A method and system for maintaining afforestation tree survival

Abstract

The invention discloses a method and system for maintaining the survival of afforestation trees, and relates to the technical field of remote sensing monitoring and tree disease prevention and control. The method comprises the following steps: collecting forest canopy hyperspectral images by unmanned aerial vehicles, and setting a forest band sensitive combination strategy to obtain a single plant chlorophyll concentration characterization value; dividing the forest area into regions based on the spectral characteristics of the image and the direct identification of the plant; setting a monitoring window to capture detailed image information of the partition boundary, dividing the monitoring window into window regions according to the divided forest areas, and locating the initial danger window region; expanding the initial danger window region according to the region expansion logic, combining with time series images, judging the influence trend of the central region disease on the surrounding trees, and issuing an early warning to the staff; the invention reduces the demand for complex algorithms, saves computing resources, and can efficiently extract forest information from massive hyperspectral image data.

Description

Technical Field

[0001] The invention relates to the technical field of remote sensing monitoring and tree disease prevention and control, and in particular to a method and system for maintaining afforestation tree survival. Background Art

[0002] At present, forest ecosystems are facing unprecedented challenges. Traditional methods of forest health monitoring mostly rely on manual inspections, which have problems such as low efficiency, high cost, and difficulty in comprehensive coverage.

[0003] In recent years, the application of remote sensing technology, especially UAV hyperspectral imaging technology, in the forestry field has made significant progress. Hyperspectral imaging can provide rich spectral information, and through the analysis of this information, the physiological state of plants can be effectively evaluated. Combined with machine learning and artificial intelligence technology, it can not only achieve real-time monitoring of forest health, but also predict potential disease risks, which is of great significance for forest resource management and protection.

[0004] Although there are some forest health monitoring methods based on remote sensing technology, most of them still have some limitations.

[0005] For example, the existing Chinese patent with publication number CN112634212A discloses a disease-hidden tree detection method and system based on a hyperspectral drone. The method includes taking the forest area infected with the disease as the sample area, randomly selecting multiple sample plants; regularly collecting leaves, obtaining the spectral data of the leaves, judging the health status of each sample plant, combining the satellite remote sensing image of the sample area to determine the disease stress area, and identifying the disease type of the sample plant; establishing a disease data model; determining the target area, collecting the hyperspectral image of the target area; preprocessing the hyperspectral image and inputting it into the disease data model to obtain the diseased plant detection results. The disease data model is established based on the sample data collected in the early stage, and the hyperspectral camera is used to collect on-site data by using the drone, and the disease data model is used to detect diseased plants, which provides data support for the use of hyperspectral remote sensing technology for large-scale forest pest and disease trends, disease monitoring and early warning.

[0006] The above method has a poor degree of automation, and there are still many problems in the prior art, such as inaccurate identification of individual trees, insufficient prediction of disease transmission trends, and huge amount of calculations. In addition, how to efficiently extract useful information from massive hyperspectral image data is also a problem to be solved. To this end, the present invention provides a method and system for afforestation tree survival maintenance. Summary of the invention

[0007] The object of the present invention is to provide a method and system for maintaining the survival of afforestation trees, so as to solve the existing problems raised in the above-mentioned background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: a method for maintaining the survival of afforestation trees, comprising the following steps:

[0009] S1. Collect hyperspectral images of forest canopy by drone, and set up a sensitive combination strategy for forest bands to obtain the chlorophyll concentration characterization value of a single plant;

[0010] S2, combining the spectral characteristics of the image and the direct identification of the plants to divide the forest area;

[0011] S3, setting a monitoring window to capture detailed image information of the partition boundary, dividing the monitoring window into window areas according to the divided forest areas, and locating the initial dangerous window area;

[0012] S4. Expand the initial danger window area according to the area expansion logic, combine the time series images, determine the impact trend of the disease in the central area on the surrounding trees, and issue an early warning to the staff.

[0013] The present invention is further improved in that the forest band sensitive combination strategy specifically comprises the following steps:

[0014] S11. Measure the chlorophyll concentration of sample plants in the forest area using a chlorophyll meter and include them in the chlorophyll concentration sample set;

[0015] S12, setting a chlorophyll sensitive band combination strategy, and extracting a chlorophyll characterization band combination;

[0016] S13, obtaining the chlorophyll concentration representation value of all pixels in the forest area through the chlorophyll representation band combination;

[0017] S14. Divide the area of ​​a single plant in the forest area by the chlorophyll concentration characterization value to obtain the chlorophyll concentration characterization value of each plant.

