KR101489468B1 - camera image calibration apparatus and method based on homography calculation - Google Patents

Abstract

The present invention relates to a homography-based camera image correction apparatus and method, and more particularly, to an apparatus and method for correcting a camera image based on homography, So that the homography value can be calculated, thereby improving the user's convenience.

Description

Technical Field [0001] The present invention relates to a homography-based camera image correction apparatus and method,

The present invention relates to an image calibration technique, and more particularly, to a homography-based camera image correction apparatus and method.

Intrinsic parameters and extrinsic parameters must be obtained for camera image calibration. The internal parameter refers to the camera internal information caused by the structural error of the camera caused by lens distortion or the like, the focal length by zooming in or zooming out, and the like.

On the other hand, the external parameter is information on how much the camera is moved from the origin of the real world and how much it is rotated, and this external parameter is not determined when the camera is manufactured.

How much the camera moves from the origin of the real world and how much it rotates can be mathematically represented by a matrix and a vector. The origin of the real world can be arbitrarily set. The three-dimensional coordinates of the real world can be represented by the coordinates of the two-dimensional image plane using a homogeneous coordinate system.

Assuming that the Z axis value of the three-dimensional coordinates of the real world is 0, it becomes a two-dimensional plane having XY coordinates, and this two-dimensional plane can be a two-dimensional image photographed by a camera. You can find the feature points in 2D images and change them as you see them from different angles. This is called homography.

There are a number of methods for acquiring three-dimensional information from an image depending on its accuracy and purpose of use. A typical 3D reconstruction is possible. 3D reconstruction uses three or more cameras or mobile cameras to obtain three-dimensional coordinates corresponding to feature points on an image. However, there are disadvantages that it is difficult to use in a real-time processing because a large number of cameras are used and a calculation amount is large.

If only one stationary camera is used, three-dimensional coordinates can not be obtained. In this case, although not the accurate three-dimensional information, it is assumed that the field of view of the camera is assumed as a plane (referred to as a ground plane) and that the detected object on the ground plane There may be a method of estimating the coordinates of the target object. This can be achieved with a projective transformation between the image plane and the ground plane. This transformation is defined as plane to plane homography as shown in FIG.

1 is a view for explaining homography. Homography H is defined as a 3x3 matrix. The feature point x included in the original image can be converted to x 'by the homography H. This homography technique is disclosed in Korean Patent Publication No. 10-2012-0021666 (Mar. 03, 2012).

A commonly used homography method is the DLT (Direct Linear Transformation) algorithm. DLT can be used to obtain homography when four pairs of corresponding points between two planes are given (exact solution for four and least mean squared error solution for four or more). Even if a camera parameter is given, the homography between the image plane and the ground plane can be calculated using the camera projection as shown in Equation 1 below.

(Equation 1)

Where k is the camera parameter matrix, R is the rotation matrix, t is the translation vector, and r _i is the value of the i th row of the rotation matrix, where z and x are the points on the image plane and the ground plane, respectively.

However, the above methods are not suitable for use in video surveillance systems. In DLT, it is difficult for the user to input the coordinates on the ground plane. In the second method, it is difficult for the user to know the information about the camera installation environment (installation height, tilt angle, focal distance, etc.) accurately. These incompatibilities are all due to the fact that the user is hard to input or difficult to input information on three dimensions.

Therefore, the present inventor has found that, by simply placing an icon for acquiring coordinates on an image including an object of which the actual size is known, without inputting the camera installation information including the camera installation height and angle, the homography value And to study the technology that can calculate.

Korean Patent Publication No. 10-2012-0021666 (March 03, 2012)

Disclosure of the Invention The present invention has been made in view of the above-described circumstances, and it is an object of the present invention to provide an image capturing apparatus and method for capturing an image by simply arranging an icon for acquiring coordinates on an image including an object whose actual size is known, And a homography-based camera image correction apparatus and method capable of calculating a homography value.

According to one aspect of the present invention, there is provided a homography-based camera image correction apparatus including: an image display unit displaying an image including an object whose actual size is photographed by a camera; An icon arrangement unit for arranging an icon for acquiring coordinates at a position of a target included in the image displayed by the image display unit and adjusting the size and angle of the icon according to the size and tilted angle of the target; A size information input unit for inputting actual size information of an object included in the image displayed by the image display unit; A coordinate information acquiring unit for acquiring a coordinate pair including an upper coordinate and a lower coordinate of the icon arranged at the position of the object included in the image by the icon arranging unit; A homography ratio calculation unit for calculating a homogrargy value for camera image correction using size information of at least two objects input by the size information input unit and at least two coordinate pairs obtained by the coordinate information obtaining unit, A calculation unit; And the like.

