KR102648844B1 - Method and system for detecting areas where wetland component changes using time-series satellite images and GIS data - Google Patents

Abstract

The present invention relates to a method and system for detecting wetland component change areas using time-series satellite images and GIS data. The method for detecting wetland component change areas using time-series satellite images and GIS data according to an embodiment of the present invention is optional. An artificial intelligence model that learns learning data related to wetland components, including local wetland hydrology, wetland vegetation, and wetland soil, is applied to individual satellite images of the band that makes up the time-series satellite image to produce a land cover image based on satellite image. a step of defining the area of wetland components in a land cover image (pre-) produced using GIS data, and a step of detecting areas where wetland components have changed by comparing the areas of the defined wetland components in time series. It can be included.

Description

{Method and system for detecting areas where wetland component changes using time-series satellite images and GIS data}

The present invention relates to a method and system for detecting areas of change in wetland components using time-series satellite images and GIS data. More specifically, the present invention relates to a method and system for detecting wetland component changes using time-series satellite images taken by season or year, for example, to detect wetland components (e.g. hydrology). , vegetation, and soil) is about a method and system for detecting wetland component change areas using time-series satellite images and GIS data with automated technology to detect areas where changes have been made.

For example, a coastal wetland is a place where vegetation is formed under the influence of seawater due to periodic tides. These coastal wetlands are rich in inorganic and organic nutrients supplied from land and the ocean, and are characterized by the highest productivity among various types of ecosystems. On the southwestern coast of Korea, coastal wetlands are continuously disappearing due to long-term land reclamation projects, but conservation of coastal wetlands has recently been highlighted as an important value due to their ecological importance. Halophytic communities are high primary producers and at the same time play an important role as habitats for various organisms and purification of various pollutants. These halophytic plants are especially useful as indicators of the health of coastal wetlands, helping to understand the current status of distribution patterns and areas. Continuous monitoring is urgent.

In Korea, research related to halophytes has focused on investigating species distribution in local areas and the influence of growth environments. However, due to large tidal differences, a full-fledged survey on the distribution pattern and area of halophytes along the entire southwest coast, where large salt marshes are distributed, has not been conducted in earnest. This is because halophytes have the characteristic of being distributed in relatively inaccessible areas such as mudflats (wetlands), which limits the ability to investigate their distribution through field surveys.

Of course, the existing method of monitoring vegetation existing in wetlands was to acquire images by mounting a still image camera on a flying platform such as a drone, and to classify and map vegetation objects through post-processing on the ground after the shooting was completed. there is.

However, this has limitations in managing rapid vegetation changes due to micro-climate changes in wetland areas. In addition, existing wetland vegetation object recognition technology is aimed at classifying wide areas, so there are limitations in identifying the type and location of vegetation.

Meanwhile, the existing method identifies the components of wetlands by period through field visits and detects areas where the components have changed, but this is inefficient because it consumes a lot of time and cost for field visits when investigating large-scale wetlands with a large area. .

In addition, technology regarding automated methods and systems for reservoir detection using 3D spatial image data has been previously known. Although a method for detecting boundary changes in reservoirs, a type of wetland, is being developed using time-series satellite images, this method has limitations in detecting continuous changes in various wetland components.

Korean Patent Publication No. 10-2366267 (2022.02.17) Korean Patent Publication No. 10-1744662 (2017.06.01) Korean Patent Publication No. 10-2533928 (2023.05.15)

Embodiments of the present invention utilize time-series satellite images and GIS data using automated technology to detect areas where wetland components (e.g., hydrology, vegetation, and soil) have changed using time-series satellite images taken by season or year. The purpose is to provide a method and system for detecting areas of wetland component change.

The method for detecting wetland component change areas using time-series satellite images and GIS data according to an embodiment of the present invention is an artificial artificial intelligence that learns learning data related to wetland components including wetland hydrology, wetland vegetation, and wetland soil in a random area. Applying an intelligent model to individual satellite images received at designated times (e.g. daily, seasonal, etc.) from time series satellite images to produce a land cover image based on the satellite image including the wetland component, the produced land cover Defining the area of the wetland component using GIS (Geographic Information System) data from the image, comparing the area of the defined wetland component in a time-series to detect areas where the wetland component has changed. Includes.

