As shown, define the XOY, XOZ, and YOZ planes as the three projection planes, project the canopy point cloud data onto the XOY, XOZ, and YOZ planes, and calculate the point cloud ratio of the number of point cloud pixels containing the point cloud data (PCD) to the total number of pixels in the projection image for each projection image, denoted as R XY 、R XZ and R YZ .

[0089] CHI is the product of these three point cloud ratios, expressed as:

[0090] CHI = R XY R XZ R YZ

[0091] For a uniform canopy, CHI is equal to 1, while for an aggregated canopy, CHI is less than 1.

[0092] S105. Modify the leaf area index under the assumption of a uniform canopy according to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space.

[0093] Specifically, use the ratio of the leaf area index under the assumption of a uniform canopy to the canopy heterogeneity index to replace the leaf area index in the quantitative relationship between the fractal dimension and the leaf area index of the uniform canopy. At the same time, since the fractal dimension is positively correlated with the leaf area index of the uniform canopy, after replacing the leaf area index in the quantitative relationship, it is equivalent to calculating the ratio of the leaf area index under the assumption of a uniform canopy to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space.

[0094] In the embodiments of the present disclosure, a quantitative relationship between the fractal dimension and the leaf area index of the uniform canopy in the canopy space is constructed based on the canopy point cloud data of the canopy space; the leaf area index of the canopy space under the assumption of a uniform canopy is calculated according to the canopy point cloud data and the quantitative relationship; the canopy point cloud data is projected onto multiple preset projection planes of the canopy space to obtain a projection image corresponding to each preset projection plane; the canopy heterogeneity index of the canopy space is determined according to the ratio of the number of point cloud pixels in the projection image to the number of pixels in the projection image, where the point cloud pixels are pixels containing the canopy point cloud; the leaf area index under the assumption of a uniform canopy is modified according to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space; by constructing a quantitative relationship between the fractal dimension and the leaf area index of the uniform canopy in the canopy space based on the point cloud data of the canopy space and defining the canopy heterogeneity index to modify the leaf area index of the uniform canopy assumption, the influence of leaf aggregation in three-dimensional space is considered, and the accuracy of leaf area index calculation is improved.

[0095] Based on the above embodiments, calculating the leaf area index of the canopy space under the assumption of a uniform canopy according to the canopy point cloud data and the quantitative relationship includes: calculating the fractal dimension of the canopy space according to the canopy point cloud data; calculating the leaf area index under the assumption of a uniform canopy based on the fractal dimension of the canopy space and the quantitative relationship.

[0096] Among them, calculating the fractal dimension of the canopy space according to the canopy point cloud data includes: determining a plurality of preset side length values; for each preset side length value, dividing the canopy space into a plurality of sample cubes with side lengths equal to the preset side length value, and counting the number of point cloud cubes containing the point cloud data among the plurality of sample cubes; performing a linear regression on the plurality of preset side length values and the natural logarithm values of the number of point cloud cubes corresponding to each preset side length value to obtain a linear regression equation; determining the absolute value of the slope of the linear regression equation as the fractal dimension of the canopy space.

[0097] The box-counting method (BCM) is often used to calculate the fractal dimension of an image because it is simple and easy to understand.

[0098] First, the three-dimensional canopy space defined by the canopy point cloud is uniformly divided into sample cubes with a side length of the preset side length value L, and the number of point cloud cubes containing the point cloud is counted as N. By changing the preset side length value L of the sample cube, the number of point cloud cubes N corresponding to each preset side length value L is obtained. By performing a linear regression on the natural logarithm values of these different Ls and their corresponding Ns, the linear regression equation is obtained:

[0099] ln(N) = -FD·ln(L) + b

[0100] Among them, the absolute value of the slope of the linear regression equation is the fractal dimension FD of the canopy space, and b is the intercept.

[0101] According to the linear regression equation, the differential solution form of the fractal dimension FD can be known as:

[0102]

[0103] Constructing the quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space based on the canopy point cloud data of the canopy space includes: dividing the canopy space into multiple sample cubes; calculating the maximum movement range when a single leaf intersects with a single sample cube; calculating the probability of intersection between the sample cube and the single leaf according to the maximum movement range and the volume of the sample cube; and constructing the quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space according to the leaf distribution characteristics in the canopy space and the probability of intersection between the sample cube and the single leaf.

