A smart management method for garden seedlings based on stereoscopic perception

Abstract

The present invention relates to the field of intelligent management technology, specifically to an intelligent management method for garden seedlings based on stereoscopic perception. The present invention obtains three-dimensional point cloud data of the garden based on binocular stereoscopic vision technology, and combines image segmentation and clustering algorithms to analyze the distribution of cherry trees in the garden to obtain various cherry management areas; by obtaining environmental data and image data of each area, the regional cherry water demand coefficient is calculated according to the first cherry data; secondly, a cherry abnormality recognition model is constructed, and the regional image data is processed using a deep convolutional neural network to identify the abnormal characteristics of regional pests and calculate the regional cherry abnormality coefficient; finally, according to the water demand coefficient and the regional cherry abnormality coefficient, the regional cherry scheduling urgency coefficient is comprehensively calculated, and then intelligent resource allocation and scheduling are realized according to the regional cherry scheduling urgency coefficient. The present invention can realize the intelligent scheduling of garden resources, thereby improving the management efficiency of the garden.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent management, and in particular to an intelligent management method for garden seedlings based on stereoscopic perception. Background Art

[0002] Garden seedling management involves multiple tasks, including irrigation, fertilization, and pest control. With the diversification of garden seedling management needs and the continuous improvement of intelligence, traditional garden management methods are difficult to meet the requirements of intelligence and efficiency in garden seedling management. Therefore, intelligent management of garden seedlings is of great significance to improving the efficiency of garden seedling management and the rational use of garden resources.

[0003] As a type of garden seedlings, traditional seedling management methods do not fully consider the scheduling of management resources between different cherry tree areas, which easily leads to limited efficiency in garden seedling management.

[0004] Therefore, a smart management method for garden seedlings based on three-dimensional perception was proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a garden seedling intelligent management method based on stereoscopic perception. The present invention relates to the field of intelligent management technology, specifically a garden seedling intelligent management method based on stereoscopic perception. The present invention obtains the three-dimensional point cloud data of the garden based on binocular stereo vision technology, and combines image segmentation and clustering algorithms to analyze the distribution of cherry trees in the garden to obtain various cherry management areas; by obtaining the environmental data and image data of each area, the regional cherry water demand coefficient is calculated according to the first cherry data; secondly, a cherry abnormality recognition model is constructed, and the regional image data is processed using a deep convolutional neural network to identify the abnormal characteristics of regional pests and calculate the regional cherry abnormality coefficient; finally, according to the water demand coefficient and the regional cherry abnormality coefficient, the regional cherry scheduling urgency coefficient is comprehensively calculated, and then the intelligent resource allocation and scheduling is realized according to the regional cherry scheduling urgency coefficient. The present invention can realize the intelligent scheduling of garden resources, thereby improving the management efficiency of the garden.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for intelligent management of garden seedlings based on stereoscopic perception, comprising:

[0008] S10. Obtain the three-dimensional point cloud data of the garden based on binocular stereo vision technology, and analyze the distribution of cherry trees in the garden by combining image segmentation and clustering algorithms, divide the garden into regions according to the distribution of cherry trees, and obtain various cherry management areas;

[0009] S20. Obtaining first cherry blossom data and second cherry blossom data of each cherry blossom management area; constructing a regional cherry blossom water demand coefficient according to the first cherry blossom data;

[0010] S30. Construct a cherry blossom anomaly recognition model to recognize the second cherry blossom data, and sequentially pass through the preprocessing layer, regional pest feature extraction layer, regional pest anomaly analysis layer and output layer of the cherry blossom anomaly recognition model; the preprocessing layer is used to preprocess the second cherry blossom data to obtain the third cherry blossom data; the regional pest feature extraction layer is used to perform image recognition on the third cherry blossom data through a deep convolutional neural network to obtain regional pest anomaly features; the regional pest anomaly analysis layer is used to calculate the regional cherry blossom anomaly coefficient according to the regional pest anomaly features; the output layer is used to output the regional cherry blossom anomaly coefficient;

[0011] S40. Construct a regional cherry blossom scheduling urgency coefficient based on the regional cherry blossom water demand coefficient and the regional cherry blossom anomaly coefficient; and perform resource allocation scheduling for each cherry blossom management area based on the regional cherry blossom scheduling urgency coefficient.

[0012] Preferably, the first cherry blossom data includes the temperature, soil moisture, cherry tree height, cherry tree trunk diameter and air humidity of each cherry blossom management area; the second cherry blossom data includes regional cherry blossom tree image data obtained by drone cruising.

