Estimation device, program, and estimation method

Abstract

To grasp the shape of a ground surface with simple processing.SOLUTION: An estimation device (100) comprises: an information acquisition unit (101) which acquires a first infrared image (Ii1) and a second visible light image (Iv2); a temperature value specification unit (104) which specifies a temperature value (T); a model generation unit (102) which generates a second object DSM (3); a change amount calculation unit (103) which calculates an altitude change amount (ΔE); and an altitude value estimation unit (105) which estimates a second altitude value based on a first correlation.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an estimation device, program, and estimation method for estimating elevation values ​​in each of a plurality of grids constituting a digital elevation model. [Background technology]

[0002] Various technologies related to land topographic surveying have been researched and developed over time. A typical example is the technology that acquires 3D point cloud data representing the land's topography and generates a 3D model by gridding this 3D point cloud data. 3D point cloud data consists of X and Y coordinate data representing horizontal position, and Z coordinate data representing elevation.

[0003] Incidentally, if the land subject to a 3D model includes geographical features, this 3D model becomes a DSM (Digital Surface Model) that includes the shapes of these features. Geographical features include, for example, artificial structures such as buildings and bridges, as well as vegetation such as trees, and are objects that obscure the ground surface. In this case, if we want to understand the shape of the land after removing the shapes of the geographical features from the topography, that is, the shape of the ground surface, we apply a filtering process to the DSM to generate a DEM (Digital Elevation Model). Filtering is the process of removing the measurement data of the geographical features from the DSM. A DEM is a 3D model that represents the shape of the ground surface.

[0004] As a technique for generating a DEM through filtering, for example, Patent Document 1 discloses an information processing device, etc., that generates filtered elevation mesh data from filtered image data output by an image generation model. The filtered elevation mesh data corresponds to a DEM. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Specification of Patent No. 6762636

Summary of the Invention

Problems to be Solved by the Invention

[0006] However, when generating filtered elevation mesh data using the information processing apparatus disclosed in Patent Document 1 or the like, it is necessary to continue machine learning until the image generation model outputs filtered image data with a level of accuracy that can be used. Therefore, compared with the generation of DSM without performing filtering processing, the generation of filtered elevation mesh data takes time and effort. In addition, the selection of learning data required for this machine learning requires a high level of proficiency equivalent to that required for filtering processing by other methods. From these facts, there is a problem that the information processing apparatus disclosed in Patent Document 1 or the like cannot easily grasp the shape of the ground surface.

[0007] One aspect of the present invention aims to enable the shape of the ground surface to be grasped by simple processing.

Means for Solving the Problems

[0008] To solve the aforementioned problems, an estimation device according to one aspect of the present invention estimates the shape of the ground surface by virtually dividing the ground surface and features into a grid, with each of the multiple unit spaces obtained being a virtual grid, and designating an arbitrary virtual grid among the multiple virtual grids as a grid of interest. The estimation device comprises an information acquisition unit that acquires infrared images of the target ground surface and features, which are the ground surface to be estimated, and optical information of reflected light reflected by the target ground surface and features; a temperature value identification unit that identifies the temperature value in the first grid corresponding to the grid of interest among the multiple grids obtained by dividing the infrared image into a grid; and an information acquisition unit that uses the optical information to determine the shape of the target ground surface and the surface of the features. The system comprises: a model generation unit that generates a numerical surface model as a three-dimensional model of the surface to be measured; a change amount calculation unit that calculates the change in elevation in the second grid by subtracting a first reference value from a first elevation value, which is the elevation value in the second grid corresponding to the grid of interest, among a plurality of grids obtained by dividing the numerical surface model into a grid; and an elevation value estimation unit that identifies the correlation between the change in elevation in the second grid and the temperature value in the first grid as a first correlation, and estimates the elevation value in the third grid corresponding to the grid of interest, among a plurality of grids obtained by dividing the numerical elevation model, which is a three-dimensional model of the target ground surface, into a grid, as the second elevation value based on the first correlation.

[0009] In one aspect of the present invention, the elevation value estimation unit calculates a first relational expression which is an expression showing the first correlation, and the second elevation value is obtained by subtracting the amount of change obtained by substituting the temperature value into the first relational expression from the first elevation value.

[0010] In one aspect of the present invention, the estimation device may have a model generation unit that performs filtering to extract points whose elevation value is equal to or greater than a second reference value from among a plurality of points constituting the three-dimensional point cloud data of the surface that forms the basis of the numerical surface model, and generate the numerical surface model from the three-dimensional point cloud data from which the points extracted by the filtering have been removed.

[0011] An estimation device according to one aspect of the present invention further comprises a solar radiation estimation unit that estimates the amount of solar radiation in the second grid as estimated solar radiation, and a difference calculation unit that calculates the difference between a reference amount and the estimated solar radiation, wherein the temperature value identification unit corrects the temperature value in the first grid using the difference, and the elevation value estimation unit considers the corrected temperature value in the first grid as the temperature value in the first grid to identify the first correlation.

[0012] In one aspect of the present invention, the estimation device may include a temperature value identification unit that calculates a correction value for correcting the temperature value in the first grid, calculates a second correlation equation as a second relational equation which is the correlation between the difference in the second grid and the correction value in the first grid, and adds the correction value obtained by substituting the difference into the second relational equation to the temperature value, and use this value as the corrected temperature value.

[0013] A program according to one aspect of the present invention is a program for causing a computer to function as the estimation device, and is a program for causing the computer to function as the information acquisition unit, the temperature value identification unit, the model generation unit, the change amount calculation unit, and the altitude value estimation unit.

[0014] To solve the aforementioned problems, an estimation method according to one aspect of the present invention estimates the shape of the ground surface by virtually dividing the ground surface and features into a grid, with each of the multiple unit spaces obtained being a virtual grid, and designating any virtual grid among the multiple virtual grids as a grid of interest, the estimation method comprising: an information acquisition step of acquiring an infrared image of the target ground surface and features which are the ground surface to be estimated, and optical information of reflected light reflected by the target ground surface and features; a temperature value identification step of identifying the temperature value in the first grid corresponding to the grid of interest among the multiple grids obtained by dividing the infrared image acquired in the information acquisition step into a grid; and using the optical information acquired in the information acquisition step, the surface of the target ground surface and features is composed of The method includes: a model generation step of generating a numerical surface model as a three-dimensional model of the surface; a change amount calculation step of dividing the numerical surface model generated in the model generation step into a grid to obtain a plurality of grids, and subtracting a first reference value from a first elevation value, which is the elevation value in a second grid corresponding to the grid of interest, as the change in elevation in the second grid; and an elevation value estimation step of identifying the correlation between the change in elevation in the second grid calculated in the change amount calculation step and the temperature value in the first grid as a first correlation, and estimating the elevation value in a third grid corresponding to the grid of interest, which is the elevation value in a plurality of grids, obtained by dividing the numerical elevation model, which is a three-dimensional model of the target ground surface, into a grid, as the second elevation value based on the first correlation. [Effects of the Invention]

[0015] According to one aspect of the present invention, the shape of the ground surface can be determined with a simple process. [Brief explanation of the drawing]