[0104] If the probability of intersection between the sample cube and the leaf is P cl , then the number of point cloud cubes N is expressed as:

[0105] N = P cl ·N c

[0106] where N c is the number of sample cubes in the canopy space and can be expressed as:

[0107]

[0108] where V c is the volume of the canopy space.

[0109] Denote the probability of intersection between the sample cube and the single leaf as P cl-s , and denote the maximum movement range when a single leaf intersects with a single sample cube as V m , P cl-s can be expressed as:

[0110]

[0111] The following introduces the derivation process of P cl-s .

[0112] Figure 3 is the schematic diagram of the three-dimensional coordinate system space provided by the embodiment of the present disclosure. As Figure 3 shown, a three-dimensional coordinate system is established based on the sample cube, and the leaf azimuth angle and the leaf elevation angle θ are defined in the three-dimensional coordinate system. Define the positive direction of the X-axis as the 0° azimuth angle direction, and define the positive direction as the direction of clockwise rotation around the Z-axis. Therefore, the is 270°. Define the XOY plane as the plane corresponding to θ = 0°, and the positive rotation direction about the Z-axis is the positive direction of θ. Therefore, the θ values of the X-axis and Y-axis are both 0°, and the θ values of the positive and negative directions of the Z-axis are 90° and -90° respectively. Additionally, define the vertex E of the sample cube to be located at the origin O of the coordinate system, and the vertex H of the sample cube is on the negative X-axis direction, which is not shown in the figure.

[0113] In the embodiments of the present disclosure, a single leaf is fitted to a circular plane, and the leaf radius r is defined. r can be determined by on-site measurement or by selecting the leaf point cloud in the canopy point cloud data and fitting the leaf plane.

[0114] The coordinates of a circle in 3D space can be represented by the unit normal vector and two mutually perpendicular unit vectors that are also perpendicular to the normal vector ( and ). Based on the θ of the leaf, can be expressed as Define as and The dot product of is 0, indicating that and are mutually perpendicular. Then, by taking the cross product of and can be obtained That is, If the center of the leaf moves to a point P(p 1 , p 2 , p 3 ) in space, the coordinates of any point on the leaf circumference can be expressed as:

[0115]

[0116] where α defines the relative position of the point within the plane of the circle. The α of is 0°, and α increases counterclockwise within the plane formed by and until 2π.

[0117] Next, the volume of the movable range of the leaf when a single leaf slides along the surface of the sample cube and intersects the sample cube is derived. Taking the face ABFE of the sample cube as an example, when the leaf has an intersection with the face ABFE, the maximum distance between the center of the leaf and the face is the distance when the leaf is tangent to the face.

[0118] Figure 4 is a schematic diagram of the leaf tangent to the surface of the sample cube provided by the embodiments of the present disclosure. As shown in Figure 4 , the black ellipse represents the leaf, and the point M 1is the tangent point between the blade and the plane ABFE. At this time, the X-axis coordinates of all points on the blade circumference are not less than 0. Therefore, holds for any α belonging to [0, 2π]. Based on this, p 1 The minimum value of is Considering that M 1 can be any point on the plane ABFE, then when the blade remains intersecting with the plane ABFE, the movable range of the blade center is a parallelepiped with the plane ABFE as the base, and its volume is expressed as:

[0119]

[0120] Similarly, when the blade remains intersecting with the other five planes respectively, the moving ranges of the blade center also form five parallelepipeds. The total volume of the six parallelepipeds is:

[0121]

[0122] In addition to translating along the surface of the sample cube, the blade can also rotate around the edges of the cube. Figure 5 is a schematic diagram of the blade tangent to the edge of the cube provided by the embodiment of the present disclosure. As Figure 5 shown, taking the side AB of the sample cube as an example, the point M 2 is the tangent point between the blade and the side AB. Assuming the blade has angles and θ, when the blade center remains intersecting with the cube, its movable range is the dotted part shown in Figure 5 . P 1 and P 3 are the center points of the blade when the blade is tangent to the plane ABCD at points A and B respectively. P 2 and P 4 are the center points of the blade when the blade is tangent to the plane ABFE at points A and B respectively. It can be seen that P 1 , P 2 , P 3 , P 4 and points A and B form a part of an oblique cylinder.