According to a further aspect of the present invention, the icon arrangement section comprises: an icon storage section for storing an icon for obtaining a coordinate; An icon selection unit for selecting an icon for acquiring coordinates stored in the icon storage unit according to a user operation; An icon overlay unit for overlaying an icon for coordinate acquisition selected by the icon selection unit on a position of an object included in the image according to a user operation; An icon adjusting unit for changing the size and angle of the icon for obtaining the coordinates overlaid on the position of the target included in the image by the icon overlay unit according to the size of the target and the tilted angle according to the user operation; .

According to a further aspect of the present invention, the icon for obtaining the coordinates includes two feature points for obtaining the top coordinate and the bottom coordinate for determining the icon size.

According to a further aspect of the present invention, the homography-based camera image correction apparatus calculates coordinates of the image plane of the image displayed by the image display unit using a homograhy value calculated by the homography calculation unit, An image correcting unit for converting the image into coordinates on a plane and outputting the converted coordinates; And further comprising:

According to still another aspect of the present invention, there is provided a homography-based camera image correction method comprising: displaying an image including an object whose actual size is photographed by a camera; An icon placement step of arranging an icon for acquiring coordinates at a position of a target included in the image displayed by the image display step and adjusting the size and angle of the icon according to the size and tilted angle of the target; A size information input step of inputting actual size information of an object included in the image displayed by the image display step; A coordinate information acquiring step of acquiring a coordinate pair including an upper coordinate and a lower coordinate of the icon arranged at the position of the object included in the image by the icon disposing step; A homography calculation step of calculating a homography value for camera image correction using size information of at least two objects input by the size information input step and at least two coordinate pairs; And the like.

According to a further aspect of the present invention, the homography-based camera image correction method further comprises a step of calculating the coordinates on the image plane of the image displayed by the image display step using the homograhy value calculated by the homography calculation step Converting the coordinates into coordinates on a ground plane and outputting the coordinate;

And further comprising:

In the present invention, a homography value is calculated by simply placing an icon for obtaining coordinates on an image including an object whose actual size is known, without the hassle of inputting camera installation information including a camera installation height and an angle So that user convenience can be improved.

1 is a view for explaining homography.
2 is a block diagram showing the configuration of an embodiment of a homography-based camera image correction apparatus according to the present invention.
3 is a diagram showing an example in which icons for coordinate acquisition are arranged in an image.
4 is a diagram showing a coordinate system of a homography calculation algorithm.
5 is a flowchart illustrating a configuration of an embodiment of a homography-based camera image correction method according to the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout.

In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

The terms used throughout the specification of the present invention have been defined in consideration of the functions of the embodiments of the present invention and can be sufficiently modified according to the intentions and customs of the user or operator. It should be based on the contents of.

2 is a block diagram showing the configuration of an embodiment of a homography-based camera image correction apparatus according to the present invention. 2, a homography-based camera image correction apparatus 100 according to the present invention includes an image display unit 110, an icon arrangement unit 120, a size information input unit 130, (140), and a homography calculation unit (150).

The image display unit 110 displays an image including an object whose actual size is photographed by a camera. To change a two-dimensional image from a different angle, it is necessary to obtain a homography value. As a precondition for obtaining a homography value, the present invention takes an object of which the actual size is known by using a camera, and displays the object through the image display unit 110.

The icon placement unit 120 arranges an icon for obtaining coordinates at a position of an object included in the image displayed by the image display unit 120 and displays the icon corresponding to the angle of the target Adjust the size and angle.

Conventionally, in order to calculate a homography value, it is troublesome to input camera installation information including a camera installation height and an angle. However, in the present invention, without inputting camera installation information including a camera installation height and an angle, It is possible to simply place the icon for acquiring the coordinates on the image including the object whose actual size is known through the arrangement unit 120, thereby avoiding troublesome user input.

At this time, as shown in FIG. 3, the icon for obtaining the coordinates may be embodied to include two feature points for obtaining the top coordinate and the bottom coordinate for determining the icon size. 3 is a diagram showing an example in which icons for coordinate acquisition are arranged in an image.

