KR101704471B1 - Fall detection apparatus and method thereof - Google Patents

Abstract

The present invention relates to a fall detection apparatus and a method thereof. The fall detection apparatus comprises: a photographing unit which obtains an image for a photographing area; and a monitoring module which represents a dynamic object existing within an image as an operation area bounding box, and recognizes the variation in the size of the operation area bounding box by using an operation line and a reference line to monitor the fall detection of the dynamic object.

Description

[0001] Fall detection apparatus and method [

The present invention relates to a fall detection apparatus and a method thereof, and more particularly, to a fall detection apparatus and a method thereof, in which a dynamic object is detected in an image obtained from a camera, a bounding box is set in the obtained dynamic object, The present invention relates to a fall detection apparatus and a method thereof.

The development of medical care, the improvement of living standards, and the expansion of medical services have contributed to an increase in the life span and an increase in the elderly population. As the proportion of the elderly population increases, the number of elderly people living alone has increased, and the importance of health for the elderly who need protection has increased.

In the case of falls, it accounts for a high percentage of the elderly accidents and poses a serious social problem.

Since elderly people are less able to cope with dangerous situations, measures are needed to deal with dangerous situations, such as falls, as they can cause serious conditions if measures are not taken.

In a situation where it is widely acknowledged that falling is a dangerous factor to the elderly, research activities on fall related topics have been continuing in various fields. Conventional fall-related technologies are mainly proposed using devices.

The device is attached to the body like an acceleration sensor or a gyro sensor, and is recognized as a fall when a specific value or more appears. However, it is troublesome to install a sensor or attach it to the body every place.

Domestic Publication No. 10-2012-0114629 Korean Publication No. 10-2013-0094489 Domestic Publication No. 10-2009-91539

 SUMMARY OF THE INVENTION Accordingly, the present invention has been made keeping in mind the above problems occurring in the prior art, and an object of the present invention is to provide a fall detection apparatus and method for detecting a dynamic object using an image obtained from a camera.

Another object of the present invention is to provide a fall detection apparatus and a method thereof that can alert a danger by generating a warning event in real time when a fall is detected.

According to an aspect of the present invention, there is provided an image pickup apparatus including: And when a dynamic object appears in the image obtained through the photographing unit, a dynamic object existing in the image is represented by an operation region bounding box, and the size change of the operation region bounding box is grasped by using an operation line and a reference line And a monitoring module for monitoring occurrence of falling of the dynamic object.

According to another aspect of the present invention, when a dynamic object appears in an image acquired through the photographing unit, the monitoring module recognizes the dynamic object, and uses the dynamic range of the recognized dynamic object and the dynamic position of the dynamic object, A setting unit configured to set an operation region bounding box of the apparatus; And a determination unit for determining a size change of the bounding box set by the setting unit by using an operation line and a reference line and determining whether a dynamic object falls in accordance with a size change of the operation region bounding box.

According to another aspect of the present invention, there is provided a surveillance module including: a detection unit that detects dynamic object appearance image frames from consecutive color image frames constituting a color image transmitted from the photographing unit based on a background image frame; And an extracting unit for extracting dynamic object feature point data through binarization in the dynamic object appearance image frame detected through the detecting unit, wherein the setting unit is configured to extract the dynamic object feature point data using the dynamic object feature point data extracted through the extracting unit, And recognizes dynamic objects existing in each of the detected dynamic image object image frames.

Further, the determination unit of one aspect of the present invention determines that the fall occurs when the operation line moves below the reference line with respect to the ground.

The monitoring module may further include an acceleration detector for measuring an acceleration, which is a change in size of the bounding box with respect to time, and the determining unit determines that the acceleration falls when the acceleration measured by the acceleration detector is equal to or greater than a threshold value .

According to another aspect of the present invention, there is provided a surveillance module comprising: an abnormal operation image storage unit for storing an image corresponding to a fall occurring in an image transmitted from the photographing unit; And an output unit outputting an alarm for informing occurrence of a fall when the fall of the dynamic object occurs.

