JPH04263395A - Equipment monitoring system - Google Patents

Abstract

PURPOSE:To accurately detect an abnormality in temperature at an early stage without being influenced by the meteorological conditions such as environmental condition or sunshine by detecting the temperature distribution difference between the average temperature and the maximum temperature on each object to be monitored dividing a monitoring area into plural parts. CONSTITUTION:A picture fetching means 2 fetches the image pickup picture from an infrared ray image pickup means 1 at every constant interval of time, and stores the previously fetched picture defined as the reference picture and the next fetched picture defined as an input picture to be monitored, respectively. A difference arithmetic means 4 calculates the difference picture between unmasked divided pictures in the input pictures and reference pictures divided by a mask control means 3. A temperature distribution difference detection means 5 detects the average temperature and maximum temperature in the temperature distribution on an object to be monitored 10 in the difference pictures obtained successively and calculates the temperature distribution difference as the deviation. When the calculated temperature distribution difference is higher than the prescribed level and the number of times for processing becomes more than the prescribed number of times, an abnormality symptom detection means 6 determines the average value for the prescribed number of times. Furthermore, the means 6 calculates the average value by repeating the determination of the average value continuously within a fixed time.

Description

[Detailed description of the invention]

0001

[Field of Industrial Application] The present invention relates to an equipment monitoring system, and more particularly to an equipment monitoring system that constantly monitors the temperature generated in equipment installed in a given space and issues an alarm when an abnormal temperature occurs. It is something.

[0002] In recent years, electric power equipment such as large transformer transformers at electric power companies and substations, refining equipment at oil plants, etc. have been equipped with the following measures to prevent fires and related explosions, improve maintenance efficiency, and save labor. There is a need for an equipment monitoring system that can appropriately and accurately notify equipment temperature status, inspection timing information, etc.

0003

[Prior Art] Conventional equipment monitoring systems use contact-type temperature sensors such as thermocouples and thermistors to detect temperature abnormalities that occur in equipment installed in any space, such as indoors or outdoors. There is.

[0004] In an equipment monitoring system using such temperature sensors, these sensors are brought into contact with and fixed on the main body of the equipment to be monitored, and the detected temperature data is sent to a monitoring device installed at a remote location. When an abnormally high temperature is detected, for example, if the temperature suddenly rises to 10
Temperature data was processed simply by outputting an abnormality display or warning because it could cause a fire or the like when the temperature rose to an abnormal (dangerous) temperature of 0°C.

[0005] However, since the contact type temperature sensor is a point sensor type, it must be used in large numbers when there are multiple pieces of equipment to be monitored or to simultaneously detect the temperature of a wide range (area) of the equipment. Moreover, by the time an abnormally high temperature is detected, it is on the verge of an emergency situation such as a fire, explosion, or destruction, and in many cases it is already too late. The disadvantage is that it is difficult to detect.

On the other hand, instead of such a temperature sensor method, a time difference detection method of temperature images is used as an equipment monitoring system for early detecting temperature abnormalities (temperature abnormality signs) especially in large equipment installed in a large space. A method using .

[0007] In this method, first, the temperature on the equipment to be monitored is captured as an infrared image using an infrared camera, and two infrared images representing temperature information taken at different times are used as an input image and a reference image, respectively. The difference images obtained from the simple difference calculation results between the stored images are displayed on a histogram graph for temperature and the number of pixels, and are compared with preset ranges for both the detected temperature data and the number of detected pixels, By detecting the temperature abnormality of the equipment only when the temperature exceeds this set range, an early warning or display of temperature abnormality signs can be issued.

[0008]

[Problems to be Solved by the Invention] In such a conventional facility monitoring system that uses simple time difference detection between infrared images, temperature detection by an infrared camera is a surface sensor, so unlike a point-type temperature sensor, it is difficult to detect temperature. Multi-point observation processing that detects a lot of temperature information at once through contact is possible, but if the equipment to be monitored is outdoors, sunlight and
There is a problem in that early signs of abnormal temperature cannot be reliably detected due to the influence of environmental conditions such as the presence or absence of wind and ambient temperature fluctuations due to seasonal changes.