[0018] The present invention is further improved in that the chlorophyll sensitive band combination strategy specifically comprises the following steps:

[0019] S121, extracting the near infrared band reflectance data set of all pixels in the forest canopy hyperspectral image , red band reflectance dataset and green band reflectance dataset ,in, represents the reflectivity of the ath near-infrared band, represents the reflectivity of the b-th red light band, represents the reflectance of the cth green light band, a represents the number of near-infrared bands, b represents the number of red light bands, and c represents the number of green light bands;

[0020] S122. Establish a chlorophyll sensitive band combination set , where the first chlorophyll sensitive band combination , the second chlorophyll sensitive band combination , the third chlorophyll sensitive band combination ;

[0021] S123. Calculate the chlorophyll concentration characterization value :

[0022] ;

[0023] in, represents the weight of the second chlorophyll-sensitive band combination, represents the weight of the first chlorophyll-sensitive band combination, Represents the weight of the third chlorophyll-sensitive band combination;

[0024] S124, setting a chlorophyll characterization band extraction algorithm, and selecting a band combination by calculating the correlation coefficient between the chlorophyll concentration of the sample plant area and the chlorophyll concentration characterization value.

[0025] The present invention is further improved in that the specific steps of the chlorophyll characterization band extraction algorithm include: calculating the correlation coefficient between the chlorophyll concentration of the sample plant area and the chlorophyll concentration characterization value, and the calculation formula is:

[0026] ;

[0027] Among them, the function Represents the correlation coefficient, function Indicates the maximum The spectral band combination, achl represents the average chlorophyll concentration of all sample plants, Represents the chlorophyll mean of all sample plants, and obtains the chlorophyll characterization band combination , Indicates the number of band combinations.

[0028] A further improvement of the present invention is that S14 includes setting a chlorophyll characterization threshold, extracting pixels whose chlorophyll concentration characterization values ​​of all pixels in the forest area are greater than the chlorophyll characterization threshold, performing connected component analysis, setting a component size threshold, and selecting components greater than the component size threshold among the marked connected components to select areas representing single plants in the forest area.

[0029] A further improvement of the present invention is that the forest area division includes extracting the number of pixels of a single plant in the forest area in step S1, setting pixel number thresholds Tpi1, Tpi2, and Tpi3, and classifying the single plants in the forest area according to the pixel number thresholds, taking the plant in the upper left corner of the forest canopy hyperspectral image as the first initial traversal plant, starting to traverse all plants with the first initial traversal plant as the radius, and then calculating the mean of the pixels of the traversed plants, checking the category to which the mean belongs, until traversing to a plant that does not belong to the same category as the mean, stopping the traversal, and listing the traversed plants together with the first initial traversal plant as the first plant area; locating the plant that caused the traversal to stop as the second initial traversal plant, and starting the traversal again in the same way as above to obtain the second plant area, until all plants are traversed to obtain the Nth plant area.

[0030] A further improvement of the present invention is that setting a monitoring window to capture detailed image information of partition boundaries includes: defining the size of the monitoring window as H×W; traversing the connection points of all areas by overlapping the center point of the monitoring window with the connection point, dividing the monitoring window into areas through the connection lines of different areas, calculating the mean of the chlorophyll concentration characterization values ​​of all plants in different monitoring window partitions, and extracting the window partition corresponding to the maximum value of the mean of the chlorophyll concentration characterization value as the initial danger window area.

[0031] The present invention is further improved in that the regional expansion logic is set according to the characteristics of tree disease transmission, and the specific steps include: collecting historical wind speed data and disease transmission data between plants in the same period, obtaining the average wind speed data of any month and the number of diseased plants per day in the same period, and establishing a regression model: , and the wind speed influence coefficient is obtained ,in, represents the disease spread index, which is obtained by calculating the difference between the number of diseased plants on the last day and the first day of the same month. represents the daily average wind speed in the same month. is the intercept term, Represents the error term; collect the angle between the historical wind direction and the disease transmission direction, recorded as the wind direction impact angle ; Calculate the window expansion factor , get the Danger window range:

[0032] ;

[0033] in, Indicates the number of window expansions.

