A method for quantitatively evaluating peach tree shape characteristics using point cloud

Abstract

The present invention provides a method for quantitatively evaluating the tree shape characteristics of peach trees using point clouds, and relates to the field of measuring the growth status of fruit trees using three-dimensional point cloud technology. The method constructs a series of indicators for quantifying the canopy characteristics of peach trees, including: step 1: ground-based laser radar collects point clouds of peach trees during dormancy and leaf curtain periods; step 2: pre-processes the collected point clouds by filtering and denoising; step 3: extracts branch structure characteristic indicators such as the number of first-level main branches, the angle between first-level main branches, the second-level branch analogy, the height position of the starting point of the second-level branch, and the height position of the center point during the dormancy period of the peach tree, and uses the indicator set [N1, #imgabs0#, R2, H s , H m ]; Step 4: Extract canopy characteristic indicators such as canopy area, aspect ratio, effective leaf area, leaf area index, coverage, and vertical distribution of effective leaf area during the leaf curtain period, using the indicator set [S, AR, ELA, LAI, C, ELA m ] said. Through these 11 characteristic indicators, the point cloud can be used to accurately express and quantitatively evaluate the canopy morphological characteristics of peach trees during the critical period.

Description

Technical Field

[0001] The present invention relates to the field of quantitative evaluation of fruit tree shape characteristics using three-dimensional point cloud technology, and specifically to evaluating peach tree shape by constructing an index of peach tree canopy characteristics using point cloud, and optimizing peach tree cultivation management and production efficiency by accurately analyzing peach tree canopy structure characteristics. Background Art

[0002] Peach trees are widely planted fruit trees in my country, with the characteristics of being light-loving and having strong branching ability. Due to the dense and intertwined canopy branches and leaves and the complex structure, there are obvious differences in the canopy characteristics of different tree shapes. The canopy structures with different morphological characteristics have an important impact on the growth, pest and disease control and yield of fruit trees. Common peach tree shapes include trunk shape, Y-shaped shape and open-heart shape. The trunk shape is compact and easy to manage, suitable for densely planted orchards in plains and hilly areas, with medium to high yields; the Y-shaped shape has good ventilation and light transmission performance, suitable for orchards in windy and rainy areas, with high yields and excellent fruit quality; the open-heart shape is open and suitable for low-density planting orchards, especially in areas with fertile soil and less precipitation, and can obtain high-yield and large fruits. In order to deeply analyze these tree shape differences, a series of indicators were constructed using lidar point cloud data to quantitatively evaluate the tree shape characteristics of peach trees. During the dormant period, the number of primary branches at level 1, the angle between primary branches at level 1 (this indicator is only for primary branches with a number greater than 2), the branch analogy at level 2, the height of the starting position of primary branches at level 2, and the height of the center point were analyzed, with the indicator set [N 1 , , R 2 , H s , H m ] is used to express the branch and trunk structural characteristics of different tree shapes; in the leaf curtain period, the crown area, aspect ratio, effective leaf area, leaf area index, coverage and vertical distribution of effective leaf area are calculated, and the index set [S, AR, ELA, LAI, C, ELA m ] indicates that the overall characteristics of the canopy and the vertical distribution characteristics of the leaf area are quantitatively described. Through these 11 characteristic indicators, the point cloud can be used to accurately express and quantitatively evaluate the canopy morphological characteristics of peach trees during the critical period.

[0003] LiDAR is a commonly used device for detecting the geometric parameters of fruit tree canopies and is used in digital orchard management. LiDAR is a non-contact method that quickly acquires high-density, high-precision three-dimensional point cloud data without causing damage to the research object. The data obtained is highly accurate and rich in information, and has great potential in extracting fruit tree structural parameters.