[0016] [Figure 1] This is a block diagram showing the functional configuration of an information processing apparatus according to the first to third embodiments of the present invention. [Figure 2] This is a photograph used as a substitute for a diagram, illustrating an example of a grid of interest. [Figure 3] This is a flowchart showing an example of an estimation method according to the first embodiment of the present invention. [Figure 4] Reference numeral 401 denotes a diagram showing an example of a first visible light image. Reference numeral 402 denotes a diagram showing an example of a second visible light image. [Figure 5] This is a photograph used as a substitute for a drawing, showing an example of a second target DSM where elevation change data is superimposed. [Figure 6] This is a photograph used as a substitute for a diagram, showing an example of the first infrared image. [Figure 7] This box plot shows an example of the correlation between temperature values ​​in the first grid and elevation changes in the corresponding second grid. [Figure 8] This is an example of a graph showing the first correlation. [Figure 9] This is a photograph used as a substitute for a diagram, showing an example of a second target DSM where the elevation change amount obtained using the first relational equation is superimposed. [Figure 10] Reference numeral 1001 is a substitute photograph showing an example of the basic 3D point cloud data before filtering. Reference numeral 1002 is a substitute photograph showing an example of the basic 3D point cloud data after filtering. [Figure 11] This is a photograph used as a substitute for a diagram, showing an example of a second infrared image. [Figure 12] This is a flowchart showing an example of an estimation method according to the third embodiment of the present invention. [Figure 13] This is an example of a graph showing the correlation between the temperature value in the first grid and the average value of the estimated basic solar radiation in the second grid corresponding to the first grid. [Figure 14] This is a photograph used as a substitute for a diagram, showing an example of a second target DSM with estimated solar radiation superimposed. [Figure 15] This is a photograph used as a substitute for a drawing, showing an example of a second target DSM with the difference superimposed. [Figure 16] This is an example of a graph showing the correlation between average solar radiation loss and temperature loss. [Figure 17]This is a photograph used as a substitute for a diagram, showing an example of a second infrared image with correction values ​​superimposed. [Figure 18] This is a photograph used as a substitute for a diagram, showing an example of a second infrared image in which the corrected temperature values ​​are superimposed. [Modes for carrying out the invention]

[0017] The first to third embodiments of the present invention will be described in detail below with reference to Figures 1 to 18. In the first to third embodiments, a field will be used as an example of the land to be estimated by the estimation device according to one aspect of the present invention. However, the land to be estimated is not limited to a field, and various types of land such as forests, pastures, urban areas, and industrial parks can also be used as targets for estimation.

[0018] [First Embodiment] <Configuration of the information processing device> Referring to Figures 1 and 2, the configuration of the information processing device 1 according to the first embodiment of the present invention will be described. The information processing device 1 is a device capable of processing various types of information. The information processing device 1 may be, for example, a tablet terminal or a smartphone. Alternatively, the information processing device 1 may be a stationary personal computer. As shown in Figure 1, the information processing device 1 includes an input unit 11, a display unit 12, a storage unit 13, a communication unit 14, and a control unit 15.

[0019] The input unit 11 is an interface that accepts various operations. For example, if the information processing device 1 is a stationary personal computer, the keyboard and mouse become the input unit 11. The display unit 12 is an output unit that displays various information. For example, if the information processing device 1 is a stationary personal computer, the monitor becomes the display unit 12. Also, for example, if the information processing device 1 is a smartphone or tablet terminal, the information processing device 1 may have a touch panel that integrates the input unit 11 and the display unit 12.

[0020] The memory unit 13 stores various types of information used by the information processing device 1. Examples of memory units 13 and the memory unit 22 described later include RAM (Random Access Memory), flash memory, and hard disk. The communication unit 14 is a unit for the information processing device 1 to send and receive various types of information with the aircraft 2 described later, or other information processing devices (not shown). The control unit 15 is, for example, a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), and controls all parts of the information processing device 1. The control unit 15 also executes processing to realize the various functions of the information processing device 1. As shown in Figure 1, the control unit 15 is equipped with an estimation device 100.

[0021] Estimation device 100 is a device for estimating the shape of the target ground surface. The target ground surface is the ground surface that the estimation device 100 is intended to estimate. Here, the term "ground surface" in this specification does not refer only to surfaces where there are no features at all and soil or sand is exposed throughout, nor to paved surfaces. Even if part or all of a surface is covered with features, if the height of the features from the reference plane (mean sea level of Tokyo Bay) is low enough to be considered the same as the elevation of the ground surface directly beneath the features, the surface of the features shall also be considered part or all of the "ground surface" in this specification.

[0022] In the first to third embodiments, the sloping surface of a field levee is defined as the target ground surface, and vegetation is assumed to cover the target ground surface as a geographic feature. In the first and second embodiments, the land composed of the target ground surface and the vegetation as a geographic feature is defined as the target land LT (see the polygon enclosed in the red frame in Figure 4, reference numeral 401). In the third embodiment, the land composed of the target ground surface and the vegetation as a geographic feature is defined as the target land LT' (see the polygon enclosed in the black frame in Figure 11). Hereinafter, target land LT and LT' will be collectively referred to as "target land".

[0023] The estimation device 100 estimates the shape of the target ground surface by selecting any virtual grid from among multiple virtual grids as the grid of interest. The grid of interest will be explained below using Figure 2. For example, when the shape of the ground surface of the land LS enclosed by the black frame in Figure 2 (hereinafter referred to as "reference land LS") is to be estimated, the estimation device 100 virtually divides the reference land LS into a grid.

[0024] Note that the division intervals, or grid sizes, shown in Figure 2 are merely examples. The estimation device 100 can arbitrarily change the grid size according to the size of the target land, the performance (resolution, resolution) of the visible light camera and infrared camera in the imaging unit 21 described later, the distance between the target land and these cameras, etc. The same applies to the first to third grids described later.

[0025] Next, the estimation device 100 divides the reference land LS into a virtual grid, and each of the resulting unit spaces is designated as a virtual grid GK'. Then, the estimation device 100 extracts an arbitrary virtual grid GK' from among the multiple virtual grids GK' that constitute the reference land LS, and designates it as the grid of interest GT'. This method of obtaining the grid of interest GT' from the reference land LS is also applicable to the target land.

[0026] Note that Figure 2 shows a photograph of the reference land LS taken in a plan view, and therefore the virtual grid GK' and the grid of interest GT' are represented in two dimensions (latitude and longitude) in Figure 2. However, the virtual grid GK' and the grid of interest GT' are actually three-dimensional concepts that include elevation.

[0027] The estimation device 100 may be provided, for example, outside the control unit 15 within the information processing device 1, or it may be provided on the aircraft 2. Alternatively, the estimation device 100 may be provided on another information processing device.

[0028] As shown in Figure 1, the estimation device 100 includes an information acquisition unit 101, a model generation unit 102, a change amount calculation unit 103, a temperature value identification unit 104, and an altitude value estimation unit 105.

[0029] The information acquisition unit 101 acquires an infrared image of the target land (see Figure 6) and optical information of the reflected light reflected by the target land. In the first to third embodiments, a visible light image of the target land (see Figure 4) is used as an example of optical information.

[0030] The optical information may also be the distance between the laser scanner and the surface, obtained from the time difference between when the laser scanner emits laser light onto the target land and when the reflected laser light returns. The surface is the surface of the target land, in other words, the surface composed of the target land surface and the surface of the vegetation. Alternatively, the optical information may be the optical information of the reflected light when infrared light is reflected by the target land.

