KR101582348B1 - Method for monitoring parking lot and apparatus thereof - Google Patents

Abstract

The present invention relates to a method for monitoring a parking garage which includes: a step of a control unit for photographing the parking garage by using multiple photographing units which are located on a moving path of the parking garage and include a fish-eye lens; a step for extracting a reference line from an image photographed by the photographing units by using an image processing unit; a step for calculating an angle of the curved reference line extracted by using the image processing unit; a step for dewarping a plurality of images photographed by using the photographing units based on the angle calculated by the image processing unit; and a step for connecting the dewarped images and outputting the dewarped images through an output unit.

Description

METHOD FOR MONITORING PARKING LOT AND APPARATUS THEREOF BACKGROUND OF THE INVENTION [0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention [0002] The present invention relates to a parking lot monitoring method and apparatus, and more particularly, to a parking lot monitoring method and apparatus that can easily monitor a wide range of parking lots through a plurality of cameras that photograph a designated area.

The parking lot is set up to build a car. As the number of supply of automobiles increases recently, the airport, large shopping centers, large-scale amusement facilities and large buildings are constructed with large parking lots.

These parking lots are divided into ground parking, which is located on the ground, and underground parking, which is located in the basement of the building. The underground parking lot is dark and the human being is rare.

Thus, landlords have separate caretakers to monitor the parking lot, but it is practically difficult for the caretaker to monitor the parking lot 24 hours a day, and in the case of large caretakers, it is impossible for the caretaker to monitor all areas.

For this reason, in recent years, many cameras have been installed in a parking lot for 24 hours, and a manager has been developing a technique for monitoring images photographed through each camera in the office.

Prior art related to the present invention is Korean Patent Laid-Open Publication No. 10-2007-0032971 (published Mar. 23, 2007, entitled " Parking lot monitoring system and monitoring method using the same).

According to the conventional technique of photographing all the parking lots by installing a plurality of cameras, a large number of cameras must be installed in the large parking lot.

That is, since the conventional box-type camera has a small angle of view and a small surveillance area, the camera has to be arranged closely in order to prevent blind spots.

Therefore, there is a problem that installation of a surveillance system for management of a large parking lot is costly due to installation of a large number of cameras.

In addition, when the parking lot is photographed with too many cameras, the administrator has to check too many photographed images.

SUMMARY OF THE INVENTION The present invention has been made in order to solve the above-mentioned problems, and it is an object of the present invention to provide an image pickup apparatus capable of capturing an extensive parking lot through a plurality of cameras including a fish- The present invention provides a parking lot monitoring method and apparatus that can easily monitor a parking lot on a single screen.

A parking lot monitoring method according to an aspect of the present invention includes the steps of photographing a parking lot through a plurality of photographing units including a fish-eye lens in which a control unit is located on a moving path of a parking lot; Extracting a reference line from an image photographed through the plurality of photographing units through an image processing unit; Calculating a bent angle of the extracted reference line through the image processing unit; Dewarping a plurality of images photographed through the plurality of photographing units based on the calculated angles through the image processing unit; And connecting the plurality of diverted images to output through an output unit.

In the step of photographing a parking lot through the plurality of photographing units according to the present invention, the plurality of photographing units are spaced apart from each other by a preset reference distance corresponding to each photographing range.

The method may further include extracting an image of a predetermined area from the plurality of images through the image processing unit before extracting the reference line.

The extracting of the reference line may include extracting a corner point from the photographed image, And extracting a reference line through connection of the extracted corner points.

In the step of extracting the corner point of the present invention, the image processing unit extracts the corner point based on a change in brightness of a pixel included in the photographed image.

In the diverting of the plurality of images of the present invention, the image processing unit diverges the plurality of images based on a Perspective Projection Transformation technique.

In the step of extracting the baseline of the present invention, the baseline is a parking line.

According to another aspect of the present invention, there is provided a parking lot monitoring apparatus comprising: a plurality of photographing units including a fish-eye lens located on a moving path of a parking lot; A reference line is extracted from an image photographed through the plurality of photographing units, a bent angle of the extracted reference line is calculated, and a plurality of images photographed through the plurality of photographing units based on the calculated angle are dewarped, A video processor; A control unit for controlling the plurality of images to be connected and output; And an output unit for outputting the connected image based on the control of the control unit.

In the present invention, the plurality of photographing units are spaced apart from each other by a preset reference distance corresponding to each photographing range.

In the present invention, the image processor extracts an image having a predetermined area from the plurality of images, and extracts a reference line from the extracted image.

