Ground surface composite image creation method, ground surface composite image creation system, and ground surface composite image creation program

Abstract

[Problem] To provide a ground surface composite image creation method, a ground surface composite image creation system, and a ground surface composite image creation program that can make it possible to relatively easily understand a use state of a ground surface, such as vegetation, by using height information. [Solution] The present invention acquires an ortho-image of a target area and a digital surface model, and creates, on the basis of the digital surface model, a ground height image pertaining to the height of an object on the ground. Each pixel having a ground height equal to or greater than a prescribed threshold is extracted from the created ground height image. A grayscale image is created in which a grayscale value of each of the extracted pixels is set on the basis of the corresponding ground height or set to a constant value. The created grayscale image is superimposed on the ortho-image to create a composite image.

Description

Earth's surface composite image creation method, earth's surface composite image creation system, and earth's surface composite image creation program

[0001] The present invention relates to a method for creating a composite earth surface image, a system for creating a composite earth surface image, and a program for creating a composite earth surface image.

[0002] In recent years, as symbolized by the increasing number of designated Globally Important Agricultural Heritage Systems and Japanese Agricultural Heritage Systems, attention has been paid to the functions of rural areas, such as culture, landscape, and biodiversity. For example, residential forests are one of the main elements that form the rural landscape, and also play a role in the surrounding environment, such as preserving biodiversity. For this reason, there is a need to quantitatively evaluate and understand the current state of rural areas, including residential forests.

[0003] Conventionally, in order to understand the use of the ground surface by vegetation and the topography, it is common to use manned aircraft or unmanned aerial vehicles (UAVs or drones) to continuously photograph a target area from different viewpoints in the sky, and then process the photographed images using techniques such as Structure from Motion (SfM), or measure the target area from the sky using a Light Detection and Ranging (LiDAR) to obtain digital image data and measurement data, thereby obtaining ground surface images (orthoimages) and digital surface models (DSMs).

[0004] In addition, methods have been proposed for detecting plants on the ground surface and understanding the utilization status of vegetation on the ground surface using the orthoimages and DSMs obtained in this way. For example, a method has been proposed in which a DSM image of a target area and a DSM image of a target object such as a plant are obtained, and then the target object is detected from the DSM using artificial intelligence (see, for example, Patent Document 1); a method has been proposed in which a slope map and a curvature map are obtained based on the DSM and then combined with an orthoimage to generate a composite image that is more three-dimensional and has higher visibility than an orthoimage (see, for example, Patent Document 2); a method has been proposed in which a distribution of a second type of plant is extracted from a DSM of a field having a first type of plant and a second type of plant based on the difference in plant height (see, for example, Patent Document 3); and a method has been proposed in which, for a target area including a forest, feature quantities are extracted from laser measurement data and aerial photographic images, and pixel values ​​of the color image at each pixel are determined based on the value of each feature quantity corresponding to each pixel, thereby visualizing differences in forest types in a color image (see, for example, Patent Document 4).

[0005] Japanese Patent Application Laid-Open No. 2018-92628 Japanese Patent No. 6700519 Japanese Patent Application Laid-Open No. 2022-66907 Japanese Patent Application Laid-Open No. 2015-141118

[0006] However, the method described in Patent Document 1 uses only DSM image information and does not use orthoimages, resulting in poor target object detection efficiency. The method described in Patent Document 2 generates a three-dimensional, highly visible composite image that is effective for understanding topography, but has the problem of difficulty in understanding the use of the ground surface, such as vegetation. The method described in Patent Document 3 is designed for fields with only two types of plants, making it difficult to understand the use of target areas other than fields, such as rural areas. The method described in Patent Document 4 allows for various settings for extracted features, but has the problem of complex procedures and formulas for determining pixel values ​​of color images.

[0007] The present invention has been made in light of these problems, and aims to provide a method for creating a ground surface composite image, a ground surface composite image creation system, and a ground surface composite image creation program that can relatively easily grasp the utilization status of the ground surface, such as vegetation, using height information.