[0123] Among them, the coordinates of P 1 are expressed as:

[0124]

[0125] The coordinates of P 2 are expressed as:

[0126]

[0127] On this basis, when considering edges CD, EF, and GH simultaneously, the movement range of the blade center can form a complete oblique cylinder. The base radius of this oblique cylinder is equal to the distance between points A and P 1 (or P 2 ), and the length of the side axis is L. The volume of the oblique cylinder is the product of the area of the base circle (i.e., the circle passing through points A, P 1 and P 2 ) and the distance (D oc ) from a certain point on the base circle (such as point B) to the other base circle. D oc is equal to the dot product of the normal vector 1 P 2 of plane AP and vector :

[0128] The volume of the oblique cylinder is:

[0129]

[0130] Similarly, when the blade rotates around the other eight edges of the sample cube and intersects with the cube, the movement range of the blade center forms two oblique cylinders. One oblique cylinder is formed based on edges AD, BC, FG, and EH, and the other oblique cylinder is formed based on edges AE, BF, CG, and DH. The volumes of these two oblique cylinders are respectively expressed as πr 2 Lsinθ, Therefore, when the blade rotates around the edge of the sample cube and intersects with the sample cube, the volume of the movable range of the blade center is:

[0131]

[0132] When the blade center is inside the cube, the blade always intersects with the cube. Therefore, while maintaining the intersection with the cube, the volume of the maximum movable range of the blade center is the volume of the cube, V m-faces , V m-edges The sum of the three, that is, the volume of the maximum movement range when the blade intersects with a single sample cube is expressed as:

[0133]

[0134] Based on this, the probability that the sample cube intersects with the single blade is substantially the ratio of the volume of the maximum movement range to the volume of the sample cube, expressed as:

[0135]

[0136] Furthermore, if n blades are randomly distributed in the canopy space and have the same θ and This means that each leaf has the same P cl-s . Then according to the Boolean model, P cl can be expressed as:

[0137] P cl = 1 - (1 - P cl-s ) n

[0138] In the real vegetation canopy space, the leaves have different θ and Therefore, different leaves should have different P cl-s values. If combined with the leaf angle distribution (LAD) function of the tree crown, calculate the average probability of the leaf intersecting with the sample cube Use this value to replace the P of all leaves with different θ and , that is: cl-s where

[0139]

[0140] is the LAD function of different leaf distribution characteristics. The leaf distribution characteristics include: spherical, uniform, planophile, erectophile, plagiophile, and extremophile.

[0141] Assume that the leaves are evenly distributed in the azimuth direction. Taking the tree crown with a spherical LAD as an example, its

[0142] calculation formula is: where S

[0143]

[0144] is the sample area corresponding to the canopy space; H p is the crown thickness, that is, the height difference between the highest point and the lowest point in the canopy point cloud data. c In summary, P

[0145] is a function of the leaf radius r, the cube side length L, and the number of leaves n in the canopy. For natural objects, FD is only stable within a certain scale range. In the embodiments of the present disclosure, preferably, when L is set to 4r, the FD accuracy is the highest. cl In addition, the uniform canopy leaf area index (hereinafter referred to as uniform LAI) can be expressed by r and n:

[0146]

[0147] ​

[0148] Combined with the following formula:

[0149]

[0150] N = P cl ·N c

[0151]

[0152] The quantitative relationship between the uniform canopy FD with a typical LAD and the uniform LAI can be obtained. Taking the uniform canopy with a spherical LAD as an example, the quantitative expression between FD and the uniform LAI is:

[0153]

[0154] Furthermore, project the canopy onto the XOY, XOZ, and YOZ planes with a resolution of 4r to obtain CHI.

[0155] Using the ratio of the uniform LAI to CHI to replace the uniform LAI in the above quantitative expression, the quantitative expression between FD and the uniform LAI is:

[0156]

[0157] In some embodiments, the method further includes: stratifying the canopy space according to a preset layer height to obtain a plurality of canopy sub-spaces; for each canopy sub-space, calculating the actual leaf area index of the canopy sub-space based on the point cloud data of the canopy sub-space corresponding to the canopy sub-space; using the actual leaf area index of the canopy sub-space as the index weight of the canopy sub-space; and calculating the stratified leaf area index of the canopy sub-space according to the index weight of the canopy sub-space and the actual leaf area index of the canopy space.

[0158] Optionally, based on the stratification height used in the Global Ecosystem Dynamics Investigation (GEDI), starting from 1.5 meters (the installation height of the TLS), the canopy space is stratified according to a preset layer height of 5 meters to obtain a plurality of canopy sub-spaces.

