KR102442723B1 - A measuring device, method and program that provides measured and estimated values of water level, volume, inflow, and pollution through satellite images using artificial intelligence models - Google Patents

Abstract

The present invention relates to a measurement device for providing measured and estimated values of water level, volume, inflow amount, and pollution level through a satellite image using an artificial intelligence model, which can reduce errors that may occur due to regional variation, and to a method and a program thereof. The measurement device of the present invention comprises: at least one processor; and a memory for storing instructions instructing the processor to perform at least one step.

Description

A MEASURING DEVICE, METHOD AND PROGRAM THAT PROVIDES MEASURED AND ESTIMATED VALUES OF WATER LEVEL, VOLUME, INFLOW providing measurements and estimates of water level, volume, inflow, and pollution levels through satellite imagery using artificial intelligence models. , AND POLLUTION THROUGH SATELLITE IMAGES USING ARTIFICIAL INTELLIGENCE MODELS}

The present invention relates to a measuring device, method and program for providing measured and estimated values of water level, volume, inflow, and pollution through satellite images using an artificial intelligence model.

Unless otherwise indicated herein, the material described in this section is not prior art to the claims of this application, and inclusion in this section is not an admission that it is prior art.

As global warming accelerates, damage due to rapid climate change is rapidly increasing, and accordingly, the importance of water resource management is increasing. In addition, as environmental issues become increasingly important, interest in water quality is also increasing.

Accordingly, the development of a technology for managing water resources by measuring water level, volume, inflow, pollution, etc. has been smoothly performed.

Mainly, technologies using measurement values measured through observatories are being developed, but there is a problem in that it is difficult to accurately provide measurements for areas separated from the observatories.

Korean Patent Registration No. 10-1751020 (registered on June 20, 2017)

One object of the present invention is to cluster the training datasets collected by region into a plurality of clusters, and to generate a measurement model and an estimation model corresponding to each of the plurality of clusters, the water level through a satellite image using an artificial intelligence model. , to provide measurement devices, methods, and programs that provide measurements and estimates of volume, inflow, and pollution levels.

Another object of the present invention is to calculate water level, volume, inflow, and pollution based on feature vectors derived from satellite images, and to provide measured and estimated values of water level, volume, inflow and pollution through a satellite image using an artificial intelligence model. To provide a measuring device, method, and program for

Another object of the present invention is to classify a satellite image into an observable area and an unobservable area, calculate the water level, volume, inflow, and pollution level corresponding to the observable area through the satellite image of the observable area, and to calculate the unobservable area. It provides measured and estimated values of water level, volume, inflow, and pollution through satellite images using artificial intelligence models that calculate water level, volume, inflow and pollution levels corresponding to unobservable areas through satellite images and corresponding geographic information of To provide a measuring device, method and program.

Another object of the present invention is a measuring device that provides measurement values and estimates of water level, volume, inflow and pollution level through a satellite image using an artificial intelligence model, which can detect excessive discharge of pollutants generated in a set monitoring area, To provide methods and programs.

According to an embodiment of the present invention, there is provided an apparatus for measuring water level, volume, inflow, and pollution level through a satellite image of the earth's surface.

The measuring device, at least one processor (processor); and a memory for storing instructions instructing the at least one processor to perform at least one step.

In the at least one step, the satellite image is divided into a plurality of preset regions, and the plurality of regions are classified into an observable region and an unobservable region based on a ratio of a cloud area to a total area of the region. to do; A plurality of pixels included in the observable area are classified into a water area and a land area, and a water area image composed of pixels classified as a water area among the plurality of pixels is generated. Based on the water area image, the water level, volume, calculating at least one of an inflow amount and a pollution level; and obtaining a predicted value of at least one of a water level, a volumetric amount, an inflow amount, and a pollution degree through a regression analysis of the unobservable region and the corresponding geographic information.

The calculating of at least one of water level, volume, inflow, and pollution based on the water domain image may include: extracting a plurality of color histograms corresponding to each of a plurality of different colors from the water domain image; and determining a plurality of first feature vectors from the plurality of color histograms, and calculating a water level corresponding to the observable region based on the plurality of first feature vectors, wherein the observable region and the The corresponding water level may be calculated through a random forest regressor (RFR) algorithm.

The calculating of at least one of water level, volume, inflow, and pollution based on the water domain image may include: extracting a plurality of color histograms corresponding to each of a plurality of different colors from the water domain image; and determining a plurality of second feature vectors from the plurality of color histograms, and calculating a volume corresponding to the observable region based on the plurality of second feature vectors, wherein the observable region and The corresponding volumetric amount may be calculated through a random forest regressor (RFR) algorithm.

In addition, the step of calculating at least one of the water level, the volumetric amount, the inflow amount and the pollution level based on the water area image may include a first water level that is a water level corresponding to the observable area of the first satellite image, which is a satellite image taken at a first time. calculating a first volumetric amount corresponding to the observable area of the first satellite image; A second water level that is a water level corresponding to an observable area of a second satellite image, which is a satellite image captured at a second time different from the first time, is calculated, and a second water level that is a volume corresponding to the observable area of the second satellite image is calculated. 2 calculating the volume; and calculating an inflow corresponding to a time between the first time and the second time through multiple regression analysis on the first water level, the first volumetric amount, the second water level, and the second volumetric amount. can

In addition, the step of obtaining a predicted value of at least one of water level, volume, inflow, and pollution through regression analysis of the unobservable area and the geographic information corresponding to the unobservable area comprises: calculating a first predicted value for the water level corresponding to the unobservable region through regression analysis on one independent variable; and calculating a first residual corresponding to the first predicted value through Kriging analysis of the water level corresponding to the observable region and the first predicted value, and based on the first predicted value and the first residual It may include calculating a water level corresponding to the unobservable region.

In addition, the step of obtaining a predicted value of at least one of water level, volume, inflow, and pollution through regression analysis of the unobservable area and the geographic information corresponding to the unobservable area comprises: calculating a second predicted value for a volume corresponding to the unobservable region through regression analysis on two independent variables; calculating a second residual corresponding to the second predicted value through Kriging analysis of the volume corresponding to the observable region and the second predicted value; and calculating a volume amount corresponding to the unobservable region based on the second predicted value and the second residual.