[0008] In order to achieve the above-mentioned object, the method for creating a composite image of the earth's surface according to the present invention is characterized by comprising: an image acquisition step of acquiring an orthoimage of a target area and a digital surface model (DSM; Digital Surface Model); a ground height image creation step of creating a ground height image relating to the height of objects on the ground based on the digital surface model acquired in the image acquisition step; a pixel extraction step of extracting pixels having a ground height equal to or greater than a predetermined threshold from the ground height image created in the ground height image creation step; a grayscale image creation step of creating a grayscale image in which the grayscale value of each pixel extracted in the pixel extraction step is set based on the corresponding ground height or set to a constant value; and a composite image creation step of creating a composite image by superimposing the grayscale image created in the grayscale image creation step on the orthoimage.

[0009] The composite image creation system for the earth's surface according to the present invention is characterized by having an image acquisition means for acquiring an orthoimage and a digital surface model of a target area, a ground height image creation means for creating a ground height image relating to the height of objects on the ground based on the digital surface model acquired by the image acquisition means, a pixel extraction means for extracting pixels having a ground height equal to or greater than a predetermined threshold from the ground height image created by the ground height image creation means, a grayscale image creation means for creating a grayscale image in which the grayscale value of each pixel extracted by the pixel extraction means is set based on the corresponding ground height or set to a constant value, and a composite image creation means for creating a composite image by superimposing the grayscale image created by the grayscale image creation means on the orthoimage.

[0010] The earth's surface composite image creation program of the present invention is characterized by causing a computer to function as an image acquisition means for acquiring an orthoimage and a digital surface model of a target area, a ground height image creation means for creating a ground height image relating to the height of objects on the ground based on the digital surface model acquired by the image acquisition means, a pixel extraction means for extracting pixels having a ground height equal to or greater than a predetermined threshold from the ground height image created by the ground height image creation means, a grayscale image creation means for creating a grayscale image in which the grayscale value of each pixel extracted by the pixel extraction means is set based on the corresponding ground height or set to a constant value, and a composite image creation means for superimposing the grayscale image created by the grayscale image creation means on the orthoimage to create a composite image.

[0011] The method, system, and program for generating a composite image of the earth's surface according to the present invention can easily generate a composite image that reflects height information by setting the shading of an orthoimage of a target area based on ground height, which represents the height of objects on the ground. Furthermore, by using this composite image, it is possible to determine not only the appearance from above but also differences in height, making it relatively easy to grasp the utilization status of the ground surface, such as vegetation. Furthermore, even if the resolution of the orthoimage is low, the accuracy of extracting vegetation, etc. can be improved. Ground height refers to the height of objects on the ground, such as buildings and trees, from the ground.

[0012] The earth's surface composite image creation method, earth's surface composite image creation system, and earth's surface composite image creation program according to the present invention create a composite image that reflects height information in the shading of an orthoimage, and therefore the composite image can be obtained as a three-channel image (red, green, and blue). For example, general image processing programs using machine learning and the like are developed on the premise of processing three-channel RGB images, and images captured by a standard camera mounted on a UAV are also three-channel images. Therefore, the earth's surface composite image creation method, earth's surface composite image creation system, and earth's surface composite image creation program according to the present invention are highly versatile, as they do not require converting three-channel images into multi-channel images of four or more channels or modifying image processing programs for use with four or more channels.

[0013] The composite image created by the earth's surface composite image creation method, earth's surface composite image creation system, and earth's surface composite image creation program according to the present invention may be used as training data for machine learning, together with the earth's surface utilization status, such as vegetation, determined from the composite image, to build a system for understanding the utilization status of the earth's surface, such as vegetation. Furthermore, the system thus built may be used to understand the utilization status of the earth's surface, such as vegetation, from the newly created composite image. In the earth's surface composite image creation method, earth's surface composite image creation system, and earth's surface composite image creation program according to the present invention, it is preferable that each pixel of the ground height image or grayscale image corresponds 1:1 to each pixel of the orthoimage, but this does not necessarily have to be the case as long as there is no positional deviation between the images.