[0009] To explain this more specifically, for example,
As shown in Figure 2(a), the average temperature (TAV: indicated by the horizontal line) on the equipment 10 is affected by fluctuations in the ambient environment temperature.
However, compared to the reference image taken in advance under normal temperature conditions,
From the input image when the temperature rises, 2 is calculated based on the difference calculation result.
A difference image of ℃ is obtained.

[0010] Also, as shown in Fig. 2(b), the input when only the average temperature of a specific (local) area indicated by diagonal lines on the equipment increases by 2°C due to the influence of sunlight etc. with respect to the reference image. From the difference between the image and the reference image, the average temperature is 0°C,
A difference image is obtained in which only the shaded area is 2°C.

[0011] Normally, this is done by detecting any abnormal temperature rise due to an actual equipment abnormality, etc., and comparing it with the temperature data on the histogram and the set range of the number of pixels, as shown in FIG. , If the above-mentioned temperature changes due to environmental temperature fluctuations or sunlight exist in the difference image, these changes will affect the comparison with the setting range and cause an error by that amount, making it difficult to accurately identify the temperature abnormality. This indicates that symptoms become undetectable.

[0012] As a result, there is a problem in that early detection of temperature abnormality signs can only be carried out under limited monitoring operation conditions such as at night or indoors where there is little influence from weather conditions.

Therefore, the present invention provides equipment that can accurately detect temperature abnormalities (signs of temperature abnormality) at an early stage without being influenced by environmental conditions such as seasonal fluctuations or meteorological conditions such as sunlight and wind. The purpose is to realize a monitoring system.

[0014]

[Means for Solving the Problems] In the equipment monitoring system according to the present invention for achieving the above object, as shown in principle in FIG. an infrared imaging means 1 for taking an image of the image; and an image taking means 2 for taking in images taken by the imaging means 1 at regular time intervals and storing the previous taken image as a reference image and the next taken image as an input image. , each of the captured and stored reference image and input image is divided into a plurality of parts, and any one of them is divided into multiple parts.
mask control means 3 for masking other than the two divided images and sequentially moving the unmasked divided images at regular time intervals; and a difference between the unmasked divided images among the divided input image and the reference image. Difference calculation means 4 for calculating an image, and temperature distribution difference detection means for detecting the maximum temperature TMAX and average temperature TAV of the temperature distribution on the object 10 from within the difference image and calculating the temperature distribution difference ΔT which is the deviation thereof. 5, when the calculated temperature distribution difference ΔT is equal to or higher than a predetermined level, the average value ΔTAV for a predetermined number of times is continuously calculated within a predetermined time to confirm the temperature continuous increase property of the temperature distribution difference ΔT. Abnormality sign detection means that detects an abnormality sign by using the input image, and updates the input image to the reference image when the abnormality sign is not detected or before the average value ΔTAV is calculated, and repeatedly captures the next image as the input image. 6.

[0015] Furthermore, an abnormality prediction detection means 7 is further provided,
By repeatedly detecting the abnormality symptoms, the predicted time until the average temperature distribution difference ΔTAV reaches a dangerous abnormality level may be calculated.

Furthermore, a dangerous abnormality detecting means 8 may be further provided together with the abnormality prediction detecting means 7 to detect a dangerous abnormality when the average temperature distribution difference ΔTAV exceeds the dangerous abnormality level.

[0017]

[Operation] In general, in order to quickly and accurately detect temperature abnormalities that occur in equipment and equipment, it is better to use an infrared camera than to use the temperature sensor method (point sensor method) as described in the conventional example. The image processing method used is more advantageous.

To explain this using FIG. 3, first, when a minute temperature abnormality occurs in equipment or equipment, the temperature abnormality spreads rapidly and the heat accumulated inside reaches the ignition point, causing a fire. This may lead to an explosion, but in this case, the temperature does not suddenly change from a normal state to a dangerous abnormal high temperature, but a preliminary temperature rise, that is, an abnormal symptom, always occurs as an intermediate step.