According to an embodiment of the present invention, the water level, volume, inflow, and pollution level are calculated through a measurement model and an estimation model matching an area. Through this, errors that may be generated due to regional variations may be reduced.

According to another embodiment of the present invention, since the water level, volume, inflow, and pollution are measured through a satellite image, the water level, volume, inflow, and pollution for an area without an observatory may be measured and provided to the user.

According to another embodiment of the present invention, since the water level, volume, inflow, and pollution are estimated through the estimation model, the water level, volume, inflow, and pollution for the cloud-formed area can be estimated and provided to the user.

According to another embodiment of the present invention, it is possible to continuously monitor the large monitoring area for the degree of pollution.

1 is a schematic diagram of a water level, volume, inflow, and pollution level measuring system according to an embodiment.
FIG. 2 is a block diagram exemplarily showing a functional module of the apparatus for measuring water level, volume, inflow, and pollution according to FIG. 1 .
FIG. 3 is a flowchart illustrating a process in which the model generator according to FIG. 2 generates a measurement model.
4 is a flowchart illustrating a process in which the model generator according to FIG. 2 generates a measurement model.
FIG. 5 is a diagram conceptually illustrating a process in which the model generator according to FIG. 2 classifies a water domain image from a satellite image.
6 is a diagram conceptually illustrating a process in which the model generator according to FIG. 2 extracts a feature vector from a water domain image.
FIG. 7 is a diagram conceptually illustrating a process in which the model generator according to FIG. 2 generates a measurement model based on a feature vector, a weather vector, and a geographic vector.
8 is a flowchart illustrating a process in which the model generator according to FIG. 2 generates an estimated model.
FIG. 9 is a diagram conceptually illustrating a process in which the region classifier according to FIG. 2 classifies an observable region and an unobservable region.
10 is a flowchart illustrating a process in which the measuring unit according to FIG. 2 measures at least one of a water level, a volumetric amount, an inflow amount, and a pollution degree.
11 is a diagram conceptually illustrating a process of measuring an inflow amount by the measuring unit according to FIG. 2 .
12 is a flowchart illustrating a process in which the estimator according to FIG. 2 estimates a water level.
13 is a flowchart illustrating a process of estimating the volume by the estimator according to FIG. 2 .
14 is a flowchart illustrating a process of estimating an inflow amount by the estimator according to FIG. 2 .
15 is a flowchart illustrating a process in which the estimator according to FIG. 2 estimates a pollution level.
FIG. 16 is a flowchart illustrating a process in which the abnormal pattern determination unit according to FIG. 2 determines whether or not an abnormal pattern of contamination level occurs.
FIG. 17 is a diagram exemplarily showing a hardware configuration of the device according to FIG. 1 .

Since the present invention can have various changes and can have various embodiments, specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and should be understood to include all modifications, equivalents and substitutes included in the spirit and scope of the present invention. In describing each figure, like reference numerals have been used for like elements.

Terms such as first, second, A, and B may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, a second component may also be referred to as a first component. and/or includes a combination of a plurality of related listed items or any of a plurality of related listed items.

When a component is referred to as being “connected” or “connected” to another component, it may be directly connected or connected to the other component, but it is understood that other components may exist in between. it should be On the other hand, when it is said that a certain element is "directly connected" or "directly connected" to another element, it should be understood that the other element does not exist in the middle.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present application, terms such as “comprise” or “have” are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features It is to be understood that this does not preclude the possibility of the presence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in a commonly used dictionary should be interpreted as having a meaning consistent with the meaning in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present application. does not

'Pollution degree' used below is a concentration for dissolved oxygen (DO), biochemical oxygen demand (BOD), and suspended matter (SS), and the unit of pollution degree may be ppm. In addition, the 'pollution degree' is a concentration with respect to hydrogen ions, and the unit of the contamination degree may be pH.

As used below, 'water level' refers to the depth of a stream. 'Capacity' as used below means the volume of rivers.

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

1 is a schematic diagram of a system for measuring water level, volume, inflow, and pollution according to an embodiment.

Referring to FIG. 1 , the water level, volume, inflow and pollution measurement system is a device 100 , a user terminal 200 , a satellite server 300 , an observatory 400 , a weather server 500 and a Geographic Information System (GIS). ) may include a server 600 .

The user terminal 200 is a terminal of a user who wants to use the service for providing water level, volume, inflow and pollution, and registers user information in the device 100 and provides water level, volume, inflow and pollution through the device 100 . Several features of the service are available.

The satellite server 300 stores a satellite image photographed with respect to the earth's surface by an artificial satellite operating spaced apart from the earth's surface. In an embodiment, the artificial satellite may be an artificial satellite that is spaced apart from the earth's surface at a predetermined interval and operates. In one embodiment, the satellite server 300 may receive a satellite image from the satellite at preset time intervals. In an embodiment, the satellite image may be an image of a preset resolution or a preset size.

The satellite server 300 transmits the satellite image received from the satellite to the device 100 .

The observatory 400 measures the water level, volume amount, inflow amount, and pollution degree of the installed location. A plurality of observatories 400 may be installed in each of a plurality of regions. In an embodiment, different numbers of observation stations 400 may be installed in each region. The observatory 400 may be installed in a place adjacent to a river and configured to measure at least one of water level, volume, inflow, and pollution.

The plurality of observatories 400 transmits at least one of the measured water level, volume, inflow, and pollution to the device 100 . In one embodiment, if there is a separate observatory server (not shown) that receives all of at least one of water level, volume, inflow, and pollution level measured at a plurality of observatories 400, the device 100 may include the observatory server ( At least one of a water level, a volumetric amount, an inflow amount, and a pollution degree may be received from).