[0031] The model generation unit 102 generates the target DSM (see Figure 5, etc.) using a visible light image. The target DSM is a numerical surface model generated as a three-dimensional model of the surface. Details of the target DSM generation process by the model generation unit 102 will be described later.

[0032] The change amount calculation unit 103 calculates the change in elevation in the second grid (see Figure 5, etc.; hereinafter referred to as "elevation change amount"). The second grid (see Figure 5) is the grid corresponding to the grid of interest in the target land, among the multiple grids obtained by dividing the target DSM into a grid.

[0033] Specifically, the change amount calculation unit 103 subtracts the first reference value from the first elevation value, which is the elevation value in the second grid, as the elevation change amount in the second grid. Here, "elevation value in the second grid" refers to the average of the elevation values ​​in the point cloud, which is aggregated and transformed into an arbitrary second grid from among multiple points that constitute the 3D point cloud data of the surface layer that forms the basis of the target DSM. Details of the first reference value will be described later.

[0034] The temperature value identification unit 104 identifies the temperature value in the first grid (see Figure 6, etc.). The first grid (see Figure 6) is a grid corresponding to the grid of interest in the target land, among a plurality of grids obtained by dividing the infrared image into a grid. Here, "temperature value in the first grid" refers to the average value of each temperature value in the point cloud, which is aggregated and transformed into an arbitrary first grid from a plurality of points that constitute the 3D point cloud data of the surface layer that forms the basis of the infrared image.

[0035] The elevation estimation unit 105 identifies the correlation between the elevation change and the temperature value as the first correlation. The elevation estimation unit 105 also estimates the elevation value in the third grid (not shown) as the second elevation value based on the identified first correlation. The third grid is a grid corresponding to the grid of interest on the target land, among multiple grids obtained by dividing the target DEM (not shown) into a grid. The target DEM is a numerical elevation model generated as a three-dimensional model of the target land surface.

[0036] <Configuration of the flying object> Referring to Figure 1, the configuration of the aircraft 2 according to the first to third embodiments of the present invention will be described. The aircraft 2 is an unmanned, autonomously flying aircraft such as a drone or UAV (Unmanned Aerial Vehicle), and captures visible light images and infrared images. The aircraft 2 may also be a manned aircraft such as an airplane or helicopter. As shown in Figure 1, the aircraft 2 comprises an imaging unit 21, a storage unit 22, a communication unit 23, and a control unit 24.

[0037] The imaging unit 21 has a visible light camera and an infrared camera (neither shown), and captures visible light images and infrared images simultaneously or synchronously. The storage unit 22 may record the various images captured by the imaging unit 21. The communication unit 23 is a unit for the aircraft 2 to send and receive various information with the information processing device 1 or other information processing devices. The aircraft 2 transmits the visible light images and infrared images captured by the imaging unit 21 to the information processing device 1 via the communication unit 23. The control unit 24, like the control unit 15, is for example a CPU or GPU, and controls all parts of the aircraft 2. The control unit 24 also executes processing to realize the various functions that the aircraft 2 is equipped with.

[0038] The subject of visible light and infrared image capture is not limited to the flying object 2. For example, if any mobile body capable of moving on the surface of the target ground and on the surface of a ground object is equipped with an imaging unit 21 and a communication unit 23, the mobile body may capture at least one of the visible light and infrared images. In this case, the aforementioned mobile body may transmit the captured images to the information processing device 1 via the communication unit 23. Examples of the aforementioned mobile body include buggies, tractors, and lawnmowers equipped with a camera that includes an imaging unit 21.

[0039] <Main processing functions of information processing equipment> The main processing of the information processing device 1 will be explained with reference to Figures 3 to 9. As a prerequisite, it is assumed that the aircraft 2 has transmitted the first visible light image Iv1 shown by reference numeral 401 in Figure 4, the second visible light image Iv2 shown by reference numeral 402 in Figure 4, and the first infrared image Ii1 (see Figure 6) to the information processing device 1.

[0040] The first visible light image Iv1 (light information) is a visible light image of the target land LT taken by the aircraft 2 at a specific point in time (the first time). The second visible light image Iv2 (light information) is a visible light image of the target land LT taken by the aircraft 2 at a point in time (the second time), which is later than the first time. The first infrared image Ii1 is an infrared image of the target land LT taken by the aircraft 2 at the second time.

[0041] In the first and second embodiments, the first time point is defined as the point immediately after mowing the target land LT. Therefore, at the second time point, the vegetation on the target land LT has grown compared to the first time point. In other words, at the second time point, the elevation value is higher in at least a portion of the surface compared to the first time point.

[0042] First, in step S501 (information acquisition step) of the flowchart shown in Figure 3, the information acquisition unit 101 acquires the first and second visible light images Iv1 and Iv2 and the first infrared image Ii1 via the communication unit 14. The information acquisition unit 101 then transmits the acquired first and second visible light images Iv1 and Iv2 to the model generation unit 102. The information acquisition unit 101 also transmits the acquired first infrared image Ii1 to the temperature value identification unit 104. The first visible light image Iv1 may be stored in the storage unit 13 beforehand, in which case the information acquisition unit 101 acquires the first visible light image Iv1 from the storage unit 13.

[0043] Next, in S502 (model generation step), the model generation unit 102 generates a first target DSM (not shown) using the acquired first visible light image Iv1. The model generation unit 102 also generates a second target DSM3 (see Figure 5) using the acquired second visible light image Iv2. The first target DSM according to the first and second embodiments is a numerical surface model generated as a three-dimensional model of the surface surface of the target land LT at a first time point. The second target DSM3 is a numerical surface model generated as a three-dimensional model of the surface surface of the target land LT at a second time point.

[0044] In the first and second embodiments, the model generation unit 102 generates a first target DSM and a second target DSM3 based on the 3D point cloud data generated by SfM-MVS (Structure From Motion - Multi-View stereo) processing. SfM-MVS processing is a known process for creating 3D point cloud data of the surface shape of a target object, and the MVS processing is performed after the SfM processing. The generation of the second target DSM3 will be explained below as an example.

[0045] The model generation unit 102 calculates the position of the imaging unit 21 at the time of imaging from the feature points of the target land LT, which is the subject of the second visible light image Iv2, using SfM processing, and generates low-density 3D point cloud data. Next, the model generation unit 102 uses MVS processing to increase the density of the low-density 3D point cloud data based on the position of the imaging unit 21 calculated by SfM processing, and generates 3D point cloud data of the surface layer that forms the basis of the second target DSM3 (hereinafter referred to as "basic 3D point cloud data").

[0046] The model generation unit 102 then converts the basic 3D point cloud data into a second target DSM3. Specifically, the model generation unit 102 divides the basic 3D point cloud data into a grid using a grid size set in advance by the user. The grid size setting may be stored in advance in the storage unit 13, or it may be set by an input operation received by the input unit 11.

[0047] The model generation unit 102 then performs statistical processing (or may average) on each elevation value of the point cloud within each grid for all grids obtained by the division, and uses the resulting average value as the elevation value in the grid. Through this series of processes, the elevation values ​​of multiple points that make up the basic 3D point cloud data are aggregated into a grid, thereby converting the basic 3D point cloud data into the second target DSM3.