[0004] At present, the main method for obtaining fruit tree canopy characteristic information by LiDAR is to reconstruct the canopy point cloud in three dimensions or extract the overall structure information of the canopy. The quantitative tree model constructed by point cloud data is a collection of a large amount of raw data, which requires further calculation to describe the overall characteristic indicators of the canopy. However, there are few studies on quantitative evaluation of canopy characteristics of fruit trees of different tree shapes based on point cloud data. Summary of the invention

[0005] In view of the shortcomings of the existing technical methods, the present invention aims to provide a method for quantitatively evaluating the tree shape characteristics of peach trees through specific indicators. The method constructs a series of specific indicators, including indicators that express branch characteristics: the number of primary branches N at level 1 1 , 1st level main branch angle (only for main branches greater than 2), 2nd level branch analogy R 2 , Height position of the starting point of the second-level branch H s And the center point height position H m , with the index set [N 1 , , R 2 , H s , H m ]; Indicators expressing the total characteristics of the canopy and the vertical distribution characteristics of the leaf area: canopy area, aspect ratio, effective leaf area, leaf area index, coverage, vertical distribution of leaf area, with the indicator set [S, AR, ELA, LAI, C, ELA m These 11 characteristic indicators are used to characterize the tree shape characteristics of peach trees, and to achieve differentiated quantitative evaluation of the canopy morphological characteristics of different tree shapes.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is: by analyzing the point cloud data of dormant branches, a quantitative three-dimensional model is constructed to extract and calculate the number of primary branches N 1 , 1st level main branch angle (only for main branches greater than 2), 2nd level branch analogy R 2 , Height position of the starting point of the second-level branch H s And the height position H of the center point of the second-level branch m Indicators such as 1 , , R 2 , H s , H m ]; at the same time, the geometric parameters of the canopy are obtained through plane projection of the canopy point cloud during the leaf curtain period, contour polygon fitting, calculation of the major and minor axes of the contour circumscribed rectangle, and voxel method, and are used to extract indicators such as crown area, aspect ratio, effective leaf area, leaf area index, and coverage. The indicator set [S, AR, ELA, LAI, C, ELA m]; the crown is layered from bottom to top along the vertical direction of the tree height, and the effective leaf area of ​​each layer of plane point cloud is calculated to represent the vertical distribution of leaf area. Based on the above 11 indicators, the crown characteristics of peach trees with different tree shapes are quantitatively evaluated.

[0007] The present invention provides a method for quantitatively evaluating peach tree shape characteristics by using specific indicators, comprising the following steps:

[0008] Step 1, obtaining peach tree point cloud data by ground-based laser radar equipment during the dormant period and leaf curtain period of peach trees;

[0009] Step 2, preprocessing the peach tree point cloud data obtained in step 1, removing point clouds irrelevant to the research target and noise point clouds caused by inevitable errors in the measurement process, and obtaining the canopy point cloud data of a single peach tree;

[0010] Step 3, obtain the characteristic index of the dormant peach tree trunk through the dormant peach tree branch point cloud, and reconstruct the three-dimensional quantitative model of the dormant peach tree branch point cloud obtained by step 2 to calculate the number of primary branches N 1 , 1st level main branch angle (only for main branches greater than 2), 2nd level branch analogy R 2 , Height position of the starting point of the second-level branch H s And the height position H of the center point of the second-level branch m Indicators such as 1 , , R 2 , H s , H m ] indicates that the primary branch index expresses the tree shape, and the secondary branch index expresses the number and distribution characteristics of the functional branches that bloom and bear fruits;

[0011] Step 4: Obtain the characteristic indicators of the peach tree canopy during the leaf curtain period through the point cloud of peach branches and leaves during the leaf curtain period. The point cloud data of peach branches and leaves during the leaf curtain period are evenly divided into multiple layers along the vertical direction of the tree height. Then, the overall point cloud and the layered point cloud of each tree are projected respectively. The projected plane point cloud is processed by contour polygon fitting, calculation of the major and minor axes of the circumscribed rectangle, and voxel method to obtain the geometric parameters of the canopy, and extract indicators such as crown area, aspect ratio, effective leaf area, leaf area index, and coverage. The effective leaf area of ​​each layer of the plane point cloud during the leaf curtain period is calculated separately to obtain the distribution indicators of the crown leaf area in the vertical direction, and the indicator set [S, AR, ELA, LAI, C, ELA m ]express.

