KR20210129472A - 3d spatial object location correction method based on road linearity - Google Patents

Abstract

The present invention relates to a 3D spatial object location correction method based on road linearity. The method includes the steps of: (a) generating a 3D spatial object model and a 2D orthogonal correction image for a specific area based on an image captured during the flight of an unmanned aerial vehicle; (b) generating a road linearity vector by extracting a road line from the 2D orthogonal correction image based on a learning model about a road pattern; (c) determining road linearity information corresponding to the road line from pre-established spatial information for the specific area; (d) generating transformation information for matching between the road linearity vector and the road linearity information; and (e) performing location correction by applying the transformation information to the 3D spatial object model and the 2D orthogonal correction image, respectively.

Description

3D spatial object position correction method based on road alignment {3D SPATIAL OBJECT LOCATION CORRECTION METHOD BASED ON ROAD LINEARITY}

The present invention relates to a three-dimensional spatial object position correction technology, and more particularly, to derive a matching coefficient by comparing features extracted from raw data with pre-established spatial information data, and to apply a three-dimensional spatial object generated through UAV. It relates to a method for correcting the position of a 3D spatial object based on a road alignment that can be applied.

Various value-added information using high-resolution UAV (Unmanned Aerial Vehicle) images is attracting attention. This is because, unlike satellites that are photographed from hundreds of kilometers in the sky and aircraft that are photographed from several kilometers, the UAV recording altitude is lowered to around 100 m, so images with a resolution of several centimeters can be acquired. However, the area covered by one image is narrow, and since a digital camera or an old camera with similar specifications is generally mounted, there is a disadvantage in that a distortion phenomenon occurs outside the acquired image.

Therefore, in order to use the UAV image with aerial images or satellite images, it is necessary to correct the undulation displacement generated according to the height of the topography/feature in each image and go through a registration process.

However, the UAV shakes more than satellites or aircraft when shooting, and the accuracy of GPS (Global Positioning System) and IMU (Inertial Measurement Unit) sensors is relatively low, so there is a high probability that external expression elements are inaccurate.

In order to improve the positioning accuracy and correct the geometry, a ground control point (GCP) is generally used. However, there may exist cases in which shooting is required faster than setting a ground reference point, such as shooting in a disaster situation, or shooting is not possible depending on topographical conditions.

Korean Patent No. 10-1160454 (2012.06.21)

An embodiment of the present invention is to provide a method for correcting the position of a three-dimensional spatial object based on a road alignment that can be applied to a three-dimensional spatial object by deriving a matching coefficient by comparing features extracted from raw data and spatial information data. .

An embodiment of the present invention can minimize the matching error between the realistic 3D spatial object and the national spatial information and improve the positioning error of the low-cost 3D building object by performing the position correction of the 3D spatial object using the road alignment. We intend to provide a method for correcting the position of a three-dimensional spatial object based on a road alignment.

Among the embodiments, the method for correcting the position of a three-dimensional spatial object based on a road alignment includes the steps of (a) generating a 3D spatial object model and a 2D orthogonal correction image for a specific area based on an image captured during the flight of an unmanned aerial vehicle; (b) generating a road line vector by extracting a road line from the 2D orthogonal correction image based on a learning model about the pattern of the road, (c) from the spatial information previously built for the specific area to the road line determining corresponding road alignment information; (d) generating transformation information for matching between the road alignment vector and the road alignment information; and (e) the 3D spatial object model and the 2D orthogonal correction image. and performing position correction by applying each of the transformation information.

In the step (a), when the photographed image is composed of a plurality of partial images photographed at equal intervals, generating the 2D orthographic correction image as one image relating to the specific region by merging the plurality of partial images may include.

The step (c) may include extracting and storing only the layer corresponding to the road line from among a plurality of layers included in the numerical map of the specific area.

The step (d) includes (d1) determining a plurality of matching points within the specific area, (d2) extracting first position coordinates corresponding to each of the plurality of matching points from the road linear vector; (d3) extracting second position coordinates corresponding to each of the plurality of matching points from the road alignment information, and (d4) a transformation matrix for forward transformation from the second position coordinates to the first position coordinates It may include the step of generating.

