Plant form phenotype information extraction method

Abstract

The invention discloses a plant form phenotype information extraction method, and belongs to the technical field of computer vision. The method comprises the following steps: providing a local neighborhood structure plant node searching and positioning algorithm, calculating a geometric distance and a direction vector angle between adjacent nodes, analyzing a local neighborhood topological structure between the nodes, and judging a node category and a branch to which the node category belongs; proposing a plant region growth communication algorithm, performing dimension reduction on a plant part mask contour of similar adjacent nodes, searching a point with the largest distance from a direction axis on a target contour along a line parallel to the direction axis, determining boundary points of a target plant part, and connecting adjacent boundary points in sequence to form a continuous curve segment, the path is close to the path of adjacent nodes in the real plant image; the operation is repeated on all adjacent nodes of the same kind in sequence, all parts of the plant are connected, the complete plant morphology is constructed, plant morphology phenotypic information is rapidly and effectively extracted, and the method has the advantages of being robust in output result, rapid and stable in convergence and the like.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and more specifically to a method for extracting plant morphological phenotype information. Background Art

[0002] Plant phenotype plays an important role in genetics, botany and agronomy. Traditional methods of measuring and analyzing plant phenotypes, such as manual observation and recording, are time-consuming and labor-intensive and cannot meet the requirements of large-scale crop phenotyping. They have certain limitations in terms of cost, performance and spatiotemporal coverage. With the rapid development of imaging technology, computing power and algorithms, computer vision has gradually changed plant phenotyping and has now become the main tool for phenotyping analysis.

[0003] Traditional machine vision technology mainly identifies and predicts plant phenotypic traits based on features such as texture and color, which requires a lot of professional knowledge. With the rapid development of computer vision, various deep learning-based models have been widely used in the field of plant phenotypic identification due to their excellence in feature extraction, which can effectively improve recognition efficiency and accuracy.

[0004] Guo et al. proposed a detection region search algorithm based on improved YOLOv5 in the Journal of Agricultural Engineering to obtain phenotypic parameters of soybean plants. This method uses the detected soybean branch area as the search area and reconstructs the soybean plant skeleton in combination with the stem node coordinate information to obtain phenotypic parameters such as plant height, stem node number, and branch number. However, this method is mainly applicable to the plant types that appear in the experiment, which limits its scope of application. In addition, the reconstructed soybean plant skeleton is not fine enough in details, and there will be large errors when measuring phenotypes such as plant height.

[0005] Fu et al. proposed a method for 3D reconstruction of bean plants using images acquired by RGB-D depth cameras in Computers and Electronics in Agriculture. This method extracts skeleton points through hierarchical clustering of point clouds, connects the main trunk skeleton points, and selects the growth path according to the morphological characteristics of the plant, ultimately achieving digital modeling of the plant. However, there is still room for improvement in the universality and robustness of this method, especially when dealing with plants with dense branches and leaves and almost no gaps between branches and leaves.

[0006] Chen et al. proposed a method based on an improved single multi-frame detector and ant colony optimization algorithm in Frontiers in Plant Science to extract phenotypic information of the whole soybean plant. This method uses an improved SSD network structure to detect and locate pods and stems, and further proposes a stem extraction method based on the ant colony optimization algorithm to construct a complete soybean plant stem. However, when the ant colony optimization algorithm processes large-scale images or complex plant structures, it may cause ants to fall into infeasible points or plan detours, affecting the efficiency of the algorithm and the accuracy of stem extraction. In addition, during the data set annotation process, the annotation range of the pod tip depends on the occlusion situation, and the unoccluded part of the stem is annotated in segments. This annotation method may introduce a certain degree of subjectivity, and different annotators may have differences, which will affect the stability of model training and the final detection results.

[0007] From the above technologies, we can see that although the existing plant morphology extraction methods have made significant progress in accuracy and efficiency, they still face some challenges, such as adaptability to specific plant types, precision of skeleton reconstruction, and measurement errors. Therefore, it is particularly important to further improve the universality of plant morphology extraction methods, reduce errors, and enhance the ability to handle complex plant structures. Summary of the invention

[0008] 1. Technical issues to be solved

[0009] In view of the problem that the current plant phenotypic information extraction methods in the prior art have certain limitations in identifying branches, stem nodes and plant heights, the present invention provides a plant morphological phenotypic information extraction method. On the one hand, all plant nodes are detected through a local neighborhood structure plant node search and positioning algorithm, and the plant node category and the branch to which it belongs are determined. On the other hand, a plant regional growth connectivity algorithm is proposed based on the plant node classification results to obtain a path close to adjacent nodes in the actual plant image, thereby realizing the connection between each branch and the trunk, constructing the plant morphology, and realizing fast and effective extraction of plant morphological phenotypic information. The method has the characteristics of robust output results, fast and stable convergence, etc.

[0010] 2. Technical solution

[0011] The purpose of the present invention is achieved through the following technical solutions.

[0012] A method for extracting plant morphological phenotype information comprises the following steps:

[0013] Collect and input plant morphology image dataset;

[0014] Based on the target detection model, the plant node detection module is configured, and the block structure of the model feature extraction layer in the backbone network of the target detection model is replaced with the EFViT structure, which is then combined with the multi-scale linear attention mechanism to form the MST-EFViT structure; all plant nodes in the plant morphological image are detected through the configured plant node detection module;

[0015] Configure the plant node classification module, traverse all plant nodes, and determine the category of the plant node and the branch to which it belongs through the local neighborhood structure plant node search and positioning algorithm;

[0016] Configure the plant growth connectivity module, and calculate the path of adjacent plant nodes in the real plant morphology image based on the plant node classification results through the plant region growth connectivity algorithm; repeat this operation for all adjacent plant nodes of the same type, connect various parts of the plant, and construct the plant morphology;

[0017] Extract and output plant morphological phenotype information based on plant morphology.

[0018] As a further improvement of the present invention, the MST-EFViT structure is expressed as:

[0019] F MST-EFViT (x) = A MST (F EFViT (x)

[0020] Among them, x represents the plant morphology image, F EFViT (x) represents the result of feature extraction of plant morphological images by EFViT structure, A MST represents the multi-scale linear attention mechanism, F MST-EFViT (x) represents the plant node obtained by extracting features from the plant morphological image using the MST-EFViT structure.