[0034] A further improvement of the present invention is that the area expansion logic further comprises: The first danger window range is obtained, the mean chlorophyll concentration characterization value of all plants in the first danger window range is calculated, and the chlorophyll characterization change standard value is set. When the difference between the mean chlorophyll concentration characterization value of all plants in the first danger window range and the mean chlorophyll concentration characterization value of all plants in the initial danger window range is less than or equal to the chlorophyll characterization change standard value, the window is expanded again to obtain the second danger window range, and the difference between the mean chlorophyll concentration characterization value of all plants in the second danger window range and the mean chlorophyll concentration characterization value of all plants in the initial danger window range is calculated again, until the first danger window range is obtained. When the difference between the mean chlorophyll concentration characteristic value of all plants within the danger window and the mean chlorophyll concentration characteristic value of all plants within the initial danger window is greater than the chlorophyll characteristic change standard value or the When the edge of the danger window exceeds the edge of the monitoring window, the expansion stops.

[0035] A further improvement of the present invention is that the judgment rule for issuing an early warning to the staff includes: When the difference between the mean chlorophyll concentration characterization value of all plants within the danger window and the mean chlorophyll concentration characterization value of all plants within the initial danger window is greater than the chlorophyll characterization change standard value, the first The danger window meets the conditions { Danger Window - Danger Window The plant with the smallest chlorophyll concentration within the danger window is regarded as the disease transmission center area, and the window expansion coefficient is used to calculate the disease transmission center area. The radius is used to issue an early warning to the staff; when the When the edge of the danger window exceeds the edge of the monitoring window, the plant with the smallest chlorophyll concentration in the initial danger window area is taken as the disease transmission center area, and the disease transmission center area is expanded by the window expansion coefficient Issue early warning to staff for the radius.

[0036] In another aspect, the present invention provides a system for maintaining the survival of afforestation trees, comprising:

[0037] The data acquisition module collects hyperspectral images of forest canopies through drones and sets a sensitive combination strategy for forest bands to obtain the chlorophyll concentration representation value of a single plant;

[0038] The regional division module combines the spectral characteristics of the image and the direct identification of the plants to divide the forest area;

[0039] The danger location module sets a monitoring window to capture detailed image information of the partition boundary, divides the monitoring window into window areas according to the divided forest areas, and locates the initial danger window area;

[0040] The disease warning module expands the initial danger window area according to the area expansion logic, combines the time series images, determines the impact trend of the disease in the central area on the surrounding trees, and issues a warning to the staff.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1. The present invention firstly utilizes image processing technology and the spectral characteristics of plants to automatically divide forest areas, thus simplifying the plant identification process, reducing the need for complex algorithms, and saving computing resources; by setting a threshold to distinguish between plant and non-plant areas, complex edge detection algorithms are avoided, further improving efficiency;

[0043] 2. Setting up monitoring windows to capture forest boundary image information and locating the initial danger zone through window area division helps to timely discover the spread trend of the disease; combining time-series images and extended logic to dynamically evaluate the impact of the disease on surrounding plants can predict the direction and speed of disease development, which is conducive to taking preventive measures;

[0044] 3. By using drones to obtain hyperspectral images of the forest canopy, specific band combinations are used to accurately assess the chlorophyll concentration of plants, thereby improving the accuracy of monitoring; the chlorophyll-sensitive band combination strategy can effectively identify and quantify the health status of plants, thereby detecting potential diseases at an early stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of a method for maintaining afforestation tree survival according to the present invention;

[0046] Figure 2 It is a flow chart of a forest band sensitive combination strategy of a forest area of ​​a method for maintaining the survival of afforestation trees of the present invention;

[0047] Figure 3 A chlorophyll sensitive band combination strategy flow chart of a method for maintaining afforestation tree survival according to the present invention;

[0048] Figure 4 A schematic diagram of the regional expansion logic of a method for maintaining afforestation tree survival according to the present invention;

[0049] Figure 5 This is a framework diagram of a afforestation tree survival maintenance system of the present invention. DETAILED DESCRIPTION

[0050] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.

[0051] The term "and / or" is only a description of the association relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the related objects are in an "or" relationship.