The weather server 500 is a server that collects weather information, and is connected to sensors for measuring weather information so as to enable information communication. In one embodiment, the meteorological information includes rainfall, cloud formation area, ultra-precipitable amount, atmospheric instability, atmospheric motion vector, aerosol optical thickness, yellow sand optical thickness, aerosol particle size, temperature and humidity profile, cloud volume, cloud top, cloud top temperature, It may include barometric pressure, altitude, sea surface temperature, aerosol visibility, cloud optical thickness, effective radius of cloud particles, short-wave radiation, long-wave radiation, and the like.

The weather server 500 transmits weather information to the device 100 .

The GIS server 600 stores geographic information on a preset area. The preset region may be divided into a plurality of detailed regions (Grids) having a preset size, and geographic information may be stored for each of the plurality of detailed regions. In an embodiment, the geographic information may include, with respect to the detailed area, the width for each area of the river, the depth for each area of the river, information on structures installed in the river, information on discharge facilities adjacent to the river, and the like. The geographic information is not limited to the above-described examples, and various types of information related to water level, volume, inflow, and pollution may be included in the geographic information.

The GIS server 500 transmits the geographic information to the device 100 .

For example, the user terminal 200, the satellite server 300, the observatory 400, the weather server 500 and the GIS server 600, which can communicate with a desktop computer (desktop computer), a laptop computer (laptop computer), a notebook (notebook), smart phone, tablet PC, mobile phone, smart watch, smart glass, e-book reader, PMP (portable multimedia player) , portable game machine, navigation device, digital camera, DMB (digital multimedia broadcasting) player, digital audio recorder, digital audio player, digital video recorder ), a digital video player, a personal digital assistant (PDA), or the like.

The device 100 calculates at least one of water level, volume, inflow and pollution based on the information received from the satellite server 300, the observatory 400, the weather server 500, and the GIS server 600, and the calculated It may be at least one server capable of providing at least one of water level, volume, inflow, and pollution to the user terminal 200 .

The device 100 may provide at least one of the water level, volume, inflow, and pollution to the user terminal 200 by operating a water level, volume, inflow, and pollution level providing service.

The device 100 classifies a satellite image for the earth's surface corresponding to the observatory 400 among a plurality of satellite images received from the satellite server 300 as a training satellite image, and corresponding to the training satellite image and the training satellite image. A model for measuring any one of the water level, the volumetric amount, the inflow amount, and the pollution degree may be generated based on any one of the water level, the volumetric amount, the inflow amount, and the pollution degree.

In addition, the device 100 classifies a satellite image for the ground surface corresponding to the observatory 400 among a plurality of satellite images received from the satellite server 300 as a satellite image for learning, a satellite image for learning, a satellite image for learning, and Based on any one of the corresponding water level, volume, inflow, and pollution, and at least one of geographic information and weather information corresponding to the satellite image for training, a model for measuring any one of water level, volume, inflow, and pollution can be generated. .

The device 100 classifies the geographic information corresponding to the location of the observatory 400 among the plurality of geographic information received from the GIS server 600 as geographic information for learning, and the level corresponding to the geographic information for learning and the geographic information for learning , a model for estimating any one of a water level, a volumetric amount, and an inflow amount may be generated based on any one of the volumetric amount and the inflow amount.

The apparatus 100 may divide each of the plurality of satellite images received from the satellite server 300 into a plurality of regions, and classify the plurality of regions into an observable region and an unobservable region. The apparatus 100 may classify an area in which the ratio of the cloud area to the total area is equal to or greater than a preset standard among the plurality of areas as the unobservable area.

The device 100 receives a satellite image from the satellite server 300 and receives any one of water level, volume, inflow, and pollution from the observable area of the satellite image through a model that measures any one of water level, volume, inflow, and pollution. You can get one.

The device 100 receives the satellite image from the satellite server 300, and through an estimation model for any one of the water level, the volume and the inflow, the predicted value of any one of the water level, the volume, and the inflow from the unobservable area of the satellite image can be obtained In addition, the apparatus 100 may calculate a residual for the predicted value through a Kriging analysis on any one of the predicted value of the water level and the volume of the unobservable area and the water level and the volume of the observable area. have. The apparatus 100 may correct any one predicted value of a water level, a volume amount, and an inflow amount for the unobservable region based on the predicted value and the residual.

The apparatus 100 may receive a satellite image from the satellite server 300 and obtain a predicted value of the pollution degree from an unobservable area of the satellite image through Kriging analysis on the pollution degree.

The apparatus 100 calculates a distribution map of any one of water level, volume, inflow, and pollution for the entire area in which the satellite image is received, based on any one of water level, volume, inflow, and pollution in the observable area and the unobservable area. can create

The apparatus 100 may set a monitoring area among all areas collected through the satellite image and generate a time-series contamination level pattern based on the contamination level of the monitoring area collected for a preset period.

The apparatus 100 may determine that an abnormal pattern has occurred when the pollution level corresponding to the monitoring area deviates from the time-series pollution level pattern.

FIG. 2 is a block diagram exemplarily showing a functional module of the device according to FIG. 1 .

Referring to FIG. 2 , the apparatus 100 includes a model generation unit 101 for generating a measurement model and an estimation model, a region classification unit 102 for classifying a received satellite image into an observable area and an unobservable area, and measurement A measurement unit 103 that measures any one of the water level, volume, inflow, and pollution level of the observable area through the model, and a weight for estimating the predicted value of any one of the water level, volume, inflow and pollution level of the unobservable area through the estimation model Government 104, a distribution map generating unit 105 that generates a distribution map of any one of water level, volume, inflow, and pollution based on any one of water level, volume, inflow, and pollution in the observable area and the unobservable area, and the monitoring area It may include an abnormal pattern determination unit 106 for detecting the occurrence of an abnormal pattern in the .

In an embodiment, the model generator 101 may generate a plurality of clusters through clustering of a plurality of training satellite images, and may generate a measurement model corresponding to each of the plurality of clusters. As a clustering method, K-Means Clustering, Mean-Shift Clustering, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), Expectation-Maximization (EM) Clustering using Gaussian Mixture Models (GMM), Agglomerative Hierarchical Clustering, etc. may be used. .