In the present invention, the image processing unit extracts a corner point from the photographed image, and extracts a reference line through connection of the extracted corner points.

The image processing unit may extract the corner point based on a brightness change of a pixel included in the photographed image.

In the present invention, the image processing unit diverges the plurality of images based on a Perspective Projection Transformation technique.

In the present invention, the image processing unit extracts a parking line from the photographed image.

According to the present invention, it is possible to monitor a parking lot without installing an excessive number of cameras by photographing an extensive parking lot through a plurality of cameras including a fish-eye lens.

In addition, the present invention can improve the surveillance concentration of the manager by connecting the images photographed through the plurality of cameras to one screen.

1 is a block diagram of a parking lot monitoring apparatus according to an embodiment of the present invention.
FIG. 2 is a view showing a position where a plurality of photographing units are installed and a photographing range of each photographing unit of the parking lot monitoring apparatus according to an embodiment of the present invention. FIG.
3 is a view showing an image taken through a photographing unit of a parking lot monitoring apparatus according to an embodiment of the present invention and an image obtained through image processing, respectively.
FIGS. 4 to 6 are diagrams for comparing brightness of pixels for extracting a reference line from a captured image of a parking lot monitoring apparatus according to an embodiment of the present invention. FIG.
FIG. 7 is a flowchart illustrating an implementation process of a parking lot monitoring method according to an embodiment of the present invention. Referring to FIG.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of a parking lot monitoring method and apparatus according to the present invention will be described in detail with reference to the accompanying drawings. In this process, the thicknesses of the lines and the sizes of the components shown in the drawings may be exaggerated for clarity and convenience of explanation. In addition, the terms described below are defined in consideration of the functions of the present invention, which may vary depending on the intention or custom of the user, the operator. Therefore, definitions of these terms should be made based on the contents throughout this specification.

1 is a block diagram of a parking lot monitoring apparatus according to an embodiment of the present invention.

Referring to FIG. 1, a parking lot monitoring apparatus according to an embodiment of the present invention includes a plurality of photographing units 10, an image processing unit 30, a control unit 50, and an output unit 70.

The plurality of photographing units 10 are devices that photograph a parking lot including a fish-eye lens.

As described above, in this embodiment, a parking lot is photographed by a plurality of photographing portions 10 including a fish-eye lens for monitoring all the parking lots having a wide range.

That is, since the fisheye lens is an ultra-wide angle lens having an angle of view of more than 180 degrees, it can take a wide range, so that the photographing unit 10 includes a fisheye lens in this embodiment.

However, since the fisheye lens pops out like a pupil of a fish and the four corners distort the image as if the diagonal line goes down, it is necessary to normalize the distorted image into a rectangular area. For a detailed description of the normalization operation, .

Further, in the present embodiment, the entire area of the parking lot is photographed through the plurality of photographing units 10, and a plurality of photographed images are connected to output a video image so that the user can monitor the entire parking lot at a glance A process of connecting the respective images photographed through the plurality of photographing units 10 is required.

To this end, in the present embodiment, the plurality of photographing sections 10 are provided so as to be spaced apart from each other by a preset reference distance corresponding to the photographing range.

That is, the plurality of photographing units 10 have a fixed photographing range set at the time of installation. In this embodiment, when a region having a specific area is extracted from a plurality of photographed images, And the corners are spaced apart from each other by a distance corresponding to each other.

Therefore, in this embodiment, the reference distances provided apart from the plurality of photographing sections 10 may be set to the distances between the photographing sections 10 where the edges of the regions extracted from the plurality of photographed images can coincide with each other.

However, the present embodiment is not limited to this case. In this case, not only a case in which all the images are automatically connected by extracting a specific region from a plurality of photographed images and connecting extracted regions, but also a control section 50, A plurality of images can be connected to each other by a method of extracting a plurality of feature points located in the same position in a plurality of images, and a plurality of images can be seamlessly connected in various other ways.

FIG. 2 is a view showing a position where a plurality of photographing units are installed and a photographing range of each photographing unit of the parking lot monitoring apparatus according to an embodiment of the present invention. FIG.

Referring to FIG. 2, it can be seen that the plurality of photographing units 10 are located on the moving path of the parking lot, and are spaced apart from each other by a predetermined distance as described above.

That is, in this embodiment, a plurality of photographing units 10 are located on the moving path of the parking lot in order to more accurately track where the vehicle that has entered the inside of the parking lot moves more accurately than the vehicle parked in the parking lot.