The cause is abnormal internal heat generation due to some cause, which always appears as a change in temperature pattern different from the normal state, and progresses gradually and slowly as shown by the curved portion (a).

That is, the temperature sensor method is curved part (b)
Whereas only the abnormally high temperature area shown in (a) was detected, if the upward trend in curve area (a) is understood and detected in advance, it will be possible to prevent fires and explosions. .

[0021] Therefore, the present invention also targets the detection of abnormality signs as an intermediate stage shown by the curve part (a) by using an image processing method using infrared rays (time difference image of temperature image). It is something.

However, in view of the purpose of detecting such early signs of temperature abnormality, how to avoid being affected by environmental conditions, weather conditions, etc.
As shown in (a), the temperature pattern change that occurs on the object 10 in a certain monitoring space is expressed as the difference between the area of the average temperature TAV and the area of the maximum temperature TMAX existing on the object 10 to be monitored. The present inventor has focused on the fact that it can be understood as a distribution difference ΔT.

That is, in this case, the temperature distribution difference ΔT is as shown in the figure (
b) As shown in b), if temperature pattern changes are detected not as absolute values of the average temperature TAV or maximum temperature TMAX, but as a deviation between them, it will not change (irrelevant) even under the influence of environmental (ambient) temperature fluctuations. ) (Note that this temperature distribution difference ΔT actually has a different value depending on the type of each facility or device).

For example, as shown in FIG. 2(c), when an abnormality symptom occurs where the average temperature TAV is 15°C and the maximum temperature TMAX is 30°C, the input image shows the difference from the reference image in the normal state. As a result, a difference image in which the average temperature TAV is 5°C and the maximum temperature TMAX is 20°C is obtained, and the temperature distribution difference ΔT is 20°C - 5°C = 15°C. In other words, this value does not change even under the influence of environmental (ambient) temperature fluctuations.

Therefore, in the equipment monitoring system according to the present invention, as shown in principle in FIG.
In this system, infrared images, which are temperature data, are constantly captured from monitored objects 10 such as equipment and equipment installed in an arbitrary space.

At this time, the image capturing means 2 captures (samples) captured images from the infrared imaging means 1 at regular time intervals, uses the previously captured image as a reference image, and monitors the next captured image. Each input image is stored as an input image.

Next, in the mask control means 3, as shown in FIG.
As shown in (c), when a plurality of monitored objects 10 are monitored on the same monitoring screen, each temperature distribution difference ΔT to be detected is usually different for each monitored object 10, so However, when monitoring, areas that are not subject to monitoring that inevitably occur on the monitoring screen do not require calculation of the temperature distribution difference ΔT and must be masked. The captured and stored reference image and input image are each divided into a plurality of regions (n) to form mask regions indicated by diagonal lines.

In this case, as shown in FIG. 2(d), it is difficult to calculate a uniform ΔT even for a single monitored object 10 due to its internal structure and characteristics, and at the same time, it is difficult to calculate a uniform ΔT due to weather conditions such as sunlight and wind. When the influence can be avoided, the same monitoring area is further subdivided into 1/N to n/N.

For this reason, the mask control means 3
), leave only one arbitrary divided image among the divided images and mask the other divided images,
Any one unmasked divided image at time t1
Masking control is performed such that the divided images are sequentially alternated with other masked (n-1) divided images every ~tn.

The difference calculating means 4 calculates a difference image between the unmasked divided image of the input image and the unmasked divided image of the reference image among the input image and the reference image divided in this way. , the temperature distribution difference detection means 5 detects the average temperature TAV and maximum temperature TM of the temperature distribution on the monitoring object 10 from within the difference images sequentially obtained from the mask control means 3.
AX is detected and its deviation is the temperature distribution difference Δ
Calculate T.

Next, the abnormality symptom detection means 6 determines whether the calculated temperature distribution difference ΔT is less than a predetermined level (here, a level corresponding to a preset abnormality symptom detection) or the number of times the temperature distribution difference ΔT is calculated is a predetermined number of times. If the difference in temperature distribution ΔT exceeds a predetermined level and the number of times of processing reaches a predetermined number or more, the reference image captured by the infrared light imaging device 1 is operated on the image capture means 2. and input images are repeatedly captured.