In an embodiment, the model generator 101 may generate a plurality of clusters through clustering of a plurality of geographic information for learning, and may generate an estimation model corresponding to each of the plurality of clusters.

In an embodiment, clustering may be performed on datasets grouped by region. For example, by grouping a plurality of satellite images for training or a plurality of geographic information for training by region, regional datasets may be generated, and clustering may be performed on regional datasets.

In an embodiment, clustering may be performed on datasets grouped by season. For example, by grouping a plurality of training satellite images or a plurality of training geographic information by season, seasonal datasets may be generated, and clustering may be performed on the seasonal datasets.

In an embodiment, clustering may be performed on datasets grouped by season and region. For example, by grouping a plurality of satellite images for training or a plurality of geographic information for training by season and region, and generating data sets grouped by season and region, clustering is performed on the datasets grouped by season and region can be

Satellite images and geographic information on the earth's surface at a location corresponding to the observatory 400 are used as learning data.

The model generating unit 101 determines a satellite image for the ground surface at a location corresponding to the plurality of observation stations 400 as a training satellite image, and any one of the water level, volume and pollution level corresponding to the determined satellite image for training and the training image. Create training data by labeling the measured values of Also, the model generator 101 may generate a measurement model for measuring any one of a water level, a volumetric amount, and a pollution degree through machine learning for a plurality of learning data. In one embodiment, the model generation unit 101, the determined satellite image for training, at least one of weather information and geographic information corresponding to the training image, a measurement value of any one of water level, volume and pollution level corresponding to the training image can be labeled to generate training data.

The model generating unit 101 determines a pair of satellite images taken at two different viewpoints with respect to a plurality of observation stations 400 and the ground surface of a corresponding position as a training satellite image pair, and the determined training satellite image pair and training Data for training is generated by labeling the measurement value of the inflow corresponding to the image pair. In addition, the model generator 101 may generate a measurement model for measuring an inflow through machine learning for a plurality of learning data. In one embodiment, the model generating unit 101, the determined training satellite image pair, at least one of weather information and geographic information corresponding to the training satellite image pair, the measurement value for the inflow corresponding to the training satellite image pair You can create data for training by labeling.

In addition, the model generator 101 determines at least one of geographic information and weather information of a location corresponding to the plurality of received observation stations 400 as learning information, and water level and volume corresponding to the determined learning information and learning information. By labeling any one of the measured values, data for training is generated. Also, the model generator 101 may generate an estimation model for any one of a water level and a volumetric amount through machine learning on a plurality of learning data.

In addition, the model generation unit 101 is configured to: the predicted value of the first water level at the first time determined from the water level estimation model, the predicted value of the first volumetric quantity at the first time determined from the volumetric estimation model, and the estimation of the water level The predicted value of the second water level at the second time determined from the model and the predicted value of the second volumetric quantity at the second time determined from the estimation model for the volumetric amount are determined as learning information, and the determined learning information and the learning information and the corresponding inflow amount are measured Create training data by labeling values. Also, the model generation unit 101 may generate an estimation model for the inflow through machine learning for a plurality of training data. The first time and the second time mean two different times.

Also, the model generator 101 may generate an estimated model of the pollution level through a Kriging analysis on the pollution level values of the received plurality of observation stations 400 and corresponding locations.

In an embodiment, the model generating unit 101 may determine, as noise, a training satellite image in which a ratio of a cloud area to a total area area among the training satellite images is equal to or greater than a preset standard as noise.

The model generator 101 groups a plurality of training data by region and season to generate regional and seasonal datasets, and generates a plurality of clusters through clustering of regional and seasonal datasets, and a plurality of clusters A measurement model and an estimation model corresponding to each can be generated.

Since a plurality of measurement models and estimation models corresponding to each of the clusters are generated, regional characteristics and seasonal characteristics may be well reflected in the measurement models and estimation models. Through this, it is possible to reduce errors that may occur due to regional and seasonal variations when calculating water level, volume, inflow, and pollution.

The number of the plurality of observatories 400 installed to measure any one of water level, volume, inflow, and pollution is different for each region. Since the satellite image, weather information, and geographic information on the earth's surface at a location corresponding to the observatory 400 are used as learning data, the data amount of the learning data varies by region.

The model generator 101 classifies a plurality of training satellite images to correspond to each of a plurality of regions, generates a plurality of datasets corresponding to each of the plurality of regions, and the minimum data among the plurality of datasets. Based on the data amount of the dataset having the amount, under sampling may be performed on the plurality of datasets. The model generator 101 may generate a plurality of clusters by clustering a plurality of under-sampled datasets. The model generator 101 may generate a measurement model and an estimation model corresponding to each of the plurality of clusters. Since the amount of data of the datasets is uniformed through undersampling, the characteristics of the dataset in an area with a small amount of data can be reflected in the measurement model and the estimation model.

In an embodiment, the clustering of the training satellite images may be performed based on a plurality of feature vectors determined from the training satellite image. The plurality of feature vectors may be a plurality of feature vectors used for learning the measurement model.

3 is a flowchart illustrating a process in which the model generating unit 101 according to FIG. 2 generates a measurement model.

First, the model generator 101 receives a plurality of training satellite images including a ground surface image of a location corresponding to each of the plurality of observation stations 400 (S110).

Also, the model generator 101 classifies the plurality of training images to correspond to each of the plurality of regions, and generates a plurality of datasets corresponding to each of the plurality of regions ( S120 ). In an embodiment, the dataset may include weather information and geographic information corresponding to the training image in addition to the training image.

In addition, the model generator 101 removes the training satellite image determined as noise from among the plurality of training satellite images included in each of the plurality of data sets ( S130 ).

In an embodiment, the model generating unit 101 may determine, as noise, a training satellite image in which a ratio of a cloud area to a total area among a plurality of training satellite images is equal to or greater than a preset standard as noise.

Also, the model generator 101 performs under-sampling on the plurality of data sets based on the data amount of the data set having the minimum data amount among the plurality of data sets from which the noise has been removed ( S140 ).

