JP2017138928A - Environment pollution prediction system and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an environment pollution prediction system, an environment pollution prediction method and a computer program.SOLUTION: An environment pollution prediction system comprises: a satellite viewdata collection module for collecting plural pieces of satellite viewdata in which a ground surface of an object area is imaged for every unit time during a prescribed collection period; a satellite viewdata analysis module for analyzing the collected satellite viewdata, creating satellite video analysis data, for creating a satellite video analysis data set; an actual measurement data collection module for collecting plural pieces of actual measurement data which are sensed by a sensor in the object area; an actual data analysis module for analyzing the actual measurement data, creating environment pollution data, for creating an environment pollution data set; a learning module for creating association information between the satellite video analysis data set and the environment pollution data set; and a prediction module for using the association information between the satellite video data set and the environment pollution data set, for creating prediction environment pollution data.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an environmental pollution prediction system and method, and more particularly, to an environmental pollution prediction system and method capable of predicting environmental pollution from satellite images and / or postings posted on social network services.

  The present invention has been derived in the course of the development of “national environmental space scanning monitoring technology” conducted in the “development of analysis and prediction technology of multi-type video and composite sensor database for early detection of environmental disasters”.

  In recent years, interest in the environment has increased and the recognition that it can have a direct impact on our lives has increased so that various environmental pollution such as water pollution and air pollution can be accurately measured and predicted. The need to do is increasing.

  Conventionally, many sensors (eg, green tide sensors, atmospheric sensors, etc.) that can collect data related to environmental pollution have been placed in the measurement target area, and environmental factors are identified through information collected through the sensors. The method of judgment was used. However, such a method is merely a method for analyzing and judging environmental factors at the time of collection, and there is a problem that results on how environmental factors will change in the future are not presented.

  In addition, in order to improve the analysis accuracy of environmental factors, a large number of sensors must be arranged, which causes a problem of high cost, and it is often difficult to arrange the sensors depending on the characteristics of the region.

  Therefore, there is an increasing demand for a method for predicting environmental pollution through satellite images that can be collected by a relatively simple method and SNS that is frequently used in recent years.

  The problem to be solved by the present invention is to provide a system capable of accurately predicting environmental pollution from satellite images and / or postings posted by a user on a social network.

  According to one aspect of the present invention, during a predetermined collection period, a satellite image data collection module that collects a plurality of satellite image data obtained by photographing a ground surface of a target area for each unit time, and the plurality of satellite images For each data, the satellite image data is analyzed to generate corresponding satellite image analysis data, and a satellite image analysis data set including the plurality of generated satellite image analysis data is generated. A data analysis module, an actual data collection module that collects a plurality of actual measurement data sensed from sensors provided in the target area for each unit time during the collection period, and each of the plurality of actual measurement data The actual measurement data is analyzed to generate environmental pollution data corresponding thereto, and the plurality of generated environmental pollution data are generated. An actual measurement data analysis module that generates an environmental pollution data set, a learning module that generates related information between the satellite video analysis data set and the environmental pollution data set, the satellite video data set, and the environmental pollution data There is provided an environmental pollution prediction system including a prediction module that generates predicted environmental pollution data from a prediction target satellite image obtained by capturing the target area at a predetermined prediction time using related information between sets.

  In one embodiment, the learning module performs machine learning on the satellite video analysis data of the unit time and the environmental pollution data of the unit time for each unit time during the collection period, and the satellite video analysis data set. And related information between the environment data set can be generated.

  In one embodiment, the environmental pollution data may be water quality data of a river located in the target area, and the sensor may be a green tide sensor.

  In one embodiment, the satellite image data analysis module may analyze a hue of a river portion indicated in the satellite image data to generate satellite image analysis data corresponding to the satellite image data.

  In one embodiment, the environmental pollution data may be air pollution degree data of the target area, and the sensor may be an air sensor.

  In one embodiment, the satellite image data analysis module may analyze the transparency of the satellite image data to generate satellite image analysis data corresponding to the satellite image data.

  In one embodiment, the environmental pollution prediction system further includes an air environment data collection module that collects air environment data of the target area for each unit time during the collection period, and the satellite image data analysis module includes: The satellite image data is corrected based on the atmospheric environment data corresponding to the unit time when the satellite image data was captured, and then the satellite image data is analyzed to generate corresponding satellite image analysis data. Can do.

  In one embodiment, the environmental pollution prediction system collects environmental pollution related keywords from postings corresponding to the target area posted on a predetermined social network service for each unit time during the collection period, The learning module further includes a keyword collection module that generates an environmental pollution-related keyword set including the collected environmental pollution-related keywords for each unit time. The learning module includes the satellite image analysis data set and the environmental pollution-related keyword set. Generating information related to the environmental pollution data set, and the prediction module uses the satellite video data set and the related information between the environmental pollution keyword related set and the environmental pollution data set to determine predetermined information. Predicted satellite image capturing the target area at the predicted time, and The predicted environmental pollution data from serial bulletin environmental pollution associated keywords included in the corresponding to the target area which is posted on the social networking service to the predicted time can be generated.

  In one embodiment, the learning module performs machine learning on the satellite video analysis data of the unit time, the environmental pollution related keywords of the unit time, and the environmental pollution data of the unit time for each unit time during the collection period. In operation, the satellite image analysis data set and the related information between the environmental pollution related keyword set and the environmental pollution data set may be generated.

  In one embodiment, the posting corresponding to the target area posted on the social network service is posted on the social network service by a terminal of a user located in the target area or an area within a certain range from the target area. Can be a posted post.