[0048] It should be noted that the generation of the first target DSM and the second target DSM3 by SfM-MVS processing is merely an example, and the model generation unit 102 may generate the first target DSM and the second target DSM3 by known processing other than SfM-MVS processing. Examples of such known processing include manual generation processing using a digital stereo plotting machine and generation processing using satellite images.

[0049] The model generation unit 102 then transmits the generated first target DSM and second target DSM3 to the change amount calculation unit 103. The model generation unit 102 also transmits the generated second target DSM3 to the elevation value estimation unit 105.

[0050] Next, in S503 (change amount calculation step), the change amount calculation unit 103 calculates the elevation change amount ΔE using the acquired first target DSM and second target DSM3. The elevation change amount ΔE is the elevation change amount in the second grid G2, which corresponds to the grid of interest GT, among the multiple grids obtained by dividing the second target DSM3 into a grid. The grid of interest GT is the grid of interest in the target land LT. For each of the multiple grids constituting the second target DSM3, the change amount calculation unit 103 calculates the elevation change amount ΔE when that grid is the second grid G2.

[0051] Specifically, the change amount calculation unit 103 subtracts the first reference value from the first elevation value in the second grid G2 to determine the elevation change amount ΔE in the second grid G2. In the first and second embodiments, the elevation value in the second grid corresponding to the grid of interest GT, among the multiple grids obtained by dividing the first target DSM into a grid, is used as the first reference value. In other words, a first reference value exists for each of the multiple grids that constitute the first target DSM. The change amount calculation unit 103 obtains the first reference value by dividing the first target DSM into a grid.

[0052] Alternatively, the first reference value may be determined in advance and stored in the memory unit 13. In this case, the change amount calculation unit 103 obtains the first reference value from the memory unit 13. Alternatively, the first reference value may be a value different from the elevation value in the grid, which is a constituent unit of the first target DSM. For example, the elevation value in the grid, which is a constituent unit of the target DEM, may be used as the first reference value.

[0053] The change amount calculation unit 103 then transmits the calculated elevation change amount ΔE to the elevation value estimation unit 105. In the first and second embodiments, the input unit 11 accepts user input, causing the change amount calculation unit 103 to transmit the second target DSM3 and the calculated elevation change amount ΔE to the display unit 12. The display unit 12 then displays the second target DSM3 with the image of the elevation change amount ΔE superimposed, as shown in Figure 5.

[0054] Next, in S504 (temperature value identification step), the temperature value identification unit 104 identifies the temperature value T in the first grid G1 from the acquired first infrared image Ii1. The first grid G1 is the grid corresponding to the grid of interest GT among a plurality of grids obtained by dividing the first infrared image Ii1 into a grid. For each of the plurality of grids constituting the first infrared image Ii1, the temperature value T is identified when that grid is designated as the first grid G1.

[0055] The temperature value identification unit 104 then transmits the identified temperature value T to the altitude value estimation unit 105. In the first and second embodiments, the input unit 11 accepts user input, causing the temperature value identification unit 104 to perform SfM-MVS processing on the first infrared image Ii1. This processing generates the first infrared image Ii1 in which the imaged temperature value T is superimposed, as shown in Figure 6. The temperature value identification unit 104 transmits the first infrared image Ii1 after SfM-MVS processing to the display unit 12, and the display unit 12 displays the first infrared image Ii1 after SfM-MVS processing.

[0056] The order of the temperature value identification process performed by the temperature value identification unit 104 to identify the temperature value T is not limited to the flowchart example shown in Figure 3. For example, the temperature value identification unit 104 may perform the identification process between S501 and S502, or at the same time as S502. Alternatively, the temperature value identification unit 104 may perform the identification process between S502 and S503.

[0057] Next, in S505 (altitude value estimation step), the altitude value estimation unit 105 identifies the correlation between the altitude change amount ΔE and the temperature value T as the first correlation. In the first to third embodiments, the altitude value estimation unit 105 identifies the first correlation by calculating a first relational expression, which is an expression representing the first correlation.

[0058] Specifically, the elevation estimation unit 105 generates a box plot, as shown in Figure 7, which shows the correlation between the temperature value T in the first grid G1 and the elevation change ΔE in the second grid G2 corresponding to the first grid G1. Then, the elevation estimation unit 105 generates the graph shown in Figure 8 using the generated box plot. The graph shown in Figure 8 shows the correlation between the temperature value T in the first grid G1 and the median elevation change ΔE for each temperature value T (hereinafter abbreviated as "median"). Then, the elevation estimation unit 105 uses the generated graph to calculate an approximation formula that approximates the distribution trend of each point plotted on the graph. This approximation formula corresponds to the first relational formula.

[0059] Equation (1) below is an example of the first relation, and is obtained when the elevation change ΔE is the result shown in Figure 5 and the temperature value T is the specific result shown in Figure 6. In equation (1) below, x is the temperature value T [°C] and y is the median [m].

[0060] (Math 1) y = 23.459e -0.173x ...(1) Next, in S506 (elevation value estimation step), the elevation value estimation unit 105 estimates the second elevation value in the third grid using the calculated first relational expression. In this case, the third grid refers to the grid corresponding to the grid of interest GT, which is obtained by dividing the second target DEM (not shown) into a grid. The second target DEM is a numerical elevation model generated as a three-dimensional model of the target ground surface at the second time point.

[0061] Specifically, the elevation estimation unit 105 substitutes the temperature value T at the first grid G1 into the first relational expression to calculate the median value at the second grid G2 corresponding to the first grid G1. Here, the elevation estimation unit 105 considers the calculated median value to be the elevation change ΔE at the second grid G2. Then, the elevation estimation unit 105 subtracts the calculated median value (i.e., elevation change ΔE) from the first elevation value at the second grid G2 to be the second elevation value at the third grid corresponding to the second grid.

[0062] In the first and second embodiments, the input unit 11 receives user input, and the elevation value estimation unit 105 transmits the elevation change amount ΔE calculated from the second target DSM3 and the first relational expression to the display unit 12. The display unit 12 then displays the second target DSM3 with the imaged elevation change amount ΔE superimposed, as shown in Figure 9. With the completion of the process in S506, all major processes performed by the information processing device 1 are completed.

[0063] <Variation> In estimating the second elevation value by the elevation value estimation unit 105, it is not essential to use the first relational expression. The elevation value estimation unit 105 may, for example, estimate the second elevation value using the first estimation model instead of the first relational expression.

[0064] The first estimation model is a trained model constructed by machine learning using the first basic dataset as training data, with temperature values ​​as input data and elevation change amounts as output data. There are no particular limitations on the machine learning method, and known methods such as convolutional neural networks can be used. The first basic dataset is a dataset consisting of past temperature values ​​and past elevation change amounts. Past temperature values ​​are the temperature values ​​at each of the multiple first grids that make up the infrared image at a time before the first time point. Past elevation change amounts are the elevation change amounts at the second grid corresponding to each of the multiple first grids mentioned above, and are elevation change amounts calculated by the change amount calculation unit 103.

[0065] In this case, the elevation estimation unit 105 inputs the temperature value T in the first grid G1 into the first estimation model, thereby obtaining the elevation change ΔE in the second grid G2 corresponding to the first grid G1 from the first estimation model. The elevation estimation unit 105 then subtracts the obtained elevation change ΔE from the first elevation value in the second grid G2 and uses this value as the second elevation value in the third grid corresponding to the second grid.