[0052] Example 1

[0053] Figure 1 A flowchart of a method for maintaining the survival of afforestation trees disclosed in this embodiment is shown, and the steps are as follows:

[0054] S1. Collect hyperspectral images of forest canopy by drone, and set up a sensitive combination strategy for forest bands to obtain the chlorophyll concentration characterization value of a single plant;

[0055] S2, combining the spectral characteristics of the image and the direct identification of the plants to divide the forest area;

[0056] S3, setting a monitoring window to capture detailed image information of the partition boundary, dividing the monitoring window into window areas according to the divided forest areas, and locating the initial dangerous window area;

[0057] S4. Expand the initial danger window area according to the area expansion logic, combine the time series images, determine the impact trend of the disease in the central area on the surrounding trees, and issue an early warning to the staff.

[0058] Figure 2 A flow chart of a forest band sensitive combination strategy for a method for maintaining tree survival in afforestation disclosed in this embodiment is shown, and the specific steps include:

[0059] S11. The staff selected several healthy plants in the forest area as sample plants, measured the chlorophyll concentration of the sample plants in the forest area using a chlorophyll meter, and included them in the chlorophyll concentration sample set;

[0060] S12. Set a chlorophyll-sensitive band combination strategy to extract chlorophyll characterization band combinations; due to the extremely large number of hyperspectral band reflectances, the vegetation index of the spectral combination can only characterize one reflected light combination. The chlorophyll-sensitive band combination strategy can couple the chlorophyll-related reflected light combination, eliminating the influence of the singleness of the result caused by a single combination. Each reflected light in the hyperspectral reflected light combination contains a large number of bands, so it is necessary to select the best reflected band combination in the best reflected light combination;

[0061] S13, obtaining the chlorophyll concentration representation value of all pixels in the forest area through the chlorophyll representation band combination;

[0062] S14. The areas of individual plants in the forest area are divided by the chlorophyll concentration characterization value to obtain the chlorophyll concentration characterization value of each plant. The plants in the forest area are relatively scattered, and the boundary with the bare soil area on the ground is very obvious, resulting in a large difference between the chlorophyll concentration characterization value of the bare soil area and the chlorophyll concentration characterization value of the plant area. Therefore, a single plant is divided by setting a threshold. This design does not need to distinguish vegetation through edge detection, threshold segmentation and other algorithms, saving computing time and system running memory, that is, the plant division can be achieved by using the existing data of the system.

[0063] Figure 3 A chlorophyll-sensitive band combination strategy flow chart of a method for maintaining tree survival in afforestation disclosed in this embodiment is shown, and the specific steps include:

[0064] S121, extracting the near infrared band reflectance data set of all pixels in the forest canopy hyperspectral image , red light band reflectance dataset and green band reflectance dataset ,in, represents the reflectivity of the ath near-infrared band, represents the reflectivity of the b-th red light band, represents the reflectance of the cth green light band, a represents the number of near-infrared bands, b represents the number of red light bands, and c represents the number of green light bands;

[0065] S122. Establish a chlorophyll sensitive band combination set , where the first chlorophyll sensitive band combination , the second chlorophyll sensitive band combination , the third chlorophyll sensitive band combination ;

[0066] S123. Calculate the chlorophyll concentration characterization value :

[0067] ;

[0068] in, represents the weight of the second chlorophyll-sensitive band combination, represents the weight of the first chlorophyll-sensitive band combination, Represents the weight of the third chlorophyll-sensitive band combination;

[0069] S124, manually selecting a sample plant area, setting a chlorophyll characterization band extraction algorithm, and selecting a band combination by calculating a correlation coefficient between the chlorophyll concentration of the sample plant area and the chlorophyll concentration characterization value.

[0070] The specific steps of the chlorophyll characterization band extraction algorithm include: calculating the correlation coefficient between the chlorophyll concentration of the sample plant area and the chlorophyll concentration characterization value, and the calculation formula is:

[0071] ;

[0072] Indicates the number of band combinations. There are a total of combinations, so , where the function Represents the correlation coefficient, function Indicates the maximum The spectral band combination, achl represents the average chlorophyll concentration of all sample plants, Represents the chlorophyll mean of all sample plants. By calculating the correlation coefficient of the mean, the influence of several plants being unrepresentative due to artificial selection is reduced, and the chlorophyll characterization band combination is obtained. .

[0073] The S14 includes setting a chlorophyll characterization threshold, extracting pixels whose chlorophyll concentration characterization values ​​of all pixels in the forest area are greater than the chlorophyll characterization threshold, performing connected component analysis, setting a component size threshold, and selecting components greater than the component size threshold among the marked connected components to select areas representing single plants in the forest area.

