Information Processing Apparatus, Information Processing Method, and Program

Abstract

To provide an information processing device, information processing method and program which can easily and accurately extract a water-filled region in a visible image by using section information that does not necessarily match actual information.SOLUTION: An information processing device comprises: a data acquisition unit 123 which acquires a visible image, section information corresponding to the visible image and a digital elevation model; a region extraction unit 124 which extracts corresponding regions in the visible image, the section information and the digital elevation model; and a water-filled region extraction unit 125 which extracts a water-filled region from the region of the section information and the region of the digital elevation model, plots the first point of the region of the section information on the region of the digital elevation model and extracts as the water-filled region the aggregate of pixels whose elevation difference with the elevation of the plotted point is within a prescribed range.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

Background Art

[0002] In recent years, machine learning technologies have been utilized in various fields, and their application to the agricultural field is also expected. When the farmland is paddy field, in order to make a production plan, it is necessary to grasp the flooded area and the bank area and estimate the cost corresponding to their areas. However, surveying for grasping the flooded area and the bank area requires a lot of time and labor, and it is not realistic to conduct manual surveying over a vast area.

[0003] On the other hand, the utilization of machine learning technologies requires the preparation of learning data for constructing a learning model. Generally, a large amount of data is required to construct a learning model, and the work of labeling the correct answers to this data takes a lot of time and labor. Patent Document 1 discloses a technique for generating a large amount of learning data at low cost for the learning of an apparatus for identifying a target object from an image or detecting a target object.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] In order to generate learning data for constructing a learning model for identifying a flooded area and a ridge area from an image, the utilization of publicly available data is considered. The Ministry of Agriculture, Forestry and Fisheries has published pen polygons, which are land parcel information of farmland created based on satellite images and the like. In addition, the Geospatial Information Authority of Japan has published a Digital Elevation Model (DEM), which is a set of elevation data that divides the ground surface into equally spaced squares and assigns the elevation value of the center point to each square.

[0006] The pen polygon is very useful as basic information regarding farmland. On the other hand, since the pen polygon determines the presence or absence of farmland from satellite images and the like, it does not always match the actual information depending on the imaging time of the image and the like. In addition, the satellite image used for creating the pen polygon is taken, processed, and manufactured from visible light reflected from the ground surface in space. Although the influence of terrain undulation and elevation is corrected, the deviation and distortion due to errors in the position and attitude of the satellite at the time of imaging, errors in the terrain data used for correction, etc. cannot be completely corrected, so there may be a deviation between the corresponding background image and the pen polygon. Furthermore, many polygons include a flooded area and a ridge area, but there are also some polygons that do not include a ridge area.

[0007] Therefore, an object of the present invention is to provide an information processing apparatus, an information processing method, and a program that can easily and accurately extract a flooded area in a visible image by using section information that does not always match the actual information.

Means for Solving the Problems

[0008] An information processing apparatus according to an aspect of the present invention includes: a data acquisition unit that acquires a visible image, section information corresponding to the visible image, and a digital elevation model; a region extraction unit that extracts corresponding regions in the visible image, the section information, and the digital elevation model; and a flooded area extraction unit that extracts a flooded area from the region of the section information and the region of the digital elevation model, the flooded area extraction unit plots a first point of the region of the section information on the region of the digital elevation model, and extracts a set of pixels whose elevation difference from the elevation of the plotted point is within a predetermined range as the flooded area.

[0009] According to this aspect, by using the section information corresponding to the visible image and the information of the digital elevation model, the flooded area in the visible image can be easily extracted.

[0010] The information processing apparatus may further include a ridge area extraction unit that classifies pixels belonging to the extracted area, which has color information obtained from a visible image and elevation information obtained from a digital elevation model, using a classification method, and extracts a ridge area from the classified pixel group. According to this aspect, not only the flooded area in the visible image but also the ridge area in the visible image can be easily extracted.

[0011] In the information processing apparatus, the ridge area extraction unit may expand the area of the section information extracted by the area extraction unit according to a predetermined condition, and classify the pixels belonging to the expanded area of the section information using a classification method. According to this aspect, by using the pixels belonging to the expanded area, the ridge area can be effectively extracted even for some areas in the section information that do not include the ridge area.