[0014] In the ground surface composite image creation method according to the present invention, the grayscale image creation step may set each grayscale value to a constant value so that each pixel extracted in the pixel extraction step becomes white. In the ground surface composite image creation system and ground surface composite image creation program according to the present invention, the grayscale image creation means may set each grayscale value to a constant value so that each pixel extracted by the pixel extraction means becomes white. In this case, parts of the created composite image that are above ground level at a height equal to or greater than a predetermined threshold become white, making it easy to extract parts that are higher than the threshold. Therefore, for example, when a target area is a rural area, it is very difficult to distinguish between vegetation such as hedges that are lower than houses and residential woods that have grown higher than houses in an image taken from above, but in a composite image obtained by setting the predetermined threshold higher than the height of houses, it is easy to extract only residential woods.

[0015] In the earth's surface composite image creation method according to the present invention, the grayscale image creation step may set the grayscale value of each pixel extracted in the pixel extraction step to a value obtained by converting the corresponding ground height using a predetermined monotonically increasing or monotonically decreasing function. In the earth's surface composite image creation system and earth's surface composite image creation program according to the present invention, the grayscale image creation means may set the grayscale value of each pixel extracted by the pixel extraction means to a value obtained by converting the corresponding ground height using a predetermined monotonically increasing or monotonically decreasing function. In this case, the ground height image is converted in tone according to the ground height of each pixel and reflected in the grayscale of the composite image, making it easy to grasp height information from the created composite image. Therefore, the composite image makes it easy to extract differences in vegetation, etc., that are difficult to grasp from the appearance alone from above.

[0016] In the composite earth surface image creation method according to the present invention, the ground height image creation step may create a digital elevation model (DEM) based on the digital surface model, and create the ground height image from the difference between the digital surface model and the created digital elevation model. Alternatively, the image acquisition step may also acquire a digital elevation model of the target area, and the ground height image creation step may create the ground height image from the difference between the digital surface model and the digital elevation model acquired in the image acquisition step. In the composite earth surface image creation system and composite earth surface image creation program according to the present invention, the ground height image creation means may create a digital elevation model based on the digital surface model, and create the ground height image from the difference between the digital surface model and the created digital elevation model. Alternatively, the image acquisition means may also acquire a digital elevation model of the target area, and the ground height image creation means may create the ground height image from the difference between the digital surface model and the digital elevation model acquired by the image acquisition means.

[0017] When using this digital elevation model, when creating a digital elevation model based on a digital surface model, for example, the digital elevation model can be created by filtering the digital surface model to remove the heights of ground objects such as buildings and trees. Furthermore, when acquiring a digital elevation model in the image acquisition step or image acquisition means, for example, a digital elevation model provided by the Geospatial Information Authority of Japan or the like can be used. Furthermore, by using the digital elevation model, a ground height image can be easily created from the difference between a digital surface model (DSM) and a digital elevation model (DEM). The created ground height image is, for example, an image based on a digital height model (DHM; Digital Height Model).

[0018] In the composite earth surface image creation method according to the present invention, the image acquisition step may acquire three-dimensional point cloud data from image data of the target area photographed from the air, and create the orthoimage and the digital surface model from the acquired three-dimensional point cloud data. In the composite earth surface image creation system and composite earth surface image creation program according to the present invention, the image acquisition means may acquire three-dimensional point cloud data from image data of the target area photographed from the air, and create the orthoimage and the digital surface model from the acquired three-dimensional point cloud data. In this case, the three-dimensional point cloud data can be easily acquired by processing image data photographed continuously from different viewpoints in the air by a UAV using Structure from Motion (SfM) or the like.

