JPH0456482A - Infrared ray monitoring system - Google Patents

Abstract

PURPOSE:To instantaneously and certainly detect abnormality even when the generated abnormality changes slowly by comparing a picture obtained the last time at the normal time with a picture obtained this time, at an interval of a picture processing period towards the other infrared ray cameras. CONSTITUTION:Each output of infrared ray cameras 11-13 is operated for sampling according to the controls of write-in controlling parts 21-23, supplied through transmitting devices 31-33 to a video switcher 4, and time-divisionally switched successively for each channel in the video switched 4. Then, it is supplied to a TV monitor 5 and also supplied to a picture processor 6 in order to be processed as a picture. In this case, picture switch-over is operated so that the inputted picture of the last sampling time at the normal time can be this time reference picture, and the inputted picture of the last sampling time is compared with that of the this time sampling time. Thus, the abnormality can be detected certainly and speedily even when the abnormality changes slowly.

Description

[Detailed Description of the Invention] [Summary] Regarding an infrared monitoring system that monitors abnormalities in equipment and the environment using an infrared camera, the present invention aims to instantly and reliably detect abnormalities that occur even if the speed of change is slow. From the image acquisition of multiple infrared cameras, the previously acquired normal image acquired after the image processing period for other infrared cameras is used as the current reference image, and this reference image is compared with the currently acquired image. The configuration is such that a means is provided for each camera.

[Industrial application field]

The present invention relates to an infrared monitoring system that monitors abnormalities in equipment or the environment using an infrared camera.

For example, in outdoor substations, etc., there are many large pieces of equipment such as transformers and circuit breakers, and if these devices rise in temperature or catch fire for some reason, or if there is an intruder inside the substation, etc. , it is necessary to detect the occurrence of such an abnormality and take appropriate measures. For this reason, there is a need for a highly accurate system that installs multiple infrared cameras in substations to constantly monitor the surroundings of equipment and promptly and reliably detect any abnormalities that occur.

[Conventional technology]

The infrared monitoring system as described above samples images obtained by an infrared camera based on a frame sampling period (for example, 1 second), and detects an abnormality by comparing two consecutive images before and after.

Figure 6 shows the frame sampling period and the change state of abnormality (
For example, it is a characteristic diagram showing the relationship with temperature rise, magnitude of abnormality, etc. △t is frame sampling time, △T
s(△T (alarm detection threshold)) is the amount of abnormal change during low-speed change, and △TH (>△T) is the amount of abnormal change during high-speed change. On the other hand, FIG. 7 is a conceptual diagram of the operation of the conventional system, showing the first channel (CH,) and the second channel (CH2).
), which shows the relationship between image capture and image processing timing by, for example, three infrared cameras of the first channel (CH, ).

In Figure 7, for example, the infrared camera of CH,
A reference image is captured at the frame sampling time to shown in A), and an input image is captured at the frame sampling time t1 as shown in the same figure (B), and the image is processed at the frame sampling time t1 as shown in the same figure (C). . C
H.

The image capturing and image processing for the infrared camera of CH is also carried out in accordance with this, and the image capturing and image processing are sequentially and repeatedly performed in a time-sharing manner by a plurality of infrared cameras. For example, if an abnormality occurs at sampling time t4 in a device monitored by an infrared camera on CH, for example, the reference image captured at sampling time t6 and the input image captured at sampling time t7, etc. The image is processed at a sampling time t. In this case, if the rate of change of the abnormality that has occurred is fast, the abnormality change amount △T shown in FIG. 6 will be detected at sampling time t7; The abnormal change amount ΔTs shown in FIG. 6 is detected.

[Problem to be solved by the invention]

However, in the conventional monitoring system, if the rate of change of the abnormality that has occurred is fast, for example, the sampling time ja, j7
Either the occurrence of an abnormality is detected because there is a large change between the two images captured respectively, or if the speed of change of the abnormality that has occurred is slow, there is almost no change between the two consecutive images above. Therefore, abnormal occurrence cannot be detected. For this reason, the conventional system has the problem that it can be applied only when the rate of change of abnormality is fast, and the field of application is limited.

