KR102391852B1 - Method for detecting motion, camera and surveillance system adopting the method - Google Patents

Abstract

A motion detection method for detecting a motion region of an image of a specific target on a screen to be monitored, including steps (a) to (c). In step (a), information on the reference height and reference width of the image of the specific object is obtained for each pixel of the screen to be monitored. In step (b), when a moving area is detected on the screen to be monitored, it is determined whether the specific object is in the moving area according to the reference height and reference width of the image of the specific object. In step (c), if it is determined that the specific object is in the movement area, the detected movement area is displayed on the monitoring target screen.

Description

Method for detecting motion, camera and surveillance system adopting the method}

The present invention relates to a motion detection method, a camera and a monitoring system employing the method, and more particularly, to a motion detection method for detecting a motion region of an image of a specific target on a monitoring target screen, and a camera employing the method and to a monitoring system.

A camera or a host device of a monitoring system may detect a movement region of an image of a specific target on a screen to be monitored. Here, as an example of a specific object, a person or a vehicle may be mentioned. When a motion area is detected, the motion area of the detection result is displayed on the screen to be monitored, while an alarm signal is output.

As such a motion detection method, there are various methods such as comparison of before and after images or probabilistic modeling. Nevertheless, the problem of detection error still remains.

For example, the movement area displayed on the screen to be monitored may include only a part of the moving person, not the whole. In addition, when only the motion of a person is detected as a specific object, an error occurs in that not only the person but also the motion of the vehicle is detected.

The problem with the background art is that the inventor possessed for the purpose of derivation of the present invention or acquired during the derivation process of the present invention, and it cannot necessarily be said that the content is known to the general public prior to the filing of the present invention.

Republic of Korea Patent Publication No. 2014-0106362 (published on September 3, 2014, Applicant: Samsung Techwin Co., Ltd., Seoul National University Industry-Academic Cooperation Foundation, Title of Invention: Abnormal Motion Detection Apparatus and Method).

An embodiment of the present invention is to provide a motion detection method, camera, and monitoring system capable of reducing the number of occurrences of motion detection errors.

According to one aspect of the present invention, there is provided a motion detection method for detecting a motion region of an image of a specific target on a screen to be monitored, including steps (a) to (c).

In step (a), information on the reference height and reference width of the image of the specific object is obtained for each pixel of the screen to be monitored.

In step (b), when the moving area is detected on the monitoring target screen, it is determined whether the specific object is in the moving area according to information on the reference height and reference width of the image of the specific object.

In step (c), if it is determined that the specific target is in the moving area, the detected moving area is displayed on the monitoring target screen.

According to the second aspect of the present invention, there may be provided a camera that detects a movement region of an image of a specific target on a screen to be monitored. This camera adopts the above motion detection method.

According to a third aspect of the present invention, there may be provided a monitoring system in which a host device is connected to a camera, and the host device detects a movement region of an image of a specific target on a screen to be monitored. The host device employs the motion detection method.

According to the motion detection method of the embodiment of the present invention, the camera and the monitoring system employing the method, information on the reference height and reference width of the image of the specific object is applied to each pixel of the screen to be monitored.

Accordingly, a change in the size of an image according to a distance between the camera and the specific object cannot act as a cause of a motion detection error. Accordingly, the number of occurrences of a motion detection error may be reduced.

For example, the number of occurrences of an error of detecting only a part of a moving person may be reduced. In addition, when only the movement of a person is detected as a specific object, the number of occurrences of an error of detecting an object other than the person may be reduced.

1 is a diagram showing a camera or surveillance system employing a method according to an embodiment of the present invention.
2 is a diagram of a screen to be monitored for explaining a method according to an embodiment of the present invention.
3 is a flowchart illustrating an information acquisition method performed by each camera or a host device of the surveillance system of FIG. 1 according to a method according to an embodiment of the present invention.
FIG. 4 is a diagram for explaining step a2a of FIG. 3 .
5 is a diagram showing that information on a reference height and reference width of an image of a specific object is stored for each pixel of a screen to be monitored.
6 is a flowchart showing that each camera of FIG. 1 or a host device of the surveillance system detects motion according to the information of FIG. 5 .
FIG. 7 is a diagram illustrating a first example of a detection error that may appear when motion is detected without using the information of FIG. 5 .
FIG. 8 is a diagram showing that the detection error of FIG. 7 does not occur by using the information of FIG. 5 .
9 is a diagram illustrating a second example of a detection error that may appear when motion is detected without using the information of FIG. 5 .
FIG. 10 is a diagram showing that the detection error of FIG. 9 does not occur by using the information of FIG. 5 .