The detection method includes the steps of selecting wetland components that do not change and are located in a wetland or surrounding area on a pixel basis from time-series satellite images taken by season or year, and selecting the time-series satellite images for the selected wetland components in pixel units. It may further include obtaining brightness value information from the multispectral band and using it as training data for the artificial intelligence model.

In the step of producing the land cover image, the image may be produced by extracting the wetland area including the wetland component from the time series satellite image.

The detection method further includes the step of calculating the area of the defined wetland component in pixel units, and the area of the individual wetland component at the designated time is the number of pixels × PSR ² (where PSR is It can be calculated to satisfy the relationship between spatial resolution of satellite image pixels.

In the detecting step, if wetland components are different in the same area in the pixel unit, an area in which wetland components have changed may be selected.

In addition, the wetland component change area detection system using time-series satellite images and GIS data according to an embodiment of the present invention provides external time-series satellite images of any area including wetland hydrology, wetland vegetation, and wetland components of wetland soil. Applying the device and the artificial intelligence model that learned the learning data related to the wetland components to individual satellite images received at a designated time from the time series satellite image to produce a land cover image based on the satellite image including the wetland components; , defines the area of the wetland component using GIS data from the land cover image produced above, and compares the area of the defined wetland component in time series to detect areas where the wetland component has changed. A wetland change area detection device. Includes.

The wetland change area detection device selects wetland components that do not change and are located in wetlands or surrounding areas on a pixel basis from time-series satellite images taken by season or year, and selects wetland components in pixel units selected from the time-series satellite images. Brightness value information can be obtained from the multispectral band of the image and used as training data for the artificial intelligence model.

The wetland change area detection device may produce an image by extracting a wetland area including the wetland component from the time-series satellite image to produce the land cover image.

The wetland change area detection device calculates the area of the wetland component area defined above in pixel units, and the area of the individual wetland component in the designated city area is the number of pixels × PSR ² (where PSR is the satellite image pixel) It can be calculated to satisfy the relational expression of (spatial resolution of).

The wetland change area detection device may select an area in which wetland components have changed when wetland components are different in the same pixel area.

The embodiment of the present invention is very effective in wetland management work because it can determine the status of wetland components that continuously change due to seasonal and rainfall characteristics and climate change.

In addition, there are a total of 2,704 inland wetlands in Korea, and the total area of inland wetlands is 1,153.4㎢, accounting for about 1% of the country's land area, so the area for wetland management is very large. Therefore, through embodiments of the present invention, it is possible to detect the current status of components that make up a large-scale wetland and areas of change in wetland components that occur by time/season without visiting the site.

1 is a diagram showing a wetland component change area detection system according to an embodiment of the present invention;
Figures 2a to 2g are diagrams for explaining the detection method of the wetland change area detection device of Figure 1;
Figure 3 is a block diagram illustrating the detailed configuration of the wetland change area detection device of Figure 1, and
Figure 4 is a flowchart showing the driving process of the wetland change area detection device of Figure 1.

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

Figure 1 is a diagram showing a wetland component change area detection system according to an embodiment of the present invention.

As shown in Figure 1, the wetland component change area detection system 90 according to an embodiment of the present invention includes a satellite 100, a communication network 110, a wetland change area detection device 120, and a third-party device 130. ) includes part or all of.

Here, “including part or all” means that the wetland component change area detection system 90 is configured by omitting some components, such as the third-party device 130, or the wetland component change area detection system 90, such as the wetland change area detection device 120. It means that some or all of the components can be integrated into a network device (e.g., wireless exchange device, etc.) constituting the communication network 110, and it is explained as including all to facilitate a sufficient understanding of the invention. do.

The artificial satellite 100 is equipped with a satellite camera that observes the Earth from outer space. The satellite camera may consist of or include an optical sensor for acquiring multispectral satellite imagery (landsat imagery). The artificial satellite 100 according to an embodiment of the present invention may be called an external device when considering the wetland change area detection device 120. Satellite cameras, referred to as 'eyes of the satellite', are divided into optical cameras, infrared cameras, and radar. Among them, optical cameras (or optical sensors) take pictures using sunlight reflected from objects that can best see objects on the ground.