  According to another aspect of the present invention, during a predetermined collection period, environmental pollution-related keywords are collected from postings corresponding to target areas posted on a predetermined social network service for each unit time. A keyword collection module for generating a set of environmental pollution-related keywords including the collected environmental pollution-related keywords for each unit time, and sensing from sensors provided in the target area for each unit time during the collection period. A sensor data collection module for collecting a plurality of measured data, and analyzing the measured data for each of the plurality of measured data to generate environmental pollution data corresponding to the generated data. An actual measurement data analysis module for generating an environmental pollution data set including the environmental pollution data, and the environmental pollution The learning module for generating the related information between the continuous keyword set and the environmental pollution data set, and the related information between the environmental pollution related keyword set and the environmental pollution data set, and An environmental pollution prediction system is provided that includes a prediction module that generates predicted environmental pollution data from environmental pollution related keywords included in a posting corresponding to the target area posted on a social network service.

  According to another aspect of the present invention, the environmental pollution prediction system collects a plurality of satellite video data obtained by photographing a ground surface of a target area for each unit time during a predetermined collection period. And the environmental pollution prediction system analyzes the satellite image data for each of the plurality of satellite image data to generate corresponding satellite image analysis data, and the generated plurality of satellites. Satellite image data analysis stage for generating a satellite image analysis data set including image analysis data, and the environmental pollution prediction system is detected from sensors provided in the target area for each unit time during the collection period. Sensor data collection stage for collecting a plurality of actually measured data, and the environmental pollution prediction system, for each of the plurality of actually measured data, Analyzing the measured data, generating environmental pollution data corresponding to the generated data, and generating an environmental pollution data set including the plurality of generated environmental pollution data, and the environmental pollution prediction system, A learning step of generating related information between the satellite image analysis data set and the environmental pollution data set, and the environmental pollution prediction system calculates related information between the satellite video data set and the environmental pollution data set. And a prediction step of generating predicted environmental pollution data from a prediction target satellite image obtained by photographing the target area at a predetermined prediction time.

  In one embodiment, the environmental pollution data may be water quality data of a river located in the target area, and the sensor may be a green tide sensor.

  In one embodiment, the satellite image data analysis step may include generating a satellite image analysis data corresponding to the satellite image data by analyzing a hue of a river portion indicated in the satellite image data.

  In one embodiment, the environmental pollution data may be air pollution degree data of the target area, and the sensor may be an air sensor.

  In one embodiment, the satellite image data analyzing step may include analyzing the transparency of the satellite image data to generate satellite image analysis data corresponding to the satellite image data.

  In one embodiment, the environmental pollution prediction method collects environmental pollution-related keywords from postings corresponding to the target area posted on a predetermined social network service for each unit time during the collection period, The method further includes a keyword collecting step of generating an environmental pollution related keyword set including the collected environmental pollution related keywords for each unit time, wherein the learning step includes the satellite image analysis data set and the environmental pollution related keyword set; Generating relevant information between the environmental pollution data set, wherein the prediction step uses the satellite video data set and the relevant information between the environmental pollution keyword related set and the environmental pollution data set. , A prediction target satellite image obtained by photographing the target area at a predetermined prediction time and the prediction time The may include the step of generating the predicted environmental pollution data from environmental pollution associated keywords included in the bulletin corresponding to the target area which is posted on a social network service.

  According to another aspect of the present invention, the environmental pollution prediction system is configured to perform environmental pollution related from a posting corresponding to a target area posted on a predetermined social network service for each unit time during a predetermined collection period. A keyword collecting stage for collecting keywords and generating an environmental pollution-related keyword set including the collected environmental pollution-related keywords for each unit time; and the environmental pollution prediction system includes each unit during the collection period. A sensor data collection stage for collecting a plurality of actual measurement data detected from sensors provided in the target area according to time, and the environmental pollution prediction system, for each of the plurality of actual measurement data, Analyzing the generated environmental pollution data corresponding to the generated environmental pollution data, and including the generated environmental pollution data. An actual measurement data analysis stage for generating a pollution data set; a learning stage in which the environmental pollution prediction system generates related information between the environmental pollution related keyword set and the environmental pollution data set; and the environmental pollution prediction system. The environmental pollution included in the posting corresponding to the target area posted on the social network service at a predetermined prediction time using the related information between the environmental pollution related keyword set and the environmental pollution data set. An environmental pollution prediction method is provided that includes a prediction step of generating predicted environmental pollution data from related keywords.

  According to another aspect of the present invention, there is provided a computer program provided in a data processing apparatus and stored in a recording medium for performing the above-described method.

  According to another aspect of the present invention, an environmental pollution prediction system includes a processor and a memory that stores a computer program executed by the processor, and the computer program is executed by the processor. In this case, an environmental pollution prediction system is provided in which the environmental pollution prediction system performs the above-described method.

  According to an embodiment of the present invention, environmental pollution can be accurately predicted from satellite images. In addition, it is possible to accurately predict environmental pollution from the postings posted by the user on the social network.

  In order to more fully understand the drawings cited in the detailed description of the invention, a brief description of each drawing is provided.

1 is a diagram schematically illustrating a driving environment of an environmental pollution prediction system according to an exemplary embodiment of the present invention. 1 is a diagram schematically illustrating a configuration of an environmental pollution prediction system according to an embodiment of the present invention. It is drawing which shows an example of satellite image data. It is a flowchart which shows the environmental pollution prediction method by one Embodiment of this invention. 5 is a flowchart illustrating an environmental pollution prediction method according to another embodiment of the present invention. 6 is a flowchart illustrating an environmental pollution prediction method according to still another embodiment of the present invention.