[0013] Preferably, the water demand coefficient of cherry blossoms in the region is:

[0014] ;

[0015] in, Indicates Regional cherry blossom water demand coefficient for each cherry blossom management area; It represents the temperature influence coefficient, which is used to indicate the effect of temperature on water demand; Indicates the temperature; It represents the soil moisture influence coefficient, which is used to indicate the influence of soil moisture on water demand; Indicates soil moisture; Indicates The number of cherry trees in each cherry management area; The influence coefficient of cherry tree height is used to indicate the influence of cherry tree height on water demand; Indicates No. 1 in the cherry blossom management area The height of a cherry tree; It represents the influence coefficient of cherry tree trunk diameter, which is used to represent the influence of cherry tree trunk diameter on water demand; Indicates The first The trunk diameter of a cherry tree; Indicates the air humidity influence coefficient, which is used to indicate the influence of air humidity on water demand; Indicates air humidity.

[0016] Preferably, the cherry blossom anomaly recognition model includes a preprocessing layer, a regional pest feature extraction layer, a regional pest anomaly analysis layer and an output layer;

[0017] The preprocessing layer is used to preprocess the second cherry blossom data to obtain third cherry blossom data; the preprocessing includes image denoising and image enhancement;

[0018] The regional pest feature extraction layer is used to perform image segmentation on the third cherry blossom data through a deep convolutional neural network, identify pest areas and non-pest areas, and identify the pixel area containing pests, the pest types and the number of various types of pests based on the pest area; identify regional pest abnormal features according to the pixel area containing pests, the pest types and the number of various types of pests and the total pixel area;

[0019] The regional pest anomaly analysis layer is used to calculate the regional cherry blossom anomaly coefficient according to the regional pest anomaly characteristics;

[0020] The output layer is used to output the cherry blossom anomaly coefficient of the region.

[0021] Preferably, the regional pest abnormal characteristics include regional pest density, regional pest coverage coefficient and regional pest risk coefficient; the regional pest risk coefficient is obtained by analyzing the risk scores of various types of pests and the number of various types of pests in each cherry blossom management area.

[0022] Preferably, the regional cherry blossom anomaly coefficient is:

[0023] ;

[0024] in, Indicates Regional cherry blossom anomaly coefficient of each cherry blossom management area; represents the regional pest density impact coefficient; Indicates Regional pest density in each cherry blossom management area; It represents the regional pest coverage impact coefficient; Indicates Regional pest coverage coefficient of each cherry blossom management area; It represents the regional pest risk impact coefficient; Indicates Regional pest risk factor for each cherry blossom management area;

[0025] Preferably, the regional pest risk factor is:

[0026] ;

[0027] in, Indicates Regional pest risk factor for each cherry blossom management area; Indicates Number of pest species in each cherry blossom management area; Indicates The first in the cherry blossom management area Number of pest species; Indicates The first in the cherry blossom management area Pest risk rating.

[0028] Preferably, the regional cherry blossom scheduling urgency coefficient is:

[0029] ;

[0030] in, Indicates Regional cherry blossom scheduling urgency coefficient of each cherry blossom management area; Indicates the influencing factor of the regional cherry blossom water demand coefficient; represents the regional cherry blossom water demand coefficient; Indicates the influencing factor of regional cherry blossom anomaly coefficient; Represents the regional cherry blossom anomaly coefficient.

[0031] Preferably, the regional cherry blossom scheduling urgency coefficient performs resource allocation scheduling on the various cherry blossom management areas, including: prioritizing each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient, obtaining a resource scheduling ranking for each area, and performing resource scheduling according to the resource scheduling ranking.

[0032] Preferably, the resource scheduling includes allocating irrigation amounts and scheduling pest control for each cherry blossom management area.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] 1. The present invention obtains three-dimensional point cloud data of the garden based on binocular stereo vision technology, and combines image segmentation and clustering algorithms to analyze the distribution of cherry trees in the garden, divides the garden into regions according to the distribution of cherry trees, and obtains various cherry management areas; obtains the first cherry data of each cherry management area; including the temperature, soil moisture, cherry tree height, cherry tree trunk diameter and air humidity of each cherry management area; constructs the regional cherry water demand coefficient based on these data; by constructing the regional cherry water demand coefficient, the demand for water of each cherry tree can be effectively interpreted, which provides a data basis for further realizing intelligent scheduling of cherry trees in the later stage, thereby further improving the management efficiency of the garden.

[0035] 2. The present invention identifies the second cherry blossom data by constructing a cherry blossom abnormality recognition model, and sequentially passes through the preprocessing layer, regional pest feature extraction layer, regional pest abnormality analysis layer and output layer of the cherry blossom abnormality recognition model; the regional pest feature extraction layer performs image recognition on the preprocessed second cherry blossom data through a deep convolutional neural network to obtain regional pest abnormality features; the regional pest abnormality analysis layer calculates the regional cherry blossom abnormality coefficient according to the regional pest abnormality features; the regional cherry blossom abnormality coefficient can effectively analyze the pest abnormality of cherry blossom trees in various areas of the garden, and provides a model basis for further realizing the intelligent scheduling of cherry trees, thereby further improving the management efficiency of the garden.