In addition, the model generator 101 generates a plurality of clusters by clustering a plurality of under-sampled data sets ( S150 ).

Also, the model generating unit 101 generates a measurement model corresponding to each of the plurality of clusters ( S160 ).

4 is a flowchart illustrating a process in which the model generating unit 101 according to FIG. 2 generates a measurement model.

First, the model generating unit 101 extracts a water domain image from the training satellite image (S161).

FIG. 5 is a diagram conceptually illustrating a process in which the model generating unit 101 according to FIG. 2 classifies a water domain image from a satellite image. Referring to FIG. 5 , the model generator 101 may classify pixels of a satellite image for training into a water region (flooded) and other regions (unflooded) through semantic segmentation. The model generator 101 may classify pixels classified as flooded as water domain images. In an embodiment, a UNet neural network or an FCN neural network may be used as a method of semantic segmentation.

Referring back to FIG. 4 , the model generator 101 extracts a plurality of color histograms for a plurality of different colors from the water domain image ( S162 ). Also, the model generator 101 may determine a plurality of feature vectors from the plurality of color histograms ( S163 ).

6 is a diagram conceptually illustrating a process in which the model generator 101 according to FIG. 2 extracts a feature vector from a water domain image. In the illustrated embodiment, the model generator 101 may extract a color histogram for blue, a color histogram for red, and a color histogram for yellow from pixels constituting the water domain image. The model generator 101 determines a plurality of feature vectors from at least one of a color histogram for blue, a color histogram for red, and a color histogram for yellow.

The model generator 101 may determine a plurality of first feature vectors from a plurality of color histograms in order to generate a measurement model for the water level. Also, the model generator 101 may determine a plurality of second feature vectors from a plurality of color histograms in order to generate a measurement model for the volumetric quantity. Also, the model generator 101 may determine a plurality of third feature vectors from a plurality of color histograms in order to generate a measurement model for the pollution degree.

Referring back to FIG. 4 , the model generating unit 101 generates training data by matching information measured at the observation station 400 corresponding to the plurality of feature vectors and the training satellite image ( S164 ).

The model generator 101 may generate the first training data by matching the plurality of first feature vectors with the satellite image for training and the water level measured at the corresponding observatory 400 . Also, the model generator 101 may generate second training data by matching the plurality of second feature vectors and the volume measured at the observation station 400 corresponding to the satellite image for training. In addition, the model generator 101 may generate third training data by matching the plurality of third feature vectors and the contamination level measured at the observation station 400 corresponding to the satellite image for training.

The model generator 101 uses first and second feature vectors determined from the satellite image for training at a first time, and first and second feature vectors determined from the satellite image for training at a second time, The fourth learning data may be generated by matching the inflow amount measured at the observation station 400 corresponding to the satellite image for learning between the first time and the second time.

In one embodiment, the model generating unit 101 receives the plurality of feature vectors, at least one of weather information and geographic information corresponding to the satellite image for training, and information measured at the observation station 400 corresponding to the satellite image for training. Learning data can be generated by matching.

The model generator 101 measures at least one of a plurality of first feature vectors, a first weather vector determined from weather information, and a first geographic vector determined from geographic information, at an observation station 400 corresponding to a satellite image for training. It is possible to generate the first learning data by matching the water level. A first meteorological vector may be generated based on information related to a water level among information included in the meteorological information. A first geographic vector may be generated based on information related to a water level among information included in the geographic information. A plurality of first weather vectors and first geographic vectors may be used to generate the first learning data.

The model generator 101 measures at least one of a plurality of second feature vectors, a second weather vector determined from weather information, and a second geographic vector determined from geographic information, at an observation station 400 corresponding to a satellite image for training. It is possible to generate the second learning data by matching the capacity amount. A second meteorological vector may be generated based on information related to a volumetric amount among information included in the meteorological information. A second geographic vector may be generated based on information related to a volumetric amount among information included in the geographic information. A plurality of second weather vectors and second geographic vectors may be used to generate the second learning data.

In addition, the model generator 101 generates a measurement model for any one of the water level, the volumetric amount, and the pollution level through a random forest analysis of the training dataset composed of the training data (S165).

FIG. 7 is a diagram conceptually illustrating a process in which the model generating unit 101 according to FIG. 2 generates a measurement model based on a feature vector, a weather vector, and a geographic vector. The model generator 101 may generate a measurement model for any one of a water level, a volumetric amount, and a pollution degree based on the feature vector. In addition, the model generator 101 may generate a measurement model for any one of the water level and the volume based on at least one of a characteristic vector, a weather vector, and a geographic vector.

The model generator 101 may generate a water level measurement model through a random forest analysis on the first training dataset composed of the first training data. In an embodiment, the model generator 101 trains the water level measurement model so that the water level is calculated from the plurality of first feature vectors through a random forest regressor (RFR) algorithm. In an embodiment, the model generator 101 measures the water level so that the water level is calculated from at least one of a plurality of connected first feature vectors and a first weather vector and a first geographic vector through a random forest regressor (RFR) algorithm. train the model.

The model generator 101 may generate a volumetric measurement model through a random forest analysis on the second training dataset composed of the second training data. In an embodiment, the model generator 101 trains the volumetric measurement model so that the volumetric amount is calculated from the plurality of second feature vectors through a random forest regressor (RFR) algorithm. In an embodiment, the model generator 101 measures the volumetric amount so that the volumetric amount is calculated from at least one of a plurality of connected second feature vectors and a second weather vector and a second geographic vector through a random forest regressor (RFR) algorithm. train the model.

The model generator 101 may generate a pollution degree measurement model through a random forest analysis on the third training dataset composed of the third training data. In an embodiment, the model generator 101 trains the pollution degree measurement model to calculate the pollution degree from the plurality of third feature vectors through a random forest regressor (RFR) algorithm.

RFR (Random Forest Regressor) is a kind of ensemble learning method used in regression analysis. In addition, the model generator 101 may evaluate the learned measurement model through a test satellite image, and may adjust a hyperparameter of a random forest regressor (RFR) based on the evaluation result.

