KR20140036637A - Method for setting region for video surveillance, apparatus therefor - Google Patents

Abstract

A method for setting a region for video surveillance and a video surveillance apparatus therefor are disclosed. The video surveillance apparatus comprises: an image acquisition unit for acquiring images; an interface unit for receiving commands by an operation of a user; a segment assignment unit which marks a plurality of segments on one side and the other side of the acquired image based on the commands; a region of interest setting unit which generates an external connection line connecting both ends of the plurality of segments to each other, and sets the area within the external connection line as a region of interest; and an attribute setting unit which sets attributes of the region of interest based on the commands, and changes the shape of the region of interest in accordance with the attributes set. [Reference numerals] (112) Interface unit; (114) Camera; (116) Display unit; (122) Segment assignment unit; (124) Region of interest setting unit; (126) Connection line generating unit; (128) Adjustment point generating unit; (132) Attribute setting unit; (134) Moving direction setting unit; (142) Object detecting unit; (144) Object direction detecting unit; (146) Notification unit

Description

Method for setting region for video surveillance and apparatus for video surveillance {Method for Setting Region for Video Surveillance, Apparatus Therefor}

The present embodiment relates to an area setting method for video surveillance and a video surveillance apparatus therefor. More specifically, for video surveillance, a video surveillance system can be configured to detect various moving objects by setting various regions of interest of an image acquired through a video surveillance apparatus and setting an object moving direction without having to constantly monitor the images. It relates to a region setting method for and a video surveillance apparatus therefor.

It should be noted that the following description merely provides background information related to the present embodiment and does not constitute the prior art.

In recent years, as crime increases, security in crime prevention is becoming important. To enhance this security, there is a surveillance system that includes a camera (eg Closed-Circuit Television) installed in a public place (eg building, road). Such a surveillance system may generally be a digital video recorder (DVR) consisting of a computer, a camera, a data transmission cable and a display. In this case, the camera included in the image storage device is fixedly mounted to capture a specific place, and captures and transmits the captured image sequences to appear on the display. Such an image storage device is generally operated by specific security personnel, who manage the image storage device and record the captured image in the storage device.

In the conventional image storage device, a lot of labor is required to continuously monitor the captured image, and when hundreds of cameras are installed, it is impossible to monitor all the images. Accordingly, there is a need for an intelligent video surveillance system that detects an event in a specific video and generates an alarm.

This embodiment provides a video surveillance that can detect a moving object by setting various regions of interest of the image acquired through the video surveillance apparatus and setting the moving direction of the object without having to constantly monitor the video for the video surveillance. The main purpose is to provide an area setting method and a video surveillance apparatus therefor.

According to an aspect of the present embodiment, an image acquisition unit for acquiring an image; An interface unit for receiving a command by a user's operation; A line segment designator which marks a plurality of line segments on one side and the other side of the image acquired based on the command; A region of interest setting unit configured to generate an external connection line connecting both ends of the plurality of line segments to each other, and to set a region within the external connection line as a region of interest; And an attribute setting unit configured to set an attribute of the ROI based on the command and change the form of the ROI into a form according to the set attribute.

According to another aspect of the present embodiment, there is provided a method for setting an area by a video surveillance apparatus, the method comprising: a user input process of receiving a command by a user's operation; A line segment designation process of marking a plurality of line segments on one side and the other side of the image acquired based on the command; A region of interest setting step of generating an external connection line connecting both ends of the plurality of line segments to each other and setting an area within the external connection line as a region of interest; And setting an attribute of the region of interest based on the command and changing an attribute of the region of interest to a form according to the set attribute. .

As described above, according to the present embodiment, the administrator does not need to continuously monitor the image to set the various ROIs of the image acquired through the video surveillance apparatus, and sets the moving direction of the object to set the moving object. There is a detectable effect.

In addition, according to the present embodiment, a specific event (for example, by analyzing the CCTV video through an algorithm, rather than the conventional simple video surveillance system that requires the administrator to continuously monitor the video surveillance field, especially CCTV video that is made by monitoring the video Object detection event) can be detected automatically. In addition, according to the present embodiment, the intelligent video surveillance system has an effect of increasing the accuracy of the event detection or recognition algorithm.

1 is a block diagram schematically showing a video surveillance apparatus according to the present embodiment;
2 is a block diagram schematically illustrating an object detection unit according to the present embodiment;
3 is a flowchart for explaining a region setting method for video surveillance according to the present embodiment;
4 is a flowchart illustrating a method of detecting an object in which a learning rate is varied according to a frame rate for moving object detection according to the present embodiment;
FIG. 5 is a flowchart illustrating a method of generating a background image for detecting an object moving by varying a learning rate according to a frame rate for detecting an object according to an embodiment of the present disclosure; FIG.
6 is an exemplary view for explaining a setting value of a learning rate that varies according to a frame rate according to the present embodiment;
7 is an exemplary view showing line segments defining both ends of a region of interest according to the present embodiment;
8 is an exemplary view illustrating a center connection line of a region of interest according to the present embodiment;
9 is an exemplary view showing an adjustment point on a center connection line according to the present embodiment;
10 is an exemplary view illustrating that the ROI is changed into an arc shape according to the present embodiment;
11 is an exemplary view illustrating that the ROI is changed into a polygonal shape according to the present embodiment;
12 is an exemplary view illustrating an object movement direction setting according to the present embodiment;
13 is an exemplary view showing a change of interest area through a plurality of adjustment points according to the present embodiment;
14 is an exemplary view illustrating that the region of interest according to the present embodiment is applied to detecting a U-turn vehicle.