[0032] Then, when the calculated temperature distribution difference ΔT is equal to or higher than a predetermined level and the number of processing times exceeds a predetermined number, the average value ΔTA of the temperature distribution difference ΔT for the predetermined number of times is
After finding V, the calculation is repeated continuously within a certain period of time.

Therefore, the continuously calculated temperature distribution difference Δ
As shown in Fig. 6, the average value ΔTAV of T forms a curve over time, but since weather changes and equipment load fluctuations do not last long, this curve does not change over a certain period (time t0). ~ t1), if it is confirmed that the temperature change continuously follows the upward trend shown by α,
This means that rather than short-term temperature fluctuations, a persistent phenomenon, that is, a true sign of temperature abnormality, has been detected.

In addition, in this case, the average value ΔTAV of the temperature distribution difference corresponds to the curve between time t0 and t1 in FIG. Exist within the level TA1 range.

In this way, the monitoring area is divided into a plurality of parts, and the average temperature TAV and maximum temperature TM on each monitoring object 10 are determined.
By detecting the temperature distribution difference ΔT between AX and statistically repeatedly processing it, it is possible to reliably detect temperature abnormality signs at an early stage without being influenced by environmental conditions or meteorological conditions such as sunlight.

Furthermore, an abnormality prediction and detection means 7 shown in broken lines is further provided in the equipment monitoring system configuration shown in principle in FIG. 1, and the average value ΔTAV of the temperature distribution difference ΔT is repeatedly calculated to detect abnormality symptoms. By repeatedly detecting
As shown in FIG. 7, when the average value ΔTAV of the temperature distribution difference and the period from time t1 to t2 is in the range from the primary abnormality level TA1 to the dangerous abnormality level TA2, if the rising curve is continuously rising, it is dangerous. Predicted time tA until it is considered abnormal and reaches dangerous abnormality level TA2 from its upward trend
It is possible to predict abnormalities by calculating the

[0037] The temperature rise (predicted) curve shown by the broken line within this range has different characteristics depending on the equipment to be monitored, and generally has the characteristics shown by A, B, and C, respectively. Both curves show an abnormal rise in temperature. A is the simplest example, and the scale of the abnormality is small and in the initial stage in normal equipment. This type of abnormal progression may continue as an extension of the past course for a certain period of time. Note that the predicted time tA in this case is t0
It can be approximately calculated from the characteristic curve between t1 and the dangerous abnormality level TA2. Similarly, the predicted time tA for curves B and C can also be calculated.

Furthermore, in the present invention, a dangerous abnormality detection means 8 shown by a dashed line is further provided in the equipment monitoring system provided with the above-mentioned abnormality prediction detection means 7, and the average value ΔT of the temperature distribution difference ΔT is
When AV exceeds the dangerous abnormality level TA2 shown in FIG. 7, it is possible to detect a dangerous abnormality and notify of a fire, an explosion, etc., or a situation immediately before such a situation.

[0039]

[Embodiment] FIG. 8 shows an embodiment of the equipment monitoring system according to the present invention shown in FIG.
101 to 10n are infrared cameras for n channels as the infrared light imaging means 1 shown in FIG.
A rotating base for supporting and rotating 1 to 1n, 111 to
11n is a controller that controls each imaging state of the infrared cameras 11 to 1n, 121 to 12n is a transmission system that transmits each captured image from the controllers 111 to 11n, and 13 is a video that switches each captured image from the transmission system 121 to 12n. The switcher 14 is an image processing device including the image capture means 2, mask control means 3, difference calculation means 4, and temperature distribution difference detection means 5 shown in FIG.
15 similarly indicates an abnormality symptom detection means 6 and an abnormality prediction detection means 7.
The video switcher 13 is controlled by a data processing device including the dangerous abnormality detection means 8 and the transmission system 121.
12n controls each of the rotating bases 101 to 10n, 16 is an alarm device that issues an alarm for temperature abnormality signs, dangerous abnormalities, etc. based on the processing results of the data processing device 15, and 17 is an image processing device 14. A TV monitor 18 is connected to the data processing device 15 and each rotating base 101 displays the difference image etc. output as a result of the difference calculation.
-10n are shown respectively. Note that the control line from the base operation console 18 is connected to each of the rotating bases 101 to 10n via the data processing device 15 and transmission systems 121 to 12n. In the figures, the same reference numerals indicate the same or corresponding parts.