[0036] 3. The present invention constructs a regional cherry blossom scheduling urgency coefficient based on the regional cherry blossom water demand coefficient and the regional cherry blossom anomaly coefficient. The regional cherry blossom scheduling urgency coefficient can be used to more intuitively and clearly analyze the resource demand situation of each cherry blossom management area in the garden, so as to perform resource allocation scheduling for each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient, thereby realizing intelligent scheduling of garden resources and improving garden management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A schematic flow chart of a method for intelligent management of garden seedlings based on stereoscopic perception provided by an embodiment of the present invention;

[0038] Figure 2 A schematic diagram of the structure of a cherry blossom anomaly recognition model provided by an embodiment of the present invention;

[0039] Figure 3 A schematic diagram of an intelligent management interface for garden seedlings provided in this embodiment. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Embodiment 1

[0042] In order to improve the management efficiency of cherry trees, Garden A applied a smart management method for garden seedlings based on three-dimensional perception.

[0043] refer to Figure 1 , which is a flow chart of a method for intelligent management of garden seedlings based on stereoscopic perception, including:

[0044] S10. Obtain the three-dimensional point cloud data of the garden based on binocular stereo vision technology, and analyze the distribution of cherry trees in the garden by combining image segmentation and clustering algorithms, divide the garden into regions according to the distribution of cherry trees, and obtain various cherry management areas;

[0045] Furthermore, the process of acquiring each cherry blossom management area is as follows: based on binocular stereo vision technology, the garden area is photographed by binocular camera devices installed at different positions in the garden to ensure that each area of ​​the garden can be photographed and obtain two sets of image data; then stereo matching is performed based on the two sets of image data to calculate the disparity value of each pixel point. The disparity value represents the position difference of the same object in the two sets of images and reflects the relative distance between the object and the camera. Through the disparity calculation, a disparity map, that is, the disparity value of each pixel in the image is obtained; then depth calculation is performed based on the disparity map and the camera's internal reference information to generate a depth map, and each pixel point is mapped to a three-dimensional space coordinate in combination with the camera's internal reference information to generate three-dimensional point cloud data for processing and analysis, and an image segmentation algorithm based on depth information is applied to extract the area of ​​cherry blossom trees in the garden, and the extracted cherry blossom tree area is clustered by DBSCAN. By analyzing the density of cherry blossom trees, different areas are identified, and the garden area is partitioned to form each cherry blossom management area; this embodiment includes 5 cherry blossom management areas; they are area 1, area 2, area 3, area 4 and area 5;

[0046] S20. Obtaining first cherry blossom data and second cherry blossom data of each cherry blossom management area; constructing a regional cherry blossom water demand coefficient according to the first cherry blossom data;

[0047] Furthermore, the first cherry blossom data includes the temperature, soil moisture, cherry blossom tree height, cherry blossom tree trunk diameter and air humidity of each cherry blossom management area; the temperature is obtained by an air temperature sensor; the soil moisture is measured by arranging soil moisture sensors at multiple locations in each cherry blossom management area and taking the average value; the air humidity is measured by arranging air humidity sensors at multiple locations in each cherry blossom management area and taking the average value; the cherry blossom tree height and trunk diameter can be measured by using a laser rangefinder for each tree in each cherry blossom management area; the cherry blossom data of the A garden part are shown in Table 1;

[0048] Table 1 Data of cherry blossoms in some gardens of A

[0049]

[0050] Furthermore, the water demand coefficient of cherry blossoms in the area is:

[0051] ;

[0052] in, Indicates Regional cherry blossom water demand coefficient for each cherry blossom management area; It represents the temperature influence coefficient, which is used to indicate the effect of temperature on water demand; Indicates the temperature; It represents the soil moisture influence coefficient, which is used to indicate the influence of soil moisture on water demand; Indicates soil moisture; Indicates The number of cherry trees in each cherry management area; The influence coefficient of cherry tree height is used to indicate the influence of cherry tree height on water demand; Indicates No. 1 in the cherry blossom management area The height of a cherry tree; It represents the influence coefficient of cherry tree trunk diameter, which is used to represent the influence of cherry tree trunk diameter on water demand; Indicates No. 1 in the cherry blossom management area The trunk diameter of a cherry tree; Indicates the air humidity influence coefficient, which is used to indicate the influence of air humidity on water demand; Indicates air humidity.

[0053] The water demand coefficients of cherry blossoms in each region are shown in Table 2;

[0054] Table 2 Water demand coefficients of cherry blossoms in different regions

[0055]

[0056] This embodiment obtains three-dimensional point cloud data of the garden based on binocular stereo vision technology, and combines image segmentation and clustering algorithms to analyze the distribution of cherry trees in the garden, divides the garden into regions according to the distribution of cherry trees, and obtains various cherry management areas; obtains the first cherry data of each cherry management area; including the temperature, soil moisture, cherry tree height, cherry tree trunk diameter and air humidity of each cherry management area; constructs a regional cherry blossom water demand coefficient based on these data; by constructing the regional cherry blossom water demand coefficient, the demand for water of each cherry tree can be effectively interpreted, which provides a data basis for further realizing intelligent scheduling of cherry trees in the later stage, thereby further improving the management efficiency of the garden.