[0012] The information processing apparatus may further include a non-paddy field removal unit that performs histogram analysis on the visible image corresponding to the flooded area extracted by the flooded area extraction unit, and removes the area determined to be a non-paddy field from the flooded area based on the characteristics of the analysis result. According to this aspect, even when using section information that does not necessarily match the actual information, the non-paddy field can be reliably removed to match the actual information.

[0013] In the information processing apparatus, the characteristics may include the sharpness and smoothness of the curve. According to this aspect, by using the sharpness and smoothness of the curve in the histogram analysis, the non-paddy field area can be easily determined.

[0014] The information processing apparatus may further include a learning data generation unit that generates learning data with labels indicating flooded areas and ridge areas in a visible image. According to this aspect, by using a visible image, sectional information, and a digital elevation model, it is possible to easily generate learning data with labels indicating flooded areas and ridge areas in the visible image.

[0015] In the information processing apparatus, the first point may be the centroid. According to this aspect, for extracting the flooded area, a point that can be surely distinguished from ridge areas or the like that may exist around the flooded area can be used.

[0016] A method according to another aspect of the present invention includes: a step of acquiring a visible image, sectional information corresponding to the visible image, and a digital elevation model; a step of extracting corresponding regions in the visible image, the sectional information, and the digital elevation model; and a step of extracting a flooded area from the region of the sectional information and the region of the digital elevation model, the step of plotting a first point of the region of the sectional information on the region of the digital elevation model and extracting, as the flooded area, a set of pixels whose elevation difference from the elevation of the plotted point is within a predetermined range.

[0017] A program according to another aspect of the present invention causes one or more computers to execute: a process of acquiring a visible image, sectional information corresponding to the visible image, and a digital elevation model; a process of extracting corresponding regions in the visible image, the sectional information, and the digital elevation model; and a process of extracting a flooded area from the region of the sectional information and the region of the digital elevation model, the process of plotting a first point of the region of the sectional information on the region of the digital elevation model and extracting, as the flooded area, a set of pixels whose elevation difference from the elevation of the plotted point is within a predetermined range.

Advantages of the Invention

[0018] According to the present invention, it is possible to provide an information processing apparatus, an information processing method, and a program that can easily and accurately extract a flooded area in a visible image by using sectional information that does not necessarily match the actual information.

Brief Description of the Drawings

[0019]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Embodiments for Carrying Out the Invention

[0020] Embodiments of the present invention will be described with reference to the accompanying drawings. Note that the following embodiments are for facilitating the understanding of the present invention and are not for limiting and interpreting the present invention. Further, the present invention can be variously modified without departing from its gist. Furthermore, those skilled in the art can adopt embodiments in which each element described below is replaced with an equivalent one, and such embodiments are also included in the scope of the present invention.

[0021] (System Configuration)

[0022] The outline of the present invention will be described with reference to FIG. 1. FIG. 1 is a schematic diagram for explaining the processing of an information processing apparatus according to an embodiment of the present invention.

[0023] The information processing apparatus 100 is a device that generates learning data for constructing a learning model for identifying a flooded area and a ridge area from an image. As shown in FIG. 1, the information processing apparatus 100 generates desired learning data using a visible image I, a pen polygon P, and a digital elevation model M. In one embodiment, the information processing apparatus 100 generates learning data including a visible image and a mask image composed of a flooded area and a ridge area in the visible image.

[0024] The pen polygon P is plot information of farmland created based on a satellite image or the like. The digital elevation model E is a set of elevation data obtained by dividing the ground surface into equally spaced squares and assigning an elevation value of the center point to each square.

[0025] The information processing apparatus 100 acquires the visible image I of the area where the learning data is to be generated, the pen polygon P corresponding to this visible image, and the digital elevation model M, and uses the color information of the visible image I, the plot information of the farmland of the pen polygon P, and the elevation information of the digital elevation model M to generate high-precision learning data for constructing a learning model for identifying a flooded area and a ridge area from the image.

[0026] (Functional Configuration)

[0027] FIG. 2 is a block diagram of an information processing apparatus according to an embodiment of the present invention. In FIG. 2, a single information processing apparatus 100 is assumed and only the necessary functional configuration is shown, but the information processing apparatus 100 can also be configured as part of a multi-functional distributed system by a plurality of computer systems.

[0028] The information processing apparatus 100 includes an input unit 110, a control unit 120, a storage unit 130, and a communication unit 140.

[0029] The input unit 110 is configured to receive an operation from an administrator of the information processing apparatus 100 and can be realized by a keyboard, a mouse, a touch panel, or the like.

