Vegetation canopy structure complexity evaluation method based on multi-dimensional and multi-layer canopy entropy

Abstract

The invention relates to the technical field of laser radar point cloud application, in particular to a vegetation canopy structure complexity evaluation method based on multi-dimensional and multi-layer canopy entropies, and the vegetation canopy structure complexity is quantitatively evaluated according to values of the multi-dimensional and multi-layer canopy entropies provided by the invention. The method provided by the invention is a simple and effective vegetation canopy structure complexity measuring method, is suitable for LiDAR point cloud data acquired by multiple platforms such as manned aerial vehicles, unmanned aerial vehicles and foundations, and is also suitable for evaluating the canopy structure complexity of various vegetation types such as forests and crops.

Description

Technical Field

[0001] This application relates to the field of lidar point cloud application technology, specifically to a method for evaluating the complexity of vegetation canopy structure based on multi-dimensional and multi-layer canopy entropy. Background Technology

[0002] Vegetation is a crucial component of terrestrial ecosystems, and the canopy structure constitutes the top-level spatial composition of plant communities. Canopy structural complexity is a comprehensive parameter describing the spatial distribution of branches and leaves within the canopy, also known as canopy structure diversity or complexity. It plays a vital role in regulating ecosystem functions, particularly in light and water resources, microclimate, forest productivity, and ecosystem stability. Specifically, the distribution of branches and leaves in the canopy directly affects the intensity and ratio of direct and diffused light, thus influencing photosynthetic efficiency; the canopy spatially redistributes precipitation by intercepting and diverting snowfall; and canopy structure influences forest microclimate by regulating the distribution of light and water resources. Different canopy structures exhibit significantly different mechanisms of influence on microclimate. Horizontal structures primarily affect nighttime radiation through radiation absorption and evapotranspiration, while vertical structures reduce wind speed within the forest by altering radiation flux and airflow. The impact of forest canopy structure on productivity and carbon sequestration capacity has also received widespread attention. Studies have found that complex canopy structures help improve light capture, thereby promoting productivity. In addition, forest stability is an important indicator of an ecosystem's ability to resist external disturbances and recover. Research shows that forests with complex structures perform better in resisting disturbances and recovering. Complex canopy structures can enhance forests' resistance to forest fires and hurricanes and accelerate the recovery process after disturbances.

[0003] The complexity of canopy structures is typically evaluated quantitatively using the following indicators: (1) Horizontal distribution index: The heterogeneity of canopy elements in the horizontal direction is measured by statistical parameters of tree diameter at breast height (such as average diameter at breast height, standard deviation of diameter at breast height, and diameter at breast height diversity index).

[0004] (2) Vertical distribution index: mainly tree height statistical parameters (such as tree height mean, standard deviation, leaf height diversity index) to reflect the distribution characteristics of canopy elements in the vertical direction.

[0005] (3) Integrated distribution index: Simultaneously consider the distribution heterogeneity in both horizontal and vertical directions. Representative indicators include canopy structure complexity based on the product of tree height, chest height sectional area, tree density, and number of canopy species; canopy structure complexity index based on the spatial horizontal location and tree height; canopy wrinkle degree; fractal dimension; integrated fractal dimension and effective number of canopy layers (SSCI); and canopy entropy applicable to multiple lidar platforms.

[0006] With the development of LiDAR technology, it has become possible to rapidly acquire three-dimensional structural information of forest canopies. However, domestic research in this area is relatively limited and still in its early stages. Canopy entropy is an existing index for evaluating the complexity of forest canopy structure based on LiDAR point cloud data. This method calculates an entropy value to measure canopy complexity by resampling LiDAR point cloud data and estimating kernel density. However, this method is complex, computationally intensive, and has specific requirements for point cloud density; theoretically, it is only applicable to ground-based LiDAR point clouds. Summary of the Invention

[0007] This application provides a method for evaluating the complexity of vegetation canopy structure based on multidimensional and multi-layer canopy entropy, in order to solve or partially solve the problems mentioned in the background art.

