KR20230130895A - Method and apparatus for evaluating activity vitality of conifer based on infection symptom of wilt disease - Google Patents

Abstract

According to the present invention, a needle-leaf tree physiology vitality evaluation method based on a pine wilt disease infection symptom comprises: a step of obtaining a hyperspectral image of a target forest area; a step of preprocessing data of the obtained hyperspectral image; a step of extracting an infection determination index of the pine wilt disease from the preprocessed data; a step of determining whether a pine wilt disease infection suspect tree exists in accordance with the extracted infection determination index; a step of determining whether browning occurs if the pine wilt disease infection suspect tree is determined; a step of classifying the pine wilt disease infection suspect tree as a pine wilt disease infection tree if an occurrence of the browning is determined; and a step of evaluating the needle-leaf tree physiology vitality of the target forest area in accordance with classification results of the pine wilt disease infection tree.

Description

Method and device for evaluating conifer physiological vitality based on wilt disease infection symptoms {METHOD AND APPARATUS FOR EVALUATING ACTIVITY VITALITY OF CONIFER BASED ON INFECTION SYMPTOM OF WILT DISEASE}

The present invention relates to a technique for evaluating the physiological vitality of coniferous trees. More specifically, the present invention relates to a technique for evaluating the physiological vitality of coniferous trees based on wilt disease infection symptoms determined based on hyperspectral video images obtained from the target forest area. This relates to a method and device for evaluating coniferous physiological vitality based on wilt disease infection symptoms.

As is known, Pine wood nematode (Bursaphelenchus xylophilus) refers to a harmful pest that causes Pine Wilt Disease in pine trees, pine trees, and Japanese pine trees.

In the case of the Republic of Korea, 65.2% of the land is forest, and it has great public value not only as a source of wood or landscaping materials, but also as a source of greenhouse gas absorption and storage, provision of forest scenery, prevention of sediment runoff, supply of oxygen, air purification, and biodiversity conservation. It's a situation.

According to the Forestry Statistics Yearbook of the Korea Forest Service (2016), the forest area of pine trees is approximately 24.6% of the total forest area as of 2015, and the area of pine wilt disease is 9,048 ha as of 2015, which accounts for approximately 0.58% of pine trees. Damage is being suffered.

It has been reported that pine tree nematode disease is caused by nematodes about 1 mm in size that invade pine trees through the pine warbler or northern pine warbler and cause death by interfering with the passage of nutrients and moisture. However, the exact cause of death is not yet known.

Since pine wilt disease began in North America, it has spread throughout the world, including Japan, Far East Asia, including Korea, China, and Russia, and even Spain and Portugal in Europe.

When infected with pine nematode disease, the leaves droop downward and gradually turn dark red and die after 3 to 6 months.

Pine wilt disease has a 100% mortality rate upon infection and the risk is very serious due to its strong spread through insect vectors.

Since pine trees, which have a red color that can be identified with the naked eye, cannot be revived, the only way to minimize the spread is to cut down damaged dead trees.

Early diagnosis of pine wilt disease in pine species (pine, white pine, cypress, etc.) is essential for surveillance and early control, but it is necessary to observe nematodes under a microscope at a time when they can be identified with the naked eye or to use genetic analysis to check for the expression of pine wilt disease-specific genes. It depends.

However, the existing microscopy method or PCR test method for pine wilt nematodes has the problem of making early diagnosis impossible because the diagnostic results may be misjudged depending on the location where the sample is collected.

Korean Patent No. 10-2239392 (Announcement Date: 2021. 04. 12.)

The present invention aims to provide a method and device for evaluating the physiological vitality of conifers based on wilt disease infection symptoms, which can evaluate the vitality of coniferous trees based on wilt disease infection symptoms determined based on hyperspectral video images obtained from a target forest area. do.

The problems to be solved by the present invention are not limited to those mentioned above, and other problems to be solved that are not mentioned can be clearly understood by those skilled in the art from the description below. will be.