According to another aspect of the present invention, there is provided a method for controlling an image capturing apparatus, the method comprising: (A) (B) if a dynamic object appears in an image acquired by the monitoring module through the photographing unit, expressing a dynamic object existing in the image by a bounding box; And (C) monitoring the change of size of the bounding box of the operation area by using the operation line and the reference line to monitor occurrence of a fall of the dynamic object.

The step (B) of the other aspect of the present invention may further comprise: (B-1) recognizing a dynamic object when the dynamic object appears in the image acquired by the monitoring module through the photographing unit; And (B-2) setting an operation region bounding box of the dynamic object using the operation region size and the operation position of the recognized dynamic object by the monitoring module.

In another aspect of the present invention, the step (C) includes the steps of: (C-1) tracking the size change of the bounding box set by the monitoring module using an operation line and a reference line; And (C-2) a step in which the monitoring module determines whether a dynamic object falls in accordance with a size change of an operation region bounding box.

According to another aspect of the present invention, in the step (B-1) of the present invention, the monitoring module detects a dynamic object appearance image frame among consecutive color image frames constituting a color image transmitted from the photographing unit, ; Extracting dynamic object feature point data through binarization within a dynamic object appearance image frame in which the monitoring module is detected; And recognizing a dynamic object existing in each detected dynamic object appearance image frame using the dynamic object feature point data extracted by the monitoring module.

Further, the monitoring module according to another aspect of the present invention determines fall if the operation line moves below the reference line with respect to the ground.

According to another aspect of the present invention, there is also provided a method of measuring an acceleration, comprising the steps of: (D) measuring an acceleration in which a monitoring module changes in size with respect to time of the bounding box; And (E) determining if the acceleration measured by the monitoring module is equal to or greater than a threshold value.

According to another aspect of the present invention, there is provided a surveillance module, comprising: (F) storing an image corresponding to a fall occurring in an image transmitted from the photographing unit; And (G) the monitoring module further includes the step of outputting an alarm to notify occurrence of a fall when the fall of the dynamic object occurs.

Through the present invention, it is possible to provide convenience to the user by judging whether or not a fall occurs through the image processing in the room, and it is possible to accurately determine whether a dangerous behavior such as a fall occurs to the elderly living alone.

1 is a configuration diagram of a fall detection apparatus according to a preferred embodiment of the present invention.
FIG. 2 shows the changes of the X and Y coordinates in the fall state, which is used as an example of explanation for deriving a feature value.
FIG. 3 also shows changes in the X and Y coordinates in the fall state used as an example of explanation for deriving the feature value.
FIG. 4 also shows changes in the X and Y coordinates in the fall state, which is used as an example of explanation for deriving a feature value.
5 is a flowchart of a fall detection method according to a preferred embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is capable of various modifications and various embodiments, and specific embodiments will be described in detail below with reference to the accompanying drawings.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

The terms first, second, etc. may be used to describe various components, but the components are not limited by the terms, and the terms are used only for the purpose of distinguishing one component from another Is used.

1 is a configuration diagram of a fall detection apparatus according to a preferred embodiment of the present invention.

Referring to FIG. 1, a fall detection apparatus according to a preferred embodiment of the present invention includes a photographing unit 10 and a monitoring module 20. The photographing unit 10 and the monitoring module 20 can communicate with each other through a network such as the Internet network.

The photographing unit 10 acquires an image of the photographing region through the photographing of a specific region in the installation space and delivers the image to the monitoring module 20 in real time. Here, the image may be a color image or an infrared image.

The photographing unit 10 includes a color camera that operates in the daytime to acquire a color image, an infrared camera that operates at night to acquire an infrared image, a light sensor or a timer to check whether there is illumination or daytime / A controller for controlling the operation of the camera and the infrared camera, and the like.