The step (d1) may include determining, as the matching point, a change point including at least one of cutoff, reduction, and expansion of a road based on road information included in the numerical map of the specific area.

The step (d4) may include generating coefficients of the polynomial as the transformation matrix in a second-order polynomial transformation using the first position coordinate as an independent variable and the second position coordinate as a dependent variable.

In step (e), (e1) extracting a 3D spatial object and 2D position coordinates from the 3D spatial object model, respectively, (e2) applying the reverse transformation of the forward transformation to the 2D position coordinates to obtain 2D correction coordinates and (e3) generating a corrected 3D spatial object model by integrating the 3D spatial object and the 2D correction coordinates.

The 3D spatial object position correction method includes (f) a distance error before and after position correction for the 2D orthogonal correction image based on a plurality of ground control points (GCPs) preset for the specific region. ) may further include verifying the accuracy of the position correction by comparing.

The disclosed technology may have the following effects. However, this does not mean that a specific embodiment should include all of the following effects or only the following effects, so the scope of the disclosed technology should not be construed as being limited thereby.

The method for correcting the position of a three-dimensional spatial object based on a road alignment according to an embodiment of the present invention can compare features extracted from raw data and spatial information data to derive a matching coefficient and apply it to a three-dimensional spatial object.

The method for correcting the position of a 3D spatial object based on a road alignment according to an embodiment of the present invention minimizes the matching error between the realistic 3D spatial object and the national spatial information by performing the position correction of the 3D spatial object using the road alignment. and can improve the alignment error of low-cost 3D building objects.

1 is a block diagram illustrating a three-dimensional spatial object position correction system according to the present invention.
FIG. 2 is a view for explaining a physical configuration of the position correction device of FIG. 1 .
FIG. 3 is a view for explaining a functional configuration of the position correction device of FIG. 1 .
4 is a flowchart illustrating a process of correcting the position of a three-dimensional spatial object based on a road alignment according to the present invention.
5 is a conceptual diagram illustrating an embodiment of a method for correcting a position of a three-dimensional spatial object based on a road alignment.
6 is a view for explaining an embodiment of data format conversion of 3D point cloud data according to the present invention.
7 is a view for explaining an embodiment of the position coordinates corresponding to the matching point according to the present invention.
8 is a view for explaining an embodiment of a distance error before and after position correction according to the present invention.
9 is a flowchart illustrating a method for correcting the position of a three-dimensional spatial object based on a road alignment according to the present invention.

Since the description of the present invention is merely an embodiment for structural or functional description, the scope of the present invention should not be construed as being limited by the embodiment described in the text. That is, since the embodiment may have various changes and may have various forms, it should be understood that the scope of the present invention includes equivalents capable of realizing the technical idea. In addition, since the object or effect presented in the present invention does not mean that a specific embodiment should include all of them or only such effects, it should not be understood that the scope of the present invention is limited thereby.

On the other hand, the meaning of the terms described in the present application should be understood as follows.

Terms such as “first” and “second” are for distinguishing one component from another, and the scope of rights should not be limited by these terms. For example, a first component may be termed a second component, and similarly, a second component may also be termed a first component.

When a component is referred to as being “connected” to another component, it may be directly connected to the other component, but it should be understood that other components may exist in between. On the other hand, when it is mentioned that a certain element is "directly connected" to another element, it should be understood that the other element does not exist in the middle. Meanwhile, other expressions describing the relationship between elements, that is, “between” and “immediately between” or “neighboring to” and “directly adjacent to”, etc., should be interpreted similarly.

The singular expression is to be understood to include the plural expression unless the context clearly dictates otherwise, and terms such as "comprises" or "have" refer to the embodied feature, number, step, action, component, part or these It is intended to indicate that a combination exists, and it should be understood that it does not preclude the possibility of the existence or addition of one or more other features or numbers, steps, operations, components, parts, or combinations thereof.

Identifiers (eg, a, b, c, etc.) in each step are used for convenience of description, and the identification code does not describe the order of each step, and each step clearly indicates a specific order in context. Unless otherwise specified, it may occur in a different order from the specified order. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

The present invention can be embodied as computer-readable codes on a computer-readable recording medium, and the computer-readable recording medium includes all types of recording devices in which data readable by a computer system is stored. . Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like. In addition, the computer-readable recording medium may be distributed in a network-connected computer system, and the computer-readable code may be stored and executed in a distributed manner.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Terms defined in general used in the dictionary should be interpreted as being consistent with the meaning in the context of the related art, and cannot be interpreted as having an ideal or excessively formal meaning unless explicitly defined in the present application.

