Single tree segmentation method, electronic equipment and computer readable storage medium
Patent Information
- Application Number
- CN202410345091.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-22
- Publication Date
- 2025-03-14
Smart Images
Figure CN119625512A_ABST
Abstract
Description
Technical Field
[0001] The present application belongs to the technical field of point cloud processing, and in particular, relates to a single tree segmentation method, an electronic device, and a computer-readable storage medium. Background Art
[0002] Single tree segmentation is an important step in forest point cloud processing, which is of great significance for forest survey, monitoring and management. Single tree segmentation is used to extract each individual tree from complex point cloud data. However, existing single tree segmentation methods have problems such as recognition errors or calculation errors, which lead to single tree segmentation errors and affect the segmentation effect. Summary of the invention
[0003] In view of the above, the present application provides a single tree segmentation method, an electronic device and a computer-readable storage medium, which can effectively reduce single tree segmentation errors, improve the accuracy and efficiency of single tree segmentation, avoid erroneous segmentation, and improve the single tree segmentation effect.
[0004] In a first aspect, the present application provides a single tree segmentation method, the method comprising: obtaining forest point cloud data to be processed; processing the forest point cloud data to obtain trunk area point cloud data; processing the trunk area point cloud data to obtain trunk points; determining the trunk point cloud corresponding to the trunk point; determining the crown area point cloud data based on the forest point cloud data and the trunk point cloud; determining the crown point cloud corresponding to the trunk point based on the trunk points and the crown area point cloud data; merging the trunk point cloud and the crown point cloud corresponding to the same trunk point to obtain a single tree point cloud.
[0005] In another possible implementation, the point cloud data of the trunk area is processed to obtain the trunk points, including: obtaining the trunk point cloud cluster and the center of gravity of the trunk point cloud cluster based on the point cloud data of the trunk area; projecting the center of gravity of the trunk point cloud cluster on the horizontal plane to obtain the center of gravity point cloud cluster; clustering the center of gravity point cloud cluster to obtain individual center of gravity clusters; extracting the center of gravity of each center of gravity cluster to obtain the trunk points.
[0006] In another possible implementation, obtaining trunk point cloud clusters and the centroid of trunk point cloud clusters based on trunk area point cloud data includes: stratifying the trunk area point cloud data according to the first layer height to obtain multiple layers of trunk point cloud data; clustering the trunk point cloud data for each layer of trunk point cloud data to obtain trunk point cloud clusters; and extracting the centroid of each trunk point cloud cluster to obtain the centroid of the trunk point cloud cluster.
[0007] In another possible implementation, determining the trunk point cloud corresponding to the trunk point includes: determining a centroid cluster corresponding to the trunk point; determining a trunk point cloud cluster corresponding to each centroid in the centroid cluster; and using the determined trunk point cloud cluster as the trunk point cloud corresponding to the trunk point.
[0008] In another possible implementation, determining the crown point cloud corresponding to the trunk point based on the trunk point and crown area point cloud data includes: obtaining the crown point cloud cluster and the center of gravity of the crown point cloud cluster for each layer based on the crown area point cloud data; for each layer, taking the center of gravity of the crown point cloud cluster with the smallest height value in the layer as the target center of gravity of the layer; for each layer, setting the height value of each trunk point to the height value of the target center of gravity of the layer to obtain the trunk point corresponding to the layer; for the crown point cloud cluster in each layer, calculating the distance between the center of gravity of the crown point cloud cluster and the trunk point corresponding to the layer; for each trunk point, taking the crown point cloud cluster corresponding to the smallest distance in each layer as the crown point cloud corresponding to the trunk point.
[0009] In another possible implementation, obtaining the crown point cloud clusters and the centroid of each layer of the crown point cloud clusters based on the crown area point cloud data includes: stratifying the crown area point cloud data according to the second layer height to obtain multiple layers of crown point cloud data; clustering the crown point cloud data for each layer to obtain crown point cloud clusters; and extracting the centroid of each crown point cloud cluster to obtain the centroid of the crown point cloud cluster.
[0010] In another possible implementation, determining the crown area point cloud data based on the forest point cloud data and the trunk point cloud includes: obtaining the remaining point clouds in the forest point cloud data except the trunk point clouds; and using the remaining point clouds as the crown area point cloud data.
[0011] In another possible implementation, processing the forest point cloud data to obtain the trunk area point cloud data includes: dividing the forest point cloud data according to a preset height range, and using the forest point cloud data within the preset height range as the trunk area point cloud data.
[0012] An embodiment of the present application provides an electronic device, which includes a processor and a memory, wherein the processor is configured to implement the single tree segmentation method when executing a computer program stored in the memory.
[0013] An embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the single tree segmentation method is implemented.
[0014] In summary, the single tree segmentation method provided in the embodiment of the present application obtains the forest point cloud data to be processed, processes the forest point cloud data, and obtains the trunk area point cloud data. The trunk area point cloud data is processed to obtain the trunk point. The trunk point cloud corresponding to the trunk point is determined. The crown area point cloud data is determined according to the forest point cloud data and the trunk point cloud. According to the trunk point and the crown area point cloud data, the crown point cloud corresponding to the trunk point is determined. The trunk point cloud and the crown point cloud corresponding to the same trunk point are merged to obtain a single tree point cloud. By extracting the trunk points from the trunk area point cloud data, and then determining the trunk point cloud and the crown point cloud according to the trunk points. Extracting the trunk points based on the trunk area point cloud data can reduce the situation of recognition errors, ensure the number of trees, and provide a reliable basis for subsequent accurate segmentation. Then, the trunk point cloud and the crown point cloud corresponding to the trunk point are determined according to the trunk point, which can effectively reduce segmentation errors, improve the accuracy and efficiency of single tree segmentation, avoid erroneous segmentation, and improve the single tree segmentation effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] Figure 1 It is a flow chart of a single wood segmentation method provided in one embodiment of the present application.
[0016] Figure 2 This is a schematic diagram of forest point cloud data provided by an embodiment of the present application.
[0017] Figure 3 Yes Figure 2 The diagram shows the point cloud data of the trunk area obtained after elevation filtering of the forest point cloud data.
[0018] Figure 4 It is a schematic diagram of multi-layer tree trunk point cloud data obtained after layering the tree trunk area point cloud data provided by an embodiment of the present application.