The following description and accompanying drawings are for understanding the operation according to the present invention, and parts that can be easily implemented by those skilled in the art may be omitted. That is, in describing the present invention, if it is determined that a detailed description of a related known technology may obscure the gist of the present invention, the detailed description thereof will be omitted.

In addition, the specification and drawings are not provided for the purpose of limiting the present invention, and the scope of the present invention should be defined by the claims. The terms used in this specification should be interpreted with meanings and concepts consistent with the technical idea of the present invention so that the present invention can be most appropriately expressed.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Of course, the motion detection method of this embodiment can be used together with various conventional detection algorithms.

1 shows a camera 101a to 101n or a surveillance system 10 employing the method of an embodiment of the present invention.

Referring to FIG. 1 , cameras 101a to 101n communicate with a host device 102 while exchanging communication signals Sco, and transmit a live-view video signal Svid to the host device. Send to 102.

The video signal Svid received by the host device 102 is displayed through a display device and stored in a recording device, for example, a hard disk drive.

Here, the method according to an embodiment of the present invention may be performed by each of the cameras 101a to 101n or may be performed by the host device 102 . That is, each of the cameras 101a to 101n or the host device 102 detects a movement region of an image of a specific target on the monitoring target screen.

Contents related thereto will be described in detail with reference to FIGS. 2 to 10 .

2 is a diagram of a monitoring target screen 201 for explaining a method according to an embodiment of the present invention. In FIG. 2, reference numeral 202 denotes a horizon on the screen, X denotes a horizontal direction on the screen, Y denotes a vertical direction on the screen, 203 denotes an image of a person in the first position, and 203e denotes a ground landing point. , 203b is the reference target area corresponding to the pixel, H is the reference height, W is the reference width, D is the distance to the horizon, 204 is the image of a person in the second position, and 204e is the ground A pixel corresponding to the landing point, and 204b indicate a reference target area, respectively.

The motion detection method of the present embodiment will be described with reference to FIGS. 1 and 2 .

In each of the cameras 101a to 101n or the host device 102, the reference height H of the images 203 and 204 of the specific target for each pixel 203e, 204e, etc. of the monitoring target screen 201, and Information on the reference width W is stored. A specific subject in this embodiment is a person. That is, in this embodiment, a moving person is detected.

When each of the cameras 101a to 101n or the host device 102 detects a movement area on the monitoring target screen 201, the reference height H and the reference width W of the images 203 and 204 of the specific target ), it is determined whether the specific object exists in the movement area.

When it is determined that the specific target is in the moving area, each of the cameras 101a to 101n or the host device 102 displays the detected moving area on the monitoring target screen and then outputs an alarm signal.

According to the motion detection method of this embodiment, the cameras 101a to 101n and the monitoring system 10 employing this method, a specific target is detected for each pixel 203e, 204e, etc. of the monitoring target screen 201 . Information on the reference height (H) and the reference width (W) of the images 203 and 204 is applied. Accordingly, a change in the size of an image according to a distance between the cameras 101a to 101n and a specific object cannot act as a cause of a motion detection error. Accordingly, the number of occurrences of a motion detection error may be reduced.

For example, the number of occurrences of an error of detecting only a part of a moving person may be reduced. In addition, when only the movement of a person is detected as a specific object, the number of occurrences of an error of detecting an object other than the person may be reduced.