[0030] The control unit 120 includes an arithmetic processing unit 121 such as a CPU or MPU and a memory 122 such as a RAM. The arithmetic processing unit 121 operates various functional units by executing a program recorded in the storage unit 130 based on various inputs. This program may be stored in a recording medium such as a CD-ROM, distributed via a network, or installed on a computer. The memory 122 is for temporarily storing various data necessary for operations such as arithmetic during the execution of processing in various programs.

[0031] The storage unit 130 is composed of a storage device such as a hard disk, and records various programs necessary for the execution of processing in the control unit 120, data necessary for the execution of various programs, and the like. In the present embodiment, it is desirable that the storage unit 130 has an image storage unit 131, a pen polygon storage unit 132, a digital elevation model (DEM) storage unit 133, and a learning data storage unit 134.

[0032] Visible images including paddy fields are stored in the image storage unit 131.

[0033] Parcel information of farmland created based on satellite images and the like is stored in the pen polygon storage unit 132. In one embodiment, the pen polygon published by the Ministry of Agriculture, Forestry and Fisheries is used as the parcel information of farmland.

[0034] A set of elevation data in which the ground surface is divided into equally spaced squares and each square has an elevation value at the center point is stored in the DEM storage unit 133. In one embodiment, the digital elevation model in the basic map information published by the Geospatial Information Authority of Japan is used as the set of elevation data.

[0035] Learning data with labels indicating the flooded areas and ridge areas in the visible image is stored in the learning data storage unit 134. In one embodiment, it is desirable that the learning data storage unit 134 stores learning data including a visible image and a mask image composed of the flooded areas and ridge areas in the visible image.

[0036] The communication unit 140 is configured to connect the information processing apparatus 100 to a network. For example, the communication unit 140 can be realized from a LAN card, an analog modem, an ISDN modem, etc., and an interface for connecting these to a processing unit via a transmission path such as a system bus.

[0037] Furthermore, as shown in FIG. 2, the arithmetic processing unit 121 includes, as functional units, a data acquisition unit 123, a region extraction unit 124, a flooded region extraction unit 125, a non-paddy field removal unit 126, a ridge region extraction unit 127, a learning data generation unit 128, and an output unit 129.

[0038] The data acquisition unit 123 acquires a visible image of the area where learning data is generated, a pen polygon corresponding to this visible image, and a DEM. In the present embodiment, the data acquisition unit 123 acquires data from the image storage unit 131, the pen polygon storage unit 132, and the DEM storage unit 133, respectively.

[0039] The region extraction unit 124 extracts regions of a predetermined shape corresponding to the acquired visible image, pen polygon, and DEM. In the present embodiment, the region extraction unit 124 overlays the visible image, pen polygon, and DEM, selects the region of 1 in the pen polygon, and extracts a region that sufficiently contains the selected region from the visible image and DEM. FIG. 3 shows the corresponding regions in the visible image, pen polygon, and DEM extracted by the region extraction unit 124. FIG. 3(a) illustrates the pen polygon, FIG. 3(b) illustrates the visible image, and FIG. 3(c) illustrates the DEM.

[0040] The flooded region extraction unit 125 extracts a flooded region based on elevation information from the extracted pen polygon region and DEM region. In the present embodiment, the flooded region extraction unit 125 plots the centroid point of the pen polygon region on the DEM region, and extracts a set of pixels whose elevation difference from the elevation of the plotted point is within a predetermined range as the flooded region. Specifically, the flooded region extraction unit 125 first determines the centroid point pixel x ij of the 16 pixels around it x i+1j-1 、x i+1j 、x i+1j+1 、x ij-1 、xij+1 , x i-1j-1 , x i-1j , x i-1j+1 , x i+4j-4 , x i+4j , x i+4j+4 , x ij-4 , x ij+4 , x i-4j-4 , x i-4j , x i-4j+4 Among them, pixels with a height difference from the center of gravity point pixel x ij within ±0.1998 m are extracted as the flooded area. The flooded area extraction unit 125 further extracts adjacent pixels with a height difference within ±0.1998 m from the pixels already extracted as the flooded area as the flooded area, and repeats this process to extract the flooded area.

