JPH027195A - Abnormality monitoring system - Google Patents

Abstract

PURPOSE:To prevent a malfunction and to save a laber by successively registering a picture at normal time photographed by a video camera and pictures photographed after that, extracting each changed amount, and acacating an alarm device when the changed amount exceeds a prescribed abnormality limit. CONSTITUTION:A video camera 1 is previously set to an area A to be monitored such as a hospital and an old people's home, the picture photographed at the normal time is stored in a computer 2 such as a personal computer, the stored picture is successively registered with the pictures photographed after that, arithmetic processing for obtaining the difference between a currently registered picture and the preceding picture is executed, and the changed amount is extracted. Next, when the changed amount exceeds the previously determined prescribed abnormality limit value, a signal is emitted from the computer 2, and an alarm device 4 in an alarm center B is actuat ed. For this alarm device 4, an alarm, an alarm lamp, etc., which are suitable, are to be used, and by the actuation of the alarm device 4, a monitoring person confirms the abnormal state of the area A by a CRT6 as a picture display device for a monitor provided in the monitoring center B and executes a countermeasure for the abnormal ity.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention can be used, for example, in nursing homes, hospitals, and other similar facilities to monitor abnormal behavior, fires, etc.
The present invention relates to an abnormality monitoring system that issues an alarm and notifies a supervisor when the state exceeds an abnormality limit.

[Conventional technology]

In recent years, nursing homes, hospitals, etc. have been preparing for fire outbreaks, abnormal behavior of lazy elderly people, or intruders from outside.
A large number of video cameras are installed in the monitored area, and the images are monitored using a CRT at a monitoring center.

In this case, even when there is no abnormal situation, the supervisor should always monitor the CR.
It is difficult to continue monitoring in front of the CRT, and it is extremely wasteful and costly to keep personnel stationed in front of the CRT at all times, especially at night when abnormal situations are most likely to occur.

On the other hand, conventional fire alarm devices detect smoke ions and hot air, which detect abnormalities when smoke or hot air enters the sensor, but depending on the direction of the airflow, these may There are problems with early fire alarms because they cannot be detected because they do not work properly, and cigarette smoke, etc., can be mistaken for fire. In particular, it was impossible to detect with conventional sensors because air currents diffuse outdoors.

Furthermore, conventional fire alarm devices often suffer from malfunctions due to poor insulation.

[Problem to be solved by the invention]

The present invention has been made to solve these conventional problems, and uses a video camera that does not malfunction due to poor sensing or insulation as in the past, to constantly photograph the monitored area, and to detect abnormal situations and abnormalities. The object of the present invention is to provide an abnormality monitoring system that does not require the supervisor to constantly monitor images by alerting the supervisor when behavior exceeds a predetermined limit value.

[Means to solve the problem]

In order to achieve the above object, the abnormality monitoring system of the present invention stores normal images taken by a video camera installed in the monitored area, sequentially superimposes them with images taken after that, and performs arithmetic processing. Extract the change by
It is characterized by activating a warning device such as an alarm or a warning light when the variable exceeds an abnormal limit value, and the abnormal behavior or abnormal situation that occurs in the monitored area is a predetermined abnormal limit value. The computer analyzes whether the limit has been exceeded and transmits the information to the supervisor using an alarm device, and the supervisor then checks the status through a display device such as a CRT and makes a corresponding decision.
There is no need to monitor.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic diagram of an abnormality monitoring system according to the present invention, in which a video camera 1 is installed in a monitored area A such as a hospital or nursing home, and images taken during normal conditions are sent to a computer 2 such as a personal computer. The image is stored, and then sequentially photographed images are superimposed and arithmetic processing is performed to calculate the difference from the previous image, and the amount of change is extracted.

Next, when the amount of change exceeds a predetermined abnormality limit value, the computer 2 issues a signal to activate the alarm device 4 in the alarm center B.

As this alarm device 4, an appropriate device such as an alarm or a warning light is used.
6, the supervisor confirms the abnormal condition and takes countermeasures.

Note that a plurality of video cameras 1 can be arranged in the monitored area A as shown in FIG. 1, and images of each video camera 1 can be monitored by switching one CRT 6.
may be provided separately for each video camera 1.