1 is a block diagram illustrating a three-dimensional spatial object position correction system according to the present invention.

Referring to FIG. 1 , the 3D spatial object position correction system 100 may include an unmanned aerial vehicle 110 , a position correction device 130 , and a database 150 .

The unmanned aerial vehicle 110 may correspond to an aircraft (aircraft) capable of flying without an actual pilot on board. For example, the unmanned aerial vehicle 110 may include an unmanned aerial vehicle (UAV) or a drone. The unmanned aerial vehicle 110 may be implemented by including various sensors for automatic flight, and may include various modules for flight capability and control capability. That is, the unmanned aerial vehicle 110 may be implemented including a control and ground support system, a communication and navigation system, and the like.

In addition, the unmanned aerial vehicle 110 may be connected to the position correction device 130 through a network, and the plurality of unmanned aerial vehicles 110 may be simultaneously connected to the position correction device 130 . The unmanned aerial vehicle 110 may be implemented by including various camera devices capable of photographing the ground during flight, and may transmit a captured image to the position correction device 130 in real time during flight. The unmanned aerial vehicle 110 may perform an operation of correcting and matching the basic displacement occurring according to the height of the topography/feature in the captured image in order to utilize the captured image as an aerial image or satellite image, and for this purpose, image processing It may be implemented including software.

The position correction device 130 may be implemented as a server corresponding to a computer or program capable of deriving a matching coefficient by comparing features extracted from raw data with spatial information data and processing the position correction of a three-dimensional spatial object. The location correction device 130 may be wirelessly connected to the unmanned aerial vehicle 110 through Bluetooth, WiFi, a communication network, etc., and may exchange data with the unmanned aerial vehicle 110 through the network.

In an embodiment, the position correction apparatus 130 may store data necessary for correcting the position of a three-dimensional spatial object based on a road alignment in conjunction with the database 150 . Meanwhile, unlike FIG. 1 , the position correction device 130 may be implemented by including the database 150 therein. In addition, the position correction device 130 may be implemented including a processor, a memory, a user input/output unit, and a network input/output unit, which will be described in more detail with reference to FIG. 2 .

The database 150 may correspond to a storage device for storing various types of information required in the operation process of the position correction device 130 . The database 150 may store information about an image captured by the unmanned aerial vehicle 110 , and may store various spatial information data for position correction, but is not necessarily limited thereto, and the position correction device 130 is unmanned. Information collected or processed in various forms may be stored in the process of generating a three-dimensional spatial object model based on the image captured by the vehicle 110 and performing position correction.

FIG. 2 is a view for explaining a physical configuration of the position correction device of FIG. 1 .

Referring to FIG. 2 , the position correction device 130 may be implemented including a processor 210 , a memory 230 , a user input/output unit 250 , and a network input/output unit 270 .

The processor 210 may execute a procedure for processing each step in the process in which the position correction device 130 operates, and manage the memory 230 that is read or written throughout the process, and the memory 230 ) can schedule the synchronization time between volatile and nonvolatile memory in The processor 210 may control the overall operation of the position correction device 130 , and is electrically connected to the memory 230 , the user input/output unit 250 , and the network input/output unit 270 to control the data flow therebetween. can The processor 210 may be implemented as a central processing unit (CPU) of the position correction device 130 .

The memory 230 is implemented as a non-volatile memory, such as a solid state drive (SSD) or a hard disk drive (HDD), and may include an auxiliary storage device used to store overall data required for the position correction device 130, It may include a main memory implemented as a volatile memory such as random access memory (RAM).

The user input/output unit 250 may include an environment for receiving a user input and an environment for outputting specific information to the user. For example, the user input/output unit 250 may include an input device including an adapter such as a touch pad, a touch screen, an on-screen keyboard, or a pointing device, and an output device including an adapter such as a monitor or a touch screen. In an embodiment, the user input/output unit 250 may correspond to a computing device accessed through a remote connection, and in such a case, the location correction device 130 may be implemented as a server.

