JP2015125092A - Consistency determination method of measurement result and consistency determination apparatus of measurement result - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a consistency determination method of a measurement result and a consistency determination apparatus of a measurement result in which when consistency is determined regarding two and more different results of measuring a same area, determination elements can include a range in a horizontal direction in addition of a range in a height direction.SOLUTION: A consistency determination method of a measurement result determines consistency by comparing two measurement point groups obtained in a same area. The method includes a topographic quantity calculation process and a topographic image generation process. In the topographic quantity calculation process a tomographic quantity is calculated for each mesh constituting a topographic model, and in the topographic image generation process a topographic image based on the topographic quantity is generated. And, consistency of two measurement point groups is determined by collating both topographic images.

Description

  The present invention relates to a technique for comparing two measurement point groups acquired for the same region and determining the consistency thereof, and more specifically, based on the amount of topography obtained from two measurement results. In particular, the present invention relates to a measurement result consistency determination method and a measurement result consistency determination apparatus that determine the consistency between the images by comparing the images.

  In the past, when performing topographic measurements over a wide range, such as creating topographic maps, it was common to use aerial photographs taken from aircraft, but nowadays, aerial laser measurements, measurements using satellite photographs, Alternatively, various measurement methods such as measurement using synthetic aperture radar have been put into practical use, and a suitable method can be selected according to the situation.

  Among these, as shown in FIG. 1, the aviation laser measurement is performed by flying the aircraft P over the terrain F to be measured and receiving a reflection signal of the laser pulse L irradiated to the terrain F. Since the aircraft P is normally equipped with a positioning device such as GPS (Global Positioning System) and an inertial measurement device such as IMU (Internal Measurement Unit), the irradiation position (x, y, z) and irradiation of the laser pulse L The posture (ω, φ, κ) can be grasped, and as a result, the three-dimensional coordinates of the measurement point (the point where the laser pulse L is reflected) can be obtained from the time difference between the irradiation time and the reception time.

  The laser pulse L reflected by the terrain F is received by a sensor mounted on the aircraft P, acquires the intensity of the laser pulse L when returning (hereinafter referred to as “reflection intensity”), and is combined with the reception time. To be recorded. This reflection intensity is the so-called energy level of the received laser pulse L, and is directly measured as a voltage, and the energy level can be obtained by converting the voltage. As shown in FIG. 1, a large number of laser pulses L are irradiated in one measurement (flight), and the number of irradiation intensities corresponding to the laser pulses L is recorded.

  Aviation laser measurement usually covers a wide range, and therefore covers a range to be measured by flying a plurality of courses (routes). Then, the adjacent courses are measured to some extent (side wrap) so that the measurement omission area does not occur. In other words, in the side lap range, two types of measurements are performed on different courses.

  By the way, it is known that aviation laser measurement has an error depending on equipment such as GPS and IMU and an error due to the laser pulse L passing through the atmosphere. Therefore, there is a difference between the two types of measurement results obtained in the side wrap range, that is, the two measurement results may not match. Therefore, when performing aerial laser measurement, cross-course verification is performed to determine consistency by comparing two measurement results in the side lap range. If it is determined that they do not match, both measurement results are aligned. Coordination between courses is performed.

  The results obtained from aero laser measurements are a large number of discrete points. When comparing two types of measurement results consisting of such point clouds, the comparison has been made after drawing the topography based on point clouds such as cross-sections and vertical sections. Two types of terrain are compared, and the difference in height at the same plane position is seen. In other words, the consistency between the two was determined only by the height difference among the measurement differences, and the mismatch due to the horizontal difference was not included in the determination.

  When performing inter-course verification, it is desirable to perform consistency determination after including a horizontal difference in addition to a height difference. However, not only between course verifications but also when comparing measurement results of two periods, to compare two discrete point clouds as in aerial laser measurement, compare the heights with reference to the plane position Was the mainstream.

  Therefore, in Patent Document 1, when comparing the terrain of two periods, a terrain model is created from discrete point clouds, the terrain amount is calculated for each mesh constituting the terrain model, and an image corresponding to the terrain amount is displayed. Propose to create. By imaging the topography of the two periods, the partial topographic change becomes clear, and a three-dimensional (plane direction and height direction) displacement vector can be obtained by the effect of creating a topographic model. That is, not only the change in the height direction but also the change in the plane direction can be confirmed.