The artificial satellite 100 may include an optical camera or various types of sensors (e.g., R, G, B sensors constituting a CCD camera, etc.) for observing the Earth, and the captured satellite image can be adjusted according to the performance of these components. Characteristics may be somewhat different. Therefore, in the embodiment of the present invention, it is possible to determine whether the same product is used in the sensor of the satellite 100 to detect changes in the components that make up the wetland, that is, hydrology, vegetation, and soil. Of course, this operation is performed in the wetland. This may be possible by previously storing and confirming spec information related to the performance of the satellite 100 in the change area detection device 120. Above all, the satellite 100 according to an embodiment of the present invention can provide multispectral images.

Here, a multispectral (or multispectral) image is an image that acquires image data belonging to a specific wavelength range of the electromagnetic spectrum. Wavelengths can be separated using filters or instruments sensitive to specific wavelengths, including light from frequencies outside the visible range, such as microwaves, infrared, ultraviolet, and X-rays. In other words, multispectral images are a collection of multiple monochromatic images of the same scene, taken with different sensors. Here, each monochromatic image is called a band. The well-known multispectral (or multiband image) is an RGB color image consisting of red, green, and blue images, each captured by a sensor sensitive to different wavelengths. Such spectral imaging may allow extraction of additional information that the human eye does not capture with its receptors for red, green, and blue. Multispectral imaging typically measures light in 3 to 15 spectral bands (or bands). However, in multispectral imaging, imaging with higher radiometric resolution and finer spectra, including hundreds or thousands of bands, is called hyperspectral. Therefore, in the embodiment of the present invention, there will be no particular limitation on the type of spectroscopic imaging performed.

The satellite 100 in FIG. 1 may include various types of satellite devices for acquiring satellite images according to an embodiment of the present invention. For example, in an embodiment of the present invention, in order to collect satellite images to detect changes in the components that make up a wetland, images may be collected by accessing a server such as the National Meteorological Satellite Center. For example, the wetland change area detection device 120 of FIG. 1 can be linked to the server of the National Meteorological Center through API (Application Programming Interface) to receive satellite images related to wetland change area detection in real time. Of course, the satellite 100 periodically transmits captured images through communication with the National Meteorological Center. Therefore, according to an embodiment of the present invention, the wetland change area detection device 120 can access the server operated by the National Meteorological Center to receive and use multi-spectral images captured by the satellite 100.

The communication network 110 includes both wired and wireless communication networks. For example, a wired or wireless Internet network may be used or linked as the communication network 110. Here, the wired network includes Internet networks such as cable networks and public switched telephone networks (PSTN), and the wireless communication network includes CDMA, WCDMA, GSM, EPC (Evolved Packet Core), LTE (Long Term Evolution), and Wibro networks. It means including. Of course, the communication network according to the embodiment of the present invention is not limited to this, and can be used as an access network for a next-generation mobile communication system to be implemented in the future, for example, a cloud computing network in a cloud computing environment, a 5G network, etc. For example, if the communication network is a wired communication network, the access point in the communication network 110 can connect to the telephone company's exchange office, etc., but in the case of a wireless communication network, data is processed by connecting to the SGSN or GGSN (Gateway GPRS Support Node) operated by the communication company, or Data can be processed by connecting to various repeaters such as BTS (Base Transceiver Station), NodeB, and e-NodeB.

Communication network 110 may include an access point. Access points include small base stations such as femto or pico base stations that are often installed in buildings. Here, femto or pico base stations are classified according to the maximum number of wetland change area detection devices (120) or third-party devices (130) that can be connected, according to the classification of small base stations. Of course, the access point includes a short-distance communication module for performing short-distance communication such as Zigbee and Wi-Fi with the wetland change area detection device 120 or a third-party device 130. Access points can use TCP/IP or RTSP (Real-Time Streaming Protocol) for wireless communication. Here, in addition to Wi-Fi, short-range communication can be performed using various standards such as Bluetooth, Zigbee, infrared (IrDA), RF (Radio Frequency) such as UHF (Ultra High Frequency) and VHF (Very High Frequency), and ultra-wideband communication (UWB). You can. Accordingly, the access point extracts the location of the data packet, specifies the best communication path for the extracted location, and forwards the data packet along the designated communication path to the next device, such as the wetland change area detection device 120. . Access points can share multiple lines in a typical network environment and include, for example, routers, repeaters, and repeaters.