[0159] Furthermore, using the method described in the above embodiments (hereinafter referred to as FDTLS), according to the FD, r, and H of each layer c , CHI to calculate the actual leaf area index of this layer, and the actual leaf area index of the i-th layer is denoted as LAI_FDTLS id ; the actual leaf area index of the entire canopy space is denoted as LAI_FDTLS.

[0160] The actual leaf area index of each layer is used as the weight factor of that layer, and the actual leaf area index of the canopy space is distributed to each sub-canopy space of each layer, and finally the stratified leaf area index of each sub-canopy space is obtained:

[0161]

[0162] where m is the number of layers into which the canopy space is stratified.

[0163] It should be noted that the laser beam emitted by the TLS sensor has a certain divergence angle. As the propagation distance of the laser beam increases, the size of the light spot generated by the divergence angle also increases. When the center of the laser beam is not aligned with the leaf, due to the existence of the light spot, the laser beam will still generate an echo, which will lead to an overestimation of the leaf area. In the embodiments of the present disclosure, the method proposed by Mkaouar and Kallel is used to correct the influence of laser beam divergence on LAI estimation.

[0164] In the embodiments of the present disclosure, by introducing the fractal theory into the field of estimating vegetation leaf area index based on ground-based lidar point cloud, through mathematical derivation, a universal relationship between the fractal dimension of the canopy in three-dimensional space and the canopy leaf area index is established, that is, the quantitative relationship between the fractal dimension in the canopy space and the leaf area index under the assumption of a uniform canopy, and the canopy heterogeneity index is defined to correct the leaf area index under the assumption of a uniform canopy, considering the influence of leaf aggregation in three-dimensional space, and improving the accuracy of leaf area index calculation.

[0165] The accuracy test data of the leaf area index calculation method provided by the present disclosure are given below.

[0166] The large-scale emulation system (LESS) model is used to simulate point cloud data, and the pre-configured RIEGL VZ-400 scanner in LESS is selected. The scanning parameters mainly include the horizontal position of the scanner, the vertical height of the scanner, the scanning range and resolution in the azimuth and zenith directions, and the minimum and maximum scanning distances. The specific parameters are shown in the following table:

[0167] Table 1 Scanning parameter table for simulating point cloud

[0168] Scanning parameters Value Scanner installation height (m) 1.5 Zenith angle range (degrees) 0-90 Azimuth angle range (degrees) 0-360 Scanning resolution in the zenith direction (degrees) 0.06 Scanning resolution in the azimuth direction (degrees) 0.06 Nearest scanning distance (m) 2 Farthest scanning distance (m) 50

[0169] Figure 6 is a schematic diagram of the TLS site settings provided for the embodiments of the present disclosure. As Figure 6 shown, for the uniform plot and the Radiation transfer Model Intercomparison (RAMI) sub-plot, the scanning stations are arranged respectively.

[0170] As Figure 6 shown, nine scanning stations were set up for each quadrat, and the black rectangles represent the boundaries of the two types of quadrats. The area of the homogeneous quadrat is 10 m × 10 m; the area of the RAMI sub-plot is 30 m × 30 m.

[0171] In LESS, a 3D object creation module was used to generate ten homogeneous tree crowns with circular leaves, and the size of each tree crown is 10 m × 10 m. Among these ten homogeneous tree crowns, three have leaf radii of 2 cm, four have leaf radii of 5 cm, and three have leaf radii of 8 cm. The leaf area index (LAI) values of the plots with leaf radii of 2 cm are 0.5, 1.0, and 1.5 respectively; the LAI values of the plots with leaf radii of 5 cm are 0.5, 1.5, 2.0, and 3.0; the LAI values of the plots with leaf radii of 8 cm are 0.5, 1.0, and 1.5. The thickness range of these tree crowns is from 1 m to 5 m. The positions of the homogeneous tree crowns are 7 m or 8 m above the ground.

[0172] Two RAMI quadrats, namely (JBS) and Wytham Wood (WWO), were used to verify the method. The JBS and WWO quadrats were constructed based on field survey data and high-precision TLS point cloud data, and the PAI values are 5.22 and 9.27 respectively. The sizes of the JBS and WWO plots are 106.15 m × 105.51 m and 100 m × 100 m respectively. The range of 90 m × 90 m from the upper left corner of the JBS and WWO plots was divided into nine sub-plots, and the size of each sub-plot is 30 m × 30 m. The scanning settings of each sub-plot are as Figure 6 shown in the schematic diagram of the TLS site settings for the RAMI sub-plot.