FIG. 8 is a flowchart illustrating a process in which the model generating unit 101 according to FIG. 2 generates an estimated model.

The model generator 101 may generate an estimated model corresponding to each of the generated clusters.

First, the model generator 101 determines geographic information for a location corresponding to a satellite image for training included in the cluster as geographic information for training (S210).

In addition, the model generator 101 selects at least one independent variable from the geographic information for learning, and selects a measurement value measured at an observation station corresponding to the satellite image for training as a dependent variable (S220).

Also, the model generator 101 generates a training dataset based on at least one independent variable and a dependent variable (S230). A training dataset is generated based on each of the plurality of training images included in the cluster and the plurality of training geographic information corresponding to the plurality of training images.

Also, the model generator 101 generates an estimated model through regression analysis on the generated training dataset (S240).

In an embodiment, multiple regression analysis may be used for regression analysis.

The generated estimation model may be defined by Equation 1 below.

In Equation 1, s ₀ means a target position at which the water level or volume is to be estimated, Z _R (s ₀ ) means the water level or volume estimated from s0, and q _k (s ₀ ) is independent at s ₀ . variables, β _k is the regression coefficient for the independent variables, and p-1 is the number of independent variables. q ₀ (s ₀ ) is 1.

In an embodiment, an estimation model for the water level may be generated through multiple regression analysis. Various pieces of information included in geographic information for learning can be set as independent variables, and independent variables having a relatively large influence on the water level, which is a dependent variable, can be determined through regression analysis on the learning dataset.

In an embodiment, an estimation model for the volume may be generated through multiple regression analysis. Various pieces of information included in geographic information for learning can be set as independent variables, and independent variables that have a relatively large influence on volume, a dependent variable, can be determined through regression analysis on the learning dataset.

9 is a diagram conceptually illustrating a process in which the region classifying unit 102 according to FIG. 2 classifies an observable region and an unobservable region.

When the satellite image is received from the satellite server 300, the region classification unit 102 divides the received satellite image into a plurality of regions, and classifies the divided regions into an observable region and an unobservable region. The region classifying unit 102 may classify an area in which a ratio of a cloud area to a total area among a plurality of areas is equal to or greater than a preset standard as an unobservable area.

In an embodiment, the area classification unit 102 may classify the plurality of areas into an observable area and an unobservable area based on the weather information received from the weather server 500 . The meteorological information may include information about a region created in the cloud.

In an embodiment, the region classification unit 102 may classify a plurality of regions into an observable region and an unobservable region through a previously learned region classification model to classify the cloud region. The image classification model can be generated through supervised learning using a CNN (Convolution Neural Network) algorithm.

The region classification unit 102 transmits the classified observable region to the measurement unit 103 .

The region classification unit 102 transmits the classified unobservable region to the estimation unit 104 .

10 is a flowchart illustrating a process in which the measuring unit 103 according to FIG. 2 measures at least one of a water level, a volume amount, an inflow amount, and a pollution level.

*First, the measurement unit 103 extracts a plurality of feature vectors from the observable region (S310).

In an embodiment, when measuring the water level, the measuring unit 103 extracts a plurality of first feature vectors from the observable area. When measuring the volume, the measuring unit 103 extracts a plurality of second feature vectors from the observable area. When measuring the pollution level, the measuring unit 103 extracts a plurality of third feature vectors from the observable area. When measuring the inflow, the measurement unit 103 extracts a plurality of first feature vectors and a plurality of second feature vectors from the observable region.

In an embodiment, when measuring the water level, the measurement unit 102 extracts a plurality of first feature vectors from the observable area, determines the first weather vector from weather information corresponding to the observable area, and is observable. The first geographic vector may be determined from geographic information corresponding to the region. When measuring the volume, the measurement unit 102 extracts a plurality of second feature vectors from the observable area, determines a second weather vector from weather information corresponding to the observable area, and includes geographic information corresponding to the observable area. A second geographic vector may be determined from When measuring the inflow, the measuring unit 103 determines a plurality of first feature vectors and a plurality of second feature vectors in the observable area, and the first and second weather vectors from the weather information corresponding to the observable area. A weather vector may be determined, and a first geographic vector and a second geographic vector may be determined from geographic information corresponding to the observable area.

Also, the measurement unit 103 measures at least one of a water level, a volume amount, an inflow amount, and a pollution degree corresponding to the observable area based on the plurality of feature vectors ( S320 ).

In an embodiment, when measuring the water level, the measurement unit 103 inputs the first feature vector to the water level measurement model, and obtains the water level corresponding to the observable area as an output value. Specifically, the water level is calculated through the RFR (Random Forest Regressor) algorithm of the water level measurement model. In another embodiment, when measuring the water level, the measurement unit 103 may input at least one of a plurality of first feature vectors, a first weather vector, and a first geographic vector to the water level measurement model, and observe as an output value Obtain the water level corresponding to the area. A plurality of first weather vectors or first geographic vectors may be input.

In an embodiment, when measuring the volumetric amount, the measurement unit 103 inputs the second feature vector to the volumetric measurement model, and obtains the volumetric amount corresponding to the observable area as an output value. Specifically, the volume is calculated through the RFR (Random Forest Regressor) algorithm of the water level measurement model. In another embodiment, when measuring the volumetric amount, the measurement unit 103 may input at least one of a plurality of second feature vectors, a second weather vector, and a second geographic vector to the water level measurement model, and observe as an output value A volume corresponding to an area is obtained. A plurality of second weather vectors or second geographic vectors may be input.

In an embodiment, when measuring the pollution level, the measurement unit 103 inputs the third feature vector to the pollution level measurement model, and obtains the pollution level corresponding to the observable area as an output value. Specifically, the pollution level is calculated through the RFR (Random Forest Regressor) algorithm of the water level measurement model.

11 is a diagram conceptually illustrating a process in which the measurement unit 103 according to FIG. 2 measures the inflow amount.