According to one aspect, the present invention includes the steps of acquiring a hyperspectral image of a target forest area, preprocessing data of the obtained hyperspectral image, and extracting an index for determining infection of wilt disease from the preprocessed data. A step of determining whether a tree suspected of being infected with the wilt wilt disease exists based on the extracted infection determination index; If it is determined that the tree is suspected of being infected with the wilt disease, determining whether a browning phenomenon occurs; If it is determined that an outbreak occurs, classifying the suspected tree nematode infection as a tree nematode-infected tree, and evaluating the physiological vitality of conifers in the target forest area based on the classification results of the tree nematode-infected tree. Based on the wilt disease infection symptoms. It can provide a method for evaluating the physiological vitality of coniferous trees.

The hyperspectral video image of the present invention may be acquired from a hyperspectral camera mounted on an aircraft flying over the target forest area.

In the preprocessing step of the present invention, the reflectance spectrum of the data of the hyperspectral video image may be corrected to a white board standard.

In the preprocessing step of the present invention, an area containing a minimum nonvegetation background from the original image of the preprocessed data may be selected as an evaluation target.

The indicator for determining infection of the wilt disease of the present invention may include a hyperspectral vegetation index and a hyperspectral end member extracted from the preprocessed data.

In the step of determining the presence or absence of a tree suspected of being infected with T. nematode of the present invention, if the mean and variance values of the hyperspectral vegetation index are within a preset range, it can be determined to be a tree suspected of being infected with T. nematode.

The step of determining whether a tree suspected of being infected with wilt elegans of the present invention is performed by classifying the hyperspectral index, including the hyperspectral vegetation index, using a machine learning algorithm, and then determining that the tree is suspected of being infected with wilt disease if preset conditions are met. can do.

The preset conditions of the present invention may include accuracy, predictability, and kappa index.

The step of determining the presence or absence of a suspected wilt disease infection of the present invention involves calculating the reflectance spectrum and similarity index of the hyperspectral end member, and if the calculation result is less than a preset reference value, it is determined that the elegans infection is suspected. can do.

The physiological vitality of conifers evaluated according to the present invention can be utilized in post-control work in the target forest area.

According to another aspect, the present invention is a computer-readable recording medium storing a computer program, wherein the computer program, when executed by a processor, acquires a hyperspectral image of a target forest area, and the obtained Preprocessing the data of the hyperspectral image, extracting an index for determining wilt disease infection from the preprocessed data, and determining whether a suspected elegans infection exists based on the extracted index for determining elegans infection; If it is determined that the tree is suspected of being infected with wilt elegans, determining whether a browning phenomenon has occurred; If it is determined that the browning phenomenon has occurred, classifying the tree suspected of being infected with wilt elegans as an order of wilt elegans infection; and a classification result of the order of infection by wilt elegans. Based on this, a computer-readable recording medium containing instructions for causing the processor to perform an operation including evaluating the physiological vitality of conifers in the target forest area may be provided.

According to another aspect, the present invention is a computer program stored in a computer-readable recording medium, wherein the computer program, when executed by a processor, acquires a hyperspectral image of a target forest area, and the obtained Preprocessing the data of the hyperspectral image, extracting an index for determining wilt disease infection from the preprocessed data, and determining whether a suspected elegans infection exists based on the extracted index for determining elegans infection; , if it is determined that the tree elegans is suspected of being infected, determining whether a browning phenomenon has occurred; if it is determined that the browning phenomenon has occurred, classifying the tree suspected of being infected with wilt elegans as a tree infected with wilt elegans; and classifying the tree infected with wilt elegans. A computer program including instructions for causing the processor to perform an operation including evaluating the physiological vitality of conifers in the target forest area based on the results may be provided.