When the dynamic object appears in the image transmitted from the photographing unit 10, the monitoring module 20 displays the dynamic object existing in the image as a bounding box of the motion region, and changes the size of the bounding box of the motion region bounding box And monitoring the occurrence of abnormal motion of the dynamic object according to the acceleration change. At this time, it is preferable that the abnormal operation includes an operation of suddenly falling down, that is, a falling operation and a falling operation.

The monitoring module 20 includes an abnormal operation image storing unit 21, a background image storing unit 22, a background image managing unit 23, a detecting unit 24, an extracting unit 25, a setting unit 26, A determination unit 28, and an acceleration detection unit 29. [0033]

Here, the abnormal operation image storage unit 21 stores the image corresponding to the abnormal operation occurrence from the image transmitted from the photographing unit 10. [

The abnormal operation image storage unit 21 may include a buffer module and a storage module.

The buffer module stores an image transmitted from the photographing unit 10, and stores only a recent image. That is, when the buffer module receives an image from the photographing unit 10, the buffer module stores consecutive image frames constituting the image in accordance with the reproduction order, and stores the image frame that has elapsed from the initial storage time of the stored image frames Only the recent image frames are saved while continuously removing.

When the abnormal operation of the dynamic object occurs, the storage module acquires the corresponding image from the buffer module for a predetermined time before and after the occurrence of the abnormal operation, and stores the corresponding image in the occurrence of the abnormal operation

Meanwhile, the background image storage unit 22 stores a background image frame.

The background image management unit 23 acquires a background image frame from the image transferred from the photographing unit 10 through a background subtraction method and stores the acquired background image frame in the background image storage unit 22. [ As the background separation technique, an adaptive Gaussian Mixture Model and an improved Double Frame Difference technique, which are statistical background models, can be used.

That is, the background image managing unit 23 designates any one of continuous image frames constituting the image transmitted from the photographing unit 10 as a temporary background image frame, and then outputs the temporary background image frame and the photographing unit 10, (Luminance, etc.) between consecutive image frames constituting an image transmitted from the moving object, and distinguishes a dynamic object having motion from the background without motion in each image frame, And generates a background image frame composed only of the background, and stores the generated background image frame in the background image storage unit 22. FIG.

The background image management unit 23 compares the background image frame stored in the background image storage unit 22 and the consecutive image frames constituting the image transmitted from the photographing unit 10, A dynamic object in motion and a background without motion are distinguished from each other, and then a dynamic object is extracted from the image frame to generate a new background image frame composed of only background, and the generated new background image frame is stored in the background image storage unit 22 , The background image frame stored in the background image storage unit 22 can be updated periodically.

In addition, when the background image management unit 23 extracts a dynamic object from an image frame and generates a background image frame only of the background, a morphological operation and a labeling algorithm are performed on the extracted area of the dynamic object Noise such as a small hole (Hole) can be removed.

The detection unit 24 detects dynamic object appearance image frames from consecutive image frames constituting an image transmitted from the photographing unit 10 on the basis of the background image frame stored in the background image storage unit 22.

For example, the detection unit 24 calculates the pixel value difference x for each pixel between the background image frame stored in the background image storage unit 22 and each image frame transmitted from the photographing unit 10, A pixel having a value difference equal to or greater than a threshold value T is determined as a pixel in which the motion has occurred, and a pixel having a threshold value or less is determined as a background image.

The detection unit 24 performs binarization by displaying f (x) as a pixel value of a pixel serving as a background image and denoting a pixel value of a pixel where motion occurs as f (x) = 255.

Binarization is an operation that shows the image as 0 and 1, as shown in Equation (1). In the present invention, an object detected through a difference image is represented by 255 pixel values instead of 1, and a background is represented by 0 to classify objects.

(1)

The extraction unit 25 extracts dynamic object feature point data including the position, color, pixel value, and the like of the motion-detected pixels detected through the detection unit 24 based on the background image frame stored in the background image storage unit 22 .