[0066] <Summary> The information processing device 1 uses the elevation value estimation unit 105 of the estimation device 100 to identify the first correlation and estimate the second elevation value. This makes it possible to obtain a value similar to the second elevation value in the third grid without generating the target DSM. Therefore, the shape of the ground surface can be grasped with simple processing.

[0067] Furthermore, according to the information processing device 1, by subtracting the elevation change obtained by substituting the temperature value in the first grid into the first relational equation from the first elevation value in the second grid, a value similar to the second elevation value in the third grid can be obtained. As a result, the second elevation value in all third grids can be estimated simply by uniformly applying the first relational equation to all target grids GT. Therefore, the shape of the ground surface can be understood more easily.

[0068] [Second Embodiment] A second embodiment of the present invention will be described below. For the sake of convenience, components having the same function as those described in the first embodiment will be denoted by the same reference numerals, and their descriptions will not be repeated. The same applies to the third embodiment described later.

[0069] <Characteristic Configuration of Information Processing Devices> Referring to Figures 1 and 10, the characteristic configuration of the information processing device 1a according to the second embodiment of the present invention will be described. As shown in Figure 1, the information processing device 1a includes an estimation device 200 having a model generation unit 102a instead of the estimation device 100. In this respect, the information processing device 1a differs from the information processing device 1 according to the first embodiment of the present invention.

[0070] The model generation unit 102a performs filtering on the basic 3D point cloud data. Filtering is a process that extracts points from among multiple points that make up the basic 3D point cloud data whose elevation value is equal to or greater than the second reference value. The second reference value is an elevation value that can be arbitrarily set by the user, and may be stored in advance in the storage unit 13, or it may be set by the input unit 11 accepting user input.

[0071] If you want to extract point clouds corresponding to perennial grasses from the basic 3D point cloud data, you could consider setting the second reference value to, for example, 50-300 [cm]. If you want to extract point clouds corresponding to trees from the basic 3D point cloud data, you could consider setting the second reference value to, for example, 300 [cm] or higher.

[0072] Furthermore, the model generation unit 102a generates new 3D point cloud data (hereinafter referred to as "processed 3D point cloud data") by removing points extracted by filtering from the basic 3D point cloud data. Then, the model generation unit 102a generates the target DSM from the processed 3D point cloud data.

[0073] For example, if the basic 3D point cloud data is the basic 3D point cloud data P shown by reference numeral 1001 in Figure 10, and the second reference value is set to 300 [cm], the model generation unit 102a extracts the point cloud PG corresponding to trees (the part enclosed by the yellow dashed line in the same figure) from the basic 3D point cloud data P by filtering. Then, the model generation unit 102a removes the point cloud PG corresponding to trees extracted by filtering from the basic 3D point cloud data P to generate the processed 3D point cloud data P' shown by reference numeral 1002 in Figure 10.

[0074] The model generation unit 102a may also extract point clouds to be filtered based on values ​​other than elevation values ​​at each of the multiple points that make up the basic 3D point cloud data. For example, the model generation unit 102a may extract point clouds to be filtered based on the slope value between two adjacent points. In this case, the second criterion value will be a slope value that can be arbitrarily set by the user.

[0075] The slope value is the angle between the straight-line distance between two adjacent points and the horizontal straight-line distance between those two points. The horizontal straight-line distance between any two adjacent points is the same regardless of which two points (assuming they are adjacent) are extracted from the basic 3D point cloud data. Therefore, the slope value is essentially determined by the difference in elevation values ​​between two adjacent points.

[0076] <Summary> If there are many target grid GTs with elevation values ​​above a predetermined value, the temperature values ​​of multiple grids obtained from infrared images will converge (saturate) to a certain value, making it difficult to accurately identify the first correlation.

[0077] In this respect, the information processing device 1a has a model generation unit 102a that removes factors that lower the accuracy of identifying the first correlation from the basic 3D point cloud data through filtering. Therefore, the first correlation can be identified with high accuracy, and the shape of the ground surface can be accurately grasped.

[0078] [Third Embodiment] A third embodiment of the present invention will be described below. As shown in Figure 1, the information processing device 1b according to the third embodiment of the present invention includes an estimation device 300 instead of estimation devices 100 and 200. In this respect, the information processing device 1b differs from the information processing devices 1 and 1a according to the first and second embodiments of the present invention.

[0079] <Characteristic Configuration of the Estimation Device> Referring to Figure 1, the characteristic configuration of the estimation device 300 according to the third embodiment of the present invention will be described. Unlike the estimation devices 100 and 200, the estimation device 300 has a solar radiation estimation unit 106, a difference calculation unit 107, a temperature value identification unit 104a, and an altitude value estimation unit 105a, as shown in Figure 1. The estimation device 300 is the same as the estimation device 100 in that it has an information acquisition unit 101, a model generation unit 102, and a change amount calculation unit 103.

[0080] The solar radiation estimation unit 106 estimates the solar radiation in the second grid as the estimated solar radiation (see Figure 12). Here, "solar radiation in the second grid" refers to the average value of the solar radiation in each point cloud, which is aggregated and transformed into an arbitrary second grid from among multiple points that constitute the 3D point cloud data of the surface layer that forms the basis of the target DSM. Furthermore, "solar radiation" in this specification refers to the maximum amount of solar radiation that can illuminate the grid of interest during the time period when the aircraft 2 captures the visible light image, and is calculated based on the solar orbit defined by the latitude, time of year (season), etc., of the target land.

[0081] The difference calculation unit 107 calculates the difference between the reference quantity and the estimated solar radiation. The reference quantity is a solar radiation that can be arbitrarily set by the user, and may be stored in advance in the storage unit 13, or it may be set by the input unit 11 receiving user input. Details of the reference quantity will be described later.

[0082] The temperature value determination unit 104a corrects the determined temperature value using the difference calculated by the difference calculation unit 107. Other functions of the temperature value determination unit 104a are the same as those of the temperature value determination unit 104. The altitude value estimation unit 105a considers the corrected temperature value corrected by the temperature value determination unit 104a as the temperature value in the first grid and determines the first correlation relationship (specifically, calculates the first relational expression). Other functions of the altitude value estimation unit 105a are the same as those of the altitude value estimation unit 105.

[0083] <Main processing functions of information processing equipment> Referring to Figures 11 to 18, the main processing of the information processing device 1b will be explained. As a prerequisite, it is assumed that the aircraft 2 has transmitted the third visible light image, the fourth visible light image Iv4, and the second infrared image Ii2 to the information processing device 1.

[0084] The third visible light image (not shown; light information) is a visible light image of the target land LT' taken by the flying object 2 at a third time point, which is an arbitrary time point. The fourth visible light image Iv4 (light information) is a visible light image of the target land LT' taken by the flying object 2 at a fourth time point, which is a time point later than the first time point.

[0085] The second infrared image Ii2 (infrared image) is an infrared image of the target land LT' captured by the aircraft 2 at the fourth time point. In this embodiment, the second infrared image Ii2 is in a state where the temperature values ​​T' in each first grid G1' are imaged by SfM-MVS processing and superimposed on the image portion of the target land LT' in the second infrared image Ii2 (see Figure 11). This SfM-MVS processing is performed in the aircraft 2, but may also be performed in, for example, the temperature value identification unit 104a.