[0041] Note that the range of the height difference used for extracting the flooded area is for illustration, and the flooded area can be extracted using other appropriate ranges of height differences. Furthermore, the point (the first point) used for extracting the flooded area is not limited to the center of gravity point, and any inner point distinguishable from the center point or the ridge area that may exist around the flooded area can be used. The flooded area extraction unit 125 may perform, for example, morphological transformation to remove noise such as water channels. Note that the method for removing noise is not limited to morphological transformation, and any other arbitrary method can be used.

[0042] The non-paddy field removal unit 126 removes the non-paddy field area based on the color information of the visible image. In the present embodiment, the non-paddy field removal unit 126 performs histogram analysis on the visible image corresponding to the flooded area extracted by the flooded area extraction unit 125, and removes the area determined as the non-paddy field from the flooded area based on the characteristics of the analysis result.

[0043] As shown in the upper part of FIG. 4, the histogram of the paddy field has a larger kurtosis and shows a smoother curve compared to the histogram of the non-paddy field in the lower part of the same figure. The non-paddy field removal unit 126 determines the non-paddy field area based on the kurtosis and smoothness of the curve, that is, the number of times the curve passes through 0 on the Y-axis when the curve is differentiated.

[0044] The ridge area extraction unit 127 classifies the area including the flooded area and the ridge area based on the color information of the visible image and the elevation information of the DEM, and extracts the ridge area. In the present embodiment, the ridge area extraction unit 127 expands the area of the pen polygon extracted by the area extraction unit 124 according to a predetermined condition, and classifies the pixels belonging to the expanded area of the pen polygon having the color information obtained from the visible image and the elevation information obtained from the DEM using K-means. For example, the ridge area extraction unit 127 may perform a morphological transformation to expand the area of the pen polygon. In the present embodiment, a morphological transformation is performed using a 30×30 size kernel, but any other kernel size can be used.

[0045] Next, the ridge area extraction unit 127 estimates which of the flooded area, the ridge area, and other areas the classified pixel group belongs to, and extracts the ridge area. For example, the ridge area extraction unit 127 can estimate which of the flooded area, the ridge area, and other areas the classified pixel group belongs to by using the area and color distribution of the classified pixel group.

[0046] Each point in the scatter diagram shown in FIG. 5 means a pixel, the color of the point represents the elevation, and the position of the point represents the color information.

[0047] In this way, by expanding the area of the pen polygon, it is possible to effectively extract the ridge area even for some polygons that do not include the ridge area in the pen polygon. Note that the classification method used is not limited to K-means, and other classification methods can be used. Also, in the present embodiment, the color information in the Lab color space is used, but the color information used for classification is not limited to the Lab color space, and other color spaces can be used.

[0048] The learning data generation unit 128 generates learning data with labels indicating the flooded areas and ridge areas in the visible image. In the present embodiment, the learning data generation unit 128 generates a mask image composed of the flooded areas extracted by the flooded area extraction unit 125 and the ridge areas extracted by the ridge area extraction unit 127, and stores the generated mask image and the corresponding visible image in the learning data storage unit 134.

[0049] The output unit 129 outputs the learning data stored in the learning data storage unit 134.

[0050] (Learning data generation process)

[0051] Referring to FIG. 6, the learning data generation process according to the embodiment of the present invention will be described in detail. In the present embodiment, it is assumed that each data is stored in the image storage unit 131, the pen polygon storage unit 132, and the DEM storage unit 133 under the management of the administrator of the information processing apparatus 100 before the learning data generation process described in FIG. 6 is performed.

[0052] In step S601, the data acquisition unit 123 of the information processing apparatus 100 acquires a visible image of the area for generating learning data, a pen polygon corresponding to this visible image, and a DEM. In the present embodiment, the data acquisition unit 123 acquires data from the image storage unit 131, the pen polygon storage unit 132, and the DEM storage unit 133, respectively.

[0053] In step S602, the area extraction unit 124 of the information processing apparatus 100 extracts areas of a corresponding predetermined shape in the acquired visible image, pen polygon, and DEM. In the present embodiment, the area extraction unit 124 overlays the visible image, the pen polygon, and the DEM, selects an area of 1 in the pen polygon, and extracts an area that sufficiently contains the selected area from the visible image and the DEM. FIG. 3 shows the corresponding areas in the visible image, pen polygon, and DEM extracted by the area extraction unit 124. FIG. 3(a) is an example of a pen polygon, FIG. 3(b) is an example of a visible image, and FIG. 3(c) is an example of a DEM.