[0019] Figure 5 Schematic diagram of the center of gravity of a tree trunk point cloud cluster provided in one embodiment of the present application.
[0020] Figure 6 Schematic diagram of the center of gravity of a center of gravity cluster provided in one embodiment of the present application.
[0021] Figure 7 It is a schematic flow chart of a method for determining a tree trunk point cloud corresponding to a tree trunk point provided in an embodiment of the present application.
[0022] Figure 8 It is a schematic diagram of a tree trunk point cloud corresponding to each tree trunk point provided in an embodiment of the present application.
[0023] Fig. 9 It is a schematic diagram of tree crown point cloud data provided by an embodiment of the present application.
[0024] Fig.10It is a schematic diagram of layer crown point cloud data provided by an embodiment of the present application.
[0025] Fig.11 It is a schematic diagram of the centroid of a tree crown point cloud cluster provided in an embodiment of the present application.
[0026] Fig.12 This is a schematic diagram of the tree crown point cloud corresponding to each trunk point provided in yet another embodiment of the present application.
[0027] Fig.13 It is a schematic diagram of a single tree point cloud provided in one embodiment of the present application.
[0028] Fig.14 It is a schematic flow chart of a single wood splitting method provided in another embodiment of the present application.
[0029] Fig.15 It is a schematic diagram of the structure of an electronic device provided in one embodiment of the present application.
[0030] Fig.16 It is a schematic diagram of the structure of a single wood splitting device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0031] It should be noted that, in the specification, claims and drawings of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "multiple" in the specification, claims and drawings of the present application refers to two or more than two.
[0032] It should also be noted that the method disclosed in the embodiments of the present application or the method shown in the flowchart includes one or more steps for implementing the method. Without departing from the scope of the claims, the execution order of multiple steps can be interchangeable with each other, and some of the steps can also be deleted.
[0033] Single tree segmentation is to divide the input point cloud into single tree point clouds of corresponding quantity and / or shape. In other words, single tree segmentation refers to the process of extracting each individual tree from the point cloud data and segmenting it when the point cloud data set is concentrated. Single tree segmentation can be applied in the fields of precision forestry survey, forest health monitoring, forestry resource management, ecological research, etc., and the embodiments of this application do not specifically limit this.
[0034] The following describes the single wood segmentation method and existing problems in the prior art:
[0035] Methods based on airborne scanning: This type of method mainly projects the airborne scanning point cloud, determines the crown seed point based on the top of the crown as the highest area, and then clusters the surrounding crown point clouds through the seed points to complete the single tree segmentation of the tree. The above method is aimed at the canopy model obtained from the airborne point cloud, which lacks the trunk point cloud information, and in densely populated forest areas, it is easy to have the problem of extracting the crown seed point incorrectly, which leads to segmentation errors.
[0036] Methods based on handheld and backpack scanning: First, linear feature judgment is performed on the input point cloud to determine the trunk seed point, and then the nearby point cloud is divided into the seed point through growth through the trunk seed point, and finally the single tree segmentation of the input point cloud is achieved. However, due to the large number of input point clouds, false seed points will be calculated based on linear feature judgment alone, resulting in segmentation errors and over-segmentation.
[0037] Methods based on deep learning: This type of method mainly collects a large number of tree samples, calculates the hyperparameter model of the tree through a deep learning network, and distinguishes through input point clouds, and finally achieves single tree segmentation of point clouds. This type of method has high requirements for training point clouds, and different trees have different point cloud features, so a large amount of sample data is required. In addition, the quality of the point clouds obtained is different for different collection methods or equipment. The generalization ability of the hyperparameter model is weak, and currently it cannot segment single trees well.
[0038] As mentioned above, the single tree segmentation method in the prior art has problems such as incorrect extraction of crown seed points, calculation of false seed points, and weak generalization ability of the deep learning network model, which in turn leads to single tree segmentation errors and affects the segmentation effect.
[0039] In view of this, the embodiment of the present application provides a single tree segmentation method, by extracting trunk points from the trunk area point cloud data, and then determining the trunk point cloud and the crown point cloud based on the trunk points. Extracting trunk points based on the trunk area point cloud data can reduce the situation of recognition errors, ensure the number of trees, and provide a reliable basis for subsequent accurate segmentation. Then, the trunk point cloud and the crown point cloud are determined based on the trunk points, effectively reducing segmentation errors, improving the accuracy and efficiency of single tree segmentation, avoiding erroneous segmentation, and improving the single tree segmentation effect.
[0040] The single wood splitting method provided in the embodiment of the present application can be applied to and executed by an electronic device, which may include but is not limited to: a desktop computer, a portable computer, a laptop computer, a tablet computer, a mobile phone, etc., which is not specifically limited in the embodiment of the present application.
[0041] In other embodiments, the single tree segmentation method provided in the embodiments of the present application can be applied to a hardware environment composed of a server and the above-mentioned electronic device. The server is connected to the above-mentioned electronic device through a network, and can be used to provide services for the electronic device or a client installed on the electronic device. A database can be set on the server or independently of the server to provide data storage services for the server, such as storing point cloud data to be processed or processed point cloud data. The above-mentioned network may include but is not limited to: a wide area network, a metropolitan area network or a local area network. The single tree segmentation method provided in the embodiments of the present application can be executed by the server, or by the electronic device and the server together.
[0042] The following describes some embodiments of the present invention by using an electronic device to perform the single wood splitting method provided in the embodiments of the present invention, and combines the accompanying drawings to illustrate some embodiments. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.
[0043] See also Figure 1 , a single wood segmentation method provided in an embodiment of the present application is exemplarily introduced, and the single wood segmentation method can be applied to electronic devices, including steps S101 to S118.
[0044] Step S101, obtaining the forest point cloud data to be processed.
[0045] The forest point cloud data can be the point cloud data collected by scanning the forest with a scanning device. Point cloud data refers to a set of vectors in a three-dimensional coordinate system. Point cloud data can be recorded in the form of points, and each point contains three-dimensional coordinates, such as coordinates on the X-axis, Y-axis, and Z-axis, which can be used to indicate the real-world coordinate position.
[0046] The above-mentioned scanning equipment may include but is not limited to: Laser Radar, contact scanner, structured light and depth camera, which is not specifically limited in the embodiments of the present application.