[0008] This application provides a method for evaluating the complexity of vegetation canopy structure based on multidimensional and multi-layer canopy entropy, including the following steps: The complexity of vegetation canopy structure is quantitatively evaluated using the values ​​of multidimensional and multi-layer canopy entropy; The calculation method for the multidimensional, multi-layer canopy entropy is as follows: For the point cloud data of the vegetation canopy in the target area, let the point cloud data contain N points. Divide the point cloud data into segments along the x, y, and z directions at certain intervals h. , , Layer, a total of Point cloud data of the layer; Then, the number of points contained in each layer in the x, y, and z directions are counted respectively. The number of points in each layer in the X direction is recorded as follows: ; The number of points in each layer along the y-direction is denoted as: ; The number of points in each layer along the z-direction is denoted as: ; The formulas for calculating multidimensional and multi-layer canopy entropy are as follows: (1) (2) (3) (4) In the formula, MDMLCE represents the multidimensional, multilayer canopy entropy of the target region. Let be the entropy along the x-axis. Let be the entropy in the y-axis direction. Let be the entropy along the z-axis, and let i be an intermediate variable whose values ​​range from 0 to 1 when calculating the x-axis, y-axis, and z-axis directions, respectively. 0 to 0 to .

[0009] Preferably, the method for calculating the multidimensional, multi-layer canopy entropy further includes the following steps: S1: Acquire LiDAR point cloud data for the target area; S2: Classify LiDAR point cloud data to obtain high vegetation point cloud data; S3: Calculate the multidimensional and multi-layer canopy entropy of the target area based on the high vegetation point cloud data.

[0010] Preferably, in step S1, the point cloud data of the target area includes one or more combinations of LiDAR point cloud data collected by multiple platforms such as manned airborne LiDAR, unmanned airborne LiDAR, and ground-based LiDAR. In step S1, the acquired LiDAR point cloud data is preprocessed by using cropping and block operations to obtain LiDAR point cloud data that accurately matches the boundary of the target area.

[0011] Preferably, in step S2, the specific steps for classifying LiDAR point cloud data and obtaining high-vegetation point cloud data are as follows: S201: Ground point identification from LiDAR point cloud data; S202: Construct a normalized digital surface model (nDSM) of the target area based on the identified ground points; S203: Classify the points in the point cloud based on the nDSM values ​​of each point in the point cloud data, obtain high vegetation points, and then generate a high vegetation point set; S204: Replace the original z-values ​​of high-vegetation points with nDSM values ​​to generate the final high-vegetation point cloud.

[0012] Preferably, in step S201, the classic irregular triangular mesh progressive encryption filtering algorithm is used to identify ground points in the point cloud.

[0013] Preferably, in step S202, the method for constructing the normalized digital surface model (nDSM) is as follows: Using the horizontal coordinates of ground points, an irregular triangular network (TIN) is constructed to generate a digital elevation model (DEM) of the survey area based on the TIN representation. For any non-ground point, find the triangle that the point falls into, and use bilinear interpolation to obtain the DEM value corresponding to the point; By subtracting the DEM value from the Z value at that point, the corresponding nDSM value can be obtained, such as... Figure 5 The image shown is a rendering of nDSM.

[0014] Preferably, in step S203, the specific method for obtaining high vegetation points and generating a set of high vegetation points is as follows: Elevation thresholds are set based on the size of the nDSM value. h, nDSM value greater than Point h is determined to be a high vegetation point; Based on the characteristic that building points have a very high probability of belonging to a certain plane, we further distinguish between high vegetation points and building points in the point cloud, and save the high vegetation point set as a new point cloud.

[0015] Preferably, in step S3, the specific method for calculating the multidimensional and multi-layer canopy entropy of the target area based on the high vegetation point cloud data is as follows: S301: Obtain the 3D bounding box of the point cloud of high vegetation in the target area; S302: Calculate the entropy in the z-axis direction; S303: Calculate the entropy in the x-axis direction; S304: Calculate the entropy in the y-axis direction; S305: Calculate the multidimensional, multilayer canopy entropy.