[0057] S30. Construct a cherry blossom anomaly recognition model to recognize the second cherry blossom data, and sequentially pass through the preprocessing layer, regional pest feature extraction layer, regional pest anomaly analysis layer and output layer of the cherry blossom anomaly recognition model; the preprocessing layer is used to preprocess the second cherry blossom data to obtain the third cherry blossom data; the regional pest feature extraction layer is used to perform image recognition on the third cherry blossom data through a deep convolutional neural network to obtain regional pest anomaly features; the regional pest anomaly analysis layer is used to calculate the regional cherry blossom anomaly coefficient according to the regional pest anomaly features; the output layer is used to output the regional cherry blossom anomaly coefficient;

[0058] The second cherry blossom data includes regional cherry blossom tree image data obtained by drone cruise shooting of each cherry blossom management area; the distribution density of cherry blossom trees in each area is photographed, wherein 5 pictures are taken in area 1; the images are numbered A1, A2, A3, A4 and A5; 3 pictures are taken in area 2, and the images are numbered A6, A7, A8; 4 pictures are taken in area 3, and the images are numbered A9, A10, A11, A12; 3 pictures are taken in area 4, and the images are numbered A13, A14, A15; 3 pictures are taken in area 5; the images are numbered A16, A17, A18;

[0059] Furthermore, the cherry blossom anomaly recognition model includes a preprocessing layer, a regional pest feature extraction layer, a regional pest anomaly analysis layer and an output layer; the schematic diagram of the cherry blossom anomaly recognition model structure is as follows: Figure 2 As shown;

[0060] The preprocessing layer is used to preprocess the second cherry blossom data to obtain third cherry blossom data; the preprocessing includes image denoising and image enhancement;

[0061] The regional pest feature extraction layer is used to perform image segmentation on the third cherry blossom data through a deep convolutional neural network, identify pest areas and non-pest areas, and identify the pixel area containing pests, the pest types and the number of various types of pests based on the pest area; identify regional pest abnormal features according to the pixel area containing pests, the pest types and the number of various types of pests and the total pixel area;

[0062] The regional pest anomaly analysis layer is used to calculate the regional cherry blossom anomaly coefficient according to the regional pest anomaly characteristics;

[0063] The output layer is used to output the cherry blossom anomaly coefficient of the region.

[0064] Furthermore, the regional pest abnormality characteristics include regional pest density, regional pest coverage coefficient and regional pest risk coefficient; the regional pest risk coefficient is obtained by analyzing the risk scores of various pests and the number of various pests in each cherry blossom management area; the risk score is obtained through expert experience;

[0065] Furthermore, the cherry blossom anomaly coefficient of the area is:

[0066] ;

[0067] in, Indicates Regional cherry blossom anomaly coefficient of each cherry blossom management area; represents the regional pest density impact coefficient; Indicates Regional pest density in each cherry blossom management area; It represents the regional pest coverage impact coefficient; Indicates Regional pest coverage coefficient of each cherry blossom management area; It represents the regional pest risk impact coefficient; Indicates Regional pest risk factor for each cherry blossom management area;

[0068] The pest density in the area is:

[0069] ;

[0070] in, Indicates Regional pest density in each cherry blossom management area; Indicates Regional pest counts in each cherry blossom management area; Indicates The total pixel area of ​​the cherry blossom management area;

[0071] The pest coverage coefficient of the area is:

[0072] ;

[0073] in, Indicates Regional pest coverage coefficient of each cherry blossom management area; Indicates The pixel area containing pests in each cherry blossom management area; Indicates The total pixel area of ​​the cherry blossom management area;

[0074] Furthermore, the pest risk factor of the area is:

[0075] ;

[0076] in, Indicates Regional pest risk factor for each cherry blossom management area; Indicates Number of pest species in each cherry blossom management area; Indicates The first in the cherry blossom management area Number of pest species; Indicates The first in the cherry blossom management area Pest risk rating.

[0077] For Garden A, the specific data output by the cherry blossom anomaly recognition model is shown in Table 3;

[0078] Table 3 Model output data table for Garden A

[0079]

[0080] This embodiment identifies the second cherry blossom data by constructing a cherry blossom abnormality recognition model, and sequentially passes through the preprocessing layer, regional pest feature extraction layer, regional pest abnormality analysis layer and output layer of the cherry blossom abnormality recognition model; the regional pest feature extraction layer performs image recognition on the preprocessed second cherry blossom data through a deep convolutional neural network to obtain regional pest abnormality features; the regional pest abnormality analysis layer calculates the regional cherry blossom abnormality coefficient according to the regional pest abnormality features; the regional cherry blossom abnormality coefficient can effectively analyze the pest abnormality of cherry trees in various areas of the garden, and provides a model basis for further realizing the intelligent scheduling of cherry trees, thereby further improving the management efficiency of the garden.