The setting unit 26 recognizes a dynamic object existing in each dynamic object appearance image frame detected through the detection unit 24 using the dynamic object feature point data extracted through the extraction unit 25, And then sets the bounding box of the dynamic range of the dynamic object for each dynamic object appearance image frame using the extracted dynamic range of the dynamic object and the operation position. Here, the operation region bounding box preferably includes a minimum rectangular shape including a dynamic object.

The setting unit 26 sets the operation line A and the reference line B to extract the dynamic range of the dynamic object existing in each dynamic object appearance image frame and the operation position.

2, X _max denotes the maximum value of the X axis in the rectangular coordinates of the bounding box, and Y _min denotes the minimum value of the Y axis in the rectangular coordinates of the bounding box . These operating lines are represented by green lines (Green) in Figs. 3 and 4.

(2)

Then, the reference line following and represented by Equation 3, X _min to refer to Figure 2 means the X-axis minimum value in the Cartesian coordinates of the bounding box, Y _max refers to the maximum Y-axis in an orthogonal coordinate axis of the bounding box .

These baselines are represented by the red lines in FIGS. 3 and 4.

(3)

FIG. 2 shows the changes of the X and Y coordinates in the fall state, which is used as an example of explanation for deriving a feature value. For the left fall, X _min and X _max decrease and Y _min and Y _max increase. In case of right fall, X _min and X _max increase, Y _min and Y _max also increase. When the fall state is the distance Y _min and Y _max was distant narrower distance X _min and X _max when compared to a normal state.

The operating line of Equation (2) is composed of larger values of the increase amount in the fall state as compared with the steady state.

On the other hand, the reference line of Equation (3) is composed of values that decrease and decrease in the fall state as compared with that in the steady state.

On the other hand, the output unit 27 outputs an alarm to notify the occurrence of an abnormal operation when an abnormal operation of the dynamic object occurs.

The output unit 27 may include a sound module, a display module, and the like.

The sound module outputs an abnormal operation occurrence alarm in the form of an alarm sound or an alarm alarm when an abnormal operation of a dynamic object occurs.

The display module displays at least one of the alarm message phrase / glyph and the abnormal operation occurrence image stored in the abnormal operation image storage unit 21 when the abnormal operation of the dynamic object occurs.

The determination unit 28 determines whether or not an abnormal operation of the dynamic object is generated by tracking the change of the operation line based on the reference line of the bounding box set in each dynamic object occurrence image frame and outputs the determination result to the abnormal operation image storage unit 21. [ And an output unit 27, as shown in Fig.

3 (a), when the operation line is above the reference line, the determination unit 28 determines that the operation line is in a normal state (when the value of the operation line is larger than the value of the reference line) If the motion line is below the reference line (if the value of the motion line is smaller than the value of the reference line), it is judged as a fall state.

Such a fall condition is similar to FIG.

On the other hand, when the operation line and the reference line according to Equations 2 and 3 are used, the determination unit 28 may perform erroneous detection when the range is large. In order to solve this problem, acceleration is applied in the present invention.

The acceleration detector 29 measures a change in size of the bounding box on the Y axis over time and determines the acceleration, and provides the determined acceleration to the determiner 28. [

The acceleration detecting section 29 can calculate the acceleration V _miny by the following equation (4). Here, t is time, and k and k-1 is the time means the order of frame, t _k is the time of k-th frame to be photographed, t _k-1 is the k-1 frames is taken, Y _mink is k and Y _min values in the frame, Y _{min (k-1)} is a Y _min values in the k-1 frames.

(4)

The determination unit 28 determines the acceleration threshold T _v is more than a fall time. That is, the determination unit 28 determines that the operation line is below the reference line and falls when the acceleration becomes equal to or greater than the threshold value.

In this regard, since the acceleration uses the two-dimensional pixel information of the image, the magnitude of the varying value varies depending on the distance from the camera. For this problem, the threshold value T _v is also defined the following equation (5), so as to be changed according to the distance.

