CN115984359A - Ground-based laser point cloud single-tree crown volume extraction method based on spherical coordinate integral - Google Patents
Ground-based laser point cloud single-tree crown volume extraction method based on spherical coordinate integral Download PDFInfo
- Publication number
- CN115984359A CN115984359A CN202211710833.8A CN202211710833A CN115984359A CN 115984359 A CN115984359 A CN 115984359A CN 202211710833 A CN202211710833 A CN 202211710833A CN 115984359 A CN115984359 A CN 115984359A
- Authority
- CN
- China
- Prior art keywords
- crown
- point cloud
- tree
- volume
- space
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The invention discloses a method for extracting volume of a single-tree crown of a foundation laser point cloud based on spherical coordinate integration, and belongs to the technical field of forestry engineering. Firstly, crown point cloud is extracted by adopting a point cloud segmentation method of a visual optimal threshold selection strategy, so that crown point extraction and operability are improved; then projecting crown points from the three-dimensional point cloud space to a three-dimensional spherical coordinate space, and innovatively providing a new crown form fine description mode; and finally, the crown is cut into a plurality of quadrangular pyramid-shaped micro elements and integrated to realize the volume extraction of the crown, so that the extraction precision and universality are improved. The method provides a complete single tree crown volume extraction technical process based on the foundation laser point cloud, is beneficial to improving the automation and intelligence level of forestry production and operation, and has important theoretical significance and engineering application reference value for forestry scientific research and forestry production such as forest three-dimensional form simulation, forest resource investigation, aboveground biomass estimation and the like.
Description
Technical Field
The invention belongs to the technical field of forestry engineering, and particularly relates to a method for extracting the volume of a single-tree crown of a foundation laser point cloud based on spherical coordinate integration.
Background
The real-time and accurate extraction of forest structural parameters is the key to understanding of land carbon reserves and ecological system health diagnosis, wherein the volume of a crown expresses the size of the geometrical volume of a space occupied by a forest crown layer, the value of the volume can comprehensively reflect the competition degree among tree groups and the photosynthesis intensity of the trees, and the forest structural parameters are important factors for ground surface ecological monitoring research such as forest three-dimensional form simulation, forest resource investigation, ground surface biomass estimation and the like. The traditional method for measuring the volume of the single tree crown mainly adopts measuring tools such as a steel tape and a height measuring instrument, the crown width and the crown height are obtained manually, the volume of the tree crown is calculated according to an empirical model of a specific tree species, but the method is extensive in technology, the measurement precision is limited by human factors and the form of the tree crown layer, and the requirements on accuracy and efficiency of forest parameter extraction cannot be met. The ground-based laser radar measurement technology (Li light Detect i ng And ranging, liDAR) adopts a non-contact laser measurement mode to quickly acquire forest three-dimensional space structure information-point cloud data with high precision And high density, has the advantages of high scanning speed And high automation degree, and opens up a new technical means for forest crown volume extraction, so that the development of laser point cloud data-based single tree crown volume extraction research has important theoretical significance And engineering reference value.
According to the difference of the space scale of the minimum differential unit, the laser point cloud single-wood crown extraction method can be divided into three types: (1) a crown volume extraction method based on point cloud boundary detection. The volume extraction method based on point cloud boundary detection is characterized in that a forest canopy is regarded as an irregular geometric body with continuous space boundaries, three-dimensional space boundary points are detected, a triangular net with irregular space on the surface of a tree crown is constructed, and the space volume of the closed triangular net is directly calculated to obtain the volume of the tree crown. According to the method, crown point cloud contour boundary points are identified and a closed space triangular network or convex hull set is formed, the space range of a crown is expressed more completely, the difference of the space morphological distribution of different crown varieties can be considered better, however, the complexity of the crown volume extraction process is increased due to the detection of the crown contour boundary point cloud and the reconstruction of the space triangular network, the large-scale popularization and use are not facilitated, and meanwhile, the rationality of the crown boundary triangular network is to be subjected to negligence due to the factors such as point cloud shortage and uneven density; (2) a crown volume extraction method based on point cloud layering. The method comprises the steps of uniformly dividing laser point cloud into a plurality of layers along the height of a canopy based on the point cloud layering, regarding each layer as a space cylinder with an irregular bottom surface, detecting the boundary of the bottom surface by adopting an alpha-shape algorithm, calculating the area of the boundary, and accumulating the volumes of the layers to obtain the volume of the canopy. According to the method, the local crown form is equivalent to a regular frustum pyramid model, so that the difficulty of crown volume extraction is reduced, but the automation degree of crown volume extraction is reduced due to the selection of the hierarchical segmentation scale of the point cloud. Some scholars try to start with the aspects of setting the optimal segmentation scale, improving the detection of layered ground boundary points of the alpha-shape algorithm and the like, and the self-adaption degree and the precision of the extraction of the laser point cloud crown are improved to a certain extent; (3) a crown volume extraction method based on point cloud voxel segmentation. The method is characterized in that a crown volume extraction method based on point cloud voxel segmentation uniformly segments crown laser point clouds into space voxel units with a plurality of sizes mainly according to spatial form distribution of a crown layer, the effectiveness of the space voxel units is detected through the number of points in the voxel units, and the crown volume is indirectly extracted through effective voxel integration. In conclusion, the existing single-tree crown volume extraction method based on laser point cloud data has the defects of low precision, low universality and the like.