[0021] As a further improvement of the present invention, the steps of determining the category of the plant node and the branch to which it belongs by using the local neighborhood structure plant node search and positioning algorithm include:

[0022] Establish a rectangular coordinate system and store the coordinates of all plant node positions in the node array;

[0023] Traverse the node array, extract the coordinates of each plant node and its three adjacent plant nodes, calculate the angle between the plant node and the first two adjacent plant nodes and the angle between the plant node and the second two adjacent plant nodes, and determine whether the plant has branches by calculating whether the difference between the two angles meets the preset conditions;

[0024] If the plant has branches, the node array is traversed to calculate the Euclidean distance and angle between each plant node and the previous plant node, and whether the plant node is the main stem node of the plant is determined according to preset conditions; the main stem node array of the plant is screened, and the branch node arrays on the left and right sides of the main stem of the plant are divided by the region judgment function according to the coordinate information in the main stem node array of the plant; for non-main stem nodes, the characteristics of their K nearest neighbor plant nodes are analyzed to determine the branch to which they belong; if the plant has no branches, the original node array is the main stem node array of the plant.

[0025] As a further improvement of the present invention, the step of determining whether a plant has branches comprises:

[0026] The plant node position coordinates are stored in the node array Node = [X, Y], where Node represents an n×2 node array containing the center points of all prediction boxes, X represents the horizontal coordinate of the center point, Y represents the vertical coordinate of the center point, and n represents a natural number;

[0027] For plant node Node[i], calculate the angle α formed by plant node Node[i], plant node Node[i+1] and plant node Node[i+2] i And the angle α formed by plant node Node[i], plant node Node[i+1] and plant node Node[i+3] i+1 , the calculation formula is:

[0028] α i =arctan 2(y i+2 -y i ,x i+2 -+x i )-+arctan 2(y i+1 -y i , x i+1 -x i )

[0029] α i+1 =arctan 2(y i+3 -y i+1 , x i+3 -x i+1 )-arctan 2(y i+2 -y i+1 , x i+2 -x i+1 )

[0030] Where i represents a natural number, (x i ,y i ) represents the coordinates of the plant node Node[i], (x i+1 ,y i+1) represents the coordinates of the plant node Node[i+1], (x i+2 ,y i+2 ) represents the coordinates of the plant node Node[i+2], (x i+3 ,y i+3 ) represents the coordinates of the plant node Node[i+3];

[0031] Set the error angle preset value. If the angle α i and angle α i +1 is less than the preset error angle, then it is judged that the plant does not have branches, and the original node array is the main stem node array of the plant; if the angle α i and angle α i If the absolute value difference of +1 is greater than or equal to the preset error angle value, it is determined that the plant has branches.

[0032] As a further improvement of the present invention, if the plant has branches, a region judgment function is defined to delimit a rectangular region to exclude plant nodes that are not on the main stem of the plant, and the formula is expressed as:

[0033] left=min(Node[0,1],Node[-1,1])-pixel

[0034] right=max(Node[0,1],Node[-1,1])+pixel

[0035] bottom=image height ;top=0

[0036] Area=(left,right.bottom,top)

[0037] Among them, Area represents a rectangular area defined by the area judgment function, left, right, bottom and top represent the left, right, bottom and top coordinate values ​​of the rectangular area respectively, pixel represents the pixel value required to increase the width of the rectangular area from left to right, image height Indicates the height of plant morphology image.

[0038] As a further improvement of the present invention, the position coordinates of the plant nodes in the rectangular area are stored in the node array, the first plant node in the node array is selected as the initial plant main stem node, and the angle between the two consecutive plant nodes after the first plant node in the plant node array and the previous plant main stem node is calculated, and the calculation formula is:

[0039] Angle1=(Node1[j,:], Node1[j+1,:], Node1[j+2,:]), j=0, 1,..., n-3

[0040] Angle2=(Node1[j,:], Node1[j+1,:], Node1[j+3,:]), j=0, 1,..., n-3

[0041] Angle3=(Node1[j,:], Node1[j+1,:], Node1[j+4,:]), j=0, 1,..., n-3

[0042] max a1 , max p1 =Max(Angle1,Angle2,Angle3)

[0043] max a2 , max p2 =SeMax(Angle1,Angle2,Angle3)

[0044] Where n represents the length of the coordinate array of the node array, Angle1, Angle2 and Angle3 represent the three-point angle obtained by the cosine theorem, j represents the index of the current plant node, j+1 represents the index of the next plant node, where j=0, 1, ..., n-3, Max represents finding the largest angle among the three angles and corresponding points, SeMax represents finding the second largest angle among the three angles and corresponding points, max a1 Indicates the maximum angle value among Angle1, Angle2 and Angle3 calculated by the Max function. a2 Indicates the second largest angle value among Angle1, Angle2 and Angle3 calculated by SeMax function, max p1 Corresponding to max a1 The plant node position, max p2 Corresponding to max a2 The location of plant nodes;

[0045] Set the error angle, if max a1 and max a2 If the difference is less than the error angle, then calculate max p1 to the node array Node1[j+1,:] and max p2 The Euclidean distance to the node array Node1[j+1,:], the point with the smallest Euclidean distance is the main stem node of the plant. Follow this step to traverse the plant node array in sequence until all plant nodes are evaluated and the main stem node array of the plant is selected;

[0046] After determining all the main stem nodes of the plants, continue to execute the area judgment function to divide the branch node arrays on the left and right sides of the main stem of the plants, judge the plant nodes in each area according to the above steps, and store the plant branch nodes in the plant branch node array until all the plant nodes are judged and the categories of all the plant nodes and the branches they belong to are determined.

[0047] As a further improvement of the present invention, the steps of calculating the paths of adjacent plant nodes in the real plant morphology image by the plant region growth connectivity algorithm include:

[0048] Mapping the plant node information in the plant main stem node array and the plant branch node array back to the plant morphological image, each plant node is represented as a rectangular box, the center point of the rectangular box corresponds to the plant node coordinate, and the width and height of the rectangular box are determined according to the size of the plant node;

[0049] Perform mask processing on the plant morphology image, traverse the plant main stem node array and the plant branch node array, for each group of adjacent plant nodes of the same type, determine the plant part boundary between two adjacent plant nodes according to their corresponding rectangular frames, calculate the minimum boundary rectangle of the plant part boundary between two adjacent plant nodes, and obtain the path of adjacent plant nodes close to the real plant morphology image;

[0050] After determining the path of each group of adjacent plant nodes of the same type, connect the main stem nodes of each plant in turn, extract the lowest point of each branch node, connect the lowest point of each branch node in turn, and construct the plant morphology.

