KR100574697B1 - Camera calibration method - Google Patents

Abstract

The present invention provides a camera correction method for automatically expanding and aligning a rotating camera image to all positions within a fixed camera image, wherein the pan (P), tilt (T) corresponding to all points of the image input through the fixed camera, The first step of generating zoom (Z) and focus (F) lookup table values; correcting the pan (P), tilt (T), zoom (Z), and focus (F) lookup table values to suit the rotating camera. A second step of designating a position to be enlarged by the rotating camera with respect to the image input through the fixed camera, and the corrected pan (P), tilt (T), zoom (Z), and focus ( F) a fourth step of automatically aligning and expanding the rotatable camera to the designated position by using a look-up table value, which is a technical feature.

Therefore, the present invention significantly reduces the operation time by automating a series of processes for enlarging and aligning the rotatable camera to an arbitrary point with respect to all points within the image input through the fixed camera. Efficient and also, by calculating the actual distance for every point in the image acquired through the fixed camera to obtain the values of pan (P), tilt (T), zoom (Z) and focus (F) for There is an effect that can improve the accuracy in aligning the rotary camera to the position of.

Rotating camera, camera calibration, auto zoom alignment

Description

Camera calibration method for automatically expanding and aligning a rotating camera image to all positions inside a fixed camera image {Camera Calibration Method}

1A is a photograph showing an image input through a fixed camera.

1B is a photograph showing an image input through a rotating camera.

FIG. 2A is a photograph showing that an arbitrary point is selected inside an image input through a fixed camera according to an exemplary embodiment of the present invention.

2B is a photograph showing a state in which the rotatable camera is automatically enlarged and aligned in an embodiment of the present invention.

3 is a flowchart illustrating an overall operation of a camera calibration method for automatically expanding and aligning an image input through a fixed camera according to an exemplary embodiment of the present invention.

4 is a diagram showing the resolution and angle of view of the fixed camera according to an embodiment of the present invention.

5 is a view showing a relationship between the fixed camera and the Y-axis distance according to an embodiment of the present invention.

6 is a view showing a relationship between the fixed camera and the X-axis distance according to an embodiment of the present invention.

FIG. 7 is a diagram illustrating a relationship between an arbitrary point of an image input through a stationary camera and a pan and tilt of a rotatable camera according to an embodiment of the present invention.

8 is a flowchart of a camera correction unit for generating a lookup table according to an embodiment of the present invention.

9 is a flowchart illustrating an automatic alignment unit of two cameras using correction data according to an exemplary embodiment of the present invention.

10 is a control block of a camera calibration apparatus for performing a camera calibration operation according to an exemplary embodiment of the present invention.

<Explanation of symbols for the main parts of the drawings>

40, 71: fixed camera 50, 72: rotating camera

73: rotating camera driver 74: lookup table (Lookup Table)

75: camera correction unit 76 software controller

The present invention relates to a surveillance or control system using a camera, and more particularly, in the case of using a fixed camera and a rotating camera, in order to enlarge a predetermined point inside an image input through the fixed camera, The present invention relates to a camera calibration method that moves to a predetermined point and automatically enlarges and aligns.

In general, a surveillance and control system using a camera is composed of a fixed camera for detecting a certain area and a rotational (PTZF) camera for enlarging a predetermined point of the detected area. In order to control the rotating camera, only the desired area of the image input through the fixed camera is enlarged and photographed.

The performance of such a system is most important to accurately detect an image, but it is also important to precisely align the designated area by moving the camera to the correct position for the area to be enlarged.

Therefore, in order to accurately control the rotating camera to specific coordinates of the fixed camera, calibration of the fixed camera and the rotating camera is required.

Conventionally, a hardware controller or a software controller has been used to enlarge and align the rotatable camera to any designated position in the image acquired through the fixed camera. In this case, a parameter of the rotatable camera is composed of Pan, Tilt, Zoom, and Focus, and the conventional hardware and software controllers are all Pan, Tilt. By manually manipulating (T), zoom (Z), and focus (F), a process of moving the center of the rotatable camera to a designated position of an image acquired through the fixed camera was performed to enlarge and align.

However, in the above method, it is difficult to control the speed and time when aligning the pan (P), tilt (T), zoom (Z), and focus (F). There is a disadvantage that is reduced.