JP 2010-266419 A

  Patent Document 1 is a suitable technique for grasping a change in topography in two periods, but does not propose a technique used for inter-course verification. In other words, it is based on the premise that the measurement results of the two periods are probable (within an allowable error range), and the difference between the two types of measurement results that should be consistent with each other assuming that various errors occur. It is not a technique for determining (or the extent). An object of the present invention is to solve the above-mentioned problem, that is, when performing consistency determination for two or more different results obtained by measuring the same region, in addition to the difference in the height direction, the difference in the horizontal direction is also determined. An object of the present invention is to provide a measurement result consistency determination method and a measurement result consistency determination apparatus that can be included in an element.

  The present invention uses a topographic model based on three-dimensional spatial information, such as DSM (Digital Surface Model) and DEM (Digital Elevation Model), when determining consistency between two types of measurement results that should be consistent, such as cross-course verification. The invention is based on an idea that has not been used in the past, focusing on the fact that it is used and compared based on the calculated amount of topography.

  The measurement result consistency determination method of the present invention is a method for determining consistency by comparing two measurement point groups obtained in the same region, and includes a terrain amount calculation step and a terrain image creation step. It is. In practicing the present invention, two models composed of a plurality of meshes are prepared. One is a first terrain model created based on the first measurement point group, and the other is a second terrain model created based on the second measurement point group. In the terrain amount calculation step, the terrain amount is calculated for each mesh constituting the first terrain model, and the terrain amount is calculated for each mesh constituting the second terrain model. In the terrain image creation step, a first terrain image obtained by imaging the first terrain model based on the terrain amount and a second terrain image obtained by imaging the second terrain model based on the terrain amount are also used. Create Then, the consistency between the two measurement point groups is determined by performing image matching between the first terrain image and the second terrain image.

  The measurement result consistency determination method according to the present invention may further include a partial verification step, a difference vector calculation step, and a mismatch region detection step. In the partial collation step, the first topographic image is divided to obtain a plurality of search areas, and image collation is performed based on each search area to extract a collation area from the second topographic image. In the difference vector calculation step, a difference vector is calculated from the search area and the collation area. In the inconsistent area detecting step, an inconsistent area in which the first measurement point group and the second measurement point group are not matched is detected based on the difference vector.

  The apparatus for determining consistency of measurement results according to the present invention includes a terrain model reading unit, a terrain amount calculating unit, a terrain image creating unit, and an image collating unit. The terrain model reading means reads two models composed of a plurality of meshes. That is, a first terrain model created based on the first measurement point group and a second terrain model created based on the second measurement point group. The terrain amount calculation means calculates the terrain amount for each mesh constituting the first terrain model, and calculates the terrain amount for each mesh constituting the second terrain model. The terrain image creating means creates a first terrain image obtained by imaging the first terrain model based on the amount of terrain, and creates a second terrain image obtained by imaging the second terrain model. The image collating means collates the first terrain image and the second terrain image. Then, the consistency of the two measurement point groups is determined based on the result of image collation.

  The measurement result consistency determination apparatus according to the present invention may further include a partial verification unit, a difference vector calculation unit, and an inconsistency region detection unit. The partial matching unit divides the first terrain image to obtain a plurality of search areas, and performs image matching based on each search area to extract a matching area from the second terrain image. The difference vector calculation means calculates a difference vector from the search area and the collation area. The inconsistent area detecting means detects an inconsistent area in which the first measurement point group and the second measurement point group are not matched based on the difference vector.

The measurement result consistency determination method and measurement result consistency determination apparatus of the present invention have the following effects.
(1) For example, when performing inter-course verification, in addition to the difference in the height direction, the difference in the horizontal direction can be included in the consistency determination. As a result, more appropriate course-to-course verification can be performed, and as a result, a high-quality measurement result can be obtained.
(2) The consistency can be determined easily and appropriately even in a forest or the like that does not have a clear calibration site.
(3) Since sex determination is performed by image matching, inconsistent portions can be clearly extracted.

Explanatory drawing which shows the measurement condition by an aviation laser. The model figure explaining "surface". The model figure explaining "the ground surface". The block diagram which shows the structure of a consistency determination apparatus. (A) is a model diagram explaining a search area, (b) is a model diagram explaining a search range. The model figure which shows the "area | region equivalent to a search area" and the "collation area | region" in a 2nd topographic image. The graph figure which shows the view which correct | amends the position of a collation area | region. Explanatory drawing which shows the template elected from the search area | region. (A) is the model figure which showed the image of the search area | region, (b) is the model figure which showed the 2nd topographic image. The flowchart which shows the process sequence performed with a consistency determination apparatus.