[0051] As a further improvement of the present invention, the method of representing each plant node as a rectangular frame includes: extracting the coordinates of each plant node from the plant main stem node array and the plant branch node array in sequence; for each plant node, setting a coordinate pair (x, y), wherein x and y represent the horizontal position and vertical position of the plant node in the plant morphological image, respectively, setting the width of the plant node to w and the height to h, and obtaining the coordinates (x, y) of the upper left corner of the rectangular frame of the plant node in the real plant morphological image; lu ,y lu ) and the lower right corner coordinate (x ru ,y ru ), determine the position and size of the rectangular frame of the plant node, and the calculation formula is:

[0052]

[0053]

[0054] Among them, xl u Represents the horizontal coordinate of the upper left corner of the rectangular box of the plant node in the real plant morphology image, y lu represents the ordinate of the upper left corner of the rectangular box of the plant node in the real plant morphology image, x ru Represents the horizontal coordinate of the lower right corner of the rectangular box of the plant node in the real plant morphology image, y ru It represents the vertical coordinate of the lower right corner of the rectangular box of the plant node in the real plant morphology image, i represents a natural number, and n represents the total number of plant nodes.

[0055] As a further improvement of the present invention, the steps of obtaining the path of adjacent plant nodes in the image close to the real plant morphology include:

[0056] The plant morphology image is processed to obtain the plant mask morphology image, and the coordinates of the upper left corners of the rectangular boxes of the same type of adjacent plant nodes in the real plant morphology image (x lu ,y lu ) and the lower right corner coordinate (x ru ,y ru ), calculate the maximum coordinates (x lmin ,y lmin )、(x rmax ,y rmax ), determine the plant part boundary between two adjacent plant nodes;

[0057] The boundary mask contour of the plant part is processed by dimensionality reduction to obtain the center point, main direction and two orthogonal direction axes of the target contour, and then it is moved up and down in the orthogonal direction of the direction axis. After the movement, the tangent point between the direction axis and the target contour is the boundary point of the minimum boundary rectangle of the target contour; the contour tangent is the direction axis after the movement, that is, y=k1x+b * , y=k2x+b * , the specific formula is:

[0058]

[0059] Among them, contour represents the set of boundary points, k1 and k2 represent the slopes of the two direction axes respectively, and b * represents the intercept of the line, b * The maximum and minimum values ​​of correspond to the points farthest from the central axis along two orthogonal directions;

[0060] Find the point with the largest distance from the direction axis along the contour line, take the arithmetic mean of the x-coordinate of the point with the largest distance from the axis to get the x-coordinate C of the center point x , take the arithmetic mean of the y coordinates of the points with the maximum distance from the axis to get the y coordinate C of the center point y , get the center point (C x ,C y), calculate the maximum off-axis distance point corresponding to the k value, determine the width and height of the target minimum bounding rectangle, and obtain the path of adjacent plant nodes in the real plant morphology image.

[0061] As a further improvement of the present invention, the steps of constructing the plant morphology include:

[0062] Determine the lowest point of each plant branch node, and for the node on the main stem of the plant, calculate the Euclidean distance between it and the lowest point of the plant branch node;

[0063] Determine two direction vectors, including the branch direction vector and the main stem direction vector of the plant, and calculate the angle of the direction vector between the lowest point of the branch and the main stem node of the plant. The calculation formula is:

[0064]

[0065] in, represents the branch direction vector of the plant, represents the main stem direction vector of the plant, θ represents the angle of the direction vector between the lowest point of the branch of the plant and the main stem node;

[0066] From the calculated Euclidean distance, the plant node with the shortest distance is selected as the candidate plant node, and it is checked whether the angle difference between the main stem direction vector and the branch direction vector of the candidate plant node is within θ. If the condition is met, the plant branch node is connected to the plant main stem node, and each plant branch node is connected in turn to construct the plant morphology.

[0067] 3. Beneficial effects

[0068] Compared with the prior art, the advantages of the present invention are:

[0069] (1) A method for extracting plant morphological phenotype information of the present invention adds an adaptive training sample selection strategy to the plant node detection module. The adaptive training sample selection strategy reduces the influence of invalid or redundant samples on the plant node detection module by screening effective samples, avoids the excessive dependence of the plant node detection module on certain samples, balances the distribution of data, and reduces the risk of overfitting the plant node detection module on the training data. In addition, a dynamic class distribution optimization algorithm is integrated into the loss function of the plant node detection module. By setting a weight factor to adjust the weight of each plant node category, the difference in the number of samples of each category in the training data is balanced, and the deep characteristics of the plant nodes are processed more accurately and balanced without increasing the computational cost.

[0070] (2) In a method for extracting plant morphological phenotype information of the present invention, the bounding box regression loss of the plant node detection module is replaced by SDTIOU Loss. SDTIOU Loss takes into account all relevant factors that affect the similarity of two bounding boxes, such as overlapping or non-overlapping areas, center point distance, and width and height deviation, and simplifies the calculation process. Using SDTIOU Loss as a measure of the similarity of two bounding boxes can effectively improve the accuracy and efficiency of bounding box regression.

[0071] (3) A method for extracting plant morphological phenotypic information of the present invention proposes a local neighborhood structure plant node search and positioning algorithm, selects an initial stem node from a preprocessed plant morphological image as the starting point of the trunk, and gradually evaluates subsequent stem nodes. The nodes on the trunk are identified by calculating the angle formed by three consecutive nodes and combining the error angle judgment. Furthermore, the Euclidean distance and angle information are used to evaluate and determine the nodes of each branch, thereby significantly improving the efficiency and accuracy of extracting plant stem-related phenotypic features.