[0012] As a preferred embodiment of the present invention, the specific steps of "obtaining characteristic indexes of dormant peach tree trunks through dormant peach tree branch point clouds" in step 3 are:

[0013] Step 3-1, read the point cloud data of a single peach tree in the dormant period, and determine the trunk path by expanding the cover set from the tree base and identifying the maximum connected component;

[0014] Step 3-2, based on the connectivity change principle of the covering set, local topology analysis is used to identify bifurcation nodes and perform initial segmentation on the point cloud;

[0015] Step 3-3, analyze the tip farthest from the branch base in all the sub-segments of the initial segmentation, correct the topological structure or branch structure of the segment, and extend the segment to the farthest tip to ensure that each segment corresponds to the actual branch as accurately as possible. Determine the starting points of the 1st and 2nd level branches respectively through the connection points between the 1st level branch and the trunk and the connection points between the 2nd level branch and the 1st level main branch. According to the density and position of the point cloud, calculate its centroid to find the center position of the 2nd level branch. In this way, the angle between the 1st level main branches, the height position of the starting point of the 2nd level branch, and the height position of the center point are obtained.

[0016] In steps 3-4, least squares fitting was used for geometric reconstruction to obtain the number of first-level branches and the number of long branches (>60 cm), medium branches (>40 cm and <60 cm), and short branches (<40 cm) of the second-level branches, and the branch analogy was calculated.

[0017] As a preferred embodiment of the present invention, the specific steps of "obtaining quantitative indicators through peach trees in the leaf curtain period" in step 4 are:

[0018] Step 4-1, read the point cloud data of a single peach tree during the canopy period, calculate the maximum and minimum values ​​of the point cloud Z, set the single layer thickness to 5 cm, and segment the point cloud of the single peach tree to obtain canopy layered point clouds of 5 cm thickness at different heights;

[0019] Step 4-2, perform horizontal projection of the crown width and each layer point cloud on the XOY plane to obtain the two-dimensional projection data of the crown width and each layer point cloud, that is, the two-dimensional point cloud of the crown and the layered crown width.

[0020] Step 4-3, crown area (S), is represented by the crown polygon area formed by connecting the outer points of the canopy plane point cloud. First, extract the contour points of the canopy projection horizontal plane, then use the Douglas-Peucker algorithm to obtain the approximate contour polygon of the crown, and use the vector cross product to calculate the area of ​​the contour polygon;

[0021] Step 4-4, the crown aspect ratio (AR) is expressed by the ratio of the major axis (a) to the minor axis (b) of the smallest adjacent rectangle of the crown, and the formula is: AR=b / a.

[0022] Step 4-5, the effective leaf area (ELA) of a single peach tree is obtained by accumulating and summing the effective leaf areas of each layer. The single-layer effective leaf area is calculated by multiplying the leaf area density by the crown area. First, the leaf area density of each layer is calculated, and then the crown projection area of ​​each layer is calculated according to the calculation method of the crown area, so as to calculate the effective leaf area of ​​each layer of a single peach tree;

[0023] Step 4-6, single-layer leaf area density (LAD) is calculated by voxel method. First, the point cloud data is divided into voxel grids to determine whether there are points in the grid. The voxel units with points are assigned a value of 1, and the units without points are assigned a value of 0. The units without points are considered to be gaps in the canopy. The specific voxelization formula is as follows:

[0024] ; ; ; .

[0025] Where: i, j, k—the number of voxels in the X, Y, and Z directions of the pth layer of a single-layer point cloud; Int—rounding function; X min , Y min , Z min —The minimum value of the X, Y, and Z directions of the pth layer of a single-layer point cloud; max , Y max , Z max —The maximum value of the p-th layer of the single-layer point cloud in the X, Y, and Z directions; Δi, Δj, Δk —The voxel size in the X, Y, and Z directions; i×j×k—The total number of voxel unit grids in the p-th layer of the single-layer point cloud; n 1 (p) - is the number of voxel units containing point clouds in the pth layer of a single-layer point cloud, n T (p) — The total number of voxel units in the pth layer of a single-layer point cloud.