[0047] It can be understood that when the scanning device is a laser radar, the collected point cloud data is the laser radar scanning point cloud.
[0048] In the embodiment of the present application, the tree point cloud data to be processed obtained in step S101 may be the tree point cloud data obtained by preprocessing the point cloud data (raw data) collected by the scanning device. By preprocessing the forestry point cloud data, the point cloud data to be calculated and processed can be effectively reduced, the subsequent processing of the point cloud data can be accelerated, and the efficiency of subsequent single tree segmentation can be effectively improved.
[0049] The above preprocessing may include but is not limited to: filtering of point clouds, removing ground points and noise points, which is not specifically limited in the present embodiment. The above point cloud filtering may include but is not limited to: voxel filtering and statistical filtering, which is not specifically limited in the present embodiment.
[0050] Exemplarily, the scanning device collects forestry point cloud data and inputs the collected forestry point cloud data into the electronic device. The electronic device performs voxel downsampling on the forestry point cloud data input by the scanning device, such as voxelizing all input forest point cloud data with a voxel side length of 0.1m, and using the center of each voxel to represent all point clouds within the voxel, thereby significantly reducing the point cloud data to be calculated and processed and improving the efficiency of single tree segmentation.
[0051] In another possible implementation, the forest point cloud data to be processed obtained in step S101 may be: point cloud data collected by a scanning device, that is, point cloud data that has not been preprocessed. The specific implementation may be determined based on actual conditions, and the embodiment of the present application does not make any specific limitations on this.
[0052] Step S102, dividing the tree point cloud data according to a preset height range to obtain tree trunk area point cloud data.
[0053] The preset height range is used to indicate the distribution range of the tree trunks. The preset height range may be [0, 3.9], [1.3, 3.9], etc. The preset height range may be set according to the tree species, the height of the current forest, etc., and the present embodiment of the application does not specifically limit this.
[0054] For example, if the trunks of real trees corresponding to the tree point cloud data obtained in step S101 are distributed between 0 and 3.9 meters, the preset height range can be set to [0, 3.9]. Figure 2 is the tree point cloud data obtained in step S101. The electronic device compares the tree point cloud data with the tree point cloud data according to the preset height range [0, 3.9]. Figure 2 The height values of the forest point cloud data shown in the figure are filtered, and only the point cloud data with height values in the interval [0, 3.9] are retained, that is, the point cloud data with Z values in the interval [0, 3.9] are retained, and the following is obtained: Figure 3 Point cloud data of the tree trunk area is shown.
[0055] Step S103, layering the tree trunk area point cloud data according to the first layer height to obtain multi-layer tree trunk point cloud data.
[0056] The first layer height can be used to indicate the layer height for performing point cloud layering operations on the trunks of trees. The first layer height can be 0.3m, 0.4m or 0.5m, etc. The first layer height can be set according to the tree species, the height of the current trees, etc., and the embodiment of the present application does not specifically limit this.
[0057] In the embodiment of the present application, the electronic device divides the tree trunk area point cloud data obtained in step S102 according to the first layer height to obtain multiple layers of tree trunk point cloud data. The height of each layer of tree trunk point cloud data is the first layer height.
[0058] For example, taking the first floor height as 0.3 meters, Figure 3 The point cloud data of the trunk area shown in the figure is divided into 13 layers with a layer height of 0.3 m per layer, and the following is obtained: Figure 4 The 13 layers of tree trunk point cloud data are shown. Figure 4 Each color fill represents a layer of tree trunk point cloud data, and each layer of tree trunk point cloud data corresponds to a real-world height of 0.3m.
[0059] Step S104: clustering the tree trunk point cloud data at each layer to obtain a tree trunk point cloud cluster.
[0060] Among them, clustering may include but is not limited to: DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering, KMeans clustering, Euclidean clustering, and regional growth, which are not specifically limited in the embodiments of the present application. Among them, the parameters of DBSCAN clustering include the neighborhood radius r and the minimum number of points MinPt, the neighborhood radius r can take values of 0.4 or 0.5, etc., and the minimum number of points MinPt can take values of 19 or 20, etc. In the specific implementation process, the optimal values of the two parameters of the neighborhood radius r and the minimum number of points MinPt of this DBSCAN clustering can be determined by cross-validation and other techniques, which are not specifically limited in the embodiments of the present application.
[0061] In the embodiment of the present application, each layer of tree trunk point cloud data may include tree trunk point cloud data of multiple different trees. The electronic device may traverse all tree trunk point cloud data of each layer, cluster all tree trunk point cloud data in each layer, and obtain clustered tree trunk point cloud clusters of each layer.
[0062] For example, taking DBSCAN clustering of tree trunk point cloud data as an example, the neighborhood radius r=0.5 and the minimum number of points MinPts=20.
[0063] It can be understood that when there are eight trees in the forest point cloud data obtained in step S101, eight trunk point cloud clusters can usually be obtained after clustering each layer of trunk point cloud data. When the number of layers in step S103 is 3, a total of 3*8=24 trunk point cloud clusters can usually be obtained. It can be understood that the number of trunk point cloud clusters actually obtained is related to the clustering operation and the actual point cloud data. For example, the neighborhood radius r and the minimum number of points MinPts mentioned above have an important influence on the clustering results.
[0064] Step S105 , extracting the centroid of each tree trunk point cloud cluster to obtain the centroid of the tree trunk point cloud cluster.
[0065] In the embodiment of the present application, after clustering each layer of tree trunk point cloud data, multiple tree trunk point cloud clusters can be obtained. For each tree trunk point cloud cluster in each layer, the center of gravity of the tree trunk point cloud cluster is extracted to obtain the center of gravity of the tree trunk point cloud cluster, and then the center of gravity of all layers of tree trunk point cloud clusters can be obtained.
[0066] In the embodiment of the present application, for each tree trunk point cloud cluster obtained in step S104, the average value of the XYZ coordinates of all point clouds in each tree trunk point cloud cluster is calculated to obtain the centroid of the corresponding tree trunk point cloud cluster. Where i is the tree trunk point cloud cluster of the i-th layer, and j is the tree trunk point cloud cluster of the j-th layer. i is greater than or equal to 1 and less than or equal to the number of layers after stratification in step S103. For example, if step S103 obtains 13 layers of tree trunk point cloud clusters, then i can be greater than or equal to 1 and less than or equal to 13.