Hereinafter, the present embodiment will be described in detail with reference to the accompanying drawings.

1 is a block diagram schematically showing the video surveillance apparatus according to the present embodiment.

The video surveillance apparatus 100 according to the present exemplary embodiment may include an interface unit 112, a camera 114, a display unit 116, a line segment designation unit 122, an ROI setting unit 124, a connection line generator 126, An adjustment point generator 128, an attribute setting unit 132, a moving direction setting unit 134, an object detector 142, an object direction detector 144, and a notification unit 146 are included. In the present embodiment, the video surveillance apparatus 100 includes the interface unit 112, the camera 114, the display unit 116, the line segment designation unit 122, the region of interest setting unit 124, the connection line generator 126, and the adjustment unit. It is described as including only the point generating unit 128, the attribute setting unit 132, the moving direction setting unit 134, the object detecting unit 142, the object direction detecting unit 144 and the notification unit 146, As merely illustrative of the technical spirit of the embodiment, those skilled in the art to which this embodiment belongs will be described with respect to the components included in the video surveillance apparatus 100 without departing from the essential characteristics of this embodiment Various modifications and variations will be applicable.

The video surveillance apparatus 100 according to the present embodiment refers to a device capable of capturing and recording an image, and a CCTV camera capable of capturing an image for preventing crime or monitoring a traffic offense such as crime prevention and abnormal behavior monitoring. It can be implemented with a surveillance camera. The image monitoring apparatus 100 detects and stores a moving object in the captured image, and divides the captured image into a plurality of frames.

The interface unit 112 receives an instruction by a user's manipulation or input. Here, the user command may be a setting command for controlling the video surveillance apparatus 100. The interface unit 112 may be implemented as a type of user input unit including a mouse, a keyboard, a joystick, a touch screen, and the like. The camera 114 refers to a kind of imaging unit equipped with a kind of imaging device (eg, a complementary metal-oxide semiconductor (CMOS) or a charge-coupled device (CCD)) and a lens unit. Such a camera 114 obtains an image. In addition, the camera 114 acquires an image for a fixed area (predetermined place). The display unit 116 is screen display means for displaying various operation state information of the image or the image monitoring apparatus 100 received from the camera 114. In addition, the camera 114 may transmit the obtained image to the line segment designation unit 122 by a wired or wireless communication method. That is, the camera 114 and the line segment designator 122 may be connected through a wired cable, or a wireless communication module may be mounted in each of the camera 114 and the line segment designator 122 to enable wireless communication with each other. The image obtained at 114 may be transmitted to the line segment designator 122 through a wireless communication method. Such a wireless communication method may be applied to various communication methods including near field communication. For example, the video surveillance apparatus 100 operates by monitoring a video captured by the camera 114 as a live stream, but the video acquired through the camera 114 is recorded on the DVR / NVR. In this case, the region of interest may be set and the object may be detected. In this case, data communication using a communication method connected between the camera 114 and the line segment designation unit 122 may be used. In addition, the display unit 116 displays messages generated while executing various programs stored in the storage unit when the image surveillance apparatus 100 is driven.

The line segment designator 122 marks the plurality of line segments 710 and 712 on one side and the other side of the image acquired based on the command through the interface unit 112. That is, when an image acquired through the camera 114 is displayed through the display unit 116, a line segment is marked at a position corresponding to a command through the interface unit 112 on the corresponding image. At this time, the length of the line segment is determined by the command through the interface unit 112. Meanwhile, as described above, the line segment designation unit 122 may be connected to the camera 114 in a wired or wireless communication manner, and may include an image acquisition unit that is a separate module for receiving an image obtained from the camera 114. have. Here, the image acquisition unit refers to a kind of reception module that receives an image acquired by the camera 114 by wire or wirelessly. The ROI setting unit 124 generates external connection lines 720 and 722 connecting both ends of the plurality of line segments 710 and 712 marked through the line segment designation unit 122, and the external connection line 720, An area in 722 is set as a Region of Interest 730. The external connection lines 720 and 722 may be automatically made when a plurality of line segments 710 and 712 are marked on an image obtained regardless of a command through the interface unit 112. In addition, the ROI setting unit 124 sets the ROI 730 by connecting the external connection lines 720 and 722 in the form of a closed traverse.

When the plurality of line segments 710 and 712 are marked on the image acquired through the line segment designator 122, the connection line generator 126 connects the center lines 810 to which the midpoints of the plurality of line segments 710 and 712 are connected. ) The center connection line 810 may be automatically formed when a plurality of line segments 710 and 712 are marked on an image obtained regardless of a command through the interface unit 112. The adjustment point generation unit 128 generates the adjustment point 910 at an arbitrary point on the center connection line 810 based on the command through the interface unit 112, and based on the additional command through the interface unit 112. When the adjustment point 910 is moved, the ROI 730 is maintained via the adjustment point 910. That is, the adjustment point generator 128 controls the ROI 730 to maintain the closed polygonal shape through the moved adjustment point 910. In addition, the adjustment point generation unit 128 additionally generates a plurality of new adjustment points 910 based on a command through the interface unit 112, and when a plurality of adjustment points 910 are moved, The region of interest 730 is maintained via all of the adjustment points 910. That is, the adjustment point generator 128 controls the ROI 730 to maintain the closed polygonal shape through the plurality of moved adjustment points 910.