According to another aspect, the present invention includes a video image acquisition unit that acquires a hyperspectral video image of a target forest area, an image preprocessor that preprocesses data of the acquired hyperspectral video image, and a wilt disease detection unit from the preprocessed data. A judgment index extraction unit for extracting the infection judgment index, a suspect tree judgment unit for determining the presence or absence of a tree suspected of being infected with wilt elegans based on the extracted infection judgment indicators, and, if it is determined to be a tree suspected of being infected with wilt elegans, a browning phenomenon. A browning phenomenon determination unit that determines whether the browning phenomenon has occurred, an infected tree determination unit that classifies the tree suspected of being infected with wilt wilt as an infected tree if it is determined that the browning phenomenon has occurred, and the target forest based on the classification results of the tree infected with wilt wilt. It is possible to provide a coniferous physiological vitality evaluation device based on wilt disease infection symptoms that includes a physiological vitality evaluation unit that evaluates the physiological vitality of coniferous trees in the region.

The video image acquisition unit of the present invention can acquire the hyperspectral video image from a hyperspectral camera mounted on an aircraft flying over the target forest area.

The image preprocessor of the present invention can correct the reflectance spectrum of data of the hyperspectral image image to a white board standard.

The image preprocessor of the present invention may select an area containing a minimum nonvegetation background from the original image of the preprocessed data as an evaluation target.

The judgment index extraction unit of the present invention may extract the hyperspectral vegetation index and hyperspectral end member extracted from the preprocessed data as the infection judgment index.

The suspicious tree determination unit of the present invention can determine the hyperspectral vegetation index. If the mean and variance values are within a preset range, it can be determined that the tree is suspicious for wilt disease infection.

The suspicious tree determination unit of the present invention may classify the hyperspectral index including the hyperspectral vegetation index using a machine learning algorithm and, if preset conditions are met, determine the tree to be suspicious for wilt disease infection.

The suspect tree determination unit of the present invention calculates the reflectance spectrum and similarity index of the hyperspectral end member, and if the calculation result is less than a preset reference value, it can be determined that the tree is suspected of being infected with the nematode infection.

According to an embodiment of the present invention, early diagnosis of pine wilt disease can be made in a non-destructive and real-time manner by evaluating the physiological vitality of coniferous trees based on symptoms of wilt disease infection determined based on hyperspectral video images obtained from the target forest area. This can minimize forest damage caused by pine wilt disease.

According to an embodiment of the present invention, by applying hyperspectral image analysis and diagnosis techniques using flying vehicles in forests that are difficult for humans to easily access, large-area pine wilt disease damage can be prevented early and a rapid and accurate surveillance system can be established. can do.

Figure 1 is a block diagram of a coniferous physiological vitality evaluation device based on wilt disease infection symptoms according to an embodiment of the present invention.
Figure 2 is a flowchart showing the main process of evaluating the physiological vitality of conifers based on the wilt disease infection symptom index according to an embodiment of the present invention.
Figure 3 is an example of the CIRI index representing the triangular area connecting three points in the NIR (R750), R (R670), and G (R550) regions from the reflectance graph and CIRI values according to the wilt disease symptom stages (R1, R2, R3). This is a graph showing.
Figure 4 is an example of a representative image of the abundance of RBG, R, dR, LogR, and vegetation index of pine and pine trees showing PWS.
Figure 5 is an example of RGB and CIR images of pine seedlings in the early (green, G), middle (yellow-brown, YB), and late (reddish-brown, RB) stages of pine wilt disease infection.

First, the advantages and features of the present invention and methods for achieving them will become clear by referring to the embodiments described in detail below along with the accompanying drawings. Here, the present invention is not limited to the embodiments disclosed below and may be implemented in various different forms. The present embodiments are merely provided to ensure that the disclosure of the present invention is complete and to provide a general understanding of the technical field to which the present invention pertains. Since this is provided as an example to enable those with knowledge to clearly understand the scope of the invention, the technical scope of the present invention should be defined by the claims.

In addition, in the following description of the present invention, if it is determined that a detailed description of a known function or configuration may unnecessarily obscure the gist of the present invention, the detailed description will be omitted. In addition, the terms described below are terms defined in consideration of the functions in the present invention, and of course, they may vary depending on the intention or custom of the user, operator, etc. Therefore, the definition should be made based on the technical ideas described throughout this specification.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings.