  An example of embodiments of a measurement result consistency determination method and a measurement result consistency determination apparatus according to the present invention will be described with reference to the drawings.

1. Overall Outline The present invention is mainly used for measurement results carried out over a certain wide range (hereinafter referred to as “target region”), that is, uses three-dimensional spatial information acquired from a wide range. Therefore, first, three-dimensional spatial information will be described.

  The three-dimensional spatial information is information composed of points, lines, surfaces, or combinations thereof having plane coordinate values and height information. Further, the plane coordinate value is represented by latitude and longitude or X and Y coordinates, and the height means a vertical distance from a predetermined reference horizontal plane such as an altitude. This three-dimensional spatial information can be created by various means. For example, it can be created based on a set of two stereo aerial photographs (satellite photographs), can be created by aerial laser measurement or satellite radar measurement, or can be created by surveying the site directly.

  The three-dimensional spatial information created on the basis of stereo aerial photographs and aerial laser measurements usually represents a “surface”. Here, as shown in FIG. 2, the surface means the upper surface of a feature that stands on the ground, such as a green object such as a forest or farmland, or a building. On the other hand, the “ground surface” means a surface after removing green objects or buildings, that is, the ground as shown in FIG. Here, the “surface” is represented by three-dimensional spatial information, the “surface model”, and the “surface” is represented by three-dimensional spatial information, “the ground model”. Note that DSM is known as a typical surface layer model, and DEM is known as a typical surface model.

  In order to create the ground surface model, processing for removing green objects and buildings from the surface layer model, so-called filtering processing is performed. The filtering process is a well-known technique, and recognizes and excludes objects that meet a predetermined condition as green objects or buildings. This process is generally executed by a computer using general-purpose software.

  Next, an outline of the present invention will be described. In the present invention, a “terrain model” created based on a surface layer model or a ground surface model (or a surface layer model or a ground surface model itself) is used. This terrain model includes a large number of grids and meshes separated by the grids. In other words, the terrain model is composed of a plurality of meshes. In addition, since the present invention determines consistency for two or more different results obtained by measuring the same region, two different terrain models are used. For convenience, one of the two terrain models is referred to as a “first terrain model” and the other is referred to as a “second terrain model”.

  In order to determine consistency between two or more different results, the first terrain model and the second terrain model are compared. At this time, both terrain models are imaged and compared. In order to image the terrain model, the terrain amount is calculated for each mesh, and pixel information corresponding to the terrain amount is given. By comparing the image of the first terrain model and the image of the second terrain model, a difference amount and a direction of the terrain (hereinafter referred to as a “difference vector”) are obtained, and matching is performed based on the difference vector. Sex determination can be made.

  Hereinafter, each element will be described in detail. Here, the technical contents of the present invention will be described using an example of a measurement result consistency determination apparatus (hereinafter simply referred to as “consistency determination apparatus”), and the contents specific to the measurement result consistency determination method will be described. This will be explained later.

2. Consistency judgment device (configuration means)
FIG. 4 is a block diagram illustrating a configuration of the consistency determination device 100. As shown in this figure, the consistency determination apparatus 100 is constituted by various means, and these means are preferably executed by using a computer. Hereinafter, each means will be described in detail. The input unit 110 gives an instruction to execute processing to various units described later, and a keyboard, a pointing device, or the like can be used.

  The terrain model storage means 120 shown in FIG. 4 stores the first terrain model and the second terrain model, and is a storage medium such as a computer hard disk or CD-ROM. That is, these terrain models are formed in a data format that can be processed by a computer. The terrain model storage unit 120 may be prepared separately for the first terrain model and for the second terrain model, or may be configured as an integral unit. Further, as shown in FIG. 4, data registration means 130 is provided, and the data registration means 130 can store the first terrain model and the second terrain model in a hard disk of a computer or the like.

  As described above, the terrain model is composed of a plurality of meshes. The mesh is formed by being divided into, for example, orthogonal grids, and each mesh has a representative point. Since the laser measurement point is usually random data, geometric calculation is often performed to give a height to the representative point of the mesh. As this calculation method, in addition to a method by TIN (Triangulated Irregular Network) for obtaining height by an irregular triangle network formed from random data, a method by nearest neighbor method (Nearest Neibor) employing the nearest laser measurement point 4, Various conventionally used methods such as an inverse distance weighting method (IWD), a Kriging method, and an averaging method can be employed. In addition, the grid which comprises a mesh does not necessarily need to orthogonally cross, and can be made into the grid by arbitrary crossing angles. In addition, the first terrain model and the second terrain model used here may be created for the present invention, and of course, if there is a ready-made one, it can be used.