Therefore, the invention aims to project tree crown points from a three-dimensional point cloud space to a three-dimensional spherical coordinate space by using ground laser point cloud data of single trees, and realize high-precision extraction of the volume of the single tree crown by using a spherical coordinate integration method.
Disclosure of Invention
The invention provides a ground laser point cloud single tree crown extraction method based on spherical coordinate integration, aiming at the defects of lower precision, low universality and the like of the traditional laser point cloud data single tree crown volume extraction method.
The invention adopts the following technical scheme that the method for extracting the volume of the single-tree crown of the foundation laser point cloud based on the spherical coordinate integral comprises the following steps:
step (1), acquiring laser point cloud data of a single tree foundation;
performing elevation information statistics on the laser point cloud of the single tree foundation, extracting the single tree crown points by adopting visual optimal threshold elevation segmentation, and reducing misclassification and missed separation caused by unreasonable threshold setting;
step (3), mapping the tree crown points of the monomer tree from a three-dimensional point cloud space to a three-dimensional spherical coordinate space, and determining an integral infinitesimal body calculation model;
and (4) in a three-dimensional spherical coordinate space, taking the infinitesimal body model as an integral function, and taking the range of a horizontal angle and a vertical angle as an integral interval, and obtaining a monomer tree volume estimation result by adopting an integral method.
Further, the laser point cloud data of the single tree foundation is obtained in the step (1), and the complete single tree crown laser point cloud is obtained in a multi-view survey station scanning mode, so that the cloud loss phenomenon of the single tree crown points caused by factors such as shielding and sensor performance is compensated, and the integrity of the point cloud data is improved. The specific implementation steps are as follows,
(11) Taking a single tree as a center, uniformly arranging 3 foundation laser scanning survey stations around the single tree, and fixedly arranging spherical targets at the sight positions of the 3 survey stations;
(12) And (3) registering the laser point cloud data of 3 different stations by using the spherical target to obtain complete single tree laser point cloud data.
Further, in the step (2), according to the distribution characteristics that the single trees are spatially and three-dimensionally distributed and the elevation of the crown point is greater than that of the trunk point, the crown point and the trunk point of the single trees are separated by adopting threshold segmentation; according to the distribution characteristic that the number of crown points is far larger than that of trunk points, an elevation-point curve is drawn, and the inflection point of the curve is selected as the optimal elevation threshold.
The concrete implementation mode is as follows,
21 Dividing the laser point cloud height values of the single tree crown into intervals, counting the number of the laser point clouds in each interval, and drawing an elevation-point curve graph;
22 Because the single tree structure has connectivity in space and the number of crown points is far greater than that of trunk points, along the direction of increasing the elevation value, the distribution curve of the elevation-point cloud number has a trend of increasing first and then decreasing, and the inflection point of the curve change corresponds to the connection point of the crown and the trunk. Manually and visually judging the inflection point of the curve and the corresponding elevation value according to the elevation-point curve graph, wherein the elevation value corresponding to the inflection point is the optimal segmentation threshold value for crown point cloud extraction;
23 The points with the single tree point cloud elevation value larger than or equal to the optimal threshold value are marked as tree crown points, and the points with the single tree point cloud elevation value smaller than the optimal threshold value are marked as tree trunk points, so that the rapid and fine extraction of the single tree crown of the laser point cloud is realized.
Further, the specific implementation manner of step (3) is as follows:
31 In the three-dimensional point cloud space, performing de-center processing on the crown point to obtain a three-dimensional coordinate value of the crown point relative to the center of gravity point;
(x i ′,y i ′,z i ') is an arbitrary crown point P i Center of gravity normalized coordinate value in three-dimensional point cloud space, (x) i ,y i ,z i ) Is an arbitrary crown point P i Rectangular coordinates in three-dimensional point cloud space, i =1,2.
32 Selecting a crown center of gravity point as an origin of the three-dimensional spherical coordinate system, establishing a projection model between the space coordinate system and the spherical coordinate system according to the space geometric transformation relation, and mapping the crown point from the three-dimensional point cloud space to the three-dimensional spherical coordinate space.