[0067] For example, a tree trunk point cloud cluster A (A1, A2...A n ), A k is the point cloud in the tree trunk point cloud cluster A, A k The coordinates of (X k , Y k , Z k ), k is an integer greater than or equal to 1 and less than or equal to n, and n is the number of all point clouds in the tree trunk point cloud cluster A. Calculate the average value of the XYZ coordinates of all point clouds in the tree trunk point cloud cluster A to obtain the centroid P of the corresponding tree trunk point cloud cluster A. c (X mean , Y mean , Z mean ). Among them, X mean =(X1+X2+X3+…+X n ) / n,Y mean =(Y1+Y2+Y3+…+Y n ) / n,Z mean =(Z1+Z2+Z3+…+Z n ) / n.
[0068] As in the above example, Figure 4 The tree trunk point cloud data of each layer shown in the figure are clustered to obtain each tree trunk point cloud cluster, and the center of gravity of each tree trunk point cloud cluster is extracted to obtain the center of gravity of each tree trunk point cloud cluster, as shown in FIG. Figure 5 shown.
[0069] Step S106, projecting the centroids of all layers of tree trunk point cloud clusters onto a horizontal plane to obtain centroid point cloud clusters.
[0070] In the embodiment of the present application, after the centroids of the tree trunk point cloud clusters of each layer are obtained in step S105, the centroids of all layers of tree trunk point cloud clusters are projected on the horizontal plane (XOY plane) to obtain the centroid point cloud cluster. The coordinates of each centroid in the centroid point cloud cluster are Where i is the i-th layer of tree trunk point cloud cluster, j is the j-th tree trunk point cloud cluster, and the relevant contents of i and j can be referred to above and will not be repeated here.
[0071] Specifically, the electronic device changes the Z values of the centroids of all tree trunk point cloud clusters obtained in step S105 to 0, and retains their original coordinate values on the X-axis and Y-axis.
[0072] Step S107: clustering the centroid point cloud clusters to obtain various centroid clusters.
[0073] In the embodiment of the present application, the clustering of the centroid point cloud clusters may include but is not limited to: DBSCAN clustering, KMeans clustering, Euclidean clustering, and regional growth, which are not specifically limited in the embodiment of the present application. Among them, in the parameters of DBSCAN clustering of the centroid point cloud clusters, the neighborhood radius r can take a value of 0.2 or 0.3, etc., and the minimum number of points MinPt can take a value of 1 or 2, etc. In the specific implementation process, the optimal values of the two parameters, the neighborhood radius r and the minimum number of points MinPt, can be determined by cross-validation and other techniques, which are not specifically limited in the embodiment of the present application.
[0074] In the embodiment of the present application, the clustering method for the centroid point cloud cluster and the clustering method for the trunk point cloud data may be the same as or different from each other.
[0075] For example, taking DBSCAN clustering of the centroid point cloud cluster as an example, the neighborhood radius r=0.2 and the minimum number of points MinPts=1.
[0076] In the embodiment of the present application, each centroid cluster can be obtained after clustering the centroid point cloud clusters. Generally, a tree can uniquely correspond to a centroid cluster.
[0077] Step S108, extracting the centroid of each centroid cluster to obtain the corresponding trunk point.
[0078] In the embodiment of the present application, after clustering the centroid point cloud clusters, multiple centroid clusters can be obtained. For each centroid cluster, the centroid of the centroid cluster is extracted to obtain the centroid of the centroid cluster, and the centroid of the centroid cluster is the trunk point.
[0079] In the embodiment of the present application, the average value of the XYZ coordinates of all point clouds in each centroid cluster is calculated to obtain the centroid corresponding to each centroid cluster. For example, the centroid cluster B (B z1 , B z2 …B zm ), B zk is the centroid point cloud in the centroid cluster B, B zk The coordinates of (X meank , Y meank , 0), k is an integer greater than or equal to 1 and less than or equal to m, and m is the number of all point clouds in the centroid cluster B. zmean =(X mean1 +X mean2 +X mean3 +…+X meanm ) / m,Y zmean =(Y mean1 +Y mean2 +Y mean3 +…+Y meanm ) / m, where L is the Lth centroid cluster. L is a positive number greater than or equal to 1, and L is less than or equal to the number of all centroid clusters obtained in step S107.
[0080] For example, after extracting the centroid of each centroid cluster obtained in step S107, the centroid of each centroid cluster can be obtained, such as Figure 6 shown. Figure 6 The centroids of each centroid cluster on the XOZ plane are shown. The centroid of the centroid cluster is the trunk point, which means that each trunk point can uniquely represent a tree in the real world.
[0081] Step S109, determining the tree trunk point cloud corresponding to each tree trunk point.
[0082] In an embodiment of the present application, after the trunk point is determined in step S108, the trunk point cloud cluster corresponding to the centroid cluster in each layer is obtained by reverse deduction based on the centroid cluster determined for the trunk point, and finally the determined trunk point cloud cluster can be used as the trunk point cloud corresponding to the trunk point.
[0083] See also Figure 7 , an example is introduced to introduce a method for determining a trunk point cloud corresponding to a trunk point provided in an embodiment of the present application, the method comprising steps S701 to S703.
[0084] Step S701: for each trunk point, determine the centroid cluster corresponding to the trunk point.
[0085] In step S108, for each centroid cluster, the centroid of the centroid cluster is extracted to obtain the trunk point corresponding to the centroid cluster. In step S701, for each trunk point, the centroid cluster corresponding to the trunk point is determined, that is, the centroid cluster of the trunk point determined in step S108 is used as the centroid cluster corresponding to the trunk point.
[0086] Exemplarily, the centroid of the centroid cluster B1 is extracted to obtain the centroid P1 of the centroid cluster B1. The centroid P1 is the trunk point P1. For the trunk point P1, it can be determined that the centroid cluster corresponding to it is the centroid cluster B1.
[0087] Step S702: determining the tree trunk point cloud cluster corresponding to each centroid in the centroid cluster.