[0026] Steps 4-7: Calculate the effective leaf area ELA of the layered canopy m , which is used to represent the vertical distribution of leaf area in the crown. The calculation formula is as follows: ; Where: LAD m —is the leaf area density of the mth layer of point cloud, S m is the crown projection area of ​​the mth layer of point cloud.

[0027] Step 4-8, calculate the effective leaf area of ​​a single tree using the following formula: ; Where: n is the total number of layers of a single tree in the Z direction.

[0028] Step 4-9, the leaf area index (LAI) calculation formula is as follows: ;

[0029] In step 4-10, the crown cover (C) is expressed as the percentage of the canopy projection area to the canopy allocation area, and the calculation formula is as follows: Where D r and D p Indicates the row spacing and plant spacing for planting peach trees.

[0030] As a preferred embodiment of the present invention, step 5 is also included: performing example verification; specifically, it includes obtaining 11 tree shape characteristic indicators for the Y-shaped peach tree and evaluating its morphological characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A flow chart of a method for calculating a tree canopy index of a method for quantitatively evaluating peach tree shape characteristics using point clouds according to the present invention;

[0032] Figure 2 A schematic diagram of a tree canopy index of a method for quantitatively evaluating peach tree shape characteristics using point clouds according to the present invention;

[0033] Figure 3 (a) is a data collection diagram of an example of a peach tree dormancy period in a method for quantitatively evaluating peach tree shape characteristics using point clouds according to the present invention;

[0034] Figure 3 (b) is a data collection diagram of an example of peach tree leaf canopy period in a method for quantitatively evaluating peach tree shape characteristics using point cloud according to the present invention;

[0035] Figure 4 It is an example of a method for quantitatively evaluating peach tree shape characteristics using point cloud of the present invention, and a level 2 branch position height map;

[0036] Figure 5 This is a vertical distribution characteristic diagram of effective leaf area, an example of a method for quantitatively evaluating peach tree shape characteristics using point clouds in the present invention. DETAILED DESCRIPTION

[0037] In order to make the objectives, technical solutions and advantages of the present invention clearer, the implementation method of the present invention will be described in detail below in conjunction with the accompanying drawings and embodiments, so that the implementation process of how the present invention applies technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0038] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0039] The present invention takes a Y-shaped peach tree as an example to specifically illustrate the method. Figure 1-5 The present invention is described in further detail.

[0040] Step 1: During the dormant and leaf curtain period of peach trees with a row spacing of 5mx2.5m, choose a clear and windless day and do the following: From 9:00 am to 17:00 pm, take the main trunk of the peach tree as the center and extend from the ground to the east, west, south and north. l distance( l A ground-based laser radar device (Trimble X7 3D laser scanner) was placed at a height of 0.5 m (0.75 m). The horizontal angle of the instrument was set to 120°, the vertical angle was set to 292°, and the scanning mode was set to quickly acquire point cloud data.

[0041] Step 2: Use CloudCompare software to remove point clouds that are not related to the experiment and perform statistical filtering on the peach tree point cloud to remove noise.

[0042] Step 3: Read the dormant peach tree branch point cloud obtained in step 2, as shown in the attached figure. Figure 3 (a). The number of primary branches at level 1, the angle between primary branches at level 1 (this indicator is only for the number of primary branches greater than 2), the branch analogy at level 2, the height position of the starting point of primary branches at level 2, and the height position of the center point of primary branches at level 2 are calculated by using the three-dimensional modeling of laser point clouds. The indicator set [N 1 , , R 2 , H s , H m ] indicates that the specific values ​​are shown in Table 1 below:

[0043] Table 1. Peach tree trunk characteristic indexes

[0044]

[0045] See the attached data chart of the starting and center heights of all level 2 branches Figure 4 .