[0072] (4) A method for extracting plant morphological phenotypic information of the present invention proposes a plant regional growth connectivity algorithm. First, for each group of adjacent nodes of the same type, the plant part boundary between two adjacent nodes is determined according to the corresponding rectangular frame, and the minimum boundary rectangle of the plant part between two adjacent nodes is calculated using the contour dimension reduction algorithm, thereby minimizing the distance error between adjacent nodes in the real plant morphological image, and obtaining a path close to the adjacent nodes in the real plant image. For branch nodes, the Euclidean distance from each point in the branch node array to all points in the trunk skeleton is calculated, and the point with the shortest distance is selected as the connection candidate point, and then the branch direction vector and the main stem direction vector are calculated. By calculating the direction vector angle between the lowest point of the branch and each node on the main stem, it is determined whether it meets the established conditions, thereby determining the connection point of each branch, and then completing the connection from the branch to the main stem, and finally constructing a complete plant morphology, and realizing the rapid and effective extraction of plant morphological phenotypic information. This has important application value in the fields of plant phenotypic research, plant growth monitoring, and precise management of crops, and has strong practicality and wide applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 This is a flow chart of a method for extracting plant morphological phenotype information according to an embodiment of the present invention;

[0074] Figure 2 This is a schematic diagram of the MST-EFViT structure of an embodiment of the present invention;

[0075] Figure 3 A schematic diagram of a dynamic class distribution optimization algorithm according to an embodiment of the present invention;

[0076] Figure 4A schematic diagram of a local neighborhood structure plant node search and positioning algorithm according to an embodiment of the present invention;

[0077] Figure 5 Schematic diagram of a plant region growth connectivity algorithm according to an embodiment of the present invention. DETAILED DESCRIPTION

[0078] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0079] Example

[0080] like Figure 1 As shown, a method for extracting plant morphological phenotype information provided by this embodiment. The method includes: collecting and inputting a plant morphological image dataset; constructing a plant node detection module based on a target detection model, replacing the block structure of the model feature extraction layer in the target detection model backbone network with an EFViT structure, and then combining it with a multi-scale linear attention mechanism to form an MST-EFViT structure; detecting all plant nodes in the plant morphological image through the plant node detection module; constructing a plant node classification module, traversing all plant nodes, and determining the plant node category and its branch through a local neighborhood structure plant node search and positioning algorithm; constructing a plant growth connectivity module, and calculating the path of adjacent plant nodes in the real plant morphological image through a plant regional growth connectivity algorithm according to the plant node classification result; repeating this operation for all adjacent plant nodes of the same type, connecting various parts of the plant, and constructing a plant morphology; extracting and outputting plant morphological phenotype information according to the plant morphology.

[0081] Specifically in this embodiment, plant morphology images are collected, and a plant morphology image data set is constructed and input.

[0082] Furthermore, a plant node detection module is constructed based on the target detection model, and the block structure of the model feature extraction layer in the backbone network of the target detection model is replaced with the EFViT structure, which is then combined with the multi-scale linear attention mechanism to form the MST-EFViT structure. Figure 2 As shown, it is a schematic diagram of the MST-EFViT structure. In this embodiment, the MST-EFViT structure is represented as:

[0083] F MST-EFViT (x) = A MST (F EFViT (x)

[0084] Among them, x represents the plant morphology image, F EFViT (x) represents the result of feature extraction of plant morphological image x by EFViT structure. It should be noted that F EFViT The (x) function contains all the feature extraction and transformation steps in the EFViT structure.MST represents the multi-scale linear attention mechanism, F MST-EFViT (x) represents the plant node obtained by extracting features from the plant morphological image x using the MST-EFViT structure.

[0085] An adaptive training sample selection strategy is introduced in the target detection model training process of the plant node detection module. The adaptive training sample selection strategy selects valid samples from the original data to reduce the impact of invalid or redundant samples on the target detection model, thereby reducing the risk of overfitting the target detection model on the training data.

[0086] In this embodiment, the adaptive training sample selection strategy is expressed as:

[0087]

[0088]

[0089] σIoU threshold =IoU avg +x×σIoU

[0090] Among them, i represents a natural number, N represents the total number of anchor points, and IoU avg represents the average value of the intersection-over-union ratio of N anchor points and the true bounding box, IoU(i) represents the intersection-over-union ratio of the i-th anchor point and the true bounding box, σ represents the standard deviation of the intersection-over-union ratio, and x represents the empirical parameter that determines which anchor points will be selected as positive samples. threshold Represents the threshold, which is used to determine whether a candidate anchor point can be selected as a positive sample. When the IoU between the anchor point and the true bounding box is greater than IoU threshold , the anchor point is selected as a positive sample, otherwise it is a negative sample.

[0091] like Figure 3 As shown, in this embodiment, a dynamic class distribution optimization algorithm is integrated into the loss function of the target detection model, and a weight factor is set to adjust the class weight to balance the number difference of samples of each category in the training data, so as to process the deep features of the plant nodes more accurately and balancedly without increasing the computational cost. In this embodiment, the weight factor is expressed as:

[0092]

[0093] Among them, W c represents the weight factor, pc represents the proportion of the cth class, that is, the proportion of the number of samples to the total samples, and β represents a hyperparameter, which is in the range of [0,1] and is used to adjust the strength of the weight. In this embodiment, when pc is very small, that is, the class is rare, the weight factor W c will increase, thereby increasing the loss value of the category; conversely, when the category is very common, the weight factor W cWill decrease.

[0094] It is worth noting that, in this embodiment, SDTIOU Loss is used as the bounding box regression loss of the plant node detection module. Therefore, the bounding box regression loss of the plant node detection module is replaced by SDTIOU Loss. SDTIOU Loss takes into account all relevant factors that affect the similarity of two bounding boxes, such as overlapping or non-overlapping areas, center point distance, and width and height deviation, and simplifies the calculation process. Using SDTIOU Loss as a measure of the similarity of two bounding boxes can effectively improve the accuracy and efficiency of bounding box regression. In this embodiment, the bounding box regression loss is expressed as:

[0095]

[0096]

[0097]

[0098] Among them, x c Indicates the horizontal coordinate of the center point of the prediction box, y c Represents the ordinate of the center point of the prediction box, Indicates the horizontal coordinate of the center point of the target frame, represents the ordinate of the center point of the target box, t represents the general term for the width and height of the target box, which is used to calculate the shape loss, w represents the width of the prediction box, h represents the height of the prediction box, γ represents the dynamic parameter, which is used to adjust the distance loss, ρt represents the predicted value in the regression process, Λ represents the angle loss, which is used to calculate the minimum angle between the line connecting the center points of the prediction box and the real box and the coordinate axis, ρ x Represents the relative deviation between the center point of the predicted box and the center point of the real box in the x direction, ρ y Indicates the relative deviation between the center point of the predicted box and the center point of the real box in the y direction, w c represents the width of the minimum bounding box covering the target box and the predicted box, h c represents the height of the minimum bounding box covering the target box and the predicted box, SDTIoU represents the bounding box regression loss, IoU represents the intersection over union ratio, and Δ and Ω both represent additional terms for considering the angular difference of the center point and aspect ratio of the bounding box.