[0088] In step S107, the centroid cluster is obtained by clustering the centroid point cloud clusters, that is, the centroid point cloud of the centroid cluster comes from the centroid of the centroid point cloud cluster. In step S106, the centroids of all layers of tree trunk point cloud clusters are projected on the horizontal plane to obtain the centroid point cloud cluster, that is, the centroid point cloud cluster of the centroid point cloud cluster is obtained based on the centroids of all layers of tree trunk point cloud clusters. In step S105, the centroid of the tree trunk point cloud cluster is obtained by extracting the centroid of each tree trunk point cloud cluster, so that the tree trunk point cloud cluster corresponding to each centroid in the centroid cluster can be determined based on the centroid cluster.
[0089] Step S703: taking the determined tree trunk point cloud cluster as the tree trunk point cloud corresponding to the tree trunk point.
[0090] In the embodiment of the present application, the centroid cluster corresponding to the trunk point is determined in step S701, and the trunk point cloud cluster corresponding to each centroid in the centroid cluster is determined in step S702, and then the determined trunk point cloud cluster can be used as the trunk point cloud corresponding to the trunk point in step S703. Thus, for each trunk point, the trunk point cloud clusters of each layer corresponding to the trunk point can be determined based on steps S701 and S702.
[0091] As in the above example, in step S701, the centroid cluster corresponding to the trunk point P1 is determined as the centroid cluster B1, and the trunk point cloud cluster corresponding to each centroid in the centroid cluster B1 is used as the trunk point cloud of the trunk point P1. If the centroid cluster B1 includes the centroid b and the centroid c, the centroid b is determined by the trunk point cloud cluster a of the first layer, and the centroid c is determined by the trunk point cloud cluster d of the second layer, then the trunk point cloud clusters a and b both belong to the trunk point cloud of the trunk point P1.
[0092] based on Figure 6 The trunk points shown can determine the trunk point cloud corresponding to each trunk point, such as Figure 8 shown.
[0093] It can be understood that for the same tree, the number of point clouds of the trunk point cloud determined based on steps S701 to S703 may be different from that of the forest point cloud data about the tree obtained in step S101, for example, less.
[0094] Step S110, determining the tree crown area point cloud data according to the forest point cloud data and the tree trunk point cloud.
[0095] In the embodiment of the present application, after obtaining the trunk point cloud corresponding to each trunk point in step S109, the remaining point cloud data except the trunk point cloud in the forest point cloud data to be processed is obtained, and the remaining point cloud data is used as the crown area point cloud data.
[0096] For example, Figure 2 The tree point cloud data to be processed is split into the trunk point cloud and the remaining point cloud. Figure 2 The tree point cloud data to be processed is removed Figure 8 The tree trunk point cloud shown, Figure 2 The remaining point cloud data in Fig. 9 As shown, the remaining point cloud data can be regarded as tree crown point cloud data.
[0097] Step S111, layering the tree crown area point cloud data according to the second layer height to obtain multi-layer tree crown point cloud data.
[0098] The second layer height is used to indicate the layer height for performing point cloud layering operations on the tree crowns. The second layer height can be 1m, 0.8m, or 1.2m, etc. The second layer height can be set according to the tree species, the height of the current trees, etc., and the embodiment of the present application does not specifically limit this.
[0099] In the embodiment of the present application, the electronic device divides the tree crown area point cloud data obtained in step S110 according to the second layer height to obtain multiple layers of tree crown point cloud data. The height of each layer of tree crown point cloud data is the second layer height.
[0100] For example, Fig. 9 The tree crown point cloud data shown is layered according to the second layer height, and the following is obtained: Fig.10 The multi-layer tree crown point cloud data shown. Fig.10 Each color filling represents a layer of tree crown point cloud data, and the height of each layer of tree crown point cloud data corresponding to the real world is the second layer height.
[0101] Step S112: clustering the tree crown point cloud data of each layer to obtain tree crown point cloud clusters.
[0102] In the embodiment of the present application, the clustering of the crown point cloud data may include but is not limited to: DBSCAN clustering, K-means clustering, and the embodiment of the present application does not specifically limit this. Among them, in the parameters of DBSCAN clustering of the crown point cloud data, the neighborhood radius r can take a value of 0.3 or 0.2, etc., and the minimum number of points MinPt can take a value of 5 or 4, etc. In the specific implementation process, the optimal values of the two parameters, the neighborhood radius r and the minimum number of points MinPt, can be determined by cross-validation and other technologies, and the embodiment of the present application does not specifically limit this.
[0103] In the embodiment of the present application, the method for clustering the crown point cloud data and the method for clustering the centroid point cloud clusters and the method for clustering the trunk point cloud data may be the same as or different from each other.
[0104] In the embodiment of the present application, each layer of tree crown point cloud data may include tree crown point cloud data of multiple different trees. Each layer of tree crown point cloud data is traversed, and each layer of tree crown point cloud data is clustered to obtain each layer of clustered tree crown point cloud clusters.
[0105] For example, taking DBSCAN clustering of tree crown point cloud data as an example, the neighborhood radius r=0.2 and the minimum number of points MinPts=5.
[0106] Step S113, extracting the centroid of each tree crown point cloud cluster to obtain the centroid of the tree crown point cloud cluster.
[0107] In the embodiment of the present application, after clustering each layer of tree crown point cloud data, multiple tree crown point cloud clusters can be obtained. For each tree crown point cloud cluster in each layer, the center of gravity of the tree crown point cloud cluster is extracted to obtain the center of gravity of the tree crown point cloud cluster.
[0108] In the embodiment of the present application, for the tree crown point cloud cluster obtained in step S112, the average value of the XYZ coordinates of all point clouds in each tree crown point cloud cluster is calculated to obtain the centroid of the corresponding tree crown point cloud cluster. For example, a tree trunk point cloud cluster C (C1, C2...C h ), C k is the point cloud in the tree trunk point cloud cluster C, C k The coordinates of (X k , Y k , Z k ), k is an integer greater than or equal to 1 and less than or equal to h, h is the number of all point clouds in the tree trunk point cloud cluster C. dmean =(X1+X2+X3+…+X h ) / h,Y dmean =(Y1+Y2+Y3+…+Y h ) / h,Z dmean =(Z1+Z2+Z3+…+Zh ) / h, where i is the i-th layer and j is the j-th point cloud cluster. The relevant contents of i and j can be referred to above and will not be repeated here.