Taking a crown center of gravity point as an origin O of a three-dimensional spatial spherical coordinate system, taking a spatial Euclidean distance of a connecting line between a crown point P and the origin O as a spherical radius R, taking a horizontal plane projection of an OP connecting line and a positive included angle of an X axis as a horizontal angle theta, and if the clockwise direction is positive, the theta is epsilon to 0,2 pi]The projection of the OP connecting line on the vertical plane and the positive included angle of the Z axis are vertical angles
Then->
According to the space geometric relationship, the projection transformation model of the tree crown points of the monomer tree from the three-dimensional point cloud space to the three-dimensional spherical coordinate space is as follows:
33 In a three-dimensional spherical coordinate space, cutting crown points into a plurality of rectangular pyramid micro-elements along a horizontal angle and a vertical angle, searching all points in each rectangular pyramid, taking the maximum value of the spherical coordinate radius of the points in the micro-elements as the height of each rectangular pyramid micro-element, and calculating the volume of each micro-element according to a space geometric volume calculation model of the rectangular pyramid.
The quadrangular pyramid infinitesimal body contains a crown point set
And j λ 1 ≤θ i ＜(j+1)λ 1 ，
Wherein j is a crown point P i Corresponding horizontal division number, l is crown point P i Corresponding vertical division number, j =1.2 ...m, l =1.2 ...n>
Is the maximum value of the coordinate radius of a point sphere in a quadrangular pyramid infinitesimal body, lambda 1 ,λ 2 The method comprises the following steps of respectively obtaining horizontal angle resolution and vertical angle resolution, respectively obtaining m and n, respectively obtaining total horizontal segmentation number and total vertical segmentation number of crown points, and according to a space rule rectangular pyramid volume calculation method, obtaining a rectangular pyramid micro-element volume dV calculation model as follows:
further, the specific implementation manner of obtaining the monomer tree volume estimation by the integration method in the step (4) is as follows:
41 The volumes of the micro elements are accumulated, and the sum is the volume extraction result of the single tree.
Has the advantages that:
the invention provides a complete ground-based laser point cloud single-tree crown volume extraction method based on spherical coordinate integration, which comprises the steps of point cloud data acquisition scheme, crown point extraction of an optimal threshold value, crown volume calculation by a spherical coordinate integration method and the like; the operability and reliability of crown point extraction are improved by adopting a visual optimal threshold selection principle; by projecting the tree crown points from the three-dimensional point cloud space to the three-dimensional spherical coordinate space and adopting the rectangular pyramid micro element model, the three-dimensional spatial form in the local range of the tree crown can be described finely, the volume extraction efficiency and precision of the laser point cloud tree crown are improved, and meanwhile, the method has better universal applicability to background environments such as different tree species, point cloud density and the like.
Drawings
Fig. 1 is a flow chart for extracting volume of a single-tree crown of a ground-based laser point cloud based on spherical coordinate integration according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of single-wood foundation laser point cloud data acquisition according to an embodiment of the present invention;
fig. 3 is a graph of a single-tree laser point cloud segmentation effect of visualized optimal threshold selection provided by an embodiment of the present invention;
fig. 4 is a schematic diagram of a relationship between a three-dimensional point cloud and a spherical coordinate system according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of an integrating micro-element according to an embodiment of the present invention;
FIG. 6 is a set of laser point cloud data for a single tree test provided by an embodiment of the present invention; (a) laser point cloud of osmanthus; (b) cedar laser point cloud; (c) camphorwood laser point cloud; (d) yunnan plain laser point cloud; e) Gingko laser point cloud; (f) cherry blossom laser point cloud;
fig. 7 is an absolute error distribution diagram of a single-tree crown extraction result according to experimental data provided by an embodiment of the present invention;
fig. 8 is a diagram illustrating a relative error distribution of the single-tree crown extraction result according to the experimental data provided in the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and experimental examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the invention.
In a natural state, the single-tree crown extends towards the space by taking the trunk as the center, the outline of the single-tree crown in a point cloud space is represented as an irregular space body with a closed boundary, but the local range can be regarded as a regular spherical curved surface, if the laser point cloud of the crown layer is converted into a three-dimensional spherical coordinate space from a three-dimensional rectangular coordinate space, and the crown layer is cut into a plurality of conical unit infinitesimal bodies which are tightly connected with each other, the volume of the crown layer can be obtained by the volume division of the conical infinitesimal bodies. Firstly, carrying out elevation distribution statistics on single-wood laser point clouds, drawing an elevation-point curve graph, and extracting canopy laser point clouds by adopting a visual optimal threshold elevation segmentation method; then defining a three-dimensional space spherical coordinate system, and projecting the single-wood canopy laser point cloud to a three-dimensional spherical coordinate space; and finally, cutting the single tree canopy points into a plurality of cone-shaped infinitesimal points along the horizontal angle and the vertical angle, and determining the volume of the single tree canopy according to the infinitesimal volume. The infinitesimal body cutting process can better express the spatial distribution characteristic of the local range of the crown outline, the infinitesimal volume division process better solves the problem of irregular description of the whole range of the crown outline, and compared with the existing laser point cloud monomer tree volume extraction method, the method has the advantages of high single-tree volume precision and good universality, and has important theoretical significance and engineering application reference value for researches such as forest three-dimensional shape simulation, forest resource investigation, aboveground biomass estimation and the like.