The wetland change area detection device 120 performs an automated operation to detect areas where wetland components (e.g., hydrology, vegetation, and soil) have changed using time series satellite images captured by season or year. Wetlands are very complex ecosystems. The elements that make up wetlands include wetland hydrology, wetland soil, and wetland vegetation, and the components of wetlands change periodically according to seasonal characteristics. In an embodiment of the present invention, artificial intelligence technology is applied to the multispectral bands of time-series satellite images taken in the same wetland by season or year, and the wetland is analyzed from individual satellite images, that is, satellite images by period (e.g., day, month, season, etc.). Create a map of wetland components by period classifying the components. And by comparing wetland component maps by period, if the components of the same area are different, it can be selected as an area where wetland component changes occur.

As will be discussed in more detail later, the wetland change area detection device 120 according to an embodiment of the present invention detects the change area of wetland components using time series satellite images and GIS data. To this end, the wetland change area detection device 120 selects wetland components that do not change and are located in wetlands or surrounding areas on a pixel basis from time-series satellite images taken by season or year, and determines the brightness value (e.g., brightness value) of the selected individual component pixel. : After obtaining information (gray value, etc.) from multispectral bands, apply this to individual satellite images, that is, time-series satellite images, to satellite images of a specific period such as day, month, or season to produce land cover images based on satellite images. do. Land cover images can refer to images that classify wetland components by applying an artificial intelligence model from individual satellite images of time series satellite images. At this time, GIS data can be used to extract land cover images corresponding to wetland areas and define them as wetland component images. In other words, wetland components can be defined by expressing (or recording) the area corresponding to the wetland area in the land cover image as GIS data. GIS data can be provided from a third-party device 130 and can be generated in the form of a GIS shape file (Shapefile), and the shape file is a vector type used to store the location and attribute information of points, lines, and surfaces. This refers to the file structure.

In addition, the wetland change area detection device 120 can determine the status of wetland components by period by producing wetland component images (or area images) using GIS data and then calculating the area of the components in pixel units. there is. The wetland change area detection device 120 compares wetland component images produced from individual satellite images (e.g., spring of a year ago and spring of this year, etc.) in time series and configures wetland components when wetland components are different in the same area in pixel units. You can select an area where elements have changed and express the changed area. The wetland change area detection device 120 can segment the area in which wetland components have changed, display the area in which wetland components have changed in detail, and calculate statistics. Of course, the display can be provided by a third-party device 130. For example, if a specific area is vegetation in the first wetland component image of a wetland and the same area is soil in the second wetland component image, the area is defined as a 'vegetation → soil' area and displayed in a unique color. And statistics related to this can be calculated in terms of area in pixel units.

The third-party device 130 may include various types of devices operated by the National Meteorological Center, or may include various devices from program manufacturers that develop programs according to embodiments of the present invention. It may also include a management device for officials to detect areas of wetland change. For example, in the case of the National Meteorological Center, satellite images captured through the satellite 100 can be provided, stored, and provided in real time upon request by the wetland change area detection device 120. Additionally, in the case of a program manufacturer, a program for detecting wetland component change areas according to an embodiment of the present invention can be developed and mounted on the wetland change area detection device 120.

As a result of the above configuration, it is very effective in wetland management work as it is possible to identify the status of wetland components that continuously change due to seasons, rainfall characteristics, and climate change. There are a total of 2,704 inland wetlands in Korea, and the total area of inland wetlands is 1,153.4㎢. Since it occupies about 1% of the country's area, the area for wetland management is very large. Therefore, through embodiments of the present invention, it will be possible to detect the current status of components that make up a large-scale wetland and areas of change in wetland components that occur by time/season without visiting the site.

FIGS. 2A to 2G are diagrams for explaining the detection method of the wetland change area detection device of FIG. 1.

For convenience of explanation, referring to FIGS. 2A to 2G together with FIG. 1, the wetland change area detection device 120 according to an embodiment of the present invention is located in a wetland or surrounding area using time series satellite images taken by season or year. You can create training samples, or learning data, for unchanged wetland components, such as hydrology, vegetation, and soil, and perform operations to learn them.