[0046] Step 4: Read the canopy branch and leaf point cloud obtained in step 2, as shown in the attached figure. Figure 3 (b). The canopy geometric parameters were obtained through data processing such as plane projection of peach canopy point cloud, contour polygon fitting, calculation of major and minor axes of contour circumscribed rectangle, and voxel method. The canopy area, aspect ratio, effective leaf area, leaf area index, coverage, vertical distribution characteristics of leaf area, and other indicators were calculated. The index set [S, AR, ELA, LAI, C, ELA m ] is represented by [S, AR, ELA, LAI, C], and its specific values ​​are shown in Table 2:

[0047] Table 2. Peach tree canopy characteristic indicators

[0048]

[0049] The peach tree canopy was layered from bottom to top along the vertical direction of the tree height, and the vertical distribution characteristics of the canopy leaf area (ELA) were calculated. m ), see the attached Figure 5 .

[0050] The above results show that the indicators calculated using high-precision point cloud computing can accurately reflect the canopy characteristics of peach trees with different tree shapes.

[0051] Although the present invention is described herein with reference to the illustrative embodiments of the present invention, the above embodiments are only preferred embodiments of the present invention, and the embodiments of the present invention are not limited to the above embodiments. It should be understood that those skilled in the art can design many other modifications and embodiments, which will fall within the scope and spirit of the principles disclosed in this application.