[0054] In step S603, the flooding area extraction unit 125 of the information processing apparatus 100 extracts a flooding area from the area of the extracted pen polygon and the area of the DEM based on elevation information. Details will be described later with reference to FIG. 7.

[0055] In step S604, the non-paddy field removal unit 126 of the information processing apparatus 100 removes the area of non-paddy fields based on the color information of the visible image. In the present embodiment, the non-paddy field removal unit 126 performs histogram analysis on the visible image corresponding to the flooding area extracted by the flooding area extraction unit 125, and removes from the flooding area the area determined to be non-paddy fields based on the characteristics of the analysis result.

[0056] As shown in the upper part of FIG. 4, the histogram of paddy fields has a greater sharpness and shows a smoother curve compared to the histogram of non-paddy fields in the lower part of the same figure. The non-paddy field removal unit 126 determines the non-paddy field area based on the sharpness and smoothness of the curve, that is, the number of times the curve passes through 0 on the Y-axis when the curve is differentiated.

[0057] In step S605, the ridge area extraction unit 127 of the information processing apparatus 100 classifies the area including the flooding area and the ridge area based on the color information of the visible image and the elevation information of the DEM, and extracts the ridge area. In the present embodiment, the ridge area extraction unit 127 performs morphological transformation on the area of the pen polygon extracted by the area extraction unit 124 to expand it, and classifies the pixels belonging to the expanded area of the pen polygon, which has color information obtained from the visible image and elevation information obtained from the DEM, using K-means. In the present embodiment, morphological transformation is performed using a 30×30 size kernel, but any other arbitrary kernel size can be used.

[0058] Each point in the scatter diagram shown in FIG. 5 means a pixel. The color of the point represents the elevation, and the position of the point represents the color information. When only the flooding area and the ridge area are included in the expanded area, the ridge area extraction unit 127 classifies the pixels into two classes. On the other hand, when other areas such as greenhouses and roads are included in the expanded area in addition to the flooding area and the ridge area, the ridge area extraction unit 127 classifies the pixels into three classes.

[0059] Next, the ridge region extraction unit 127 estimates which of the waterlogged region, the ridge region, and other regions the classified pixel group belongs to, and extracts the ridge region. For example, the ridge region extraction unit 127 can estimate which of the waterlogged region, the ridge region, and other regions the classified pixel group belongs to by using the area and color distribution of the classified pixel group.

[0060] In this way, by expanding the area of the pen polygon, it is possible to effectively extract the ridge region even for some polygons that do not include the ridge region in the pen polygon. Note that the method used for classification is not limited to K-means, and other classification methods can be used. Also, in this embodiment, the color information in the Lab color space is used, but the color information used for classification is not limited to the Lab color space, and other color spaces can be used.

[0061] In step S606, the learning data generation unit 128 of the information processing apparatus 100 generates learning data with labels indicating the waterlogged region and the ridge region in the visible image. In this embodiment, the learning data generation unit 128 generates a mask image composed of the waterlogged region extracted by the waterlogged region extraction unit 125 and the ridge region extracted by the ridge region extraction unit 127, and stores the generated mask image and the corresponding visible image in the learning data storage unit 134.

[0062] As described above, according to this embodiment, the information processing apparatus 100 can generate learning data for constructing a learning model for identifying the waterlogged region and the ridge region from an image by using a pen polygon. By utilizing the existing pen polygon, it is possible to reduce the time and labor required for preparing the learning data from data collection to labeling. In particular, it is suitable when the area where the learning data is generated is flat and rectangular paddy fields are arranged side by side.

[0063] (Waterlogged Region Extraction Process)

[0064] Referring to FIG. 7, the waterlogging area extraction process in step S603 will be described in detail. FIG. 7 is a flowchart showing the waterlogging area extraction process according to an embodiment of the present invention.

[0065] In step S701, the waterlogging area extraction unit 125 obtains the centroid of the area of the pen polygon. Next, in step S702, the waterlogging area extraction unit 125 plots the centroid point of the area of the pen polygon on the area of the DEM, and in step S703, pixels whose elevation difference from the plotted point is within a predetermined range are filled.