Therefore, in order to overcome these disadvantages, a technique of adopting a method of automating the above operation process and moving to an approximate position using a preset function of a rotating camera has been proposed. If the number of points to be specified is only greater than 200 to 300 than this, there is a lot of difficulties in the correction, there is an inefficient problem.

Therefore, in order to solve the above problem, when a random point is selected for every point in the image input through the fixed camera, the two cameras are calibrated so that the rotary camera is automatically enlarged and aligned to the point. The purpose is to provide a method.

According to the present invention, when using a hardware controller or a software controller for aligning a rotating camera to an arbitrary position within an image acquired through a fixed camera, a pan, tilt, zoom, and focus are used. Its purpose is to automate a series of processes to move the center of the rotatable camera to a designated position inside the image input through the fixed camera by manipulating.

Another object of the present invention is to provide a method for shortening the time for expanding and aligning a rotary camera to an arbitrary position with respect to all points in an image acquired through a fixed camera.

It is still another object of the present invention to provide a method for improving accuracy in expanding and aligning a rotary camera to an arbitrary position with respect to all points in an image acquired through a fixed camera.

1A is a photograph showing an image input through a fixed camera, and FIG. 1B is a photograph showing an image input through a rotating camera.

1A to 2B, when an arbitrary point 10 is designated as shown in FIG. 2A for all points in an image acquired through a fixed camera as shown in FIG. 1A, The rotatable camera is automatically magnified and aligned to the arbitrary point 10 to show the result 20 of being magnified and aligned as shown in FIG. 2B.

The camera correction method for automatically expanding and aligning a rotating camera image to all positions within the fixed camera image according to the present invention includes pan (P), tilt (T), and zoom corresponding to all points of the image input through the fixed camera. (Z), a first step of generating a focus (F) lookup table value; correcting the pan (P), tilt (T), zoom (Z), and focus (F) lookup table values to be suitable for a rotating camera. In a second step, a third step of designating a position to be enlarged with the rotatable camera with respect to the image input through the fixed camera and the corrected pan (P), tilt (T), zoom (Z), and focus (F). And a fourth step of automatically aligning and expanding the rotatable camera to the designated position using a lookup table value.

Preferably, the first step of generating the pan (P), tilt (T), zoom (Z), and focus (F) lookup table values may be performed by using the fixed camera to adjust pan (P) and tilt (T) values. Obtaining a zoom (Z) and focus (F) value using the rotating camera.

Preferably the step of obtaining the pan (P), the tilt (T) value, calculating the X-axis distance and Y-axis distance of the image input through the fixed camera, the input through the fixed camera according to the user's settings Designating a representative correction point within the captured image, calculating a pan (P) value and a tilt (T) value with respect to the representative correction point, and a pan (P) value and a tilt for all points of the input image. Correcting the value (T).

Preferably, the step of obtaining the zoom (Z) and focus (F) values may include calculating an X-axis distance and a Y-axis distance of an image input through the fixed camera, using the X-axis distance and the Y-axis distance. Calculating the actual diagonal length from the rotatable camera to the floor, storing the zoom (Z) and focus (F) values for the diagonal length, and zooming (Z) for all points of the input image, Interpolating the focus F value.

Prior to this, terms or words used in the specification and claims should not be construed as having a conventional or dictionary meaning, and the inventors should properly explain the concept of terms in order to best explain their own invention. Based on the principle that can be defined, it should be interpreted as meaning and concept corresponding to the technical idea of the present invention.

Therefore, the embodiments described in the specification and the drawings shown in the drawings are only the most preferred embodiment of the present invention and do not represent all of the technical idea of the present invention, it is possible to replace them at the time of the present application It should be understood that there may be various equivalents and variations.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

3 is a view showing the flow of a camera correction method for automatic magnification alignment which is a function to be achieved in the present invention. As shown in FIG. 3, pan (P), tilt (T), zoom (Z), and focus (F) lookup table (LUT) values corresponding to all points of the image input through the fixed camera are generated. The first step (S10), the second step (S20), the fixed camera to correct the pan (P), tilt (T), zoom (Z), focus (F) look-up table values suitable for the rotating camera A third step (S30) of designating a position to be enlarged with the rotatable camera with respect to the image input through the step S30 and the corrected pan (P), tilt (T), zoom (Z), and focus (F) lookup table values Using a fourth step (S40) for automatically aligning and expanding the rotatable camera to the designated position.

A method of obtaining the actual distance of all points of an image acquired by the fixed camera of the present invention will be described with reference to FIGS. 4 to 6.