  The terrain model reading unit 140 reads the first terrain model and the second terrain model from the terrain model storage unit 120 in accordance with the instruction input by the input unit 110. The terrain amount calculation unit 150 calculates the terrain amount for each mesh of the terrain model read by the terrain model reading unit 140. Here, the terrain amount is a so-called index representing the feature of the terrain, and is calculated based on the representative points of the mesh, the original random data, and the like.

  Examples of the terrain amount include an altitude value, a Laplacian value, a ground opening value, an underground opening value, an inclination amount, or a combination thereof. Here, the Laplacian value is generally used for drawing a Laplacian diagram that represents the rate of change in inclination. This Laplacian diagram has a feature that it is positive in a depressed terrain, negative in a protruding terrain, and has an absolute value where the change in the terrain is large.

  The ground opening value and the underground opening value are generally for drawing an opening diagram. Among the opening diagrams, the ground opening diagram represents the extent of the sky that can be seen within a certain distance from the point of interest, and the ground opening value increases as the point protrudes from the surroundings. The ridge shows a large ground opening value, and as a result, it has a feature that the protruding peak and ridge are emphasized. On the other hand, in the opening map, the underground opening map shows the size of the basement within a certain distance when looking down from the ground surface, contrary to the ground opening map. The underground opening value indicates a large value, for example, a large underground opening value is indicated at a depression or a valley bottom, and as a result, the depression or valley is emphasized.

  The tilt amount is generally used for drawing a tilt amount diagram. The inclination amount map indicates the degree of inclination of the terrain. The larger the inclination, the larger the inclination value, and vice versa.

  The terrain image creating unit 160 creates a terrain image based on the terrain amount calculated by the terrain amount calculating unit 150. Specifically, the pixel corresponding to the terrain amount with respect to pixels (pixels) for image creation. A terrain image is created by adding information and collecting pixels having the pixel information. This pixel is formed on the basis of a mesh. For example, one pixel is one pixel, four meshes are one pixel, and nine meshes are one pixel. Can be formed.

  Pixel information given to a pixel represents brightness and color (hue, saturation, and brightness), and color models such as RGB, CMYK, and NCS can be used. Pixel information is given according to the terrain amount of the mesh. For example, red, orange, yellow, green, and blue are set in the order in which the terrain amount shows a large value. The altitude value can be arbitrarily set, for example, the altitude value can be a color and the slope amount can be a combination of brightness.

  By the above procedure, a “first terrain image” is created based on the first terrain model, and a “second terrain image” is created based on the second terrain model. When the first terrain image and the second terrain image are created, one is selected as the “search image” and the other is selected as the “searched image”. Here, for the sake of convenience, the case where the first terrain image is used as a search image and the second terrain image is used as a search target image will be described.

  The image matching unit 170 performs image matching (matching) between the first terrain image and the second terrain image created by the terrain image creation unit 160. At this time, matching can be performed for the entire terrain image, but the terrain image can be divided into several parts, and matching can be performed in divided units. This is executed by the partial matching means 171. Specifically, the “search area” that is a predetermined partial area is cut out from the first terrain image that is the search image, and the second terrain that is the search target image. The image is scanned, and a “collation area” that matches the search area is detected.

  5A and 5B are diagrams for explaining a partial search by the partial matching means 171. FIG. 5A is a model diagram for explaining a search area, and FIG. 5B is a model diagram for explaining a search range. As shown in FIG. 5A, the partial matching unit 171 divides the first topographic image into regions of an arbitrary size, cuts out a plurality of search regions, and performs matching in each search region. At this time, the second topographic image is scanned, but only a specific range of the second topographic image can be scanned. In this range, an area corresponding to a search area cut out from the first terrain image is extracted from the second terrain image, and an area expanded by a predetermined width (buffer) around the area (hereinafter referred to as a search area) "Searchable range"). By scanning only this search range, matching can be performed efficiently.