(5)

Equation (5) is a threshold value T _v is changed according to the distance between the person. The Y _max coordinates at the end of the human foot are used in the image. The threshold T _v increases as the distance between the person and the camera increases, and the threshold T _v decreases as the distance increases.

If the equation (1) is equal to or greater than Equation (2), that is, when _{B is} equal to or greater than T _B and the Equation (3) is equal to or greater than T _v , the determination unit (28) determines a fall and generates a warning event to the user.

In the above-described configuration, a fall detection method according to an embodiment of the present invention will be described with reference to FIG.

First, the photographing unit 10 transmits an image of the photographing region to the monitoring module 20 in real time through photographing a specific region in the installation space (S100).

The image transmitted to the monitoring module 20 through the step S100 may be a color image or an infrared image depending on whether there is illumination or whether it is daytime / nighttime.

After the step S100, the monitoring module 20 acquires a background image frame from the image through the background separation technique (S200).

In step S200, the monitoring module 20 designates one of consecutive image frames constituting an image as a temporary background image frame, and then forms a temporary background image frame and an image transmitted through the step S100 The dynamic object in motion and the motionless background are separated and recognized in each image frame through the comparison of pixel values between consecutive image frames. Then, the dynamic object is extracted from the image frame to generate a background image frame composed of only background And stores the generated background image frame therein.

In step S200, the monitoring module 20 compares pixel values between a background image frame stored therein and consecutive image frames constituting an image transmitted through the step S100, The dynamic object is separated from the motionless background, and then the dynamic object is extracted from the image frame to generate a new background image frame composed only of the background, and the generated new background image frame is stored therein, It is possible to periodically update the background image frame.

In step S300, the monitoring module 20 detects the dynamic object appearance image frames from the consecutive image frames constituting the image transmitted through the step S100 on the basis of the background image frame.

In step S300, the monitoring module 20 calculates a pixel value difference between the background image frame obtained through the step S200 and each image frame transmitted through the step S100. If the calculated pixel value difference is equal to or greater than the threshold value The pixel is determined as a pixel in which the motion has occurred, and the image frame having the pixel in which the motion has occurred is detected as the dynamic object appearance image frame.

After step S300, the monitoring module 20 recognizes dynamic objects existing in each dynamic object appearance image frame (S400).

In step S400, the monitoring module 20 extracts the dynamic object feature point data including the position, color, and pixel value of the motion-detected pixels within the detected dynamic object appearance image frame through the above-described step S300, The dynamic object existing in the dynamic object appearance image frame can be recognized by using the dynamic object feature point data.

Next, in step S400, the monitoring module 20 extracts an operation area size and an operation position of a dynamic object existing in each dynamic object appearance image frame, and then uses the size and the operation position of the extracted dynamic object The bounding box of the dynamic range of the dynamic object for each dynamic object appearance image frame is set (S500).

In the above step S500, the operation area bounding box preferably includes a minimum rectangular shape including a dynamic object.

In step S500, the monitoring module 20 may use a binarization method to extract the dynamic range of the dynamic object existing in each image frame and the operation position.

After the above step S500, the monitoring module 20 sets the operation line and the reference line in the bounding box set in each dynamic object appearance image frame, calculates the acceleration indicating the movement of the bounding box by time, In operation S600, a size change of the operation area bounding box is tracked by a change in acceleration, and a fall occurs in the dynamic object according to a size change of the operation area bounding box.

In step S600, the monitoring module 20 tracks the movement of the motion line based on the reference line of the bounding box set in each dynamic object appearance image frame to determine whether a dynamic object falls.

3 (a), when the operation line is above the reference line, the monitoring module 20 determines that the operation line is in the normal state. Referring to FIG. 3 (b) State.

On the other hand, if the operation line and the reference line according to Equations 2 and 3 are used, the monitoring module 20 may perform erroneous detection when the range is large. In order to solve this problem, acceleration is applied in the present invention.