[0074] The forest area division includes: extracting the number of pixels of a single plant in the forest area in step S1, setting pixel number thresholds Tpi1, Tpi2, and Tpi3, and classifying the single plants in the forest area according to the pixel number thresholds, taking the plant in the upper left corner of the forest canopy hyperspectral image as the first initial traversal plant, starting to traverse all plants with the first initial traversal plant as the radius, and then calculating the mean of the pixels of the traversed plants, checking the category to which the mean belongs, until traversing to a plant that does not belong to the same category as the mean, stopping the traversal, and listing the traversed plants together with the first initial traversal plants as the first plant area; locating the plant that caused the traversal to stop as the second initial traversal plant, and starting the traversal again in the same way as above to obtain the second plant area, until all plants are traversed to obtain the Nth plant area. Since the sizes of plants in the same area are similar, this method can be used to achieve zoning according to the growth of the plants;

[0075] The detailed image information of the partition boundary captured by the monitoring window is set as follows: the size of the monitoring window is defined as H×W; a connection point is defined when three or more regions are connected; the connection points of all regions are traversed by coinciding the center point of the monitoring window with the connection point; the monitoring window is divided into regions by the connection lines of different regions; the mean of the chlorophyll concentration characterization values ​​of all plants in different monitoring window partitions is calculated; and the window partition corresponding to the maximum value of the mean of the chlorophyll concentration characterization value is extracted as the initial danger window area.

[0076] The area expansion logic is set according to the tree disease propagation characteristics, Figure 4 A schematic diagram of the regional expansion logic of a method for maintaining the survival of afforestation trees disclosed in this embodiment is shown. When planting in a forest area, trees of the same type are relatively concentrated, and the diseases between trees of the same type are contagious, and the types of high-incidence diseases between different tree species are also different. Therefore, before analysis, the plant type should be determined, and the susceptible disease type corresponding to the type and the corresponding disease transmission data between plants should be retrieved. The specific steps include: collecting historical wind speed data and disease transmission data between plants in the same period, obtaining the average wind speed data of any month and the number of diseased plants per day in the same period, and establishing a regression model: , and the wind speed influence coefficient is obtained ,in, represents the disease spread index, which is obtained by calculating the difference between the number of diseased plants on the last day and the first day of the same month. represents the daily average wind speed in the same month. is the intercept term, Represents the error term; collect the angle between the historical wind direction and the disease transmission direction, recorded as the wind direction impact angle ; Calculate the window expansion factor , get the Danger window range:

[0077] ;

[0078] in, Indicates the number of window expansions.

[0079] The area expansion logic further includes: The first danger window range is obtained, the mean chlorophyll concentration characterization value A1 of all plants in the first danger window range is calculated, and the chlorophyll characterization change standard value is set. When the difference between the mean chlorophyll concentration characterization value A2 of all plants in the first danger window range and the mean chlorophyll concentration characterization value AT of all plants in the initial danger window range is less than or equal to the chlorophyll characterization change standard value, the window is expanded again to obtain the second danger window range, and the difference between the mean chlorophyll concentration characterization value A2 of all plants in the second danger window range and the mean chlorophyll concentration characterization value AT of all plants in the initial danger window range is calculated again, until the first danger window range is obtained. The average chlorophyll concentration of all plants within the danger window When the difference between the average chlorophyll concentration of all plants within the initial danger window is greater than the standard value of chlorophyll concentration change or the When the edge of the danger window exceeds the edge of the monitoring window, the expansion stops; among them, the standard value of chlorophyll characterization change approaches 0, achieving the purpose of monitoring the health changes of plants within the window.

[0080] The judgment rules for issuing warnings to staff members include: When the difference between the mean chlorophyll concentration characterization value of all plants within the danger window and the mean chlorophyll concentration characterization value of all plants within the initial danger window is greater than the chlorophyll characterization change standard value, the first The danger window meets the conditions { Danger Window - Danger Window The plant with the smallest chlorophyll concentration within the danger window is regarded as the disease transmission center area, and the window expansion coefficient is used to calculate the disease transmission center area. The radius is used to issue an early warning to the staff; when the When the edge of the danger window exceeds the edge of the monitoring window, it indicates that the health status of the surrounding plants is close to that of the plants in the initial danger window. In order to ensure that as many diseased plants as possible are covered, the plant with the smallest chlorophyll concentration in the initial danger window area is extracted as the disease transmission center area. The window expansion coefficient is used to calculate the disease transmission center area. Issue early warning to staff for the radius;

[0081] The threshold and weight settings can be set by the operator.