The network input/output unit 270 includes an environment for connecting with an external device or system through a network, for example, a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), and a VAN (Wide Area Network) (VAN). It may include an adapter for communication such as Value Added Network).

FIG. 3 is a view for explaining a functional configuration of the position correction device of FIG. 1 .

Referring to FIG. 3 , the position correction device 130 includes an image processing unit 310 , a matching road alignment generating unit 320 , a reference road alignment determining unit 330 , a conversion information generating unit 340 , and a position correction performing unit ( 350 ), an accuracy verification unit 360 , and a control unit 370 .

The image processing unit 310 may generate a 3D spatial object model and a 2D orthogonal correction image for a specific region based on the image captured during the flight of the unmanned aerial vehicle 110 . Here, the 3D spatial object model may correspond to a 3D spatial model corresponding to an area photographed during the flight process of the unmanned aerial vehicle 110, and the 2D orthogonal correction image is a undulation by center projection with respect to an image photographed from above. It may correspond to a two-dimensional image in which displacement and displacement due to camera posture are removed. Accordingly, the 2D orthographic correction image may correspond to a two-dimensional image having characteristics such as a map.

Meanwhile, the image processing unit 310 may receive and utilize the 3D spatial object model independently generated by the unmanned aerial vehicle 110 and the 2D orthogonal correction image, and in this case, the 3D spatial object model is 3D point cloud data. may correspond to

In an embodiment, the image processing unit 310 generates a 2D orthogonal correction image as one image for a specific region by merging a plurality of partial images when the captured image is composed of a plurality of partial images photographed at equal intervals. can The unmanned aerial vehicle 110 may periodically photograph an image in the process of flying over a specific area, and thus may generate a plurality of partial images photographed at equal intervals. The image processing unit 310 may receive a plurality of partial images as original data, and may generate one integrated orthogonal corrected image through bundle adjustment and image matching.

The matching road alignment generator 320 may generate a road linear vector by extracting a road line from the 2D orthogonal correction image based on a learning model related to the pattern of the road. To this end, the position correction device 130 may perform learning about the application pattern based on aerial images and actual road information of various regions, and as a result of learning, a specific image as input data and a road included in the image It is possible to create a learning model that generates information about lines as output data. The matching road linearity generation unit 320 may acquire vector-type data about a road in a region corresponding to the 2D orthogonal correction image by using the built-up learning model.

The reference road alignment determiner 330 may determine road alignment information corresponding to the road line from spatial information previously established for a specific area. To this end, the location correction device 130 can access the related pre-established spatial information in conjunction with the road name address guidance system, which is an external system, and can keep the database 150 up to date through periodic collection and update. .

In an embodiment, the reference road alignment determiner 330 may extract and store only a layer corresponding to a road line from among a plurality of layers included in the numerical map of a specific area. For example, the reference road alignment determining unit 330 selects a road line (road) as actual information about a road in a specific area from among various information provided by a numerical map (in particular, a road name map) of national spatial information data. It can be determined as road alignment information.

The transformation information generator 340 may generate transformation information for matching between the road linear vector and the road linear information. That is, the road linear vector may be used as matching data, and the road linear information may be used as reference data. Roads are included in most images and can be easily identified, and since there is not much change compared to buildings, it can be suitable as a standard for position correction in that extraction and matching point detection are easy even in automation development.

Accordingly, in order to convert the 2D orthogonal correction image to the reference image, the position correction apparatus 130 may select a road with a low possibility of change from among the information included in each image as a reference, and the transformation information generating unit 340 may As characteristic information about the road, the road linear vector and the road linear information may be used respectively.

In one embodiment, the transformation information generating unit 340 determines a plurality of matching points within a specific area, extracts first position coordinates corresponding to each of the plurality of matching points from the road linear vector, and from the road linear information It is possible to extract second position coordinates corresponding to each of the plurality of matching points, and generate a transformation matrix for forward transformation from the second position coordinates to the first position coordinates. In this case, the matching point has a clear characteristic in the image, and thus a location that is easier to identify than the surroundings can be determined.