), 직각으로 교차되는 직선 속성(

), 곡선 속성(

), 자유 곡선 속성(

)을 포함한다. 이때, 속성 설정부(132)는 인터페이스부(112)를 통한 명령에 의해 곡선 속성, 자유 곡선 속성이 설정된 경우 관심 영역(730)을 원호 형태(

)로 변경하며, 직선 속성, 직각으로 교차되는 직선 속성이 설정된 경우 관심 영역(730)을 다각형(Polygon) 형태로 변경한다. 이동방향 설정부(134)는 인터페이스부(112)를 통한 명령에 근거하여 객체에 대한 객체 이동방향(1210)을 설정한다. 여기서, 객체 이동방향(1210)은 동, 서, 남, 북 방향(예컨대, →, ↓, ↑, ←)뿐 아니라, 동북, 동남, 서북, 서남 방향과 같은 대각선 방향(예컨대, ↗, ↙, ↖, ↘)도 설정이 가능할 것이다.The attribute setting unit 132 sets an attribute of the ROI 730 based on a command through the interface unit 112, and changes the shape of the ROI 730 into a form according to the set attribute. These properties are linear properties (

), Straight line attributes that intersect at right angles (

), Curve properties (

), Freehand line properties (

). In this case, the attribute setting unit 132 may be configured to form the arc of interest (730) when the curve attribute and the free curve attribute are set by the command through the interface unit 112.

), And the ROI 730 is changed into a polygonal shape when a straight line property and a straight line property that cross at right angles are set. The movement direction setting unit 134 sets the object movement direction 1210 for the object based on the command through the interface unit 112. Here, the object movement direction 1210 is not only east, west, south and north directions (eg, →, ↓, ↑, ←), but also diagonal directions (eg, ↗, ↙, ↖, ↘) can also be set.

The object detector 142 detects an object passing through the ROI 730. Hereinafter, the operation of the object detector 142 will be described. In addition, the object detector 142 checks the frame rate of the current frame (t) of the image, and sets a learning rate for adjusting the image sensitivity according to the frame rate when the frame rate is different from the preset frame rate. The background image of the current frame is generated using the background image, the current frame t, and the learning rate of the previous frame t-1, and the object is detected by comparing the background image of the current frame and the current frame. . The object detector 142 will be described in more detail with reference to FIG. 2. The object direction detector 144 detects the object movement direction 1210 by comparing the object of the current frame with the object of the previous frame. That is, the object direction detector 144 may detect the direction in which the object moves by using the object detected by the object detector 142. The notifier 146 displays the object detected by the object detector 142 and the object movement direction 1210 detected by the object direction detector 144 on the display 116. The notification unit 146 may control whether or not the notification through the display unit 116 according to a predetermined condition (notification cycle, notification only when detecting the object, notification until the object movement direction, etc.) through the interface unit 112. will be.

2 is a block diagram schematically illustrating an object detector according to an exemplary embodiment.

The object detector 142 according to the present exemplary embodiment includes an image receiver 210, an image checker 220, a learning rate setting unit 230, a background image generator 240, and an object recognizer 250. In the present exemplary embodiment, the object detector 142 includes only the image receiving unit 210, the image checking unit 220, the learning rate setting unit 230, the background image generating unit 240, and the object recognizing unit 250. However, this is merely illustrative of the technical idea of the present embodiment, and a person having ordinary knowledge in the technical field to which the present embodiment belongs is included in the object detecting unit 142 without departing from the essential characteristics of the present embodiment. Various modifications and variations to the elements will be applicable.

The object detecting unit 142 includes an image receiving unit 210, an image checking unit 220, a learning rate setting unit 230, a background image generating unit 240, and an object recognizing unit 250. The image receiving unit 210 receives an image captured by the camera 114 and transmits the image to the image checking unit 220.

The image checking unit 220 checks the frame rate of the captured image received from the camera 114. Here, the image checking unit 220 checks whether the frame rate of the captured image is the same based on the preset frame rate. For example, when the frame rate preset in the image checking unit 220 is 30 fps, it is checked whether the frame rate for the captured image is equal to 30 fps or different from 15 fps or 10 fps. Here, the frame rate of the captured image is preferably the same as the preset frame rate, but the object detector 142 simultaneously processes a plurality of images or performs additional functions (for example, garbage dumping, a service specialized in object storage, etc.). Frame rate can be changed by increasing throughput. Here, the frame rate may be calculated by comparing the input time of the current frame and the first previous frame.

The learning rate setting unit 230 adjusts and sets the learning rate according to the frame rate of the captured image. As such, the learning rate setting unit 230 determines that the captured image has a different frame rate than the preset frame rate, and sets the learning rate according to the frame rate of the captured image, and sets the set learning rate. By using the background image generator 240 to generate the updated background image. Here, the learning rate is a value for adjusting the sensitivity of the background image and is set to a constant value between 0 and 1 (for example, 0.99, 0.998, etc.). The sensitivity of the background image that changes according to the learning rate means that the moving object of the captured image can be quickly absorbed as the background image, or the moving object of the captured image can be slowly absorbed. Since the background included in the image is different, the criteria for detecting a moving object are the same.

The background image generator 240 generates a background image using the learning rate set by the learning rate setting unit 230. Here, the background image refers to an image representing a background rather than a moving object. In addition, the background image may be implemented as an image generated using at least one or more of a pixel value, a gray-level value, and a boundary value of the image. The background image generator 240 transmits the background image to the object recognizer 250 and detects a different part as a moving object compared to the current frame. In addition, even if the frame rate of the captured image is changed, the background image generator 240 performs the following process to generate a background image equal to a preset result value based on a predetermined function.

Hereinafter, a variable learning rate and a process for generating a background image according to this equation will be described. First, in general, the background image M (t) at time t generated using a fixed learning rate is expressed by Equation 1 below.

