Vegetation management system and vegetation management method

Abstract

To suppress increase of cost and precisely analyze the risk that vegetation gets in contact with a facility.SOLUTION: A vegetation management system 1 comprises: a vegetation classification unit 13 which classifies photographed vegetation in remote sensing data; a long-term change prediction unit 15 which predicts a long-term wide-area fluctuation which is a range fluctuation of the vegetation in a predetermined long-term time-series, on the basis of a classification result of the vegetation and the remote sensing data; a short-term change prediction unit 16 which predicts a short-term wide-area fluctuation which is the range fluctuation of the vegetation in a predetermined short-term time-series, on the basis of the classification result of the vegetation and the remote sensing data; a risk determination unit 17 which determines the risk due to contact between the vegetation and a facility, on the basis of a prediction result by the long-term change prediction unit 15 and the prediction result by the short-term change prediction unit 16; and a visualization unit 18 which visualizes a determination result of the risk.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a vegetation management system and a vegetation management method, and is suitable for use in a vegetation management system and a vegetation management method that use remote sensing measurement information to support maintenance work on power facilities regarding vegetation contact. [Background technology]

[0002] In conventional power facility maintenance work, the routes along which power facilities such as distribution and transmission lines are located are periodically surveyed manually, and in places where there is a risk of problems such as trees coming into contact with the power facilities, tree branches are removed and herbicides are used. However, in recent years, due to labor shortages and other reasons, efforts are being made to automate the survey work.

[0003] In efforts to automate this survey work, attention is being paid to remote sensing technology that can monitor power facilities and trees (vegetation) remotely, taking into consideration that many power facilities such as power distribution and transmission lines are installed in places that are difficult for people to access, such as mountainous regions. Representative methods of remote sensing technology include the use of artificial satellites, aircraft, drones, etc. Furthermore, research and development is underway to determine vegetation contact through three-dimensional measurement using LiDAR (Light Detection and Ranging) sensors as a method of determining contact between vegetation and facilities.

[0004] For example, Patent Document 1 discloses a system for analyzing plant growth based on remote sensing images captured by remote sensing technology. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2016-123369 A Summary of the Invention [Problem to be solved by the invention]

[0006] However, in conventional technology, when analyzing the contact between facilities and vegetation using remote sensing data, it is necessary to judge the contact with vegetation by three-dimensional measurement using remote sensing sensors such as LiDAR sensors, and in order to make an accurate judgment, the shooting range becomes wide and the shooting period becomes frequent, which causes the problem of very high costs.In addition, in order to support the visualization and intuitive operation for users, it is an essential process to make the facilities and vegetation three-dimensional, and the fact that it requires a large amount of heterogeneous and time-series geographic information for this is also a factor in increasing costs.

[0007] The present invention has been made in consideration of the above points, and aims to propose a vegetation management system and method that can suppress increases in costs and accurately analyze the risk of contact between vegetation and facilities. [Means for solving the problem]

[0008] In order to solve such problems, the present invention provides a vegetation management system that uses remote sensing data obtained by remote sensing of a facility and vegetation to be analyzed to analyze the risk of the vegetation coming into contact with the facility, and includes a vegetation classification unit that classifies the vegetation photographed in the remote sensing data, a long-term change prediction unit that predicts long-term wide-area variations, which are range variations of vegetation in a specified long-term time series, based on the vegetation classification results by the vegetation classification unit and the remote sensing data, a short-term change prediction unit that predicts short-term wide-area variations, which are range variations of vegetation in a specified short-term time series, based on the vegetation classification results by the vegetation classification unit and the remote sensing data, a risk assessment unit that assesses the risk of contact between the vegetation and the facility based on the prediction results by the long-term change prediction unit and the prediction results by the short-term change prediction unit, and a visualization unit that visualizes the risk assessment results by the risk assessment unit.

[0009] In addition, in order to solve such problems, the present invention provides a vegetation management method by a vegetation management system that uses remote sensing data photographed by remote sensing of a facility and vegetation to be analyzed to analyze the risk of the vegetation coming into contact with the facility, the vegetation management system comprising: a vegetation classification step in which the vegetation management system classifies the vegetation photographed in the remote sensing data; a long-term change prediction step in which the vegetation management system predicts long-term wide-area variations, which are range variations of vegetation in a predetermined long-term time series, based on the vegetation classification result of the vegetation classification step and the remote sensing data; a short-term change prediction step in which the vegetation management system predicts short-term wide-area variations, which are range variations of vegetation in a predetermined short-term time series, based on the vegetation classification result of the vegetation classification step and the remote sensing data; a risk determination step in which the vegetation management system determines the risk of contact between the vegetation and the facility based on the prediction result of the long-term change prediction step and the prediction result of the short-term change prediction step; and a visualization step in which the vegetation management system visualizes the risk determination result of the risk determination step. Effect of the Invention

[0010] According to the present invention, it is possible to suppress increases in costs and to accurately analyze the risk of contact between vegetation and facilities. [Brief description of the drawings]

[0011] [Figure 1] 1 is a block diagram showing an example of the configuration of a vegetation management system 1 according to an embodiment of the present invention. [Diagram 2] 1 is a block diagram showing an example of a hardware configuration of a vegetation management system 1. FIG. [Diagram 3] 10 is a flowchart showing the procedure of the overall processing in the vegetation management system 1. [Figure 4] FIG. 2 is a diagram for explaining an image of data stored in a database 20. [Diagram 5] 13 is a flowchart showing an example of a processing procedure for vegetation classification processing. [Figure 6] FIG. 1 is a diagram illustrating an example of a vegetation classification map. [Figure 7] 13 is a flowchart showing an example of a processing procedure for tree height estimation processing. [Figure 8] FIG. 13 is a diagram illustrating an example of a tree height map. [Figure 9] 13 is a flowchart illustrating an example of a processing procedure for long-term change prediction processing. [Figure 10] FIG. 1 is a diagram for explaining an image of generating a long-term wide-area vegetation map and a long-term tree height prediction map. [Figure 11] 13 is a flowchart illustrating an example of a processing procedure for short-term change prediction processing. [Figure 12] 1 is a diagram illustrating an image of generating a short-term wide-area vegetation map and a short-term tree height prediction map. FIG. [Figure 13] 13 is a flowchart showing an example of a processing procedure of a risk determination process. [Figure 14] FIG. 13 is a diagram illustrating an example of growth risk determination using a growth risk model. [Figure 15] FIG. 11 is a diagram illustrating an example of a contact risk determination using a contact risk model. [Figure 16] FIG. 13 is a diagram illustrating an example of damage risk determination based on a damage simulation. [Figure 17] 13 is a flowchart illustrating an example of a processing procedure for visualization processing. [Figure 18] FIG. 13 is a diagram showing an example of a visualization image. [Figure 19] 13 is a flowchart showing an example of a processing procedure for a maintenance instruction process; [Figure 20] FIG. 13 is a diagram showing an example of a display output of a maintenance instruction. [Figure 21] FIG. 13 is a diagram illustrating an example of a maintenance support screen. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

[0013] 1 is a block diagram showing an example of the configuration of a vegetation management system 1 according to one embodiment of the present invention. The vegetation management system 1 according to this embodiment is a vegetation management system for power facility maintenance that analyzes the risk of contact between vegetation and power facilities by using remote sensing data for an area to be analyzed (target area) specified by a user.