[0040] In this embodiment, n infrared cameras are prepared, but this is useful for monitoring areas that cannot be imaged with just one surveillance camera, or for monitoring a wider area.
In order to support high-precision detection, CH1 to CHn are provided with different imaging positions, imaging angles, etc., and the base operation console 18
It is selected by This infrared camera may hereinafter be collectively referred to as "1".

In such an embodiment, first, the infrared cameras 11 to 1n capture the temperature on the monitored object 10 as an infrared image, and the dedicated controllers 111 to 1
1n to transmission systems 121 to 12n,
The video switcher 13 appropriately switches and selects each captured image from the transmission systems 121 to 12n according to a command from the data processing device 15, and sends the selected image to the image processing device 14.

In addition, the image processing device 14 calculates the average temperature TAV and the maximum temperature TMAX from the difference image between the input image and the reference image, each of which is divided into a plurality of input images taken at different points in time, and As a result, a temperature distribution difference ΔT is calculated.

Further, the data processing device 15 repeatedly performs statistical data processing based on the calculated temperature distribution difference ΔT to detect signs of temperature abnormality, predict dangerous abnormalities, and detect dangerous abnormalities.

Furthermore, the alarm device 16 generates an alarm from each detection signal as a result of processing these data, and the TV monitor 17 monitors each captured image and the difference image processed by the image processing device 14. ing.

FIG. 9 shows a detailed configuration example of the image processing device 14, in which numeral 21 denotes a write controller that controls storage of captured images from the monitored object 10 selected by the video switcher 13. , 22 and 23 are frame memories each storing one frame of captured images instructed by the write controller 21, and 24 is a frame memory that performs image division and mask control for each captured image read out from each frame memory 22 and 23. 25 is a difference calculator that calculates a difference image between each divided image sent from the mask generator 24; 27 is a difference calculator that binarizes each pixel of the difference image from the difference calculator 25; A binarization circuit 29 is a histogram calculation circuit that performs a histogram calculation, which will be described later, on the binarized data from the binarization circuit 27; 30 is a 2
A projection calculation circuit performs a projection process to be described later on the binarized data from the digitization circuit 27, and 31 calculates the average value TAV and maximum value TMAX of temperature data from the calculation results of the histogram calculation circuit 29 and the projection calculation circuit 30. Calculation circuits for calculating are shown in each figure.

Operation of Image Processing System The operation of the image processing device 14 as described above is shown in FIG.
First, the image processing system will be explained based on the processing flowchart shown in FIG.

First, the captured image from the monitored object 10 captured by the infrared camera 1 is subjected to frame sampling at fixed time intervals (step S3 in FIG. 10), so that the earlier sampled image is used as the reference image. ,
The images immediately after that are stored as input images in the frame memories 22 and 23 by the write controller 21, respectively (steps S1 and 2). In addition, in step S3,
When calculating the normal temperature distribution difference ΔT, slow sampling TS is used for the time interval at which the infrared camera 1 takes images, but when detecting a fire, high-speed sampling FS is used to quickly detect sudden abnormalities. .

Next, the mask generation circuit 24 divides each image stored in the frame memories 22 and 23 into a plurality of images as shown in FIG. 5, and masks unnecessary portions (step S31, 4), difference calculator 25
A difference image between (each divided image of) the input image and (each divided image of) the reference image is calculated by (step S6)
. Note that the mask shape created in the mask generation in step S4 can be set to any shape by image processing.

Thereafter, the calculated difference image is subjected to binarization processing by the binarization circuit 27 (step S7).