According to the present invention, the distance (P), tilt (T), zoom (Z), and focus (P) points are obtained by using distance information between two cameras and actual distance information of each pixel in an image using Inigo's distance measurement method. F) Calculate the value.

Inigo's proposed three-dimensional actual distance calculation method for each point in a two-dimensional image includes the focal length of the camera, the size of the CCD, the angle of view for the X and Y axes, the height of the camera (h), and the distance to the bottom of the image. If D _miny is known, the actual distance of all the points in the image acquired from the fixed camera 40 can be obtained.

Referring to FIG. 4 of the present invention, the figure shown in FIG. 4A shows the CCD 30 and the focus in the camera. The magnitude of the CCD is a product (a × b) of the horizontal length (a) of the CCD and the vertical length (b) of the CCD, f is the focal point, and λ represents the focal length. With these values, the angles of view Φ _x and Φ _y can be calculated using the following equation.

[제 1 수학식]

[First Equation]

4 (b) shows the real world range taken by the angles of view of Φ _x and Φ _y , where r _x , r _y represents the resolution of the X and Y axes, respectively.

5 is a method for calculating a Y-axis distance (dy) for calculating the distance away from the lowest point of the fixed camera of the present invention.

As shown in FIG. 5, the height h of the fixed camera 40 is measured, and the distance Dminy to the bottom end of the image displayed through the calibrated camera is measured. Subsequently, the height h and the distance Dminy from the bottom of the image are calculated by applying the following second equation to obtain the Y-axis distance of the image displayed through the fixed camera.

[제 2 수학식]

[Second equation]

Referring to FIG. 6, calculating a distance (X-axis distance) away from the image center of the fixed camera of the present invention may be obtained through the following third equation.

[제 3 수학식]

[Third Equation]

After calculating the actual X-axis distance and Y-axis distance for all the areas inside the image of the fixed camera through the above method, as shown in Figure 7 the pan of the rotating camera at an arbitrary point C (r, s) ( P) and the tilt (T) value are obtained. The pan (P) value corresponds to the horizontal angle of the rotatable camera and corresponds to θ of FIG. 7, and the tilt (T) value corresponds to the vertical angle of the rotatable camera and corresponds to Φ of FIG. 7. The pan value θ and the tilt value Φ are calculated by the following Equation 4 below, and each pan (P) and tilt (T) is corrected by correcting an error of the installation environment using one representative point. Create a lookup table.

[제 4 수학식]

[Fourth Equation]

Then, using the actual diagonal length (d _g ) from the rotatable camera to the floor, the zoom (Z), focus (F) values of the rotatable camera are sampled according to the diagonal length (d _g ), By using this, the zoom (Z) and focus (F) values of all the areas of the fixed camera are interpolated to be generated and stored in a lookup table for each zoom (Z) and focus (F).

Therefore, the pan (P), tilt (T), zoom (Z), and focus (F) values of the rotating camera are stored for all points of the fixed camera.

All the above processes will be described in detail through a flowchart of a camera correction unit for generating the lookup table of FIG. 8 of the present invention. The distance between the actual Y-axis and the X-axis is calculated for all regions of the image input through the fixed camera (S101). In order to correct an error of the installation environment, a representative correction point is designated (S102), and the pan (P) and tilt (T) values of the representative correction point are stored, and the image of the image input through the fixed camera is stored. The pan (P) and tilt (T) values for all points are corrected (S104).

On the other hand, using the actual Y-axis and X-axis distance calculated in step S101 to calculate the actual diagonal length from the rotating camera to the floor (S106), and the zoom (Z), focus (F) value for the diagonal length After storing (S107), the zoom (Z) and focus (F) values for all points in all regions of the image input through the fixed camera are interpolated (S108).

Thereafter, the pan (P) and tilt (T) values obtained through the step S104 and the zoom (Z) and focus (F) values obtained through the step (S108) are pans (P), tilt (T), and zoom ( Z), a lookup table for each focus F is generated.

9 is a flowchart illustrating an automatic alignment operation of the rotary camera according to the present invention.

 A position to be enlarged and captured by the rotating camera is specified in all areas of all images input through the fixed camera (S121). If this is described in detail with reference to FIG. 2A of the present invention, if a point corresponding to the reference numeral 10 is designated using a mouse or the like, the coordinates (x, y) of the point are read.