[0173] The coordinates of the simulated point clouds are all based on the same coordinate origin. Therefore, the multi-site cloud data simulated by LESS can be directly stitched without complex processing. After stitching, the point clouds beyond the boundaries of each sub-quadrat are cropped. The FDTLS and PATHTLS methods estimate PAI based on the processed point cloud data respectively.

[0174] In addition to the simulated point cloud data, fisheye images of each sub-quadrat were also simulated. Comparative methods such as CC, CLX, PATH, and 1DFD estimate the PAI of the sub-quadrat based on the fisheye images respectively. The mounting height of the fisheye camera is fixed at 1.5 m, which is the same as the mounting height of the TLS. The sampling points of the fisheye images are evenly distributed within the sub-quadrat. Figure 7 This is an example diagram of the distribution of sampling points of the fisheye image within the sub-quadrat in the embodiment of the present disclosure.

[0175] The resolution of the fisheye image is 4000×4000 pixels, and it adopts equal solid projection and a field of view angle of 180°. To better distinguish the vegetation pixels and sky pixels in the simulated fisheye image, the reflectivity and transmittance of leaves and branches are set to 0 in LESS, and the sky is set to a white sky. The PAI of the sub-plot is the average value of the PAI estimated from nine images taken within the sub-plot.

[0176] To mitigate the problem of incomplete canopy point clouds caused by the mutual occlusion of branches and leaves, multiple TLS scans were conducted at different positions within the plot. However, the point clouds of the upper canopy will still inevitably be missing. To analyze the impact of the incomplete upper canopy point clouds on the accuracy of the FDTLS method, additional scans were carried out. In addition to the conventional upward scan from below the canopy, the TLS was also set to scan downward from above the canopy. Except for setting the height (adjusted to 35 meters) and the scanning zenith angle range (adjusted to 0° to -90°, i.e., the lower hemisphere space), the other parameters were the same as those used in the upward scan (see Table 1). Subsequently, the point clouds from the upward and downward scans were stitched together to obtain a complete canopy point cloud. Then, the FDTLS method was used to estimate the PAI of the canopy based on the complete point cloud.

[0177] To analyze the impact of different L on the LAI estimation during the FD calculation, L was set to 3r, 4r, and 5r respectively. Figure 8 The schematic diagram of the LAI results of the uniform plot estimated in the embodiments of the present disclosure. The results show that when the side length of the cube is set to 3 times (i.e., K = 3), 4 times (i.e., K = 4), and 5 times (i.e., K = 5) the leaf radius during the FD calculation, the LAI estimated using the FDTLS method is highly consistent with the true LAI (LAIt) of the plot when L is set to 3r, 4r, or 5r. In contrast, when L = 4r, the RMSE is the lowest (0.18), and the R 2 is the highest (0.98). Therefore, preferably, L = 4r is set when using the FDTLS method.

[0178] Figure 9 The schematic diagram of the PAI results of the RAMI sub-plot estimated using six comparison methods in the embodiments of the present disclosure. The horizontal axis represents the true PAI (PAIt) value calculated by LESS. The circles represent the results of 9 JBS sub-plots, and the squares represent the results of 9 WWO sub-plots. The black dashed line represents the linear regression, and the solid line represents the 1:1 line. Among them, the six comparison methods include CC, CLX, PATH, 1DFD, PATHTLS, and FDTLS. As Figure 9As shown, the PAI estimated based on fisheye images is underestimated, especially in the 9 subplots of the WWO scenario. In the WWO scenario, the canopy is dense, with a high LAI and a low gap fraction (<0.10), which makes it difficult for the estimation method using two-dimensional projection information based on images to accurately correct the influence of the clumping effect. In contrast, the PAI estimated by the PATH method is in the best agreement with the true PAI, followed by the 1DFD method and the CLX method, while the CC method shows the most serious underestimation (RMSE = 2.72, RRMSE = 36.27%).