[0014] As shown in FIG. 1, the vegetation management system 1 is a system to which a plurality of heterogeneous input data 10 (remote sensing image data 10a, vegetation information data 10b, geographic data 10c, environmental data 10d, and management data 10e) are input, and is configured to include a data acquisition unit 11 (remote sensing data acquisition unit 11a, geographic information acquisition unit 11b, environmental information acquisition unit 11c, and management information acquisition unit 11d), a database generation unit 12, a vegetation classification unit 13, a tree height estimation unit 14, a long-term change prediction unit 15, a short-term change prediction unit 16, a risk determination unit 17, a visualization unit 18, a maintenance instruction unit 19, and a database 20. Details of each unit will be described appropriately in the following description. In FIG. 1, data movement and processing transitions are represented by solid lines with arrows, but the arrows are omitted for the database 20 because it is accessed from almost all components.

[0015] Each piece of data constituting the input data 10 will now be described.

[0016] The remote sensing image data 10a is time-series image data obtained by observing the earth's surface with a remote sensing sensor, and specifically, for example, satellite photos or aerial photos by RapidEye, MODIS (Moderate Resolution Imaging Spectroradiometer), LANDSAT, Sentinel, ICESat (Ice, Cloud, and Land Elevation Satellite), etc. The remote sensing image data 10a may be, for example, time-series low-resolution remote sensing images and high-resolution remote sensing images, but the type of data of the remote sensing image data 10a in this embodiment is not limited to these, and for example, time-series images of the same remote sensing image may be used. In this description, the high and low resolutions are not based on an absolute standard but on a relative comparison.

[0017] The vegetation information data 10b is data acquired regarding physiological characteristics of vegetation, and the type of data is not limited, but may be, for example, vegetation distribution information, spectrum information, vegetation type information, and tree height (crown height) information.

[0018] Among the above data, the vegetation distribution information is information indicating the latitude, longitude, and distribution shape of the location. Spectral data is data from a spectrum library obtained by an optical sensor (particularly, a passive sensor that obtains information from light from an object) observing vegetation. A passive sensor basically has the same structure and function as the eye (naked eye), and in detail, an optical system such as a lens (corresponding to the crystalline lens of the eye) collects light from an object and forms an image on a detection system (corresponding to the retina of the eye) to obtain spectral and spatial information (e.g., color and shape). In particular, regarding the spectrum (color), while the eye only captures visible light, an optical sensor can detect a wide range of wavelengths from visible light to infrared light, and this allows the acquisition of a large amount of useful information such as the identification of minerals, rocks, and vegetation that cannot be read by the eye, the temperature of the earth's surface, land use status, and water resources and plankton resources of seas and lakes. Furthermore, an optical sensor can also obtain this useful information as a two-dimensional image over a wide area. Vegetation type information and tree height information are obtained by on-site sampling and actual measurement.

[0019] The geographic data 10c is geographic information data provided by a geographic information system (GIS), and includes information on the positions and shapes (e.g., polygons) of electric power facilities (power distribution lines, power transmission lines, etc.) and vegetation. The geographic data 10c is essential data when data acquired on-site and data acquired from satellites, aircraft, etc. are used together in the vegetation management system 1.

[0020] The environmental data 10d is data related to the local (regional) environment, and includes, for example, soil data, meteorological data, elevation data, slope data, etc. Specific examples of meteorological data include, but are not limited to, AMeDAS data, MODIS ground surface temperature data, and meteorological station data.

[0021] The management data 10e is data to be referred to when the maintenance instruction unit 19 determines the contents of the maintenance instruction, and includes information on the management and safety of the power facility. Specifically, for example, the management data 10e is operational data such as operation and safety memos, logs, and history of the power company, and is provided by the power company.

[0022] 2 is a block diagram showing an example of the hardware configuration of the vegetation management system 1. In FIG. 2, a state in which remote sensing image data 10a provided from a remote sensing observation device 2 is input to the vegetation management system 1 is shown.

[0023] 2, the remote sensing observation device 2 is an observation satellite or an aircraft equipped with a remote sensing sensor (such as LiDAR), and the type is not limited. The remote sensing observation device 2 periodically provides remote sensing image data 10a as an observation result by the remote sensing sensor. The vegetation management system 1 acquires this remote sensing image data 10a as one of the input data 10. In this embodiment, a satellite image acquired by an observation satellite is used as the remote sensing image data 10a for explanation.

[0024] The vegetation management system 1 can be realized by a general computer system (PC) with a calculation function, and is configured to include a CPU (Central Processing Unit) 1a, a RAM (Random Access Memory) 1b, and a storage device 1c, for example, as shown in Fig. 2. Note that the vegetation management system 1 may be configured in a manner other than that shown in Fig. 2, and for example, it is considered to include an output device such as a display as an output destination for maintenance instructions.

[0025] In the vegetation management system 1, each of the functional units (data acquisition unit 11, database generation unit 12, vegetation classification unit 13, tree height estimation unit 14, long-term change prediction unit 15, short-term change prediction unit 16, risk determination unit 17, visualization unit 18, and maintenance instruction unit 19) shown in FIG. 1 is implemented as a combination of multiple CPUs 1a and RAMs 1b, each of which is divided according to its role, in order to perform calculation processing. Each of these functional units can read / write data by accessing the storage device 1c. In addition, the database 20 of the vegetation management system 1 shown in FIG. 1 is implemented by the storage device 1c. Specifically, the storage device 1c can include not only storage media such as a hard disk drive (HDD) or a solid state drive (SSD) built into a PC, but also storage media such as a universal serial bus (USB) memory externally connected to the PC. In addition, a storage means such as a database connected to the PC via a network so as to be able to communicate with the PC may also be included in one form of the storage device 1c.

[0026] Fig. 3 is a flowchart showing the procedure of the overall processing in the vegetation management system 1. Prior to the start of the processing shown in Fig. 3, a region to be analyzed (target region) is designated in the vegetation management system 1 by a user.

[0027] First, in step S1, the data acquisition unit 11 inputs the input data 10 to the vegetation management system 1. In terms of the individual components of the data acquisition unit 11, the remote sensing data acquisition unit 11a acquires the remote sensing image data 10a and the vegetation information data 10b, the geographic information acquisition unit 11b acquires the geographic data 10c, and the environmental information acquisition unit 11c acquires the environmental data 10d. In this embodiment, the environmental data 10d includes meteorological data related to temperature, rainfall, or sunshine, but may be other types of data. In addition, the management information acquisition unit 11d acquires the management data 10e. In this embodiment, the management data 10e includes a monitoring log, a felling log, a maintenance location, and a time history, but may be other types of data. Then, the input data 10 acquired by the data acquisition unit 11 in step S1 is transmitted to the database generation unit 12.

[0028] In the next step S2, the database generation unit 12 stores the input data 10 transmitted from the data acquisition unit 11 in step S1 in the database 20, and also performs the data generation process described below.

[0029] The data generation process is mainly a process for preparing intermediate data necessary for each unit to perform the process described later, and in detail, the following process is performed. In the data generation process, the database generation unit 12 first performs masking of a target area designated by a user for the past time-series remote sensing image data 10a (for example, a satellite image) using the position and shape information included in the geographic data 10c. Furthermore, the database generation unit 12 generates mapping data of the target area by combining "remote sensing data" obtained by extracting the target area, which is the masking part, from the remote sensing image data 10a and "geographic data" obtained by extracting information about the target area from the geographic data 10c. Note that in this example, the remote sensing data extracted above includes at least remote sensing data indicating a digital surface model. Furthermore, the database generation unit 12 also performs masking of the target area for the vegetation information data 10b and the environmental data 10d, and generates "vegetation information data" and "environmental data" from which information about the target area is extracted. The database generating unit 12 may also mask the target area from the management data 10e and extract "management data" from which information related to the target area is extracted. Then, the database generating unit 12 stores the above-mentioned generated various data in the database 20. Furthermore, in step S2, the database generating unit 12 performs mapping by appropriately combining the remote sensing data, vegetation information data, geographic data, and environmental data stored in the database 20 in addition to the above-mentioned data, and stores the mapping data in the database 20.