[0016] Preferably, in step S301, the specific method for obtaining the 3D bounding box of the high-vegetation point cloud of the target area is as follows: Iterate through every point in the high-vegetation point cloud and find the minimum value along the x-axis. Maximum value Minimum value along the y-axis Maximum value minimum value along the z-axis Maximum value The three-dimensional coordinates of the lower left corner of the three-dimensional bounding box of the high vegetation point cloud to be determined ( ), the three-dimensional coordinates of the upper right corner ( ).

[0017] Preferably, in step S302, the specific method for calculating the entropy in the z-axis direction is as follows: Vertical layering is performed on the high-vegetation point cloud, and entropy is calculated, such as... Figure 7 As shown.

[0018] First, determine the number of vertical layers. The calculation formula is as follows: , Secondly, calculate the number of points in each layer in the vertical direction. Calculation formula: , Next, the proportion of points contained in each layer in the vertical direction to the total number of points contained in the entire high-vegetation point cloud is calculated. The calculation formula is as follows: , Finally, calculate the entropy in the z-axis direction. The calculation formula is shown in formula (4). Based on the method and formulas (2) and (3) for calculating the entropy in the z-axis direction, the entropy in the x and y-axis directions is calculated respectively, and then the multidimensional and multi-layer canopy entropy is calculated according to formula (4).

[0019] Compared with the prior art, the beneficial effects of this application are as follows: The method proposed in this application is a simple and effective way to measure the complexity of vegetation canopy structure. It is applicable to LiDAR point cloud data acquired by multiple platforms such as manned aircraft, unmanned aircraft, and ground-based systems, and is also suitable for evaluating the complexity of canopy structure of various vegetation types such as forests and crops. Attached Figure Description

[0020] The present application will be further described below with reference to the accompanying drawings and embodiments.

[0021] Figure 1 This is an example of point cloud data obtained by the method described in this application. Figure 2 The target area LiDAR point cloud after cropping in this application, Figure 3 This is a rendering of the ground points in the target area of ​​this application and the generated digital elevation model (DEM). Figure 4 This is a point cloud classification effect diagram for the target area of ​​this application. Figure 5 This is a rendering of the nDSM for the target region of this application. Figure 6 This is a schematic diagram of the point cloud effect for high vegetation in the target area of ​​this application. Figure 7 This is a layered effect diagram along the z-axis of this application. Figure 8 This is a layered effect diagram along the x-axis of this application (only the point clouds of odd-numbered layers are shown to facilitate the demonstration of the differences between the data of each layer). Figure 9This is a layered effect diagram along the y-axis of this application (only the point clouds of even-numbered layers are shown to facilitate the demonstration of the differences between the data of each layer). Figure 10 Manned airborne LiDAR point cloud data for forest areas in Toyama Prefecture, Japan. Figure 11 This is a thematic map of multidimensional and multi-layered canopy entropy in the forest region of Toyama Prefecture, Japan. Figure 12 This is a schematic diagram of the vegetation at a location in the forest area of ​​Toyama Prefecture, Japan, where the multidimensional and multi-layered canopy entropy equals 6.8458. Figure 13 This is a schematic diagram of vegetation at a location in the forest area of ​​Toyama Prefecture, Japan, where the multidimensional and multi-layered canopy entropy equals 6.2787. Figure 14 This is a schematic diagram of the vegetation at a location in the forest area of ​​Toyama Prefecture, Japan, where the multidimensional and multi-layered canopy entropy equals 5.8851. Figure 15 This refers to airborne LiDAR point cloud data from drones in the Huangshan forest area of ​​Anhui Province. Figure 16 This is a thematic map of multidimensional and multi-layered canopy entropy in the Huangshan forest area of ​​Anhui Province. Figure 17 This is a schematic diagram of vegetation at a location in the Huangshan forest area of ​​Anhui Province where the multidimensional and multi-layered canopy entropy equals 6.3524. Figure 18 This is a schematic diagram of the vegetation at a location in the Huangshan forest area of ​​Anhui Province where the multidimensional and multi-layered canopy entropy equals 5.9740. Figure 19 This is a schematic diagram of the vegetation at a location in the Huangshan forest area of ​​Anhui Province where the multidimensional and multi-layered canopy entropy equals 5.7850. Figure 20 For ground-based LiDAR point cloud data of forest areas, Figure 21 The multidimensional, multi-layered canopy entropy of 8.3526 is used to illustrate the forest region scene. Figure 22 For drone-borne LiDAR point cloud data of crop (flue-cured tobacco) planting areas, Figure 23 A multi-dimensional, multi-layered thematic map of canopy entropy for crop (flue-cured tobacco) planting areas. Figure 24 This is a schematic diagram of the vegetation at a location in a crop (flue-cured tobacco) planting area where the multidimensional, multi-layered canopy entropy equals 8.5445. Figure 25 A schematic diagram of vegetation at a location in a crop (flue-cured tobacco) planting area where the multidimensional, multi-layered canopy entropy equals 8.3554. Figure 26 A schematic diagram of vegetation at a location in a crop (flue-cured tobacco) planting area where the multidimensional, multi-layered canopy entropy equals 8.008. Detailed Implementation