Figure 1 is a block diagram of a coniferous physiological vitality evaluation device based on wilt disease infection symptoms according to an embodiment of the present invention.

Referring to FIG. 1, the conifer physiological vitality evaluation device according to this embodiment includes an image acquisition unit 102, an image preprocessing unit 104, a judgment index extraction unit 106, a suspicious tree determination unit 108, and a browning It may include a phenomenon determination unit 110, an infection determination unit 112, a physiological vitality evaluation unit 114, and an evaluation information DB 116.

The video image acquisition unit 102 acquires hyperspectral video images of the target forest area. These hyperspectral video images are acquired, for example, from a hyperspectral camera mounted on an aircraft (e.g., drone, etc.) flying over the target forest area. It can be.

Depending on the geopolitical location of the target forest area, hyperspectral images may include coniferous trees such as sea pine, Japanese pine, Japanese pine, pine, strobe pine, and Rigida pine.

The image pre-processing unit 104 may perform functions such as pre-processing data of a hyperspectral image acquired through the image acquisition unit 102.

Here, preprocessing may mean correcting the reflectance spectrum of the data of the hyperspectral video image to a white board standard, and minimally containing nonvegetation background from the original image of the preprocessed data. Areas may be selected for evaluation.

The judgment index extraction unit 106 may perform functions such as extracting an infection judgment index of wilt disease from data preprocessed through the image preprocessor 104. Here, the indicator for determining wilt disease infection may include, for example, a hyperspectral vegetation index and a hyperspectral end member (unique spectral spectrum representing each pixel) extracted from preprocessed data.

The Color InfraRed Index (CIRI) of the vegetation index, which is sensitive to the early stage before the onset of pine wilt disease symptoms, can be derived as shown in Equation 1 below.

[Formula 1]

CIRI = 1/2((53(R550-R670)+9x(R670-R750)+121(R750-R550))

In Equation 1 above, R550, R670, and R750 represent reflectance values at 550, 670, and 750 nm, respectively.

Figure 3 is an example of the CIRI index representing the triangular area connecting three points in the NIR (R750), R (R670), and G (R550) regions from the reflectance graph and CIRI values according to the wilt disease symptom stages (R1, R2, R3). This is a graph showing.

First derivative transformation (dR), logarithimic transformation (logR), and hyperspectral vegetation indices (hVI) for the pixel reflectance spectrum (R) of ROI (region of interest) and endmembers can be extracted based on the Pixel Purity Index (PPI). Here, dR and logR are known to reflect the characteristics of R to a greater extent.

The suspect tree determination unit 108 performs functions such as determining the presence or absence of a tree suspected of being infected with wilt disease based on the infection judgment index (hyperspectral vegetation index and hyperspectral end member) extracted through the judgment index extraction unit 106. It can be done.

Figure 4 is an example of a representative image of the abundance of RBG, R, dR, LogR, and vegetation index of pine and pine trees showing PWS. In other words, the degree of wilt disease infection can be calculated based on the abundance per pixel of the two spectra.

As an example, the suspect tree determination unit 108 determines that the mean and variance values of the hyperspectral vegetation index are within a preset range and determines that the tree is suspected of being infected with wilt disease.

For example, if the CIRI mean and variance values are within +-10% of the values of 8.05 and 24.35 for pine trees and 5.86 and 21.81 for pine trees, the suspect tree determination unit 108 determines the tree as a tree suspected of being infected with wilt disease. You can judge.

As another example, the suspect tree determination unit 108 may classify a hyperspectral index including a hyperspectral vegetation index using a machine learning algorithm and then, if preset conditions are met, determine the tree as a tree suspected of being infected with wilt disease.