[0019] In the earth's surface composite image creation method according to the present invention, the image acquisition step may create the orthoimage and the digital surface model based on aerial photographs of the target area taken from above and measurement data obtained by measuring the target area from above with a laser ranging apparatus (LiDAR). In the earth's surface composite image creation system and earth's surface composite image creation program according to the present invention, the image acquisition means may create the orthoimage and the digital surface model based on aerial photographs of the target area taken from above and measurement data obtained by measuring the target area from above with a laser ranging apparatus (LiDAR). In this case, the orthoimage and the digital surface model can be created using aerial photographs and measurement data obtained by a laser ranging apparatus (LiDAR) without using a UAV.

[0020] The earth's surface composite image creation method according to the present invention preferably includes a composite image display step of displaying the composite image created in the composite image creation step. The earth's surface composite image creation system according to the present invention preferably includes composite image display means for displaying the composite image created by the composite image creation means. The earth's surface composite image creation program according to the present invention preferably causes a computer to further function as composite image display means for displaying the composite image created by the composite image creation means. In this case, the displayed composite image can be used to understand the utilization status of the earth's surface, such as vegetation.

[0021] The ground surface composite image creation method, ground surface composite image creation system, and ground surface composite image creation program according to the present invention are preferably used in areas with little elevation difference, such as rural areas, because they use height information to grasp the utilization status of the ground surface, such as vegetation. The ground surface composite image creation method, ground surface composite image creation system, and ground surface composite image creation program according to the present invention can be used, for example, to grasp the distribution of trees and plantings in residential woods and parks, to grasp the distribution of artificial objects such as buildings and solar panels, to grasp fallen trees due to disasters, and to manage them. Therefore, they are suitable for use in environmental consulting, environmental information processing services, construction, aerial surveying, agriculture, forestry, and fisheries, among others.

[0022] According to the present invention, it is possible to provide a method for creating a ground surface composite image, a ground surface composite image creation system, and a ground surface composite image creation program that can relatively easily grasp the utilization status of the ground surface, such as vegetation, using height information.

[0023] Fig. 1 is a block diagram showing the configuration of a ground surface composite image creation system according to an embodiment of the present invention; Fig. 2 is a flowchart showing the processing procedure of the ground surface composite image creation system according to an embodiment of the present invention; Fig. 3 shows (a) an orthoimage created with a certain rural area as a target area, (b) a grayscale image of (1) created in the target area of ​​(a), (c) a grayscale image of (2) created in the target area of ​​(a), and (d) a composite image created by superimposing the orthoimage of (a) and the grayscale image of (b), of the ground surface composite image creation system according to an embodiment of the present invention.

[0024] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Figures 1 to 3 show a method for creating a composite image of the earth's surface, a system for creating a composite image of the earth's surface, and a program for creating a composite image of the earth's surface according to an embodiment of the present invention. The method for creating a composite image of the earth's surface according to an embodiment of the present invention is preferably implemented by the system for creating a composite image of the earth's surface according to an embodiment of the present invention, and can be executed by a computer using the program for creating a composite image of the earth's surface.

[0025] As shown in Figure 1, a composite earth's surface image creation system 10 according to an embodiment of the present invention comprises a computer, and has a storage means 11, an input means 12, a display means 13, and a main control unit 14. The storage means 11 comprises a memory capable of storing various data. The input means 12 comprises a keyboard, mouse, etc., which allow a user to input various data and information. The display means 13 comprises a monitor (display). The main control unit 14 comprises a CPU and has calculation and control functions. The main control unit 14 is connected to the input means 12, storage means 11, and display means 13, and is configured to be able to control each of them.

[0026] The storage unit 11 is configured to store digital aerial image data of the target area captured in advance by a UAV, aerial photographs of the target area captured in advance from the air, and measurement data of the target area captured in advance from the air using a laser distance meter (LiDAR). The aerial image data captured by the UAV consists of multiple images captured consecutively from different viewpoints in the sky. The storage unit 11 is also configured to store 3D point cloud data, orthoimages, digital surface models (DSMs), ground elevation images, grayscale images, composite images, and other data created by the main control unit 14. The storage unit 11 may also be configured to store a digital elevation model (DEM) of the target area provided by the Geospatial Information Authority of Japan or other organizations, acquired in advance via the Internet or the like.