Figure 2a shows the process of selecting unchanged wetland components located in a wetland or surrounding area on a pixel-by-pixel basis from time-series satellite images captured by season or year. When looking at the images in (a) and (b) of Figure 2A, the red, green, and blue squares in the time series image show unchanged wetland components.

The wetland change area detection device 120 in FIG. 1 acquires the brightness value information of individual component pixels selected in pixel units in FIG. 2A from the multispectral band, defines and labels this as a wetland component training sample of the individual satellite image. do. That is, as shown in Figure 2b, training samples, or learning data, can be generated through pixel brightness value information acquired from each band image for wetland hydrology and wetland vegetation, and then trained on an artificial intelligence program or deep learning program. According to this learning operation, the artificial intelligence program mounted on the wetland change area detection device 120 produces learning results based on the previously learned learning data for each individual satellite image of the time series satellite image that has been input and stored or is input later. It will be printed. In the case of artificial intelligence programs, various methods such as supervised learning, unsupervised learning, and semi-supervised learning methods that combine two learning methods can be used, so embodiments of the present invention will not be particularly limited to any one method.

In addition, the wetland change area detection device 120 applies the learned model, that is, the artificial intelligence program, to individual satellite images by learning the training samples of wetland components through artificial intelligence/deep learning algorithms, that is, to daily or seasonal satellite images. Thus, land cover images based on satellite images can be produced. Using individual satellite images, we produce classified images of wetland areas. Figure 2c shows this process. Figure 2a shows a time series satellite image, while Figure 2c shows a land cover image. In other words, the land cover image can be viewed as an image created by classifying wetland areas from individual satellite images at a specific point in time series satellite images according to the learning results of an artificial intelligence model.

Next, the wetland change area detection device 120 can use GIS data to extract a land cover image corresponding to the wetland area and define it as a wetland component image. In other words, the area corresponding to the wetland area in the land cover image can be expressed or recorded as GIS data to define the area of the wetland component, and the image where this definition is made can be viewed in the wetland component image. Figures 2d and 2e show wetland component images. In Figure 2, the wetland component image shows wetland hydrology, wetland vegetation, and wetland soil in an integrated form. Of course, the wetland component image according to an embodiment of the present invention may be an image defining an area limited to the area of the wetland component that changes.

The wetland change area detection device 120 can determine the status of wetland components by period (e.g., by season or year) by producing an image of wetland components and calculating the area of the components in pixel units. The area of a component can be expressed as <Relational Equation 1>.

[Relational Expression 1]

Individual wetland component area: number of pixels ×

(Here, PSR is the spatial resolution of satellite image pixels (e.g. in m, km units))

Based on <Relational Equation 1>, for example, if the spatial resolution is 50cm, an area of 2500cm horizontally and vertically can be considered to be detected as 1 pixel.

In addition, the wetland change area detection device 120 compares wetland component images produced from individual satellite images in time series, and if wetland components are different in the same area in pixel units, selects this as an area with changed wetland components and selects the changed area. It can be expressed. Figure 2f is a diagram showing areas of wetland component change.

Furthermore, the wetland change area detection device 120 can subdivide the area in which wetland components have changed, as shown in FIG. 2g, to specifically display the area in which wetland components have changed and calculate statistics. For example, if a specific area is vegetation in the first wetland component image of a wetland and the same area is soil in the second wetland component image, the area is defined as a 'vegetation → soil' area and displayed in a unique color. And statistics related to this can be calculated in terms of area in pixel units. Figure 2g shows the wetland components changed from vegetation to soil and the wetland components changed from hydrology to vegetation, respectively.

In addition to the above, further details may be given regarding the satellite 100, wetland change area detection device 120, and third-party device 130 of FIG. 1, and other detailed information will be replaced by those contents.

Figure 3 is a block diagram illustrating the detailed configuration of the wetland change area detection device of Figure 1.

As shown in FIG. 3, the wetland change area detection device 120 of FIG. 1 according to an embodiment of the present invention includes a communication interface unit 300, a control unit 310, a wetland change area detection unit 320, and a storage unit. Includes part or all of (330).