For example, when the hyperspectral indicator including CIRI is classified by a machine learning algorithm (e.g., Ensemble), if the conditions of accuracy of 80%, predictability of 70%, and kappa index of 0.7 are met, the suspicious item judgment unit 108 The tree can be judged to be suspected of being infected with wilt disease.

As another example, the suspect tree determination unit 108 calculates the reflectance spectrum and similarity index of the hyperspectral end member, and if the calculation result is less than a preset reference value, the tree may be judged to be a tree suspected of being infected with wilt. .

For example, when calculating the reflectance spectrum and similarity index of pure member end members of pine trees and pine trees, if an end member with a value of 0.05 or less is extracted, the suspect tree determination unit 108 determines that the tree is suspected of being infected with wilt disease. You can judge by the neck. Here, the CIRI value of the pine tree end member is 5.31. There are three types: 6.48 and 7.52, and in the case of pine, there are two types of end members: 6.99 and 11.61.

When the browning phenomenon determination unit 110 is notified by the suspect tree determination unit 108 that the tree is judged to be suspected of being infected with wilt disease, the browning phenomenon determination unit 110 may perform functions such as determining whether browning phenomenon has occurred in the tree.

For example, when the R value increases from 0.12 to 0.18 in the triple code of RGB or CIR color, the browning phenomenon determination unit 110 may determine that browning has occurred in the corresponding wood.

When the browning phenomenon determination unit 110 notifies the occurrence of the browning phenomenon, the infected tree determination unit 112 may perform functions such as classifying the tree suspected of being infected with wilt elegans as an elegans infection tree.

Figure 5 is an example of RGB and CIR images of pine seedlings in the early (green, G), middle (yellow-brown, YB), and late (reddish-brown, RB) stages of pine wilt disease infection.

Referring to Figure 5, you can check the hierarchical distribution, reflectance spectrum (Spectra), and distribution status (Histogram) of the NDVI and CIRI vegetation indices derived from each image, and the numbers at the bottom of the RGB and CIR images are 650 nm and 550 nm. , reflectance values at 450 nm, 750 nm, 670 nm, and 550 nm are shown, respectively.

The physiological vitality evaluation unit 114 performs functions such as evaluating the physiological vitality (physiological vitality level) of coniferous trees in the target forest area based on the classification results of wilt nematode-infected trees of each tree in the target forest area. The physiological vitality of conifers in the forest area evaluated in real time here can be stored in the evaluation information DB 116 along with coordinate information of the forest area.

In addition, the physiological vitality of coniferous trees evaluated by the physiological vitality evaluation unit 114 is transmitted in real time to the jurisdictional server in charge of the forest area, so that it can be used in real time for post-pest control work in the forest area.

Figure 2 is a flowchart showing the main process of evaluating the physiological vitality of conifers based on the wilt disease infection symptom index according to an embodiment of the present invention.

Referring to FIG. 2, the video image acquisition unit 102 acquires a hyperspectral video image of the target forest area (step 202). Here, hyperspectral video images may be acquired, for example, from a hyperspectral camera mounted on an aircraft flying over a target forest area.

The image preprocessing unit 104 preprocesses data of the hyperspectral image acquired through the video image acquisition unit 102 (step 204). Here, preprocessing the data may mean correcting the reflectance spectrum of the data of the hyperspectral video image to a white board standard.

The judgment index extraction unit 106 extracts the infection judgment index of wilt disease from the data preprocessed through the image preprocessor 104 (step 206). Here, the indicator for determining wilt disease infection may include, for example, a hyperspectral vegetation index and a hyperspectral end member (unique spectral spectrum representing each pixel) extracted from preprocessed data.

The suspect tree determination unit 108 determines the presence or absence of a tree suspected of being infected with wilt elegans based on the infection judgment index (hyperspectral vegetation index and hyperspectral end member) extracted through the judgment index extraction unit 106 (step 208). .

As an example, the suspect tree determination unit 108 may determine that the mean and variance values of the hyperspectral vegetation index are within a preset range and that the tree is suspected of being infected with wilt disease.