[0027] The input unit 12 is configured to input a set threshold value and to instruct how to create a grayscale image. The display unit 13 is configured to be able to display the aerial image data stored in the storage unit 11, the orthoimage, the digital surface model, the digital elevation model, the ground height image, the grayscale image, the composite image, etc.

[0028] The main control unit 14 includes an image acquisition means 21, a ground height image creation means 22, a pixel extraction means 23, a grayscale image creation means 24, a composite image creation means 25, and a composite image display means 26. The image acquisition means 21 is configured to acquire three-dimensional point cloud data by processing a plurality of aerial image data stored in the storage means 11 using Structure from Motion (SfM), and to create an orthoimage and a digital surface model of the target area from the three-dimensional point cloud data, or to create a digital surface model from LiDAR measurement data stored in the storage means 11. The image acquisition means 21 may also be configured to acquire a digital elevation model from the storage means 11. The created orthoimage is a three-channel image of RGB.

[0029] The ground height image creation means 22 is configured to create a ground height image relating to the height of objects on the ground based on the digital surface model acquired by the image acquisition means 21. The ground height image creation means 22 may, for example, create a digital elevation model based on the digital surface model, and create a ground height image from the difference between the digital surface model and the created digital elevation model. Alternatively, the ground height image creation means 22 may create a ground height image from the difference between the digital surface model and the digital elevation model acquired by the image acquisition means. In these cases, the created ground height image is an image based on a digital height model (DHM; Digital Height Model).

[0030] The pixel extraction means 23 is configured to extract pixels having a ground height equal to or greater than the predetermined threshold input by the input means 12 from the ground height image created by the ground height image creation means 22 .

[0031] The grayscale image creation means 24 is configured to set the grayscale value of each pixel extracted by the pixel extraction means 23 from the ground height image created by the ground height image creation means 22 according to the following (1) and (2), and create each grayscale image. (1) The grayscale value of each pixel extracted by the pixel extraction means 23 is set to a constant value so that each pixel becomes white. (2) The grayscale value of each pixel extracted by the pixel extraction means 23 is set to a value obtained by converting the corresponding ground height using a predetermined monotonically increasing function or monotonically decreasing function.

[0032] The composite image creation means 25 is configured to create a composite image by superimposing the grayscale image (1) or the grayscale image (2) created by the grayscale image creation means 24 on the orthoimage created by the image acquisition means 21, based on the selection input by the input means 12. The composite image display means 26 is configured to display the composite image created by the composite image creation means 25 on the display means 13.

[0033] The earth's surface composite image creation system 10 according to the embodiment of the present invention is configured so that each pixel not extracted by the pixel extraction means 23 is set to black (pixel value is zero), and the gradation value of the orthoimage created by the image acquisition means 21 becomes the gradation value of the composite image as is. Note that when creating the gradation image of (1), it is also possible to set each pixel extracted by the pixel extraction means 23 to be black, and each pixel not extracted by the pixel extraction means 23 to be white.

[0034] Next, the processing procedure and operation of the ground surface composite image creation system 10 according to an embodiment of the present invention will be described with reference to FIG. 2. As shown in FIG. 2, when using a UAV (drone), first, a target area is photographed continuously from different viewpoints in the sky, and the acquired multiple aerial image data sets are stored in the storage unit 11 (step 31). Next, the aerial image data sets stored in the storage unit 11 are processed by the image acquisition unit 21 using SfM to create 3D point cloud data (step 32). Furthermore, the image acquisition unit 21 creates an orthoimage of the target area (step 33) and a digital surface model (DSM image) (step 34) from the 3D point cloud data.

[0035] The ground height image creation means 22 uses the digital surface model and a digital elevation model (DEM image) (step 35) provided by a provider such as the Geospatial Information Authority of Japan and stored in the storage means 11 to obtain differential data (DSM-DEM) between the DSM image and the DEM image (step 36), and creates a ground height image (DHM) (step 37).