SUMMARY OF THE INVENTION An object of the present invention is to provide an infrared monitoring system that can instantly and reliably detect an abnormality that has occurred even if the rate of change is slow.

[Means to solve the problem]

FIG. 1 shows a diagram of the principle of the present invention. When multiple infrared cameras 20, to 203 are used as shown in FIG.
Normally acquired image (P1) acquired last time (t1) after the image processing period (12 to t5) for 02,202
A means is provided for each of the plurality of infrared cameras 203 to 203 to compare this reference image (P1) with the acquired image (P7) of this time (t7), with the reference image (P1) of this time (t6) being set as the reference image (P1) of this time (t6). . In addition, when multiple different monitoring targets are set for one infrared camera 20 as shown in FIG. ) was acquired last time (t1) during normal operation (P1
) is set as the reference image (P1) of the current time (t6), and means is provided for comparing this reference image (P1) and the acquired image (P7) of the current time (t7) for each of a plurality of different monitoring targets.

[Effect]

In the present invention, for example, a normal image (P1) acquired last time in CH, and an image (P7) acquired this time are compared. Therefore, in the monitoring range of CH, for example, the difference between images P1 and P7 is large not only when the abnormality that occurs at timing t4 is changing rapidly, but also when the changing speed is slow. Abnormalities can be detected reliably and quickly even when the speed is slow.

〔Example〕

Fig. 2 is a block diagram of one embodiment of the present invention, and Fig. 3 is a block diagram of an embodiment of the present invention.
3A and 3B show image processing flowcharts of the systems shown in the figures. In FIG. 2, infrared cameras 11 of CH, to CH3
~1. The outputs are respectively output from the write control unit 2. ~2. As in the conventional example, sampling is performed as shown in FIG. 4(B) at every frame sampling period (for example, 1 second) shown in FIG. 4(A) according to the control of FIG.

100b) is supplied to the video switcher 4 via the transmission devices 31 to 33, where it is sequentially switched for each channel in a time-divisional manner, and supplied to the TV monitor 5, where it is displayed on the monitor, while the image processing device 6 Similarly to the conventional example, the image is processed at the timing shown in FIG. 4(C). In this case, when comparing two images, the present invention does not compare two successive images before and after, as in the past, but compares the previous normal sampling time (CH , is 1+) input image this time (
The image is exchanged so that it becomes the reference image of t5 in CH, (step 100c in Figure 3), and the input image of the previous sampling time (1+ in CH) and the input image of the current sampling time (ti in CH, ti) are used. (step 102 in FIG. 3). At this time, offset addition is performed in advance (Step 1 in Figure 3).
O1), the abnormal temperature information stored this time is higher than the background temperature information stored last time (before the abnormality occurred) so that the difference calculation result will always be a 10 value (the abnormal temperature may be lower than the background temperature, If this continues, the difference calculation result will appear as both a 10 value and a - value, which is inconvenient for processing).

Further, the image processing device 6 performs binarization processing, and detects only abnormal temperature areas in the image (step 1 in Fig. 3).
03), and an AND operation is performed to fit the actual temperature value to the binarized information (step 104).

The output of the image processing device 6 is supplied to a data processing device 7,
Here, a histogram calculation is performed and the data is processed into temperature vs. number of pixels (step 105), and if the information is higher than a predetermined temperature and the number of pixels is greater than a predetermined number, this is regarded as an abnormal occurrence and step 106) is carried out, and the alarm device 8 A warning will be issued. The output of the data processing device 7 is displayed on a display 9 and printed out on a printer 10.

As described above, in the present invention, when looking at CH, for example, two consecutive images (sampling times t0 and 1+, ja and 17) are compared as in the conventional example shown in column (A) of FIG. Instead, as shown in column (B) of Fig. 5, the input image from the previous time (t1) during normal operation is used this time (t6).
Image processing is performed at timing t7 as a reference image.

Therefore, the abnormality can be detected reliably and at high speed not only when the abnormality changes at a fast rate, but also when the abnormality changes at a slow rate. Similar processing is performed on other channels as well.