The plurality of photographing units 10 according to the present embodiment are spaced apart from each other such that the photographing range S is partially overlapped as shown in Fig.

As described above, since the four corners of the image photographed through the fisheye lens are severely distorted, in the present embodiment, rather than diverging the entire photographed image, a part of the image close to four corners is removed, Only a specific region is extracted.

Therefore, the plurality of photographing units 10 are installed so as to have overlapping photographing ranges as shown in FIG.

The image processing unit 30 extracts a reference line from a plurality of images photographed through the plurality of photographing units 10, calculates a bent angle of the extracted reference line, and derives a plurality of images photographed based on the calculated angle, (Dewarping).

That is, as described above, since the plurality of photographing units 10 according to the present embodiment photographs a parking lot including a fish-eye lens, the photographed image includes distorted portions, and the image processing unit 30 includes a distorted portion To a rectangular area.

3 is a view showing an image taken through a photographing unit of a parking lot monitoring apparatus according to an embodiment of the present invention and an image obtained through image processing, respectively.

3 (A) shows an image taken through a photographing unit 10 including a fish-eye lens, and FIG. 3 (B) shows an image taken through an image processing of the image processing unit 30. FIG. Image.

Specifically, the image photographed through the photographing unit 10 according to the present embodiment has a wide angle of view, but distortion at four corners is significant. Therefore, as described above, It is possible to acquire a distortion-free image as in (B) by first extracting the region and diverging the extracted S1 region.

That is, as described above, since the plurality of photographing units 10 are provided so as to have overlapping photographing ranges, the image processing unit 30 can extract only the S1 region except the region where the distortion is severe in the photographed image, The image of the parking area can be obtained.

Therefore, the image processing section 30 extracts an area other than the photographing range overlapping based on the photographing range S of the plurality of photographing sections 10, so that the region where the distortion is too strong can be filtered out first.

The image processing unit 30 extracts a reference line from the extracted image and diverges a plurality of captured images based on the bent angle of the extracted reference line.

That is, since the criteria for diverging varies according to how much the image is distorted, the image processing unit 30 determines the degree of distortion based on the reference line included in the extracted image and diverts the degree of distortion.

In particular, in the present embodiment, the image processing section 30 sets the parking line as a reference line and extracts the parking line from a plurality of images.

That is, in the parking lot, the parking line clearly marked for distinguishing the parking area may be the most accurate reference. Therefore, in this embodiment, the image processing unit 30 extracts the parking line from the photographed image to judge the degree of distortion of the image .

Specifically, in this embodiment, the image processing unit 30 extracts the parking line based on the brightness change of the pixels included in the photographed image.

FIGS. 4 to 6 are diagrams for comparing brightness of pixels for extracting a reference line from a captured image of a parking lot monitoring apparatus according to an embodiment of the present invention. FIG.

As described above, since the plurality of photographing units 10 are located on the moving path of the parking lot, a parking line is included on the side of the photographed image as shown in Fig.

Accordingly, the image processing unit 30 extracts the parking line from the image photographed through the plurality of photographing units 10. Specifically, the image processing unit 30 extracts the corner point of the parking line, You can extract lines.

At this time, the corner point of the parking line means a point where three parking lines are gathered like the P1 area of Fig. 4, or a point where two parking lines are gathered like the P2 area of Fig.

That is, when a point where three parking lines such as the P1 area are gathered and a point where two parking lines such as the P2 area are gathered are connected, the image processing unit 30 first extracts the corner point of the parking line from the image And the parking line is extracted through the connection of the corner points.

Particularly, in this embodiment, the image processing unit 30 can extract the corner point based on the brightness change of the pixel included in the photographed image.

FIG. 5 is a diagram showing an example of extracting a corner point included in the P1 region of FIG. 4, and FIG. 6 is an example of extracting a corner point included in the P2 region of FIG.

Specifically, the image processing unit 30 searches for the brightness of the pixel in the counterclockwise direction on the basis of an arbitrary point included in the image, and repeats the above-described process while continuously moving the arbitrary point.

At this time, when the specific point included in the image is selected as the corner point included in the P1 region, the result of searching the brightness of the pixel in the counterclockwise direction based on the corner point is shown in the graph of FIG.

Corner points included in the P1 area are formed at angles of 90 °, 90 °, and 180 ° with respect to the other parking lines because three parking lines are gathered.

Therefore, as shown in FIG. 5, after the brightness of the pixels changes at intervals of 90 degrees, the brightness of the pixels changes again at intervals of 180 degrees.