Here, “including part or all” means that the wetland change area detection device 120 of FIG. 1 is configured by omitting some components such as the storage unit 330, or the wetland change area detection unit 320 is configured by the control unit ( 310), it means that it can be integrated and configured with other components such as

The communication interface unit 300 communicates with the satellite 100 and the third-party device 130 of FIG. 1 via the communication network 110 of FIG. 1, respectively. The communication interface unit 300 can perform operations such as modulation/demodulation, encoding/decoding, muxing/demuxing, etc. during communication. This is obvious to those skilled in the art, so further explanation will be omitted.

The communication interface unit 300 can communicate with the satellite 100 or a third-party device 130 to receive time-series satellite images captured by season or year and transmit them to the control unit 310. For example, when the communication interface unit 300 is interconnected with the third-party device 130 and API, it can periodically receive time-series satellite images of an arbitrary area including a wetland and transmit them to the control unit 310. Of course, when the wetland change area detection device 120 provides an online service, that is, a platform-type service, the communication interface unit 300 can provide related services to wetland-related officials or general users if necessary.

The control unit 310 is responsible for the overall control operation of the communication interface unit 300, wetland change area detection unit 320, and storage unit 330 of FIG. 3. The control unit 310 can receive time-series satellite images through the communication interface unit 300, temporarily store them in the storage unit 330, retrieve them, and provide them to the wetland change area detection unit 320 for image analysis. In addition, when there is a request from the wetland change area detection unit 320, the control unit 310 displays the area where wetland components have changed as shown in FIG. 2f and provides the communication interface unit 300 to the third party device 130 of FIG. 1. ) or to control the communication of the communication interface unit 300 to specifically display the area where wetland components have changed, calculate statistics, and provide the statistics to the third-party device 130 of FIG. 1, as shown in FIG. 2g. You can.

In addition, when time-series satellite images are provided through the communication interface unit 300, the control unit 310 can systematically classify and store them in the DB 120a of FIG. 1. In this case, the data is stored through an internal dedicated network such as an intranet. It can be transmitted and saved. Upon request from the wetland change area detection unit 320, the control unit 310 may retrieve and provide time series satellite images stored in the DB 120a for data analysis. Additionally, the control unit 310 may store analysis results based on data analysis of time-series satellite images in the DB 120a of FIG. 1. Of course, in the DB 120a of FIG. 1, data can be classified and stored to detect the current status of components that make up a large-scale wetland and areas of change in wetland components that occur by time/season, and the corresponding data may be graphically processed by the wetland change area detection unit 320 and provided to the manager device of the person involved in managing the wetland constituting the third-party device 130 of FIG. 1.

The wetland change area detection unit 320 operates to easily detect wetland components that make up a large-scale wetland. To this end, the wetland change area detection unit 320 is equipped with an artificial intelligence program and can use it to produce land cover images based on satellite images. In other words, the wetland change area detection unit 320 can acquire time-series satellite images taken by season or year, that is, time-series satellite images of the wetland or its surrounding areas, and use them to generate learning data for detecting wetland change areas. there is. Here, the learning data can be called a training sample, and the unchanging wetland components located in the wetland or the surrounding area in the time series satellite image can be considered as the standard. Unchanged wetland components are selected in pixel units, and brightness information of individual components, such as wetland hydrology or wetland vegetation pixels, is acquired from the multispectral bands of time-series satellite images, and these are used as training samples for learning artificial intelligence learning models. It can be created from data. The wetland change area detection unit 320 can label this, such as a supervised learning method.

When learning of the artificial intelligence learning model is completed, the wetland change area detection unit 320 applies the learning model to satellite images at a specific point in time (e.g., by season, year, etc.) from individual satellite images, that is, time-series satellite images, and uses the satellite image as the base. Produce land cover video. The satellite image-based land cover image is an image generated by classifying wetland areas from time series satellite images and is clearly shown in Figure 2c. And the wetland change area detection unit 320 can generate an image with defined wetland components using GIS data from the satellite image-based land cover image. This is clearly shown in Figure 2d. In a land cover image, an image in which the wetland area is expressed or recorded as GIS data can be a wetland component image.