[0081] S40. Constructing a regional cherry blossom scheduling urgency coefficient according to the regional cherry blossom water demand coefficient and the regional cherry blossom abnormality coefficient; performing resource allocation scheduling for each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient;

[0082] Furthermore, the urgency coefficient of cherry blossom scheduling in the region is:

[0083] ;

[0084] in, Indicates Regional cherry blossom scheduling urgency coefficient of each cherry blossom management area; Indicates the influencing factor of the regional cherry blossom water demand coefficient; represents the regional cherry blossom water demand coefficient; Indicates the influencing factor of regional cherry blossom anomaly coefficient; represents the regional cherry blossom abnormality coefficient; further, the cherry blossom scheduling urgency coefficient of each region is shown in Table 4;

[0085] Table 4. Regional cherry blossom scheduling urgency coefficient table for each region

[0086]

[0087] Furthermore, the regional cherry blossom scheduling urgency coefficient performs resource allocation scheduling on each cherry blossom management area, including: prioritizing each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient, obtaining a resource scheduling ranking of each area, and performing resource scheduling according to the resource scheduling ranking; the resource scheduling includes allocating irrigation volume and scheduling pest control for each cherry blossom management area;

[0088] Reference Figure 3 A schematic diagram of an intelligent management interface for garden seedlings provided for the present embodiment; comprising three modules: water demand coefficient identification, cherry blossom anomaly identification and analysis, and scheduling management and resource allocation; the water demand coefficient identification module realizes the division of cherry tree areas and the identification of the water demand coefficient of each cherry tree management area; the cherry blossom anomaly identification and analysis module realizes the identification of the regional cherry tree anomaly coefficient of each cherry tree management area in the garden; the scheduling management and resource allocation module obtains the regional scheduling urgency coefficient based on the water demand coefficient and the regional cherry tree anomaly coefficient, and performs resource scheduling based on the regional scheduling urgency coefficient.

[0089] In order to verify the effectiveness of the smart management method for garden seedlings based on stereoscopic perception proposed in this embodiment, comparative experiments were conducted using method one and method two. Method one is a smart management method for garden seedlings based on stereoscopic perception proposed in this embodiment; method two is a smart management method for garden seedlings based on stereoscopic perception proposed in this embodiment without considering the resource scheduling and sorting of each cherry blossom management area; the two methods were compared based on the irrigation water resource utilization efficiency and the completion of the pest control task. The specific comparison results are shown in Table 5.

[0090] Table 5 Comparison of the effects of different methods on the management of garden cherry trees

[0091]

[0092] It can be seen from Table 5 that the intelligent management method for garden seedlings based on stereoscopic perception proposed in this embodiment has a better management effect on garden cherry trees, which reflects the effectiveness of the intelligent management method for garden seedlings based on stereoscopic perception proposed in this embodiment;

[0093] This embodiment constructs a regional cherry blossom scheduling urgency coefficient based on the regional cherry blossom water demand coefficient and the regional cherry blossom anomaly coefficient. The regional cherry blossom scheduling urgency coefficient can be used to more intuitively and clearly analyze the resource demand situation of each cherry blossom management area in the garden, so as to allocate resources for each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient, thereby realizing intelligent scheduling of garden resources and improving garden management efficiency.

[0094] Embodiment 2

[0095] B Garden is currently studying how to improve the management efficiency of cherry trees in the garden, and has applied a smart management method for garden seedlings based on three-dimensional perception;

[0096] refer to Figure 1 , which is a flow chart of a method for intelligent management of garden seedlings based on stereoscopic perception, including:

[0097] S10. Obtain the three-dimensional point cloud data of the garden based on binocular stereo vision technology, and analyze the distribution of cherry trees in the garden by combining image segmentation and clustering algorithms, divide the garden into regions according to the distribution of cherry trees, and obtain various cherry management areas;

[0098] Furthermore, the process of acquiring each cherry blossom management area is as follows: based on binocular stereo vision technology, the garden area is photographed by binocular camera devices installed at different positions in the garden to ensure that each area of ​​the garden can be photographed and obtain two sets of image data; then stereo matching is performed based on the two sets of image data to calculate the disparity value of each pixel point. The disparity value represents the position difference of the same object in the two sets of images and reflects the relative distance between the object and the camera. Through the disparity calculation, a disparity map, that is, the disparity value of each pixel in the image is obtained; then depth calculation is performed based on the disparity map and the camera's internal reference information to generate a depth map, and each pixel point is mapped to a three-dimensional space coordinate in combination with the camera's internal reference information to generate three-dimensional point cloud data for processing and analysis, and an image segmentation algorithm based on depth information is applied to extract the area of ​​cherry blossom trees in the garden, and the extracted cherry blossom tree area is clustered by DBSCAN. By analyzing the density of cherry blossom trees, different areas are identified, and the garden area is partitioned to form each cherry blossom management area; this embodiment includes 5 cherry blossom management areas; they are area 1, area 2, area 3, area 4 and area 5;