Monitoring module 20 determines the acceleration by measuring the size change with time of the bounding box on the Y-axis, and the acceleration is determined as a threshold value T when _v is more than falling.

That is, the monitoring module 20 determines that the operation line is below the reference line, and falls when the acceleration reaches the threshold value or more.

In this regard, since the acceleration uses the two-dimensional pixel information of the image, the magnitude of the varying value varies depending on the distance from the camera. Because of this problem, the threshold value T _v also needs to be changed according to the distance. Therefore, the Y _max coordinate located at the tip of the human foot is used in the image. The threshold value T _V increases as the distance between the person and the camera increases, It has a characteristic that the threshold value T _v is reduced.

The monitoring module 20 determines that the fall occurs when Equation 1 is equal to or greater than Equation 2, that is, when _{B is} equal to or greater than T _B and Equation 3 is equal to or greater than T _v .

As a result of the determination in step S600, when the abnormal operation of the dynamic object occurs, the monitoring module 20 outputs an alarm for notifying of occurrence of an abnormal operation (S700).

In step S700, the surveillance module 20 stores an abnormality occurrence image in the inside, outputs an alarm message phrase / pattern, and an abnormally stored abnormality occurrence while outputting an alarm sound or an alarm alarm when an abnormal operation of a dynamic object occurs And at least one of the images can be displayed.

According to the present invention as described above, it is possible to provide convenience to the user by judging whether or not a fall occurs through image processing in the room, and it is possible to precisely determine whether a dangerous behavior such as a fall occurs to the elderly living alone.

10: photographing part 20: monitoring module
21: abnormal operation image storage unit 22: background image storage unit
23: Background video management unit 24:
25: extracting unit 26: setting unit
27: output unit 28:
29: Acceleration detector

Claims (19)

A photographing unit for acquiring an image of a photographing area; And
When a dynamic object appears in the image obtained through the photographing unit, a dynamic object existing in the image is expressed by a bounding box of the motion region, and the size change of the motion region bounding box is grasped by using the motion line and the reference line, And a monitoring module for monitoring a fall occurrence of the object,
The monitoring module includes:
A setting unit for recognizing a dynamic object when the dynamic object appears in the image acquired through the photographing unit and setting an operation region bounding box of the dynamic object by using the size and the operation position of the recognized dynamic object; And
And a determination unit that tracks a size change of the bounding box set through the setting unit using an operation line and a reference line and determines whether a dynamic object falls in accordance with a size change of the operation region bounding box,
Wherein the operation line used by the determination unit is expressed by the following equation (2), and the reference line is expressed by the following equation (3).

(2)

Where A is the operating line, X _max is the maximum value of the X axis in the Cartesian coordinates of the bounding box, and Y _min is the minimum value of the Y axis in the Cartesian coordinates of the bounding box.

(3)

Where B is the operating line, X _min is the minimum value of the X axis in the Cartesian coordinates of the bounding box, and Y _max is the Y axis maximum value in the Cartesian coordinates of the bounding box. delete The method according to claim 1,
The monitoring module includes:
A detector for detecting dynamic object appearance image frames among consecutive color image frames constituting a color image transmitted from the photographing unit on the basis of a background image frame; And
And an extraction unit for extracting dynamic object feature point data through binarization in the dynamic object appearance image frame detected through the detection unit,
Wherein the setting unit recognizes a dynamic object existing in each dynamic object occurrence image frame detected through the detection unit using the dynamic object feature point data extracted through the extraction unit. delete The method according to claim 1 or 3,
Wherein the determination unit determines that the fall occurs when the operation line moves below the reference line based on the ground surface. The method of claim 5,
The monitoring module
And an acceleration detector for measuring an acceleration, which is a magnitude change with respect to time of the bounding box,
Wherein the determination unit determines that the acceleration is fall if the acceleration measured by the acceleration detection unit is equal to or greater than a threshold value. The method of claim 6,
_Wherein the acceleration detecting unit calculates the acceleration V _miny by the following equation (4).
(4)

Here, t is time, and k and k-1 is the time means the order of frame, t _k is the time of k-th frame to be photographed, t _k-1 is the k-1 frames is taken, Y _mink is k In the frame, Y _{min is the} value, and Y _{min (k-1)} is the Y _min value in the k-1 frame. The method of claim 6,
The determining unit fall detection device of the threshold T _v is to be changed according to the distance by Equation (5) below.