[0179] The PAI estimated based on point cloud data is more accurate than that estimated based on fisheye images. The basic principles of the PATH and PATHTLS methods are the same, both by correcting the clumping effect caused by inconsistent path lengths in the canopy. However, the ways they extract the path lengths are different. The PATH method estimates the path length distribution by segmenting the sub-sampling lines and based on the gaps of these sub-sampling lines. While the PATHTLS method extracts the path length distribution by reconstructing the outer envelope based on the point cloud data. Compared with estimating the path length distribution from the sampling line data, the path length distribution can be more accurately extracted from the point cloud data. Therefore, the PAI estimated by the PATHTLS method based on the point cloud is more consistent with the PAIt (RMSE = 1.34, RRMSE = 17.87%), while the PAI estimated by the PATH method based on the fisheye image (RMSE = 1.56, RRMSE = 20.80%) is slightly less accurate. Although both the 1DFD method and the FDTLS method calculate the FD of the canopy based on field measurement data (the 1DFD method uses 2D images and 1D sampling lines; the FDTLS method uses 3D point clouds), there are significant differences in principle. For example, the 1DFD method relies on Beer-Lambertian and requires knowledge of the leaf projection function G and the gap fraction, while the FDTLS method does not require this information. The PAI estimated by the FDTLS method is more accurate than that by the 1DFD method. The RMSE and RRMSE of the 1DFD method and the FDTLS method are 1.77 and 23.60%, and 1.14 and 15.20% respectively.

[0180] Figure 10 It is a schematic diagram of the PAI results of the RAMI subplots estimated based on the complete canopy point cloud, where the circles represent the results of the JBS subplots and the squares represent the results of the WWO subplots. Although multi-station scanning within the subplots reduces the missing point clouds, the upper canopy point clouds are still incomplete. Therefore, by simulating the point clouds scanned from above the canopy and stitching them with the traditional point clouds of the canopy scanned from below, complete canopy point clouds are obtained. Based on this complete canopy point cloud, the results of the canopy PAI estimated using the FDTLS method are as Figure 10 shown. Compared with Figure 9Compared with the estimation results of the FDTLS method, the PAI estimated based on the complete point cloud did not show systematic underestimation (especially for the sub-plots in the WWO scenario, i.e., Figure 10 the PAI values represented by squares in

[0181] Figure 11 ), and was closer to the true PAI of the plot (RMSE = 0.85, RRMSE = 11.33%). Figure 11 Figure 6 shows the comparison results between the average stratified PAI calculated using the FDTLS method and the accurate average stratified PAI calculated using LESS. To further analyze the impact of the missing point cloud in the upper canopy, the average stratified PAI (PAIt, black square line) of 9 sub-plots in the WWO and JBS scenarios was calculated, as Figure 11 shown, when estimating the stratified LAI using only the point cloud scanned from below the tree canopy upward, the missing point cloud mainly led to the underestimation of the stratified PAI in the upper canopy ( Figure 11 the black circle line in

[0182] ). For the WWO scenario, the stratified PAI between 16.5 m and 21.5 m estimated using the FDTLS method was significantly lower than the true PAI (PAIt), with an absolute deviation of 1.27; for the JBS scenario, the stratified PAI between 21.5 m and 26.5 m was also lower than the true value, with an absolute deviation of 0.59. Using the complete point cloud to estimate the stratified PAI effectively alleviated the underestimation problem, as

[0183] Figure 12 shown by the black triangle line in Figure 12 . For the WWO scenario, the absolute deviation of the stratified PAI between 16.5 m and 21.5 m estimated by the FDTLS method decreased from 1.27 to 0.15; for the JBS scenario, the absolute deviation between 21.5 m and 26.5 m decreased from 0.59 to 0.32. Therefore, in a dense forest canopy, if some methods are adopted to supplement the missing point cloud in the upper canopy, such as fusing the canopy point cloud obtained by unmanned aerial vehicle laser scanning with the point cloud obtained by TLS scanning, the accuracy of estimating PAI or stratified PAI using the FDTLS method can be significantly improved, especially for the PAI in the upper canopy. It can be seen that compared with the existing methods, the leaf area index calculation method provided by the embodiments of the present disclosure can calculate the LAI and stratified LAI at the plot scale more accurately and is insensitive to the leaf inclination distribution. Figure 7 is a schematic structural diagram of the leaf area index calculation device provided by the embodiments of the present disclosure. The leaf area index calculation device provided by the embodiments of the present disclosure can execute the processing flow provided by the embodiments of the leaf area index calculation method, as Figure 12As shown in the figure, the leaf area index calculation device 120 includes: a construction module 121, a first calculation module 122, a projection module 123, a determination module 124, and a correction module 125; the construction module 121 is used to construct a quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space based on the canopy point cloud data in the canopy space; the first calculation module 122 is used to calculate the leaf area index of the canopy space under the assumption of a uniform canopy according to the canopy point cloud data and the quantitative relationship; the projection module 123 is used to project the canopy point cloud data onto a plurality of preset projection planes in the canopy space to obtain a projection image corresponding to each preset projection plane; the determination module 124 is used to determine the canopy heterogeneity index of the canopy space according to the ratio of the number of point cloud pixels to the number of pixels in the projection image in the projection image, where the point cloud pixels are pixels containing canopy point cloud; the correction module 125 is used to correct the leaf area index under the assumption of a uniform canopy according to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space.