[0099] In this embodiment, after the plant node detection module is constructed based on the target detection model, the plant morphology image is input into the plant node detection module, and all plant nodes in the plant morphology image are detected by the plant node detection module.

[0100] like Figure 4As shown, a plant node classification module is constructed to traverse all plant nodes and determine the plant node category and its branch through the local neighborhood structure plant node search and positioning algorithm. The specific steps include:

[0101] S11. Establish a rectangular coordinate system and store the position coordinates of all plant nodes in the node array Node=[X,Y], where Node represents an n×2 node array containing the center points of all prediction boxes, X represents the horizontal coordinate of the center point, and Y represents the vertical coordinate of the center point.

[0102] S12. Traverse the node array, extract the rectangular coordinates of each plant node and its three adjacent plant nodes, calculate the angle between the plant node and the first two adjacent plant nodes and the angle between the plant node and the last two adjacent plant nodes, calculate whether the difference between the two angles meets the preset conditions, and determine whether the plant has branches.

[0103] Specifically, the coordinates of each group of three consecutive plant nodes in the node array are extracted, and for each plant node Node[i], the angle α formed by it and the two adjacent plant nodes Node[i+1] and Node[i+2] is calculated. i And the angle α formed by Node[i], Node[i+1] and Node[i+3] i+1 , and its calculation formula is:

[0104] α i =arctan 2(y i+2 -y i , x i+2 -x i )-arctan 2(y i+1 -y i , x i+1 -x i )

[0105] α i+1 =arctan 2(y i+3 -y i+1 , x i+3 -x i+1 )-arctan 2(y i+2 -y i+1 , x i+2 -x i+1 )

[0106] Among them, (x i ,y i ) represents the coordinates of the current plant node Node[i], (x i+1 ,y i+1 ) represents the coordinates of the plant node Node[i+1], (x i+2 ,yi+2 ) represents the coordinates of the plant node Node[i+2], (x i+3 ,y i+3 ) represents the coordinates of the plant node Node[i+3].

[0107] S13. If the plant has branches, traverse the node array, calculate the Euclidean distance and angle between each plant node and the previous plant main stem node, and determine whether the plant node is the plant main stem node according to preset conditions.

[0108] Specifically, set the error angle preset value Angle k , if the angle α i and angle α i The absolute value difference of +1 is less than the preset error angle Angle k , then traverse all elements in the node array in turn. If all elements meet the preset conditions, it is determined that the plant has no branches, and the node array Node is the main stem node array. If there is an angle α i and angle α i +1 The absolute value difference is greater than or equal to the error angle preset value Angle k If the plant has branches,

[0109] S14. Filter and obtain the main stem node array of the plant, and divide the branch node arrays on the left and right sides of the main stem of the plant through the area judgment function according to the coordinate information in the determined main stem node array of the plant; for non-main stem nodes of the plant, analyze the characteristics of its K nearest neighboring plant nodes to determine the branch to which it belongs.

[0110] Specifically, if the plant has branches, a region judgment function is defined to delimit a rectangular region to exclude plant nodes that are not on the main stem. The formula is expressed as:

[0111] left=min(Node[0,1],Node[-1,1])-pixel

[0112] right=max(Node[0,1],Node[-1,1])+pixel

[0113] bottom=image height ;top=0

[0114] Area=(left,right.bottom,top)

[0115] Among them, Area represents the rectangular area delineated by the area judgment function, left, right, bottom, and top represent the left, right, bottom, and top coordinate values ​​of the rectangular area respectively, and pixel represents the pixel value required to increase the width of the rectangular area from left to right.

[0116] S15. The node coordinates of the rectangular area are stored in the node array Node1. The first node in the node array is selected as the initial main stem node. The initial operation begins. The angle Angle between the two consecutive nodes after the first node of the node array and the previous main stem node is calculated. The calculation formula is:

[0117] Angle1=(Node1[j,:], Node1[j+1,:], Node1[j+2,:]), j=0, 1,..., n-3

[0118] Angle2=(Node1[j,:], Node1[j+1,:], Node1[j+3,:]), j=0, 1,..., n-3

[0119] Angle3=(Node1[j,:], Node1[j+1,:], Node1[j+4,:]), j=0, 1,..., n-3

[0120] max a1 , max p1 =Max(Angle1,Angle2,Angle3)

[0121] max a2 , max p2 =SeMax(Angle1,Angle2,Angle3)

[0122] Where n is the length of the coordinate array of Node1, Angle1, Angle2, and Angle3 are the angles of three points obtained by the cosine theorem, j is the index of the current node, and j+1 is the index of the next node, where j=0, 1, ..., n-3 (n is the length of the coordinate array of Node1), Max means finding the largest angle among the three angles and corresponding points, SeMax means finding the second largest angle among the three angles and corresponding points, and max al Indicates the maximum angle value among Angle1, Angle2, and Angle3 calculated by the Max function. a2 Indicates the second largest angle value among Angle1, Angle2, and Angle3 calculated by the SeMax function, max p1 Corresponding to max a1 The node position, max p2 Corresponding to max a2 The node location.

[0123] S16. If max a1 and max a2If the difference is less than the error angle a, then calculate max p1 to Node1[j+1,:] and max p2 The Euclidean distance to Node1[j+1,:], the point with the smallest distance is the main stem node. Follow this step to traverse the node array Node1 in sequence until all nodes are evaluated, and the main stem node array MNode is screened out.

[0124] S17. After all the main stem nodes are determined, continue to execute the region judgment function to divide the branch node arrays on the left and right sides of the main stem, judge the nodes in each region according to the above steps, and store the branch nodes in the corresponding branch array BNode j In , j represents the corresponding plant branch number, until all nodes are judged, the categories of all plant nodes and their branches are determined.

[0125] Therefore, in this embodiment, through the local neighborhood structure plant node search and positioning algorithm, the initial stem node is selected from the preprocessed plant morphological image as the starting point of the trunk, and the subsequent stem nodes are gradually evaluated. The nodes on the trunk are identified by calculating the angle formed by three consecutive nodes and combining the error angle judgment. Furthermore, the Euclidean distance and angle information are used to evaluate and determine the nodes of each branch, thereby significantly improving the efficiency and accuracy of extracting plant stem-related phenotypic characteristics.