[0030] In the next step S3, the vegetation classification unit 13 determines the type of vegetation based on the remote sensing data and geographic data (or mapping data combining geographic data) and the vegetation information data generated in step S2 and stored in the database 20, generates a "vegetation classification map" in which the determined vegetation type is reflected in the mapping data, and stores the map in the database 20 (vegetation classification process). The details of the vegetation classification process will be described later with reference to FIG. 5 etc.

[0031] In the next step S4, the tree height estimation unit 14 uses the vegetation classification map, remote sensing data (multi-band satellite image data of remote sensing data showing a digital surface model), geographic data, environmental data (weather data, altitude data), etc. stored in the database 20 to construct a tree height estimation model that estimates the crown height of the vegetation photographed in the remote sensing data for each vegetation classified in step S3, estimates the tree height of each vegetation using the model, generates a tree height map by mapping the estimation results to the geographic data, and stores the tree height estimation model and tree height map in the database 20 (tree height estimation process). Details of the tree height estimation process will be described later with reference to FIG. 7 etc.

[0032] In the next step S5, the long-term change prediction unit 15 predicts long-term wide-area variations, which are range variations of vegetation in a predetermined (e.g., user-specified) long-term time series, using the vegetation classification map, environmental data, and remote sensing data based on a past long-term time series stored in the database 20, and further predicts long-term tree height variations, which are tree height variations of vegetation in the same predetermined long-term time series, using the tree height estimation model generated in step S4, and stores these prediction results in the database 20 (long-term change prediction process). Details of the long-term change prediction process will be described later with reference to Fig. 9 etc.

[0033] In the next step S6, the short-term change prediction unit 16 predicts short-term wide-area variations, which are range variations of vegetation in a predetermined (e.g., user-specified) short-term time series, using the vegetation classification map, environmental data, and past short-term time-series remote sensing data stored in the database 20, and further predicts short-term tree height variations, which are tree height variations of vegetation in the same predetermined short-term time series, using the tree height estimation model generated in step S4, and stores these prediction results in the database 20 (short-term change prediction process). Details of the short-term change prediction process will be described later with reference to Fig. 11 etc.

[0034] In the next step S7, the risk determination unit 17 determines (calculates) the risk of contact between vegetation and the power facility due to time-series fluctuations based on the long-term / short-term wide-area fluctuation and tree height fluctuation prediction results by the long-term change prediction unit 15 and the short-term change prediction unit 16, and the tree height map generated by the tree height estimation unit 14, and stores the determination result (risk determination result) in the database 20 (risk determination process). In the risk determination process, the risk determination unit 17 first maps the long-term / short-term wide-area vegetation map, tree height map, and geographic data of the power facility stored in the database 20 to remote sensing data. Next, the risk determination unit 17 evaluates the two-dimensional positional relationship between the time-series fluctuation of the wide-area vegetation map and the power facility, and evaluates the three-dimensional positional relationship between the time-series fluctuation of the vegetation tree height and the power facility. Then, the risk determination unit 17 determines the vegetation growth risk, the contact risk between vegetation and the power facility, and the damage risk to the power facility based on the results of both evaluations, and stores the determination result in the database 20. The risk determination process will be described in detail below with reference to FIG.

[0035] In the next step S8, the visualization unit 18 performs two-dimensional or three-dimensional visualization (visualization process) at a time interval of a specific time (or a time designated by the user) using the prediction results of the wide-area variation and tree height variation and the risk assessment results stored in the database 20. The details of the visualization process will be described later with reference to FIG.

[0036] Then, in the final step S9, the maintenance instruction unit 19 presents a maintenance instruction to the user based on the analysis results and visualization results up to step S8 and the management data 10e (or management data) stored in the database 20 (maintenance instruction processing). Details of the maintenance instruction processing will be described later with reference to FIG.

[0037] By carrying out the above steps S1 to S9, the vegetation management system 1 can use remote sensing data to predict long-term and short-term fluctuations in vegetation for a target area specified by the user, accurately and at low cost even if the target area is wide, analyze the risk of contact between power facilities and vegetation, and instruct the user on maintenance work based on the analysis results.In addition, the vegetation management system 1 stores various data (input data, intermediate data, output data) input or generated in the above series of analyses in database 20, so that this data can be used for subsequent analyses or other analyses.

[0038] Fig. 4 is a diagram for explaining an image of the data stored in the database 20. In Fig. 4, as examples of data used as intermediate data, images of satellite images 201, which are remote sensing data, vegetation information data 202, geographic data 203, environmental data 204, and management data 205 are visually shown.

[0039] As described above in step S2, the satellite image 201 is an example of time-series remote sensing image data 10a from which a target area has been extracted by the database generator 12. Similarly, the vegetation information data 202 is an example of vegetation information data 10b from which information on the target area has been extracted, the geographic data 203 is an example of geographic data 10c from which information on the target area has been extracted, the environmental data 204 is an example of environmental data 10d from which information on the target area has been extracted, and the management data 205 is an example of management data 10e from which information on the target area has been extracted. However, for the sake of simplicity, detailed information on each piece of information is omitted from FIG. 4.

[0040] 5 is a flowchart showing an example of a processing procedure of the vegetation classification process. As described above, the vegetation classification process corresponds to the process of step S3 in FIG.

[0041] According to Fig. 5, first, the vegetation classification unit 13 acquires mapping data of the target area (step S11) generated by the database generation unit 12 by combining the remote sensing data and the geographic data in step S2 of Fig. 3. Next, the vegetation classification unit 13 acquires mapping data combining the environmental data, vegetation information data, and geographic data generated in step S2 of Fig. 3 (step S12). The processing of steps S11 and S12 may be executed in reverse order or may be executed in parallel.

[0042] In steps S11 and S12, the vegetation classification unit 13 may directly obtain the target data from the database generation unit 12, or may obtain the target data stored in the database 20. This is because, in the vegetation management system 1 according to this embodiment, basically all data, including intermediate data, is stored in the database 20, and similarly in the processing by other functional units described later, desired data may be obtained from the functional unit that generated the data, or from the database 20.

[0043] Next, the vegetation classification unit 13 uses the mapping data and vegetation information data of the target area acquired in steps S11 and S12 to determine the type of vegetation in the mapping range (step S13). A specific method of "vegetation classification" for determining the type of vegetation is, for example, to use a machine learning technique to learn spectral library data contained in the vegetation information data and estimate the classification from the spectral features of the vegetation. Note that the method of vegetation classification is not limited to the above example, and other known methods may be used.

[0044] Next, the vegetation classification unit 13 generates a vegetation classification map by recording geographic data for each type of vegetation classified in step S13, and provides the vegetation classification map to the database 20 (step S14).

[0045] The processes of steps S11 to S14 may be repeatedly executed on past mapping data in time series, and as a result, the vegetation classification unit 13 can generate a past vegetation classification map in time series.

[0046] FIG. 6 is a diagram for explaining an example of a vegetation classification map. FIG. 6(A) shows a pre-classification map image 211 as an example of mapping data of a target area before vegetation classification in step S13 in FIG. 5 is performed, and FIG. 6(B) shows a vegetation classification map image 212 as an example of a vegetation classification map generated in step S14 in FIG. 5 after vegetation classification. As shown in FIG. 6(A), the pre-classification map image 211 is a normal satellite image (remote sensing data). In contrast, the vegetation classification map image 212 shown in FIG. 6(B) is a mapping image generated by coloring the pre-classification map image 211 by vegetation classification for each type of vegetation. Note that in the vegetation classification map image 212, various types of geographic information in the target area may be color-coded by range or shape based on geographic data.