  Since the present invention can be modified in various ways and have various embodiments, specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this should not be construed as limiting the invention to the specific embodiments, but should be understood to include any transformations, equivalents or alternatives that fall within the spirit and scope of the invention. . In describing the present invention, when it is determined that a specific description of a related known technique may obscure the gist of the present invention, a detailed description thereof will be omitted.

  The terms such as “first” and “second” are used to describe various components, but the components are not limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

  The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. A singular expression includes the plural expression unless the context clearly indicates otherwise.

  In this specification, terms such as “including” or “having” are intended to indicate that there is a feature, number, step, action, component, part, or combination thereof as described in the specification. It should be understood that it does not exclude the presence of, or the possibility of adding, one or more other features or numbers, steps, actions, components, parts or combinations thereof. .

  Also, in this specification, when any one component 'transmits' data to another component, the component directly transmits the data to the other component. It means that the data can also be transmitted to the other component through at least one further component. On the other hand, when any one component 'directly transmits' data to another component, the data is transmitted to the other component without passing through the other component by the component. Means.

  Hereinafter, the present invention will be described in detail with reference to the accompanying drawings, centering on embodiments of the present invention. The same reference numerals in each drawing denote the same members.

  FIG. 1 is a diagram schematically illustrating a driving environment of an environmental pollution prediction system according to an embodiment of the present invention. In order to implement an environmental pollution prediction method according to an embodiment of the present invention, an environmental pollution prediction system 100 may be provided.

  The environmental pollution prediction system 100 can predict environmental pollution in a specific area (hereinafter referred to as “target area”) that is a target of environmental pollution prediction.

  As shown in FIG. 1, the environmental pollution prediction system 100 can collect satellite video data on the ground surface of the target area captured by the satellite 200 above the target area.

  In addition, the environmental pollution prediction system 100 can collect environmental pollution related keywords included in a posting corresponding to the target area posted on a predetermined social network service providing server 300. The environmental pollution-related keyword may be a predetermined keyword such as “odor” or “pollution”, for example.

  On the other hand, the posting corresponding to the target area is a keyword related to the target area, and may be, for example, a place name of the target area or a place name of an area within a certain range from the target area.

  The posting corresponding to the target area may be a posting posted by a user located in the vicinity of the target area. That is, the terminal of the user located in the target area or an area within a certain range from the target area may be a posting posted on the social network service.

  The environmental pollution prediction system 100 can collect actual measurement data measured by a plurality of sensors 400-1 to 400-M provided in the target area. The actual measurement data is data that can confirm the degree of environmental pollution, and may include a green tide sensor, a fine dust sensor, a chlorophyll-a sensor, a turbidity sensor, and the like.

  The environmental pollution prediction system 100 can collect phase video data, environmental pollution-related keywords, and actual measurement data for each unit time during a certain period (collection period). That is, the environmental pollution prediction system 100 can collect phase video data, environmental pollution related keywords, and actual measurement data every unit time during the collection period. Accordingly, the phase image data corresponding to the specific unit time, the environmental pollution related keyword, and the actual measurement data may be data collected at the same time (that is, the specific unit time).

  Meanwhile, the environmental pollution prediction system 100 can predict the degree of environmental pollution at a certain time point (hereinafter referred to as “prediction time”) after the collection period has elapsed. The environmental pollution prediction system 100 can generate predicted environmental pollution data indicating the degree of environmental pollution predicted in the prediction time.

  More specifically, the environmental pollution prediction system 100 analyzes each of a plurality of satellite video data collected every unit time within the collection period, and obtains satellite video analysis data corresponding to the satellite video data. And generating a satellite image analysis data set including the plurality of generated satellite image analysis data.

  The environmental pollution prediction system 100 may generate an environmental pollution keyword set including environmental pollution related keywords collected every unit time within the collection period.

  In addition, the environmental pollution prediction system 100 analyzes each of a plurality of actually measured data collected every unit time within the collection period, and generates environmental pollution data corresponding to the actually measured data. An environmental pollution data set including a plurality of environmental pollution data can be generated.

  Meanwhile, the environmental pollution prediction system 100 generates the satellite image data set and / or related information between the environmental pollution keyword related set and the generated environmental pollution data set, and the generated information is continuously generated. Using related information, a satellite image (predicted satellite image) collected at a predetermined predicted time and / or a post corresponding to the target area posted on the SNS provided by the social network providing system 300 at the predicted time. Predictive environmental pollution data can be generated from environmental pollution-related keywords included in objects. That is, the predicted environmental pollution data is stored in a satellite image (predicted satellite image) collected at a predetermined predicted time and / or in the target area posted on the SNS provided by the social network providing system 300 at the predicted time. The information may be information on the degree of environmental pollution predicted from the environmental pollution-related keywords included in the corresponding posting.

  Hereinafter, the detailed configuration of the environmental pollution prediction system 100 will be described in more detail with reference to FIG. FIG. 2 is a diagram schematically showing a configuration of an environmental pollution prediction system according to an embodiment of the present invention.

  As shown in FIG. 2, the environmental pollution prediction system 100 includes a satellite video data collection module 110, a satellite video data analysis module 120, an actual data collection module 130, an actual data analysis module 140, a learning module 150, a prediction module 160, An air environment data collection module 170 and a keyword collection module 180 may be included.