[0184] Optionally, the construction module 121 includes a division unit 1211, a first calculation unit 1212, a second calculation unit 1213, and a construction unit 1214; the division unit 1211 is used to divide the canopy space into a plurality of sample cubes; the first calculation unit 1212 is used to calculate the maximum movement range when a single leaf intersects with a single sample cube; the second calculation unit 1213 is used to calculate the probability of intersection between the sample cube and the single leaf according to the maximum movement range and the volume of the sample cube; the construction unit 1214 is used to construct a quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space according to the leaf distribution characteristics in the canopy space and the probability of intersection between the sample cube and the single leaf.

[0185] Optionally, the canopy space is a hexahedron, and the preset projection planes are three pairwise adjacent planes in the hexahedron; the determination module 124 includes a third calculation unit 1241 and a fourth calculation unit 1242; the third calculation unit 1241 is used to calculate the ratio of the number of point cloud pixels to the number of pixels in the projection image for each projection image to obtain the point cloud ratio corresponding to the projection image; calculate the product of the point cloud ratios corresponding to each projection image to obtain the canopy heterogeneity index of the canopy space.

[0186] Optionally, the correction module 125 is specifically used to calculate the ratio of the leaf area index under the assumption of a uniform canopy to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space.

[0187] Optionally, the first calculation module 122 includes a fifth calculation unit 1221 and a sixth calculation unit 1222; the fifth calculation unit 1221 is configured to calculate the fractal dimension of the canopy space according to the canopy point cloud data; the sixth calculation unit 1222 is configured to calculate the leaf area index based on the uniform canopy assumption based on the fractal dimension of the canopy space and the quantitative relationship.

[0188] Optionally, the fifth calculation unit 1221 is specifically configured to determine a plurality of preset side length values; for each preset side length value, divide the canopy space into a plurality of sample cubes with side lengths equal to the preset side length value, and count the number of point cloud cubes containing the point cloud data among the plurality of sample cubes; perform a linear regression on the plurality of preset side length values and the natural logarithm values of the number of point cloud cubes corresponding to each preset side length value to obtain a linear regression equation; determine the absolute value of the slope of the linear regression equation as the fractal dimension of the canopy space.

[0189] Optionally, the leaf area index calculation device 120 includes a second calculation module 126, configured to layer the canopy space according to a preset layer height to obtain a plurality of canopy sub-spaces; for each canopy sub-space, calculate the actual leaf area index of the canopy sub-space based on the canopy sub-space point cloud data corresponding to the canopy sub-space; use the actual leaf area index of the canopy sub-space as the index weight of the canopy sub-space; calculate the stratified leaf area index of the canopy sub-space according to the index weight of the canopy sub-space and the actual leaf area index of the canopy space.

[0190] Figure 12 The leaf area index calculation device of the illustrated embodiment can be used to execute the technical solutions of the above method embodiments, and its implementation principle and technical effects are similar, which will not be elaborated here.

[0191] Figure 13 It is a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure. The electronic device provided by the embodiment of the present disclosure can execute the processing flow provided by the leaf area index calculation method embodiment, as Figure 13 shown, the electronic device 130 includes: a memory 131, a processor 132, a computer program, and a communication interface 133; wherein, the computer program is stored in the memory 131 and is configured to be executed by the processor 132 to perform the leaf area index calculation method as described above.

[0192] In addition, an embodiment of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored, and the computer program is executed by a processor to implement the leaf area index calculation method described in the above embodiments.

[0193] In addition, an embodiment of the present disclosure further provides a computer program product, which includes a computer program or instruction. When the computer program or instruction is executed by a processor, the above-mentioned leaf area index calculation method is implemented.

[0194] It should be noted that, in this document, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the existence of additional identical elements in the process, method, article or device comprising the element.