  When performing matching by the search area, scanning is performed while moving the search area in units of one pixel within the range of the second topographic image (or within the search range). That is, the matching area matched with the search area is configured by pixels of the second terrain image, that is, the mesh of the second terrain model, and the matching area is surrounded by the grid of the terrain model. FIG. 6 is a model diagram showing an “area corresponding to a search area” and a “collation area” in the second topographic image. However, the area most matched with the search area is not necessarily located at the position surrounded by the grid.

  FIG. 7 is a graph showing the idea of correcting the position of the collation area. In this graph, the horizontal axis is the XY plane shown in FIG. 6, and the vertical axis is the sum of squares of the difference in the height direction (the sum of squares in the matching region). In this graph, the horizontal axis is originally represented by the XY plane, and the square sum of the difference is also plotted in a plane, but here, for convenience, it is represented by a cross section of Y = 0. From this figure, the difference is the smallest between when the search area is moved by -1 in the X direction and when it is moved by -2 in the X direction, and it is appropriate to extract the collation area at this position. I understand. In this way, it is possible to extract the collation region position by correcting from the grid position.

When matching by the search area, a template composed of a combination of a plurality of pixels can be used. This template is formed by a plurality of pixels selected from the search area, and a template that forms a particularly characteristic shape is suitable. FIG. 8 is an explanatory diagram showing the template 200 selected from the search area. The extracted template 200 is scanned to find a similar image in the second terrain image. Here, the reason for matching by the template 200 will be described. FIG. 9A is a model diagram showing an image of a search area, and FIG. 9B is a model diagram showing a second topographic image. Focusing on one pixel V ₀ in FIG. 9A and searching for a corresponding pixel in FIG. 9B, pixel V ₁ , pixel V ₂ , and pixel V having the same luminance (brightness) are obtained. ₃ is matched. In other words, three patterns of the difference of the pixel V ₀ can be considered and cannot be specified as any one. On the other hand, by focusing on the template 200 comprised of four pixels including the pixel V ₁ of FIG. 9 (a) (a range in surrounded by the broken line drawing), search for images corresponding thereto among shown in FIG. 9 (b) Then, the template 200 surrounded by the broken line in FIG. 9B can be specified. In this way, using the template 200 makes it easier to collate two images.

  When the template 200 is extracted, the template 200 is scanned to find a similar image in the second terrain image. Specifically, based on the pixel information included in the template 200, a combination that has the same pixel information arrangement is searched from the second topographic image.

  Note that the same image as the template 200 does not always exist in the second topographic image. Therefore, a certain degree of redundancy can be given to the degree of collation. For example, a threshold is provided for the residual of the compared pixel information, and if it is within this threshold, it is determined as an equivalent image (pixel), or if the number (or ratio) of different pixels is equal to or less than a predetermined threshold Differences in images can be allowed under various conditions, such as determining to collate with the template 200 or determining by combining them.

  The difference vector calculation means 172 calculates a difference amount and a difference direction based on the combination of the search area and the matching area extracted by the partial matching means 171 to obtain a difference vector. The search area and the collation area are partial areas of the first terrain image and the second terrain image, respectively. Since these terrain images are composed of meshes of the terrain model, three-dimensional coordinate calculation is easily performed. be able to.

  The inconsistent area detection unit 173 detects a partial area that is inconsistent between the first terrain model and the second terrain model according to the difference vector obtained by the difference vector calculation unit 172. If the difference amount of the difference vector exceeds a predetermined threshold value, it is determined that there is a mismatch, and the search area (or matching area) that is the basis of the difference vector is detected as a mismatch area. The mismatched area and the difference vector detected here are output by the output means 180 which is an output device such as a display or a printer.

(Process flow)
Next, the flow of processing executed by the consistency determination device 100 will be described with reference to FIG. First, the terrain model reading unit 140 reads the first terrain model and the second terrain model from the terrain model storage unit 120 that stores the terrain model (Step 11 and Step 12). For each of the read terrain models, the terrain amount calculation unit 150 calculates a terrain amount for each mesh (Step 21 and Step 22), and the terrain image creation unit 160 calculates the first terrain image and the first terrain model based on the terrain amount. 2 terrain images are created (Step 31, Step 32).

  The first collation unit 171 divides the first topographic image into regions of an arbitrary size, and sets a plurality of (here, n) search regions (Step 41). One search area is selected from the n search areas (Step 51), and a search range is set from the second topographic image (Step 52).

  With the selected search area, the search range is scanned in the second topographic image (Step 60), and the collation area to be collated is specified (Step 70). Then, the difference vector calculation means 172 obtains a difference vector based on the search area and the collation area (Step 80), further determines the matching / mismatch of both based on the difference vector, and the search area determined to be inconsistent (Or the collation area) is detected as an inconsistent area by the inconsistent area detecting means 173 (Step 90).