Referring to FIG. 11 , the measurement unit 103 obtains a first water level and a first volume corresponding to the observable area at a first time through the water level measurement model and the volumetric measurement model, and is observable at a second time A second water level and a second volume corresponding to the area are obtained. The first time and the second time are random times different from each other. The measurement unit 103 may calculate the inflow amount between the first time and the second time based on the difference between the first water level and the second water level and the difference between the first volumetric amount and the second volumetric amount. In an embodiment, the inflow may be calculated by multiple regression analysis using the first water level, the first volumetric amount, the second water level, and the second volumetric amount as independent variables and the inflow amount as the dependent variable. An inflow measurement model may be generated through multiple regression analysis on the fourth training dataset composed of the fourth training data.

12 is a flowchart illustrating a process in which the estimator 104 according to FIG. 2 estimates a water level.

First, when the unobservable area is received, the estimator 104 obtains a predicted value for the water level in the unobservable area from geographic information corresponding to the unobservable area through the water level estimation model ( S410 ).

The estimator 104 extracts independent variables having a relatively strong correlation with the water level, determined in the learning process of the water level estimation model, from geographic information corresponding to the unobservable region. Also, the estimator 104 inputs the extracted independent variables to the water level estimation model, and obtains a predicted value for the level of the unobservable region output from the water level estimation model.

Also, the estimator 104 calculates a residual corresponding to the predicted value through kriging analysis ( S420 ).

When the residual is calculated, the estimator 104 calculates the level of the unobservable region based on the predicted value and the residual ( S430 ).

13 is a flowchart illustrating a process in which the estimator 104 according to FIG. 2 estimates a volumetric amount.

First, when the unobservable area is received, the estimator 104 obtains a predicted value for the volumetric amount of the unobservable area from geographic information corresponding to the unobservable area through the volume estimation model ( S510 ).

The estimator 104 extracts independent variables having a relatively strong correlation with the volumetric amount determined in the learning process of the volumetric volume estimation model from the unobservable region and corresponding geographic information. In addition, the estimator 104 inputs the extracted independent variables to the volumetric amount estimation model, and obtains a predicted value for the volumetric amount of the unobservable region output from the volumetric amount estimation model.

Also, the estimator 104 calculates a residual corresponding to the predicted value through kriging analysis ( S520 ).

When the residual is calculated, the estimator 104 calculates the volume of the unobservable region based on the predicted value and the residual ( S530 ).

14 is a flowchart illustrating a process of estimating the inflow amount by the estimator 104 according to FIG. 2 .

First, when the unobservable area is received, the estimator 104 performs a first water level predicted value and a first value of the unobservable area from geographic information corresponding to the unobservable area at a first time through the water level estimation model and the volumetric volume estimation model. A capacity prediction value is obtained (S610).

Also, the estimator 104 calculates a residual corresponding to each of the first predicted water level and the first predicted volume through kriging analysis ( S620 ).

In addition, the estimator 104 is configured to: the first water level and the first water level of the unobservable region at the first time based on the first water level predicted value, the first volumetric amount predicted value, and the residuals corresponding to each of the first water level predicted value and the first volumetric amount predicted value 1 Calculate the volume (S630).

In addition, the estimator 104 obtains a second predicted water level and a second predicted volume of the unobservable region from geographic information corresponding to the unobservable region at the second time through the water level estimation model and the volumetric amount estimation model (S640) .

Also, the estimator 104 calculates a residual corresponding to each of the second predicted water level and the second predicted volume through kriging analysis ( S650 ).

In addition, the estimator 104 is configured to generate the second water level and the second water level of the unobservable region at the second time based on the second water level predicted value, the second predicted volume, and the residual corresponding to each of the second predicted value and the second predicted value. 2 Calculate the volume (S660).

In addition, the estimator 104 calculates the inflow amount between the first time and the second time corresponding to the unobservable region based on the first water level, the first volumetric amount, the second water level, and the second volumetric amount ( S670 ). In an embodiment, the inflow may be calculated by multiple regression analysis using the first water level, the first volumetric amount, the second water level, and the second volumetric amount as independent variables and the inflow amount as the dependent variable.

The residual described in FIGS. 12, 13 and 14 may be defined by Equation 2 below.

In Equation 2, s ₀ denotes a target position for which a value is to be estimated, e(s ₀ ) denotes a residual at s ₀ , and w _i (s ₀ ) denotes a kriging corresponding to each of a plurality of sample positions. It means a weight, and e( s _i ) means a regression residual corresponding to each of a plurality of sample positions. n means the number of sample positions.

In an embodiment, a plurality of adjacent areas adjacent to the unobservable area may be used as a plurality of sample locations.

A kriging weight corresponding to each of the plurality of sample positions may be determined based on a variogram function. The variogram function is a measure of the similarity of data at a certain distance.

The empirical variogram function may be defined by Equation 3 below.

In Equation 3, γ(h) denotes a variogram function, Z(x) denotes a water level or volume of sample positions, and h denotes a distance between sample positions. n means the number of sample position pairs. In an embodiment, the distance between the sample locations may be set based on a size of a preset observable area.

The theoretical variogram function is modeled from the empirical variogram function, and the Kriging weight can be determined through the covariance derived from the theoretical variogram function. As the theoretical variogram function, a spherical function, an exponential function, a Gaussian function, etc. may be used.

The water level and volume of the unobservable region described in FIGS. 12, 13 and 14 may be defined by Equation 4 below.

In Equation 4, s ₀ means a target position at which the water level or volume is to be estimated, Z _RK (s ₀ ) means the water level or volume estimated at s0, and q _k (s ₀ ) is an independent variable in s ₀ . , β _k means the regression coefficient for the independent variables, and p-1 means the number of independent variables. q ₀ (s ₀ ) is 1. w _i (s ₀ ) denotes a Kriging weight corresponding to each of the plurality of sample positions, and e( s _i ) denotes a regression residual corresponding to each of the plurality of sample positions . n means the number of sample positions.

15 is a flowchart illustrating a process in which the estimator 104 according to FIG. 2 estimates a pollution level.