[0022] The specification and claims use certain terms to refer to specific components. Those skilled in the art will understand that hardware manufacturers may use different names to refer to the same component. This specification and claims do not distinguish components based on differences in name, but rather on differences in function. The term "comprising" throughout the specification and claims is an open-ended term and should be interpreted as "comprising but not limited to." "Approximately" means that within an acceptable margin of error, those skilled in the art can solve the technical problem and substantially achieve the technical effect within a certain margin of error.

[0023] In the description of this application, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "horizontal", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0024] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0025] The spatial distribution of branches and leaves within the vegetation canopy needs to consider both horizontal and vertical distribution, i.e., three-dimensional spatial distribution. The more complex the vegetation canopy, the more homogeneous the branches and leaves will be in both the horizontal and vertical directions. The degree of homogeneity can be measured by entropy, a core concept in information theory used to quantify the uncertainty of information. That is, entropy is an indicator of the uncertainty of a random variable. According to entropy theory, entropy reaches its maximum value when all possible values ​​of a random variable are uniformly distributed. Therefore, entropy can be used to measure the structural complexity of the vegetation canopy.

[0026] This application provides a method for evaluating the complexity of vegetation canopy structure based on multidimensional and multi-layer canopy entropy, which quantitatively evaluates the complexity of vegetation canopy structure using the values ​​of multidimensional and multi-layer canopy entropy.

[0027] Specifically, the calculation method for the multidimensional, multi-layer canopy entropy is as follows: For the point cloud data of the vegetation canopy in the target area, let the point cloud data contain N points. Divide the point cloud data into segments along the x, y, and z directions at certain intervals h. , , Layer, a total of Point cloud data of the layer; Then, the number of points contained in each layer in the x, y, and z directions are counted respectively. The number of points in each layer in the X direction is recorded as follows: ; The number of points in each layer along the y-direction is denoted as: ; The number of points in each layer along the z-direction is denoted as: ; The formulas for calculating multidimensional and multi-layer canopy entropy are as follows: (1) (2) (3) (4) In the formula, MDMLCE represents the multidimensional, multilayer canopy entropy of the target region. Let be the entropy along the x-axis. Let be the entropy in the y-axis direction. Let be the entropy along the z-axis, and let i be an intermediate variable whose values ​​range from 0 to 1 when calculating the x-axis, y-axis, and z-axis directions, respectively. 0 to 0 to .

[0028] The higher the value of the multidimensional and multi-layer canopy entropy, the higher the complexity of the vegetation canopy structure in the corresponding target area.

[0029] Preferably, the method for calculating the multidimensional, multi-layer canopy entropy further includes the following steps: S1: Acquire LiDAR point cloud data for the target area; S2: Classify LiDAR point cloud data to obtain high vegetation point cloud data; S3: Calculate the multidimensional and multi-layer canopy entropy of the target area based on the high vegetation point cloud data.