If no abnormality is detected during the determination in the data processing device 7 described above (step 106 in FIG. 3), image replacement as described in step 100c described above is performed, and similar processing is performed thereafter.

In addition, infrared camera 1. ~l, respectively, are rotating devices 1, a~
It is configured to be turned by a predetermined angle at La, and the turning angle, turning direction, etc. can be freely controlled by operating the turning device console 11.

Further, the above embodiment includes three infrared cameras 1. to 13, the surroundings of each device are monitored to detect that the device has become abnormal due to a rise in temperature, but the present invention is not limited to this. In the camera, for example, three monitoring modes may be provided to detect the presence or absence of a fire, the presence or absence of a temperature rise, and the presence or absence of an intruder, respectively. In this case, respective threshold values are provided for each of the three monitoring modes, and image capture and image processing are sequentially and repeatedly performed in a time-sharing manner for each mode in the same manner as described in FIG. 4.

Alternatively, a plurality of infrared cameras may be provided and a plurality of modes as described above may be set for each infrared camera.

〔Effect of the invention〕

As explained above, according to the present invention, since the previously acquired normal image is compared with the currently acquired image, even if the rate of change of the abnormality that has occurred is slow, this can be reliably confirmed.
Moreover, it can be detected at high speed, and can be applied to a wider range of fields than the conventional method, which could only be applied when the rate of change is fast.

[Brief explanation of drawings]

Fig. 1 is a principle diagram of the present invention, Fig. 2 is a block diagram of an embodiment of the present invention, Fig. 3 is an image processing flowchart of the present invention, Fig. 4 is a conceptual diagram of the operation of the present invention, and Fig. 5 is a diagram of the operation of the present invention. A diagram comparing the present invention and a conventional example, FIG. 6 is a characteristic diagram showing the relationship between the rate of change of abnormality occurrence and frame sampling, and FIG. 7 is a conceptual diagram of the operation of the conventional example. 8 is the alarm device, Pl is the normal image acquired last time (current reference image), and P7 is the image acquired this time. Patent applicant Fujitsu Ltd. In the diagram, 1, ~Is, 201-203 are infrared cameras, 21-2
．． 4 is a video switcher; 6 is an image processing device; 7 is a data processing device Figure 5 ÷△↑← Figure 6

Claims (3)

[Claims] (1) Images acquired by a plurality of infrared cameras (20_1 to 20_3) that capture images of respective monitoring target ranges are captured at different timings (t_1, t_3, t_5, t_7) for each of the plurality of infrared cameras (20_1 to 20_3). ,...) in a sequential time-sharing manner, and the plurality of infrared cameras (2
0_1 to 20_3) In an infrared monitoring system that detects anomalies by comparing two images acquired at different times, from the image acquisition this time (t_6, t_7), Image processing period (t
The normal image (P_1) acquired last time (t_1) after _2 to t_5) is used as the reference image (t_6) of this time (t_6).
P_1), the reference image (P_1) and the current (t_7) acquired image (P
_7) The above-mentioned multiple infrared cameras (
20_1 to 20_3) An infrared monitoring system characterized by being provided with means for each of them. (2) Infrared camera (20_1) that images the monitoring target area
) at different timings (t_1, t_3, t_5, t_7,...) for different monitoring targets.
) in an infrared monitoring system that detects anomalies by sequentially processing in a time-sharing manner and comparing two images acquired at different times for each of the different monitoring targets. The normally acquired image (P_1) acquired last time (t_1) after an image processing period (t_2 to t_5) for other monitoring targets after acquisition is set as the reference image (P_1) of this time (t_6), and the reference image is (P_1) and the acquired image (P_1) and this time (t_7)
_7) An infrared monitoring system characterized by comprising means for comparing each of the plurality of different monitoring targets. (3) The plurality of different monitoring targets include whether or not a fire has occurred;
3. The infrared monitoring system according to claim 2, wherein the infrared monitoring system detects the presence or absence of a temperature rise and the presence or absence of an intruder.