The icon arranging unit 120 for arranging icons for acquiring coordinates includes an icon storing unit 121, an icon selecting unit 122, an icon overlaying unit 123, and an icon adjusting unit 124 Lt; / RTI >

The icon storage unit 121 stores an icon for obtaining coordinates. For example, the icon for acquiring the coordinates may be an icon capable of adjusting the size and the placement angle of the human figure as shown in FIG.

The icon selecting unit 122 selects an icon for obtaining coordinates stored in the icon storing unit 121 according to a user operation. For example, the icon selecting unit 122 may be configured to provide a user interface for selecting an icon for acquiring coordinates in a template manner, and to select an icon from the user interface according to a user operation through a mouse or the like.

The icon overlaying unit 123 overlays an icon for obtaining a coordinate selected by the icon selecting unit 122 on a position of an object included in the image according to a user operation.

For example, when a user selects an icon from a user interface that selects an icon for coordinate acquisition in a template manner using a mouse, and dragged the selected icon with a mouse to display an image of an object included in the image displayed by the image display unit 120 Position to overlay the icon for coordinate acquisition on the position of the object included in the image.

The icon adjusting unit 124 adjusts the size and angle of the icon for obtaining the coordinates overlaid on the position of the target included in the image by the icon overlay unit 123 according to the size of the target and the tilted angle Change it.

For example, when the user clicks an icon for acquiring the coordinates overlaid on the position of the object included in the image with the mouse, the icon adjusting unit 124 adjusts the size and angle of the icon to the outside of the icon for coordinate acquisition by dragging the mouse And the marker is manipulated by the user using the mouse, the size and angle of the icon for obtaining the coordinates overlaid on the position of the target included in the image are adjusted according to the size and inclined angle of the target Can be implemented.

The size information input unit 130 inputs actual size information of an object included in the image displayed by the image display unit 110. [ Unlike the conventional method, in order to calculate the homography value, the camera installation information including the camera installation height and angle is not input, and the actual size photographed and displayed by the camera through the size information input unit 130 is known Enter the actual size information of the subject

For example, the size information input unit 130 may provide a user interface for inputting the actual size information of the target included in the image, and receive the actual size information of the target included in the image from the user. For example, when the object is a person, the actual size information of the object may be the key size of the object.

The coordinate information obtaining unit 140 obtains a pair of coordinates including an upper coordinate and a lower coordinate of the icon arranged at the position of the object included in the image by the icon arranging unit 120. [

For example, when the coordinate information obtaining unit 140 detects the coordinates of two feature points for obtaining the upper and lower coordinates, which determines the icon size from the icon placed at the position of the target included in the image as shown in FIG. 3 To obtain a pair of coordinates including an upper coordinate and a lower coordinate of an icon disposed at a position of an object included in the image.

The homography calculation unit 150 calculates the size information of at least two objects input by the size information input unit 130 and the coordinates of the camera using at least two coordinate pairs obtained by the coordinate information obtaining unit 140. [ Calculate the homograhy value for image correction.

Meanwhile, according to a further aspect of the invention, the homography-based camera image correction apparatus 100 may further include an image correction unit 160. The image correcting unit 160 corrects the coordinates on the image plane of the image displayed by the image display unit 110 by using the homograhy value calculated by the homography calculating unit 150 as coordinates on the ground plane And outputs it.

As can be seen from Equation 1 above, if the camera parameters are known, homography can be easily calculated. In the present invention, a camera (a person, a tree, a column, or the like) standing vertically on the ground in a three-dimensional space is photographed by a camera and displayed through the image display unit 110, Icon, and size information for the object is input through the size information input unit 130. FIG.

Then, the coordinate pairs of the object are obtained by the coordinate information obtaining unit 140 using the camera parameters, and the homography calculating unit 150 obtains the coordinates of the coordinates of the corresponding object input by the user, Using the pair, we obtain the homography between the image plane and the ground plane. Then, the image correcting unit 160 converts the coordinates on the given image plane into the coordinates on the ground plane using the calculated homograhy value to display the image.