Claims

1. A method for quantitatively evaluating peach tree shape characteristics using point clouds, characterized in that: The steps include: Step 1: Obtain peach tree point cloud data during the dormant period and leaf curtain period of peach trees respectively; Step 2: Preprocess the peach tree point cloud data obtained in step 1; Step 3: Obtain the characteristic indexes of peach tree trunks in the dormant period through the dormant peach tree branch point cloud; Step 4: Obtain the canopy characteristic indexes of peach trees in the leaf curtain period through the branch and leaf point cloud of peach trees in the leaf curtain period; The "trunk characteristic index of peach tree in dormancy" in step 3 includes the number of primary branches, the angle between primary branches, the branch analogy of secondary branches, the height position of the starting point of secondary branches, and the height position of the center point of secondary branches, with the index set [N1, , R2, H s , H m ] indicates that the primary branch index expresses the tree shape, and the secondary branch index expresses the number and distribution characteristics of functional branches that bloom and bear fruits; In the step 3, "obtaining characteristic indexes of dormant peach tree trunks through dormant peach tree branch point clouds" specifically includes the following steps: Step 3-1, read the point cloud data of a single peach tree in the dormant period, and determine the trunk path by expanding the cover set from the tree base and identifying the largest connected component; Step 3-2, based on the connectivity change principle of the covering set, local topology analysis is used to identify bifurcation nodes and perform initial segmentation on the point cloud; Step 3-3, analyze the tip farthest from the branch base in all the sub-segments of the initial segmentation, correct the topological structure or branch structure of the segment, and extend the segment to the farthest tip to ensure that each segment corresponds to the actual branch as accurately as possible; determine the starting points of the first and second level branches respectively through the connection points between the first level branch and the trunk and the connection points between the second level branch and the first level main branch; calculate the center of the second level branch according to the density and position of the point cloud, and thereby obtain the angle between the first level main branches, the height position of the starting point of the second level branch, and the height position of the center point; Step 3-4, use least squares fitting to perform geometric reconstruction, obtain the number of first-level branches and the number of long branches, medium branches, and short branches of second-level branches, and calculate the branch analogy; construct an index set [N1, , R2, H s , H m ]Quantitatively express the characteristics of branches and trunks; the length of long branches is greater than 60cm, the length of medium branches is greater than 40cm and less than 60cm, and the length of short branches is less than 40cm; The "peach tree canopy characteristic indicators during the leaf curtain period" in step 4 include indicators: canopy area, aspect ratio, effective leaf area, leaf area index, coverage, vertical distribution of effective leaf area, and the indicator set [S, AR, ELA, LAI, C, ELA m ] indicates that it mainly expresses the overall canopy characteristics of peach trees and the details of vertical distribution of leaves; In the step 4, "obtaining characteristic indexes of peach tree canopy in leaf curtain period through the point cloud of peach tree branches and leaves in leaf curtain period" specifically includes the following steps: Step 4-1, read the point cloud data of a single peach tree during the canopy period, calculate the maximum and minimum values ​​of the point cloud Z, set the single layer thickness to 5 cm, and segment the point cloud of the single peach tree to obtain canopy layered point clouds of 5 cm thickness at different heights; Step 4-2, perform horizontal projection of the crown width and each layer point cloud on the XOY plane to obtain the two-dimensional projection data of the crown width and each layer point cloud, that is, the two-dimensional point cloud of the crown width and the layered crown width; Step 4-3, the crown area (S) is represented by the crown polygon area formed by connecting the outer points of the canopy plane point cloud; first extract the contour points of the canopy projection horizontal plane, then use the Douglas-Peucker algorithm to obtain the approximate contour polygon of the crown, and use the vector cross product to calculate the area of ​​the contour polygon; Step 4-4, the canopy aspect ratio (AR) is expressed by the ratio of the major axis (a) to the minor axis (b) of the canopy minimum adjacent rectangle, and the formula is: AR = b / a; Step 4-5, the effective leaf area (ELA) of a single peach tree is obtained by accumulating and summing the effective leaf areas of each layer; wherein the single-layer effective leaf area is calculated by multiplying the leaf area density by the crown area; firstly, the leaf area density of each layer is calculated, and then the single-layer crown projection area is calculated according to the calculation method of the crown area, so as to calculate the effective leaf area of ​​each layer of the single peach tree; Step 4-6, single-layer leaf area density (LAD) is calculated by voxel method. First, the point cloud data is divided into voxel grids to determine whether there are points in the grid. The voxel units with points are assigned a value of 1, and the units without points are assigned a value of 0. The units without points are considered to be gaps in the canopy. The specific voxelization formula is as follows: ； ； ； ； Where: i, j, k—the number of voxels in the X, Y, and Z directions of the pth layer of a single-layer point cloud; Int—rounding function; X min , Y min , Z min —The minimum value of the X, Y, and Z directions of the pth layer of a single-layer point cloud; max , Y max , Z max —The maximum value of the p-th layer of a single-layer point cloud in the X, Y, and Z directions; Δi, Δj, Δk —The voxel size in the X, Y, and Z directions; i×j×k—The total number of voxel unit grids in the p-th layer of a single-layer point cloud; n1(p)—The number of voxel units containing point clouds in the p-th layer of a single-layer point cloud, n T (p) — the total number of pixel units in the pth layer of a single-layer point cloud; Steps 4-7: Calculate the effective leaf area ELA of the layered canopy m , which is used to represent the vertical distribution of leaf area in the crown. The calculation formula is as follows: ; Where: LAD m —For m Leaf area density of layer point cloud, S m For the m The crown projection area of ​​the layer point cloud; Step 4-8, calculate the effective leaf area of ​​a single tree using the following formula: ; Where: n is the total number of layers of a single tree in the Z direction; Step 4-9, the leaf area index (LAI) calculation formula is as follows: ; In step 4-10, the crown cover (C) is expressed as the percentage of the canopy projection area to the canopy allocation area, and the calculation formula is as follows: Where D r and D p Indicates the row spacing and plant spacing for planting peach trees.

2. The method for quantitatively evaluating peach tree shape characteristics using point cloud according to claim 1, characterized in that: The step 1 of "obtaining peach tree point cloud data" specifically includes selecting a clear and windless sunny day during the dormant period and leaf curtain period of the peach tree, placing ground-based laser radar equipment in four directions of the east, west, south and north with the main trunk of the peach tree as the center, and obtaining the fruit tree point cloud data.

3. The method for quantitatively evaluating peach tree shape characteristics using point cloud according to claim 1, characterized in that: The "preprocessing" in step 2 specifically includes the removal of irrelevant point clouds, denoising and filtering of experimental point clouds.

4. The method for quantitatively evaluating peach tree shape characteristics using point cloud according to claim 1, characterized in that: It also includes step 5: conducting example verification; specifically, it includes obtaining 11 tree shape characteristic indicators for the Y-shaped peach tree and evaluating its morphological characteristics.

Images