More specifically, the conversion information generating unit 340 may determine a point where identification with the surroundings is clear based on a road within a specific area as a matching point, and a road linear vector of a 2D orthogonal correction image and a road of a pre-established spatial information It is possible to extract the position coordinates corresponding to the matching points from each of the linear information. In this case, the location coordinates may correspond to x and y coordinates on a plane, and a location coordinate table may be generated as coordinate information about the matching point. The transformation information generator 340 may utilize an nth-order polynomial transformation to derive a transformation rule between position coordinates.

In an embodiment, the conversion information generating unit 340 may determine a change point including at least one of cutoff, reduction, and expansion of a road as a matching point based on road information included in the numerical map of a specific area. The conversion information generator 340 may determine a point at which an abrupt change in road information is sensed as a matching point. For example, a point where the continuity of road information is interrupted due to a break in the road, a point at which a rate of change of road information exceeds a preset threshold value according to a reduction or expansion of a road width, etc. may be determined as a matching point.

In an embodiment, the transformation information generating unit 340 may generate coefficients of the corresponding polynomial as a transformation matrix in a second-order polynomial transformation using the first positional coordinate as the independent variable and the second positional coordinate as the dependent variable. The transformation information generator 340 may derive matching information using an nth-order polynomial transformation, and the number of matching points determined in consideration of the fact that ^{(n+1) 2 matching points are required for the nth-order polynomial transformation} It is possible to determine the order of the polynomial transformation based on . For example, when a total of 9 matching points are determined, the transform information generator 340 may select any one of a first-order polynomial and a second-order polynomial (see FIG. 7 ).

Meanwhile, the transform information generator 340 may determine the coefficients Kx(i,j) and Ky(i,j) of the polynomial function using the least squares estimate, which may be calculated using the following equation. .

[Equation]

,

을 생성할 수 있다.Here, Kx and Ky are used as input values for the polynomial, so that the coordinates xo and yo in the two-dimensional orthogonal correction image may be mapped to the coordinates xi and yi in the actual road information. The transform information generator 340 may generate the polynomial coefficients Kx and Ky as a transform matrix having a size of 2*2 in the second-order polynomial transform. In FIG. 7 , the transformation information generation unit 340 is a transformation matrix, respectively.

,

can create

The position correction performing unit 350 may perform position correction by applying transformation information to the 3D spatial object model and the 2D orthogonal correction image, respectively. The position correction performing unit 350 may perform an independent position correction operation on the 3D spatial object model and the 2D orthogonal correction image, and in the case of the 3D spatial object model, an operation of extracting only information for geometric correction may be performed separately. can The position correction performing unit 350 may perform geometric correction on the 2D orthogonal corrected image by using the transformation matrix generated by the transformation information generating unit 340 .

In one embodiment, the position correction performing unit 350 extracts the 3D spatial object and the 2D position coordinates from the 3D spatial object model, respectively, and applies the reverse transformation of the forward transformation to the 2D position coordinates to generate 2D correction coordinates, A corrected 3D spatial object model can be created by integrating 3D spatial objects and 2D correction coordinates. The 3D spatial object model may be expressed as 3D point cloud data, and the position correction performing unit 350 sequentially performs an operation for extracting the building area and outer point based on this and a texturing operation. You can create 3D spatial objects. In this case, the 3D spatial object may correspond to a 3D building model as a spatial model of a building existing in a specific area.

In addition, the position correction performing unit 350 through data format conversion based on the 3D point cloud data 6 of 'x', 'y', 'z', 'color red', 'color green' and 'color blue' Data having attribute values of dog items can be generated, and attribute values of 'x' and 'y' items can be extracted as 2D position coordinates from among them. That is, the position correction performing unit 350 may generate 2D correction coordinates as a result of position correction by applying a transformation matrix to the 'x' and 'y' items of the 3D spatial object model (see FIG. 6 ).

Also, the position correction performing unit 350 may generate a corrected 3D spatial object model by integrating the 3D spatial object and the 2D correction coordinates. The position correction performing unit 350 performs position correction using a transformation matrix for 'x' and 'y' among the attribute values of six items obtained through data format conversion, and the remaining 'z' and 'color red' , 'color green' and 'color blue' can be kept as they are. The position correction performing unit 350 may generate a corrected 3D spatial object model by replacing the existing attribute values with corrected 'x' and 'y' and then acquiring 3D point cloud data through data format conversion.