  Depending on the embodiment of the present invention, some of the above-described components may not necessarily correspond to the components essential to the implementation of the present invention. It goes without saying that the prediction system 100 may include more components. For example, the environmental pollution prediction system 100 includes other components of the environmental pollution prediction system 100 (for example, a satellite video data collection module 110, a satellite video data analysis module 120, an actual measurement data collection module 130, an actual data analysis module 140, It may further include a control module (not shown) that can also control the functions and / or resources of the learning module 150, prediction module 160, atmospheric environment data collection module 170, and / or keyword collection module 180).

  The environmental pollution prediction system 100 includes hardware resources and / or software necessary for implementing the technical idea of the present invention, and always represents one physical component or one device. Does not mean. That is, the environmental pollution prediction system 100 means a logical combination of hardware and / or software provided to embody the technical idea of the present invention, and is provided in devices separated from each other when necessary. Thus, by performing each function, it may be embodied as a set of logical configurations for embodying the technical idea of the present invention. In addition, the environmental pollution prediction system 100 may mean a set of configurations separately implemented for each function or role for implementing the technical idea of the present invention. For example, the satellite image data collection module 110, the satellite image data analysis module 120, the actual measurement data collection module 130, the actual measurement data analysis module 140, the learning module 150, the prediction module 160, the atmospheric environment data collection module 170, and / or the keyword collection module. 180 may be located on different physical devices or on the same physical device. In some embodiments, the satellite image data collection module 110, the satellite image data analysis module 120, the actual measurement data collection module 130, the actual measurement data analysis module 140, the learning module 150, the prediction module 160, the atmospheric environment data collection module 170, keywords The detail elements that make up each individual module, such as the collection module 180, are also located on different physical devices, and the detail elements located on different physical devices are organically coupled to each other, which each individual module performs. Functions can also be realized.

  In the present specification, the module means a functional and structural combination of hardware for performing the technical idea of the present invention and software for driving the hardware. For example, the module means a logical unit of hardware resources for executing a predetermined code and the predetermined code, and means a physically linked code or a kind of hardware. That is not what can be easily inferred by an average expert in the technical field of the present invention.

  The satellite image data collection module 110 can collect a plurality of satellite image data obtained by photographing the ground surface of the target area for each unit time during a predetermined collection period. As described above, the satellite video data may be video data captured by the satellite 200 above the target area. The satellite video data collection module 110 collects satellite video data for the ground surface of the target area captured by the satellite 200 every unit time, and collects satellite video data during the collection period. it can.

  FIG. 3 is a diagram illustrating an example of satellite video data. As shown in FIG. 3, the satellite image data may include a river portion. That is, in the example of FIG. 3, the target area may include a river.

  Meanwhile, the satellite image data analysis module 120 analyzes the satellite image data for each of the plurality of satellite image data, generates corresponding satellite image analysis data, and generates the generated plurality of the plurality of satellite image data. A data group including satellite image analysis data (hereinafter referred to as “satellite image analysis data set”) can be generated.

  The method in which the satellite image data analysis module 120 analyzes each satellite image data may vary depending on the type of pollution predicted by the environmental pollution prediction system 100. Depending on the embodiment, the satellite image data analysis module 120 analyzes the hue and transparency of each satellite image data, and a more specific method will be described later.

  Meanwhile, the actual measurement data collection module 130 collects a plurality of actual measurement data sensed from the sensors 400-1 to 400-M provided in the target area for each unit time during the collection period. Can do. The sensors 400-1 to 400-M vary depending on the type of pollution predicted by the environmental pollution prediction system 100. For example, when the environmental pollution prediction system 100 is a system that predicts water quality of a river, The sensors 400-1 to 400-M may be water quality detection sensors (for example, a turbidity sensor, a green tide sensor, a chlorophyll-a sensor) that can grasp the level of water quality. Alternatively, when the environmental pollution prediction system 100 is a system that predicts the degree of air pollution, the sensors 400-1 to 400-M are sensors that can predict the degree of air pollution (for example, a dust sensor). It can be.

  Meanwhile, the actual measurement data analysis module 140 analyzes the actual measurement data for each of the plurality of actual measurement data, generates environmental pollution data corresponding to the analysis data, and generates the generated environmental pollution data. A data group including the following (hereinafter referred to as “environmental pollution data set”) can be generated.

  In one embodiment, the environmental pollution data is water quality data of a river located in the target area, and the sensors 400-1 to 400-M are water quality detection sensors that can check the water quality of the river. obtain. For example, the water quality sensing may be a green tide sensor, a chlorophyll-a sensing sensor and / or a turbidity sensor. In this case, the satellite image data analysis module 120 can analyze the hue of the river portion indicated in the satellite image data and generate satellite image analysis data corresponding to the satellite image data. This is because the hue of the river portion is determined to be different depending on the water quality of the river. In the present embodiment, the analysis result of each satellite image data generated by the satellite image data analysis module 120 is given to each pixel or a certain group of pixels included in the river portion indicated by the satellite image data. (For example, a hue value of each pixel) or a predetermined value (for example, an average value) determined by a numerical value assigned to each pixel or a certain pixel group.

  In another embodiment, the environmental pollution data is air pollution level data of the target area, and the sensors 400-1 to 400-M can check the air pollution level of the target area. It can be. For example, the atmospheric sensor may be a fine dust sensor. In this case, the satellite image data analysis module 120 can analyze the transparency of the satellite image data and generate satellite image analysis data corresponding to the satellite image data. This is because the transparency of an image obtained by photographing the target area is determined depending on the degree of air pollution in the target area. In this embodiment, the analysis result of each satellite image data generated by the satellite image data analysis module 120 is a numerical value (for example, each value) assigned to each pixel or a certain group of pixels included in the satellite image data. Pixel transparency, etc.), and may be a predetermined numerical value (for example, an average value) determined by a numerical value given to each pixel or a certain group of pixels.

  On the other hand, the satellite image obtained by photographing the ground surface of the target area captured by the satellite 200 may vary depending on the atmospheric environment (for example, humidity, clouds, wind direction, etc.) of the target area. Therefore, in one embodiment, the satellite image data analysis module 120 may correct each of the plurality of satellite image data, analyze the satellite image data, and generate corresponding satellite image analysis data. it can. Therefore, the atmospheric environment data collection module 170 collects the atmospheric environment data of the target area for each unit time during the collection period, and the satellite image data analysis module 120 captures the satellite image data. The satellite image data can be corrected based on the atmospheric environment data corresponding to the unit time.

  The air environment data collection module 170 collects air environment data of the target area from a predetermined air environment data providing server (for example, the Japan Meteorological Agency server), and the air environment data is data indicating the air environment of the target area. (For example, humidity, rainfall, snowfall, airflow, cloud level, etc.).

  The satellite image data analysis module 120 can correct the brightness, hue, saturation, and the like of the satellite image data of the unit time based on the atmospheric environment data of the specific unit time.

  Meanwhile, the keyword collection module 180 collects environmental pollution related keywords from postings corresponding to the target area posted on a predetermined social network service for each unit time during the collection period. In addition, an environmental pollution related keyword set including the environmental pollution related keywords for each unit time can be generated. The keyword collection module 180 may collect environmental pollution-related keywords included in a posting corresponding to the target area from the social network service providing server 300.

  The environmental pollution-related keyword is one of predetermined keywords related to environmental pollution, and may be, for example, “odor”, “contamination”, “green tide”, “bad odor”, “floating matter”, and the like. On the other hand, the posting corresponding to the target area is a keyword related to the target area, and may be, for example, a place name of the target area or a place name of an area within a certain range from the target area.

  In one embodiment, the post corresponding to the target area may be a post posted by a user located in the vicinity of the target area. That is, the terminal of the user located in the target area or an area within a certain range from the target area may be a posting posted on the social network service. In this embodiment, since a user located in the vicinity of the target area posts on the SNS the contents experienced at the time of posting, the posting corresponding to the target area indicates the contamination status of the target area at the time of posting. It can be a material to show.

  Meanwhile, the learning module 150 may generate related information between the satellite image analysis data set and / or the environmental pollution related keyword set and the environmental pollution data set. That is, according to an exemplary embodiment, the learning module 150 may include information related to the satellite image analysis data set and the environmental pollution data set, information related to the environmental pollution related keyword set and the environmental pollution data set, and The satellite information analysis data set and the related information between the environmental pollution related keyword set and the environmental pollution data set may be generated.

In one embodiment, the learning module 150 may generate the related information by performing machine learning using individual data included in each set as an input value.
Since the data included in the satellite image analysis data set, the environmental pollution related keyword set, and the environmental pollution data set are all collected in a specific unit within the collection period, the satellite image analysis data set, Each of the data included in the environmental pollution related keyword set and the environmental pollution data set may correspond to a specific unit time. The learning module 150 may include satellite image data corresponding to the unit time and / or environmental pollution related keywords corresponding to the unit time and environmental pollution data corresponding to the unit time for each unit time in the collection period. Can be used as an input value.

  For example, the learning module 150 can perform various known machine learning algorithms such as a decision tree learning method, a related rule learning method, an artificial neural network, a support vector machine, and clustering.

  Meanwhile, after the learning module 150 generates the satellite image analysis data set and / or the related information between the environmental pollution related keyword set and the environmental pollution data set, the prediction module 150 160 is a prediction target satellite image obtained by photographing the target area at a predetermined predicted time using the generated related information and / or a post corresponding to the target area posted on the social network service at the predicted time. Predictive environmental pollution data can be generated from the environmental pollution-related keywords included in.

  The predicted environmental pollution data generated in this way is not the data measured from the sensor provided in the target area, but the satellite video (predicted target satellite video) collected at the predicted time and / or the predicted time. The information may be information on the degree of environmental pollution predicted from environmental pollution-related keywords included in a posting corresponding to the target area posted on the SNS provided by the social network providing system 300. Therefore, according to the above-described technical idea of the present invention, the degree of environmental pollution in the target area can be accurately determined from satellite images and / or postings posted on the SNS even when the target area is not equipped with a sensor. There is an effect that can be predicted.

  FIG. 4 is a flowchart illustrating an environmental pollution prediction method according to an embodiment of the present invention. FIG. 4 is a diagram illustrating a method for predicting environmental pollution data from satellite video data.

  Referring to FIG. 4, the environmental pollution prediction system 100 may generate a satellite image analysis data set SI and an environmental pollution data set SP by performing step S100.

  In step S100, the environmental pollution prediction system 100 may perform steps S101 to S108 every unit time Ti (1 <= i <= N). In FIG. 4, the total collection time for collecting the satellite image data and the actual measurement data may be (Ti + 1−Ti) * N.

In step S101, the environmental pollution prediction system 100 may collect satellite video data I _i obtained by photographing the ground surface of the target area in unit time Ti.

In step S102, the environmental pollution prediction system 100 collects atmospheric environment data E _i measured in unit time Ti, and in step S103, the satellite image data I _i can be corrected based on the atmospheric environment data E _i. .

In step S104, the environmental pollution prediction system 100 may analyze the resultant structure obtained by correcting the satellite image data I _i, and generates a satellite image analyzing data A _i, at step S105, the generated satellite image analysis data A _i Can be added to the satellite image analysis data set SI.

On the other hand, in step S106, the environmental pollution prediction system 100 may collect measured data _{D i} sensed from the sensor 400-1 to 400-M provided in the target area in a unit time Ti, at step S107, the measured analyzing the data _{D i,} it generates environmental pollution data _{P i,} at step S108, the environmental pollution data _{P i} may be added to the environment pollution dataset SP.

  If the satellite image analysis data set SI and the environmental pollution data set SP including the data collected every unit time are generated during the collection period by the above method, the environmental pollution prediction is performed in step S110. The system 100 can generate related information R between the satellite image analysis data set SI and the environmental pollution data set SP through machine learning.

  If the related information R is generated, the environmental pollution prediction system 100 may generate predicted environmental pollution data from a prediction target satellite image captured at a predetermined prediction time in step S120. More specifically, the environmental pollution prediction system 100 corrects the satellite image to be predicted based on the atmospheric environment data measured at the prediction time, and then generates predicted satellite image analysis data therefrom. Using the information R, the predicted environmental pollution data that is environmental pollution data corresponding to the predicted satellite image analysis data can be generated.

  FIG. 5 is a flowchart illustrating an environmental pollution prediction method according to another embodiment of the present invention. FIG. 5 is a diagram showing a method for predicting environmental pollution data from the postings posted on the SNS.

  Referring to FIG. 5, the environmental pollution prediction system 100 may generate an environmental pollution related keyword set SK and an environmental pollution data set SP by performing step S200.

  In step S200, the environmental pollution prediction system 100 may perform steps S201 to S205 every unit time Ti (1 <= i <= N). In FIG. 5, the total collection time for collecting the environmental pollution keyword and the actual measurement data may be (Ti + 1−Ti) * N.

In step S201, the environmental pollution prediction system 100 may collect environmental pollution related keyword _{K i} from posting corresponding to the target area that is posted on the SNS unit time Ti, at step S202, the environmental pollution prediction system 100 The environmental pollution related keyword K _i can be added to the environmental pollution related keyword set SK.

On the other hand, in step S203, the environmental pollution prediction system 100 may collect measured data _{D i} sensed from the sensor 400-1 to 400-M provided in the target area in a unit time Ti, at step S204, the measured analyzing the data _{D i,} it generates environmental pollution data _{P i,} at step S205, the environmental pollution data _{P i} may be added to the environment pollution dataset SP.

  If the environmental pollution related keyword set SK and the environmental pollution data set SP including data collected every unit time are generated during the collection period by the above method, the environmental pollution prediction is performed in step S210. The system 100 may generate related information R between the environmental pollution related keyword set SK and the environmental pollution data set SP through machine learning.

  If the related information R is generated, in step S220, the environmental pollution prediction system 100 predicts the predicted environment from the environmental pollution related keywords included in the posting corresponding to the target area posted on the SNS at a predetermined prediction time. Contamination data can be generated.

  FIG. 6 is a flowchart illustrating an environmental pollution prediction method according to still another embodiment of the present invention. FIG. 6 is a diagram illustrating a method for predicting environmental pollution data from satellite video data and a posting posted on the SNS.

  Referring to FIG. 6, the environmental pollution prediction system 100 may generate a satellite image analysis data set SI and an environmental pollution data set SP by performing step S300.

  In step S300, the environmental pollution prediction system 100 may perform steps S301 to S308 every unit time Ti (1 <= i <= N). In FIG. 6, the total collection time for collecting the satellite video data, the environmental pollution related keywords, and the actual measurement data may be (Ti + 1−Ti) * N.

In step S301, the environmental pollution prediction system 100 may collect satellite video data I _i obtained by photographing the ground surface of the target area in unit time Ti.

In step S302, the environmental pollution prediction system 100 analyzes the satellite image data _{I i,} and generates a satellite image analyzing data _{A i,} at step S303, the generated satellite image analyzing data _{A i} satellite image analysis data was It can be added to the set SI. Of course, in some implementations, the environmental pollution prediction system 100 may analyze the corrected satellite image data after correcting the satellite image data I _i as in the embodiment of FIG.

On the other hand, in step S304, the environmental pollution prediction system 100 may collect environmental pollution related keyword K _i from posting corresponding to the target area that is posted on the SNS unit time Ti, at step S305, the environmental pollution prediction system 100, it is possible to add an environment pollution-related keyword K _i on the environment pollution-related set of keywords SK.

On the other hand, in step S306, the environmental pollution prediction system 100 may collect measured data _{D i} sensed from the sensor 400-1 to 400-M provided in the target area in a unit time Ti, at step S307, the measured analyzing the data _{D i,} it generates environmental pollution data _{P i,} at S308 step, the environmental pollution data _{P i} may be added to the environment pollution dataset SP.

  If the satellite image analysis data set SI, the environmental pollution related keyword set SK, and the environmental pollution data set SP including the data collected every unit time are generated during the collection period by the above-described method, S310 is generated. The environmental pollution prediction system 100 may generate the satellite image analysis data set SI and the related information R between the environmental pollution related keyword set SK and the environmental pollution data set SP through machine learning.

  If the related information R is generated, in step S320, the environmental pollution prediction system 100 displays a prediction target satellite image captured at a predetermined prediction time and a display corresponding to the target area posted on the SNS at the prediction time. Predictive environmental pollution data can be generated from environmental pollution-related keywords included in objects.

  Meanwhile, in some implementations, the environmental pollution prediction system 100 may include a processor and a memory that stores a program executed by the processor. The processor may include a single core CPU or a multi-core CPU. The memory includes high-speed random access memory and may include non-volatile memory such as one or more magnetic disk storage devices, flash memory devices, or other non-volatile solid state memory devices. Access to the memory by the processor and other components is controlled by a memory controller. Here, when the program is executed by the processor, the environmental pollution prediction system 100 according to the present embodiment is caused to perform the environmental pollution prediction method described above.

  Meanwhile, the environmental pollution prediction method according to the embodiment of the present invention is embodied in the form of computer-readable program instructions and stored in a computer-readable recording medium, and the control program and the target program according to the embodiment of the present invention. Can also be stored on a computer-readable recording medium. Computer-readable recording media include all types of recording devices that can store data that can be read by a computer system.

  The program instructions recorded on the recording medium may be specially designed and configured for the present invention, or may be known and usable by those skilled in the software field.

  Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and floppy disks. Includes hardware devices specially configured to store and execute program instructions such as magnetic-optical media such as optical disks and ROM, RAM, and flash memory. . The medium described above may also be a transmission medium such as a light or metal line or a waveguide including a carrier wave for transmitting a signal designating a program command, a data structure, or the like. The computer-readable recording medium can be distributed to computer systems connected via a network and stored and executed as computer-readable code in a distributed manner.

  Examples of program instructions include not only machine language code such as that produced by a compiler, but also high-level language code that can be executed by a device that electronically processes information using an interpreter, for example, a computer.

  The hardware device described above is configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

  The above description of the present invention is for illustrative purposes, and those skilled in the art can easily modify the present invention in other specific forms without changing the technical idea and essential features of the present invention. You will understand that there is. Accordingly, it should be understood that the above-described embodiments are illustrative in all aspects and not limiting. For example, each component described as a single type can be implemented in a distributed manner, and components described as being similarly distributed are also implemented in a combined form.

  The scope of the present invention is indicated by the following claims rather than the above detailed description, and all modifications or variations derived from the meaning and scope of the claims and the equivalent concept thereof are disclosed. Should be construed as being included within the scope of the present invention.

  The present invention can be used in an environmental pollution prediction system and method.

Claims (20)