As shown in fig. 1, in an embodiment, a method for extracting a volume of a single tree crown based on a ground-based laser point cloud based on a spherical coordinate integral is provided, which specifically includes the following steps:
step (1), acquiring laser point cloud data of a single tree foundation;
performing elevation information statistics on the laser point cloud of the single tree foundation, and extracting single tree crown points by adopting visual optimal threshold elevation segmentation;
step (3), mapping the tree crown points of the monomer tree to a three-dimensional sphere coordinate space from a three-dimensional point cloud space, and determining an integral micro-element calculation model;
and (4) in a three-dimensional spherical coordinate space, taking the infinitesimal body model as an integral function, and taking the ranges of a horizontal angle and a vertical angle as integral intervals, and obtaining the volume estimation result of the tree crown of the monomer tree by adopting an integral method.
And (1) acquiring laser point cloud data of the single tree foundation. In the embodiment of the invention, the ground laser scanning system is influenced by factors such as ground object shielding, and the point cloud data of a single crown station is lost. In order to improve the integrity of laser point cloud, the single tree foundation laser point cloud data acquisition adopts a multi-station uniform distribution scheme: 3 scanning survey stations are uniformly distributed around a single tree as a center, spherical targets are distributed at the positions of the survey stations with a sight, and the spherical targets are used for the registration operation of point clouds of the survey stations. The multi-station uniform layout scheme is shown in figure 2.
And (2) extracting crown points by adopting an elevation segmentation method selected by a visual optimal threshold. The single tree is in a spatial three-dimensional structure, the elevation of the tree crown point is greater than that of the trunk point, the number of the trunk points is far greater than that of the trunk points, and the single-tree point cloud can be divided into the tree crown point and the trunk point by utilizing the height value of the connection point of the tree crown and the trunk. The optimal threshold is determined by adopting a visual threshold selection principle, firstly, single-tree laser point cloud elevation values are subjected to interval statistics according to the step length of 0.1m, an elevation-point curve graph is drawn, the inflection point of the elevation-point curve is judged visually, and the elevation value corresponding to the inflection point is set as the optimal threshold, as shown in an attached figure 3. Optimum threshold value H at this time threshold And =2.49m, and the individual data point cloud is divided into crown points and trunk points according to the threshold value, as shown in formulas (1-2).
The crown point discrimination principle is as follows: p is a radical of i ＝{(x i ，y i ，z i )|z i ≥H threshold } (1)
The trunk point discrimination principle is as follows: p is a radical of i ＝{(x i ，y i ，z i )|z i ＜H threshold } (2)
Step (3), projecting tree crown points from a three-dimensional point cloud space to a three-dimensional sphere coordinate space, and determining an integral infinitesimal body calculation model. Defining a three-dimensional spherical coordinate system: taking the center of gravity of the crown point as an origin O of a three-dimensional space spherical coordinate system, taking the spatial Euclidean distance between the crown point P and a connecting line of the origin as a spherical radius R, taking the horizontal plane projection of an OP connecting line and the positive included angle of an X axis as a horizontal angle theta, and specifying clockwise as positive, then theta belongs to [0,2 pi ], []The projection of the OP connecting line on the vertical plane and the positive included angle of the Z axis are vertical angles
Then->
The position relationship between the rectangular coordinate system of the three-dimensional point cloud space and the spherical coordinate system of the three-dimensional space is shown in fig. 4. According to the space geometric transformation relation, any crown point P i From its three-dimensional point cloud rectangular coordinates (x) i ,y i ,z i ) Projection to three-dimensional ball coordinate->
The specific steps for determining the integral infinitesimal body model are as follows:
31 Compute crown centroid point coordinates. The purpose of calculating the coordinates of the center of gravity point of the crown is to determine the origin of the three-dimensional spherical coordinate system of the crown, and the coordinates of the center of gravity point of the crown can be calculated from the average value of the three-dimensional space coordinates of the crown point, as shown in formula (3), wherein (x) 0 ,y 0 ,z 0 ) Is the coordinates of the center of gravity point of the crown, and N is the number of the crown points.
32 Crown point centroiding. The purpose of the crown point barycenter is to convert the crown point from three-dimensional space coordinates to coordinate values relative to the barycenter point. The process of crown point barycenter formation is shown in equation (4).
(x i ′,y i ′,z i ′)＝(x 0 -x i ，y 0 -y i ，z 0 -z i ) (4)
33 ) crown point projective transformation relationships. According to the space geometric transformation relation, a projection transformation model of the crown point from the three-dimensional point cloud space to the three-dimensional sphere coordinate space is shown as a formula (5).