The accuracy verification unit 360 compares the distance error before and after the position correction for the 2D orthogonal correction image based on a plurality of ground control points (GCPs) preset for a specific region to correct the position. accuracy can be verified. In this case, the plurality of ground reference points may be set based on an integrated reference point set near a point where an image is captured for a specific area. The accuracy verification unit 360 may verify the accuracy based on a distance error between the actual coordinates of the ground reference point and the position coordinates of the point corresponding to the ground reference point in the 2D orthogonal corrected image.

In FIG. 8 , when a total of six GCPs are set, the difference with criteria coordinates before and after position correction can be checked. In this case, the distance error may be calculated as a straight line distance based on the ground reference points (GCPs). That is, it can be confirmed that the distance, which showed an average difference of about 34.54 m before correction, is reduced to about 1.21 m after correction through position correction according to the present invention.

Meanwhile, the accuracy verification unit 360 may determine that the accuracy verification has been successful when, as a result of the accuracy verification for the 2D orthogonal corrected image, the distance error after the correction is less than the distance error before the correction. In addition, the position correction apparatus 130 may determine whether the accuracy verification is successful based on the reduction rate of the distance error before and after the correction through the accuracy verifying unit 360 . That is, if the difference between the distance error before and after the correction is not large, since the efficiency of the position correction is not high, the position correction device 130 performs accuracy verification based on the 2D orthogonal correction image and then, according to the result, the 3D spatial object model It is possible to determine whether to perform the position correction operation for the .

The control unit 370 controls the overall operation of the position correction device 130, and includes an image processing unit 310, a matching road alignment generating unit 320, a reference road alignment determining unit 330, a conversion information generating unit 340, A control flow or data flow between the position correction performing unit 350 and the accuracy verifying unit 360 may be managed.

4 is a flowchart illustrating a process of correcting the position of a three-dimensional spatial object based on a road alignment according to the present invention.

Referring to FIG. 4 , the position correction device 130 generates a 3D spatial object model and a 2D orthogonal correction image for a specific region based on an image captured during the flight process of the unmanned aerial vehicle 110 through the image processing unit 310 . It can be done (step S410). The position correction apparatus 130 may generate a road linear vector by extracting a road line from the 2D orthogonal correction image based on the learning model regarding the pattern of the road through the matching road linear generating unit 320 (step S420).

Also, the location correction apparatus 130 may determine the road alignment information corresponding to the road line from the spatial information previously established for a specific area through the reference road alignment determiner 330 (step S430). The position correction apparatus 130 may generate transformation information for matching between the road linear vector and the road linear information through the transformation information generating unit 340 (step S440).

Also, the position correction apparatus 130 may perform position correction by applying transformation information to each of the 3D spatial object model and the 2D orthogonal correction image through the position correction performing unit 350 (step S450). The position correction device 130 is a distance error (distance) before and after position correction for the 2D orthogonal correction image based on a plurality of ground control points (GCP) preset for a specific area through the accuracy verification unit 360 . error) to verify the accuracy of the position correction (step S460).

5 is a conceptual diagram illustrating an embodiment of a method for correcting the position of a three-dimensional spatial object based on a road alignment.

Referring to FIG. 5 , the position correction device 130 may improve accuracy through position correction for 3D point cloud data generated based on an image captured in a state where a ground reference point (GCP) is not set by the unmanned aerial vehicle. have. To this end, the location correction device 130 may utilize road information among the vector files provided by the public data portal as reference data for improving location accuracy.

That is, the position correction device 130 may first perform geometric correction on the two-dimensional orthogonal correction mosaic obtained based on the captured image, and apply the transformation matrix calculated in the process to the three-dimensional point cloud to perform position correction. can be done In this case, the transformation matrix may be derived as forward transformation (M) information through matching between the road linear vector extracted from the 2D orthogonal correction mosaic and the road information extracted from the national spatial information data. The position correction apparatus 130 may perform position correction by applying the inverse transformation (M-1) of the forward transformation (M) to each of the 3D point cloud and the 3D orthographic correction mosaic (step S510).

9 is a flowchart illustrating a method for correcting the position of a three-dimensional spatial object based on a road alignment according to the present invention.