[0082] Example 2

[0083] Figure 5 The framework diagram of a afforestation tree survival maintenance system of the present invention is shown. Based on the same inventive concept as that of Example 1, the present invention provides an afforestation tree survival maintenance system, including:

[0084] The data acquisition module collects hyperspectral images of forest canopies through drones and sets a sensitive combination strategy for forest bands to obtain the chlorophyll concentration representation value of a single plant;

[0085] The regional division module combines the spectral characteristics of the image and the direct identification of the plants to divide the forest area;

[0086] The danger location module sets a monitoring window to capture detailed image information of the partition boundary, divides the monitoring window into window areas according to the divided forest areas, and locates the initial danger window area;

[0087] The disease warning module expands the initial danger window area according to the area expansion logic, combines the time series images, determines the impact trend of the disease in the central area on the surrounding trees, and issues a warning to the staff.

[0088] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0089] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0090] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0091] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0092] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. A method for maintaining the survival of afforestation trees, characterized in that: The following steps are involved: S1. Collect hyperspectral images of forest canopy by drone, and set up a sensitive combination strategy for forest bands to obtain the chlorophyll concentration characterization value of a single plant; The specific steps of the forest band sensitive combination strategy include: S11. Measure the chlorophyll concentration of sample plants in the forest area using a chlorophyll meter and include them in the chlorophyll concentration sample set; S12, setting a chlorophyll sensitive band combination strategy, and extracting a chlorophyll characterization band combination; S13, obtaining the chlorophyll concentration representation value of all pixels in the forest area through the chlorophyll representation band combination; S14, dividing the area of ​​the single plant in the forest area according to the chlorophyll concentration characterization value to obtain the chlorophyll concentration characterization value of each plant; The S14 includes setting a chlorophyll characterization threshold, extracting pixels whose chlorophyll concentration characterization values ​​of all pixels in the forest area are greater than the chlorophyll characterization threshold, performing connected component analysis, setting a component size threshold, and selecting components greater than the component size threshold from the marked connected components to select an area representing a single plant in the forest area; S2, combining the spectral characteristics of the image and the direct identification of the plants to divide the forest area; The forest area division includes: extracting the number of pixels of a single plant in the forest area in step S1, setting pixel number thresholds Tpi1, Tpi2, and Tpi3, and classifying the single plants in the forest area according to the pixel number thresholds, taking the plant in the upper left corner of the forest canopy hyperspectral image as the first initial traversal plant, starting to traverse all plants with the first initial traversal plant as the radius, and then calculating the mean of the pixels of the traversed plants, checking the category to which the mean belongs, until traversing to a plant that does not belong to the same category as the mean, stopping the traversal, and listing the traversed plants together with the first initial traversal plants as the first plant area; locating the plant that caused the traversal to stop as the second initial traversal plant, and starting the traversal again in the same way as above to obtain the second plant area, until all plants are traversed to obtain the Nth plant area; S3, setting a monitoring window to capture detailed image information of the partition boundary, dividing the monitoring window into window areas according to the divided forest areas, and locating the initial dangerous window area; The detailed image information of the partition boundary captured by the monitoring window is set as follows: defining the size of the monitoring window as H×W; traversing the connection points of all regions by coinciding the center point of the monitoring window with the connection point, dividing the monitoring window into regions by the connection lines of different regions, calculating the mean of the chlorophyll concentration characterization values ​​of all plants in different monitoring window partitions, and extracting the window partition corresponding to the maximum value of the mean of the chlorophyll concentration characterization value as the initial danger window region; S4. Expand the initial danger window area according to the area expansion logic, combine the time series images, determine the impact trend of the disease in the central area on the surrounding trees, and issue an early warning to the staff.

2. A method for maintaining the survival of afforestation trees according to claim 1, characterized in that: The specific steps of the chlorophyll sensitive band combination strategy include: S121, extracting the near infrared band reflectance data set of all pixels in the forest canopy hyperspectral image , red band reflectance dataset and green band reflectance dataset ,in, represents the reflectivity of the ath near-infrared band, represents the reflectivity of the b-th red light band, represents the reflectance of the cth green light band, a represents the number of near-infrared bands, b represents the number of red light bands, and c represents the number of green light bands; S122. Establish a chlorophyll sensitive band combination set , where the first chlorophyll sensitive band combination , the second chlorophyll sensitive band combination , the third chlorophyll sensitive band combination ; S123. Calculate the chlorophyll concentration characterization value : ； in, represents the weight of the second chlorophyll-sensitive band combination, represents the weight of the first chlorophyll-sensitive band combination, Indicates the weight of the third chlorophyll-sensitive band combination; S124, setting a chlorophyll characterization band extraction algorithm, and selecting a band combination by calculating the correlation coefficient between the chlorophyll concentration of the sample plant area and the chlorophyll concentration characterization value.