34 ) crown infinitesimal models. In the three-dimensional spherical coordinate space, cutting the crown points into a plurality of quadrangular pyramid-shaped infinitesimal bodies along the horizontal direction and the vertical direction, wherein the quadrangular pyramid-shaped infinitesimal bodies contain crown point sets
And j λ 1 ≤θ i ＜(j+1)λ 1 ，/>
Wherein j is a crown point P i Corresponding horizontal division number, l is crown point P i Corresponding vertical division number, j =1.2 ..., m, l =1.2 ...n>
Is the maximum value of the coordinate radius of a point sphere in a quadrangular pyramid infinitesimal body, lambda 1 ,λ 2 The horizontal angle resolution and the vertical angle resolution are respectively, and m and n are respectively the total number of horizontal segmentation and the total number of vertical segmentation of the crown points, as shown in fig. 5. According to the method for calculating the volume of the regular rectangular pyramid, the dV calculation model of the volume of the rectangular pyramid micro element can be represented by a public expression (6) as follows:
and 4, determining the volume of the tree crown of the monomer tree by adopting a spherical coordinate integration method. Because of the number of crown points and the radius R of the tree in the body of different rectangular pyramid micro-elements jl The volume of the crown is a discontinuous variable, so that the volume of the crown is calculated by using a sum formula of the discrete variables, as shown in formula (7).
Angular resolution λ 1 ,λ 2 The method is used for inputting parameters for extracting the volume of the laser point cloud tree crown, if the value is selected to be too large, the extraction result of the volume of the tree crown is rough, and if the value is selected to be too small, the extraction efficiency of the volume of the single tree is low. And integrating the factors of precision and efficiency, wherein the angular resolution of the horizontal angle and the vertical angle is equal and is 3-5 times of the average distance of the point cloud.
In order to verify the effectiveness and feasibility of the ground-based laser point cloud single-tree crown volume extraction method based on the spherical coordinate integral, the method is verified by using a test example.
1. Test data. The test object is 6 different tree species, crown form monomer trees of osmanthus fragrans, cedar, cinnamomum camphora, yunnan magnolia, ginkgo and oriental cherry in the school zone of encyclopedia of Yunnan Master university, and the laser point cloud acquisition equipment is a lycra P40 foundation laser radar system, main parameters: the distance measurement precision is 1.2mm +10ppm; the angle measurement precision is 8'; the point position precision is 3mm @50m; the scan rate is 100 ten thousand points/second; the horizontal and vertical field angles are 360 ° and 270 °, respectively. The data acquisition adopts a multi-station scanning scheme, three scanning measuring stations are uniformly distributed around a research object as a center, a plane target is arranged in a common visual area of the three measuring stations, single-wood laser point cloud obtained by utilizing preprocessing function modules of Decaiska business software Cyc one multi-station point cloud data registration, denoising, cutting and the like is used as final experimental data, and the maximum error of the registration of the laser point cloud of two adjacent stations is 0.09m. The test data laser point clouds are shown in fig. 6 (a-f).
The selected 6 groups of foundation laser point cloud experimental data have different laser point cloud average distances, point cloud numbers and single tree heights, represent 5 different tree crown layer space forms of a space sphere, a space triangular pyramid, a space ellipsoid, a space cylinder and a space inverted triangular pyramid respectively, have universal applicability in engineering researches such as urban green land resource investigation and carbon reserves and the like, can meet experimental data requirements of feasibility and reliability analysis of a single tree crown volume extraction sphere coordinate integral method, and statistics of basic information of the experimental data laser point cloud is shown in table 1.
TABLE 1 statistical table of basic information of experimental data
2. And (5) result and analysis. Because the outline of the single-tree crown is an irregular space body, the whole space form of the single-tree crown cannot be described finely by a mathematical model, and therefore the theoretical value of the volume of the crown cannot be determined directly. Weekun et al (tree crown volume calculation method [ J ] based on three-dimensional laser scanning point cloud, 2013,44 (07): 235-240) propose a laser point cloud tree crown volume extraction method based on voxel segmentation, analyze effectiveness and reliability of an algorithm by adopting a plurality of groups of experimental data, discuss influence of voxel size selection on tree crown volume extraction precision and efficiency, and show that the smaller the voxel is, the higher the tree crown extraction precision is, and become a standard reference for evaluating the tree crown volume extraction performance of the laser point cloud data. The method is based on the laser point cloud crown volume extraction voxel segmentation method, and the estimation result is theoretical reference of the crown volume when the voxel size is 0.05 m. And (3) selecting an absolute error and a relative error as the evaluation indexes of the crown volume extraction precision as shown in a formula (8), wherein the formula represents the extracted value of the crown volume of the laser point cloud and represents a theoretical reference value of the crown volume, the absolute error describes the deviation of the crown extraction result and the theoretical value, and the relative error% describes the deviation proportion of the crown extraction result and the theoretical value.