[0109] For example, the centroid of each tree crown point cloud cluster is extracted to obtain the centroid of the tree crown point cloud cluster as follows: Fig.11 shown.
[0110] Step S114: for each layer, the centroid of the crown point cloud cluster with the smallest height value in the layer is used as the target centroid of the layer.
[0111] In step S114, for each layer divided in step S111, the centroid of the crown point cloud cluster with the smallest height value in the layer is found, and the centroid is used as the target centroid of the layer.
[0112] Step S115 , for each layer, the height value of each trunk point is set to the height value of the target center of gravity of the layer, so as to obtain the trunk point corresponding to the layer.
[0113] In step S115, for each layer divided in step S111, the height value of the trunk point obtained in step S108 is set as the height value of the target centroid of the layer to obtain the trunk point corresponding to the layer.
[0114] For example, the two trunk points obtained in step S108 are Step S111 divides the tree crown point cloud data into three layers, and step S114 obtains the height value of the target center of gravity of the first layer as Z c1 , the height value of the target center of gravity of the second layer is Z c2 And the height value of the target center of gravity of the third layer is Z c3 For example, in each layer, the height values (Z values) of the two tree trunk points obtained in step S108 are set as the height value of the target center of gravity of the layer. For example, in the first layer, the tree trunk points after setting are That is, the trunk points corresponding to the first layer are In the second layer, the trunk points after setting are That is, the trunk points corresponding to the second layer are In the third layer, the trunk points after setting are That is, the trunk points corresponding to the third layer are
[0115] Step S116, for each tree crown point cloud cluster in each layer, calculate the distance between the centroid of the tree crown point cloud cluster and the tree trunk point corresponding to the layer.
[0116] In the embodiment of the present application, for each layer, the distance between each centroid of the crown point cloud cluster of the layer and the trunk point corresponding to the layer is calculated. The method of calculating the distance between the centroid and the trunk point includes but is not limited to: Euclidean distance, Manhattan distance, Dijkstra algorithm and Astar algorithm, which are not specifically limited in the embodiment of the present application.
[0117] For example, taking the Euclidean distance as an example, the formula for calculating the nearest trunk point at the i-th layer is:
[0118]
[0119] in, as well as are the X-axis, Y-axis and Z-axis coordinates of the j-th centroid, respectively, z (X zmean ), P z (Y zmean ) and P z (Z c ) are the X-axis, Y-axis and Z-axis coordinates of the trunk point corresponding to the i-th layer.
[0120] Step S117 , for each trunk point, the crown point cloud cluster corresponding to the smallest distance in each layer is used as the crown point cloud corresponding to the trunk point.
[0121] In the embodiment of the present application, the centroid of each layer of tree crown point cloud clusters is traversed, and the distance between the centroid and all tree trunk point clouds of this layer is calculated. Then, for each tree trunk point, the tree crown point cloud cluster corresponding to the smallest distance in each layer is used as the tree crown point cloud corresponding to the tree trunk point. In other words, for each tree crown point cloud cluster, it is considered to belong to the tree trunk point with the smallest distance from it, so it can be used as the tree crown point cloud of the tree trunk point with the smallest distance from it.
[0122] For example, for Figure 6 For each trunk point shown, the distance between the trunk point and Fig.11 The trunk point cloud cluster closest to the center of gravity of the crown point cloud cluster shown in the figure is taken as the crown point cloud corresponding to the trunk point, and the crown point cloud of each trunk point is obtained as follows Fig.12 shown.
[0123] Step S118, merging the trunk point cloud and the crown point cloud corresponding to the same trunk point to obtain a single tree point cloud.
[0124] In the embodiment of the present application, the trunk point cloud of each tree is obtained in step S703, and the crown point cloud of each tree is obtained in step S117. The trunk point cloud and the crown point cloud of each tree are merged according to the trunk point to obtain a complete single tree point cloud, thereby realizing single tree segmentation. Fig.13 shown.
[0125] In the embodiment of the present application, the trunk points are extracted from the trunk area point cloud data, and then the trunk point cloud and the crown point cloud are determined based on the trunk points. Extracting the trunk points based on the trunk area point cloud data can reduce the situation of recognition errors, ensure the number of trees, and provide a reliable basis for subsequent accurate segmentation. Then, the trunk point cloud and the crown point cloud corresponding to the trunk point are determined based on the trunk points, which can effectively reduce segmentation errors, improve the accuracy and efficiency of single tree segmentation, avoid erroneous segmentation, and improve the single tree segmentation effect.
[0126] See also Fig.14 , another single tree segmentation method provided in an embodiment of the present application is exemplarily introduced, and the single tree segmentation method can be executed by an electronic device, including steps S141 to S146.
[0127] Step S141, obtaining the tree point cloud data to be processed.
[0128] The relevant contents of step S141 may refer to step S101 and will not be described again here.
[0129] Step S142, processing the forest point cloud data to obtain the trunk area point cloud data.
[0130] The processing of the forest point cloud data may include but is not limited to: preprocessing, elevation filtering, etc. The relevant contents of the preprocessing and elevation filtering can be referred to above and will not be repeated here.
[0131] Specifically, the electronic device may divide the forest point cloud data according to a preset height range, and use the forest point cloud data within the preset height range as trunk area point cloud data.
[0132] Step S143, processing the point cloud data of the trunk area to obtain the trunk points.
[0133] The processing of the point cloud data of the trunk region may include but is not limited to: layering operation, clustering, and extracting the center of gravity. The relevant contents of layering operation, clustering, and extracting the center of gravity can be referred to above and will not be repeated here.
[0134] Specifically, the electronic device can first obtain the trunk point cloud cluster and the center of gravity of the trunk point cloud cluster based on the point cloud data of the trunk area, and then project the center of gravity of the trunk point cloud cluster onto the horizontal plane to obtain the center of gravity point cloud cluster, and then cluster the center of gravity point cloud cluster to obtain each center of gravity cluster, and finally extract the center of gravity of each center of gravity cluster to obtain the trunk point.
[0135] In other embodiments, the electronic device may extract the center of the centroid cluster, and then use the extracted center as the trunk point, which is not specifically limited in the embodiments of the present application.