As another example, the suspect tree determination unit 108 may classify a hyperspectral index, including a hyperspectral vegetation index, using a machine learning algorithm, and then, if preset conditions are met, determine that the tree is a tree suspected of being infected with wilt disease.

In the suspect tree determination unit 108, as another example, after calculating the reflectance spectrum and similarity index of the hyperspectral end member, if the calculation result is less than a preset reference value, the tree in question may be judged to be a tree suspected of being infected with wilt. .

As a result of the determination in step 208, if it is determined that the tree is suspected of being infected with wilt disease, the browning phenomenon determination unit 110 determines whether browning phenomenon has occurred in the tree (step 210).

As a result of the determination in step 210, if it is determined that browning has occurred, the infected tree determination unit 112 classifies the tree suspected of being infected with wilt elegans as an elegans-infected tree (step 212).

The physiological vitality evaluation unit 114 evaluates the physiological vitality (physiological vitality level) of coniferous trees in the corresponding forest area based on the classification results of the wilt nematode-infected trees of each tree in the corresponding forest area (step 214).

In addition, the physiological vitality of conifers in the forest area evaluated in real time is stored in the evaluation information DB 116 along with coordinate information of the forest area (step 216). Here, the physiological vitality of coniferous trees evaluated by the physiological vitality evaluation unit 114 is transmitted in real time to the jurisdictional server in charge of the forest area, so that it can be used in real time for post-control work in the forest area.

The above description is merely an illustrative explanation of the technical idea of the present invention, and those of ordinary skill in the art to which the present invention pertains can make various substitutions, modifications, changes, etc. without departing from the essential characteristics of the present invention. It will be easy to see that this is possible. That is, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention, but rather to explain it, and the scope of the technical idea of the present invention is not limited by these embodiments.

Therefore, the scope of protection of the present invention should be interpreted in accordance with the claims described below, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of rights of the present invention.

102: Video image acquisition unit
104: Image preprocessing unit
106: Judgment indicator extraction unit
108: Suspicion judgment department
110: Browning phenomenon judgment unit
112: Infection tree determination unit
114: Physiological vitality evaluation unit
116: Evaluation information DB

Claims (20)