The invention discloses a method for estimating and researching crown volume based on three-dimensional laser point cloud data by using a spherical coordinate integration method and a method 1 (Liufang, von Zhongke, yanglite, and the like) [ J]Agro-mechanical journal, 2016,47 (03): 328-334), method 2 (linson, tianina, birixin, etc.. Single tree crown volume accurate calculation of three-dimensional laser scan data [ J]The volume extraction results of the crown of point cloud data obtained by the method of 2020,45 (08): 115-122) are compared, and the two existing methods respectively representAnd (3) performing point cloud boundary detection by using a crown volume extraction method and performing point cloud layering by using a crown volume extraction method. Table 2 shows the extracted values of the volume of the laser point cloud crown and the precision statistics based on the experimental data of different methods. Comparing different laser point cloud experimental data, when the spatial form of the crown is in a sphere-like shape or a triangular pyramid shape and the outline is more regular, such as sweet osmanthus flower, cedar and cinnamomum camphora experimental data, the crown volume extraction by different methods shows better effect, the crown volume extraction precision index of the experimental data is kept in a certain range, and the maximum value of the absolute error of the crown volume extraction is 1.02m 3 The maximum value of the relative error is 7.32%, as shown in bold in table 2; when the outline of the crown is irregular, such as Yunnan magnolia and ginkgo and other test data, the fluctuation of the volume extraction precision index of the crown of the test data is large, the performance of the extraction result of the test data of the Yunnan magnolia laser point cloud is most obvious, and the maximum absolute error value is 3.46m 3 . Compared with different laser point cloud crown volume extraction methods, the maximum value of the absolute error of crown extraction based on the method is 2.33m 3 The maximum value of the relative error is 3.40%, and compared with the existing method, the method has higher extraction precision and keeps better applicability to different crown forms and tree species. The analysis reason is that the point cloud boundary detection method well considers the irregularity of the external outline of the crown by detecting the crown boundary points and generating the three-dimensional convex hull, but when the crown shape has larger spurs, the unreasonable convex hull distorts the crown volume extraction result; the volume extraction of the crown of the point cloud layering is realized by cutting the crown into a plurality of faults and regarding the faults as a regular round table, and the volume calculation of the crown only considers the form of a sampling surface of the fault and does not consider the spatial characteristic of the crown between two sampling surfaces, so that the volume result of the layer elliptic crown is smaller, and the volume extraction result of the irregular-shaped crown is larger; according to the invention, the discrete three-dimensional laser point cloud is projected to the three-dimensional spherical coordinate space, the crown layer is cut into a plurality of conical micro-elements, the difference of the crown side outline can be described more finely, the micro-element volume is divided in the spherical coordinate space, the factors such as point cloud internal loss, crown shape difference and the like are taken into account better, and higher extraction precision and universality are shown.
TABLE 2 laser point cloud crown volume extraction and precision analysis
FIGS. 7-8 show the precision evaluation index distribution curves of the crown volume extraction results of the experimental data based on the method of the present invention and the existing method, wherein the diamond-shaped labeled curve represents the evaluation curve of the crown extraction precision of the experimental data for point cloud boundary detection, the square labeled curve represents the evaluation curve of the crown extraction precision of the experimental data for the point cloud layered crown volume extraction method, and the triangle labeled curve represents the evaluation curve of the crown extraction precision of the experimental data for the method of the present invention. Aiming at different crown laser point cloud experimental data, the crown extraction absolute error and relative error distribution curves of the method are relatively changed and smooth, and the precision indexes are distributed in a certain range, so that the analysis shows that the laser point cloud single-tree crown extraction method based on the method has the advantage of high stability.
3. And (6) concluding. The volume of the crown is a key index parameter for forest investigation and forestry resource monitoring and management. Aiming at the defects that the universality of the laser point cloud crown volume extraction method is not high, the application provides a ground-based laser point cloud single-tree crown volume extraction method based on the spherical coordinate integration on the basis of analyzing the single-tree structure. A multi-group foundation laser point cloud data experiment is selected, factors such as point cloud density, tree crown shape and tree species are involved, comparison is carried out through two existing laser point cloud tree crown volume extraction methods of a point cloud boundary detection method and a point cloud layering method, and experimental results and analysis show that: the method projects the laser point cloud to the three-dimensional spherical coordinate space, and changes the tree crown volume extraction ideas of the traditional methods such as disordered point cloud layering and boundary detection; by considering the local range of the crown outline as a spherical model, a crown volume calculation spherical coordinate integral strategy based on a conical micro element is designed, factors such as point cloud internal loss, irregular crown outline shape, forest growth difference and the like are well considered, and compared with other existing methods, the method has the advantages that the accuracy and the stability of crown extraction are better, and the maximum absolute error of single tree crown volume extraction is realizedA value of 2.33m 3 The maximum value of the relative error is 3.40%.