[0036] Furthermore, when aerial photographs and LiDAR measurement data are used without using a UAV (drone), the image acquisition means 21 orthorectifies the aerial photographs stored in the storage means 11 to create orthoimages (step 33), and a digital surface model (DSM image) is created from the LiDAR measurement data stored in the storage means 11. The ground height image creation means 22 filters the digital surface model to remove the heights of ground objects such as buildings and trees, creating a digital elevation model (DEM image), and obtains differential data (DSM-DEM) between the DSM image and the DEM image (step 36), creating a ground height image (DHM) (step 37). Here, when creating the orthoimage, the aerial photograph may be orthorectified using the DEM image created by the ground height image creation means 22. An example of an aerial photograph of a rural area as the target area is shown in Figure 3(a).

[0037] A ground height threshold is set with reference to the orthoimage, the created DSM image, the DEM image, the ground height image, etc., and the threshold is input from the input means 12 (step 38). The pixel extraction means 23 extracts pixels from the ground height image that have a ground height equal to or greater than the input threshold (step 39).

[0038] The grayscale image creation means 24 sets the grayscale value of each extracted pixel to a constant value so that each extracted pixel becomes white, and creates a grayscale image (1) according to this setting (step 40). Furthermore, the grayscale image creation means 24 uses a gradation conversion (contrast conversion) technique to set the grayscale value of each extracted pixel to a value obtained by converting the corresponding ground elevation using a predetermined monotonically increasing or decreasing function, and creates a grayscale image (2) according to this setting (step 41). Note that at this time, pixels not extracted by the pixel extraction means 23 are set to black (pixel value = zero) (binarization). Examples of the grayscale image (1) and the grayscale image (2) for the target area of ​​Figure 3(a) are shown in Figures 3(b) and (c), respectively. In Figures 3(b) and (c), the threshold is set to a height above ground that is higher than hedges and other vegetation and houses, but lower than residential forests. 3(b) and (c) are images obtained using an aerial photograph and LiDAR measurement data.

[0039] When creating the grayscale image (2), a monotonically increasing function or a monotonically decreasing function for converting the ground height into a grayscale value may be set in advance, or the type of monotonically increasing function or monotonically decreasing function to be used may be specified from the input means 12. The monotonically increasing function and the monotonically decreasing function may be a linear function or a nonlinear function, or may be a combination of these.

[0040] The user selects either the grayscale image (1) or the grayscale image (2) by referring to the created grayscale image (1) and grayscale image (2), and inputs the selection from the input means 12 (step 42). The composite image creation means 25 creates a composite image by superimposing the grayscale image (1) or the grayscale image (2) selected by input from the input means 12 on the orthoimage (step 43). The composite image display means 26 sends the composite image created by the composite image creation means 25 to the display means 13, and displays the composite image on the display means 13 (step 44). This allows the displayed composite image to be used to grasp the utilization status of the ground surface, such as vegetation.

[0041] An example of a composite image of the target area of ​​Figure 3(a) is shown in Figure 3(d). The composite image of Figure 3(d) is obtained by superimposing the grayscale image of (1) on an orthoimage. In such a composite image created by superimposing the grayscale image of (1) on an orthoimage, areas above a certain threshold appear white, making it easy to extract areas above that threshold. Therefore, for example, when the target area is a rural area, it is very difficult to distinguish between hedges and other vegetation that are lower than the houses and residential forests that have grown higher than the houses in an image taken from the air (e.g., the image of Figure 3(a)). However, in a composite image obtained by setting the certain threshold higher than the height of the houses, it is easy to extract only the residential forests. For example, from the composite image of Figure 3(d), the white areas can be easily extracted as residential forests.