(M (t): background image at time t, α: learning rate, M (t-1): background image at time t-1, I (t): captured image received from camera 114 at time t Pixel value of the image)

If the time t is changed to t + k to calculate the variable learning rate according to the present embodiment, the background image M (t + k) at time t + k is expressed by Equation 2.

(M (t + k): background image at time t + k, α: learning rate, M (t + k-1): background image at time t + k-1, I (t + k): time t pixel value of the captured image received from the camera 114 at + k)

Meanwhile, in [Equation 2] to [Equation 7], the pixel values of the photographed image are assumed to be the same regardless of time in order to calculate only the background image generated according to the learning rate. That is, it is assumed that I (t + k) = I (t + k-1) = I (t + k-2) = ... = I (t).

In addition, M (t + k-1) in Equation 2 is expressed by M (t + k-2), which is the background image of the previous frame, and I (t + k-2), which is a captured image pixel value. Equation 3] can be represented.

(M (t + k): background image at time t + k, α: learning rate, M (t + k-1): background image at time t + k-1, M (t + k-2): Background image at time t + k-2, I (t + k-2): pixel value of the captured image received from camera 114 at time t + k-2)

In addition, M (t + k-2) in [Equation 3] is expressed by M (t + k-3), which is the background image of the previous frame, and I (t + k-3), which is the captured image pixel value. It can be expressed as 4.

(M (t + k): background image at time t + k, α: learning rate, M (t + k-2): background image at time t + k-2, M (t + k-3): Background image at time t + k-3, I (t + k-3): pixel value of the captured image received from camera 114 at time t + k-3)

On the other hand, when generalized by repeating [Equation 2] to [Equation 4], the background image M (t + k) at the time t + k generated using the variable learning rate is the same as [Equation 5].

(M (t + k): background image at time t + k, α ^k : variable learning rate, M (t): background image at time t, I (t): received from camera 114 at time t Value of the captured captured image)

That is, when the image having the changed frame rate is received for every k frames, the object detector 142 calculates the same result as when receiving the preset frame rate using the fixed learning rate α by changing the learning rate α to α ^k . can do.

Meanwhile, when the preset frame rate is T and the changed frame rate is S, the k value becomes T / S, and the background image M (t + k) generated by using the same is expressed by [Equation 6].

(M (t + k): background image at time t + k, α ^k : variable learning rate, M (t): background image at time t, I (t): received from camera 114 at time t Value of the captured captured image, k: preset frame rate (T) / changed frame rate (S))

On the other hand, Equation 6 includes an exponential operation and has a large amount of computational throughput. Therefore, in order to approximate α ^k in [Equation 6], it can be converted into {1-k × (1-α)} and expressed as Equation 7. Here, an approximate value of α ^k and {1-k × (1-α)} will be described in detail with reference to FIG. 6.

(M (t + k): background image at time t + k, {1-k × (1-α)}: variable learning rate, α: fixed learning rate, M (t): background image at time t , I (t): pixel value of the captured image received from the camera 114 at time t, k: preset frame rate (T) / changed frame rate (S))

When the object recognizer 250 receives the current frame from the image receiver 210, the object recognizer 250 compares the current frame with the background image generated by the background image generator 240 and determines that the changed part is a moving object.

3 is a flowchart illustrating a region setting method for video surveillance according to the present embodiment.

The video surveillance apparatus 100 acquires an image using the provided camera 114 (S310). That is, in operation S310, the video surveillance apparatus 100 acquires an image of a fixed area (predetermined place). In this case, the camera 114 in the video surveillance apparatus 100 may transmit the acquired image to the line segment designation unit 122 connected by wire or wirelessly, and the line segment designation unit 122 may be obtained by using the mounted image acquisition unit. An image can be received. The video surveillance apparatus 100 receives a command by a user's manipulation or input (S320). In operation S320, the video surveillance apparatus 100 may receive a setting command for controlling the video surveillance apparatus 100 using a mouse, a keyboard, a joystick, a touch screen, and the like, which is the interface unit 112 provided therein.

The video surveillance apparatus 100 marks a plurality of line segments 710 and 712 based on a command through the interface unit 112 on one side and the other side of the image acquired through the provided camera 114 (S330). In operation S330, the image monitoring apparatus 100 displays an image acquired through the camera 114 through the display unit 116, and marks a line segment at a position corresponding to a command through the interface unit 112 on the corresponding image. . At this time, the length of the line segment is determined by the command through the interface unit 112.

The video surveillance apparatus 100 generates external connection lines 720 and 722 connecting both ends of the plurality of marked line segments 710 and 712 to each other, and the region within the external connection lines 720 and 722 is a region of interest 730. (S340). In operation S340, the external connection lines 720 and 722 may be automatically formed when the plurality of line segments 710 and 712 are marked on the image obtained regardless of the command through the interface unit 112. When the plurality of line segments 710 and 712 are marked on the acquired image, the image monitoring apparatus 100 generates a center connection line 810 connecting the midpoints of the plurality of line segments 710 and 712 with each other (S350). In operation S350, the center connection line 810 may be automatically formed when a plurality of line segments 710 and 712 are marked on an image acquired regardless of a command through the interface unit 112.