The specific embodiments described herein are merely illustrative of the feasibility of the invention. Various modifications, additions and substitutions for the specific embodiments described may be made by those skilled in the art without departing from the scope and range of equivalents of the invention.
Claims (7)
1. The method for extracting the volume of the single tree crown of the foundation laser point cloud based on the spherical coordinate integral is characterized by comprising the following steps of:
step (1), acquiring laser point cloud data of a single tree foundation;
performing elevation information statistics on the laser point cloud of the single tree foundation, and extracting single tree crown points by adopting visual optimal threshold elevation segmentation;
step (3), mapping the tree crown points of the monomer tree from a three-dimensional point cloud space to a three-dimensional spherical coordinate space, and determining an integral infinitesimal body calculation model;
and (4) in a three-dimensional spherical coordinate space, taking the infinitesimal body model as an integral function, and taking the range of a horizontal angle and a vertical angle as an integral interval, and obtaining a single tree crown volume estimation result by adopting an integral method.
2. The ground-based laser point cloud single tree crown volume extraction method based on the sphere coordinate integral as claimed in claim 1, wherein the step (1) specifically comprises:
taking a single tree as a center, uniformly arranging 3 foundation laser scanning stations around the single tree, and mounting spherical targets at the sight positions of the 3 foundation laser scanning stations;
and (3) registering the laser point cloud data of the laser scanning survey stations of the 3 different foundations by using the spherical target to obtain complete single tree laser point cloud data.
3. The method for extracting the volume of the single tree crown based on the ground based laser point cloud integrated by the spherical coordinates as claimed in claim 1, wherein the step (2) specifically comprises:
drawing an elevation-point counting curve according to the elevation value distribution of the laser point cloud of the single tree, and setting an elevation value corresponding to an inflection point of the counting curve as an optimal threshold value;
and partitioning the point cloud by using the optimal threshold value to realize the extraction of the laser point cloud monomer tree crown, wherein points with the monomer tree point cloud elevation value larger than or equal to the optimal threshold value are partitioned into tree crown points, and points with the monomer tree point cloud elevation value smaller than the optimal threshold value are partitioned into tree trunk points.
4. The ground-based laser point cloud single-tree crown volume extraction method based on the spherical coordinate integration as claimed in claim 1, wherein the step (3) specifically comprises:
establishing a three-dimensional spherical coordinate system: taking a crown center of gravity point as an origin O of a three-dimensional space spherical coordinate system, taking a space Euclidean distance between a crown point P and an origin O connecting line as a spherical radius R, taking a horizontal plane projection of an OP connecting line and a positive included angle of an X axis as a horizontal angle theta, and specifying clockwise as positive, then theta is formed by [0,2 pi ], []The projection of the OP connecting line on the vertical plane and the positive included angle of the Z axis are vertical angles
Then->
Point cloud projection transformation: according to the space geometric relationship, the projection transformation model of the tree crown points of the monomer tree from the three-dimensional point cloud space to the three-dimensional spherical coordinate space is as follows:
wherein (x' i ,y′ i ,z′ i ) Is an arbitrary crown point P i The coordinate value is normalized at the gravity center of the three-dimensional point cloud space,
is an arbitrary crown point P i Coordinate values in a three-dimensional spherical coordinate space, i =1,2.. Wherein N and N are the total number of crown points;
crown infinitesimal body model: in the three-dimensional spherical coordinate space, cutting the crown points into a plurality of quadrangular pyramid-shaped infinitesimal bodies along the horizontal direction and the vertical direction, wherein the quadrangular pyramid-shaped infinitesimal bodies contain crown point sets
And j λ 1 ≤θ i ＜(j+1)λ 1 ，/>
Wherein j is a crown point P i Corresponding horizontal division number, l is crown point P i Corresponding vertical division number, j =1.2 ...m, l =1.2 ...n>
Is the maximum value of the coordinate radius of a point sphere in a quadrangular pyramid infinitesimal body, lambda 1 ,λ 2 The method comprises the following steps of respectively obtaining horizontal angle resolution and vertical angle resolution, wherein m and n are respectively the total number of crown point horizontal segmentation and the total number of vertical segmentation, and according to a space rule rectangular pyramid volume calculation method, a rectangular pyramid micro element volume dV calculation model is as follows:
5. the method for extracting the volume of the single tree crown of the laser point cloud based on the spherical coordinate integration as claimed in claim 4, wherein the center of gravity normalized coordinate values of the tree crown points Pi in the three-dimensional point cloud space are as follows:
(x i ,y i ,z i ) Is a crown point P i Rectangular coordinates in a three-dimensional point cloud space.