[0030] Specifically, in step S1, the point cloud data of the target area includes one or more combinations of LiDAR point cloud data collected from multiple platforms such as manned airborne LiDAR, unmanned aerial vehicle airborne LiDAR, and ground-based LiDAR.

[0031] Specifically, in step S1, the acquired LiDAR point cloud data is preprocessed by using preprocessing operations such as cropping and segmentation to obtain LiDAR point cloud data that accurately matches the boundary of the target region.

[0032] Specifically, in step S2, the specific steps for classifying LiDAR point cloud data and obtaining high-vegetation point cloud data are as follows: S201: Ground point identification from LiDAR point cloud data; S202: Construct a normalized digital surface model (nDSM) of the target area based on the identified ground points; S203: Classify the points in the point cloud based on the nDSM values ​​of each point in the point cloud data, obtain high vegetation points, and then generate a high vegetation point set; S204: Replace the original z-values ​​of high-vegetation points with nDSM values ​​to generate the final high-vegetation point cloud.

[0033] Specifically, in step S201, the classic Progressive TIN Densification (PTD) algorithm is used to identify ground points in the point cloud. The core of the method is as follows: first, a sparse initial TIN is constructed using seed points as the terrain base; then, each point is checked against distance and angle thresholds to determine if it is a ground point. If the condition is met, the TIN is inserted and the topology is updated until no new points are added. Finally, the vertices of the TIN become the set of ground points, and the identification effect is as follows: Figure 3 As shown.

[0034] Specifically, in step S202, the method for constructing the normalized digital surface model (nDSM) is as follows: Using the horizontal coordinates of ground points, an irregular triangular network (TIN) is constructed to generate a digital elevation model (DEM) of the survey area based on the TIN representation. For any non-ground point, find the triangle that the point falls into, and use bilinear interpolation to obtain the DEM value corresponding to the point; By subtracting the DEM value from the Z value at that point, the corresponding nDSM value can be obtained, such as... Figure 5 The image shown is a rendering of nDSM.

[0035] Specifically, in step S203, the specific methods for obtaining high vegetation points and generating a set of high vegetation points are as follows: Elevation thresholds are set based on the size of the nDSM value. h, nDSM value greater than Point h is determined to be a high-vegetation point; Based on the characteristic that building points have a very high probability of belonging to a certain plane, we can further distinguish between high vegetation points and building points in the point cloud, such as... Figure 4As shown, this is the point cloud classification effect. The high vegetation point set is saved as a new point cloud.

[0036] Specifically, a 3D “region growing” algorithm is used to segment the high-vegetation point cloud into a plane. During the segmentation process, a key parameter, “minimum number of points contained in a plane object”, needs to be adjusted appropriately according to the scene type and point cloud density. After the plane segmentation, if a high-vegetation point belongs to a certain plane object, the high-vegetation point is reclassified as a “building point”; otherwise, the category of the high-vegetation point remains unchanged.

[0037] Specifically, in step S204, for high vegetation points, the original Z value is affected by the terrain and cannot truly reflect its vertical distribution. However, the nDSM value removes the influence of terrain undulations and can better reflect the vertical distribution information of high vegetation points. The point cloud after replacing the original Z value of the high vegetation point with the nDSM value becomes a new high vegetation point cloud.

[0038] Specifically, in step S3, the method for calculating the multidimensional and multi-layer canopy entropy of the target area based on the high vegetation point cloud data is as follows: S301: Obtain the 3D bounding box of the point cloud of high vegetation in the target area; S302: Calculate the entropy in the z-axis direction; S303: Calculate the entropy in the x-axis direction; S304: Calculate the entropy in the y-axis direction; S305: Calculate the multidimensional, multilayer canopy entropy.