The wetland change area detection unit 320 can produce an image of wetland components and calculate the area of the components in pixel units to generate the status of wetland components by period so that wetland area managers, etc. can identify them. The area of individual wetland components for each period can be calculated by multiplying the square of the spatial resolution (PSR) of the satellite image pixels and the number of pixels. In this way, wetland component images produced from individual satellite images (e.g., last year's spring and this year's spring, etc.) are compared in time series, and if wetland components are different in the same area at the pixel level, the area where wetland components have changed is selected and Information can be provided to managers who manage wetlands.

The storage unit 330 may temporarily store various types of information or data that are processed under the control of the control unit 310. The storage unit 330 may store data from, for example, time-series satellite images and then provide the data to the wetland change area detection unit 320 for image analysis.

In addition to the above, the communication interface unit 300, control unit 310, wetland change area detection unit 320, and storage unit 330 of FIG. 3 can perform various operations, and other details have been sufficiently explained previously. I would like to replace it with those contents.

The communication interface unit 300, control unit 310, wetland change area detection unit 320, and storage unit 330 of FIG. 3 according to an embodiment of the present invention are composed of hardware modules that are physically separated from each other, but each module will be able to store software to perform the above operations internally and execute it. However, the software is a set of software modules, and each module can be formed of hardware, so there will be no particular limitation on the configuration of software or hardware. For example, the storage unit 330 may be hardware, such as storage or memory. However, since it is possible to store information through software (repository), the above content will not be specifically limited.

Meanwhile, as another embodiment of the present invention, the control unit 310 may include a CPU and memory, and may be formed as a single chip. The CPU includes a control circuit, an arithmetic unit (ALU), an instruction interpretation unit, and a registry, and the memory may include RAM. The control circuit performs control operations, the operation unit performs operations on binary bit information, and the command interpretation unit includes an interpreter or compiler, which can convert high-level language into machine language and machine language into high-level language. , the registry may be involved in software data storage. According to the above configuration, for example, at the beginning of the operation of the wetland change area detection device 120, the program stored in the wetland change area detection unit 320 is copied, loaded into memory, that is, RAM, and then executed to perform data operation. Processing speed can be quickly increased. In the case of deep learning models, they can be loaded into GPU memory rather than RAM and executed by accelerating the execution speed using GPU.

Figure 4 is a flowchart showing a method for detecting wetland component change areas using time-series satellite images and GIS data according to an embodiment of the present invention.

For convenience of explanation, referring to FIG. 4 together with FIG. 1, the wetland change area detection device 120 according to an embodiment of the present invention learns learning data related to wetland hydrology, wetland vegetation, and wetland components of wetland soil in a random area. An artificial intelligence model, that is, an artificial intelligence program, is applied to individual satellite images at a designated time from time series satellite images to produce a land cover image based on satellite images that classifies wetland components (S400). Here, the time series satellite image may refer to a satellite image of a wetland or its surrounding area, as shown in Figure 2a, and the land cover image may be an image limited to the wetland area.

Additionally, the wetland change area detection device 120 can define the area of wetland components in a land cover image already produced using GIS data (S410). For example, a pre-produced land cover image may include areas of wetland components that change and areas of wetland components that do not change. For example, in an embodiment of the present invention, since we are interested in the area of changing wetland components, GIS data can be specified along the area.

Furthermore, the wetland change area detection device 120 detects areas where wetland components have changed by comparing the areas of the defined wetland components in time series (S420). For example, the wetland component area may be defined through GIS data, that is, the area of the wetland component defined by season or year, and the area of the defined area may be calculated in pixel units. In this way, by identifying or analyzing the current status of wetland components by season or year, it is possible to detect which components have changed into which components. For example, vegetation can change into soil, and hydrology can change into vegetation, and it can be said that these changes are detected.

In addition to the above, the wetland change area detection device 120 of FIG. 1 can perform various operations, and other details have been sufficiently explained previously, so these will be replaced.

Meanwhile, even though all the components constituting the embodiment of the present invention are described as being combined or operating in combination, the present invention is not necessarily limited to this embodiment. That is, as long as it is within the scope of the purpose of the present invention, all of the components may be operated by selectively combining one or more of them. In addition, although all of the components may be implemented as a single independent hardware, a program module in which some or all of the components are selectively combined to perform some or all of the combined functions in one or more pieces of hardware. It may also be implemented as a computer program having. The codes and code segments that make up the computer program can be easily deduced by a person skilled in the art of the present invention. Such computer programs can be stored in non-transitory computer readable media and read and executed by a computer, thereby implementing embodiments of the present invention.