  The above processing is repeatedly performed by the number (n) of the divided search areas (loop), and finally the mismatch area and the difference vector are output by the output unit 180 such as a display or a printer (Step 100). .

3. Measurement Result Consistency Determination Method In the measurement result consistency determination method, the content described in the consistency determination apparatus 100 is executed by a person (operator). Hereinafter, it demonstrates individually. The content overlapping with the consistency determination device 100 will not be described repeatedly here.

  The terrain amount calculation step performs the contents described in the terrain amount calculation means 150, and calculates the terrain amount for each mesh of the terrain model. In the terrain image creation step, the contents described in the terrain image creation means 160 are performed, and a first terrain image and a second terrain image are created based on the amount of terrain. The partial matching step performs the contents described in the partial matching unit 171, sets a search area from the first topographic image, and detects a matching area that matches the search area from the search range of the second topographic image. To do. The movement vector calculation step performs the contents described in the difference vector calculation means 172, and calculates the difference vector based on the difference between the search area and the collation area. The inconsistent area detecting step performs the contents described in the inconsistent area detecting means 173, and based on the difference vector, a partial area that is inconsistent between the first topographic model and the second topographic model is detected. To detect.

  The measurement result consistency determination method and the measurement result consistency determination apparatus according to the present invention can be suitably implemented in course-to-course verification when aviation laser measurement is performed over a wide range. In addition to aerial laser measurement, it can also be used when measuring the same location multiple times using aerial photographs and satellite images. Furthermore, the present invention can also be applied to the case where the consistency of results when the same range is measured by different methods is determined.

DESCRIPTION OF SYMBOLS 100 Consistency determination apparatus 110 Input means 120 Terrain model memory | storage means 130 Data registration means 140 Terrain model reading means 150 Terrain amount calculation means 160 Terrain image creation means 170 Image collation means 171 Partial collation means 172 Difference vector calculation means 173 Inconsistency area detection Means 180 Output means 200 Template F Terrain L Laser P Aircraft

Claims (4)

In a method for determining consistency by comparing two measurement point groups obtained in the same region,
A model composed of a plurality of meshes, and a first terrain model created based on the first measurement point group and a second terrain model created based on the second measurement point group are prepared,
Calculating a terrain amount for each mesh constituting the first terrain model, and calculating a terrain amount for each mesh constituting the second terrain model;
Creating a first terrain image obtained by imaging the first terrain model based on the terrain amount, and creating a second terrain image obtained by imaging the second terrain model; With
A method for determining consistency of measurement results, wherein image consistency between the first terrain image and the second terrain image is used to determine consistency between two measurement point groups. A partial matching step of dividing the first terrain image to obtain a plurality of search areas and performing image matching based on each search area to extract a matching area from the second terrain image;
A difference vector calculating step of calculating a difference vector from the search area and the matching area;
And a mismatch region detection step of detecting a mismatch region in which the first measurement point group and the second measurement point group are not matched based on the difference vector. Item 5. A method for determining consistency of measurement results according to Item 1. In an apparatus for judging consistency by comparing two measurement point groups obtained in the same region,
A terrain model reading means for reading out a first terrain model created based on the first measurement point group and a second terrain model created based on the second measurement point group, the model comprising a plurality of meshes When,
A terrain quantity calculating means for calculating a terrain quantity for each mesh constituting the first terrain model, and calculating a terrain quantity for each mesh constituting the second terrain model;
Terrain image creation means for creating a first terrain image obtained by imaging the first terrain model based on the terrain quantity, and for creating a second terrain image obtained by imaging the second terrain model; ,
Image collating means for image collating the first terrain image and the second terrain image,
An apparatus for determining consistency of measurement results, wherein the consistency between two measurement point groups can be determined based on the result of image collation. The image collating unit includes a partial collating unit, a difference vector calculating unit, and an inconsistent region detecting unit,
The partial matching unit divides the first terrain image to obtain a plurality of search areas, performs image matching based on each search area, and extracts a matching area from the second terrain image,
The difference vector calculating means calculates a difference vector from the search area and the matching area,
4. The mismatch area detection unit detects an inconsistency area where the first measurement point group and the second measurement point group do not match based on the difference vector. Device for determining consistency of the measurement results described.

Images