To explain in more detail, first, the monitored area A during normal operation
is taken with the video camera 1, and the normal image a is divided into multiple windows and stored as shown in Figure 2 (al).
.

Next, after a certain period of time has elapsed, the monitored area A is sequentially photographed to obtain images (b) (see Figure 2 (bl)).Then, this image is superimposed on normal image a to obtain a normal image. a
Perform arithmetic processing to calculate the difference between the two and extract the change. As shown in Figure 2 (bl), when smoke begins, which is a sign of a fire, most of the image (b) changes, and by taking the difference between that image and the normal image (a), Only the image C that has changed due to smoke 8, such as C1, is extracted.When the number of changed windows exceeds a predetermined abnormality limit value, an alarm is transmitted to the monitoring center, and the monitoring staff immediately identifies the scene where the abnormality has been detected. You can check visually on the CRT6 display and work on early extinguishment and government ordinances.

In this case, as the abnormality caused by smoke 8 continues to expand,
Even if you take the difference between images taken sequentially, the change remains and expands, so if the change increases to an abnormal limit setting value of 60% or more compared to the normal image, the alarm device 4 is activated. Let them do it.

On the other hand, even in the case of a person, if the person wanders around, such as a dull old man, it can be recognized because the changes remain forever.

Next, the flowchart shown in FIG. 3 will be explained.

First, at 10 in the figure, image capture 'G (1) J is carried out, at 20 the second image capture' G T214 is carried out, and then at 30 the difference between the above two images, that is, '
G (11-G (2) = F (process 21J, and if 'F(2)=0' of 40 is YES'', return from 40 to 10.

Also, ' F (If 21 = "NO", which is not OJ, the third image capture from 50 to 60 ' G (31 J
and at 70 'G (21-G (31=
F (take the difference between the second and third sheets, like 31J, and get 8
At 0, 'F f3) / G fl) When J is smaller than the ratio α between the change in the second and third images where an abnormality occurred and the first normal image, it will return to 10. .

Since this ratio α is the predetermined abnormality limit value of the abnormal part with respect to the normal image, r F (3)/G
(11>α”), further capture the next 90 images and calculate the difference from the previous image as F (n),
'F(nl/G(1)J) is calculated and compared with the above α.

At this time, a limit is set in advance as to the number of sheets to be calculated, and if α exceeds this limit, a warning device 4 such as an alarm is activated.

If the value becomes smaller than α after several calculations,
It shall be determined that it is normal and return to the initial routine.

In addition, in FIG. 3, i represents initialization of the count value,
K represents the number of determinations for alarm output.

〔Effect of the invention〕

As explained above, since the present invention uses a video camera that does not malfunction due to poor sensing or poor insulation, etc., it is possible to detect abnormalities with a higher degree of reliability, and immediately visually observe the abnormal scene. This allows for immediate response to abnormal situations without the need to go to the scene and check, unlike conventional fire alarm systems.

In addition, this system does not require a supervisor to constantly watch images even at night, when abnormal situations often occur. Supervisors are usually engaged in other tasks, are alerted to abnormalities, and can take appropriate measures immediately. This makes it possible to safely manage hospitals, nursing homes, etc. with a minimum number of personnel.

Further, the present invention is capable of not only issuing a fire alarm but also monitoring the movement of people such as illegal intruders and issuing an alarm.

[Brief explanation of the drawing]

Figure 1 is a schematic diagram of the abnormality monitoring system according to the present invention, Figure 2 is a schematic diagram of the abnormality monitoring system according to the present invention;
FIG. fa), FIG. 2(b), and FIG. 2 fc) are explanatory diagrams showing the process of image processing by a computer, and FIG. 3 is a flowchart of the abnormality monitoring system according to the present invention. 1...video camera, 2...computer, 4...
・Alarm device, 6...CRT Figure 1

Claims (1)

[Claims] An abnormality monitoring system that sequentially superimposes normal images taken by a video camera and images taken after that, extracts changes, and activates an alarm device when the changes exceed a predetermined abnormality limit.