In obtaining information on the reference height (H) and reference width (W) of the images 203 and 204 of a specific target for each pixel (203e, 204e, etc.) of the monitoring target screen 201, the monitoring target screen ( Each pixel (203e, 204e, etc.) of 201 is matched with each ground landing point of the image (203, 204) of a specific object. Here, any one pixel corresponding to any one ground landing point exists at the lowest vertical position among the pixels of the images 203 and 204 of the specific object. That is, among the pixels of the images 203 and 204 of the specific target, the lowest pixel in the vertical direction (Y) corresponds to the ground landing point.

In the information of the reference height (H) and the reference width (W) of the images 203 and 204 of a specific target for each pixel (203e, 204e, etc.) of the monitoring target screen 201, the unit of the number of pixels is used. .

In obtaining information on the reference height (H) and reference width (W) of the images 203 and 204 of the specific target for each pixel (203e, 204e, etc.) of the monitoring target screen 201, the user actually The specific subject that becomes the subject is photographed at least once. Here, each of the cameras 101a to 101n or the host device 102 may display at least a first pixel (eg, 203e) of the monitoring target screen 201 of an image (eg, 203) of the specific target. By making it at least the first ground landing point, information of the reference height H and the reference width W of the image 203 of the specific target for the at least the first pixel (eg, 203e) is obtained.

Next, each of the cameras 101a to 101n or the host device 102 determines the reference height H of the image (eg, 203) of a specific target for the at least the first pixel (eg, 203e) and Using the information on the reference width W, information on the reference height H and reference width W of the image (eg, 204 ) of a specific target for each of the remaining pixels is calculated. This calculation process is shown in FIG. 3 .

FIG. 3 shows an information acquisition method performed by each of the cameras 101a to 101n of FIG. 1 or the host device 102 of the monitoring system according to the method of an embodiment of the present invention. This will be described with reference to FIGS. 2 and 3 as follows.

First, each of the cameras 101a to 101n or the host device 102 of the monitoring system searches for the horizon 202 on the monitoring target screen (step a2a). This step a2a will be described with reference to FIG. 4 .

Next, each of the cameras 101a to 101n or the host device 102 of the monitoring system obtains a distance D from the horizon 202 for each of the remaining pixels (step a2b).

Next, each of the cameras 101a to 101n or the host device 102 of the monitoring system has a reference height of an image (eg, 203) of a specific target for the at least the first pixel (eg, 203e) According to (H) and the reference width (W) information, and the distance D from the horizon 202 for each of the remaining pixels, an image of a specific target for each of the remaining pixels (eg, 204 ) ), calculate the reference height (H) and the reference width (W) (step a2c). Of course, the calculation formula may be derived in various ways according to the characteristics of each camera 101a to 101n or the host device 102 of the monitoring system.

Next, each of the cameras 101a to 101n or the host device 102 of the monitoring system stores information of the reference height H and the reference width W obtained for each of all pixels in the recording device (step a2d).

FIG. 4 is a diagram for explaining step a2a of FIG. 3 .

In FIG. 4 , the same reference numerals as those of FIG. 2 indicate objects having the same function. In Fig. 4, reference numeral 203h denotes a pixel corresponding to the topmost point in the first image 203 of a specific object, 401 denotes an image of a person at a third position, 401e denotes a pixel corresponding to a ground landing point, and 401h denotes a pixel corresponding to the ground landing point. Each of the pixels corresponding to the uppermost point in the third image 401 of a specific target is indicated.

The process of step a2a will be described in detail with reference to FIGS. 3 and 4 as follows.

First, the user obtains a plurality of images 203 and 401 of a specific target by shooting once.

Accordingly, each of the cameras 101a to 101n or the host device 102 of the monitoring system connects the pixels 203h and 401h corresponding to the uppermost points in the plurality of specific target images 203 and 401 to each other. A first straight line 402 is obtained while connecting.

In addition, each of the cameras 101a to 101n or the host device 102 of the monitoring system connects the pixels 203e and 401e corresponding to the lowest points in the plurality of specific target images 203 and 401 to each other. while obtaining a second straight line 403 .

Then, each of the cameras 101a to 101n or the host device 102 of the monitoring system sets the vertical position (Y-axis position) of the intersection of the first straight line 402 and the second straight line 403 to the horizon 202 . Set to the vertical position (Y-axis position) of

5 shows that information 501 of a reference height and reference width of an image of a specific object is stored for each pixel of the monitoring target screen 201 . That is, reference size information 501 of an image of a specific target is stored for each pixel.