Since corner points included in the P2 area are gathered in two directional parking lines, an arbitrary parking line is formed at an angle of 90 DEG in one direction and 270 DEG in the other direction with another parking line.

Accordingly, as shown in FIG. 6, after the brightness of the pixels changes at intervals of 90 degrees, a pattern in which the brightness of the pixels changes again at intervals of 270 degrees appears.

However, since the actual parking line is formed at an angle of 90 degrees, it can be seen that the interval at which the brightness of the pixel varies in FIG. 5 is different from the interval at which the brightness of the pixel changes in FIG. 6, This is because the parking line is photographed at an angle smaller than 90 °.

Therefore, the image processing unit 30 does not determine whether the brightness of a pixel appears at 90 °, 180 °, or 270 ° intervals, but rather extracts a corner point based on the pattern of brightness change.

Then, the image processing unit 30 extracts the parking line by connecting the extracted corner points, and calculates the curved angle of the parking line. Specifically, the image processing unit 30 can calculate the curved angle of the parking line by comparing coordinate information of pixels having the same brightness.

However, the present embodiment is not limited to this, and it is possible to calculate the bent angle of the parking line through various known methods or other unknown methods.

Subsequently, the image processing section 30 diverges a plurality of photographed images based on the calculated angle of curvature of the parking line.

That is, the image processing unit 30 diverges so that the amount of change of the photographed images is relatively larger as the angle of warping of the parking line is relatively larger, and as the warping angle of the parking line is relatively smaller, So that the amount of image change is relatively small.

The diopting is a method of normalizing a distorted portion included in the photographed image into a rectangular region. For example, there is a method of diverting the coordinate system of the photographing portion 10 in advance based on the degree of distortion of the image In addition to the above, various other known or unknown methods may be used.

In particular, in the present embodiment, the image processing unit 30 diverges a plurality of images based on a Perspective Projection Transformation technique.

The perspective projection transformation technique is a transformation technique that projects a point in a three-dimensional space onto a two-dimensional image plane, and transforms the distorted four corners of the extracted image to be projected onto each corner of the rectangle.

Since the perspective projection transformation technique is a known technique, a detailed description of the implementation process will be omitted.

The control unit 50 controls the image processing unit 30 to connect and output a plurality of images that are diverted so that the user can monitor the entire area of the parking lot in one screen.

At this time, the control unit 50 may control to connect all of the plurality of images diverging the images photographed by the plurality of photographing units 10 and output them to one screen. However, the present embodiment is not limited to this.

That is, even if the parking area is wide like a large parking lot, it may be difficult to monitor some areas when all the images are connected and output to one screen. Therefore, in this embodiment, So that all or some of the areas can be monitored.

In particular, in the present embodiment, the controller 50 may extract a feature point from an image and connect the plurality of images to each other in such a manner that the same feature point is connected to the plurality of images. In this case, the feature point may be the corner point .

The output unit 70 is a device for outputting a plurality of images connected based on the control of the control unit 50. The output unit 70 displays all the images of the entire area of the parking lot connected to each other, So that it can be tracked and monitored.

FIG. 7 is a flowchart illustrating an implementation process of a parking lot monitoring method according to an embodiment of the present invention. Referring to FIG.

Referring to FIG. 7, a method for implementing a parking lot monitoring method according to an embodiment of the present invention will be described. First, the control unit 50 includes a plurality of photographing units 10 including a fish-eye lens, (S10).

As described above, in this embodiment, a parking lot is photographed by a plurality of photographing portions 10 including a fish-eye lens for monitoring all the parking lots having a wide range.

Further, in the present embodiment, the entire area of the parking lot is photographed through the plurality of photographing units 10, and a plurality of photographed images are connected to output a video image so that the user can monitor the entire parking lot at a glance A process of connecting the respective images photographed through the plurality of photographing units 10 is required.

To this end, in the present embodiment, the plurality of photographing sections 10 are set apart from each other by a predetermined distance corresponding to the photographing range.

That is, the plurality of photographing units 10 have a fixed photographing range set at the time of installation. In this embodiment, when a region having a specific area is extracted from a plurality of photographed images, And the corners are spaced apart from each other by a distance corresponding to each other.

Therefore, in the present embodiment, the separation distance at which the plurality of photographing units 10 are installed can be set to a distance at which the edges of the regions extracted from the plurality of photographed images can coincide with each other.