The video surveillance apparatus 100 generates an adjustment point 910 at an arbitrary point on the center connection line 810, and when the adjustment point 910 is moved based on an additional command through the interface unit 112, the adjustment point ( The region of interest 730 is maintained via 910 (S360). At this time, any point on the center connection line 810 is a position corresponding to the command through the interface unit 112. In operation S360, the video surveillance apparatus 100 controls the region of interest 730 to maintain the closed polygonal shape in the form of the moved adjustment point 910. In addition, the video surveillance apparatus 100 additionally generates a plurality of new adjustment points 910 based on a command through the interface unit 112, and when the plurality of adjustment points 910 is moved, a plurality of adjustments. The region of interest 730 is maintained via all of the points 910. That is, the video surveillance apparatus 100 controls the region of interest 730 to maintain the closed polygonal shape in the form of a plurality of moved adjustment points 910.

), 직각으로 교차되는 직선 속성(

), 곡선 속성(

), 자유 곡선 속성(

)을 포함한다. 이때, 영상 감시 장치(100)는 인터페이스부(112)를 통한 명령에 의해 곡선 속성, 자유 곡선 속성이 설정된 경우 관심 영역(730)을 원호 형태(

)로 변경하며, 직선 속성, 직각으로 교차되는 직선 속성이 설정된 경우 관심 영역(730)을 다각형 형태로 변경한다.The video surveillance apparatus 100 sets a property of the ROI 730 and changes the shape of the ROI 730 to a shape according to the set property (S370). At this time, the attribute is set corresponding to the command through the interface unit 112. Further, in step S370, the attribute is a linear attribute (

), Straight line attributes that intersect at right angles (

), Curve properties (

), Freehand line properties (

). In this case, when the curve property and the free curve property are set by the command through the interface unit 112, the video surveillance apparatus 100 may form an arc shape (

), And the ROI 730 is changed into a polygonal shape when a straight line property and a straight line property that cross at right angles are set.

On the other hand, the video surveillance apparatus 100 sets the object movement direction 1210 for the object. In this case, the object movement direction 1210 may be set based on a command through the interface unit 112. Here, the object movement direction 1210 is not only east, west, south and north directions (e.g., →, ↓, ↑, ←) but also diagonal directions (e.g., ↗, ↙, ↖, ↘) can also be set.

The video surveillance apparatus 100 detects an object passing through the ROI 730 (S380). In operation S370, the image monitoring apparatus 100 checks a frame rate of the current frame t among the acquired images, and if the frame rate is different from the preset frame rate, adjusts the learning rate for adjusting the image sensitivity according to the frame rate. Set, generate a background image of the current frame using the background image of the previous frame (t-1), the current frame (t), and the learning rate, and detect the object by comparing the background image of the current frame and the current frame. do.

In operation S380, the image monitoring apparatus 100 may detect the object movement direction 1210 by comparing the object of the current frame with the object of the previous frame. That is, the object direction detector 144 may detect the direction in which the object moves by using the object detected by the object detector 142. Also, the video surveillance apparatus 100 may display the object detected by the object detector 142 and the object movement direction 1210 detected by the object direction detector 144 through the display 116. In this case, the video surveillance apparatus 100 may control whether to be notified through the display unit 116 according to a preset condition (notification cycle, notification only when the object is detected, notification until the object movement direction, etc.).

In FIG. 3, steps S310 to S380 are described as being sequentially executed. However, this is merely illustrative of the technical idea of the present embodiment, and those of ordinary skill in the art to which the present embodiment pertains will be described. 3 may be modified and modified in various manners by changing the order described in FIG. 3 or executing one or more steps of steps S310 to S380 in parallel without departing from the essential characteristics, and thus, FIG. It is not limited.

As described above, the region setting method for image surveillance according to the present embodiment described in FIG. 3 may be implemented in a program and recorded in a computer-readable recording medium. The computer-readable recording medium having recorded thereon a program for implementing an area setting method for video surveillance according to the present embodiment includes all kinds of recording devices storing data that can be read by a computer system. Examples of such computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, etc., and also implemented in the form of a carrier wave (e.g., transmission over the Internet) . The computer readable recording medium may also be distributed over a networked computer system so that computer readable code is stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing the present embodiment can be easily inferred by programmers in the technical field to which the present embodiment belongs.

4 is a flowchart illustrating a method of detecting an object in which a learning rate is varied according to a frame rate to detect an object according to an exemplary embodiment.

The object detector 142 included in the video surveillance apparatus 100 receives a current frame photographed from the camera 114 (S410). The object detector 142 included in the video surveillance apparatus 100 generates an updated background image by applying a variable learning rate based on the current frame (S420). Here, the background image refers to a background image generated by using Equation 6 or Equation 7.

The object detector 142 included in the image monitoring apparatus 100 detects a moving object by comparing the background image generated in operation S410 with the current frame photographed by the camera 114 in operation S430. For example, if a bag exists in the current frame, and the background image generated in step S410 does not exist, the bag is compared with the current frame and the background image, and determined as an object moving the bag. The object detector 142 included in the video surveillance apparatus 100 transmits the detection information about the moving object detected in operation S430 to the display unit 116 so that an administrator may monitor the operation information (S440).

4, steps S410 to S440 are sequentially executed. However, this is merely illustrative of the technical idea of the present embodiment, and it is understood that any person skilled in the art will recognize that the present invention is not limited to this embodiment It will be understood that various changes and modifications may be made to the invention without departing from the essential characteristics thereof, or alternatively, by executing one or more of the steps S410 through S440, But is not limited thereto.

5 is a flowchart illustrating a method of generating a background image for detecting an object moving by varying a learning rate according to a frame rate for detecting an object according to an exemplary embodiment.