[0042] Furthermore, in the composite image created by superimposing the grayscale image (2) on the orthoimage, the range of ground height above a predetermined threshold is converted into grayscale and reflected in the grayscale of the composite image, making it easy to grasp height information from the created composite image. Therefore, even in the composite image created by superimposing the grayscale image (2) on the orthoimage, differences in vegetation, etc., which are difficult to grasp from the appearance alone from above, can be easily extracted.

[0043] The ground surface composite image creation system 10 according to the embodiment of the present invention can easily generate a composite image that reflects height information by setting the shading of the orthoimage of the target area based on the ground height. Furthermore, by using this composite image, it is possible to determine not only the appearance from above but also differences in height, making it relatively easy to grasp the use status of the ground surface, such as vegetation. Furthermore, even if the resolution of the orthoimage is low, the accuracy of extracting vegetation, etc. can be improved.

[0044] The earth's surface composite image creation system 10 according to the embodiment of the present invention creates a composite image in which height information is reflected in the shading of the orthoimage, and therefore the composite image can be obtained as a three-channel image of RGB. This eliminates the need to convert a three-channel image into an image with four or more channels or to modify an image processing program for use with four or more channels, making it highly versatile.

[0045] The composite image created by the ground surface composite image creation system 10 according to the embodiment of the present invention may be used as training data for machine learning together with data on the ground surface utilization status, such as vegetation, ascertained from the composite image, in order to construct a ground surface utilization status estimation system for grasping the ground surface utilization status, such as vegetation. Furthermore, the ground surface utilization status, such as vegetation, may be grasped from the newly created composite image using the ground surface utilization status estimation system constructed in this manner.

[0046] A composite image of multiple regions was created using the earth's surface composite image creation system 10 according to an embodiment of the present invention. Images obtained by dividing a composite image into multiple parts were separated into training data and test data. Machine learning was performed using the training data (teacher data; the actual distribution of residential woodlands), and a system for estimating the distribution of residential woodlands was constructed. The accuracy of the constructed system's ability to extract residential woodlands was then evaluated using the test data. The Dice coefficient was used for the accuracy evaluation. As a comparative example, machine learning and accuracy evaluation were performed using images obtained by dividing only an orthoimage into multiple parts, without using the earth's surface composite image creation system 10 according to an embodiment of the present invention, as training data and test data.

[0047] As a result, the Dice coefficient in the comparative example where the compound forest was extracted using only the orthoimage was 0.64. In contrast, the Dice coefficient when the compound forest was extracted using the composite image made up of the grayscale image (1) was 0.76. Furthermore, the Dice coefficient when the compound forest was extracted using the composite image made up of the grayscale image (2) was 0.80. Note that when creating the grayscale image (2), a linear transformation was used as the gradation transformation. From these results, it can be said that the use of composite images can improve extraction accuracy.

[0048] The earth's surface composite image creation program according to the embodiment of the present invention may be provided externally via a communication line, for example, or may be provided in a form recorded on a computer-readable recording medium such as a CD (CD-ROM, CD-R, CD-RW, etc.), a DVD (DVD-ROM, DVD-RAM, DVD-R, DVD-RW, DVD+R, DVD+RW, etc.), or a USB memory. In these cases, the computer can read the earth's surface composite image creation program from the communication line or recording medium, transfer it to an internal storage device of the computer, and store and use it. Furthermore, the earth's surface composite image creation program according to the embodiment of the present invention may be recorded in a storage device (recording medium) such as a magnetic disk, optical disk, or magneto-optical disk, and provided to the computer from the storage device via a communication line.

[0049] Here, the term "computer" refers to hardware that operates under the control of an operating system (OS). In addition, if an OS is not required and the hardware is operated solely by an application program, the hardware itself is considered a computer. The hardware includes at least a microprocessor such as a CPU and a means for reading computer programs stored in an internal storage device or a recording medium.

[0050] The application program as the earth's surface composite image creation program according to the embodiment of the present invention includes program code for causing the computer to execute the program as described above. Furthermore, some of the functions may be implemented by the OS rather than by the application program.