[0047] 7 is a flowchart showing an example of a processing procedure for the tree height estimation process. As described above, the tree height estimation process corresponds to the process in step S4 in FIG.

[0048] According to Fig. 7, first, the tree height estimation unit 14 acquires a multi-band satellite image of remote sensing data showing a digital surface model from the database 20 (step S21). Next, the tree height estimation unit 14 acquires geographic data, environmental data, and vegetation information data (step S22). Furthermore, the tree height estimation unit 14 acquires a vegetation classification map showing the vegetation distribution by type generated in the vegetation classification process (step S23). The processing of steps S21 to S23 may be performed in a different order or in parallel.

[0049] Next, the tree height estimation unit 14 creates a tree height estimation model that estimates the crown height for the vegetation classified in the vegetation classification process based on the data acquired in steps S21 to S23 (step S24). A known method may be adopted to create the tree height estimation model, but for example, a random forest machine learning model can be used. In this case, the tree height estimation model can be constructed by using the value of the digital surface model as the objective variable and other spectrum data, meteorological data, altitude data, vegetation information data, etc. as explanatory variables.

[0050] Next, the tree height estimation unit 14 estimates the crown height (tree height) for each type of vegetation using the tree height estimation model constructed in step S24 (step S25).Then, the tree height estimation unit 14 maps the estimation result in step S25 to the geographic data to generate a tree height map, and stores the tree height map together with the tree height estimation model constructed in step S24 in the database 20 (step S26), and ends the tree height estimation process.

[0051] FIG. 8 is a diagram for explaining an example of a tree height map. FIG. 8(A) shows a pre-estimation map image 221 as an example of mapping data of a target area before a tree height estimation process is performed, and FIG. 8(B) shows a tree height map image 222 as an example of a tree height map generated in step S26 of FIG. 7 based on the result of tree height estimation. As shown in FIG. 8(A), the pre-estimation map image 221 is a normal satellite image (remote sensing data). In contrast, the tree height map image 222 shown in FIG. 8(B) is a mapping image generated by coloring the height of the tree crowns estimated for each vegetation using a tree height estimation model on the pre-estimation map image 221. Note that in the tree height map image 222, various geographical information in the target area may be color-coded based on geographical data to indicate the range or shape.

[0052] 9 is a flowchart showing an example of the processing procedure of the long-term change prediction process. As described above, the long-term change prediction process corresponds to the process of step S5 in FIG.

[0053] The long-term change prediction process shown in Fig. 9 is a process for predicting long-term fluctuations in vegetation, but in this embodiment, "long-term" means a long-term period in which changes and fluctuations in vegetation due to the growth of trees, etc. can be confirmed, and can be considered to be a unit of, for example, about one year. In contrast, the short-term change prediction process shown in Fig. 11 is a process for predicting short-term fluctuations in vegetation, but in this embodiment, "short-term" means a short-term period in which changes and fluctuations in vegetation due to meteorological changes, such as trees falling due to weather or seasonal changes, can be confirmed, and can be considered to be a unit of, for example, about three months.

[0054] 9, first, the long-term change prediction unit 15 acquires the results of the vegetation classification process (tree height estimation model, vegetation classification map) from the database 20 (step S31). Next, the long-term change prediction unit 15 acquires multi-band satellite images of remote sensing data based on a long-term time series from the database 20 (step S32). Furthermore, the long-term change prediction unit 15 acquires environmental data, vegetation information data, and geographic data from the database 20 (step S33). The processing of steps S31 to S33 may be executed in a different order or in parallel.

[0055] Next, the long-term change prediction unit 15 generates a long-term time series change prediction model for predicting remote sensing data in the long-term future based on the long-term time series remote sensing data acquired in step S32 (step S34). The long-term time series prediction model can predict time series fluctuations by generating, for example, an RNN (recurrent neural network), but other known models may also be generated to make predictions.

[0056] Next, the long-term change prediction unit 15 uses the prediction model generated in step S34 to predict wide-area fluctuations in vegetation after a long-term designated time period (or after each designated time period) designated by the user, and generates a wide-area map that maps the prediction results (step S35). That is, the wide-area map generated in step S35 represents the prediction results of wide-area fluctuations in vegetation over a long period of time. Note that when generating the wide-area map, various data (vegetation classification map, environmental data, vegetation information data, and geographic data) acquired in steps S31 to S33 may be used as appropriate.

[0057] Next, the long-term change prediction unit 15 predicts the tree height fluctuation of the vegetation after a long-term specified time (or at each specified time lapse) by performing tree height estimation using the tree height estimation models of various vegetation generated in the tree height estimation process of Fig. 7 for the wide-area map of vegetation generated in step S35, and generates a tree height prediction map by mapping the prediction results (step S36). Note that when generating the tree height prediction map, various data (vegetation classification map, environmental data, vegetation information data, and geographic data) acquired in steps S31 to S33 may be used as appropriate.

[0058] In the following description, in order to distinguish between the wide-area map of vegetation generated in step S35 of the long-term change prediction process and the wide-area map of vegetation generated in step S45 of the short-term change prediction process described below, the former will be referred to as the long-term wide-area vegetation map and the latter will be referred to as the short-term wide-area vegetation map. Similarly, the tree height prediction map generated in step S36 of the long-term change prediction process will be referred to as the long-term tree height prediction map, and the tree height prediction map generated in step S46 of the short-term change prediction process will be referred to as the short-term tree height prediction map.

[0059] Finally, the long-term change prediction unit 15 stores the prediction results of steps S35 and S36 (the long-term wide-area vegetation map and the long-term tree height prediction map) in the database 20, and ends the long-term change prediction process.

[0060] Fig. 10 is a diagram for explaining an image of generating a long-term wide-area vegetation map and a long-term tree height prediction map. In Fig. 10, as an example of a long-term time series, the interval between times t1, t2, the present (now), and t3 is one year.

[0061] As shown in Fig. 10, in the long-term change prediction process, the long-term change prediction unit 15 predicts wide-area fluctuations in vegetation at a specified future time (e.g., time t3) from satellite images (wide-area maps 231, 232) at past times t1 and t2 using a long-term time series change prediction model, and generates a wide-area map (wide-area map 233) that maps the prediction results. In addition, although not shown in the figure, a wide-area change prediction of vegetation is also performed in a similar manner for each time after time t3 to generate a wide-area map. As a result, the wide-area map predicted for the future becomes a long-term vegetation wide-area map that represents a long-term prediction of fluctuations in vegetation in a wide range.

[0062] Next, the long-term change prediction unit 15 estimates the tree height of the vegetation for each wide-area map at each time (wide-area maps 231-233) using a tree height estimation model, and generates tree height maps (tree height maps 234-236) that map the estimation results. Of these tree height maps, the tree height maps generated for future times (time t3 and times thereafter) become long-term tree height prediction maps that represent predictions of fluctuations in the height of tree crowns over a long time series.

[0063] 11 is a flowchart showing an example of the procedure of the short-term change prediction process. As described above, the short-term change prediction process corresponds to the process of step S6 in FIG.

[0064] 11, first, the short-term change prediction unit 16 acquires the result of the vegetation classification process (vegetation classification map) from the database 20 (step S41). Next, the short-term change prediction unit 16 acquires multi-band satellite images of past short-term time-series remote sensing data from the database 20 (step S42). Furthermore, the short-term change prediction unit 16 acquires environmental data, vegetation information data, and geographic data from the database 20 (step S43). The processing of steps S41 to S43 may be performed in a different order or in parallel.