The object detector 142 provided in the video surveillance apparatus 100 receives an image captured by the camera 114 in operation S510. Here, the captured image includes the current frame t. The object detector 142 included in the video surveillance apparatus 100 checks the frame rate of the current frame t and compares the frame rate of the current frame t with a preset frame rate (S530). Here, the object detection unit 142 included in the video surveillance apparatus 100 processes the plurality of images at the same time, or increases the throughput by using additional functions (for example, garbage dumping, a service specialized in storing objects, etc.) The frame rate of the image photographed at 114 may be reduced. In order to confirm the change in the frame rate, step S530 is performed.

As a result of checking step S530, when the object detection unit 142 included in the video surveillance apparatus 100 determines that the frame rate for the current frame t is different from the preset frame rate, the frame for the current frame t is determined. The learning rate is set based on the rate (S540). Here, the learning rate is set to the learning rate based on [Equation 6] or [Equation 7]. On the other hand, as a result of confirming step S530, when the object detection unit 142 provided in the video surveillance apparatus 100 is confirmed that the frame rate for the current frame (t) is the same as the preset frame rate, by using the preset learning rate A background image is generated (S550).

The object detecting unit 142 included in the image monitoring apparatus 100 may update the background image updated by using the previously stored background image for the first previous frame t-1, the current frame t, and the learning rate set in operation S540. It generates (S560). Here, the background image is preferably a background image generated using Equation 6, but is not necessarily limited thereto. The background image may be a background image generated using Equation 7 to reduce the computational throughput. In addition, the object detection unit 142 provided in the video surveillance apparatus 100 preferably operates the background image recursively at a predetermined cycle.

In FIG. 5, steps S510 to S560 are described as being sequentially executed. However, this is merely illustrative of the technical idea of the present embodiment, and a person having ordinary knowledge in the technical field to which the present embodiment belongs may use the present embodiment. 5 may be modified and modified in various ways, such as by changing the order described in FIG. 5 or executing one or more steps of steps S510 to S560 in parallel without departing from the essential characteristics. It is not limited.

6 is an exemplary view for explaining a setting value of a learning rate that varies according to a frame rate according to the present embodiment.

The object detector 142 according to the present exemplary embodiment generates a background image to detect a moving object. Here, the object detector 142 generates a background image using Equation 6, but α ^k included in Equation 6 requires a large amount of computational processing as an exponential operation. The exponential calculation may be performed as it is, but in order to reduce the amount of calculation and the calculation time, the background image may be generated by approximating α ^k to {1-k × (1-α)} as shown in [Equation 7].

FIG. 6 is a table illustrating values comparing a learning rate of Equation 6 and a learning rate of Equation 7 with respect to a learning rate a having a value between 0.99 and 0.999. Here, the similarity between the two values is calculated as the learning rate of Equation 7 / learning rate × 100 of Equation 6 and displayed to the second decimal place.

As shown in FIG. 6, when a has a value of 0.998, it can be seen that {1-k × (1-α)} and α ^k coincide 100% at 1 to 5 k values. In addition, since the value of 0.99 or more has a similar value of 99% or more, the learning rate adjustment including the exponential operation of [Equation 6] can be calculated by replacing the learning rate adjustment using the multiplication operation of [Equation 7].

7 is an exemplary diagram illustrating line segments defining both ends of the ROI 730 according to the present embodiment.

The video surveillance apparatus 100 acquires an image as illustrated in FIG. 7 by using the provided camera 114. That is, as shown in FIG. 7, the video surveillance apparatus 100 acquires an image of a 'crossroads' which is a fixed area (predetermined place). The obtained image is displayed through the display unit 116 provided in the video surveillance apparatus 100.

Subsequently, a plurality of images based on a command through the interface unit 112 on one side and the other side of the acquired image (intersection image) in order to designate the ROI 730 to the 'intersection' image, which is the image obtained by the image monitoring apparatus 100. Mark the line segments 710 and 712. That is, the interface unit 112 provided with the video surveillance apparatus 100 receives a command for marking a line segment by a user's manipulation or input. The video surveillance apparatus 100 may receive a command for marking a line segment using a mouse, a keyboard, a joystick, a touch screen, and the like, which is the interface unit 112 provided. That is, when the image acquired through the camera 114 is displayed on the display unit 116, the image monitoring apparatus 100 marks a line segment at a position corresponding to a command through the interface unit 112 on the corresponding image. At this time, the length of the line segment is determined by the command through the interface unit 112. The video surveillance apparatus 100 generates external connection lines 720 and 722 connecting both ends of the plurality of marked line segments 710 and 712 to each other, and the region within the external connection lines 720 and 722 is a region of interest 730. Set to). In this case, the external connection lines 720 and 722 may be automatically made when a plurality of line segments 710 and 712 are marked on an image obtained regardless of the command through the interface unit 112.

That is, in order to designate the ROI 730, the video surveillance apparatus 100 may define two line segments 710 and 712 ('segment A') that define both ends of the ROI 730 as illustrated in FIG. 7. 710 and 'segment B' 712). The designation of the 'segment A' 710 and the 'segment B' 712 may be made through various interfaces such as input of coordinates or mouse click. In the video surveillance apparatus 100, a closed polygon that connects two end points of two specified line segments 'line segment A' 710 and 'line segment B' 712 is designated as the initial ROI 730.

8 is an exemplary view illustrating a center connection line of a region of interest according to the present embodiment.

When the plurality of line segments 710 and 712 are marked on the acquired image, the image monitoring apparatus 100 generates a center connection line 810 connecting the midpoints of the plurality of line segments 710 and 712 to each other. The center connection line 810 may be automatically formed when a plurality of line segments 710 and 712 are marked on an image acquired regardless of a command through the interface unit 112. That is, as shown in FIG. 8, the video surveillance apparatus 100 generates a central connection line 810 'O' connecting two line segments 'line segment A' 710 and 'line segment B' 712. . In this case, the line segment O 810 is connected to the midpoints O _A and O _B of the two line segments 'line segment A' 710 and 'line segment B' 712.