[0126] like Figure 5 As shown in the figure, a plant growth connectivity module is constructed. According to the plant node classification results, the path of adjacent plant nodes in the real plant morphology image is calculated by the plant region growth connectivity algorithm. This operation is repeated for all adjacent plant nodes of the same type, and various parts of the plant are connected to obtain the plant morphology. The specific steps include:

[0127] S21. Map the information of each plant node in the plant main stem node array and the plant branch array back to the plant morphological image, each plant node is represented by a rectangular box, the center point of the rectangular box corresponds to the coordinates of the plant node, and the width and height of the rectangular box are determined according to the actual size of the plant node.

[0128] Specifically, the coordinates of each plant node are extracted from the plant main stem node array and the plant branch node array in turn. For each plant node, a coordinate pair (x, y) is set, where x and y represent the horizontal position and vertical position of the plant node in the plant morphological image, respectively. Assuming that the width of each plant node is w and the height is h, the coordinates of the upper left corner of the rectangular box of each plant node in the real plant morphological image (x lu ,y lu ) and the lower right corner coordinate (x ru ,y ru), and then determine the position and size of the rectangular frame of each plant node, and the calculation formula is:

[0129]

[0130]

[0131] Among them, x lu Represents the horizontal coordinate of the upper left corner of the rectangular box of the plant node in the real plant morphology image, y lu Represents the ordinate of the upper left corner of the rectangular box of the plant node in the real plant morphology image, x ru Represents the horizontal coordinate of the lower right corner of the rectangular box of the plant node in the real plant morphology image, y ru It represents the vertical coordinate of the lower right corner of the rectangular box of the plant node in the real plant morphology image, i represents a natural number, and n represents the total number of plant nodes.

[0132] S22. Perform mask processing on the plant morphology image, traverse the plant main stem node array and the plant branch array, and for each group of adjacent plant nodes of the same type, determine the plant part boundary between two adjacent plant nodes according to their corresponding rectangular frames, and use the contour dimension reduction algorithm to calculate the minimum boundary rectangle of the plant part between two adjacent plant nodes, and obtain the path of adjacent plant nodes close to the real plant morphology image. The specific steps include:

[0133] S221. Convert the plant morphology image from the RGB color space to the HSV color space, define the corresponding HSV range, obtain the plant mask morphology image, and then extract the coordinates (x lu ,y lu ) and the lower right corner coordinate (x ru ,y ru ), calculate the maximum coordinates (x lmin ,y lmin )、(x rmax ,y rmax ), thereby determining the plant part boundary between two adjacent plant nodes.

[0134] S222. Use the existing principal component analysis technology to perform dimensionality reduction processing on the boundary mask contour of the plant part to obtain the center point and main direction of the target contour and two orthogonal direction axes. It should be noted that the direction axis is a linear function that moves up and down in the orthogonal direction of the direction axis. After the movement, the tangent point between the direction axis and the contour is the point farthest from the center axis and is also the boundary point of the minimum boundary rectangle of the contour. At this time, the tangent line of the contour is the direction axis after the movement, that is, y=k1x+b * , y=k2x+b * , the specific formula is:

[0135]

[0136] Among them, contour represents the set of boundary points, k1 and k2 represent the slopes of the two direction axes respectively, and b * represents the intercept of the line, b * The maximum and minimum values ​​of correspond to the points farthest from the central axis along two orthogonal directions, respectively.

[0137] S223. Find the point with the largest distance from the direction axis along the contour line points, a total of four points, based on these four points, take the arithmetic average of the x-coordinates of the four points with the largest distance from the axis to obtain the x-coordinate C of the center point x , take the arithmetic mean of the y coordinates of the four points to get the y coordinate C of the center point y , get the center point (C x ,C y ), and then calculate the maximum off-axis distance point corresponding to each k value, and then determine the width and height of the target minimum bounding rectangle, and finally obtain the path of adjacent plant nodes in the real plant morphology image.

[0138] S23. After determining the path of each group of similar plant nodes of the same type, connect the main stem nodes of each plant in sequence, extract the lowest point of each branch, find the plant stem node in the main stem array that is close to it and has a suitable angle, and connect the plant nodes of each branch in sequence to construct a complete plant morphology. The specific steps include:

[0139] S231. Determine the lowest point of each branch, that is, the first element of each branch array, and for each node on the main stem, calculate the Euclidean distance between it and the lowest point of the branch. It should be noted that in plant morphological image processing, a bottom-up scanning method is usually used to store node coordinates, so the first element is the lowest point of each branch.

[0140] S232. Determine two direction vectors, the branch direction vector is the vector from the lowest point of the branch to other nodes of the branch, and the main stem direction vector is the vector from the lowest point of the branch to the nodes on the main stem, and calculate the angle of the direction vector between the lowest point of the branch and each node on the main stem, and the calculation formula is:

[0141]

[0142] in, represents the branch direction vector, represents the main stem direction vector, and θ represents the angle between the branch direction vector and the main stem direction vector.

[0143] S233. From the calculated distances, select the node with the shortest distance as the candidate node, and check whether the angle difference between the direction vector of the candidate node and the direction vector of the branch is within the Angel L = 45°, if all conditions are met, the corresponding branch node should be connected to the main stem node, and each branch node array is connected in turn to construct a complete plant skeleton morphology.

[0144] Therefore, in this embodiment, the plant region growth connectivity algorithm is used to first determine the plant part boundary between two adjacent nodes according to the corresponding rectangular frame for each group of adjacent nodes of the same type, and the contour dimension reduction algorithm is used to calculate the minimum boundary rectangle of the plant part between two adjacent nodes, thereby minimizing the distance error between adjacent nodes in the real plant morphological image, and obtaining a path close to the adjacent nodes in the real plant image. For branch nodes, the Euclidean distance from each point in the branch node array to all points in the trunk skeleton is calculated, and the point with the shortest distance is selected as the connection candidate point, and then the branch direction vector and the main stem direction vector are calculated. By calculating the direction vector angle between the lowest point of the branch and each node on the main stem, it is determined whether it meets the established conditions, thereby determining the connection point of each branch, and then completing the connection from the branch to the main stem, and finally constructing a complete plant morphology, and realizing the rapid and effective extraction of plant morphological phenotypic information. This has important application value in the fields of plant phenotypic research, plant growth monitoring, and precise management of crops, and has strong practicality and wide applicability.