[0099] S20. Obtaining first cherry blossom data and second cherry blossom data of each cherry blossom management area; constructing a regional cherry blossom water demand coefficient according to the first cherry blossom data;

[0100] Furthermore, the first cherry blossom data includes the temperature, soil moisture, cherry blossom tree height, cherry blossom tree trunk diameter and air humidity of each cherry blossom management area; the temperature is obtained by an air temperature sensor; the soil moisture is measured by arranging soil moisture sensors at multiple locations in each cherry blossom management area and taking the average value; the air humidity is measured by arranging air humidity sensors at multiple locations in each cherry blossom management area and taking the average value; the cherry blossom tree height and trunk diameter can be measured by using a laser rangefinder to measure the tree height of each tree in each cherry blossom management area; the cherry blossom data of the garden part B is shown in Table 6;

[0101] Table 6. Cherry blossom data for part of B garden

[0102]

[0103] Furthermore, the water demand coefficient of cherry blossoms in the area is:

[0104] ;

[0105] in, Indicates Regional cherry blossom water demand coefficient for each cherry blossom management area; It represents the temperature influence coefficient, which is used to indicate the effect of temperature on water demand; Indicates the temperature; It represents the soil moisture influence coefficient, which is used to indicate the influence of soil moisture on water demand; Indicates soil moisture; Indicates The number of cherry trees in each cherry management area; The influence coefficient of cherry tree height is used to indicate the influence of cherry tree height on water demand; Indicates The first The height of a cherry tree; It represents the influence coefficient of cherry tree trunk diameter, which is used to represent the influence of cherry tree trunk diameter on water demand; Indicates The first The trunk diameter of a cherry tree; Indicates the air humidity influence coefficient, which is used to indicate the influence of air humidity on water demand; Indicates air humidity.

[0106] The table of water demand coefficients for cherry blossoms in each area of ​​Garden B is shown in Table 7;

[0107] Table 7 Water demand coefficient of cherry blossoms in different regions

[0108]

[0109] The second cherry blossom data includes regional cherry blossom tree image data obtained by drone cruise photography of each cherry blossom management area; the distribution density of cherry blossom trees in each area is photographed, wherein 4 pictures are taken in area 1; the images are numbered B1, B2, B3, and B4; 3 pictures are taken in area 2; the images are numbered B5, B6, and B7; 2 pictures are taken in area 3, and the images are numbered B8 and B9; 3 pictures are taken in area 4, and the images are numbered B10, B11, and B12; 3 pictures are taken in area 5; the images are numbered P13, P14, and P15;

[0110] S30. Construct a cherry blossom anomaly recognition model to recognize the second cherry blossom data, and sequentially pass through the preprocessing layer, regional pest feature extraction layer, regional pest anomaly analysis layer and output layer of the cherry blossom anomaly recognition model; the preprocessing layer is used to preprocess the second cherry blossom data to obtain the third cherry blossom data; the regional pest feature extraction layer is used to perform image recognition on the third cherry blossom data through a deep convolutional neural network to obtain regional pest anomaly features; the regional pest anomaly analysis layer is used to calculate the regional cherry blossom anomaly coefficient according to the regional pest anomaly features; the output layer is used to output the regional cherry blossom anomaly coefficient;

[0111] Furthermore, the cherry blossom anomaly recognition model includes a preprocessing layer, a regional pest feature extraction layer, a regional pest anomaly analysis layer and an output layer; the schematic diagram of the cherry blossom anomaly recognition model structure is as follows: Figure 2 As shown;

[0112] The preprocessing layer is used to preprocess the second cherry blossom data to obtain third cherry blossom data; the preprocessing includes image denoising and image enhancement;

[0113] The regional pest feature extraction layer is used to perform image segmentation on the third cherry blossom data through a deep convolutional neural network, identify pest areas and non-pest areas, and identify the pixel area containing pests, the pest types and the number of various types of pests based on the pest area; identify regional pest abnormal features according to the pixel area containing pests, the pest types and the number of various types of pests and the total pixel area;

[0114] The regional pest anomaly analysis layer is used to calculate the regional cherry blossom anomaly coefficient according to the regional pest anomaly characteristics;

[0115] The output layer is used to output the cherry blossom anomaly coefficient of the region.