Furthermore, in the histogram calculation circuit 29, 2
From the digitized image data, create a histogram showing the number of pixels distributed in each temperature region by taking the sections divided into temperature regions on the horizontal axis and the number of pixels on the vertical axis (
same step S9).

[0051] In this case, a set value of the temperature that becomes a dangerous abnormality is set on the histogram, and if the number of pixels exceeding this set value is a predetermined value or more, a dangerous abnormality is determined and a dangerous abnormality detection signal is sent. (step S20).

In the projection calculation circuit 30, each pixel of the difference image displayed as a histogram is plotted on the X-axis and Y-axis on the image frame.
The calculation circuit 31 calculates the maximum temperature TMAX based on the graph display on the histogram and calculates each temperature from the result of the projection calculation by the projection calculation circuit 30. The average temperature TAV is calculated by dividing the total temperature of the pixels by the total number of pixels as an area (step S11), and the temperature distribution difference ΔT is calculated from the deviation between the maximum temperature TMAX and the average temperature TAV. .

In step S3, when high-speed sampling FS is performed to detect a fire or the like, an offset adder 26 which gives a predetermined offset amount to the difference calculator 25 and An AND operator 28 for ANDing the binarized data of the binarization circuit 27 is further provided.

In this case, first the offset adder 26
In order to quickly detect a fire, an offset amount is given to the difference calculator 25 to add a predetermined offset temperature to each image data stored in the frame memories 22 and 23 pixel by pixel (step S5); AND processing is performed on the binarized image data by the AND calculator 28 (step S8). and finally the abnormal location (
In order to specify the coordinates of the location where the fire occurred, the projection calculation circuit 3
From the coordinate position resulting from the projection calculation from 0, the temperature center of gravity, that is, the address information is calculated (step S1
9).

Operation of the data processing system Next, the operation of the data processing system will be explained based on the same flowchart with reference to FIGS. 6 and 7.
The abnormality sign detection means 6, the abnormality prediction detection means 7, and the dangerous abnormality detection means 8 shown in FIG.

In order to confirm the continuous increase, the calculated temperature distribution difference ΔT is first compared with the 0th order abnormality level TA0, and if it is less than this value, the number of processing times N for calculating the temperature distribution difference ΔT is determined. is less than the predetermined number of times n1,
The processing routine from capturing the input image of the image processing system to calculating the temperature distribution difference ΔT is repeated until ΔT>TA0 and N>n1 (zero-order detection of abnormality signs: step S12).

Then, the average value ΔTAV of the calculated temperature distribution difference ΔT for a predetermined number n1 times is calculated, and this is further calculated for a predetermined processing period (time t0
~t1) to statistically calculate the average value ΔTAV(t) as a function of time change,
ΔTAV(t)>1st abnormality level TA1 (however, TA1=
T+α) and confirmed a clear continuous increase,
In addition to detecting temperature abnormality signs (confirmation and evaluation of abnormality signs: steps S13 and 14), this rising trend data Δ
Store TAV(t) in file 40.

Next, by confirming and predicting the upward trend from the data stored in the file 40, the average value ΔTAV(t) of the temperature distribution difference ΔT is determined to be at level TA2, which becomes a dangerous abnormality from time t1 to time t2. Anomaly prediction is performed by calculating the predicted time tA until the time tA is reached (step S15). Note that in step S16, step S15
The same steps as above are repeated (tracking), and the predicted time tA is reviewed as necessary.

Then, when the average value ΔTAV(t) reaches the dangerous abnormality level TA2, or even if it does not reach TA2, the monitoring time exceeds the maximum allowable time, a dangerous abnormality is detected (dangerous abnormality occurrence: Step S17), the dangerous abnormality is generated as a warning with an alarm, buzzer, etc. (step S18)

[0060] In the alarm process of step S18, an alarm is generated by detecting abnormality signs from steps S14 and 15, an alarm is generated by detecting a dangerous abnormality from step S20,
Predicted time tA until a dangerous abnormality occurs from step S15
, the address of the abnormal location from step S19, or a graph of the average value ΔTAV(t) after the occurrence of the abnormal symptom, etc. can be displayed.