Referring to the reference size information 501 for any one pixel in FIG. 5 , a specific target of motion detection is a person, the reference width of the human image in the pixel is 30 pixels, and the human image in the pixel The reference height of is 150 pixels.

FIG. 6 shows that each of the cameras 101a to 101n of FIG. 1 or the host device 102 of the monitoring system detects motion according to the information of FIG. 5 . This is explained as follows.

Each of the cameras 101a to 101n or the host device 102 of the monitoring system determines whether a motion region of an image of a specific target is detected (step S601). When the motion region is detected, each of the cameras 101a to 101n or the host device 102 of the monitoring system performs the following steps S603 to S613.

In step S603, each of the cameras 101a to 101n or the host device 102 of the monitoring system searches for a pixel existing at the lowest vertical position among pixels of an image of a specific target. That is, each of the cameras 101a to 101n or the host device 102 of the monitoring system searches for a pixel corresponding to a ground landing point of an image of a specific target whose motion is detected.

Next, each of the cameras 101a to 101n or the host device 102 of the monitoring system reads information of the reference height (H in Fig. 2) and the reference width (W in Fig. 2) corresponding to the found pixel. (Step S605).

Next, each of the cameras 101a to 101n or the host device 102 of the monitoring system obtains a difference between the height of the image of the specific target whose motion is detected and the reference height. In addition, it is determined whether the obtained height difference is greater than a set value (step S607).

If the obtained height difference is greater than the set value, each of the cameras 101a to 101n or the host device 102 of the monitoring system determines that the motion-detected target is not a specific target and performs step S615.

If the obtained height difference is not greater than the set value, each of the cameras 101a to 101n or the host device 102 of the monitoring system performs the following steps S609 to S613.

In step S609, each of the cameras 101a to 101n or the host device 102 of the monitoring system obtains a difference between the width of the image of the specific object whose motion is detected and the reference width. Further, it is determined whether or not the obtained width difference is larger than a set value (step S609).

If the obtained width difference is greater than the set value, each of the cameras 101a to 101n or the host device 102 of the monitoring system determines that the motion-detected target is not a specific target and performs step S615.

If the obtained width difference is not greater than the set value, each of the cameras 101a to 101n or the host device 102 of the monitoring system performs the following steps S611 and S613. That is, each of the cameras 101a to 101n or the host device 102 of the monitoring system displays the detected motion area on the monitoring target screen (step S611), and then outputs an alarm signal (step S613).

Steps S601 to S613 are repeatedly performed until an end signal is generated (step S615).

FIG. 7 shows a first example of a detection error that may appear when motion is detected without using the information of FIG. 5 . FIG. 8 shows that the detection error of FIG. 7 does not occur by using the information of FIG. 5 .

7 and 8, reference numeral 201 denotes a screen to be monitored. In FIG. 7 , reference numeral 701 denotes an erroneously detected motion region. In FIG. 8 , reference numeral 801 denotes a well-detected motion region.

As can be seen with reference to FIGS. 7 and 8 , when the size information for each location of FIG. 5 is used, the number of occurrences of an error of detecting only a part of a moving person can be reduced.

FIG. 9 shows a second example of a detection error that may appear when motion is detected without using the information of FIG. 5 . FIG. 10 shows that the detection error of FIG. 9 does not occur by using the information of FIG. 5 .

9 and 10, reference numeral 201 denotes a screen to be monitored. In FIG. 9 , reference numeral 901 denotes a well-detected motion region and 902 denotes an erroneously-detected motion region, respectively. In FIG. 10, reference numeral 1001 denotes a well-detected motion region.

As can be seen with reference to FIGS. 9 and 10 , when the size information for each location of FIG. 5 is used, the number of occurrences of an error in detecting an object other than a person, which is a specific object, may be reduced.

As described above, according to the motion detection method of the embodiment according to the present invention, the camera and the monitoring system employing the method, information on the reference height and reference width of the image of a specific object is obtained for each pixel of the screen to be monitored. applies.