[0039] Specifically, in step S301, the method for obtaining the 3D bounding box of the point cloud of high vegetation in the target area is as follows: Iterate through every point in the high-vegetation point cloud and find the minimum value along the x-axis. Maximum value Minimum value along the y-axis Maximum value minimum value along the z-axis Maximum value The three-dimensional coordinates of the lower left corner of the three-dimensional bounding box of the high vegetation point cloud to be obtained ( ), the three-dimensional coordinates of the upper right corner ( ).

[0040] Specifically, in step S302, the method for calculating the entropy in the z-axis direction is as follows: Vertical layering is performed on the high-vegetation point cloud, and entropy is calculated, such as... Figure 7 As shown.

[0041] First, determine the number of vertical layers. The calculation formula is as follows: , Secondly, calculate the number of points in each layer in the vertical direction. Calculation formula: , Next, the proportion of points contained in each layer in the vertical direction to the total number of points contained in the entire high-vegetation point cloud is calculated. The calculation formula is as follows: , Finally, calculate the entropy in the z-axis direction, using formula (4).

[0042] Specifically, in step S303, the method for calculating the entropy in the x-axis direction is as follows: The point cloud of high vegetation is layered along the x-axis, and its entropy is calculated, such as... Figure 8 As shown.

[0043] First, determine the number of layers in the x-axis direction. The calculation formula is as follows: , Secondly, calculate the number of points in each layer along the x-axis. Calculation formula: , Next, the proportion of points contained in each layer along the x-axis to the total number of points contained in the entire high-vegetation point cloud is calculated. The calculation formula is as follows: , Finally, calculate the entropy in the x-axis direction, using formula (2).

[0044] Specifically, in step S304, the method for calculating the entropy in the y-axis direction is as follows: The point cloud of high vegetation is layered along the y-axis, and its entropy is calculated, such as... Figure 9 As shown.

[0045] First, determine the number of layers in the y-axis direction. The calculation formula is as follows: , Secondly, calculate the number of points in each layer along the y-axis. Calculation formula: , Next, the proportion of points contained in each layer along the y-axis to the total number of points contained in the entire high-vegetation point cloud is calculated. The calculation formula is as follows: , Finally, calculate the entropy in the y-axis direction, using formula (3).

[0046] Specifically, in step S305, the multidimensional multilayer canopy entropy is calculated with reference to formula (1).

[0047] Example 1 We selected point clouds from four different scenarios for our experiment.

[0048] Experimental data 1 ( Figures 10 to 14 The selected data was airborne LiDAR point cloud data acquired from a large aircraft platform in Toyama Prefecture, Japan. The point cloud density was 12 points / square meter. The area was 1600.00 meters long and 1500.00 meters wide. Figure 10 As shown, we calculated the multidimensional and multi-layer canopy entropy of each grid cell according to the 20m*20m grid pattern, and the results are as follows. Figure 11 As shown.

[0049] Experimental data 2 ( Figures 15 to 19 The selected data is airborne LiDAR point cloud data acquired by a drone platform in the Huangshan mountainous area of ​​Anhui Province. The point cloud density is 198 points / square meter. The area is 1000.00 meters long and 900.00 meters wide. Figure 15 As shown; we calculated the multidimensional, multi-layer canopy entropy of each grid cell according to the 20m*20m grid pattern, and the results are as follows. Figure 16 As shown.

[0050] Experimental data 3 ( Figures 20 to 21 The selected data uses LiDAR point cloud data acquired from a ground-based lidar platform in a dense forest area of ​​Guangxi. The point cloud density is 21,749 points / square meter. The area is 78.00 meters long and 83.00 meters wide. Figure 20 As shown, the vegetation in this scene exhibits a complex, multi-layered distribution. We calculated the multi-dimensional, multi-layered canopy entropy of the entire scene data using a grid, and the results are as follows. Figure 21 As shown.

[0051] Experimental data 4 ( Figures 22 to 26 The study selected airborne LiDAR point cloud data from a drone platform in a hilly farmland in Yishui County, Shandong Province (where tobacco is the main crop). The point cloud density was 203 points / square meter. The area is 787.52 meters long and 456.29 meters wide. Figure 22 As shown; we calculated the multidimensional, multi-layer canopy entropy of each grid cell according to the 20m*20m grid pattern, and the results are as follows. Figure 23 As shown.