In addition, a look-up table for each of the pan (P), tilt (T), zoom (Z), and focus (F) generated and stored according to the present invention is referred to (S122) at the coordinates (x, y) to which the position to be enlarged and captured is designated. The corresponding pan (P), tilt (T), zoom (Z) and focus (F) values are called up.

The software controller of the rotating camera driver issues a command to move the pan (P), tilt (T), zoom (Z), and focus (F) values corresponding to the positions of the designated coordinates (x, y) (S123). ).

When the rotatable camera is moved by the software controller, the result is that the center of the rotatable camera is automatically aligned (S124) with the designated coordinates (x y) of the stationary camera.

Referring to Figure 10 the control configuration of the camera correction apparatus for performing the camera correction operation of the present invention will be described.

The camera correction apparatus of the present invention is a fixed camera 71, a rotating camera 72, a rotating camera driver 73, look-up table 74, camera correction unit 75, software controller 76, camera automatic alignment The unit 77 is composed of a computer (PC) 78.

When a few representative points are designated in the fixed camera 71 and the rotating camera 72 and the image inputted through the fixed camera 71, pans (P), tilts (T), and zooms are applied to the points. (Z), after obtaining the focus (F) value, the approximate value obtained for all points other than the representative point through interpolation is stored, and no matter what point is determined inside the image input through the fixed camera 71 A look-up table 74 for automatically aligning the center of the camera, and a camera correction unit for correcting the rotatable camera 72 into an image input through the fixed camera by using the look-up table 74 ( 75 and the camera image automatic alignment unit 77 and the camera image aligning unit to correct the rotatable camera image into the fixed camera image by the control program of the software controller 76 of the computer (PC, 78). Phase is automatically drives the rotatable camera 72 by the sorting unit 77 consists of a rotatable camera driving unit 73 for correcting a rotation type camera image into the stationary camera image.

Although the present invention has been shown and described with reference to the preferred embodiments as described above, it is not limited to the above embodiments and those skilled in the art without departing from the spirit of the present invention. Various changes and modifications will be possible.

As described above, the present invention significantly reduces the operation time by automating a series of processes for expanding and aligning the rotatable camera to an arbitrary point for all points within the image input through the fixed camera. Compared with the prior art, there is a very efficient effect.

The present invention obtains the actual distance for all points in the image acquired through the fixed camera, and obtains the values of pan (P), tilt (T), zoom (Z), and focus (F) to any position. There is an effect that can improve the accuracy in the enlarged alignment of the rotating camera.

Claims (4)

Generating a pan (P), tilt (T), zoom (Z), and focus (F) lookup table values corresponding to all points of the image input through the fixed camera; A second step of correcting the pan (P), tilt (T), zoom (Z), and focus (F) lookup table values to be suitable for the rotating camera; A third step of designating a position to be enlarged by the rotatable camera with respect to the image input through the fixed camera; And A fourth step of automatically aligning and expanding the rotatable camera to the designated position using the corrected pan (P), tilt (T), zoom (Z), and focus (F) lookup table values Camera correction method for automatically expanding the alignment of the rotating camera image to any position inside the fixed camera image including a. The method of claim 1, The first step of generating the pan (P), tilt (T), zoom (Z), focus (F) look-up table values, Obtaining pan (P) and tilt (T) values using the fixed camera; And Obtaining zoom (Z) and focus (F) values using the rotating camera Camera correction method for automatically expanding the alignment of the rotating camera image to any position inside the fixed camera image including a. The method of claim 2, Obtaining the pan (P), tilt (T) value, Calculating an X axis distance and a Y axis distance of an image input through the fixed camera; Designating a representative correction point in an image input through the fixed camera according to a user's setting; Calculating a pan (P) value and a tilt (T) value with respect to the representative correction point; And Compensating the pan (P) value and the tilt (T) value for all points of the input image Camera correction method for automatically expanding the alignment of the rotating camera image to any position inside the fixed camera image including a. The method of claim 2, Obtaining the zoom (Z), focus (F) value, Calculating an X axis distance and a Y axis distance of an image input through the fixed camera; Calculating an actual diagonal length from the rotatable camera to the floor using the X axis distance and the Y axis distance; Storing zoom (Z) and focus (F) values for the diagonal length; And Interpolating zoom (Z) and focus (F) values for all points of the input image Camera correction method for automatically expanding the alignment of the rotating camera image to any position inside the fixed camera image including a.

Images