[0145] Finally, plant morphology phenotype information is extracted and output according to the plant morphology. In this embodiment, the plant phenotype information includes the number of internodes, internode length, number of branches, branch angles, plant height, and main stem length. Therefore, the plant morphology phenotype information extraction method provided in this embodiment can realize the accurate measurement of plant morphology phenotype information, and has important application value in the fields of plant phenotype research, plant growth monitoring, etc.

[0146] The above schematically describes the invention and its implementation methods, which is not restrictive. Without departing from the spirit or basic features of the invention, the invention can be implemented in other specific forms. What is shown in the accompanying drawings is only one of the implementation methods of the invention. The actual structure is not limited thereto, and any figure mark in the claims should not limit the claims involved. Therefore, if a person of ordinary skill in the art is inspired by it, without departing from the purpose of the invention, a structural method and an embodiment similar to the technical solution are designed without creativity, which should all belong to the protection scope of the present invention. In addition, the word "including" does not exclude other elements or steps, and the word "one" before the element does not exclude the inclusion of "multiple" elements. The multiple elements stated in the product claim can also be implemented by one element through software or hardware. The words first, second, etc. are used to indicate the name, and do not indicate any specific order.

Claims

1. A method for extracting plant morphological phenotype information, comprising the following steps: Collect and input plant morphology image dataset; Based on the target detection model, the plant node detection module is configured, and the block structure of the model feature extraction layer in the backbone network of the target detection model is replaced with the EFViT structure, which is then combined with the multi-scale linear attention mechanism to form the MST-EFViT structure; all plant nodes in the plant morphological image are detected through the configured plant node detection module; Configure the plant node classification module, traverse all plant nodes, and determine the category of the plant node and the branch to which it belongs through the local neighborhood structure plant node search and positioning algorithm; Configure the plant growth connectivity module, and calculate the path of adjacent plant nodes in the real plant morphology image based on the plant node classification results through the plant region growth connectivity algorithm; repeat this operation for all adjacent plant nodes of the same type, connect various parts of the plant, and construct the plant morphology; Extract and output plant morphological phenotype information based on plant morphology.

2. The method for extracting plant morphological phenotype information according to claim 1, characterized in that: The MST-EFViT structure is represented as: F MST-EFViT (x)=A MST (F EEFViT (x)) Among them, x represents the plant morphology image, F EFViT (x) represents the result of feature extraction of plant morphological images by EFViT structure, A MST represents the multi-scale linear attention mechanism, F MST-EFViT (x) represents the plant node obtained by extracting features from the plant morphological image using the MST-EFViT structure.

3. The method for extracting plant morphological phenotype information according to claim 2, characterized in that: The steps of determining the type of plant nodes and the branch to which they belong by using the local neighborhood structure plant node search and positioning algorithm include: Establish a rectangular coordinate system and store the coordinates of all plant node positions in the node array; Traverse the node array, extract the coordinates of each plant node and its three adjacent plant nodes, calculate the angle between the plant node and the first two adjacent plant nodes and the angle between the plant node and the second two adjacent plant nodes, and determine whether the plant has branches by calculating whether the difference between the two angles meets the preset conditions; If the plant has branches, the node array is traversed to calculate the Euclidean distance and angle between each plant node and the previous plant node, and whether the plant node is the main stem node of the plant is determined according to preset conditions; the main stem node array of the plant is screened, and the branch node arrays on the left and right sides of the main stem of the plant are divided by the region judgment function according to the coordinate information in the main stem node array of the plant; for non-main stem nodes, the characteristics of their K nearest neighbor plant nodes are analyzed to determine the branch to which they belong; if the plant has no branches, the original node array is the main stem node array of the plant.

4. The method for extracting plant morphological phenotype information according to claim 3, characterized in that: The step of determining whether the plant has branches comprises: The plant node position coordinates are stored in the node array Node = [X, Y], where Node represents an n×2 node array containing the center points of all prediction boxes, X represents the horizontal center point, Y represents the vertical center point, and n represents a natural number; For plant node Node[i], calculate the angle α formed by plant node Node[i], plant node Node[i+1] and plant node Node[i+2] i And the angle α formed by plant node Node[i], plant node Node[i+1] and plant node Node[i+3] i+1 , the calculation formula is: a i arctan 2(and i+2 -and i ,x i+2 -x i )-arctan 2(and i+1 -and i ,x i+1 -x i ) a i+1 arctan 2(and i+3 -and i+1 ,x i+3 -x i+1 )-arctan 2(and i+2 -and i+1 ,x i+2 -x i+1 ) Where i represents a natural number, (x i ,y i ) represents the coordinates of the plant node Node[i], (X i+1 ,y i+1 ) represents the coordinates of the plant node Node[i+1], (x i+2 ,y i+2 ) represents the coordinates of the plant node Node[i+2], (x i+3 ,y i+3 ) represents the coordinates of the plant node Node[i+3]; Set the error angle preset value. If the angle α i and angle α i +1 is less than the preset error angle, then it is judged that the plant does not have branches, and the original node array is the main stem node array of the plant; if the angle α i and angle α i If the absolute value difference of +1 is greater than or equal to the preset error angle value, it is determined that the plant has branches.

5. The method for extracting plant morphological phenotype information according to claim 4, characterized in that: If the plant has branches, define a region judgment function to delimit a rectangular area to exclude plant nodes that are not on the main stem of the plant. The formula is expressed as: left=min(Node[0,1],Node[-1,1])-pixel right=max(Node[0,1],Node[-1,1])+pixel bottom=image height ;top=0 Area=(left,right.bottom,top) Among them, Area represents a rectangular area defined by the area judgment function, left, right, bottom and top represent the left, right, bottom and top coordinate values ​​of the rectangular area respectively, pixel represents the pixel value required to increase the width of the rectangular area from left to right, image height Indicates the height of plant morphology image.