Referring to FIG. 9 , the position correction device 130 may perform image correction on an image captured in a state in which the ground reference point (GCP) is not set by the drone (Drone) ( Image correction (EO)).

The location correction device 130 may use national spatial data provided by a public data portal as comparison data for improving location accuracy. That is, the position correction apparatus 130 may calculate matching information (Matching Coefficient) for position correction by comparing the two-dimensional information obtained by image correction based on the captured image with the national space-based data. For example, the position correction apparatus 130 may use the transformation matrix calculated in the corresponding process as matching information for position correction.

On the other hand, the position correction device 130 may extract a 3D point cloud based on the image correction result (Point cloud extraction), and may generate 3D space objects (BO) based on the 3D point cloud. The position correction apparatus 130 may perform position correction on the 3D spatial object by applying matching information to the 3D spatial object (step S910).

Therefore, the position correction device 130 according to the present invention provides an improvement in the position accuracy of the 2D and 3D images of the UAV image obtained without selecting a ground reference point, thereby enabling linkage and compatibility with other spatial information data, etc. It is possible to provide the effect of expanding the range of use of 3D spatial objects obtained from cloud data.

Although the above has been described with reference to preferred embodiments of the present invention, those skilled in the art can variously modify and change the present invention within the scope without departing from the spirit and scope of the present invention as set forth in the claims below. You will understand that it can be done.

100: 3D spatial object position correction system
110: unmanned aerial vehicle 130: positioning device
150: database
210: processor 230: memory
250: user input/output unit 270: network input/output unit
310: image processing unit 320: matching road linear generation unit
330: reference road alignment determining unit 340: transformation information generating unit
350: position correction performing unit 360: accuracy verification unit
370: control unit

Claims (8)

(a) generating a 3D spatial object model and a 2D orthogonal correction image for a specific area based on the image captured during the flight process of the unmanned aerial vehicle;
(b) generating a road linear vector by extracting a road line from the 2D orthogonal correction image based on a learning model about a road pattern;
(c) determining road alignment information corresponding to the road line from the spatial information previously established for the specific area;
(d) generating transformation information for matching between the road linear vector and the road linear information; and
(e) performing position correction by applying the transformation information to the 3D spatial object model and the 2D orthogonal correction image, respectively.
According to claim 1, wherein the step (a)
and generating the 2D orthogonal correction image as one image related to the specific region by merging the plurality of partial images when the captured image is composed of a plurality of partial images photographed at equal intervals. A method of correcting the position of a three-dimensional spatial object based on road alignment.
According to claim 1, wherein step (c) is
and extracting and storing only a layer corresponding to the road line from among a plurality of layers included in the numerical map of the specific area.
According to claim 1, wherein step (d) is
(d1) determining a plurality of matching points within the specific area;
(d2) extracting first position coordinates corresponding to each of the plurality of matching points from the road linear vector;
(d3) extracting second location coordinates corresponding to each of the plurality of matching points from the road alignment information; and
(d4) generating a transformation matrix for forward transformation from the second position coordinates to the first position coordinates.
5. The method of claim 4, wherein (d1) step
and determining, as the matching point, a change point including at least one of cutoff, reduction, and expansion of a road based on road information included in the numerical map of the specific area as the matching point. How to correct the position of a spatial object.
The method of claim 4, wherein step (d4) is
and generating coefficients of the polynomial as the transformation matrix in a second-order polynomial transformation using the first position coordinate as an independent variable and the second position coordinate as a dependent variable. How to correct the position of a spatial object.
5. The method of claim 4, wherein step (e) is
(e1) extracting a 3D spatial object and a 2D location coordinate from the 3D spatial object model, respectively;
(e2) generating 2D correction coordinates by applying an inverse transformation of the forward transformation to the 2D position coordinates; and
(e3) generating a corrected 3D spatial object model by integrating the 3D spatial object and the 2D correction coordinates.
According to claim 1,
(f) by comparing the distance error before and after the position correction of the 2D orthogonal correction image based on a plurality of ground control points (GCP) preset for the specific region, the position correction A method of correcting the position of a three-dimensional spatial object based on a road alignment, characterized in that it further comprises the step of verifying the accuracy.

Images