Accordingly, a change in the size of an image according to the distance between the camera and a specific object cannot act as a cause of a motion detection error. Accordingly, the number of occurrences of a motion detection error may be reduced.

For example, the number of occurrences of an error of detecting only a part of a moving person may be reduced. In addition, when only the motion of a person is detected as a specific object, the number of occurrences of an error of detecting an object other than the person may be reduced.

So far, the present invention has been focused on preferred embodiments. Those of ordinary skill in the art to which the present invention pertains will understand that the present invention can be implemented in modified forms without departing from the essential characteristics of the present invention.

Therefore, the disclosed embodiments are to be considered in an illustrative rather than a restrictive sense. The scope of the present invention is indicated in the claims rather than the above description, and the invention claimed by the claims and inventions equivalent to the claimed invention should be construed as being included in the present invention.

It is highly likely to be used not only for motion detection but also for recognition of a specific object.

101a to 101n: cameras, 102: a host device;
201: monitoring target screen, 202: horizon on the screen,
X : Horizontal direction on the screen, Y : Vertical direction on the screen,
203: an image of a person in the first position;
203e: a pixel corresponding to the ground landing point;
203b: reference target area, H: reference height,
W: reference width, D: distance from the horizon,
204: an image of a person in the second position;
204e: a pixel corresponding to the ground landing point;
204b: reference target area;
203h: the pixel corresponding to the topmost point;
401: an image of a person in a third position;
401e: a pixel corresponding to the ground landing point;
401h: the pixel corresponding to the topmost point;
501: reference size information for one pixel;
701, 902: erroneously detected motion regions;
801, 901, 1001: well-detected motion regions.

Claims (6)

A motion detection method for detecting a motion region of an image of a specific target on a monitoring target screen, the method comprising:
(a) storing information of a reference height and a reference width of the image of the specific target for each pixel of the monitoring target screen;
(b) when the motion region is detected on the monitoring target screen, finding a pixel present in the lowest vertical position of an image of a specific target for which motion is detected,
the pixel present in the lowest vertical position is the lowest pixel in the vertical direction corresponding to the ground landing point among the pixels of the image of the specific target;
(c) reading information on a reference height and a reference width of a pixel corresponding to the lowest vertical position among pixels of the monitoring target screen;
(d) obtaining a height difference between the height of the image of the specific object in which the movement is detected and the read reference height and a difference in width between the width of the image of the specific object in which the movement is detected and the read reference width;
(e) determining that the detected movement area is not the movement area of the specific object if the obtained height difference is greater than a first set value or the obtained width difference is greater than a second set value . The method of claim 1, wherein in step (a),
By associating each pixel of the monitoring target screen with each ground landing point of the specific target image, information on the reference height and reference width of the specific target image for each pixel of the monitoring target screen is obtained. A motion detection method obtained. The method of claim 2, wherein in step (a),
In the information of the reference height and reference width of the image of the specific target for each pixel of the monitoring target screen,
The unit of number of pixels is used,
In matching each pixel of the monitoring target screen to each ground landing point of the image of the specific target,
Any one pixel corresponding to any one ground landing point exists in the lowest vertical position among the pixels of the image of the specific object. The method of claim 1,
In the step (a), in storing the information of the reference height and reference width of the image of the specific target for each pixel of the monitoring target screen,
(a1) at least the first pixel of the monitoring target screen becomes at least a first ground landing point of the image of the specific target, and information of the reference height and reference width of the image of the specific target for the at least the first pixel wanted; and
(a2) calculating information on the reference height and reference width of the image of the specific object for each of the remaining pixels by using the information on the reference height and the reference width of the image of the specific object for the at least the first pixel; This is done,
The step (a2) is,
(a2a) finding a horizon on the screen to be monitored;
(a2b) obtaining a distance from the horizon for each of the remaining pixels; and
(a2c) the specific target for each of the remaining pixels according to information on the reference height and reference width of the image of the specific target for the at least the first pixel, and the distance from the horizon for each of the remaining pixels Including; calculating the reference height and reference width of the image of the motion detection method. delete delete

Images