6. The method for extracting the volume of the single tree crown based on the ground based laser point cloud integrated by the spherical coordinates as claimed in claim 1, wherein the step (4) specifically comprises:
according to the radius R of the tree crown infinitesimal body jl For non-continuous variables, the volume V of the crown is calculated by summation of discrete variables:
7. the method of claim 4 or 6, wherein the horizontal angular resolution λ is a single tree crown volume of ground based laser point cloud based on spherical coordinate integration 1 And vertical angular resolution λ 2 The values of (a) are equal and are 3-5 times of the average distance of the point clouds.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211710833.8A CN115984359A (en) | 2022-12-29 | 2022-12-29 | Ground-based laser point cloud single-tree crown volume extraction method based on spherical coordinate integral |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211710833.8A CN115984359A (en) | 2022-12-29 | 2022-12-29 | Ground-based laser point cloud single-tree crown volume extraction method based on spherical coordinate integral |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115984359A true CN115984359A (en) | 2023-04-18 |
Family
ID=85962240
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211710833.8A Pending CN115984359A (en) | 2022-12-29 | 2022-12-29 | Ground-based laser point cloud single-tree crown volume extraction method based on spherical coordinate integral |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115984359A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116595800A (en) * | 2023-06-13 | 2023-08-15 | 东北林业大学 | Tree crown space distribution simulation method, electronic equipment and storage medium |
| CN117437269A (en) * | 2023-12-22 | 2024-01-23 | 深圳大学 | Tree point cloud non-rigid registration method and related equipment |
| CN117455977A (en) * | 2023-09-27 | 2024-01-26 | 杭州市交通工程集团有限公司 | Method and system for calculating stacking volume based on three-dimensional laser scanning |
-
2022
- 2022-12-29 CN CN202211710833.8A patent/CN115984359A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116595800A (en) * | 2023-06-13 | 2023-08-15 | 东北林业大学 | Tree crown space distribution simulation method, electronic equipment and storage medium |
| CN116595800B (en) * | 2023-06-13 | 2023-10-13 | 东北林业大学 | Tree crown space distribution simulation method, electronic equipment and storage medium |
| CN117455977A (en) * | 2023-09-27 | 2024-01-26 | 杭州市交通工程集团有限公司 | Method and system for calculating stacking volume based on three-dimensional laser scanning |
| CN117437269A (en) * | 2023-12-22 | 2024-01-23 | 深圳大学 | Tree point cloud non-rigid registration method and related equipment |
| CN117437269B (en) * | 2023-12-22 | 2024-04-16 | 深圳大学 | Tree point cloud non-rigid registration method and related equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115984359A (en) | Ground-based laser point cloud single-tree crown volume extraction method based on spherical coordinate integral | |
| CN107274417B (en) | Single tree segmentation method based on airborne laser point cloud aggregation relation | |
| CN106845399B (en) | Method for extracting single-tree information from LiDAR point cloud by using hierarchical clustering mode | |
| CN110490415B (en) | Building progress assessment method for visual coordination of multiple unmanned aerial vehicles | |
| CN105488770B (en) | A kind of airborne laser radar point cloud filtering method of object-oriented | |
| CN109142679B (en) | Forest soil nutrient space prediction method based on artificial neural network kriging interpolation | |
| CN100485662C (en) | Characteristic analytical method for product point clouds surface based on dynamic data access model | |
| CN110807781B (en) | Point cloud simplifying method for retaining details and boundary characteristics | |
| CN109887015A (en) | A kind of point cloud autoegistration method based on local surface feature histogram | |
| CN105976312A (en) | Point cloud automatic registering method based on point characteristic histogram | |
| CN112381861A (en) | Forest land point cloud data registration and segmentation method based on foundation laser radar | |
| CN102628944A (en) | Stratus cloud and convective cloud automatic recognition method based on Doppler radar data | |
| CN110794424B (en) | Full-waveform airborne laser radar ground feature classification method and system based on feature selection | |
| CN102955160A (en) | Three-dimensional laser radar technology based transmission line tower parameter determination method | |
| CN108562885B (en) | High-voltage transmission line airborne LiDAR point cloud extraction method | |
| CN113269825B (en) | Forest breast diameter value extraction method based on foundation laser radar technology | |
| CN102096072A (en) | Method for automatically measuring urban parts | |
| CN109408909B (en) | Three-dimensional rough surface micro-convex body fitting method | |
| Li et al. | An iterative-mode scan design of terrestrial laser scanning in forests for minimizing occlusion effects | |
| CN107908841B (en) | Three-dimensional wall surface graspable position discrimination algorithm | |
| CN111832582B (en) | Method for classifying and segmenting sparse point cloud by utilizing point cloud density and rotation information | |
| CN115147551A (en) | Method for extracting homonymous feature points on surface of monitored object by remote ground laser scanning technology | |
| CN107705344A (en) | Plant canopy model extracting method in laser scanning environment cloud data | |
| CN108460422B (en) | Submarine geomorphy kind identification method based on depth distribution feature | |
| CN109785261A (en) | A kind of airborne LIDAR three-dimensional filtering method based on gray scale volume element model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |