Air pollution estimation algorithm using the rate of change of height of color components extracted from image data

Abstract

To provide a method and a system for a terminal having difficulty in fixing an angle of view to estimate aerosol concentration.SOLUTION: A method for estimating the aerosol concentration in a target area by using a computer includes: a physical quantity acquisition step of acquiring a physical quantity from a sample image obtained by photographing the target area; a numerical value for estimation acquisition step of acquiring a numerical value for estimation for estimating the aerosol concentration by using the physical quantity; and an aerosol concentration estimation step of estimating the aerosol concentration by using the numerical value for estimation. The physical quantity includes brightness for each of a plurality of color components at the plurality of positions of the sample image, and the height of the plurality of positions in the target area. The numerical value for estimation includes the rate of change of the ratio between the brightness of a first color and the brightness of a second color in the brightness of the plurality of color components, relative to the height.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an air pollution estimation algorithm that uses the rate of change of color components in image data. More specifically, the present invention relates to a method and system for estimating aerosol concentration based on the rate of change of height of physical quantities in captured image data. [Background technology]

[0002] Japanese Patent Application Laid-Open Publication No. 2023-061680 (Patent Document 1) describes a method and system for estimating aerosol concentration. This method narrows the image range to be analyzed to narrow ranges such as the upper sky, the lower sky, the boundary with the horizon, and building areas, and extracts physical quantities correlated with aerosol concentration for each range.

[0003] This method of estimating aerosol concentration has the problem that it requires a fixed angle of view using a fixed camera or the like. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2023-061680 Summary of the Invention [Problem to be solved by the invention]

[0005] An object of the invention described in this specification is to provide a method for estimating aerosol concentration without requiring a fixed angle of view, and a system for implementing the method. [Means for solving the problem]

[0006] The present invention is basically based on the finding that by analyzing a sample image of a target area and using the rate of change in color components, the intensity of each color component, and the ratio of these components from the sample image, the aerosol concentration can be estimated without fixing the angle of view.

[0007] One invention relates to a method for estimating aerosol concentration in a target area by a computer, which includes a physical quantity acquisition step (S101), a numerical value acquisition step (S102) for estimation, and an aerosol concentration estimation step (S103). The physical quantity acquisition step (S101) is a step for acquiring physical quantities from a sample image of a target area. The physical quantities include physical quantities for each of a plurality of color components at a plurality of positions at different heights in the sample image, and heights for the plurality of positions. The estimation value acquisition step (S102) is a step for obtaining estimation values ​​for estimating the aerosol concentration using the obtained physical quantities. The estimation values ​​include the rate of change in the magnitude of the values ​​of the physical quantities for each of the multiple color components. The aerosol concentration estimation step (S103) is a step for estimating the aerosol concentration using the estimation value.

[0008] Examples of the physical quantities for each of the plurality of color components include the luminance of the first color component and the luminance of the second color component, and examples of the estimation values ​​include the rate of change of the magnitude of the ratio of the luminance of the first color component to the luminance of the second color component.

[0009] An example of the physical quantity for each of the plurality of color components includes the intensity of the first color component, and an example of the estimation value includes the rate of change of the intensity of the first color component.

[0010] The example sample image is an image taken using a mobile device.

[0011] One invention relates to a computer-based aerosol concentration estimation system 1. This system 1 includes an imaging unit 3, a physical quantity acquisition unit 5, an estimation value acquisition unit 7, and an aerosol concentration estimation unit 9. The photographing unit 3 is an element for photographing a sample image. The physical quantity acquisition unit 5 is an element for obtaining predetermined physical quantities from the captured sample image. The physical quantities include physical quantities for each of a plurality of color components at a plurality of positions at different heights in the sample image, and heights at the plurality of positions. The estimation value acquisition unit 7 is a component for obtaining estimation values ​​for estimating the aerosol concentration using the obtained physical quantities. The estimation values ​​include the rate of change in the height of the values ​​of the physical quantities for each of the multiple color components. The aerosol concentration estimation unit 9 is an element for estimating the aerosol concentration using the estimation numerical value. [Effects of the Invention]

[0012] According to the above invention, the aerosol concentration in a target area can be calculated from a sample image captured without fixing the angle of view. Therefore, this specification can provide a method for estimating the aerosol concentration in a target area by simply capturing a sample image with a mobile device or the like, where it is difficult to fix the angle of view, and a system for implementing this method. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 1 is a flow chart illustrating an example of a method for estimating aerosol concentration in a target area. [Figure 2] FIG. 2 is a block diagram illustrating an example of a system for estimating aerosol concentration. [Figure 3] FIG. 3 is a conceptual diagram showing an example of the configuration of an aerosol concentration estimation system. [Figure 4] FIG. 4 is a conceptual diagram for explaining the aerosol concentration estimation process using machine learning. [Figure 5] FIG. 5 is a conceptual diagram showing an example of pattern matching. [Figure 6] FIG. 6 is a photograph in place of a drawing showing an example of a sample image. [Figure 7]FIG. 7 is a graph, instead of a drawing, showing the relationship between the ratio of brightness between B and G (horizontal axis) and the height of a predetermined area (position of the part excluding buildings, etc.) of a sample image. [Figure 8] Figure 8 is a graph, instead of a drawing, showing the relationship between the vertical rate of change (horizontal axis) of the blue to green brightness ratio (B / G) and the concentration (vertical axis) of suspended particulate matter (SPM) with a particle size of 10 μm or less. [Figure 9] Figure 9 is a graph, instead of a drawing, showing the relationship between the vertical rate of change (horizontal axis) of the green to red luminance ratio (G / R) and the concentration (vertical axis) of suspended particulate matter (SPM) with a particle size of 10 μm or less. [Figure 10] Figure 10 is a graph, instead of a drawing, showing the relationship between the vertical rate of change (horizontal axis) of the blue to red luminance ratio (B / R) and the concentration (vertical axis) of suspended particulate matter (SPM) with a particle size of 10 μm or less. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, embodiments of the present invention will be described with reference to the drawings. The present invention is not limited to the embodiments described below, but also includes appropriate modifications of the embodiments below within the scope obvious to those skilled in the art.

[0015] 1 is a flowchart showing an example of a method for estimating aerosol concentration in a target area. S indicates a step. One invention described in this specification relates to a method for estimating aerosol concentration in a target area using a computer.

[0016] 2 is a block diagram showing an example of a system for estimating aerosol concentration. This system 1 includes an imaging unit 3, a physical quantity acquisition unit 5, an estimation value acquisition unit 7, and an aerosol concentration estimation unit 9. This system is computer-based, and each process is performed by the computer.

[0017] The computer has an input unit, an output unit, a control unit, a calculation unit, and a memory unit, and each element is connected by a bus or the like to enable information exchange. For example, a control program or various information may be stored in the memory unit. When predetermined information is input from the input unit, the control unit reads the control program stored in the memory unit. Then, the control unit reads the information stored in the memory unit as appropriate and transmits it to the calculation unit. The control unit also transmits the input information to the calculation unit as appropriate. The calculation unit performs calculation processing using the received various information and stores it in the memory unit. The control unit reads the calculation results stored in the memory unit and outputs them from the output unit. In this manner, various processes and steps are performed. The various units and means execute these various processes. The computer may have a processor, and the processor may realize various functions and steps. The computer may be standalone. Some of the functions of the computer may be distributed between a server and a terminal. In this case, it is preferable that the server and the terminal can exchange information via a network such as the Internet or an intranet. The computer may include a processor and a memory coupled to the processor. The memory may store instructions that, when executed by the processor, cause the computer to perform various processes or function as various elements. The computer may be provided with various training data to build a learning model and perform various calculations through machine learning. In this case, the computer may perform various analyses using the learning model created through machine learning and deep learning of AI (artificial intelligence).

[0018] FIG. 3 is a conceptual diagram showing an example of the configuration of an aerosol concentration estimation system. This system may include a terminal connected to a network such as the Internet or an intranet, and a server connected to the network. Of course, a single computer or mobile terminal may function as the device of the present invention, or multiple servers may be present. In the example of FIG. 3, input terminals 31, 33, 35, and 37 are connected to a server 41 via a network 39. Examples of input terminals include a camera 31, a smartphone 33, a portable game console 35, and a laptop computer 37. Any device with a photographing unit, such as these, can function as an input terminal for this system. The captured sample images may be processed within these terminals or sent to the server 41 for processing therein. It is preferable that various information be exchanged wirelessly between the mobile terminal and the server. Information may also be exchanged between the terminal and the server using optical information.

[0019] Next, we will explain an example of a method for estimating aerosol concentration in a target area using a computer. As shown in Figure 1, this method includes a physical quantity acquisition step (S101), a numerical value acquisition step (S102) for estimation, and an aerosol concentration estimation step (S103). This series of algorithms for estimating aerosol concentration is called SNAP-CII.

[0020] The physical quantity acquisition process (S101) is a process for obtaining physical quantities from a sample image of a target area. The target area may be an area where a spatial photograph can be taken. The target area may be a certain region, or an area photographed from a certain location. The physical quantities are numerical values ​​used to obtain estimation values ​​used to estimate the aerosol concentration in the aerosol concentration estimation process. To obtain a sample image, a sample image photographed by a known photographing device may be input into a computer. The sample image may be a digital image of space obtained by photographing the target area with a photographing device. The physical quantities may be for each pixel, or may be for a specific region of the sample image.

[0021] Examples of physical quantities include physical quantities for multiple color components at multiple positions at different heights in the sample image and heights for the multiple positions. The target area is preferably, for example, an area of ​​the sample where no buildings or other structures exist. One example of a physical quantity includes information about the intensity of multiple color components at multiple positions in the sample image and heights in the target area for the multiple positions. Another example of a physical quantity is the intensity of each color component. Examples of the intensity of each color component include the luminance and reflectance of each RGB value. Color values ​​such as RGB may be corrected using gamma correction or other methods. The three axes of color components are not limited to RGB, and any other three axes may be used, such as CMY, CIE XYZ, CIE xyY, CIE u'v'Y, CIE LUV, CIE LAB, and CIE LCh.

[0022] The photographing unit 3 photographs a sample image. The sample image of the target area is input to a computer and stored in a storage unit as appropriate. At this time, one or more pieces of information from among (identification information of) the photographing location, the photographing angle, the distance from the target space, the photographing time, and the weather conditions at the time of photographing may also be input to the computer and stored in the storage unit in a state associated with the input sample image as appropriate.

[0023] The physical quantity acquisition unit 5 may divide the sample image into multiple regions and extract regions for estimating the aerosol concentration. Furthermore, the sample image may be analyzed to exclude regions with obstacles such as buildings or trees from the region for calculating the aerosol concentration. Furthermore, in the sky region, the amount of cloud cover may be analyzed to exclude the region from the region for calculating the aerosol concentration, or a coefficient or constant (correction value) for calculating the aerosol concentration may be calculated. The physical quantity acquisition unit 5 may obtain information regarding the height in the target region. The information regarding the height in the target region may be the altitude [m] at each position. In this case, the physical quantity acquisition unit 5 may obtain the altitude of each position or each pixel using an algorithm for estimating the altitude at each position or each pixel from the obtained sample image. Furthermore, the information regarding the height in the target region may be the relative position between the top and bottom edges of the target region to be analyzed. In this case, the physical quantity acquisition unit 5 may obtain the height of each position or each pixel from the relative position in the target region. Furthermore, the physical quantity acquisition unit 5 analyzes the sample image to obtain the intensity of each of multiple color components for each position or each pixel. The physical quantity acquisition unit 5 may also determine the intensity of each of a plurality of color components for each position or each pixel. Methods for determining the intensity of each of a plurality of color components for each position or each pixel and the intensity of each color component are publicly known, as described in JP 2023-061680 A. The physical quantities determined by the physical quantity acquisition unit 5 are stored in a storage unit as appropriate.

[0024] The estimation value acquisition step (S102) is a step of using the obtained physical quantities to obtain estimation values ​​for estimating aerosol concentration. The estimation values ​​are the rate of change in the magnitude of the values ​​of the physical quantities for each of a plurality of color components. An example of the estimation value is the rate of change in the magnitude of the ratio between the luminance of a first color component and the luminance of a second color component. Another example of the estimation value is the rate of change in the magnitude of the luminance of the first color component. Another example of the estimation value is the rate of change in the magnitude of the reflectance of the first color component. The computer reads the physical quantities obtained in the physical quantity acquisition step from the memory unit. Then, based on instructions from the program, the control unit reads necessary numerical values ​​from the memory unit as appropriate and uses the numerical values ​​together with the physical quantities to cause the calculation unit to perform a calculation to obtain the estimation values. The computer stores the obtained estimation values ​​in the memory unit as appropriate. Methods for obtaining the rate of change in the magnitude of a physical quantity are well known. The simplest method for obtaining the rate of change in the magnitude of a physical quantity is to divide the difference between the physical quantities at two points by the difference in the heights of the two points. Furthermore, a graph showing the relationship between the physical quantity and the height in a specific region can be created, and the rate of change in height (Δh) can be calculated by calculating the slope of the graph.

[0025] Specifically, the estimation value acquisition unit 7 uses the physical quantity for each color component of the pixel of the sample image and the height in the target area to determine the rate of change in the height of the physical quantity for each color component in the area where the sample image was captured. For example, if the position of the sample image or pixel position is (x, y), then y corresponds to the height (h). For example, the estimation value acquisition unit 7 reads out the intensity of each color for multiple points with the same x and different y from the storage unit. The estimation value acquisition unit 7 reads out the intensity of the first color and the intensity of the second color (for example, the luminance of blue (B) and the luminance of green (G) at a certain pixel) and determines the ratio (B / G) of the intensity of the first color to the intensity of the second color. For example, if R R , R G and R B Let be the brightness of each R, G, and B. An example of the ratio of the brightness of the first color to the brightness of the second color is R B / R G , R G / R R , and R R / RB The computer control unit calculates the R R R G R B value is used, and R B / R G , R G / R R , and R R / R B Then, the control unit of the computer performs a calculation to obtain the calculated R B / R G , R G / R R , and R R / R B The value of is stored in the storage unit as appropriate. The estimation value acquisition unit 7 performs calculations at multiple points where y is different. The estimation value acquisition unit 7 uses the ratio of the intensity of the first color to the intensity of the second color (B / G) for multiple points where y is different to calculate the rate of change (rate of change in height) of the ratio of the intensity of the first color to the intensity of the second color (B / G). Similar calculations may be performed for other colors. In this example, the ratio of the luminance of the first color to the luminance of the second color is used as the feature. Note that other feature values ​​can also be used as estimation values ​​if their rate of change in height is correlated with aerosol concentration. Note that in the examples, it was confirmed that there is a correlation between the rate of change of the ratio of the intensity of the first color to the intensity of the second color and the aerosol concentration. From this, it can be inferred that there is a correlation between the rate of change of the intensity of each color and the aerosol concentration. It can also be inferred that there is a correlation between the rate of change of intensity regardless of color and the aerosol concentration.

[0026] The aerosol concentration estimation step (S103) is a step for estimating the aerosol concentration using the estimation value. The computer reads the estimation value from the memory unit. The control unit of the computer estimates the aerosol concentration using the estimation value based on instructions from the program.

[0027] FIG. 4 is a conceptual diagram illustrating a process for estimating aerosol concentrations using machine learning. The computer may be equipped with, for example, a machine learning engine, and may estimate the aerosol concentration using the estimation values ​​through machine learning. The calculated aerosol concentration may be stored in a storage unit. As shown in FIG. 4, the machine learning engine receives multiple inputs of estimation values ​​(e.g., the rate of change in the luminance of the first color) as feature values ​​and correct answer data (e.g., the aerosol concentration values ​​for the target area provided by the Ministry of the Environment) as training data. By repeatedly providing this training data, the learning model is trained to create a trained model. The trained model can be created by inputting estimation values ​​obtained from sample images into the learning model and recognizing a correlation between the obtained aerosol concentrations. By repeatedly providing this training data and test data, the accuracy of the trained model can be improved.

[0028] The aerosol concentration estimation unit 9 estimates the aerosol concentration using the estimation value. For example, the aerosol concentration estimation unit 9 has the trained model described above, and inputs the estimation value read from the storage unit into the trained model. The aerosol concentration estimation unit 9 can then obtain aerosol concentration-related information. The aerosol concentration-related information obtained in this manner may be stored in the storage unit as appropriate, or may be output. Examples of the aerosol concentration-related information may include an estimated value of the aerosol concentration in the target area, information such as an index indicating the aerosol concentration (e.g., aerosol concentration is extremely high, high, normal, or low), or image information such as an icon corresponding to the aerosol concentration. The computer may also use the estimated aerosol concentration to calculate other values ​​such as visibility, visibility, viewability, air pollution level, or air quality index. The meteorological condition determination process described below may be used as appropriate.

[0029] Aerosol concentration output process (S104) The estimated aerosol concentration may be output as appropriate. For example, estimated values ​​of aerosol concentration are output to a server from multiple locations. This allows information on aerosol concentration to be collected over a wide area. As a result, for example, if the aerosol concentration is high near a certain factory and low away from the factory, the factory is estimated to be the pollution source. Also, for example, if the computer is a user's mobile terminal (including a smartphone, mobile phone, and wristwatch), the aerosol concentration may be displayed on the display of the mobile terminal. The aerosol concentration may also be output to a server that can exchange information with the computer. In this way, information on aerosol concentration is output from multiple users' mobile terminals and aggregated on the server.

[0030] The computer may also determine the level of air pollution based on the estimated aerosol concentration. In this case, the control unit may use the estimated aerosol concentration to read a threshold value from the storage unit, and determine the level of air pollution by comparing the estimated aerosol concentration with the threshold value. The determined level of air pollution may be stored in the storage unit or output as appropriate. The computer may also output the aerosol concentration and the level of air pollution together with weather forecast information.

[0031] This specification also provides a program for causing a computer (or processor) to function as the above-described aerosol concentration estimation system, and a computer-readable information recording medium storing such a program. This invention can also be used as an application installed on a mobile device such as a smartphone or tablet. This allows a user to easily measure aerosol concentration using a mobile device with the application installed.

[0032] Weather condition determination process Next, the meteorological condition determination process shown in FIG. 1 will be described. The meteorological condition determination process is an optional process for determining whether or not the meteorological conditions are suitable for estimating the aerosol concentration (whether or not the shooting location matches the clear weather conditions). Another example of the meteorological condition determination process is a process for determining the meteorological conditions to be used in determining when the method for estimating the aerosol concentration differs depending on the meteorological conditions or when the parameters for estimating the aerosol concentration differ depending on the meteorological conditions. The meteorological condition determination process is not particularly limited as long as it can achieve the above. This process may also be performed in the advance preparation process.

[0033] An example of the weather condition determination process includes a weather condition input process (S121) and a condition determination process (S122). The weather condition input process (S121) is a process for inputting information required for determination into a computer, and the condition determination process (S122) is a process for determining weather conditions based on the input information. A specific example of the weather condition determination process is determining weather conditions using global solar radiation (GSI) and diffuse solar radiation (DSI). International Publication WO2018 / 100957 describes a daylighting system. The weather detection unit of this daylighting system includes a direct solar radiation meter that detects direct solar radiation and a pyranometer that detects global solar radiation. This system then calculates diffuse solar radiation by subtracting direct solar radiation from global solar radiation, and calculates the ratio of diffuse solar radiation to direct solar radiation. As described above, a system for determining global solar radiation and diffuse solar radiation using a direct pyranometer and a pyranometer is known. The computer may be connected to the direct pyranometer and the pyranometer, or may be configured to input observation data from the direct pyranometer and the pyranometer.

[0034] An example of a weather condition determination process is GSI > 550 W / m 2In this example, the global solar radiation (GSI) and diffuse solar radiation (DSI) are input to the system. The system stores the input global solar radiation (GSI) and diffuse solar radiation (DSI) in the memory as appropriate. Based on the program's instructions, the control unit reads the global solar radiation (GSI) and a threshold value (for example, 550 W / m 2 ) is read out. The control unit causes the calculation unit to perform a calculation to compare the global solar radiation (GSI) with a threshold value, and if the global solar radiation (GSI) is greater than the threshold value, proceed to the next step. On the other hand, the control unit causes the calculation unit to perform a calculation to compare the global solar radiation (GSI) with a threshold value, and if the global solar radiation (GSI) is less than the threshold value, determine that the weather conditions are not sunny. In the next step, the control unit reads the global solar radiation (GSI) and the diffuse solar radiation (DSI) from the memory unit and causes the calculation unit to perform a calculation process to calculate DSI / GSI based on instructions from the program. If this process is performed by hardware, a reverse circuit and a multiplication circuit may be used to calculate 1 / global solar radiation (GSI) from the global solar radiation (GSI), and the DSI / GSI may be calculated by multiplying the calculated value by the diffuse solar radiation (DSI). The control unit stores the calculated DSI / GSI in the memory unit. The control unit reads the DSI / GSI and a threshold value (e.g., 0.15) from the memory unit, and causes the calculation unit to perform a calculation to compare the DSI / GSI with the threshold value based on instructions from the program. If the DSI / GSI is smaller than the threshold value, it is determined that the weather conditions are fine. Information indicating that the weather conditions are fine may be stored in the memory unit as appropriate, or may be used for subsequent processing.