[0136] The following exemplifies a method for obtaining a tree trunk point cloud cluster and the center of gravity of a tree trunk point cloud cluster: the electronic device can first layer the tree trunk area point cloud data according to the first layer height to obtain multiple layers of tree trunk point cloud data, and then cluster the tree trunk point cloud data for each layer to obtain tree trunk point cloud clusters, and finally extract the center of gravity of each tree trunk point cloud cluster to obtain the center of gravity of the tree trunk point cloud cluster.
[0137] Step S144, determining the tree trunk point cloud corresponding to the tree trunk point.
[0138] In an embodiment of the present application, after the trunk points are determined, for each trunk point, the centroid cluster corresponding to the trunk point is determined, and then the centroid corresponding to each layer can be determined based on the centroid cluster, and the trunk point cloud cluster corresponding to the center of each layer is used as its trunk point cloud.
[0139] Specifically, the electronic device determines a centroid cluster corresponding to the trunk point, then determines a trunk point cloud cluster corresponding to each centroid in the centroid cluster, and uses the determined trunk point cloud clusters of each layer as the trunk point cloud corresponding to the trunk point.
[0140] Step S145, determining the tree crown area point cloud data based on the forest point cloud data and the tree trunk point cloud.
[0141] In an embodiment of the present application, the electronic device obtains the remaining point clouds in the forest point cloud data except for the trunk point clouds, and uses the remaining point clouds as the crown area point cloud data.
[0142] Step S146, determining the crown point cloud corresponding to the trunk point based on the trunk point and crown area point cloud data.
[0143] Specifically, the electronic device obtains the crown point cloud clusters and the centroid of the crown point cloud clusters of each layer based on the crown area point cloud data. For each layer, the centroid of the crown point cloud cluster with the smallest height value in the layer is used as the target centroid of the layer. For each layer, the height value of all the trunk points in this layer is set as the height value of the target centroid of the layer to obtain the trunk point corresponding to the layer. For the crown point cloud clusters in each layer, the distance between the centroid of the crown point cloud cluster and the trunk point corresponding to the layer is calculated. For each trunk point, the crown point cloud cluster corresponding to the smallest distance in each layer is used as the crown point cloud corresponding to the trunk point.
[0144] The following is an exemplary method for obtaining a crown point cloud cluster and the center of gravity of the crown point cloud cluster: the electronic device layers the crown area point cloud data according to the second layer height to obtain multiple layers of crown point cloud data. Then, for each layer of crown point cloud data, the crown point cloud data is clustered to obtain a crown point cloud cluster. Finally, the center of gravity of each crown point cloud cluster is extracted to obtain the center of gravity of the crown point cloud cluster.
[0145] Step S147, merging the trunk point cloud and the crown point cloud corresponding to the same trunk point to obtain a single tree point cloud.
[0146] Fig.15 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Fig.15 As shown, the electronic device 10 may include a memory 101, a processor 102, an input / output (I / O) interface 103 and a bus 104. The processor 102 is coupled to the memory 101 and the I / O interface 103 through the bus 104.
[0147] The memory 101 may include one or more random access memories (RAM) and one or more non-volatile memories (NVM). The random access memory can be directly read and written by the processor 102, and can be used to store executable programs (such as machine instructions) of the operating system or other running programs, and can also be used to store user and application data. The random access memory may include static random-access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), etc.
[0148] The non-volatile memory may also store executable programs and user and application data, etc., and may be pre-loaded into the random access memory for direct reading and writing by the processor 110. The non-volatile memory may include a disk storage device and a flash memory.
[0149] The memory 101 is used to store one or more computer programs. The one or more computer programs are configured to be executed by the processor 102. The one or more computer programs include a plurality of instructions. When the plurality of instructions are executed by the processor 102, a single wood segmentation method executed on the electronic device 10 can be implemented, such as Figure 1 Steps S101 to S118 shown, Figure 7 Steps S701 to S702 and Fig.14 Steps S141 to S147 are shown.
[0150] In other embodiments, the electronic device 10 further includes an external memory interface for connecting to an external memory to expand the storage capacity of the electronic device 10 .
[0151] The processor 102 may include one or more processing units, for example, the processor 102 may include an application processor (AP), a modem processor, a graphics processor (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Different processing units may be independent devices or integrated into one or more processors.
[0152] The processor 102 provides computing and control capabilities. For example, the processor 102 is used to execute a computer program stored in the memory 101 to implement the above-mentioned single tree segmentation method, including Figure 1 Steps S101 to S118 shown, Figure 7 Steps S701 to S702 and Fig.14 Steps S141 to S147 are shown.
[0153] The I / O interface 103 is used to provide a channel for user input or output. For example, the I / O interface 103 can be used to connect various input and output devices, such as a mouse, keyboard, touch device, display screen, etc., so that the user can enter information or visualize information.
[0154] The I / O interface 103 is also used to provide a channel for data transmission with a scanning device, wherein the scanning device may be a device for collecting forest point cloud data. The electronic device 10 may obtain the forest point cloud data collected by the scanning device through the I / O interface 103 .
[0155] The bus 104 is at least used to provide a channel for mutual communication among the memory 101 , the processor 102 , and the I / O interface 103 in the electronic device 10 .
[0156] In other embodiments, the electronic device 10 may further include a communication module, based on which a communication connection is established with a scanning device to obtain the forest point cloud data collected by the scanning device.
[0157] It is to be understood that the structure illustrated in the embodiment of the present application does not constitute a specific limitation on the electronic device 10. In other embodiments of the present application, the electronic device 10 may include more or fewer components than shown in the figure, or combine some components, or split some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.
[0158] See also Fig.16 , Fig.16 It is a structural diagram of a single wood splitting device provided in one embodiment of the present application.
[0159] In some embodiments, the single wood splitting device 160 may include a plurality of functional modules composed of computer program segments. The computer programs of the various program segments in the single wood splitting device 160 may be stored in a memory of an electronic device and executed by at least one processor to perform (see Figure 1 , Figure 7 as well as Fig.14 Description) Function of single tree segmentation.
[0160] In this embodiment, the single tree segmentation device 160 can be divided into multiple functional modules according to the functions it performs. The functional modules may include: a point cloud data acquisition module 151, a trunk area point cloud data acquisition module 152, a trunk point acquisition module 153, a trunk point cloud acquisition module 154, a crown area point cloud data acquisition module 155, a crown point cloud acquisition module 156 and a merging module 157. The module referred to in this application refers to a series of computer program segments that can be executed by at least one processor and can complete fixed functions, which are stored in a memory. In this embodiment, the functional implementation method of each module in the single tree segmentation device 160 can refer to the above definition of the single tree segmentation method, and will not be repeated here.