[0066] Specifically, the waterlogging area extraction unit 125 first determines the centroid pixel x ij of the 16 pixels x i+1j-1 , x i+1j , x i+1j+1 , x ij-1 , x ij+1 , x i-1j-1 , x i-1j , x i-1j+1 , x i+4j-4 , x i+4j , x i+4j+4 , x ij-4 , x ij+4 , x i-4j-4 , x i-4j , x i-4j+4 around it, and extracts pixels whose elevation difference from the centroid pixel x ij is within ±0.1998 m as the waterlogging area. The waterlogging area extraction unit 125 further extracts adjacent pixels whose elevation difference from the pixels already extracted as the waterlogging area is within ±0.1998 m as the waterlogging area, and extracts the waterlogging area by repeating this process.

[0067] Note that the range of the elevation difference used for extracting the waterlogging area is for illustration purposes, and the waterlogging area can be extracted using other appropriate ranges of elevation differences. Furthermore, the point used for extracting the waterlogging area is not limited to the centroid point, and any inner point distinguishable from the center point, the ridge area that may exist around the waterlogging area, etc. can be used.

[0068] In step S704, the padding area extraction unit 125 performs morphological transformation to remove noise such as water channels and extracts the padding area. Note that the method for removing noise is not limited to morphological transformation, and any other arbitrary method can be used.

Explanation of Signs

[0069] 100… Information processing apparatus, 110… Input unit, 120… Control unit, 121… Arithmetic processing unit, 122… Memory, 123… Data acquisition unit, 124… Region extraction unit, 125… Padding area extraction unit, 126… Non-paddy field removal unit, 127… Ridge area extraction unit, 128… Learning data generation unit, 129… Output unit, 130… Storage unit, 131… Image storage unit, 132… Pen polygon storage unit, 133… DEM storage unit, 134… Learning data storage unit, 140… Communication unit, I… Visible image, P… Pen polygon, M… Elevation model

Claims

1. A data acquisition unit that acquires a visible image, section information corresponding to the visible image, and a digital elevation model; A region extraction unit that extracts corresponding regions in the visible image, the section information, and the digital elevation model; A flooding region extraction unit that extracts a flooding region from the region of the section information and the region of the digital elevation model, plots a first point of the region of the section information on the region of the digital elevation model, and extracts a set of pixels whose elevation difference from the plotted point is within a predetermined range as the flooding region; A non-paddy field removal unit that performs histogram analysis on the visible image corresponding to the flooding region extracted by the flooding region extraction unit and removes a region determined to be a non-paddy field from the flooding region based on the characteristics of the analysis result; A learning data generation unit that generates labeled learning data indicating the flooding region and the ridge region in the visible image; An information processing apparatus comprising the above.

2. The information processing apparatus according to claim 1, further comprising a ridge region extraction unit that classifies pixels belonging to the extracted region having color information obtained from the visible image and elevation information obtained from the digital elevation model using a classification method, and extracts a ridge region from the classified pixel group.

3. The ridge region extraction unit according to claim 2 expands the region of the section information extracted by the region extraction unit according to a predetermined condition, and classifies the pixels belonging to the expanded region of the section information using a classification method. The information processing apparatus described.

4. The information processing apparatus according to claim 1, wherein the characteristics include the sharpness and smoothness of the curve.

5. The information processing apparatus according to any one of claims 1 to 4, wherein the first point is a centroid.

6. A step of acquiring a visible image, section information corresponding to the visible image, and a digital elevation model; A step of extracting corresponding regions in the visible image, the section information, and the digital elevation model; A step of extracting a flooding region from the region of the section information and the region of the digital elevation model, plotting a first point of the region of the section information on the region of the digital elevation model, and extracting a set of pixels whose elevation difference from the plotted point is within a predetermined range as the flooding region; A step of performing histogram analysis on the visible image corresponding to the extracted flooding region and removing a region determined to be a non-paddy field from the flooding region based on the characteristics of the analysis result; A step of generating labeled learning data indicating the flooding region and the ridge region in the visible image; A method including the above.

7. On one or more computers, a process of acquiring a visible image, sectional information corresponding to the visible image, and a digital elevation model; a process of extracting corresponding regions in the visible image, the sectional information, and the digital elevation model; a process of extracting a flooded area from the region of the sectional information and the region of the digital elevation model, the process of plotting a first point of the region of the sectional information on the region of the digital elevation model and extracting, as the flooded area, a set of pixels whose elevation difference from the elevation of the plotted point is within a predetermined range; a process of performing histogram analysis on the visible image corresponding to the extracted flooded area and removing, from the flooded area, a region determined to be non-paddy field based on the characteristics of the analysis result; a process of generating labeled learning data indicating the flooded area and the ridge area in the visible image A program for execution.

Images