[0086] In this embodiment, the third time point is defined as the point immediately after mowing the target land LT'. Therefore, at the fourth time point, the vegetation on the target land LT' has grown compared to the third time point. In other words, at the fourth time point, the elevation value is higher in at least a portion of the surface compared to the third time point.

[0087] First, the processing steps S601 to S604 in the flowchart shown in Figure 12 are the same as the processing steps S501 to S504 in the flowchart shown in Figure 3. However, in S601, the information acquisition unit 101 also transmits the acquired second infrared image Ii2 to the solar radiation estimation unit 106.

[0088] Furthermore, in S602, the model generation unit 102 generates a first target DSM (not shown) and a second target DSM 31. The model generation unit 102 then transmits the generated second target DSM 31 to the solar radiation estimation unit 106. In this embodiment, the first target DSM is a numerical surface model generated as a three-dimensional model of the surface of the target land LT' at the third time point. The second target DSM 31 (see Figure 14, etc.) is a numerical surface model generated as a three-dimensional model of the surface of the target land LT' at the fourth time point.

[0089] Next, in S605, the solar radiation estimation unit 106 estimates the estimated solar radiation SE in the second grid G2' using the acquired second infrared image Ii2 and the second target DSM31. The second grid G2' (see Figure 14) is the grid corresponding to the grid of interest GT'' among the multiple grids obtained by dividing the second target DSM31 into a grid. The grid of interest GT'' is the grid of interest in the target land LT'.

[0090] In this embodiment, the solar radiation estimation unit 106 first estimates a basic estimated solar radiation SE' for each of the multiple second grids G2' constituting the second target DSM31, and then estimates an estimated solar radiation SE using the basic estimated solar radiation SE'.

[0091] The basic estimated solar radiation SE', like the estimated solar radiation SE, is an estimate of the solar radiation in the second grid G2'. There are no particular limitations on the estimation method of the basic estimated solar radiation SE' by the solar radiation estimation unit 106, and various known methods can be adopted. For known estimation methods, please refer to the literature "Fu, P., and PM Rich. 2000. The Solar Analyst 1.0 Manual. Helios Environmental Modeling Institute (HEMI), USA" and "Rich, PM, R. Dubayah, WAHetrick, and SCSaving. 1994. Using Viewshed Models to Calculate Intercepted Solar Radiation: Applications in Ecology. American Society for Photogrammetry and Remote Sensing Technical Papers, pp 524-529."

[0092] Then, the solar radiation amount estimation unit 106 extracts first grids G1' with the same temperature value T' from among the plurality of first grids G1' that make up the second infrared image Ii2. For each set of second grids G2' (hereinafter, "corresponding set") corresponding to a set of first grids G1' with the same temperature value T' (hereinafter, "reference set"), the solar radiation amount estimation unit 106 calculates the average value R of the basic estimated solar radiation amount SE'. ave to calculate.

[0093] Then, the solar radiation amount estimation unit 106 generates a graph as shown in FIG. 13 showing the correlation between the temperature value T' and the average value R ave . Using the generated graph, the solar radiation amount estimation unit 106 calculates, as the first basic approximation formula, an approximation formula that approximates the distribution tendency of each point plotted on the graph. The following formula (2) is an example of the first basic approximation formula. In the following formula (2), T' is the temperature value [°C] in the first grid G1', and R ave is the average value [Wh / m 2 of the basic estimated solar radiation amount SE'. Each coefficient (58.121, -74.501) in the following formula (2) is a value that can vary depending on the reflectance (albedo) of the target land LT', the estimated environment, and the like.

[0094] (Equation 2) R ave = 58.12ln(T') - 74.501 ···(2) Then, the solar radiation amount estimation unit 106 uses the average value R obtained by substituting the temperature value T' into the first basic approximation formula as the estimated solar radiation amount SE in the second grid G' corresponding to the first grid G1' in which the temperature value T' is stored. That is, the solar radiation amount estimation unit 106 uses the first basic approximation formula as an estimation formula to calculate the estimated solar radiation amount SE. ave

[0095] The following formula (3) is an example of an estimation formula and is an example when the first basic approximation formula represented by the above formula (2) is used as the estimation formula. In the following formula (3), R i is the estimated solar radiation amount SE [Wh / m 2 in the i-th second grid G2', and T i ​is the temperature value T'[°C] in the first grid G1' corresponding to the i-th second grid G2'. "i" is any natural number from 1 to the total number of second grids G2'.

[0096] (Math 3) R i =58.12ln(T i )-74.501 ···(3) In this way, the solar radiation estimation unit 106 estimates the estimated solar radiation SE by substituting the temperature value T' at the first grid G1' into the estimation formula and using the resulting value as the estimated solar radiation SE. The solar radiation estimation unit 106 then transmits the estimated solar radiation SE to the difference calculation unit 107.

[0097] In this embodiment, when the input unit 11 receives user input, the solar radiation estimation unit 106 transmits the second target DSM 31 and the estimated solar radiation SE to the display unit 12. The display unit 12 then displays the second target DSM 31 with the image of the estimated solar radiation SE superimposed on it, as shown in Figure 14.

[0098] Next, in S606, the difference calculation unit 107 uses the acquired estimated solar radiation SE to calculate the difference ΔS between the reference quantity and the estimated solar radiation SE in the second grid G2'. The difference calculation unit 107 applies the same reference quantity to all second grids G2'. In this embodiment, the difference calculation unit 107 calculates the average value R of all the values ​​calculated by the solar radiation estimation unit 106. ave The largest value among them is obtained from the solar radiation estimation unit 106 and used as the reference quantity. For example, the temperature value T' and the average value R ave If the correlation with the reference amount is as shown in the graph in Figure 13, the reference amount is approximately 84 [Wh / m³]. 2 ]

[0099] The difference calculation unit 107 then transmits the calculated difference ΔS to the temperature value identification unit 104a. In this embodiment, when the input unit 11 receives user input, the difference calculation unit 107 transmits the second target DSM 31 and the calculated difference ΔS to the display unit 12. The display unit 12 then displays the second target DSM 31 with the imaged difference ΔS superimposed, as shown in Figure 15.

[0100] Next, in S607, the temperature value identification unit 104a corrects the temperature value T' at the first grid G1' using the acquired difference ΔS.

[0101] Here, we assume that in the target land LT', there are no geographical features that block sunlight, such as trees, and the surface of the vegetation that constitutes the target land surface has a nearly uniform elevation, in which case the solar radiation distribution and temperature distribution on the target land surface will both be nearly uniform. In this case, the mean value R ave From the maximum value (i.e., the reference quantity) of any mean R ave The value obtained by subtracting the average solar radiation loss R in any corresponding set is l This becomes equal to (hereinafter referred to as "the first relation").

[0102] Average solar radiation loss R l This value represents the average of the lost solar radiation in all second grids G2' belonging to the corresponding set. The lost solar radiation is the amount of solar radiation lost by any grid GT'' due to the influence of features present in that grid GT''. In this case, "any grid GT''" refers to any grid GT'' that corresponds to any second grid G2' belonging to the corresponding set.

[0103] Also, the mean R ave The value obtained by subtracting the temperature value T' in an arbitrary reference set from the temperature value T' when it is maximized is the loss temperature value T in an arbitrary reference set. l It becomes equal to (hereinafter referred to as "the second relationship"). Loss temperature value T lThis represents the temperature value lost by any grid GT'' due to the influence of features present in that grid GT''. In this case, "any grid GT''" refers to any grid GT'' corresponding to any first grid G1'' belonging to the reference set.