[0061]

[Effects of the Invention] As explained above, in the equipment monitoring system according to the present invention, two infrared images taken at different times are used as an input image and a reference image, and each image is divided into a plurality of monitoring areas. The temperature distribution difference between each average temperature and maximum temperature is calculated, and input images and reference images are repeatedly captured until an abnormality symptom trend is detected, and the average value of the temperature distribution difference is calculated to statistically detect abnormality symptoms. The system is configured to detect abnormal symptoms by confirming and evaluating the
An equipment monitoring system that can reliably detect signs of temperature abnormality at an early stage is realized, and as a result, it can be used to prevent equipment fires, explosions, etc.

Furthermore, by calculating the predicted time until the calculated average value of the temperature distribution difference reaches the dangerous abnormality level, it is possible to understand the progress of the abnormality and also to predict the abnormality until it becomes a dangerous abnormality.

Furthermore, if it is detected that the average value of the temperature distribution difference exceeds the dangerous abnormality level, a dangerous abnormality alarm is issued to notify of a fire, explosion, etc., or to notify of a situation immediately before such a situation. It is also possible to do so.

[Brief explanation of the drawing]

FIG. 1 is a principle block diagram of an equipment monitoring system according to the present invention.

FIG. 2 is a diagram showing the relationship between common temperature error factors and temperature abnormalities.

FIG. 3 is a diagram showing a general abnormal temperature pattern in equipment.

FIG. 4 is a diagram conceptually showing a temperature distribution difference in the present invention.

FIG. 5 is a diagram illustrating image division and mask processing in the present invention.

FIG. 6 is a diagram conceptually showing the detection of temperature abnormality signs in the present invention.

FIG. 7 is a diagram conceptually explaining temperature abnormality prediction in the present invention.

FIG. 8 is a diagram showing an embodiment of the equipment monitoring system according to the present invention.

FIG. 9 is a diagram showing an embodiment of an image processing device used in the present invention.

FIG. 10 is a processing flowchart diagram according to an embodiment of the present invention.

[Explanation of symbols]

1 Infrared light imaging means 2 Image capture means 3 Mask control means 4 Difference calculation means 5 Temperature distribution difference detection means 6 Abnormality symptom detection means 7 Abnormality prediction detection means 8 Dangerous abnormality detection means 10 Monitored objects

Claims (3)

[Claims] 1. An infrared imaging means (1) for taking an infrared image from a monitored object (10) in an arbitrary space, and capturing images of the imaging means (1) at regular time intervals,
an image capturing means (2) for storing a previous captured image as a reference image and a next captured image as an input image, and dividing each of the captured and stored reference image and input image into a plurality of parts. mask control means (3) for masking all but one of the divided images and sequentially moving the unmasked divided images at regular time intervals; a difference calculation means (4) for calculating a difference image between divided images that are not divided;
Temperature distribution difference detection means (5) that detects the maximum temperature (TMAX) and average temperature (TAV) of the temperature distribution on the object (10) from within the difference image and calculates the temperature distribution difference (ΔT) that is the deviation thereof. and the calculated temperature distribution difference (ΔT)
When the temperature distribution difference (ΔT) is above a predetermined level, the average value for a predetermined number of times is used to confirm the continuous rise in temperature of the temperature distribution difference (ΔT).
(ΔT AV) is continuously calculated within a predetermined time to detect abnormal signs, and when the abnormal signs are not detected or before the average value (ΔT AV) is calculated, the input image is used as the reference image. An equipment monitoring system comprising: an abnormality sign detection means (6) that repeatedly updates and captures the next image as an input image. 2. The apparatus further comprises an abnormality prediction and detection means (7) that calculates a predicted time until the average temperature distribution difference (ΔT AV) reaches a dangerous abnormal level by repeatedly detecting the abnormality sign. Claim 1 characterized in that
Equipment monitoring system described in . [Claim 3] The average temperature distribution difference (ΔT AV)
3. The equipment monitoring system according to claim 2, further comprising dangerous abnormality detection means (8) for detecting a dangerous abnormality when the level exceeds the dangerous abnormality level.