The estimator 104 calculates a pollution degree corresponding to the unobservable region through kriging analysis (S710).

Pollution degree may be defined by Equation 5 below.

는 s₀에서 추정된 오염도를 의미하며, w_i(s₀)는 복수의 샘플 위치들 각각과 대응하는 크리깅 웨이트(Kriging weight)를 의미하고, Z(s_i)는 복수의 샘플 위치들 각각에서 측정된 오염도를 의미한다. n은 샘플 위치의 개수를 의미한다.In Equation 5, s ₀ means a target position to estimate a value,

denotes the pollution degree estimated at s ₀ , w _i ( s ₀ ) denotes a kriging weight corresponding to each of the plurality of sample positions, and Z( s _i ) denotes the degree of contamination at each of the plurality of sample positions. Measures the degree of contamination. n means the number of sample positions.

In an embodiment, a plurality of observable regions adjacent to an unobservable region may be used as a plurality of sample locations.

A kriging weight corresponding to each of the plurality of sample positions may be determined based on a variogram function. The variogram function is a measure of the similarity of data at a certain distance.

The empirical variogram function may be defined by Equation 3 above.

When the observable area and the unobservable area are classified by the area classification unit 102 , the water level, volume, inflow, and pollution level corresponding to the observable area are calculated by the measurement unit 103 , and the unobservable area is calculated by the estimation unit 104 . The water level, volume, inflow, and pollution level corresponding to and are calculated. Through this, the water level, volume, inflow, and pollution level for the entire area of the satellite image received from the satellite server 300 may be calculated.

The distribution map generating unit 105 generates a distribution map for the water level, volume, inflow, and pollution for the entire area in which the satellite image is received, based on the measured value corresponding to the observable area and the estimated value corresponding to the unobservable area. can

16 is a flowchart illustrating a process in which the abnormal pattern determination unit 106 according to FIG. 2 determines whether or not an abnormal pattern of pollution level occurs.

First, the abnormal pattern determination unit 106 determines a monitoring area (S410).

In addition, the abnormal pattern determining unit 106 generates a time-series contamination level pattern based on the level of pollution measured for a preset period for the monitoring area (S420). For example, based on the pollution level for 20 years, it is possible to generate a pollution degree pattern that changes in time series according to spring, summer, autumn, and winter. In an embodiment, a K-Mean Clustering method may be used to generate a time-series contamination pattern.

Also, the abnormal pattern determination unit 106 may determine that the abnormal pattern has occurred when the contamination level corresponding to the monitoring area is out of the time-series contamination level pattern (S430). For example, when an excessively high pollution degree is calculated compared to the average pollution degree in March of the monitoring area, it can be determined that an abnormal pattern has occurred.

Through this, it is possible to monitor abnormal patterns (eg, wastewater discharge exceeding the standard) occurring in the monitoring area.

FIG. 17 is a diagram exemplarily showing a hardware configuration of the device according to FIG. 1 .

Referring to FIG. 11 , the device 100 includes at least one processor 110 and a memory storing instructions instructing the at least one processor 110 to perform at least one operation. (memory) may be included.

The at least one operation may include the components 101 to 106 of the above-described apparatus 100 or other functions or operation methods.

Here, the at least one processor 110 may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed. can Each of the memory 120 and the storage device 160 may be configured as at least one of a volatile storage medium and a non-volatile storage medium.

For example, the memory 120 may be one of a read only memory (ROM) and a random access memory (RAM), and the storage device 160 is a flash-memory. , a hard disk drive (HDD), a solid state drive (SSD), or various memory cards (eg, micro SD card).

Also, the device 100 may include a transceiver 130 that performs communication through a wireless network. Also, the device 100 may further include an input interface device 140 , an output interface device 150 , a storage device 160 , and the like. Each component included in the device 100 may be connected by a bus 170 to communicate with each other.

For example, the device 100 can communicate with a desktop computer (desktop computer), a laptop computer (laptop computer), a notebook (notebook), a smart phone (smart phone), a tablet PC (tablet PC), a mobile phone (mobile phone) , smart watch, smart glass, e-book reader, PMP (portable multimedia player), portable game console, navigation device, digital camera, DMB (digital multimedia broadcasting) player , a digital audio recorder, a digital audio player, a digital video recorder, a digital video player, a personal digital assistant (PDA), or the like.

The methods according to the present invention may be implemented in the form of program instructions that can be executed by various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the computer readable medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the art of computer software.

Examples of computer-readable media may include hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions may include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as at least one software module to perform the operations of the present invention, and vice versa.

In addition, the above-described method or apparatus may be implemented by combining all or part of its configuration or function, or may be implemented separately.

Although the above has been described with reference to preferred embodiments of the present invention, those skilled in the art can variously modify and change the present invention within the scope without departing from the spirit and scope of the present invention as set forth in the claims below. You will understand that it can be done.

Claims (1)

A device for measuring water level, volume, inflow, and pollution level through satellite imagery on the earth's surface,
at least one processor; and
a memory for storing instructions instructing the at least one processor to perform at least one step;
The at least one step is
dividing the satellite image into a plurality of preset regions, and classifying the plurality of regions into observable regions and unobservable regions based on a ratio of a cloud area to a total area of the region;
A plurality of pixels included in the observable area are classified into a water area and a land area, and a water area image composed of pixels classified as a water area among the plurality of pixels is generated. Based on the water area image, the water level, volume, calculating at least one of an inflow amount and a pollution level; and
acquiring a predicted value of at least one of water level, volume, inflow, and pollution through regression analysis of the unobservable region and corresponding geographic information,
Calculating at least one of a water level, a volume amount, an inflow amount, and a pollution degree based on the water area image,
extracting a plurality of color histograms corresponding to each of a plurality of different colors from the water domain image; and
determining a plurality of first feature vectors from the plurality of color histograms, and calculating a water level corresponding to the observable region based on the plurality of first feature vectors,
The water level corresponding to the observable area is calculated through a random forest regressor (RFR) algorithm,
measuring device.

Images