[0195] The above are only specific embodiments of the present disclosure, enabling those skilled in the art to understand or implement the present disclosure. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to these embodiments described herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for calculating leaf area index, characterized in that: The method comprises: constructing a quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space based on canopy point cloud data in the canopy space; Calculate the leaf area index of the canopy space based on a uniform canopy assumption according to the canopy point cloud data and the quantitative relationship; Projecting the canopy point cloud data onto a plurality of preset projection planes in the canopy space to obtain a projection image corresponding to each preset projection plane; Determining a canopy heterogeneity index of the canopy space according to a ratio of the number of point cloud pixels in the projection image to the number of pixels in the projection image, wherein the point cloud pixels are pixels that include a canopy point cloud; The leaf area index based on the uniform canopy assumption is corrected according to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space.

2. The method according to claim 1, characterized in that The method of constructing a quantitative relationship between the fractal dimension and the uniform canopy leaf area index in the canopy space based on the canopy point cloud data in the canopy space includes: dividing the canopy space into a plurality of sample cubes; Calculating the maximum motion range of a single blade when it intersects with a single sample cube; Calculating the probability that the sample cube intersects the single leaf according to the maximum motion range and the volume of the sample cube; According to the leaf distribution characteristics in the canopy space and the probability of the sample cube intersecting with the single leaf, a quantitative relationship between the fractal dimension in the canopy space and the uniform canopy leaf area index is constructed.

3. The method according to claim 1, characterized in that The canopy space is a hexahedron, and the preset projection planes are three planes adjacent to each other in the hexahedron; determining the canopy heterogeneity index of the canopy space according to the ratio of the number of point cloud pixels in the projection image to the number of pixels in the projection image includes: For each of the projection images, calculating the ratio of the number of pixels of the point cloud to the number of pixels of the projection image, and obtaining the point cloud ratio corresponding to the projection image; The product of the point cloud proportions corresponding to each projection image is calculated to obtain the canopy heterogeneity index of the canopy space.

4. The method according to claim 1, characterized in that The step of correcting the leaf area index based on the uniform canopy assumption according to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space includes: The ratio of the leaf area index based on the uniform canopy assumption to the canopy heterogeneity index is calculated to obtain the actual leaf area index of the canopy space.

5. The method according to claim 1, characterized in that The calculating the leaf area index of the canopy space based on the uniform canopy assumption according to the canopy point cloud data and the quantitative relationship comprises: Calculating the fractal dimension of the canopy space according to the canopy point cloud data; The leaf area index based on the uniform canopy assumption is calculated based on the fractal dimension of the canopy space and the quantitative relationship.

6. The method according to claim 5, characterized in that The step of calculating the fractal dimension of the canopy space based on the canopy point cloud data comprises: Determine multiple preset side length values; For each preset side length value, the canopy space is divided into a plurality of sample cubes with side lengths of the preset side length values, and the number of point cloud cubes containing the point cloud data in the plurality of sample cubes is counted; Performing linear regression on the multiple preset side length values ​​and the natural logarithm of the number of point cloud cubes corresponding to each of the preset side length values ​​to obtain a linear regression equation; The absolute value of the slope of the linear regression equation is determined as the fractal dimension of the canopy space.

7. The method according to claim 1, characterized in that The method further comprises: The canopy space is layered according to preset layer heights to obtain a plurality of canopy subspaces; For each canopy subspace, calculating the actual leaf area index of the canopy subspace based on the canopy subspace point cloud data corresponding to the canopy subspace; Using the actual leaf area index of the canopy subspace as the index weight of the canopy subspace; The layered leaf area index of the canopy subspace is calculated according to the index weight of the canopy subspace and the actual leaf area index of the canopy space.

8. A leaf area index calculation device, characterized in that: include: A construction module for constructing a quantitative relationship between a fractal dimension and a uniform canopy leaf area index in the canopy space based on canopy point cloud data in the canopy space; A first calculation module, configured to calculate the leaf area index of the canopy space based on a uniform canopy assumption according to the canopy point cloud data and the quantitative relationship; A projection module, used for projecting the canopy point cloud data onto a plurality of preset projection planes in the canopy space to obtain a projection image corresponding to each preset projection plane; A determination module, configured to determine a canopy heterogeneity index of the canopy space according to a ratio of the number of point cloud pixels in the projection image to the number of pixels in the projection image, wherein the point cloud pixels are pixels containing a canopy point cloud; The correction module is used to correct the leaf area index based on the uniform canopy assumption according to the canopy heterogeneity index to obtain the actual leaf area index of the canopy space.

9. An electronic device, characterized in that: include: Memory; processor; as well as Computer programs; The computer program is stored in the memory and is configured to be executed by the processor to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images