Here, a non-transitory readable recording medium refers to a medium that stores data semi-permanently and can be read by a device, rather than a medium that stores data for a short period of time, such as a register, cache, or memory. . Specifically, the above-described programs may be stored and provided on non-transitory readable recording media such as CD, DVD, hard disk, Blu-ray disk, USB, memory card, ROM, etc.

In the above, preferred embodiments of the present invention have been shown and described, but the present invention is not limited to the specific embodiments described above, and may be used in the technical field to which the invention pertains without departing from the gist of the invention as claimed in the claims. Of course, various modifications can be made by those skilled in the art, and these modifications should not be understood individually from the technical idea or perspective of the present invention.

100: satellite 110: communication network
120: Wetland change area detection device 130: Third-party device
300: communication interface unit 310: control unit
320: wetland change area detection unit 330: storage unit

Claims (10)

An artificial intelligence model that learns learning data related to wetland components, including wetland hydrology, wetland vegetation, and wetland soil in a random area, is applied to individual satellite images received at a specified time from time-series satellite images to generate wetland components containing the wetland components. Producing a land cover image based on satellite imagery;
defining an area of the wetland component using GIS (Geographic Information System) data from the produced land cover image;
Comprising the areas of the wetland components defined above in a time-series to detect areas where wetland components have changed,
It further includes calculating the area in pixel units for the area of the wetland component defined above,
The area of individual wetland components at the above designated time is calculated to satisfy the relationship of number of pixels × PSR ² (where PSR is the spatial resolution of satellite image pixels). Wetland components using time series satellite images and GIS data. Change area detection method. According to paragraph 1,
Selecting, on a pixel basis, wetland components that do not change and are located in wetlands or surrounding areas from time-series satellite images captured by season or year; and
Obtaining brightness value information from the multi-spectral band of the time-series satellite image for the selected wetland component in pixel units and using it as training data for the artificial intelligence model;
Additionally, a method for detecting wetland component change areas using time series satellite images and GIS data. According to paragraph 2,
The step of producing the land cover image is,
A method for detecting wetland component change areas using time-series satellite images and GIS data, which produces images by extracting wetland areas containing the wetland components from the time-series satellite images. delete According to paragraph 1,
The detection step is,
A method for detecting wetland component change areas using time-series satellite images and GIS data to select an area with changed wetland components when wetland components are different in the same pixel-level area. An external device that provides time-series satellite images of an arbitrary area, including wetland hydrology, wetland vegetation, and wetland soil components; and
An artificial intelligence model that has learned learning data related to the wetland components is applied to individual satellite images received at a designated time from the time series satellite image to produce a land cover image based on the satellite image including the wetland component, and the production of the above It includes a wetland change area detection device that uses GIS data from a land cover image to define the area of the wetland component, and compares the area of the defined wetland component in time series to detect areas where the wetland component has changed. ,
The wetland change area detection device calculates the area of the wetland component area defined above in pixel units, and the area of the individual wetland component in the designated city area is the number of pixels × PSR ² (where PSR is the satellite image pixel) A wetland component change area detection system using time-series satellite images and GIS data, which is calculated to satisfy the relational expression of (spatial resolution of). According to clause 6,
The wetland change area detection device selects wetland components that do not change and are located in wetlands or surrounding areas on a pixel basis from time-series satellite images taken by season or year, and selects wetland components in pixel units selected from the time-series satellite images. A wetland component change area detection system using time-series satellite images and GIS data, which acquires brightness value information from the multispectral bands of the image and uses it as learning data for the artificial intelligence model. In clause 7,
The wetland change area detection device extracts a wetland area including the wetland component from the time-series satellite image to produce an image of the wetland component using time-series satellite image and GIS data to produce the land cover image. Change area detection system. delete According to clause 6,
The wetland change area detection device is a wetland component change area detection system using time-series satellite images and GIS data that selects an area where wetland components have changed when wetland components are different in the same area in the pixel unit.