[0065] Next, the short-term change prediction unit 16 generates a short-term time series change prediction model for predicting remote sensing data in the short-term future based on the short-term time series remote sensing data acquired in step S42 (step S44). The short-term time series prediction model can predict time series fluctuations by generating, for example, an RNN (recurrent neural network), but other known models may also be generated to make predictions.

[0066] Next, the short-term change prediction unit 16 uses the prediction model generated in step S44 to predict wide-area fluctuations in vegetation after a short-term designated time period (or after each designated time period) designated by the user, and generates a short-term wide-area vegetation map by mapping the prediction results (step S45). In other words, the short-term wide-area vegetation map represents the prediction results of wide-area fluctuations in vegetation over a short period of time. Note that when generating the short-term wide-area vegetation map, various data (vegetation classification map, environmental data, vegetation information data, and geographic data) acquired in steps S41 to S43 may be used as appropriate.

[0067] Next, the short-term change prediction unit 16 predicts the tree height fluctuation of the vegetation after a short-term specified time (or at each specified time lapse) by performing tree height estimation using the tree height estimation models for various vegetation generated in the tree height estimation process of Fig. 7 for the short-term vegetation wide-area map generated in step S45, and generates a short-term tree height prediction map by mapping the prediction results (step S46). Note that when generating the short-term tree height prediction map, various data acquired in steps S41 to S43 (vegetation classification map, environmental data, vegetation information data, and geographic data) may be used as appropriate.

[0068] Finally, the short-term change prediction unit 16 stores the prediction results of steps S45 and S46 (short-term wide-area vegetation map and short-term tree height prediction map) in the database 20, and ends the short-term change prediction process.

[0069] Fig. 12 is a diagram for explaining an image of generating a short-term wide-area vegetation map and a short-term tree height prediction map. In Fig. 12, as an example of a short-term time series, the interval between times t1, t2, the present (now), and t3 is three months.

[0070] As shown in Fig. 12, in the short-term change prediction process, the short-term change prediction unit 16 predicts wide-area fluctuations in vegetation at a specified future time (e.g., time t3) from satellite images (wide-area maps 241, 242) at past times t1 and t2 using a short-term time series change prediction model, and generates a wide-area map (wide-area map 243) that maps the prediction results. In addition, although not shown in the figure, a wide-area change prediction of vegetation is also performed in a similar manner for each time after time t3 to generate a wide-area map. As a result, the wide-area map predicted for the future becomes a short-term vegetation wide-area map that represents a prediction of short-term fluctuations in vegetation in a wide range.

[0071] Next, the short-term change prediction unit 16 estimates the tree height of the vegetation for each wide-area map at each time (wide-area maps 241-243) using a tree height estimation model, and generates tree height maps (tree height maps 244-246) that map the estimation results. Of these tree height maps, the tree height maps generated for future times (time t3 and times thereafter) become short-term tree height prediction maps that represent predictions of fluctuations in tree crown height in a short-term time series.

[0072] Fig. 13 is a flowchart showing an example of the processing procedure of the risk determination process. As described above, the risk determination process corresponds to the process of step S7 in Fig. 3, and is executed by the risk determination unit 17. In Fig. 13, the processes of steps S51 and S52 may be executed in a reverse order or in parallel.

[0073] 13, first, the risk determination unit 17 acquires the short-term wide-area vegetation map and the short-term tree height prediction map generated in the short-term change prediction process, for example, from the database 20 (step S51). Next, the risk determination unit 17 acquires the long-term wide-area vegetation map and the long-term tree height prediction map generated in the long-term change prediction process, for example, from the database 20 (step S52).

[0074] Next, the risk determination unit 17 uses the short-term wide-area vegetation map and the long-term wide-area vegetation map acquired in steps S51 and S52 to construct a growth risk model that evaluates a wide-area contact risk (growth risk) between vegetation and power facilities due to short-term / long-term range fluctuations in vegetation (step S53). Specifically, the risk determination unit 17 constructs a short-term growth risk model that estimates a risk of vegetation (such as trees) falling due to meteorological fluctuations such as weather and seasonal changes over a short-term time series such as several months from the short-term wide-area vegetation map, and stores the model in the database 20. In addition, the risk determination unit 17 constructs a long-term growth risk model that estimates a contact risk due to the growth of vegetation (such as trees) over a long-term time series such as one year from the long-term wide-area vegetation map, and stores the model in the database 20.

[0075] Next, the risk determination unit 17 constructs a contact risk model that evaluates a detailed contact risk between vegetation (trees, etc.) and the power facility due to short-term / long-term tree height fluctuations in the vegetation based on the short-term / long-term growth risks calculated from the growth risk model in step S53, the short-term / long-term tree height prediction maps acquired in steps S51 and S52, and geographic data related to the power facility (e.g., power transmission lines and distribution lines), and stores the evaluation result by the model in the database 20 (step S54). That is, in determining the contact risk using the contact risk model, the geographic information of the power facility is compared with the geographic information of the boundary line of the vegetation (trees, etc.) to determine the presence or absence of contact (or the contact location). The method of constructing the contact risk model is not particularly limited, and it can be constructed using, for example, a mathematical model, deep learning, machine learning, or the like.

[0076] Next, the risk assessment unit 17 executes a damage simulation to predict the risk of damage to the power facility based on the contact risk calculated in step S54 (damage risk assessment), stores the prediction result in the database 20, and terminates the risk assessment process.

[0077] Fig. 14 is a diagram for explaining an example of growth risk determination using a growth risk model. Fig. 14 shows two-dimensional maps 251, 252, and 253 in which a wide-area map of vegetation and power facilities are mapped two-dimensionally at each time as an image of a growth risk model 250 over the time series progression of times t1, t2, and t3 (t3 is in the future).

[0078] In FIG. 14, in the two-dimensional map 251, the vegetation range 251a and the power line 251b (an example of an electric power facility) do not contact each other. In the two-dimensional map 252, the vegetation range 252a and the power line 252b do not contact each other. On the other hand, in the two-dimensional map 253, which is a future prediction model, it is shown that the vegetation range 253a and the power line 253b may contact each other at the risk location 253c. From such a growth risk model 250, the growth risk of the risk location 253c is determined at time t3. In the growth risk determination using the above-mentioned growth risk model, a determination using a short-term growth risk model and a determination using a long-term growth risk model are separately executed. In the short-term growth risk determination, it is possible to determine a short-term growth risk such as a tree falling down due to meteorological changes such as weather (rain, typhoons, etc.) and seasonal changes. In the long-term growth risk determination, it is possible to determine a long-term growth risk due to the growth of trees.

[0079] Fig. 15 is a diagram for explaining an example of contact risk determination using a contact risk model. In Fig. 15, vegetation and power facilities at each time are three-dimensionally mapped as an image of a contact risk model 260 over the time series progression of times t1, t2, and t3 (t3 is in the future). Individually, tree crown models 261, 262, and 263 showing the tree crowns of the vegetation at each time, and a utility pole 264 and a power transmission line 265, which are examples of power facilities, are shown. The layout of the power facilities can be generated from geographic data.

[0080] 15, tree crown models 261, 262 at times t1, t2 are not in contact with either a utility pole 264 or a power line 265. On the other hand, tree crown model 263 at time t3, which is a future prediction model, indicates that there is a possibility of contact with power line 265 at risk location 266. From such a contact risk model 260, the contact risk of risk location 266 at time t3 is determined.