[0104] In this embodiment, the temperature value determination unit 104a corrects the temperature value T' using the first and second relationships, based on the assumptions described above. The relationship between the temperature value T' and the average value R is described below. ave Let's take the example of a case where the correlation between the two factors is as shown in the graph in Figure 13.

[0105] First, the temperature value identification unit 104a uses the first relationship to determine the average solar radiation loss R for each corresponding set. l The temperature value identification unit 104a calculates the loss temperature value T for each reference set using the second relationship. l The temperature value identification unit 104a calculates the average solar radiation loss R based on the calculation result. l and loss temperature value T l A graph like Figure 16 is generated, showing the correlation between the two.

[0106] Then, the temperature value identification unit 104a uses the generated graph to calculate a second basic approximation formula that approximates the distribution trend of each point plotted on the graph. Equation (4) below is calculated by comparing the temperature value T' with the mean value R ave This is an example of the second basic approximation formula when the correlation with is as shown in the graph in Figure 13. In equation (4) below, T l is the loss temperature value [°C] in the reference set, and R l This is the average solar radiation loss in the corresponding set [Wh / m²]. 2 ]

[0107] (Math 4) T l =-0.0022R l 2 +0.3R l +0.2502 ···(4) Then, the temperature value identification unit 104a refers to the first relationship and calculates the average value R aveThe value obtained by subtracting the estimated solar radiation SE at an arbitrary second grid G2'' from the maximum value (i.e., the reference quantity) is estimated to be the solar radiation loss at that arbitrary second grid G2''. In other words, the temperature value identification unit 104a estimates that the difference ΔS corresponds to the solar radiation loss (hereinafter, "first estimation").

[0108] Furthermore, the temperature value identification unit 104a refers to the second relationship and calculates the average value R ave It is estimated that the value obtained by subtracting the temperature value T' at any first grid G1' from the temperature value T' when is maximized will be the loss temperature value at that arbitrary first grid G1' (hereinafter, "second estimation"). The "loss temperature value at the first grid G1'" is the loss temperature value T in the reference set to which the first grid G1' belongs. l This is the same as, and below, "Loss temperature value T in the first grid G1'" l It is written as "".

[0109] To rephrase this second estimation, the temperature value T' in the first grid G1' is equal to the loss temperature value T. l By adding the mean R ave It is estimated that the temperature value T' at which is maximized, that is, the temperature value T' obtained after removing the influence of features present in the target grid GT'' corresponding to the first grid G1'', is obtained. From this, the temperature value identification unit 104a determines the loss temperature value T in the first grid G1''. l The correction value cT is used to correct the temperature value T' in the first grid G1'. i Let's assume that.

[0110] Then, the temperature value identification unit 104a calculates a second basic approximation formula based on the first and second estimations, using the lost solar radiation (i.e., the difference ΔS) in the second grid G2' and the correction value cT in the first grid G1' corresponding to the second grid G2'. i This is considered an equation that shows the correlation between and . In other words, the temperature value identification unit 104a is the correction value cT i The second basic approximation formula is used as the formula for calculating . Hereafter, the aforementioned correlation will be referred to as the "second correlation," and the formula that shows the second correlation will be referred to as the "second relation formula."

[0111] From the above, the average solar radiation loss R l and loss temperature value T l The graph in Figure 16, which shows the correlation between the difference ΔS in the second grid G2' and the correction value cT in the first grid G1' corresponding to the second grid G2', is shown. i This can be reinterpreted as a graph showing the second correlation between and .

[0112] Equation (5) below is an example of the second relation, and is an example of replacing the second basic approximation formula expressed in equation (4) above with the second relation. In equation (5) below, lR i This is the difference ΔS[Wh / m] in the i-th second grid G2'. 2 ] and cT i is the correction value [°C] in the first grid G1' corresponding to the i-th second grid G2'. "i" is any natural number from 1 to the total number of second grids G2'.

[0113] (Math 5) cT i =-0.0022lR i 2 +0.3lR i +0.2502 ···(5) Then, the temperature value determination unit 104a determines the correction value cT obtained by substituting the difference ΔS in the second grid G2' into the second relational equation. i This value is added to the temperature value T' in the first grid G1' corresponding to the second grid G2'. The temperature value identification unit 104a uses the value obtained by this addition as the corrected temperature value T'' in the first grid G1'.

[0114] Then, the temperature value identification unit 104a transmits the corrected temperature value T'' to the altitude value estimation unit 105a. In this embodiment, when the input unit 11 receives user operation, the temperature value identification unit 104a receives the imaged corrected value cT iThe temperature value identification unit 104a superimposes the image to generate a second infrared image Ii2. The temperature value identification unit 104a transmits the superimposed second infrared image Ii2 to the display unit 12, and the display unit 12 displays the superimposed second infrared image Ii2 as shown in Figure 17. Furthermore, in this embodiment, when the input unit 11 receives user operation, the temperature value identification unit 104a generates a second infrared image Ii2 in which the corrected imaged temperature value T'' is superimposed. The temperature value identification unit 104a transmits the superimposed second infrared image Ii2 to the display unit 12, and the display unit 12 displays the superimposed second infrared image Ii2 as shown in Figure 18.

[0115] Next, in S608, the elevation estimation unit 105a identifies the correlation between the elevation change ΔE and the corrected temperature value T'' as the first correlation. In other words, the elevation estimation unit 105a identifies the first correlation by considering the corrected temperature value T'' in one grid G1'' as the temperature value T'' in the first grid G1''. The other processing in S608 is the same as the processing in S505.

[0116] Next, the processing content of S609 is the same as that of S506. Upon completion of processing S609, all major processing performed by the information processing device 1b is completed.

[0117] <Variation> In correcting the temperature value T' by the temperature value determination unit 104a, it is not essential to use the second relational expression. The temperature value determination unit 104a may, for example, correct the temperature value T' using a second estimation model instead of the second relational expression.

[0118] The second estimation model is a trained model constructed by machine learning using the second basic dataset as training data, with the difference (loss of solar radiation) as input data and the correction value (loss of temperature value) as output data. There are no particular limitations on the machine learning method, and known methods such as convolutional neural networks can be used. The second basic dataset is a dataset consisting of past differences and past correction values. The past differences are the differences in each of the multiple second grids constituting the target DSM at a time before the first time point. The past correction values ​​are the correction values ​​in the first grid corresponding to each of the multiple second grids mentioned above, and correspond to the loss of temperature value calculated by the temperature value identification unit 104a.

[0119] In this case, the temperature value identification unit 104a inputs the difference ΔS in the second grid G2' to the second estimation model, thereby determining the correction value cT in the first grid G1' corresponding to the second grid G2'. i The second estimation model is used to obtain the value cT. Then, the temperature value identification unit 104a uses the correction value cT obtained from the second estimation model. i This is added to the temperature value T' in the first grid G1'. The temperature value identification unit 104a sets the value obtained by this addition as the corrected temperature value T'' in the first grid G1'.

[0120] <Summary> Depending on the time and period in which the second infrared image Ii2 was captured, and the surrounding environment of the ground surface and features targeted for information acquisition by the information acquisition unit 101, variations may occur in the amount of solar radiation received at different locations on the ground surface and features. This variation causes the influence of features not to be reflected in the temperature value T' in the first grid G1' identified by the temperature value identification unit 104a, and consequently, the elevation value estimation unit 105a becomes unable to identify the first correlation relationship that accurately reflects the influence of features.