An object standing perpendicular to the ground in three-dimensional space can be modeled as a principal axis, and this straight line is projected in a straight line to the image plane. Considering that the main object of the surveillance image is a person and the lower side of the straight line is the upper coordinate and the lower coordinate, one pair of coordinates is defined as the lower coordinate z _i on the image plane and the upper coordinate z _i ', And the height of the straight line l _i in the three-dimensional space. The heights of the objects may be different.

Allowing the user to enter a height for multiple coordinate pairs allows the algorithm to be performed. If two or more coordinate pairs are given, an exact solution can be obtained if the number is two or two according to the number. If the number is two or more, the least square method (Least Mean Squared Error solution) and the bundle adjustment An approximate solution can be obtained.

Since the user input may have noise such as an error of several pixels and the height of the object may be different from the actual one, it is preferable to perform the algorithm using several objects. The present invention takes such noise into account so that the error is minimized.

The user interface (not shown) provided by the coordinate information obtaining unit 140 may be implemented to display and obtain information on the obtained coordinates to the user. Thus, the user can determine if the acquired coordinates are appropriate and reset the coordinates if they are not.

When all the processes are completed, the homography can be calculated through the homography calculation unit 150. Using the homography, the image correction unit 160 converts the coordinates on the image plane into coordinates on the ground plane, Can be used for surveillance.

For example, the image corrector 160 transforms coordinates on the image plane into coordinates on the ground plane in a top view mode through a separate window, and displays the position of the icon input by the user . The x and y axes in the transformed image are the same as the axes of the actual ground plane, and the display unit can be a meter. In the top view mode, the non-plane part of the converted image will have many distortions. The user can check the position of the area corresponding to the ground plane and the position of the icon to see if the calculation is performed well.

For example, the homography calculation unit 150 may be implemented to obtain a homography between an image plane and a ground plane through a homography calculation algorithm as follows. First, the principal point of an internal parameter of the camera parameter is assumed to be an image center (u ₀ = W / 2, v ₀ = H / 2). Therefore, when the algorithm is performed, the coordinates of the image center are subtracted from the coordinates of the image center (u ₀ = W / 2, v ₀ = H / 2). On the other hand, the aspect ratio is assumed to be α _u = α _v = focal distance f.

On the other hand, it is assumed that the extrinsic parameter is only the rotation? _X about the x axis. The tilt angle is assumed to be 0 to 90 degrees. Since the origin of the world coordinate can be arbitrarily determined, let the coordinates of the camera center be c = [0, 0, -h] ^T. Thus, the origin of the ground plane is [0, 0, 0] ^T , that is, the projection of the camera center perpendicular to the ground plane. Knowing the rotation and the camera center can tell you the translation.

The above assumption is a general assumption in the camera calibration problem and the camera parameters to be obtained under these assumptions are focal length f, tilt angle θ _t , camera height h to be. In this case, the camera parameter matrix K, the rotation matrix R, and the translation vector r are as follows.

Also, the tilt angles θ _t and φ _x have the following relationship according to the coordinate system setting.

The range of φ _x by the assumption of the range of the tilt angle is as follows.

The coordinate system of the homography calculation algorithm is set as shown in FIG. 4 on the assumption of the camera parameter. In Fig. 4, I and G denote an image plane and a ground plane, respectively. x _c , y _c , and z _c represent the axes of the camera coordinate system. x _w , y _w , and z _w represent the axes of the world coordinate system. c represents the center of the camera. O _w represents the origin of the actual coordinate system. z ₀ represents the principal point. l represents the height of the object. z and z 'represent the top coordinates on the image plane, respectively. x and x 'represent the bottom and top coordinates in 3D space, respectively, and x = [x, y, 0,1] ^T , x' = [x, y, l, 1] ^T.

The transformation from the bottom coordinate to the top coordinate is called 'foot to head homology'. The assumption in 'foot to head homology' is that the object must be perpendicular to the ground plane, which is identical to the assumption in the present invention. The lower coordinate z and the upper coordinate z 'on the image plane have the relationship expressed by the following Equation 2.

(Equation 2)

In the present invention, it is assumed that the heights of the objects may be different from each other, so that the homology may be different for each object. Of a height of the destination l 'foot to head homology' H ^l _FH may be calculated as the following equation 3.

(Equation 3)

here,

to be.

The matrix Q is not related to the height of the object. If the matrix Q is obtained, both the homology and the camera parameter of each object can be obtained. From Equation 2 and Equation 3, the relation between the lower coordinate z = [u, v] ^T on the image plane of the object with the height l and the upper coordinate z '= [u', v '] ^T can be derived as follows.