[0081] FIG. 16 is a diagram for explaining an example of damage risk determination by damage simulation. FIG. 16 shows an image diagram of a damage risk model 270 used in the damage simulation. In the damage risk determination, a detailed determination is made for a place with a high probability of contact (for example, risk location 266 in FIG. 15) determined in the contact risk determination. Specifically, as shown in FIG. 16, the damage risk model 270 uses a three-dimensional model of the power facility and vegetation (electric pole 271, distribution line 272, vegetation 273) to determine the risk of damage to the power facility due to contact with vegetation, taking into consideration the height, wind speed, falling of trees, and the expansion of the amount of vegetation (expansion of the tree crown in this embodiment). The method of risk determination is not limited to the above example, and other known methods may be used.

[0082] As described above, in the risk determination process shown in FIG. 13, three types of determinations are performed: determination of short-term / long-term wide-area growth risk (step S53), determination of short-term / long-term contact risk (step S54), and determination of damage risk at the location where the contact risk is determined (step S55). Among these, the determination of the growth risk can be regarded as a process of predicting a rough "wide-area risk" from a growth model that estimates the future, and the determination of the contact risk can be regarded as a process of predicting a more detailed "detailed risk" using geographic data of the power facility. Furthermore, the determination of the damage risk can be regarded as a process of predicting a specific damage risk from the degree of contact for the location where the detailed risk is predicted. Therefore, the risk determination unit 17 may perform various determinations in the risk determination process in stages, for example, when the possibility of contact is roughly predicted by the determination of the growth risk (wide-area risk) using a two-dimensional model, by performing a contact risk (detailed risk) determination and damage simulation using a three-dimensional model for the predicted contact location, thereby analyzing in detail the damage risk predicted at the contact location. In the above case, the reason for using a three-dimensional model in the contact risk assessment and damage simulation is to perform highly accurate contact assessment, but this is not limited to this as long as the method can obtain highly accurate assessment results. For example, remote sensing data (satellite images) with higher accuracy than usual may be used.

[0083] Fig. 17 is a flowchart showing an example of the processing procedure of the visualization process. As described above, the visualization process corresponds to the process of step S8 in Fig. 3, and is executed by the visualization unit 18. In Fig. 17, the order of execution of the processes of steps S61 and S62 may be interchanged, or they may be executed in parallel.

[0084] 17, first, the visualization unit 18 acquires the short-term wide-area vegetation map and short-term tree height prediction map generated in the short-term change prediction process, and geographic data related to power facilities (e.g., distribution lines and transmission lines) and vegetation corresponding to these maps, from, for example, the database 20 (step S61). Next, the visualization unit 18 acquires the long-term wide-area vegetation map and long-term tree height prediction map generated in the long-term change prediction process, and geographic data related to power facilities (e.g., distribution lines and transmission lines) and vegetation corresponding to these maps, from, for example, the database 20 (step S62).

[0085] Next, the visualization unit 18 visualizes the growth risk based on the result of the growth risk determination calculated in step S53 of the risk determination process (step S63). When visualizing, the data acquired in steps S61 and S62 can be appropriately combined and used.

[0086] Next, the visualization unit 18 visualizes the contact risk based on the contact risk determination result calculated in step S54 of the risk determination process (step S64). When visualizing, the data acquired in steps S61 and S62 can be appropriately combined and used.

[0087] Next, the visualization unit 18 visualizes the damage risk based on the damage risk determination result calculated in step S55 of the risk determination process (step S65), and ends the visualization process. When visualizing, the data acquired in steps S61 and S62 can be appropriately combined and used.

[0088] 18 is a diagram showing an example of visualization of the growth risk in step S63 of FIG.

[0089] In Fig. 18, a visualized image 281 of wide-area variation and a visualized image 282 of tree height variation are shown as two-dimensional images for each of the specified times t1, t2, and t3. The visualized image 281 of wide-area variation and the visualized image 282 of tree height variation can be generated from the wide-area vegetation map and the tree height map used in the growth risk assessment at the specified time, respectively. Note that although still images are shown in Fig. 18, each visualized image may be displayed as a moving image at a specified time resolution during a specified time interval (t1 to t3).

[0090] Also, although a specific example is omitted, when visualizing the contact risk in step S64 of FIG. 17, a three-dimensional image (or video) as exemplified in FIG. 15 may be used to display the contact risk between vegetation and the power facility at the specified time. When visualizing the damage risk in step S65 of FIG. 17, a two-dimensional image (or video) as exemplified in FIG. 16 may be used to display the damage risk to the power facility due to contact with vegetation at the specified time. A three-dimensional image (or video) may also be used to visualize the damage risk. Also, the visualization of each risk determination in steps S63 to S65 of FIG. 17 may be displayed together on a management screen as in FIG. 21 described later.

[0091] 19 is a flowchart showing an example of the procedure of the maintenance instruction process. As described above, the maintenance instruction process corresponds to the process of step S9 in FIG.

[0092] 19, first, the maintenance instruction unit 19 acquires and inputs remote sensing data from the database 20 (step S71). Examples of the remote sensing data acquired in step S81 include satellite images, drone images, and vehicle-mounted camera images, but other data may also be used.

[0093] Next, the maintenance instruction unit 19 checks the determination result of the risk determination process using the remote sensing data acquired in step S81 (steps S72 and S73).

[0094] More specifically, in step S72, the maintenance instruction unit 19 checks the wide-area contact risk. Specifically, the maintenance instruction unit 19 checks the wide-area risk locations (e.g., risk location 253c in FIG. 14) determined in the growth risk determination in the risk determination process (step S53 in FIG. 13) and the wide-area risk situation (e.g., risk locations shown in the visualized image 281 in FIG. 18) visualized in the growth risk visualization in the visualization process (step S63 in FIG. 17), and selects an area of ​​interest where a wide-area contact risk is predicted between the power facility and vegetation. The area of ​​interest may be selected using a threshold based on the quantification of the risk value, but other methods may also be used.

[0095] Then, in step S73, the maintenance instruction unit 19 checks the detailed contact risk for the region of interest selected in step S72. Specifically, the maintenance instruction unit 19 checks the detailed risk locations (e.g., risk location 266 in FIG. 15) determined in the contact risk determination in the risk determination process (step S54 in FIG. 13) and the detailed risk situation visualized in the contact risk visualization in the visualization process (step S64 in FIG. 17), and identifies the detailed region where the contact risk is predicted between the power facility and vegetation. The detailed region may be identified using a threshold based on the quantification of the risk value, but other methods may also be used.

[0096] Next, the maintenance instruction unit 19 acquires the management data corresponding to the wide-area risk and the detailed risk confirmed in steps S72 and S73, and the detailed area specified in step S73, from the database 20, and determines the instruction contents of the necessary maintenance work (step S74). The instruction contents of the maintenance work determined in step S74 may include, for example, the identification of the maintenance location, the optimization of vehicle allocation, the trip planning of personnel, the transportation of equipment, and the like. In addition, the method of determining the instruction contents of the maintenance work is not limited to a specific method, but for example, the identified risk locations (detailed areas) may be listed, and the route and time for dispatching personnel and transporting equipment associated with the maintenance work may be optimized from the registered contents of the list. Then, the maintenance instruction unit 19 outputs the instruction contents of the determined maintenance work from a predetermined output device (for example, a liquid crystal display, a printer, etc.), and ends the maintenance instruction processing.

[0097] Fig. 20 is a diagram showing an example of a display output of maintenance instructions. A maintenance instruction screen 290 shown in Fig. 20 is an example of an output of the maintenance work instruction contents output in step S74 of Fig. 19, and is a display screen displayed on a display device (not shown) of the vegetation management system 1.