[0121] In this regard, the information processing device 1b, specifically the elevation value estimation unit 105a, identifies the first correlation by considering the corrected temperature value T'' in the first grid G1'' as the temperature value T'' in the first grid G1''. This eliminates the adverse effect of variations in solar radiation on the accuracy of identifying the first correlation, making it possible to identify the first correlation that accurately reflects the influence of geological features. Therefore, the shape of the ground surface can be accurately grasped with simple processing.

[0122] According to the above configuration, the correction value cT obtained by substituting the difference ΔS at the second grid G2' into the second relational equation is used for the temperature value T' at the first grid G1'. i By adding this value, the temperature value T' in the first grid G1' can be corrected. Therefore, since the first correlation that accurately reflects the influence of geological features can be identified, the shape of the ground surface can be accurately grasped with simple processing.

[0123] [Examples of implementation using software] The functions of the estimation devices 100 to 300 (hereinafter abbreviated as "devices") can be realized by programs that cause a computer to function as the device, and by programs that cause a computer to function as each control block of the device (especially each part included in the control unit 15).

[0124] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., memory) as hardware for executing the program. By executing the program using this control device and storage device, each of the functions described in each of the embodiments is realized.

[0125] The program may be recorded on one or more computer-readable recording media, not temporary ones. These recording media may or may not be provided by the device. In the latter case, the program may be supplied to the device via any wired or wireless transmission medium.

[0126] Furthermore, some or all of the functions of each control block can also be implemented by logic circuits. For example, an integrated circuit in which logic circuits functioning as each of the control blocks are formed is also included in the scope of the present invention. In addition, it is also possible to implement the functions of each control block using, for example, a quantum computer.

[0127] [Additional Notes] The present invention is not limited to the embodiments described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of symbols]

[0128] 3.31 Second Target DSM (Numerical Surface Model) 100, 200, 300 estimator 101 Information Acquisition Department 102, 102a Model generation unit 103 Change Amount Calculation Unit 104, 104a Temperature value identification section 105, 105a Elevation Estimation Section 106 Solar radiation estimation section 107 Difference calculation part G1, G1' First Grid G2, G2' 2nd grid GK' Virtual Grid GT, GT', GT'' Grid to watch II1 First infrared image (infrared image) II2 Second infrared image (infrared image) Iv1 1st visible light image (light information) Iv2 2nd visible light image (light information) Iv4 4th visible light image (light information) P: Basic 3D point cloud data (3D point cloud data of the surface that forms the basis of the numerical surface model) PG point cloud (points extracted by filtering) P' Processed 3D point cloud data (3D point cloud data with points extracted by filtering removed) SE estimated solar radiation T, T' temperature values T´´ Corrected temperature value ΔE: Change in elevation (change in elevation in the second grid) ΔS difference

Claims

1. An estimation device for estimating the shape of the ground surface by virtually dividing the ground surface and geographical features into a grid, thereby creating multiple unit spaces, each of which is designated as a virtual grid, and selecting any of these virtual grids as the grid of interest, wherein An information acquisition unit acquires an infrared image of the target ground surface, which is the ground surface to be estimated, and the aforementioned features, and optical information of the reflected light reflected by the target ground surface and the aforementioned features. A temperature value identification unit identifies the temperature value in a first grid corresponding to the grid of interest among a plurality of grids obtained by dividing the infrared image into a grid, A model generation unit generates a numerical surface model as a three-dimensional model of the surface layer composed of the target ground surface and the surface of the object, using the aforementioned optical information. A change amount calculation unit, which calculates the change in elevation in the second grid, takes the value obtained by subtracting the first reference value from the first elevation value, which is the elevation value in the second grid corresponding to the grid of interest, among the multiple grids obtained by dividing the numerical surface model into a grid, An estimation device comprising: an elevation value estimation unit that identifies the correlation between the amount of change in elevation in the second grid and the temperature value in the first grid as a first correlation, and estimates the elevation value in the third grid corresponding to the grid of interest, among a plurality of grids obtained by dividing the numerical elevation model, which is a three-dimensional model of the target ground surface, into a grid, as a second elevation value based on the first correlation.

2. The elevation value estimation unit is, We calculate the first relational equation, which is the equation showing the first correlation, The estimation device according to claim 1, wherein the second altitude value is the value obtained by subtracting the amount of change obtained by substituting the temperature value into the first relational expression from the first altitude value.

3. The aforementioned model generation unit, From among the multiple points that constitute the three-dimensional point cloud data of the surface surface which forms the basis of the numerical surface model, a filtering process is performed to extract points whose elevation value is equal to or greater than the second reference value. The estimation apparatus according to claim 1 or 2, which generates the numerical surface model from three-dimensional point cloud data from which the points extracted by the filtering described above have been removed.

4. A solar radiation estimation unit that estimates the amount of solar radiation in the second grid as estimated solar radiation, The system further comprises a difference calculation unit that calculates the difference between a reference amount and the estimated solar radiation, The temperature value determination unit corrects the temperature value in the first grid using the difference, The estimation device according to claim 1, wherein the elevation value estimation unit considers the corrected temperature value in the first grid as the temperature value in the first grid to determine the first correlation.

5. The temperature value determination unit is, A correction value is calculated to correct the temperature value in the first grid. A second correlation equation is calculated as the second relational equation, which is the correlation between the difference in the second grid and the correction value in the first grid. The estimation device according to claim 4, wherein the corrected temperature value is obtained by adding the correction value obtained by substituting the difference into the second relational expression to the temperature value.

6. A program for causing a computer to function as an estimation device according to claim 1, wherein the computer functions as the information acquisition unit, the temperature value identification unit, the model generation unit, the change amount calculation unit, and the altitude value estimation unit.

7. An estimation method for estimating the shape of the ground surface by virtually dividing the ground surface and geographic features into a grid, thereby designating each of the multiple unit spaces obtained as a virtual grid, and designating any of the multiple virtual grids as the grid of interest, wherein An information acquisition step that acquires an infrared image of the target ground surface, which is the ground surface to be estimated, and the features, and optical information of the reflected light reflected by the target ground surface and the features, A temperature value identification step involves dividing the infrared image acquired in the information acquisition step into a grid and identifying the temperature value in the first grid corresponding to the grid of interest among a plurality of grids obtained by this division, A model generation step in which a numerical surface model is generated as a three-dimensional model of the surface layer composed of the target ground surface and the surface of the object, using the optical information acquired in the information acquisition step, A change amount calculation step in which the numerical surface model generated in the model generation step is divided into a grid to obtain a plurality of grids, and the value obtained by subtracting the first reference value from the first elevation value, which is the elevation value in the second grid corresponding to the grid of interest, is defined as the change amount of elevation in the second grid. An estimation method comprising: a step of determining the correlation between the amount of change in elevation in the second grid calculated in the change amount calculation step and the temperature value in the first grid, which is identified as a first correlation; and an elevation value estimation step of determining the elevation value in the third grid corresponding to the grid of interest, from among a plurality of grids obtained by dividing the numerical elevation model, which is a three-dimensional model of the target ground surface, into a grid, as the second elevation value based on the first correlation.

Images