3. A method for maintaining the survival of afforestation trees according to claim 2, characterized in that: The specific steps of the chlorophyll characterization band extraction algorithm include: calculating the correlation coefficient between the chlorophyll concentration of the sample plant area and the chlorophyll concentration characterization value, and the calculation formula is: ； Among them, the function Represents the correlation coefficient, function Indicates the maximum The spectral band combination, achl represents the average chlorophyll concentration of all sample plants, Represents the chlorophyll mean of all sample plants, and obtains the chlorophyll characterization band combination , Indicates the number of band combinations.

4. A method for maintaining the survival of afforestation trees according to claim 3, characterized in that: The regional expansion logic is set according to the characteristics of tree disease transmission. The specific steps include: collecting historical wind speed data and disease transmission data between plants in the same period, obtaining the average wind speed data of any month and the number of diseased plants per day in the same period, and establishing a regression model: , and the wind speed influence coefficient is obtained ,in, It represents the disease spread index, which is obtained by calculating the difference between the number of diseased plants on the last day and the first day of the same month. represents the daily average wind speed in the same month. is the intercept term, Represents the error term; collect the angle between the historical wind direction and the disease transmission direction, recorded as the wind direction impact angle ; Calculate the window expansion factor , get the Danger window range: ； in, Indicates the number of window expansions.

5. A method for maintaining the survival of afforestation trees according to claim 4, characterized in that: The area expansion logic further includes: The first danger window range is obtained, the mean chlorophyll concentration characterization value of all plants in the first danger window range is calculated, and the chlorophyll characterization change standard value is set. When the difference between the mean chlorophyll concentration characterization value of all plants in the first danger window range and the mean chlorophyll concentration characterization value of all plants in the initial danger window range is less than or equal to the chlorophyll characterization change standard value, the window is expanded again to obtain the second danger window range, and the difference between the mean chlorophyll concentration characterization value of all plants in the second danger window range and the mean chlorophyll concentration characterization value of all plants in the initial danger window range is calculated again, until the first danger window range is obtained. When the difference between the mean chlorophyll concentration characteristic value of all plants within the danger window and the mean chlorophyll concentration characteristic value of all plants within the initial danger window is greater than the chlorophyll characteristic change standard value or the When the edge of the danger window exceeds the edge of the monitoring window, the expansion stops.

6. A method for maintaining the survival of afforestation trees according to claim 5, characterized in that: The judgment rules for issuing warnings to staff members include: When the difference between the mean chlorophyll concentration characterization value of all plants within the danger window and the mean chlorophyll concentration characterization value of all plants within the initial danger window is greater than the chlorophyll characterization change standard value, the first The danger window meets the conditions { Danger Window - Danger Window The plant with the smallest chlorophyll concentration within the danger window is regarded as the disease transmission center area, and the window expansion coefficient is used to calculate the disease transmission center area. The radius is used to issue an early warning to the staff; when the When the edge of the danger window exceeds the edge of the monitoring window, the plant with the smallest chlorophyll concentration in the initial danger window area is taken as the disease transmission center area, and the disease transmission center area is expanded by the window expansion coefficient Issue early warning to staff for the radius.

7. A system for maintaining the survival of afforestation trees, used to implement a method for maintaining the survival of afforestation trees as described in any one of claims 1 to 6, characterized in that: include: The data acquisition module collects hyperspectral images of forest canopies through drones and sets a sensitive combination strategy for forest bands to obtain the chlorophyll concentration representation value of a single plant; The regional division module combines the spectral characteristics of the image and the direct identification of the plants to divide the forest area; The danger location module sets a monitoring window to capture detailed image information of the partition boundary, divides the monitoring window into window areas according to the divided forest areas, and locates the initial danger window area; The disease warning module expands the initial danger window area according to the area expansion logic, combines the time series images, determines the impact trend of the disease in the central area on the surrounding trees, and issues a warning to the staff.

Images