(5)

Where Y _max is the Y axis maximum value in the Cartesian coordinate axis of the bounding box. The method according to claim 1 or 3,
The monitoring module
An abnormal operation image storage unit for storing an image corresponding to a fall occurring in the image transmitted from the photographing unit; And
And an output unit for outputting an alarm for notifying occurrence of a fall when the fall of the dynamic object occurs. (A) transmitting, in real time, an image of a shooting region in a photographing unit to a monitoring module;
(B) if a dynamic object appears in an image acquired by the monitoring module through the photographing unit, expressing a dynamic object existing in the image by a bounding box; And
(C) monitoring module monitors changes in the size of the bounding box of the operation area by using an operation line and a reference line to monitor occurrence of a fall of the dynamic object,
The step (C)
(C-1) tracking the size change of the bounding box set by the monitoring module using an operation line and a reference line; And
(C-2) The monitoring module determines whether a dynamic object falls in accordance with a size change of an operation region bounding box,
Wherein the operation line used by the monitoring module is expressed by the following equation (2), and the reference line is expressed by the following equation (3).

(2)

Where A is the operating line, X _max is the maximum value of the X axis in the Cartesian coordinates of the bounding box, and Y _min is the minimum value of the Y axis in the Cartesian coordinates of the bounding box.

(3)

Where B is the operating line, X _min is the minimum value of the X axis in the Cartesian coordinates of the bounding box, and Y _max is the Y axis maximum value in the Cartesian coordinates of the bounding box. 12. The method of claim 10,
The step (B)
(B-1) recognizing a dynamic object when a dynamic object appears in an image acquired by the monitoring module through the photographing unit; And
(B-2) setting the bounding box of the dynamic range of the dynamic object by using the dynamic range and the dynamic range of the dynamic object recognized by the monitoring module. delete 12. The method of claim 11,
Wherein the step (B-1)
Detecting, by the monitoring module, dynamic object appearance image frames from consecutive color image frames constituting a color image transmitted from the photographing unit based on a background image frame;
Extracting dynamic object feature point data through binarization within a dynamic object appearance image frame in which the monitoring module is detected; And
And the monitoring module recognizing a dynamic object existing in each detected dynamic object appearance image frame using the extracted dynamic object feature point data. delete The method according to any one of claims 10, 11 and 13,
Wherein the monitoring module determines that the fall occurs when the operation line moves below the reference line with respect to the ground. The method of claim 15,
(D) measuring an acceleration in which the monitoring module is a size change of the bounding box with respect to time; And
(E) judging the fall as a fall if the acceleration measured by the monitoring module is not less than a threshold value. The method of claim 16,
_Wherein the acceleration _Vminy measured by the monitoring module is calculated by the following equation (4).
(4)

Here, t is time, and k and k-1 is the time means the order of frame, t _k is the time of k-th frame to be photographed, t _k-1 is the k-1 frames is taken, Y _mink is k In the frame, Y _{min is the} value, and Y _{min (k-1)} is the Y _min value in the k-1 frame. The method of claim 16,
Fall detection method in which the monitoring module threshold T _v is to be changed according to the distance by Equation (5) below.

(5)

Where Y _max is the Y axis maximum value in the Cartesian coordinate axis of the bounding box. The method according to any one of claims 10, 11 and 13,
(F) the monitoring module stores an image corresponding to a fall occurring from the image transmitted from the photographing unit; And
(G) The monitoring module further includes outputting an alarm to notify occurrence of a fall when the fall of the dynamic object occurs.

Images