[0098] The maintenance instruction screen 290 in FIG. 20 shows an office, two equipment warehouses (equipment warehouses 1-2), and four risk locations (risk locations 1-4). When the maintenance instruction unit 19 determines the maintenance work instruction content (step S74 in FIG. 19), it searches for the optimal route for dispatching personnel and transporting equipment, referring to the positions of the office and equipment warehouses relative to each risk location. Furthermore, the maintenance instruction unit 19 estimates the maintenance time at each risk location by performing calculations taking into account the travel time and maintenance work time, and displays the route and time together on the maintenance instruction screen 290. Specifically, in the case of the maintenance instruction screen 290 in FIG. 20, the dates for travel to each risk location and performing maintenance work are scheduled for April 12 or April 13, and the routes for dispatching personnel and transporting equipment are shown by lines with arrows. Note that such maintenance work instruction content differs depending on the number and positions of risk locations, the positions of the offices and warehouses, and the like.

[0099] The maintenance instruction unit 19 (or the visualization unit 18) may display a maintenance support screen related to visualization and maintenance instructions on a management GUI (Graphical User Interface) or the like to support the maintenance work.

[0100] Fig. 21 is a diagram showing an example of a maintenance support screen. The maintenance support screen 300 shown in Fig. 21 is an example of a maintenance support screen displayed as a GUI on a terminal of a maintenance worker, and includes a fluctuation display block 301, a fluctuation display block 302, a risk display block 303, and a maintenance instruction block 304. Note that the detailed design and detailed items of the GUI are not limited to the example of Fig. 21.

[0101] The fluctuation display block 301 is a block that displays the predicted results of growth risk in long-term fluctuation, and the fluctuation display block 302 is a block that displays the predicted results of growth risk in short-term fluctuation. The fluctuation display blocks 301 and 302 allow risk locations and contact conditions at a desired specified time to be confirmed by operating a slide bar provided on the time axis. The risk display block 303 is a block that displays detailed information on predicted risks, and displays a list of risk locations and displays information such as latitude and longitude, risk level, and predicted time. The maintenance instruction block 304 is a block that displays instructions for maintenance work, and displays, for example, the same as the maintenance instruction screen 290 illustrated in FIG. 20.

[0102] As described above, the vegetation management system 1 according to the present embodiment uses a plurality of remote sensing data capable of photographing a wide area in a time series from the past to the present to distinguish the type of vegetation, predict long-term and short-term range changes of vegetation, and determine and visualize the risk of contact between vegetation and facilities (e.g., electric power facilities) based on the results of the long-term and short-term vegetation change prediction. Such a vegetation management system 1 utilizes two-dimensional remote sensing data to classify vegetation and predict future changes in two time series, long-term and short-term, from past measurement data (remote sensing data), and further determines and visualizes the risk of contact with facilities without necessarily using three-dimensional data. Therefore, the vegetation management system 1 according to the present embodiment can suppress the increase in cost due to the use of three-dimensional data and accurately analyze the risk of contact between vegetation and facilities.

[0103] Furthermore, the vegetation management system 1 according to this embodiment can predict not only horizontal fluctuations (wide-area fluctuations) of vegetation classified by type based on remote sensing data, but also vertical fluctuations (tree height fluctuations) of vegetation by estimating the height of tree crowns. Furthermore, the results of such fluctuation predictions performed in both long-term and short-term time series are used to determine the risk of contact with facilities, thereby enabling accurate and highly precise risk analysis to be realized at low cost.

[0104] In addition, the vegetation management system 1 of this embodiment can predict changes in vegetation volume at a time or time interval specified by the user for both horizontal variation (wide-area variation) and vertical variation (tree height variation) of vegetation classified by type based on remote sensing data, and can analyze the risk of contact with facilities, thereby providing a system that is highly convenient for users.

[0105] Furthermore, according to the vegetation management system 1 of this embodiment, the results of the growth risk assessment and contact risk assessment can be visualized in two or three dimensions, making it easier to visually judge the contact situation in areas where there is a risk of contact.

[0106] Furthermore, the vegetation management system 1 can easily allow the user to recognize appropriate maintenance work by instructing maintenance work based on the results of risk assessment and visualization. When determining the content of the maintenance work, as shown in Fig. 20, by accumulating information on multiple risk locations and making a comprehensive judgment, it is possible to optimize the introduction route of personnel and equipment, and to realize more efficient maintenance work instructions. Furthermore, by displaying both the result of risk assessment and visualization data as shown in Fig. 21, it is possible to convey necessary information to the user in an easy-to-understand manner.

[0107] Furthermore, the vegetation management system 1 according to this embodiment uses data measured in the past to predict long-term and short-term changes in tree height of vegetation, and predicts the wide area of ​​vegetation and fluctuations in tree height. At this time, by interpolating data used in tree height prediction and simulation, it is possible to prevent contact between vegetation and facilities while suppressing the frequency of actually measuring tree height on-site and without using three-dimensional data excessively (or not at all) in the analysis. As a result, the vegetation management system 1 according to this embodiment is expected to have the effect of significantly reducing the inspection costs of vegetation and facilities. Furthermore, since vegetation is analyzed based on remote sensing data that can capture a wide area, it becomes possible to monitor vegetation over a wider area than before, and it is expected to contribute greatly to the inspection and maintenance of facilities.

[0108] The present invention is not limited to the above-described embodiment, but includes various modified examples. For example, the above-described embodiment has been described in detail to clearly explain the present invention, and is not necessarily limited to those including all of the configurations described. It is also possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace a part of the configuration of an embodiment with another configuration.

[0109] In addition, the above-mentioned configurations, functions, processing units, processing means, etc. may be realized in hardware, for example, by designing a part or all of them as an integrated circuit. In addition, the above-mentioned configurations, functions, etc. may be realized in software by a processor interpreting and executing a program that realizes each function. Information such as the program, table, file, etc. that realizes each function can be stored in a memory, a recording device such as a hard disk or SSD, or a recording medium such as an IC (Integrated Circuit) card, an SD card, or a DVD (Digital Versatile Disc).

[0110] In addition, the control lines and information lines in the drawings are those that are considered necessary for the explanation, and not all control lines and information lines in the product are necessarily shown. In reality, it can be considered that almost all components are connected to each other. [Explanation of symbols]

[0111] 1 Vegetation Management System 1a CPU 1b RAM 1c storage device 2. Remote sensing observation equipment 10 Input Data 10a Remote sensing image data 10b Vegetation information data 10c Geographical Data 10d Environmental Data 10e Management Data 11 Data Acquisition Section 11a Remote sensing data acquisition section 11b Geographic Information Acquisition Department 11c Environmental Information Acquisition Department 11d Management information acquisition department 12 Database Generation Section 13 Vegetation Classification Section 14 Tree height estimation section 15 Long-term Change Prediction Department 16 Short-term Change Prediction Section 17 Risk Assessment Department 18 Visualization section 19 Maintenance instruction department 20 Database 201 Satellite Images 202 Vegetation information data 203 Geographic Data 204 Environmental Data 205 Management Data 211 Pre-classification map images 212 Vegetation Classification Map Image 221 Pre-estimation map image 222 Tree height map image 231~233,241~243 Wide Area Map 234~236,244~246 Tree height map 250 Growth Risk Model 260 Contact Risk Model 270 Damage Risk Model 281,282 Visualization image 290 Maintenance instruction screen 300 Maintenance support screen