9 is an exemplary view showing an adjustment point on the center connecting line according to the present embodiment.

The video surveillance apparatus 100 generates an adjustment point 910 at an arbitrary point on the center connection line 810 using the provided adjustment point generation unit 128, and based on an additional command through the interface unit 112. When the adjustment point 910 is moved, the ROI 730 is maintained via the adjustment point 910. At this time, any point on the center connection line 810 is a position corresponding to the command through the interface unit 112.

That is, as shown in FIG. 9, the video surveillance apparatus 100 may generate an adjustment point 910 by clicking an arbitrary point on the line segment O 810, and the position of the generated adjustment point 910. By moving the shape of the 'line segment O' 810 can be modified. In this case, the video surveillance apparatus 100 may change the shape of the 'line segment O' 810 by moving the adjustment point 910.

10 is an exemplary view showing that the ROI is changed into an arc shape according to the present embodiment.

As illustrated in FIG. 10, the image surveillance apparatus 100 is changed into a shape having an arc having the same shape as the 'segment O'. In this case, the attributes of the region of interest 730 may be set in a straight line in addition to the curve as necessary, and thus the shape of the region of interest 730 is changed.

), 직각으로 교차되는 직선 속성(

), 곡선 속성(

), 자유 곡선 속성(

)을 포함한다. 이때, 영상 감시 장치(100)는 인터페이스부(112)를 통한 명령에 의해 곡선 속성, 자유 곡선 속성이 설정된 경우 관심 영역(730)을 원호 형태(

)로 변경한다.That is, the video surveillance apparatus 100 sets the property of the ROI 730 and changes the shape of the ROI 730 to a shape according to the set property. In this case, the property is a straight property (

), Straight line attributes that intersect at right angles (

), Curve properties (

), Freehand line properties (

). In this case, when the curve property and the free curve property are set by the command through the interface unit 112, the video surveillance apparatus 100 may form an arc shape (

).

11 is an exemplary view illustrating that the ROI is changed into a polygonal shape according to the present embodiment.

), 직각으로 교차되는 직선 속성(

), 곡선 속성(

), 자유 곡선 속성(

)을 포함한다. 이때, 영상 감시 장치(100)는 인터페이스부(112)를 통한 명령에 의해 직선 속성, 직각으로 교차되는 직선 속성이 설정된 경우 관심 영역(730)을 다각형 형태로 변경한다.As shown in FIG. 11, the shape of the ROI 730 may be changed into a polygon instead of an arc. That is, the video surveillance apparatus 100 sets the property of the ROI 730 and changes the shape of the ROI 730 to a shape according to the set property. In this case, the property is a straight property (

), Straight line attributes that intersect at right angles (

), Curve properties (

), Freehand line properties (

). In this case, the video surveillance apparatus 100 changes the ROI 730 into a polygonal shape when a straight line property and a straight line property that cross at right angles are set by a command through the interface unit 112.

12 is an exemplary view illustrating an object movement direction setting according to the present embodiment.

As illustrated in FIG. 12, the video surveillance apparatus 100 may designate a moving direction of an object to be observed according to a user's needs with respect to the set ROI 730. In this case, the center connecting line 810 penetrating the region of interest 730 refers to a moving path of the object passing through the region of interest 730, and the object moving direction 1210 is designated on the center connecting line 810 to observe. You can specify the moving direction of the object. As shown in FIG. 12, the region of interest 730 may be used to detect objects (vehicles) that enter from the bottom of the screen and turn left. The object movement direction 1210 may not be set depending on the application of video surveillance, or may be set in one or two directions.

In general, when the ROI 730 is designated by a simple polygon, it is difficult to designate and monitor an object moving along various movement paths rather than a simple linear movement path. It is easy to monitor the object moving in a specific path instead of the object passing through the area 730 in a straight line. That is, as shown in FIG. 12, the left-turning vehicle (object) can be detected. The video surveillance apparatus 100 sets the object movement direction 1210 for the object based on the command through the interface unit 112. Here, the object movement direction 1210 is not only east, west, south and north directions (eg, →, ↓, ↑, ←), but also diagonal directions (eg, ↗, ↙, ↖, ↘) can also be set. Also, the video surveillance apparatus 100 detects an object passing through the ROI 730. The video surveillance apparatus 100 checks a frame rate of the current frame t among the acquired images, sets a learning rate for adjusting image sensitivity according to the frame rate when the frame rate is different from the preset frame rate. A background image of the current frame is generated using the background image of the previous frame t-1, the current frame t, and the learning rate, and the object is detected by comparing the current frame with the background image of the current frame. Thereafter, the video surveillance apparatus 100 may detect the object movement direction 1210 by comparing the object of the current frame with the object of the previous frame. That is, the video surveillance apparatus 100 may detect a direction in which the object moves by using the detected object.

13 is an exemplary view showing a change of interest area through a plurality of adjustment points according to the present embodiment.

As shown in FIG. 13, the shape of the region may be changed through one or more adjustment points 910. That is, the video surveillance apparatus 100 controls the region of interest 730 to maintain the closed polygonal shape through the moved adjustment point 910. In addition, the video surveillance apparatus 100 additionally generates a plurality of new adjustment points 910 based on a command through the interface unit 112, and when the plurality of adjustment points 910 is moved, a plurality of adjustments. The region of interest 730 is maintained via all of the points 910. That is, the video surveillance apparatus 100 controls the region of interest 730 to maintain the closed polygonal shape through the plurality of moved adjustment points 910.