In summary, the following equations 4 and 5 can be obtained.

(Equation 4)

here,

(Equation 5)

If from Equation 4 n {the lower coordinates, top coordinates, and height of a pair of coordinates} _{{z i, z i ',} l i} n i = 1 is given, it is possible to obtain a 2n linear equations for q.

here,

, Which can be obtained from a pair of coordinates.

If n = 2, then the exact solution is

ego,

If n> 2, the least squares solution is given by Equation 6.

(Equation 6)

Where A ⁺ is a pseudo inverse matrix of A.

All elements of vector q must be negative by the range condition of φ _x and by Equation 5. If the above linear solution does not satisfy this condition, it can be seen that there is a lot of errors in the user's input. Therefore, the user does not proceed to the next step and informs the user that the parameter can not be obtained. The same problem occurs when the pseudo inverse matrix of matrix A does not exist in Eq. (6).

Once the linear solution q is obtained, the camera parameters of Equation 7 can be easily determined by Equation 5 and? _X.

(Equation 7)

The camera parameters obtained in linear solutions may not be accurate if there is noise due to the coordinate pair. Therefore, it is necessary to adjust the parameters to minimize the reprojection error.

Bundle adjustment is a nonlinear optimization process for obtaining a parameter that minimizes reprojection error. For example, the LM (Levenberg Marquardt) algorithm can be used.

사이의 에러

를 최소화 하는 파라미터

를 LM 알고리즘을 통하여 구한다.In the bundle adjustment, the upper coordinate z _i 'on the image plane and the upper coordinate z _i ' obtained from the lower coordinate z _i using the homography H ^l _FH

Error between

≪ / RTI >

Is obtained through the LM algorithm.

here

ego,

to be.

The error can be redefined as follows by homology.

here,

ego,

에 대한

의 편미분 계수들은then,

For

The partial differential coefficients of

에 대한

의 기울기는ego,

For

The slope of

는

에 대한 벡터

의 자코비안(Jacobian) 행렬이다.here,

The

Vector for

It is a Jacobian procession.

에 대한

의 자코비안 행렬은Accordingly

For

The Jacobian procession of

.

When you repeat k times, the parameters are updated as follows:

The iteration continues until it converges, and μ can be set to a small value that controls the convergence stability and speed, and the existence of the inverse is assured.

Once the camera parameters are obtained through such an algorithm, homography from the ground plane to the image plane or vice versa can be obtained and the interplanar coordinate transformation can be performed.

Therefore, according to the present invention, by simply arranging the icon for acquiring the coordinates on the image including the object of which the actual size is known, without the hassle of inputting the camera installation information including the camera installation height and angle, Since the homography value can be calculated, the user's convenience can be improved.

A homography calculation process for camera image correction performed by the homography calculation apparatus as described above will be described with reference to FIG. 5 is a flowchart illustrating a configuration of an embodiment of a homography-based camera image correction method according to the present invention.

First, in the image display step 510, the homography calculation device displays an image including an object whose actual size is photographed by the camera. Since we have already explained the image display with the object of which the actual size is known, the redundant explanation is omitted.

Then, in the icon placement step 520, the homography calculation device arranges an icon for obtaining coordinates at the position of the target included in the image displayed by the image display step 510, Adjust the size and angle of the icon to fit the angle.

At this time, the icon for obtaining the coordinates may include two feature points for obtaining the upper and lower coordinates, which determine the icon size. The arrangement of the icon for acquiring the coordinates, the adjustment of the size and the angle have been described earlier, and thus, the description of the overlap is omitted.

Then, in the size information input step 530, the homography computing device inputs the actual size information of the object included in the image displayed by the image display step 510. The description of the actual size information input of the object included in the image has been described earlier, so redundant description is omitted.

Then, in the coordinate information obtaining step 540, the homography computing device obtains a pair of coordinates including the upper and lower coordinates of the icon arranged at the position of the object included in the image by the icon placing step 520 . The description of the coordinate pair acquisition is omitted here, so redundant description is omitted.

Next, in the homography calculation step 550, the homography calculation device calculates the size information of at least two objects input by the size information input step 530 and the size information of the at least two objects inputted by the coordinate information acquisition step 540 Compute the homograhy value for camera image correction using at least two coordinate pairs. The description of the homograhy value calculation for the camera image correction has been described earlier, so redundant description is omitted.