[0161] The point cloud data acquisition module 151 is used to acquire the tree point cloud data to be processed. The trunk area point cloud data acquisition module 152 is used to process the tree point cloud data to obtain the trunk area point cloud data. The trunk point acquisition module 153 is used to process the trunk area point cloud data to obtain the trunk points. The trunk point cloud acquisition module 154 is used to determine the trunk point cloud corresponding to the trunk point. The crown area point cloud data acquisition module 155 is used to determine the crown area point cloud data based on the tree point cloud data and the trunk point cloud. The crown point cloud acquisition module 156 is used to determine the crown point cloud corresponding to the trunk point based on the trunk point and the crown area point cloud data. The merging module 157 is used to merge the trunk point cloud and the crown point cloud corresponding to the same trunk point to obtain a single tree point cloud.
[0162] The present application also provides a computer program product, which includes a computer program. The computer program can be stored in a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the above-mentioned single tree segmentation method, including: Figure 1 Steps S101 to S118 shown, Figure 7 Steps S701 to S702 and Fig.14Steps S141 to S147 are shown.
[0163] The present application also provides a computer-readable storage medium, on which a computer program is stored. The computer program includes program instructions. The method implemented when the program instructions are executed can refer to the single wood splitting method in the above-mentioned embodiments of the present application, including: Figure 1 Steps S101 to S118 shown, Figure 7 Steps S701 to S702 and Fig.14 Steps S141 to S147 are shown.
[0164] The computer-readable storage medium may be an internal memory of the electronic device of the above-mentioned embodiment, such as a hard disk or memory of the electronic device. The computer-readable storage medium may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device.
[0165] In some embodiments, the computer-readable storage medium may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the electronic device, etc.
[0166] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0167] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0168] In the embodiments provided in the present application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0169] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0170] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.
Claims
1. A single wood splitting method, characterized in that: The method comprises: Obtain forest point cloud data to be processed; Processing the forest point cloud data to obtain tree trunk area point cloud data; Processing the point cloud data of the tree trunk area to obtain tree trunk points; Determine a tree trunk point cloud corresponding to the tree trunk point; Determine the tree crown area point cloud data according to the tree point cloud data and the tree trunk point cloud; Determine the tree crown point cloud corresponding to the tree trunk point according to the tree trunk point and the tree crown area point cloud data; The trunk point cloud and the crown point cloud corresponding to the same trunk point are merged to obtain a single tree point cloud.
2. The single wood splitting method according to claim 1, characterized in that: The processing of the point cloud data of the trunk area to obtain the trunk points includes: Obtaining a trunk point cloud cluster and a center of gravity of the trunk point cloud cluster according to the trunk area point cloud data; Projecting the centroid of the tree trunk point cloud cluster on a horizontal plane to obtain a centroid point cloud cluster; Clustering the centroid point cloud clusters to obtain various centroid clusters; The centroid of each centroid cluster is extracted to obtain the trunk point.
3. The single wood splitting method according to claim 2, characterized in that: The step of obtaining a trunk point cloud cluster and a center of gravity of the trunk point cloud cluster according to the trunk area point cloud data comprises: Layering the tree trunk area point cloud data according to the first layer height to obtain multi-layer tree trunk point cloud data; For each layer of tree trunk point cloud data, clustering the tree trunk point cloud data to obtain the tree trunk point cloud clusters; The centroid of each of the tree trunk point cloud clusters is extracted to obtain the centroid of the tree trunk point cloud cluster.
4. The single wood splitting method according to claim 2, characterized in that: Determining the tree trunk point cloud corresponding to the tree trunk point includes: Determining the centroid cluster corresponding to the trunk point; Determine a tree trunk point cloud cluster corresponding to each center of gravity in the center of gravity cluster; The determined tree trunk point cloud cluster is used as the tree trunk point cloud corresponding to the tree trunk point.
5. The single wood splitting method according to any one of claims 1 to 4, characterized in that: Determining the tree crown point cloud corresponding to the tree trunk point according to the tree trunk point and the tree crown area point cloud data comprises: Obtaining the crown point cloud cluster of each layer and the center of gravity of the crown point cloud cluster according to the crown area point cloud data; For each layer, the centroid of the tree crown point cloud cluster with the smallest height value in the layer is used as the target centroid of the layer; For each layer, the height value of each of the trunk points is set as the height value of the target center of gravity of the layer, so as to obtain the trunk point corresponding to the layer; For the tree crown point cloud cluster in each layer, calculating the distance between the centroid of the tree crown point cloud cluster and the tree trunk point corresponding to the layer; For each of the tree trunk points, the tree crown point cloud cluster corresponding to the smallest distance in each layer is used as the tree crown point cloud corresponding to the tree trunk point.
6. The single wood splitting method according to claim 5, characterized in that: The step of obtaining the tree crown point cloud cluster of each layer and the center of gravity of the tree crown point cloud cluster according to the tree crown area point cloud data comprises: Layering the tree crown area point cloud data according to the second layer height to obtain multi-layer tree crown point cloud data; For each layer of tree crown point cloud data, clustering the tree crown point cloud data to obtain tree crown point cloud clusters; The centroid of each of the tree crown point cloud clusters is extracted to obtain the centroid of the tree crown point cloud cluster.
7. The single wood splitting method according to claim 6, characterized in that: Determining the tree crown area point cloud data according to the tree point cloud data and the tree trunk point cloud comprises: Acquire the remaining point clouds in the forest point cloud data except the tree trunk point cloud; The remaining point cloud is used as the point cloud data of the crown area.
8. The single wood splitting method according to claim 1, characterized in that: The processing of the forest point cloud data to obtain the trunk area point cloud data includes: The forest point cloud data is divided according to a preset height range, and the forest point cloud data within the preset height range is used as the trunk area point cloud data.
9. An electronic device, characterized in that: The electronic device comprises a processor and a memory, and the processor is used to implement the single wood splitting method according to any one of claims 1 to 8 when executing a computer program stored in the memory.
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 single tree segmentation method according to any one of claims 1 to 8 is implemented.