14 is an exemplary view illustrating that the region of interest according to the present embodiment is applied to detecting a U-turn vehicle.

As illustrated in FIG. 14, the object may be used to detect an object having a specific movement path, such as the detection of a U-turn vehicle. In this case, the ROI 730 and the ROI may be intuitively set. The video surveillance apparatus 100 sets the object movement direction 1210 for the object by using the movement direction setting unit 134 provided. In this case, the object movement direction 1210 may be set based on a command through the interface unit 112. Here, the object movement direction 1210 is not only east, west, south and north directions (eg, →, ↓, ↑, ←), but also diagonal directions (eg, ↗, ↙, ↖, ↘) can also be set. Thereafter, the video surveillance apparatus 100 may detect the object movement direction 1210 by comparing the object of the current frame with the object of the previous frame. That is, the object direction detector 144 may detect the direction in which the object moves by using the object detected by the object detector 142. In addition, the image monitoring apparatus 100 displays the object detected by the object detector 142 and the object movement direction 1210 detected by the object direction detector 144 using the notification unit 146 provided with the display unit 116. Can be represented by

The foregoing description is merely illustrative of the technical idea of the present embodiment, and various modifications and changes may be made to those skilled in the art without departing from the essential characteristics of the embodiments. Therefore, the present embodiments are to be construed as illustrative rather than restrictive, and the scope of the technical idea of the present embodiment is not limited by these embodiments. The scope of protection of the present embodiment should be construed according to the following claims, and all technical ideas within the scope of equivalents thereof should be construed as being included in the scope of the present invention.

As described above, the present embodiment is applied to various video surveillance fields, and sets various regions of interest of an image acquired through the video surveillance apparatus, and sets an object moving direction to generate an effect of detecting a moving object. It is a useful invention.

112: interface unit 114: camera
116: display unit 122: line segment designator
124: region of interest setting unit 126: connection line generator
128: control point generation unit 132: property setting unit
134: movement direction setting unit 142: object detection unit
144: object direction detection unit 146: notification unit

Claims (12)

An image acquiring unit acquiring an image;
An interface unit for receiving a command by a user's operation;
A line segment designator which marks a plurality of line segments on one side and the other side of the image acquired based on the command;
A region of interest setting unit configured to generate an external connection line connecting both ends of the plurality of line segments to each other, and to set a region within the external connection line as a region of interest; And
An attribute setting unit which sets an attribute of the ROI based on the command and changes the form of the ROI into a form according to the set attribute;
Video surveillance apparatus comprising a. The method of claim 1,
An object detector detecting an object passing through the ROI; And
Movement direction setting unit for setting the object movement direction for the object
Video surveillance apparatus comprising a further. 3. The method of claim 2,
The object detector checks a frame rate of the current frame (t) of the image, and if the frame rate is different from a preset frame rate, sets the learning rate for adjusting the image sensitivity according to the frame rate. A background image of the current frame is generated using the background image of the previous frame t-1, the current frame t, and the learning rate, and the background image of the current frame is compared with the current frame. And detecting the object. 3. The method of claim 2,
An object direction detection unit configured to detect an object moving direction by comparing an object of a current frame with an object of a previous frame of the image; And
Notification unit for indicating the detected object and the object movement direction through a display unit
Video surveillance apparatus comprising a further. The method of claim 1,
A connecting line generating unit generating a central connecting line connecting the midpoints of the plurality of line segments with each other; And
An adjustment point generation unit generates an adjustment point at an arbitrary point on the center connection line based on the command, and maintains the region of interest via the adjustment point when the adjustment point is moved based on the command.
Video surveillance apparatus comprising a further. The method of claim 5, wherein
The adjustment point generation unit may further generate a plurality of new adjustment points based on the command, and maintain the region of interest via all the plurality of adjustment points when each of the plurality of adjustment points is moved. Video monitoring device. The method of claim 1,
The property is a straight property (

), Straight line attributes that intersect at right angles (

), Curve properties (

), Freehand line properties (

Wherein, the property setting unit is characterized by the circular arc shape (the arc of interest if the curve property, the free curve property is set by the command)

) And change the ROI to a polygonal shape when the straight line property and the straight line property that cross at right angles are set. The method of claim 1,
The ROI setting unit,
And setting the ROI by connecting the external connection line in the form of a closed traverse. The method of claim 1,
Wherein,
And obtaining the image with respect to the fixed area. In the method for setting the area by the video surveillance apparatus,
A user input process of receiving a command by a user's operation;
A line segment designation process of marking a plurality of line segments on one side and the other side of the image acquired based on the command;
A region of interest setting step of generating an external connection line connecting both ends of the plurality of line segments to each other and setting an area within the external connection line as a region of interest; And
An attribute setting process of setting an attribute of the region of interest based on the command and changing the form of the region of interest to a form according to the set attribute;
An area setting method for video surveillance comprising a. 11. The method of claim 10,
A connecting line generating process of generating a central connecting line connecting the midpoints of the plurality of line segments with each other; And
An adjustment point generation process of generating an adjustment point at an arbitrary point on the center connection line based on the command, and maintaining the region of interest via the adjustment point when the adjustment point is moved based on the command.
The area setting method for video surveillance, characterized in that it further comprises. 11. The method of claim 10,
An object detection process of detecting an object passing through the ROI; And
Moving direction setting process of setting an object moving direction for the object
The area setting method for video surveillance, characterized in that it further comprises.

Images