[0035] The calculation unit of the computer may read out the sample image from the storage unit, perform pattern matching processing based on instructions from the program, and classify the sample image into sunny areas, cloud areas, rain, buildings, trees, and others, and determine the weather conditions based on the calculation results obtained. In this case, for example, the weather conditions may be determined by reading out a threshold value that serves as a determination criterion from the storage unit and comparing it with the threshold value.

[0036] Weather conditions of the target area may be input from an input unit of the computer and stored in a memory unit. Then, the control unit of the computer may cause the calculation unit to perform appropriate calculation processing based on instructions from the program, and determine whether the input weather conditions (e.g., one or more of weather, temperature, humidity, and precipitation probability) are appropriate for estimating the aerosol concentration. Furthermore, a coefficient for calculating the aerosol concentration may be used appropriately based on the input weather conditions, and may be used in the step of estimating the aerosol concentration described below.

[0037] For example, if the meteorological conditions at the time of measurement are determined to be suitable for estimating the aerosol concentration in the meteorological condition determination step, the step of estimating the aerosol concentration described below may be performed.

[0038] FIG. 5 is a conceptual diagram illustrating an example of pattern matching. In the example of FIG. 5, areas including parts that are estimated to be buildings or trees (areas with obstacles) based on the RGB values ​​of each pixel in the sample image are removed from the area for estimating aerosol concentration, and the area of ​​the sample image above the area with obstacles is used to estimate aerosol concentration. In the example of FIG. 5, sample image 51 is classified into area 53, which is not used for estimating aerosol concentration due to obstacles such as buildings, area 55 above which there are no clouds, and area 57 above which there are clouds. In this example, the computer that performed the image analysis may determine the target area to be area 55 without clouds and area 57 above which there are clouds, or it may determine area 55 without clouds as the target area. [Example]

[0039] In this example, the amount of solar radiation at the time and location where the sample image was acquired was obtained from other satellite data, etc., and the conditions for clear skies with few clouds were determined. Image analysis techniques were used to extract features related to aerosol concentration from the sample image. A machine learning model was constructed using these features as input variables to calculate the aerosol concentration.

[0040] The aerosol concentration estimation method and system of the present invention were verified using image data taken in Fukuoka City over a two-year period from January 2021 to December 2022. Here, an example is shown in which the feature quantity is the rate of change in the RGB ratio. Concentrations of suspended particulate matter (SPM) with a particle size of 10 μm or less in aerosols were classified into three classes using Random Forest (RF), K-nearest Neighbor (K-NN), and Support Vector Machine (SVM). The classification threshold was set at 10 μg / m based on the annual and daily average values ​​of the environmental quality standards set by the World Health Organization. 3 , and 30 μg / m 3 The criteria for determining a clear sky were a ratio (DSI / GSI) of two variables, global solar radiation (GSI) and diffuse solar radiation (DSI), of 0.15 or less based on solar radiation data observed by the Himawari-8 satellite. Furthermore, to remove any clouds that remained visible in the image, only pixels satisfying (BR) / (B+R) > 0.2 were analyzed. For this target range, the vertical change rates of B / G, G / R, and B / R relative to the vertical relative pixel coordinates were calculated. Ground-based observation data from the Ministry of the Environment's Wide-Area Air Pollutant Monitoring System were used as the ground truth data for building the machine learning model. The dataset processed using the above method was randomly divided into training and testing sets in a 6:4 ratio. A model built using only the training data was applied to the testing data, and the accuracy rate was calculated. As a result, accuracy rates of 72%, 69%, and 72% were achieved for RF, K-NN, and SVM, respectively.