According to a further aspect of the present invention, the homography-based camera image correction method may further include an image correction step 560. In the image correcting step 560, the homography calculating device calculates the coordinates of the image on the image plane of the image displayed by the image displaying step 510 using the homograhy value calculated by the homography calculating step 550 Into coordinates on the ground plane and outputs the coordinates.

Therefore, according to the present invention, by simply arranging the icon for acquiring the coordinates on the image including the object of which the actual size is known, without the hassle of inputting the camera installation information including the camera installation height and angle, It is possible to calculate a homography value, thereby improving the user's convenience. Therefore, the above-described object of the present invention can be achieved.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those of ordinary skill in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. .

The present invention is industrially applicable in the field of image calibration technology and its application technology.

100: Homography calculating device 110:
120: icon placement unit 121: icon storage unit
122: icon selection unit 123: icon overlay unit
124: icon adjusting unit 130: size information input unit
140: Coordinate information acquisition unit 150: Homography calculation unit
160: image correction unit

Claims (7)

An image display unit for displaying an image including an object whose actual size is photographed by a camera;
An icon arrangement unit for arranging an icon for acquiring coordinates at a position of a target included in the image displayed by the image display unit and adjusting the size and angle of the icon according to the size and tilted angle of the target;
A size information input unit for inputting actual size information of an object included in the image displayed by the image display unit;
A coordinate information acquiring unit for acquiring a coordinate pair including an upper coordinate and a lower coordinate of the icon arranged at the position of the object included in the image by the icon arranging unit;
A homography ratio calculation unit for calculating a homogrargy value for camera image correction using size information of at least two objects input by the size information input unit and at least two coordinate pairs obtained by the coordinate information obtaining unit, A calculation unit;
11. A homography-based camera image correction apparatus comprising:
The icon arranging unit:
An icon storage unit for storing icons for obtaining coordinates;
An icon selection unit for selecting an icon for acquiring coordinates stored in the icon storage unit according to a user operation;
An icon overlay unit for overlaying an icon for coordinate acquisition selected by the icon selection unit on a position of an object included in the image according to a user operation;
An icon adjusting unit for changing the size and angle of the icon for obtaining the coordinates overlaid on the position of the target included in the image by the icon overlay unit according to the size of the target and the tilted angle according to the user operation;
Wherein the camera image correcting device is a homography-based camera image correcting device. delete The method according to claim 1,
The icon for obtaining the coordinates is:
And a feature point for obtaining an upper coordinate and a lower coordinate for determining the size of the icon. The method according to claim 1 or 3,
Wherein the homography-based camera image correction apparatus comprises:
An image correcting unit for converting the coordinates on the image plane of the image displayed by the image display unit into coordinates on the ground plane by using a homograhy value calculated by the homography calculating unit;
Wherein the camera image correcting device further comprises: A homography-based camera image correction method performed by a camera image correction device,
An image display step of displaying an image including an object whose actual size is photographed by a camera;
An icon placement step of arranging an icon for acquiring coordinates at a position of a target included in the image displayed by the image display step and adjusting the size and angle of the icon according to the size and tilted angle of the target;
A size information input step of inputting actual size information of an object included in the image displayed by the image display step;
A coordinate information acquiring step of acquiring a coordinate pair including an upper coordinate and a lower coordinate of the icon arranged at the position of the object included in the image by the icon disposing step;
Calculating a homograhy value for camera image correction using size information of at least two objects input by the size information input step and at least two coordinate pairs obtained by the coordinate information obtaining step, Graph computation steps;
A method for correcting a homography-based camera image,
In the icon placement step, an icon for coordinate acquisition is selected according to a user operation, an icon for selecting a coordinate is overlaid on a position of an object included in the image, and an icon for overlaying coordinates Wherein the size and the angle of the icon are changed according to the size and the inclined angle of the corresponding object. 6. The method of claim 5,
The homography-based camera image correction method includes:
An image correction step of converting the coordinates on the image plane of the image displayed by the image display step into coordinates on the ground plane by using the homograhy value calculated by the homography calculation step;
Further comprising the steps of: (a) The method according to claim 5 or 6,
The icon for obtaining the coordinates is:
And a feature point for obtaining an upper coordinate and a lower coordinate for determining an icon size.

Images