A step of acquiring a hyperspectral video image of the target forest area,
Preprocessing data of the acquired hyperspectral image;
A step of extracting infection judgment indicators of wilt disease from preprocessed data;
A step of determining the presence or absence of a suspected wilt disease infection based on the extracted infection determination index;
If it is determined that the tree is suspected of being infected with the wilt disease, determining whether browning occurs;
If it is determined that the browning phenomenon has occurred, classifying the suspected tree nematode infection as a tree infected with wilt disease;
Comprising the step of evaluating the physiological vitality of conifers in the target forest area based on the classification results of the wilt disease-infected trees.
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 1,
The hyperspectral video image is,
obtained from a hyperspectral camera mounted on an aircraft flying over the target forest area.
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 1,
The preprocessing step is,
Correcting the reflectance spectrum of the data of the hyperspectral video image to a white board standard
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 3,
The preprocessing step is,
Selecting the area containing the minimum nonvegetation background from the original image of the preprocessed data as the evaluation target.
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 1,
The indicator for determining infection of the above wilt disease is,
Containing hyperspectral vegetation index and hyperspectral end member extracted from the preprocessed data
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 5,
The step of determining whether the suspected wilt infection is present is,
The mean and variance values of the hyperspectral vegetation index are within the preset range, and the tree is judged to be suspected of being infected with the wilt disease.
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 5,
The step of determining whether the suspected wilt infection is present is,
After classifying the hyperspectral indicators including the hyperspectral vegetation index using a machine learning algorithm, if the preset conditions are met, the tree is judged to be suspected of being infected with the wilt disease.
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 7,
The preset conditions are:
including accuracy, predictability, and kappa index.
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 5,
The step of determining whether the suspected wilt infection is present is,
After calculating the reflectance spectrum and similarity index of the hyperspectral end member, if the calculation result is less than a preset reference value, the tree is judged to be suspected of being infected with the nematode wilt.
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 1,
The physiological vitality of the coniferous trees evaluated was,
Used for post-control work in the target forest area
Method for evaluating coniferous physiological vitality based on wilt disease infection symptoms. A computer-readable recording medium storing a computer program,
When the computer program is executed by a processor,
A step of acquiring a hyperspectral video image of the target forest area,
Preprocessing data of the acquired hyperspectral image;
A step of extracting infection judgment indicators of wilt disease from preprocessed data;
A step of determining the presence or absence of a suspected wilt disease infection based on the extracted infection determination index;
If it is determined that the tree is suspected of being infected with the wilt disease, determining whether browning occurs;
If it is determined that the browning phenomenon has occurred, classifying the suspected tree nematode infection as a tree infected with wilt disease;
A step of evaluating the physiological vitality of conifers in the target forest area based on the classification results of the wilt disease-infected trees.
A computer-readable recording medium including instructions for causing the processor to perform an operation including. A computer program stored on a computer-readable recording medium,
When the computer program is executed by a processor,
A step of acquiring a hyperspectral video image of the target forest area,
Preprocessing data of the acquired hyperspectral image;
A step of extracting infection judgment indicators of wilt disease from preprocessed data;
A step of determining the presence or absence of a suspected wilt disease infection based on the extracted infection determination index;
If it is determined that the tree is suspected of being infected with the wilt disease, determining whether browning occurs;
If it is determined that the browning phenomenon has occurred, classifying the suspected tree nematode infection as a tree infected with wilt disease;
A step of evaluating the physiological vitality of conifers in the target forest area based on the classification results of the wilt disease-infected trees.
A computer program including instructions for causing the processor to perform an operation including. A video image acquisition unit that acquires a hyperspectral video image of a target forest area,
an image pre-processing unit that pre-processes data of the acquired hyperspectral image;
A judgment index extraction unit that extracts an index for determining infection by wilt disease from the preprocessed data;
A suspicious tree judgment unit that determines whether a tree suspected of being infected with the wilt worm is present based on the extracted infection judgment index;
If the tree is judged to be suspected of being infected with the wilt disease, a browning phenomenon determination unit that determines whether browning phenomenon has occurred;
An infected tree determination unit that classifies the tree suspected of being infected with wilt disease as a tree infected with wilt elegans if it is determined that the browning phenomenon has occurred;
A physiological vitality evaluation unit that evaluates the physiological vitality of coniferous trees in the target forest area based on the classification results of the wilt disease-infected trees.
A device for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 13,
The video image acquisition unit,
Obtaining the hyperspectral video image from a hyperspectral camera mounted on an aircraft flying over the target forest area.
A device for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 13,
The image preprocessing unit,
Correcting the reflectance spectrum of the data of the hyperspectral video image to a white board standard
A device for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 15,
The image preprocessing unit,
Selecting the area containing the minimum nonvegetation background from the original image of the preprocessed data as the evaluation target.
A device for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 13,
The judgment indicator extraction unit,
The hyperspectral vegetation index and hyperspectral end member extracted from the preprocessed data are extracted as the infection judgment index.
A device for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 17,
The suspicion judgment department,
The mean and variance values of the hyperspectral vegetation index are within the preset range, and the tree is judged to be suspected of being infected with the wilt disease.
A device for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 17,
The suspicion judgment department,
After classifying the hyperspectral indicators including the hyperspectral vegetation index using a machine learning algorithm, if the preset conditions are met, the tree is judged to be suspected of being infected with the wilt disease.
A device for evaluating coniferous physiological vitality based on wilt disease infection symptoms. According to claim 17,
The suspicion judgment department,
After calculating the reflectance spectrum and similarity index of the hyperspectral end member, if the calculation result is less than a preset reference value, the tree is judged to be suspected of being infected with the nematode wilt.
A device for evaluating coniferous physiological vitality based on wilt disease infection symptoms.

Images