[0052] In addition, h was set to 1.616 meters in the first three forest scenes and 0.382 meters in the fourth crop scene, mainly because the crops are relatively small and the elevation interval should also be set to a smaller value.

[0053] Analysis of the test results in the above four scenarios reveals that the entropy values ​​obtained by the method proposed in this application can all reflect the complexity of the vegetation canopy structure in the region. That is, in the same test scenario, the larger the entropy value, the higher the complexity of the canopy structure; conversely, the smaller the entropy value, the lower the complexity of the canopy structure.

[0054] like Figure 12 As shown, the multidimensional, multi-layered canopy entropy at this location is 6.8458, indicating dense vegetation with a multi-layered distribution; as... Figure 13 As shown, the multidimensional, multi-layered canopy entropy at this location is 6.2787, indicating dense vegetation with a two-layered distribution; as... Figure 14 As shown, the multidimensional canopy entropy at this location is 5.8851, indicating that the vegetation canopy exhibits a single-layer distribution.

[0055] same, Figure 17 , 18 19, Figure 24 , 25 Similar patterns can be observed in 26, where the entropy of the multidimensional and multi-layered canopy shows a decreasing trend, while the complexity of the canopy structure also shows a decreasing trend.

[0056] in addition, Figure 21 The entropy value of this scenario is 8.3526. This scenario features dense vegetation covering the entire canopy space, making it the most complex canopy scenario. On the other hand, comparing the results of experimental data 4 with those of experimental data 1, 2, and 3 shows that the change in the complexity of crop canopy structure is not as significant as that of forest canopy structure. For example, regarding crops... Figure 24 , 25 The three entropies corresponding to 26 are 8.5445, 8.3554, and 8.008, respectively, while those related to forests... Figure 17 , 18 The three entropies corresponding to 19 are 6.3524, 5.9740, and 5.7850, respectively. The difference for crops is 0.5437, while the difference for forests is 0.5674.

[0057] The above four experiments demonstrate that the method proposed in this application is a simple and effective method for measuring the complexity of vegetation canopy structure. It is applicable to LiDAR point cloud data acquired by multiple platforms such as manned aircraft, unmanned aircraft, and ground-based systems, and is also suitable for evaluating the complexity of canopy structure of various vegetation types such as forests and crops.

[0058] The embodiments of this application have been described in detail above with reference to the accompanying drawings. However, this application is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of this application.

Claims

1. A method for evaluating the complexity of vegetation canopy structure based on multidimensional and multi-layer canopy entropy, characterized in that, Includes the following steps: The complexity of vegetation canopy structure is quantitatively evaluated using the values ​​of multidimensional and multi-layer canopy entropy; The calculation method for the multidimensional, multi-layer canopy entropy is as follows: For the point cloud data of the vegetation canopy in the target area, let the point cloud data contain N points. Divide the point cloud data into segments along the x, y, and z directions at certain intervals h. , , Layer, a total of Point cloud data of the layer; Then, the number of points contained in each layer in the x, y, and z directions are counted respectively. The number of points in each layer in the X direction is recorded as follows: ; The number of points in each layer along the y-direction is denoted as: ; The number of points in each layer along the z-direction is denoted as: ; The formulas for calculating multidimensional and multi-layer canopy entropy are as follows: （1） （2） （3） （4） In the formula, MDMLCE represents the multidimensional, multilayer canopy entropy of the target region. Let be the entropy along the x-axis. Let be the entropy in the y-axis direction. Let be the entropy along the z-axis, and let i be an intermediate variable whose values ​​range from 0 to 1 when calculating the x-axis, y-axis, and z-axis directions, respectively. 0 to 0 to .

2. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 1, characterized in that: The method for calculating multidimensional, multi-layer canopy entropy also includes the following steps: S1: Acquire LiDAR point cloud data for the target area; S2: Classify LiDAR point cloud data to obtain high vegetation point cloud data; S3: Calculate the multidimensional and multi-layer canopy entropy of the target area based on the high vegetation point cloud data.

3. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 2, characterized in that: In step S1, the point cloud data of the target area includes one or more combinations of LiDAR point cloud data collected by multiple platforms such as manned airborne LiDAR, unmanned airborne LiDAR, and ground-based LiDAR. In step S1, the acquired LiDAR point cloud data is preprocessed by using cropping and block operations to obtain LiDAR point cloud data that accurately matches the boundary of the target area.

4. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 2, characterized in that: In step S2, the specific steps for classifying LiDAR point cloud data and obtaining high-vegetation point cloud data are as follows: S201: Ground point identification from LiDAR point cloud data; S202: Construct a normalized digital surface model (nDSM) of the target area based on the identified ground points; S203: Classify the points in the point cloud based on the nDSM values ​​of each point in the point cloud data, obtain high vegetation points, and then generate a high vegetation point set; S204: Replace the original z-values ​​of high-vegetation points with nDSM values ​​to generate the final high-vegetation point cloud.

5. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 4, characterized in that: In step S201, the classic irregular triangular mesh progressive encryption filtering algorithm is used to identify ground points in the point cloud.

6. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 4, characterized in that: In step S202, the method for constructing the normalized digital surface model (nDSM) is as follows: Using the horizontal coordinates of ground points, an irregular triangular network (TIN) is constructed to generate a digital elevation model (DEM) of the survey area based on the TIN representation. For any non-ground point, find the triangle that the point falls into, and use bilinear interpolation to obtain the DEM value corresponding to the point; By subtracting the DEM value from the Z value of the point, the nDSM value corresponding to the point is obtained, as shown in Figure 5, which is the nDSM effect diagram.

7. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 4, characterized in that: In step S203, the specific methods for obtaining high vegetation points and generating a set of high vegetation points are as follows: Elevation thresholds are set based on the size of the nDSM value. h, nDSM value greater than Point h is determined to be a high vegetation point; Based on the characteristic that building points have a very high probability of belonging to a certain plane, we further distinguish between high vegetation points and building points in the point cloud, and save the high vegetation point set as a new point cloud.

8. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 2, characterized in that: In step S3, the specific method for calculating the multidimensional and multi-layer canopy entropy of the target area based on the high vegetation point cloud data is as follows: S301: Obtain the 3D bounding box of the point cloud of high vegetation in the target area; S302: Calculate the entropy in the z-axis direction; S303: Calculate the entropy in the x-axis direction; S304: Calculate the entropy in the y-axis direction; S305: Calculate the multidimensional, multilayer canopy entropy.

9. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 8, characterized in that: In step S301, the specific method for obtaining the 3D bounding box of the high-vegetation point cloud of the target area is as follows: Iterate through every point in the high-vegetation point cloud and find the minimum value along the x-axis. Maximum value Minimum value along the y-axis Maximum value minimum value along the z-axis Maximum value The three-dimensional coordinates of the lower left corner of the three-dimensional bounding box of the high vegetation point cloud to be determined ( ), the three-dimensional coordinates of the upper right corner ( ).

10. The vegetation canopy structure complexity evaluation method based on multidimensional and multi-layer canopy entropy according to claim 9, characterized in that: In step S302, the specific method for calculating the entropy in the z-axis direction is as follows: The point cloud with high vegetation was processed into vertical layers and its entropy was calculated, as shown in Figure 7. First, determine the number of vertical layers. The calculation formula is as follows: ， Secondly, calculate the number of points in each layer in the vertical direction. Calculation formula: ， Next, the proportion of points contained in each layer in the vertical direction to the total number of points contained in the entire high-vegetation point cloud is calculated. The calculation formula is as follows: ， Finally, calculate the entropy in the z-axis direction. The calculation formula is shown in formula (4). Based on the method and formulas (2) and (3) for calculating the entropy in the z-axis direction, the entropy in the x and y-axis directions is calculated respectively, and then the multidimensional and multi-layer canopy entropy is calculated according to formula (4).

Images