A satellite video data collection module that collects a plurality of satellite video data obtained by photographing the ground surface of the target area for each unit time during a predetermined collection period;
For each of the plurality of satellite video data, the satellite video data is analyzed to generate corresponding satellite video analysis data, and the satellite video analysis data set including the generated satellite video analysis data A satellite image data analysis module for generating
An actual measurement data collection module for collecting a plurality of actual measurement data sensed from sensors provided in the target area for each unit time during the collection period;
For each of the plurality of actual measurement data, the actual measurement data is analyzed to generate environmental pollution data corresponding thereto, and the actual measurement data for generating an environmental pollution data set including the plurality of generated environmental pollution data An analysis module;
A learning module for generating relevant information between the satellite image analysis data set and the environmental pollution data set;
A prediction module for generating predicted environmental pollution data from a predicted target satellite video obtained by photographing the target area at a predetermined predicted time using related information between the satellite video data set and the environmental pollution data set;
Environmental pollution prediction system including. The learning module
Machine learning is performed on the satellite video analysis data of the unit time and the environmental pollution data of the unit time for each unit time during the collection period, and the relationship between the satellite video analysis data set and the environmental data set The environmental pollution prediction system according to claim 1 which generates information.   The environmental pollution prediction system according to claim 1, wherein the environmental pollution data is water quality data of a river located in the target area, and the sensor is a green tide sensor. The satellite video data analysis module includes:
4. The environmental pollution prediction system according to claim 3, wherein the hue of the river portion indicated in the satellite image data is analyzed to generate satellite image analysis data corresponding to the satellite image data.   The environmental pollution prediction system according to claim 1, wherein the environmental pollution data is air pollution degree data of the target area, and the sensor is an atmospheric sensor. The satellite video data analysis module includes:
6. The environmental pollution prediction system according to claim 5, wherein the transparency of the satellite video data is analyzed to generate satellite video analysis data corresponding to the satellite video data. The environmental pollution prediction system includes:
An air environment data collection module that collects air environment data of the target area for each unit time during the collection period;
The satellite video data analysis module includes:
The satellite image data is analyzed after correcting the satellite image data based on the atmospheric environment data corresponding to the unit time when the satellite image data was captured, and satellite image analysis data corresponding to the satellite image data is generated. The environmental pollution prediction system according to 1. The environmental pollution prediction system includes:
During the collection period, environmental pollution-related keywords are collected from postings corresponding to the target area posted on a predetermined social network service for each unit time, and the collected environmental pollution-related per unit time is collected. It further includes a keyword collection module that generates a set of environmental pollution related keywords that contain keywords,
The learning module
Generating relevant information between the satellite image analysis data set and the environmental pollution related keyword set and the environmental pollution data set;
The prediction module is
Using the satellite video data set and the related information between the environmental pollution keyword related set and the environmental pollution data set, a prediction target satellite video obtained by photographing the target area at a predetermined prediction time and the social at the prediction time The environmental pollution prediction system according to claim 1, wherein the predicted environmental pollution data is generated from environmental pollution related keywords included in a posting corresponding to the target area posted on a network service. The learning module
Performing machine learning on the satellite video analysis data of the unit time, the environmental pollution related keywords of the unit time, and the environmental pollution data of the unit time for each unit time during the collection period, and the satellite video analysis data set and The environmental pollution prediction system according to claim 8, wherein related information between the environmental pollution related keyword set and the environmental pollution data set is generated. Posts corresponding to the target area posted on the social network service are:
9. The environmental pollution prediction system according to claim 8, wherein a terminal of a user located in the target area or an area within a certain range from the target area is a post posted on the social network service. During a predetermined collection period, environmental pollution related keywords are collected from postings corresponding to the target area posted on a predetermined social network service for each unit time, and the environmental pollution related by the collected unit time is collected. A keyword collection module that generates a set of environmental pollution related keywords including keywords,
A sensor data collection module for collecting a plurality of actual measurement data detected from sensors provided in the target area for each unit time during the collection period;
For each of the plurality of actual measurement data, the actual measurement data is analyzed to generate environmental pollution data corresponding thereto, and the actual measurement data for generating an environmental pollution data set including the plurality of generated environmental pollution data An analysis module;
A learning module for generating relevant information between the environmental pollution related keyword set and the environmental pollution data set;
Using the related information between the environmental pollution related keyword set and the environmental pollution data set, the environmental pollution related included in the posting corresponding to the target area posted on the social network service at a predetermined prediction time. A prediction module that generates predicted environmental pollution data from keywords;
Environmental pollution prediction system including. A satellite image data collection stage in which the environmental pollution prediction system collects a plurality of satellite image data obtained by photographing the ground surface of the target area for each unit time during a predetermined collection period;
The environmental pollution prediction system analyzes the satellite image data for each of the plurality of satellite image data, generates corresponding satellite image analysis data, and generates the generated plurality of satellite image analysis data. A satellite image data analysis stage for generating a satellite image analysis data set including:
The environmental pollution prediction system collects a plurality of measured data sensed from sensors provided in the target area for each unit time during the collection period; and
The environmental pollution prediction system analyzes the actual measurement data for each of the plurality of actual measurement data, generates environmental pollution data corresponding thereto, and includes the generated plurality of the environmental pollution data. An actual data analysis stage to generate a data set;
A learning stage in which the environmental pollution prediction system generates relevant information between the satellite image analysis data set and the environmental pollution data set;
The environmental pollution prediction system generates predicted environmental pollution data from a prediction target satellite video obtained by photographing the target area at a predetermined prediction time, using related information between the satellite video data set and the environmental pollution data set. Prediction stage to
Environmental pollution prediction method including   The environmental pollution prediction method according to claim 12, wherein the environmental pollution data is water quality data of a river located in the target area, and the sensor is a green tide sensor. The satellite image data analysis step includes:
The environmental pollution prediction method according to claim 13, further comprising: analyzing a hue of a river portion indicated in the satellite video data to generate satellite video analysis data corresponding to the satellite video data.   The environmental pollution prediction method according to claim 12, wherein the environmental pollution data is air pollution degree data of the target area, and the sensor is an atmospheric sensor. The satellite image data analysis step includes:
The environmental pollution prediction method according to claim 15, further comprising: analyzing the transparency of the satellite image data to generate satellite image analysis data corresponding to the satellite image data. The environmental pollution prediction method includes:
During the collection period, environmental pollution-related keywords are collected from postings corresponding to the target area posted on a predetermined social network service for each unit time, and the collected environmental pollution-related per unit time is collected. It further includes a keyword collection stage that generates a set of environmental pollution-related keywords that contain keywords,
The learning stage includes
Generating relevant information between the satellite image analysis data set and the environmental pollution related keyword set and the environmental pollution data set;
The prediction step includes
Using the satellite video data set and the related information between the environmental pollution keyword related set and the environmental pollution data set, a prediction target satellite video obtained by photographing the target area at a predetermined prediction time and the social at the prediction time 13. The environmental pollution prediction method according to claim 12, further comprising the step of generating the predicted environmental pollution data from environmental pollution related keywords included in a posting corresponding to the target area posted on a network service. The environmental pollution prediction system collects environmental pollution-related keywords from postings corresponding to target areas posted on a predetermined social network service for each unit time during a predetermined collection period, and the collected units. A keyword collection stage for generating an environmental pollution-related keyword set including hourly environmental pollution-related keywords,
The environmental pollution prediction system collects a plurality of measured data sensed from sensors provided in the target area for each unit time during the collection period; and
The environmental pollution prediction system analyzes the actual measurement data for each of the plurality of actual measurement data, generates environmental pollution data corresponding thereto, and includes the generated plurality of the environmental pollution data. An actual data analysis stage to generate a data set;
A learning step in which the environmental pollution prediction system generates related information between the environmental pollution related keyword set and the environmental pollution data set;
The environmental pollution prediction system uses the related information between the environmental pollution related keyword set and the environmental pollution data set, and posts corresponding to the target area posted on the social network service at a predetermined prediction time. A prediction stage for generating predicted environmental pollution data from environmental pollution-related keywords contained in
Environmental pollution prediction method including   A computer program provided in a data processing apparatus and stored in a recording medium for performing the method according to claim 12 or 18. An environmental pollution prediction system,
A processor;
A memory for storing a computer program executed by the processor,
19. The environmental pollution prediction system, when the computer program is executed by the processor, causes the environmental pollution prediction system to perform the method according to claim 12 or 18.