[0051] REFERENCE SIGNS LIST 10 Earth's surface composite image creation system 11 Storage means 12 Input means 13 Display means 14 Main control unit 21 Image acquisition means 22 Ground height image creation means 23 Pixel extraction means 24 Grayscale image creation means 25 Composite image creation means 26 Composite image display means

Claims

1. A method for creating a composite image of the earth's surface, comprising: an image acquisition step for acquiring an ortho-image and a digital surface model (DSM) of a target area; a ground height image creation step for creating a ground height image relating to the height of objects on the ground based on the digital surface model acquired in the image acquisition step; a pixel extraction step for extracting pixels having a ground height equal to or greater than a predetermined threshold from the ground height image created in the ground height image creation step; a grayscale image creation step for creating a grayscale image in which the grayscale value of each pixel extracted in the pixel extraction step is set based on the corresponding ground height or set to a constant value; and a composite image creation step for creating a composite image by superimposing the grayscale image created in the grayscale image creation step on the ortho-image.

2. A method for producing a composite image of the earth's surface according to claim 1, wherein said grayscale image producing step sets each grayscale value to a constant value so that each pixel extracted in said pixel extracting step becomes white.

3. A method for creating a composite image of the earth's surface as described in claim 1, characterized in that the grayscale image creation step sets the grayscale value of each pixel extracted in the pixel extraction step to a value obtained by converting the corresponding ground height using a predetermined monotonically increasing function or monotonically decreasing function.

4. A method for creating a synthetic ground surface image as described in claim 1, characterized in that the ground height image creation step creates a digital elevation model (DEM; Digital Elevation Model) based on the digital surface model, and creates the ground height image from the difference between the digital surface model and the created digital elevation model.

5. A method for creating a synthetic earth surface image according to claim 1, characterized in that the image acquisition step also acquires a digital elevation model of the target area, and the ground height image creation step creates the ground height image from the difference between the digital surface model and the digital elevation model acquired in the image acquisition step.

6. A method for creating a synthetic image of the earth's surface as described in any one of claims 1 to 5, characterized in that the image acquisition step acquires three-dimensional point cloud data from image data photographed from the air over the target area, and creates the orthoimage and the digital surface model from the acquired three-dimensional point cloud data.

7. A method for creating a synthetic earth surface image according to any one of claims 1 to 5, characterized in that the image acquisition step creates the orthoimage and the digital surface model based on aerial photographs taken of the target area from above and measurement data obtained by measuring the target area from above using a laser distance measuring instrument (LiDAR).

8. A method for producing a composite image of the earth's surface according to any one of claims 1 to 5, further comprising a composite image display step for displaying the composite image produced in the composite image production step.

9. A ground surface composite image creation system comprising: an image acquisition means for acquiring an ortho-image and a digital surface model of a target area; a ground height image creation means for creating a ground height image relating to the height of objects on the ground based on the digital surface model acquired by the image acquisition means; a pixel extraction means for extracting pixels having a ground height equal to or greater than a predetermined threshold from the ground height image created by the ground height image creation means; a grayscale image creation means for creating a grayscale image in which the grayscale value of each pixel extracted by the pixel extraction means is set based on the corresponding ground height or set to a constant value; and a composite image creation means for creating a composite image by superimposing the grayscale image created by the grayscale image creation means on the ortho-image.

10. A ground surface composite image creation program which causes a computer to function as: an image acquisition means for acquiring an ortho-image and a digital surface model of a target area; a ground height image creation means for creating a ground height image relating to the height of objects on the ground based on the digital surface model acquired by the image acquisition means; a pixel extraction means for extracting pixels having a ground height equal to or greater than a predetermined threshold from the ground height image created by the ground height image creation means; a grayscale image creation means for creating a grayscale image in which the grayscale value of each pixel extracted by the pixel extraction means is set based on the corresponding ground height or set to a constant value; and a composite image creation means for creating a composite image by superimposing the grayscale image created by the grayscale image creation means on the ortho-image.