[0116] Furthermore, the regional pest abnormality characteristics include regional pest density, regional pest coverage coefficient and regional pest risk coefficient; the regional pest risk coefficient is obtained by analyzing the risk scores of various pests and the number of various pests in each cherry blossom management area; the risk score is obtained through expert experience;

[0117] Furthermore, the cherry blossom anomaly coefficient of the area is:

[0118] ;

[0119] in, Indicates Regional cherry blossom anomaly coefficient of each cherry blossom management area; represents the regional pest density impact coefficient; Indicates Regional pest density in each cherry blossom management area; It represents the regional pest coverage impact coefficient; Indicates Regional pest coverage coefficient of each cherry blossom management area; It represents the regional pest risk impact coefficient; Indicates Regional pest risk factors for each cherry blossom management area.

[0120] The pest density in the area is:

[0121] ;

[0122] in, Indicates Regional pest density in each cherry blossom management area; Indicates Regional pest counts in each cherry blossom management area; Indicates The total pixel area of ​​the cherry blossom management area;

[0123] The pest coverage coefficient of the area is:

[0124] ;

[0125] in, Indicates Regional pest coverage coefficient of each cherry blossom management area; Indicates The pixel area containing pests in each cherry blossom management area; Indicates The total pixel area of ​​the cherry blossom management area;

[0126] Furthermore, the pest risk factor of the area is:

[0127] ;

[0128] in, Indicates Regional pest risk factor for each cherry blossom management area; Indicates Number of pest species in each cherry blossom management area; Indicates The first in the cherry blossom management area Number of pest species; Indicates The first in the cherry blossom management area Pest risk rating.

[0129] For Garden B, the specific data output by the cherry blossom anomaly recognition model is shown in Table 8;

[0130] Table 8 Model output data table for Garden B

[0131]

[0132] S40. Construct a regional cherry blossom scheduling urgency coefficient based on the regional cherry blossom water demand coefficient and the regional cherry blossom anomaly coefficient; and perform resource allocation scheduling for each cherry blossom management area based on the regional cherry blossom scheduling urgency coefficient.

[0133] Furthermore, the urgency coefficient of cherry blossom scheduling in the region is:

[0134] ;

[0135] in, Indicates Regional cherry blossom scheduling urgency coefficient of each cherry blossom management area; Indicates the influencing factor of the regional cherry blossom water demand coefficient; represents the regional cherry blossom water demand coefficient; Indicates the influencing factor of regional cherry blossom anomaly coefficient; Represents the regional cherry blossom anomaly coefficient.

[0136] Furthermore, the urgency coefficient of cherry blossom scheduling in each region is shown in Table 9;

[0137] Table 9: Regional cherry blossom scheduling urgency coefficient table for each region

[0138]

[0139] Furthermore, the regional cherry blossom scheduling urgency coefficient performs resource allocation scheduling on the various cherry blossom management areas, including: prioritizing each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient, obtaining a resource scheduling ranking for each area, and performing resource scheduling according to the resource scheduling ranking; the resource scheduling includes allocating irrigation volume and scheduling pest control for each cherry blossom management area.

[0140] Reference Figure 3 A schematic diagram of an intelligent management interface for garden seedlings provided for the present embodiment; comprising three modules: water demand coefficient identification, cherry blossom anomaly identification and analysis, and scheduling management and resource allocation; the water demand coefficient identification module realizes the division of cherry tree areas and the identification of the water demand coefficient of each cherry tree management area; the cherry blossom anomaly identification and analysis module realizes the identification of the regional cherry tree anomaly coefficient of each cherry tree management area in the garden; the scheduling management and resource allocation module obtains the regional scheduling urgency coefficient based on the water demand coefficient and the regional cherry tree anomaly coefficient, and performs resource scheduling based on the regional scheduling urgency coefficient.

[0141] In order to verify the effectiveness of the smart management method for garden seedlings based on stereoscopic perception proposed in this embodiment, comparative experiments were conducted using method one and method two. Method one is a smart management method for garden seedlings based on stereoscopic perception proposed in this embodiment; method two is a smart management method for garden seedlings based on stereoscopic perception proposed in this embodiment without considering the resource scheduling sorting of each cherry blossom management area; the two methods were compared based on the irrigation water resource utilization efficiency and the completion degree of the pest control task. The specific comparison results are shown in Table 10.

[0142] Table 10 Comparison of the effects of different methods on the management of garden cherry trees

[0143]

[0144] It can be seen from Table 10 that the intelligent management method for garden seedlings based on stereoscopic perception proposed in this embodiment has a better management effect on garden cherry trees, which reflects the effectiveness of the intelligent management method for garden seedlings based on stereoscopic perception proposed in this embodiment;

[0145] This embodiment constructs a regional cherry blossom scheduling urgency coefficient based on the regional cherry blossom water demand coefficient and the regional cherry blossom anomaly coefficient. The regional cherry blossom scheduling urgency coefficient can be used to more intuitively and clearly analyze the resource demand situation of each cherry blossom management area in the garden, so as to allocate resources for each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient, thereby realizing intelligent scheduling of garden resources and improving garden management efficiency.