6. The method for extracting plant morphological phenotype information according to claim 5, characterized in that: The coordinates of the plant nodes in the rectangular area are stored in the node array, the first plant node in the node array is selected as the initial plant main stem node, and the angle between the two consecutive plant nodes after the first plant node in the plant node array and the previous plant main stem node is calculated. The calculation formula is: Angle1=(Node1[j,:], Node1[j+1,:], Node1[j+2,:]), j=0, 1,..., n-3 Angle2=(Node1[j,:], Node1[j+1,:], Node1[j+3,:]), j=0, 1, ..., n-3 Angle3=(Node1[j,:], Node1[j+1,:], Node1[j+4,:]), j=0, 1,..., n-3 max a1 ,max p1 =Max(Angle1,Angle2,Angle3) max a2 ,max p2 =SeMax(Angle1,Angle2,Angle3) Where n represents the length of the coordinate array of the node array, Angle1, Angle2 and Angle3 represent the three-point angle obtained by the cosine theorem, j represents the index of the current plant node, j+1 represents the index of the next plant node, where i=0, 1, ..., n-3, Max represents finding the largest angle among the three angles and corresponding points, SeMax represents finding the second largest angle among the three angles and corresponding points, max a1 Indicates the maximum angle value among Angle1, Angle2 and Angle3 calculated by the Max function. a2 Indicates the second largest angle value among Angle1, Angle2 and Angle3 calculated by SeMax function, max p1 Corresponding to max a1 The plant node position, max p2 Corresponding to max a2 The location of plant nodes; Set the error angle, if max a1 and max a2 If the difference is less than the error angle, then calculate max p1 to the node array Node1[j+1,:] and max p2 The Euclidean distance to the node array Node1[j+1,:], the point with the smallest Euclidean distance is the main stem node of the plant. Follow this step to traverse the plant node array in sequence until all plant nodes are evaluated and the main stem node array of the plant is selected; After determining all the main stem nodes of the plants, continue to execute the area judgment function to divide the branch node arrays on the left and right sides of the main stem of the plants, judge the plant nodes in each area according to the above steps, and store the plant branch nodes in the plant branch node array until all the plant nodes are judged and the categories of all the plant nodes and the branches they belong to are determined.

7. The method for extracting plant morphological phenotype information according to claim 6, characterized in that: The steps of calculating the paths of adjacent plant nodes in the real plant morphology image by using the plant region growth connectivity algorithm include: Mapping the plant node information in the plant main stem node array and the plant branch node array back to the plant morphological image, each plant node is represented as a rectangular box, the center point of the rectangular box corresponds to the plant node coordinate, and the width and height of the rectangular box are determined according to the size of the plant node; Perform mask processing on the plant morphology image, traverse the plant main stem node array and the plant branch node array, for each group of adjacent plant nodes of the same type, determine the plant part boundary between two adjacent plant nodes according to their corresponding rectangular frames, calculate the minimum boundary rectangle of the plant part boundary between two adjacent plant nodes, and obtain the path of adjacent plant nodes close to the real plant morphology image; After determining the path of each group of adjacent plant nodes of the same type, connect the main stem nodes of each plant in turn, extract the lowest point of each branch node, connect the lowest point of each branch node in turn, and construct the plant morphology.

8. The method for extracting plant morphological phenotype information according to claim 7, characterized in that: The method of representing each plant node as a rectangular frame includes: extracting the coordinates of each plant node from the plant main stem node array and the plant branch node array in sequence; for each plant node, setting a coordinate pair (x, y), wherein x and y represent the horizontal position and vertical position of the plant node in the plant morphological image, respectively, setting the width of the plant node to w and the height to h, and obtaining the coordinates of the upper left corner of the rectangular frame of the plant node in the real plant morphological image (x lu ,y lu ) and the lower right corner coordinate (x ru ,y ru ), determine the position and size of the rectangular frame of the plant node, and the calculation formula is: Among them, x lu Represents the horizontal coordinate of the upper left corner of the rectangular box of the plant node in the real plant morphology image, y lu Represents the ordinate of the upper left corner of the rectangular box of the plant node in the real plant morphology image, x ru Represents the horizontal coordinate of the lower right corner of the rectangular box of the plant node in the real plant morphology image, y ru It represents the vertical coordinate of the lower right corner of the rectangular box of the plant node in the real plant morphology image, i represents a natural number, and n represents the total number of plant nodes.

9. The method for extracting plant morphological phenotype information according to claim 8, characterized in that: The steps of obtaining a path close to adjacent plant nodes in the real plant morphology image include: The plant morphology image is processed to obtain the plant mask morphology image, and the coordinates of the upper left corners of the rectangular boxes of the same type of adjacent plant nodes in the real plant morphology image (x lu ,y lu ) and the lower right corner coordinate (x ru ,y ru ), calculate the maximum coordinates (x lmin ,y lmin )、(x rmax ,y rmax ), determine the plant part boundary between two adjacent plant nodes; The boundary mask contour of the plant part is processed by dimensionality reduction to obtain the center point, main direction and two orthogonal direction axes of the target contour, and then it is moved up and down in the orthogonal direction of the direction axis. After the movement, the tangent point between the direction axis and the target contour is the boundary point of the minimum boundary rectangle of the target contour; the contour tangent is the direction axis after the movement, that is, y=k1x+b * ,y=k2x+b * , the specific formula is: Among them, contour represents the set of boundary points, k1 and k2 represent the slopes of the two direction axes respectively, and b * represents the intercept of the line, b * The maximum and minimum values ​​of correspond to the points farthest from the central axis along two orthogonal directions; Find the point with the largest distance from the direction axis along the contour line, take the arithmetic mean of the x-coordinate of the point with the largest distance from the axis to get the x-coordinate C of the center point x , take the arithmetic mean of the y coordinates of the points with the maximum distance from the axis to get the y coordinate C of the center point y , get the center point (C x ,C y ), calculate the maximum off-axis distance point corresponding to the k value, determine the width and height of the target minimum bounding rectangle, and obtain the path of adjacent plant nodes in the real plant morphology image.

10. The method for extracting plant morphological phenotype information according to claim 9, characterized in that: The steps of constructing the plant morphology include: Determine the lowest point of each plant branch node, and for the node on the main stem of the plant, calculate the Euclidean distance between it and the lowest point of the plant branch node; Determine two direction vectors, including the branch direction vector and the main stem direction vector of the plant, and calculate the angle of the direction vector between the lowest point of the branch and the main stem node of the plant. The calculation formula is: in, represents the branch direction vector of the plant, represents the main stem direction vector of the plant, θ represents the angle of the direction vector between the lowest point of the branch of the plant and the main stem node; From the calculated Euclidean distance, the plant node with the shortest distance is selected as the candidate plant node, and it is checked whether the angle difference between the main stem direction vector and the branch direction vector of the candidate plant node is within θ. If the condition is met, the plant branch node is connected to the plant main stem node, and each plant branch node is connected in turn to construct the plant morphology.

Images