[0041] In this example, a sample image was obtained. Figure 6 is a photograph, instead of a drawing, showing an example of the sample image. The vertical change rate of three components (RGB, CIE XYZ, etc.) and their ratios were extracted from the sample image as feature quantities. Figure 7 is a graph, instead of a drawing, showing the relationship between the B to G luminance ratio (horizontal axis) and the height of a predetermined area of ​​the sample image (the position of the area excluding buildings, etc.). As shown in Figure 7, the change rate was approximately constant from the top to the bottom of the sky. Figure 8 is a graph, instead of a drawing, showing the relationship between the vertical change rate (horizontal axis) of the blue to green luminance ratio (B / G) and the concentration of suspended particulate matter (SPM) with a particle size of 10 μm or less (vertical axis). Figure 9 is a graph, instead of a drawing, showing the relationship between the vertical change rate (horizontal axis) of the green to red luminance ratio (G / R) and the concentration of suspended particulate matter (SPM) with a particle size of 10 μm or less (vertical axis). Figure 10 is a graph (in place of a drawing) showing the relationship between the vertical rate of change (horizontal axis) of the blue-to-red luminance ratio (B / R) and the concentration of suspended particulate matter (SPM) with a particle size of 10 μm or less (vertical axis). Figures 8 to 10 demonstrate a correlation between the vertical rate of change in the luminance of color components and the vertical rate of change in the luminance ratio of color components and aerosol concentration. Therefore, for example, it is possible to construct a machine learning model using these change rates as features (training data) and obtain a trained model. Building such a machine learning model eliminates the need to specify the sky range (e.g., upper, middle, or lower part of the sky). Furthermore, because the lower the aerosol concentration, the larger the change rate, it was found to be effective as a feature for determining aerosol concentration.

[0042] In the examples, Weather Report, operated by Weathernews Inc. (registered trademark), is an example of a service that is highly compatible with the present invention. This service is a weather forecast service based on users posting photos of the sky every day. If the present invention were used in such a user-participation weather forecast service, aerosol concentrations could be estimated from the photos posted by users, ultimately enabling every citizen to observe aerosols. This is expected to increase the general public's understanding and awareness of not only air pollution but also global atmospheric observation, and has great potential for social applications in addition to academic significance. [Industrial Applicability]

[0043] The present invention can be used in fields such as weather forecasting. [Explanation of symbols]

[0044] 1 System 3. Filming Department 5 Physical quantity acquisition section 7. Estimation value acquisition section 9. Aerosol concentration estimation section

Claims

1. 1. A method for estimating aerosol concentration in a region of interest by a computer, comprising: the target area includes a clear sky portion, The computer a physical quantity acquisition step of acquiring physical quantities from a sample image obtained by photographing the target area; an estimation value acquisition step of obtaining an estimation value for estimating an aerosol concentration using the physical quantity; an aerosol concentration estimation step of estimating an aerosol concentration using the estimation value; Including, the physical quantity includes an intensity of a first color component and an intensity of a second color component at each of a plurality of pixel locations of the sample image; the estimation value is a vertical rate of change of a ratio of an intensity of the first color component to an intensity of the second color component with respect to a relative pixel coordinate in the vertical direction in the sample image; Methods for estimating aerosol concentrations in the target area.

2. 2. A method for estimating aerosol concentration in a target area according to claim 1, comprising: A method for estimating aerosol concentration in a target area, wherein the sample image is an image taken using a mobile device.

3. 1. A computer-based aerosol concentration estimation system, comprising: an image capturing unit that captures a sample image of a target area including a clear sky portion; a physical quantity acquisition unit that acquires a predetermined physical quantity from the sample image; an estimation value acquisition unit that acquires an estimation value for estimating an aerosol concentration using the physical quantity; an aerosol concentration estimation unit that estimates an aerosol concentration using the estimation value; and the physical quantity includes an intensity of a first color component and an intensity of a second color component at each of a plurality of pixel locations of the sample image; the estimation value is a vertical rate of change of a ratio of an intensity of the first color component to an intensity of the second color component with respect to a relative pixel coordinate in the vertical direction in the sample image; Aerosol concentration estimation system.

4. A program for causing a computer to function as an aerosol concentration estimation system, The aerosol concentration estimation system comprises: an image capturing unit that captures a sample image of a target area including a clear sky portion; a physical quantity acquisition unit that acquires a predetermined physical quantity from the sample image; an estimation value acquisition unit that acquires an estimation value for estimating an aerosol concentration using the physical quantity; an aerosol concentration estimation unit that estimates an aerosol concentration using the estimation value; and the physical quantity includes an intensity of a first color component and an intensity of a second color component at each of a plurality of pixel locations of the sample image; the estimation value is a vertical rate of change of a ratio of an intensity of the first color component to an intensity of the second color component with respect to a relative pixel coordinate in the vertical direction in the sample image; program.

Images