[0146] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent management of garden seedlings based on stereoscopic perception, characterized in that: include: S10. Obtain the three-dimensional point cloud data of the garden based on binocular stereo vision technology, and analyze the distribution of cherry trees in the garden by combining image segmentation and clustering algorithms, divide the garden into regions according to the distribution of cherry trees, and obtain various cherry management areas; S20. Obtain the first cherry blossom data and the second cherry blossom data of each cherry blossom management area; construct the regional cherry blossom water demand coefficient according to the first cherry blossom data; the first cherry blossom data includes the temperature, soil moisture, cherry blossom tree height, cherry blossom tree trunk diameter and air humidity of each cherry blossom management area; the second cherry blossom data includes the regional cherry blossom tree image data obtained by drone cruise photography; the regional cherry blossom water demand coefficient is: in, Indicates Regional cherry blossom water demand coefficient for each cherry blossom management area; It represents the temperature influence coefficient, which is used to indicate the effect of temperature on water demand; Indicates the temperature; It represents the soil moisture influence coefficient, which is used to indicate the influence of soil moisture on water demand; Indicates soil moisture; Indicates The number of cherry trees in each cherry management area; The influence coefficient of cherry tree height is used to indicate the influence of cherry tree height on water demand; Indicates No. 1 in the cherry blossom management area The height of a cherry tree; It represents the influence coefficient of cherry tree trunk diameter, which is used to represent the influence of cherry tree trunk diameter on water demand; Indicates No. 1 in the cherry blossom management area The trunk diameter of a cherry tree; Indicates the air humidity influence coefficient, which is used to indicate the influence of air humidity on water demand; Indicates air humidity; S30. Construct a cherry blossom anomaly recognition model to recognize the second cherry blossom data, sequentially passing through the preprocessing layer, regional pest feature extraction layer, regional pest anomaly analysis layer and output layer of the cherry blossom anomaly recognition model; the preprocessing layer is used to preprocess the second cherry blossom data to obtain the third cherry blossom data; the preprocessing includes image denoising and image enhancement; The regional pest feature extraction layer is used to perform image segmentation on the third cherry blossom data through a deep convolutional neural network, identify pest areas and non-pest areas, and identify the pixel area containing pests, pest types and the number of various types of pests based on the pest area; identify regional pest abnormality characteristics according to the pixel area containing pests, pest types and the number of various types of pests and the total pixel area; the regional pest abnormality characteristics include regional pest density, regional pest coverage coefficient and regional pest risk coefficient; the regional pest risk coefficient is obtained by analyzing the risk scores of various types of pests and the number of various types of pests in each cherry blossom management area; the regional pest abnormality analysis layer is used to calculate the regional cherry blossom abnormality coefficient according to the regional pest abnormality characteristics; the regional cherry blossom abnormality coefficient is: in, Indicates Regional cherry blossom anomaly coefficient of each cherry blossom management area; represents the regional pest density impact coefficient; Indicates Regional pest density in each cherry blossom management area; It represents the regional pest coverage impact coefficient; Indicates Regional pest coverage coefficient of each cherry blossom management area; It represents the regional pest risk impact coefficient; Indicates Regional pest risk factor for each cherry blossom management area; The output layer is used to output the cherry blossom anomaly coefficient of the region; S40. Construct a regional cherry blossom scheduling urgency coefficient based on the regional cherry blossom water demand coefficient and the regional cherry blossom anomaly coefficient; and perform resource allocation scheduling for each cherry blossom management area based on the regional cherry blossom scheduling urgency coefficient.

2. A method for intelligent management of garden seedlings based on stereoscopic perception according to claim 1, characterized in that: The pest risk factor of the area is: in, Indicates Regional pest risk factor for each cherry blossom management area; Indicates Number of pest species in each cherry blossom management area; Indicates The first in the cherry blossom management area Number of pest species; Indicates The first in the cherry blossom management area Pest risk rating.

3. The method for intelligent management of garden seedlings based on stereoscopic perception according to claim 1, characterized in that: The urgency coefficient of cherry blossom scheduling in the area is: in, Indicates Regional cherry blossom scheduling urgency coefficient of each cherry blossom management area; Indicates the influencing factor of the regional cherry blossom water demand coefficient; represents the regional cherry blossom water demand coefficient; Indicates the influencing factor of regional cherry blossom anomaly coefficient; Represents the regional cherry blossom anomaly coefficient.

4. The method for intelligent management of garden seedlings based on stereoscopic perception according to claim 1, characterized in that: The resource allocation scheduling of each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient includes: prioritizing each cherry blossom management area according to the regional cherry blossom scheduling urgency coefficient, obtaining a resource scheduling ranking for each area, and performing resource scheduling according to the resource scheduling ranking.

5. The method for intelligent management of garden seedlings based on stereoscopic perception according to claim 4 is characterized in that: The resource scheduling includes allocating irrigation amounts and scheduling pest control for each cherry blossom management area.

Images