Claims

1. A vegetation management system that uses remote sensing data obtained by photographing a facility and vegetation to be analyzed by remote sensing to analyze a risk of the vegetation coming into contact with the facility, a vegetation classification unit that classifies vegetation captured in the remote sensing data; a long-term change prediction unit that predicts long-term wide-area changes, which are range changes of vegetation in a predetermined long-term time series, based on the vegetation classification result by the vegetation classification unit and the remote sensing data; a short-term change prediction unit that predicts short-term wide-area changes, which are range changes of vegetation in a predetermined short-term time series, based on the vegetation classification result by the vegetation classification unit and the remote sensing data; a risk determination unit that determines a risk of contact between the vegetation and the facility based on a prediction result by the long-term change prediction unit and a prediction result by the short-term change prediction unit; a visualization unit that visualizes a result of the risk determination by the risk determination unit; A vegetation management system comprising:

2. a tree height estimation unit that estimates a tree height, which is a height of a tree crown of the vegetation photographed in the remote sensing data, for each vegetation classified by the vegetation classification unit; the long-term change prediction unit predicts, in addition to predicting the long-term wide-area variation, a long-term tree height variation, which is a tree height variation of vegetation in the long-term time series, based on a result of the tree height estimation by the tree height estimation unit; the short-term change prediction unit predicts short-term tree height fluctuation, which is a tree height fluctuation of vegetation in the short-term time series, based on the result of the estimation of the crown height by the tree height estimation unit, in addition to predicting the short-term wide-area fluctuation; The risk determination unit determines a risk of contact between the vegetation and the facility based on a prediction result of the long-term wide-area variation and the long-term tree height variation by the long-term change prediction unit and a prediction result of the short-term wide-area variation and the short-term tree height variation by the short-term change prediction unit. The vegetation management system according to claim 1 .

3. a maintenance instruction unit that instructs maintenance work to eliminate a risk caused by contact between the vegetation and the facility based on a result of the determination by the risk determination unit and data for visualization generated by the visualization unit. The vegetation management system according to claim 2 .

4. The risk determination unit determines the risk by: a growth risk assessment for evaluating a risk of contact between the vegetation and the facility over a wide area due to a change in the range of the vegetation predicted by the long-term wide-area change and the short-term wide-area change, for each of a long-term time series and a short-term time series that overlap at least partially; a contact risk assessment for evaluating in detail the risk of contact between the vegetation and the facility due to the tree height fluctuation of the vegetation predicted by the long-term tree height fluctuation and the short-term tree height fluctuation for each of the long-term time series and the short-term time series for which the growth risk assessment was performed; and and a damage risk assessment for predicting a damage risk of the facility by simulation based on an evaluation result of the contact risk assessment. The vegetation management system according to claim 2 .

5. The risk assessment unit, in the growth risk assessment, Using the results of the short-term wide-area fluctuation prediction, a short-term growth risk model is constructed to estimate the risk of vegetation lodging caused by meteorological fluctuation; Using the results of the long-term, wide-area predictions, a long-term growth risk model will be constructed to estimate the risks posed by vegetation growth. The vegetation management system according to claim 4 .

6. The risk assessment unit performs the growth risk assessment using a two-dimensional model, and when contact between vegetation and a facility is predicted by the assessment, performs the contact risk assessment and the damage risk assessment using a three-dimensional model for the predicted contact point. The vegetation management system according to claim 4 .

7. a data acquisition unit that acquires input data including a plurality of remote sensing data in a past time series, vegetation information data related to physiological characteristics of vegetation, geographic data including information on the positions and shapes of vegetation and facilities, environmental data related to the local environment, and management data related to the management and security of the facilities, and stores the input data in a database; the vegetation classification unit classifies vegetation captured in the remote sensing data using the input data other than the management data, and generates a vegetation classification map by mapping the results of the classification; the tree height estimation unit creates a tree height estimation model that estimates the height of a tree crown for the vegetation classified by the vegetation classification unit, using the input data other than the management data, and generates a tree height map that maps the tree height estimation results obtained by the tree height estimation model; the long-term change prediction unit uses the remote sensing data in the past long-term time series, the vegetation classification results by the vegetation classification unit, the physiological characteristics of the vegetation indicated in the vegetation information data, the geographical data, and the environmental data to generate a long-term time series change prediction model that predicts remote sensing data in the future long-term time series, generates a long-term vegetation wide-area map on which the prediction results by the long-term time series change prediction model are mapped, and further estimates tree heights for the long-term vegetation wide-area map using the tree height estimation model, and generates a long-term tree height prediction map on which the estimation results are mapped, the short-term change prediction unit uses the remote sensing data in past short-term time series, the vegetation classification results by the vegetation classification unit, the physiological characteristics of the vegetation indicated in the vegetation information data, the geographical data, and the environmental data to generate a short-term time series change prediction model that predicts remote sensing data in future short-term time series, generates a short-term vegetation wide-area map on which the prediction results by the short-term time series change prediction model are mapped, and further estimates tree heights for the short-term vegetation wide-area map using the tree height estimation model, and generates a short-term tree height prediction map on which the estimation results are mapped, The risk determination unit determines a risk of contact between the vegetation and the facility using the products of the long-term change prediction unit and the short-term change prediction unit and the geographic data; The maintenance instruction unit determines and outputs instructions for the maintenance work based on the result of the determination by the risk determination unit, the data for visualization generated by the visualization unit, and the management data. The vegetation management system according to claim 3 .

8. The predetermined long-term time series is set to a time interval that allows confirmation of changes in vegetation due to tree growth, and the predetermined short-term time series is set to a time interval that allows confirmation of changes in vegetation due to meteorological changes. The vegetation management system according to claim 1 .

9. A vegetation management method using a vegetation management system, which uses remote sensing data obtained by remotely photographing a facility and vegetation to be analyzed, to analyze a risk of the vegetation coming into contact with the facility, a vegetation classification step in which the vegetation management system classifies the vegetation photographed in the remote sensing data; a long-term change prediction step in which the vegetation management system predicts long-term wide-area changes, which are range changes of vegetation in a predetermined long-term time series, based on the vegetation classification result by the vegetation classification step and the remote sensing data; a short-term change prediction step in which the vegetation management system predicts short-term wide-area changes, which are range changes of vegetation in a predetermined short-term time series, based on the vegetation classification result by the vegetation classification step and the remote sensing data; a risk determination step in which the vegetation management system determines a risk of contact between the vegetation and the facility based on a prediction result by the long-term change prediction step and a prediction result by the short-term change prediction step; a visualization step in which the vegetation management system visualizes a result of the risk assessment step; A vegetation management method comprising:

10. The vegetation management system further includes a tree height estimation step of estimating a tree height, which is a height of a crown of the vegetation photographed in the remote sensing data, for each vegetation classified in the vegetation classification step, In the long-term change prediction step, the vegetation management system predicts long-term tree height fluctuation, which is a tree height fluctuation of vegetation in the long-term time series, based on the tree height estimation result in the tree height estimation step, in addition to predicting the long-term wide-area fluctuation, In the short-term change prediction step, the vegetation management system predicts short-term tree height fluctuation, which is a tree height fluctuation of vegetation in the short-term time series, based on the result of estimating the crown height in the tree height estimation step, in addition to predicting the short-term wide-area fluctuation; In the risk assessment step, the vegetation management system assesses a risk of contact between the vegetation and the facility based on the prediction results of the long-term wide-area variation and the long-term tree height variation in the long-term change prediction step and the prediction results of the short-term wide-area variation and the short-term tree height variation in the short-term change prediction step.

10. The vegetation management method according to claim 9.

11. The vegetation management system further includes a maintenance instruction step of instructing a maintenance work for eliminating a risk due to contact between the vegetation and the facility based on a determination result of the risk determination step and data for visualization generated in the visualization step. A vegetation management method according to claim 10.

Images