The control unit 50 extracts a reference line from a plurality of images photographed through the plurality of photographing units 10 through the image processing unit 30 (S20).

At this time, in this embodiment, the parking line indicated to distinguish the parking zone can be set as the reference line for setting clearer reference.

As described above, since the plurality of photographing units 10 according to the present embodiment photographs a parking lot including a fish-eye lens, the photographed image includes distorted portions, and the image processing unit 30 converts the distorted portions into rectangles It is necessary to perform a diagonalization to normalize the area.

At this time, since the criteria for diverging varies according to how the image is specifically distorted, a parking line is extracted to judge how distorted the image is.

Specifically, the image processing unit 30 extracts the corner points of the parking line based on the brightness change of the pixels included in the photographed image, and extracts the parking lines through the connection of the extracted corner points.

That is, since the brightness change of the pixels around the corner point of the parking line is relatively more and more regular than that of the bottom area, in this embodiment, the corner points of the parking line are first extracted based on the brightness change of the pixels, The parking line can be extracted through the connection.

Subsequently, the image processing unit 30 calculates the bent angle of the extracted reference line (S30), and diverges a plurality of photographed images based on the calculated angle (S40).

That is, since the criteria for diverging varies according to how much the image is distorted, the image processing unit 30 determines the degree of distortion based on how much the reference line included in the extracted image is warped, Diopting.

The control unit 50 connects the plurality of images that have been diverted through the image processing unit 30 in step S50 and outputs the corresponding images through the output unit 70 in step S60.

According to the present embodiment, it is possible to monitor a parking lot without installing an excessive number of cameras by photographing a wide range of parking lots through a plurality of cameras including a fish-eye lens.

In addition, the present embodiment can improve the surveillance concentration of the manager by connecting the images photographed through the plurality of cameras to one screen and outputting them.

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. I will understand. Accordingly, the technical scope of the present invention should be defined by the following claims.

10:
30:
50:
70:

Claims (14)

Photographing a parking lot through a plurality of photographing units including a fish-eye lens, the control unit being located on a moving path of the parking lot;
Extracting a reference line from an image photographed through the plurality of photographing units through an image processing unit;
Calculating a bent angle of the extracted reference line through the image processing unit;
Dewarping a plurality of images photographed through the plurality of photographing units based on the calculated angles through the image processing unit; And
And connecting the plurality of images, which are diverted, through an output unit,
The step of extracting the reference line includes:
Extracting a corner point from the photographed image; And
Extracting a reference line through connection of the extracted corner points,
In the step of extracting the corner point, the image processing unit extracts the corner point based on a brightness change of a pixel included in the photographed image,
In the step of extracting the reference line, the reference line is a parking line,
Wherein the control unit extracts feature points from the plurality of images and connects the plurality of images by connecting the extracted feature points to each other, Wherein the parking lot monitoring method comprises:
The method according to claim 1,
Wherein the plurality of photographing sections are spaced apart from each other by a predetermined reference distance corresponding to each photographing range in the step of photographing the parking lot through the plurality of photographing sections.
The method according to claim 1,
Further comprising extracting an image of a predetermined area from the plurality of images through the image processing unit before extracting the reference line.
delete delete The method according to claim 1,
Wherein the image processing unit diverges the plurality of images based on a Perspective Projection Transformation technique in diverting the plurality of images.
delete A plurality of photographing units positioned on a moving path of the parking lot, the photographing units including a fish-eye lens;
A reference line is extracted from an image photographed through the plurality of photographing units, a bent angle of the extracted reference line is calculated, and a plurality of images photographed through the plurality of photographing units based on the calculated angle are dewarped, A video processor;
A control unit for controlling the plurality of images to be connected and output; And
And an output unit for outputting the connected image based on the control of the control unit,
Wherein the image processing unit extracts a corner point from the photographed image and extracts the reference line through connection of the extracted corner point, extracts the corner point based on a brightness change of a pixel included in the photographed image, The reference line is a parking line,
Wherein the control unit extracts feature points from the plurality of images, connects the extracted feature points, and connects the plurality of images, wherein the feature points are the corner points.
9. The method of claim 8,
Wherein the plurality of photographing units are spaced apart from each other by a preset reference distance corresponding to each photographing range.
9. The method of claim 8,
Wherein the image processing unit extracts an image of a predetermined area from the photographed image, and extracts a reference line from the extracted image.
delete delete 9. The method of claim 8,
Wherein the image processing unit diverges the plurality of images based on a Perspective Projection Transformation technique.
delete

Images